ABSTRACT VARIATION BETWEEN AND WITHIN NINE EURASIAN SPECIES OF UZmus BY Gary Long The American elm, long the most popular shade tree in eastern North America, has been decimated over much of its range by the Dutch elm disease (Ceratocystis ulmi). This study represents the first step in a program to screen foreign species of UZmus for possible replacements for the American elm. The objectives of this study are as follows: 1. Determine the occurrence and extent of variation in morphological characters of the samaras of nine Eurasian Species of UZmus available for study. 2. Determine the occurrence and extent of variation in morphology, physiology and resistance to Dutch elm disease of the seedlings grown from these samaras. 3. Combine the information gained under numbers 1 and 2 above and information from the literature into a basic set of data to be used in planning and carry- ing out future elm breeding research at Michigan State University. Gary Long Seed collections from 567 individual trees, repre- senting 9 Eurasian species of UZmus were included in this study. Twenty-six characters were studied on a four samara sample from each seedlot. Forty-nine additional characters were studied on the seedlings produced. These characters were then analyzed to determine the differences between species and the occurrence and ex- tent of genetic variability within species. There were highly significant differences (.01 level) between species in each of the 75 characters studied. Eight of the nine species included in the study were distinctly different from each other. There was very little difference between U. gZabra and U. elliptica. Very little of the within-species variation could be accounted for by the geographic origin of the seed. Only leaf size in U. minor showed a recognizable geographic trend. Susceptibility to Dutch elm disease seemed to occur at random within the material studied. Selection for re- sistance will need to be on a family basis. U. UZmus minor and U. japonica showed the greatest potential for producing an acceptable substitute for the American elm. VARIATION BETWEEN AND WITHIN NINE EURASIAN SPECIES OF ULMUS BY .- ._.‘ ‘- ‘. J11 . ‘5 \ Garyitong A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1971 ACKNOWLEDGMENTS The author wishes to express his appreciation to Dr. J. W. Wright for his assistance in selecting the topic and his guidance and counsel during the course of the study. The help of the other members of the Graduate Committee; Drs. M. W. Adams, J. H. Beaman and J. H. Hanover, is also acknowledged. Special thanks to Dr. Hans M. Heybroek of the Stichting Bosboquroefstation "De Dorschkamp," Wageningen, The Netherlands, who served as advisor to the author (from January to June of 1969) while guest lecturer at Michigan State University. Thanks are also due to the many European and Asiatic collectors who supplied the seed that made this study possible. ii TABLE OF CONTENTS LIST OF TABLES O O O O O I O O O O I O O 0 Chapter I. II. III. IV. INTRODUCTION 0 O O O O O O O O O O TAXONOMY O O O O O O O O O O O O O The Genus UZmus . . . . . . . . . The Subgeneric Classification . . Phy10geny O O O O O O O O O O 0 Economic Importance . . . . . . The Species of UZmue . . . . . . Eastern North American Species of UZmus . . . . . . . . . . . The Elms of Mexico . . . . The Elms of Europe . . Elms of Eastern Asia . . . Elms of the Himalayas and Southeast Asia . . . . . . . . CYtOlogy O O O O O O O O O O O O HYBRIDIZATION IN ULMUS . . . . . . Natural Hybrids . . . . . . . . . Hybridization Techniques . . . . Crossability Patterns of UZmus . THE DUTCH ELM DISEASE . . . . . . . History . . . . . . . . . . . . . Description of th Fungus and the Disease . . . . . . . . . . . Development of th Disease After Artificial Inoculation . . . . . Environmental Factors Affecting the Disease . . . . . . . . . . . iii Page vi Chapter V. VI. VII. Physiological Factors Affecting Dutch Elm Disease . . . . . . . Internal Mechanics of Resistance to Dutch Elm Disease . . . . . THE GENETIC IMPROVEMENT OF ULMUS Europe . . . Netherlands England . . Italy . . . U.S.S.R. . Portugal . United States . . . . . University of Wisconsin . Agriculture Research Service of U.S.D.A. . . . . . . . . Cornell University . . . . University of Massachusetts South Dakota University . . Syracuse University . . . . Canada . . . . . . . . . . . RATIONAL AND CONDUCT OF THE MICHIGAN STATE UNIVERSITY ELM BREEDING PROJECT Materials and Methods . . . Seed Procurement . . . . Cultural Practices . . . Measurement of Characters Samara Characters . . . . Leaf Characters . . . . . Nursery Characters . . . RESULTS 0 O O O O O O O O O O O O Germination and Growth . Variation Between Species Variation Within Species UZmus minor . . . . . UZmus Zaevis . UZmus glabra . UZmus pumila . UZmus japonica UZmus parvifolia UZmus Zaciniata . UZmus eZZiptica . UZmus waZZiohiana . . Summary of Within-Specie Variability . . . . . . S iv Page 50 53 56 57 57 61 62 63 65 65 65 66 68 70 71 72 73 74 75 75 76 80 81 82 83 84 84 87 98 98 105 111 112 113 113 114 115 115 116 Chapter Page VIII. THE POSSIBILITIES OF BREEDING ELMS RESISTANT TO DUTCH ELM DISEASE . . . . . . . . 118 LIST OF REFERENCES 0 O O O O O I O O O O O O O O O O 123 VITA I O O O O O O I O O O O O O O O O O O O O O O O 139 APPENDICES O O O O O O O O O O O O O O O O O O O O O 140 Table 1. 5a. 5b. LIST OF TABLES The subgeneric classification of the Genus UZmus L O D O O O O O O O O O 0 Key to the sections, subsections a and series of UZmus L. . . . . . . . . . Natural interspecific hybrids (=x) of UZmus. Reported in the literature . . UZmus hybrid combinations produced by controlled pollinations . . . . . . . Number of collections of the four European species of UZmus represented in the M.S.U. collection summarized by country of origin . . . . . . . . . . . . Number of collections of the five Asian species of UZmus represented in the M.S.U. collection summarized by country of origin . . . . . . . . . . . . The number of accessions included in each part of the study summarized by speCieS O O O O O O I O O O O O O O O O O The average number of seedlings per seedlot and the percentage of seedlots producing at least 1, 10 or 20 seedlings, summarized by species . . . . . . . . . . Species means rated low, medium or high for each samara character and the per- cent of total variance accounted for by differences among species . . . . . . Results of artificial inoculation with Ceratocystis uZmi . . . . . . . . . . . . vi Page 35 41 77 78 81 81 89 96 Table Page 10. The average difference per trait between nine Eurasian species of Uzmus O O O O O O O O O O O O O I O O O O O O O 97 11. The number of accessions of U. minor included in each part of the study . . . . . . 99 12. The percent of total variance within U. minor accounted for by differences between regions and regional averages for each distinctive character. H,M,L= high, medium and low respectively . . . . . . . 100 13. The average differences among 26 geographical origins of U. minor. The underlined numbers refer to specific localities which are defined in Appendix D . . . . . . . . . . . . . 104 14. The number of accessions of U. Zaevis included in each part of the study . . . . . . 105 15. The number of characters in each part of the U. Zaevia study that were significantly different between regions or source within region . . . . . . . . . . . . 106 16. The percent of total variance within U. Zaevis accounted for by differences between regions and regional averages for each distinctive character. H,M,L= high, medium and low respectively . . . . . . . 107 17. The average differences among geograph- ical origins of U. Zaevis . . . . . . . . . . . ZZO vii CHAPTER I INTRODUCTION The American elm (UZmus amerioana L.) has been without question the most pOpular shade tree in the United States.. Its ready availability, ease of transplanting, relatively fast growth rate and ability to withstand rough treatment have all added to its popularity. However, its unique vase—shaped form, not found in any other tree grow- ing in this country, native or exotic, is the reason most often cited for it being the most planted ornamental spe- cies in the United States. The introduction of Dutch elm disease (Ceratooystis uZmi (Buis.) Moreau) into the United States in the early 1930's drastically altered the value of elm plantings in this country. American elm has been almost completely elim- inated from many areas and the planting of new elms has been curtailed in most regions. The advance of Dutch elm disease through the native range of American elm has stimulated extensive research on disease transmission by insects, on the nature of pathoge- nicity and on methods of controlling the disease. The ori- entation of most of this work has been toward slowing the rate of loss of existing elms. 1 New elms will not be planted in most parts of the United States until tree planters can be given a guarantee that these trees have a high probability of living a normal life span. One way to provide such elms for future genera- tions of Americans to enjoy is to develop cultivars which are genetically resistant to the disease. In the spring of 1968 these ideas prompted research workers at Michigan State University to initiate a program to develop varieties of elm resistant to dutch elm disease. Three possible approaches to developing genetically resis- tant cultivars of elm.were considered. They were as follows: (1) Search for resistance in the existing populations of American elm; (2) Introduce resistance into American elm by crossing it with other species of known resistance; (3) Develop cultivars with desirable characteristics combined with a high degree of resistance to Dutch elm disease from other species of elm. The first approach has been followed by several re- search organizations in the United States and Canada. They have shown that resistance to the disease does occur in American elm but in very limited quantities. The resistant trees discovered in these programs are often poor in other characteristics, which limits the possibility of using them directly as clones. More breeding work is necessary before this resistance can be of value to tree planters. The second approach has received attention by elm breeders in this country for many years (Smucker,1944). However, a strong compatability barrier prevents crossing American elm with other species. Attempts to overcome this barrier are being made (Derman and May,l966). Unless this barrier can be effectively overcome, this approach offers little potential for elm breeders. The third approach has been adopted by the elm breeding program at Michigan State University. Foreign species of elm are reported to contain a higher level of resistance to Dutch elm disease than American species (Went 1938, Heybroek 1968). They also show resistance to some of the other insects and diseases that plague American elm (Heybroek,1969). Much information is available on the re- sistance of these species to Dutch elm disease. However, there has been little information on the variability in disease resistance or other characters within species. The initial phase of the elm breeding program at Michigan State University was to obtain as much European and Asiatic UZmus material as possible, screen this mate- rial for resistance to dutch elm disease and evaluate its potential for future improvement work. The study described in this thesis represents the first portion of this screen- ing program. l. The objectives of this thesis study are as follows: Determine the occurrence and extent of variation in morphological characters of the samaras of nine Eurasian species of UZmus available for study. Determine the occurrence and extent of variation in morphology, physiology and resistance to Dutch elm disease of the seedlings grown from these samaras. Combine the information gained under numbers 1 and 2 above and information from the literature into a basic set of data to be used in planning and carry- ing out future elm breeding research at Michigan State University. CHAPTER II TAXONOMY A thorough taxonomic study is a valuable first step in any breeding program. This type of study allows the breeder to become familiar with available material, provides information on floral biology and reduces the chances of mistaken identification of parental material. It may provide insight into the evolutionary development of the group being studied, which will be valuable in planning and conducting research. If interspecific crosses are planned, it can provide worthwhile clues on the cross- ability pattern in the genus and will aid in authenticating any hybrids produced. The Genus UZmus The genus UZmus consists of about thirty species of trees widely distributed through the Northern Hemisphere. The genus is conspicuously absent from western North America today, but fossil evidence indicates that it was present dur- ing the Miocene when the genus reached its maximum distribu- tion (Berryjl923). The native ranges of several species of UZmus seem to be decreasing at the present time (Staub,l967). Central and northern Asia are generally considered to be the 5 center of distribution of the genus because the most species are native to that region. However, the elms of eastern North America are the most diverse group. No comprehensive monograph of the genus has been pub- lished. Heybroek (1968) compiled a survey of the genus but did not include taxonomic descriptions. Richens (1956) did an interesting and original survey of the genus based on only three characters. Other pertinent sources of reference for the genus include: Bean (1951), Elias (1970), Rehder (1940, 1949), Schneider (1917), Sokolov (1958). Regional floras also serve as a source of information on particular species. The Subgeneric Classification The genus has been divided into various sections, subsections and series. The classification generally fol- lowed was proposed by Schneider (1916) and revised by Rehder (1949). Five sections are recognized in this classification. A sixth section, Trichooarpus Cheng was proposed in 1963. The section UZmus (Syn. Madooarpus Dum.) is the only section which has been subdivided into subsections and series. This section is by far the largest in the genus and the only one represented on all four continents. Schneider (1916) seemed to use this section to accommodate species that did not clearly belong elsewhere. Heybroek (1968) in- creased the naturalness of the group by transferring four species to other sections. The subgeneric classification of UZmus is shown in Table l. The classification of species not previously classified was done with the aid of the key in Table 2. Most of the subgeneric grouping is natural. How- ever, some species which have recently become better known do not fit well into these groups. Heybroek (1968) stated that an accurate subdivision of the genus will not be pos- sible until more information is available on several Asiatic species. Phylogeny Bechtel (1921) studied the floral anatomy and mor- phology of several genera of the UZmaceae family and con- cluded that UZmus was the most primitive. He described the flowers of UZmus as highly reduced and specialized. Tippo (1938) showed data from wood anatomy which supports this arrangement. Grudzinskja (1966) recognized three types of inflo- rescences: paniculiform in BZepharocarpus, racemose in Chaetoptelea and Trichoptelea and capitate in UZmus and Microptelea. He concluded that evolution had passed from the first to the third type of varying degrees of reduction of flower number and shortening of main and lateral axes. Richens (1963) compiled a list of parasites specific to elm. By studying the occurrence of these parasites in 8 Table l. The subgeneric Classification of the Genus Ulmus L. Section Subsection Origin1 Series Species Blephocarpus Dum. U. laevis E U. americana Am Chaetoptelea (Leibm.) C. Schn. U. mexicana M U. thomasii Am U. alata Am Ulmus (Syn. Madocarpus Dum.) Glabrae (Moss.) C. Schn. Fulvae C. Schn. U. rubra Am U. elliptica E Euglabrae C. Schn. U. glabra E U. laciniata As U. bergmaniana As U. uyematsui As Foliaceae C. Schn. Nitentes C. Schn. U. wallichiana H U. minor E U. japonica As U. wilsoniana As U. chumila H U. procera E Pumilae C. Schn. U. pumila As Microptelea (Spach.) Pl. U. parvifolia As U. lanceaefolia H U. crassifolia Am U. monterreyensis M Trichoptelea C. Schn. U. serotina Am Trichocarpus Cheng U. kunmingensis As U. davidiana As U. macrocarpa As Insufficiently known to classify U. villosa U. lesuerri U. divaricata U. multinervosa Siifiim lOrigins are American, Asia, Europe, Mexico and Himalayas and Southeast Asia. Table 2. Key to the sections, subsections and series of UZmusL. A. Flowers vernal, appearing before or shortly after leaves. B. Flowers subsessile to very short pedicellate, not pendulous, samaras not ciliate on margins. C. Flowers produced in flower buds only. ........................ Section Ulmus D. Nutlet separated from the apical notch of the wing by a suture. ........................ Subsection Glabrae E. Samara pubescent over the locule. ....................... Series Fulvae E. Samara glabrous. ....................... Series Glabrae D. Nutlet near the apical notch of the wing. ........................ Subsection Foliaceae F. Nutlet in the apical part of the elongated samara.................. Series Nitentes F. Nutlet in the center of a round samara. ........................ Series pumiZae C. Flowers produced in mixed buds ........................ Section Trichocarpus B. Flowers long pedicellate, often unequally so, becoming pendulous,samaras marginally ciliate. G. Flowers fasciculate, the floral axis only slightly elongating, branches without corky wings. ........................ Section BZepharocarpus G. Flowers racemose, the floral axis slender and elongating, branches often with corky wings. ........................ Section Chaetoptelea A. Flowers AutumnaL.or appearing considerably after the leaves. H. Flowers fasciculate, not pendulous, leaves sub- persistent ............. Section Microptelea H. Flowers in pendulous racemes, leaves decidous. ........................ Section Trichoptelea 10 different areas and on different species, he was able to show some interesting relationships. The American, Euro- pean and Himalayan populations seemed to be more closely related to each other than to the Far East population. Five different elm populations occur in Europe. They are the Balkans, Central Europe, Italy, Scandinavia and England. The English population resembled the Central European pop- ulation more than it did material from adjacent areas of the mainland. Evidence of trans-Atlantic migration of Ulmus was also shown. Economic Importance Elm wood is hard, heavy, durable resists splitting and will last indefinitely under water. The wood of Ameri- can species of elm has been of minor economic importance but some European and Asiatic species produce valuable wood. The fact that the Dutch elm disease was imported into this country on elm veneer logs from Europe attests to their value. Today in the Netherlands veneer quality elm logs command the highest prices paid for any species grown in the country (Heybroek,l963b). Elm wood was used to make water pipes in many Euro- pean cities before the manufacture of metal products was perfected. Its resistance to splitting made it the primary wood used to make hubs for wooden wheels. Because of its military value for making bows, gun carriages and below 11 water parts for wooden ships, planting of elm was encouraged in parts of Europe and existing stands were preserved for future military use. The fibrous inner bark of elm trees was used to make cloth by peOple of Europe and Asia. It has also been used to make rope and twine in Europe, Asia (Heybroek,l966a), and at a local level in the southern United States (Harrar and Harrar, 1946) . The bark of elm trees was considered a good human food in Scandinavia (Heybroek,l963b), and elm was planted for this purpose in Norway. Chun (1921) reported that elm bark is still being used as a source of human food in mountainous areas of China. Bark, leaves and twigs of U. pumila were eaten in China in times of famine and the spe- cies was extensively planted for that purpose (Grudzinskja, 1961). The mucilaginous inner bark has been used as a binding agent in making flat bread from barley in Scandi- navia (Heybroek,l966a), as a folk remedy to cure throat ailments by the Indians and early settlers in the United States (Fowells,l965), and as an ingredient in the manu- facture of incense in China (Chun,l921). In almost all parts of Europe and Asia, elm leaves have been used, and highly prized, as a source of livestock feed. They have a high nutritive value and are highly palat- able to stock (Anonv 1947). Elms were planted in Europe since before Roman times for this purpose. 12 A recession of elms in Europe about 5,000 years ago discovered by pollen counts in peat formations,has been attributed to newly immigrated groups of farmers lopping the elm for fodder (Troels-Smith,l960). This practice is still widespread in the Himalayan Region and elms face extinction in some areas if the practice is con- tinued, because the constant lopping prevents the formation of flower buds and seed. Elm trees have been indispensible in the growing of grapes in Italy for the past 2,000 years.- Vines were trained on living trees and the foliage was cut in the summer to feed stock (Heybroekll966a). In spite of all the interesting uses of elm, its chief value in most regions of the world has been for orna- mental and shelter plantings. Elms have been planted in the parks and gardens of Europe and Asia since ancient times (Sokolov,l958). Today, even in Holland where elm wood receives premium prices, it is still considered mainly as an ornamental. The planting of elm for shade trees in the United States began soon after the first settlements were established on this continent. Elm is an important component of shelterbelt plantings in the Great Plains of the United States and in the Steppes of Russia (George 1937; Collins 1955; Grudzinskja 1956; Albensky 1969). 13 The Species of Ulmus The species of Ulmus included in the following discussion are listed in Table 1. Flora Europaea (Tutin et. alv 1964) was the authority used to determine the Euro- pean species of Ulmus recognized. The Asiatic species fol- low Heybroek (1968). Five species of elm that have been described from Mexico have been included even though the validity of some of them is questionable. Eastern North American Species of Ulmus Six species of Ulmus are native to Eastern North America. They are very diverse and distinct from each other. Five recognized sections of the genus are repre- sented, more than in any other elm growing region of the world. The species offer no special taxonomic problems. U. americana L. American Elm The native range of American elm covers almost the entire United States east of the Rocky Mountains and ex- tends into southern Canada. It has been successfully planted in almost all parts of the United States. American elm is the only species of Ulmus that is regularly a poly- ploid (4x=56). Florida elm (Ulmus americana var. floridana (Chapm.) Little) differs from the type variety only in degree of several quantitative characters (Kurz and Godfrey,l962). 14 Fernald (1945) recognized four forms of American elm based on smoothness of leaf and branch pubescence. Elias (1970) stated that attempts to distinguish these forms can be frus- trating since both smooth and rough leaves can be found on the same tree. U. thomasii Sarg. Rock Elm U. racemosa Thomas, non Borkh Rock elm is native from western New England to Minnesota and Nebraska south to Tennessee. It can repro- duce by root suckers but this reproduction is not believed to be important in the natural spread of the tree (Otis, 1931). Rock elm produces the hardest, toughest wood of all the elms (Harlow and Harrar;1958). Lumbering activi- ties have seriously reduced the species over much of its range. No races have been reported in the Species. Natu- ral hybrids with U. rubra are reported from Wisconsin (Fowells 1965) and Missouri (Hess and Dunn,1967). Hess and Dunn (1967) suggest that an ecological isolation between the two species prevents hybrid seedlings from surviving under natural conditions. U. alata Michx. Winged Elm The distribution of winged elm is primarily within the humid region of southeastern United States. The species 15 is of no particular commercial importance except that it is occasionally cut as a substitute for rock elm (Harlow and Harrarjl958). Its pleasing growth habit has made it a fa— vorite ornamental of many southern communities. The inner bark is fibrous and has been used locally for baling twine (Harrar and HarrarI1946). U. rubra Muhl. Red or slippery Elm U. fqua Michx. The native range of red elm covers most of the eastern half of the United States. It is the only elm species that shows seed dormancy. Red elm can spread by interconnected rhizomes during the seedling stage (Fowells, 1965). Natural hybrids of U. pamiZa X U. rubra have been reported from many locations where the two species have been growing together. Several of these hybrids have been selected and propagated for distribution by commercial nurserymen (Green,l964). U. crassifolia Nutt. Cedar Elm The cedar elm is native from southwestern Tennessee to southern Texas. An isolated'stand was recently reported from northern Florida (McDaniel and Swift71967). Standley (1922) believed that the species probably crossed the border 16 into Tamaulipas, Mexico. However, the species has never been definitely reported from Mexico. Cedar elm is the common elm of Texas where it is used as a street and shade tree (Harrar and Harrar,l946). U. serotina Sarg. September Elm September elm ranges from southern Illinois, Kentucky and Tennessee south to Alabama, Georgia, Arkansas and eastern Oklahoma. The species resembles American elm in general form, although it never gets as large. It has been used as a shade tree in Alabama and Georgia (Harrar and Harrar 1946). The September elm is the least known of all the elms of eastern North America. No races or hybrids of the species have been reported. The Elms of Mexico Five species of Ulmus have been described from Mexico. Only one of these, U. mexicana, occurs abundantly anywhere in the country. The other four species are known only from their original descriptions based on a few her- barium samples (Muller 1936, 1937, 1942; Standley,l936). More study is needed to determine the validity of these species. 17 U. mexicana (Leibm.) Planch. Chaetoptelea mexicana Leibm. U. mexicana is a tropical species ranging from central Mexico to Honduras (Standley and Steyermark,l946). It was originally described as a separate genus (Chaetoptelea Leibm.) because the fruit do not have true wings. However, this difference was not regarded as sufficient to establish a new genus and it was reduced to a section of Ulmus. Schneider (1916) put two other American species, U. thomasii and U. alata in the same section. U. divaricata C. H. Mull. This little known species was described in 1936 from near Monterrey, Nuevo Leon, Mexico. It was found grow- ing commonly in the type locality. The species seems to be closely related to U. americana, differing only in the shape of the samara tip. No information is available on the species except the original description. (Muller,l936.) U. multinervosa C. H. Mull. U. multinvervosa was described in 1937 from the mountains of Northern Coahuila, Mexico. At the time of collection, all of the trees were barren of flowers and fruits, but a single samara and some peduncles were found after a tedious search in some debris—filled crevices l8 (Muller,1937). The vegetative characters of the species do not distinguish it from U. divaricata, so the species was described on the basis of this single fruit. No additional information has been published on this species. U. monterreyensis C. H. Mull. U. monterreyensis, a fall blooming elm, was de- scribed from Nuevo Leon, Mexico in 1942 (C. H. Muller, 1942). It is the first definitie report of a fall bloom- ing elm from Mexico, although Standley (1922) believed that U. crassifolia, common in southeast Texas, probably crossed the border into Tamaulipas. The author stated that the species is rare in the type locality. Vegetatively the species is very similar to U. crassifolia (Elias,l970). The type specimen had im- mature fruit. Mature fruit and additional flowering ma- terial are necessary to determine if this is a valid spe- cies. U. Zesuerrii Standley This species is known only from the Rio Bonita area of Sonora (erroneously cited as Chihuahua under the original description)(Muller,l942). The type specimen was sterile so the species is known only from vegetative material (Standley, 1936) . 19 It is the only species of Ulmus reported from west of the Continental Divide in North America. The Elms of Europe The elms of Europe have been a source of consider- able confusion for taxonomists. Several factors have con- tributed to this confusion. Elms hybridize easily; the hybrids are generally fertile and hybridize further to produce swarms of intermediates. In Europe, man has in- terfered with the native elms for thousands of years (Trols-Smith 1960; Heybroek,1963b). As a result, in Europe today one can never be sure if a tree is native or not. This interference of man has added to the hybridization and the resulting confusion. The difficult natural conditions have caused con- siderable confusion in the nomenclature which has hindered the understanding of the genus. This confusion dates back to Linnaeus who named only one European Species, U. campestris L. This name was subsequently applied in most European countries to the most common, local form or spe- cies. Being a permanent source of error and confusion, the name was rejected as a "nomen ambigum" (Melville,l938). Further confusion arose from botanists and horti- culturists who described horticultural varieties (often clones) as botanical varieties or species. Flora Europaea (1964) recognized only five species of Ulmus in Europe. 20 Green (1964) compiled a list of cultivar names for Ulmus. These two publications should help to overcome much of the confusion in nomenclature. U. minor Mill. U. carpinifolia Gled. U. foliaceae Gilib. U. stricta Lindl. U. nitens Moench. U. glabra Mill. (non Huds.) U. densa Litv. U. campestris L. (pro parte) U. minor is native in central and southern Europe, from Spain to the Ural Mountains in the east and extending into North Africa and Asia Minor. It reproduces itself by root suckers as well as seed. The root suckering habit of the Species enabled it to withstand lopping for fodder bet- ter than non-suckering Species. It was planted for this purpose in many areas of Europe since ancient times (Hey- broek,1963b). U. minor has been a source of considerable confu- sion for taxonomists. Rehder (1949) lists twenty three synonyms of the Species. Many varieties and forms of the Species have been described. However, the validity of some of these is questionable. ULVQEabra Huds..Wych or Scotch Elm U. scabra Mill. U. montana With. The native range of U. gZabra covers most of Eur0pe. Its range extends north of the Artic Circle in Norway and as far east as the Ural Mountains. 21 U. glabra has been planted extensively as a park and street tree since ancient times. It occurs naturally on well-watered, fertile soils and is very intolerant of salinity and drought (Sokolov,l958). The roots of U. glabra do not produce root suckers. Because of this, it is the favorite understock used to propagate improved varieties of elm in Europe. Natural hybrids between U. glabra and U. minor U.X hoZZandica) Miller occur extensively in Europe. The hybrid seems to be more prominent in some areas than either parent. Many of the most desirable cultivated varieties of elm in Europe have been selected from these hybrids. U. procera Salisb. English Elm The origin of U. procera is unknown. In older books it is said to be native to Europe, introduced to England by the Romans (Bean,l951). However, the tree seems to be dis- tinctly different from U. minor growing on the continent and is not found growing wild anywhere except in southern England. It occurs in parks and gardens in Spain, but al- ways as a planted tree. It is the dominant tree in many parts of southern England, especially in hedgerows. How- ever, all of these trees seem to have come from root suckers, which the tree produces freely. The production of fertile seed is extremely rare in England, although it does produce 22 fertile seed in milder climates (Henry 1910). The tree may have originated as a hybrid from the other British elms (Bean,l951). Many horticultural varieties have been selected which indicates that genetic variability is present in the species (Green,l964). Richens (1967) reported natural hy- brids between U. procera and U. minor are common in southern England. U. Zaevis PaZZ. European White Elm U. pedunculata Foug. U. effusa Willd. U. racemosa Borkh. U. laevis is native from the Pyrenees to the Ural Mountains. The species has been cultivated since ancient times for shade and shelter in the U.S.S.R. (Sokolov,l958) but it has received little attention from tree planters in the rest of Europe (Heybroek,l968). A variety celtidea (often regarded as a separate species) has been described from the U.S.S.R. Its general growth habit is reported to resemble U. americana (Heybroek, 1968; Sokolov,1958). U. elliptica C. Kock U. elliptica is native in the broadleaved forests of the southern part of European U.S.S.R. This species is morphologically similar to the American Species U. rubra. 23 U. elliptica has received little attention from tree planters, although Sokolov (1958) states that it is suitable for planting within its native range. U. canescens Melville The native range of U. canescens encircles the Eastern Mediterranean basin, it occurs sporadically where suitable site conditions exist. The Species was first described in 1957 and is not well known. It is Similar to U. minor in many respects (Tutin, et. alv,1964). Elms of Eastern Asia More Species of Ulmus are native to Eastern Asia than any other region of the world. The taxonomy of many of these Species is not completely known. As these spe- cies become better known, they should increase our under- standing of the genus. U. pumila L. Siberian Elm U. androssowii Litv., U. pinnato-ramosa Dieck U. Manchurica Nakai The Siberian elm, often erroneously called Chinese elm, has a natural range extending almost completely across Asia from eastern China to Russian Turkestan, north to the Transbaikal retion. Grudzinskaja (1961) thinks that the 24 species may have originated in Sinkiang and was transported to the Far East and Russian Central Asia by man in ancient times. The species was first introduced into the United States in the early 1900's by F. N. Meyers and others (webb,l948). The drought resistance, relatively fast growth rate and ease of propagation seemed to make this species the answer to tree planters' problems, especially in the Great Plains Region. The species was promoted by several govern- ment agencies and planting was begun on an extensive scale. By 1948 U. pumila was recognized as the most planted tree species of all time in the United States (Webb,l948). The windbreaks and shelterbelts of the Great Plains Region made up the bulk of this planting. The extensive planting of Siberian elm created a strong demand for planting stock. The source of the seed did not seem important and it was generally obtained from the most readily available source. 'Most of the seed was collected by students of Nanking University from trees grow- ing around Nanking, China, which is about the same latitude as Houston, Texas. These southern sources were planted throughout the central and northern Great Plains Region. The problem of winter hardiness soon developed. The famous Armistice Day freeze of 1940 killed a high percentage of Siberian elm throughout the central and northern Great Plains. 25 (Engstrom and Matthew;l942). This provided one of the best examples of the importance of seed source in the his- tory of tree planting. Hardy strains of Siberian elm have been developed in South Dakota (Collins,l955), and Manitoba (Green,l96l). No one seems to like the Species very much because of its susceptibility to ice damage, relatively short life Span, and its prolific reproduction that have made it a weed in many areas. Some people have even advocated a ban on planting the species in certain areas (D. E. Hutchins, personal communication). However, no other species has been able to equal its fast growth rate, ease of establish- ment and drought resistance. Collins (1955) reported that it was still planted more than any other hardwood in the central and northern Great Plains region. U. kunmingensis Cheng U. kunmingensis was described from Yunnan province of mainland China in 1963. The original description is the only information available on the species (Cheng,l963). U. parvifolia Jacquin Chinese Elm U. parvifolia is native to a wide latitudinal range from the island of Hainan to Korea. It is also native in Taiwan and Japan. 26 Southern sources of Chinese elm are evergreen and the leaves usually persist on northern sources until mid- winter. Chun (1921) reported that the autumn foliage often turns a brillant red or yellow. The interesting, scaly bark adds to the ornamental value of the Species. The Chinese elm is a small tree that Should not be considered as a replacement for the American elm. However, it is a fine ornamental in its own right and should be planted more in this country (Wyman,l965). U. davidiana Planchon U. davidiana is native to northern China, Mongolia, Korea and the Amur region of far eastern U.S.S.R. The spe- cies is reported to be one of the hardiest of all elms (Weikoff,l94l). Very few specimens of this species have ever been planted outside the native range of the species. U. macrocarpa Hance U. macrocarpa is native to northern China, Manchuria, Korea, Mongolia and the southern part of far eastern U.S.S.R. Komorov (1936) stated that the tree is a rare relic of an earlier vegetation type. The species is almost unknown as a cultivated tree, but its broad spreading crown and Shiny foliage make it an attractive tree, especially when fruit- ing (Sokolov, 1958) . 27 U. wilsoniana Schneider U. wilsoniana is native to Central China. Schneider (1917) reported that the species is not common in the type locality. U. wilsoniana is growing in Arnold Arboretum from scionwood sent from Central China in 1910. Grafted speci- mens of this tree have been planted in other arboretums and botanic gardens. The glossy green foliage of this tree makes it an attractive ornamental. It is of interest to elm breeders because it is resistant not only to Dutch elm disease, but also to elm bark beetle and elm leaf beetle. U. uyematsui Hayata U. uyematsui is native to the central part of the island of Taiwan at elevations of 1500-2500 meters. The species is of no economic importance and has received little attention from researchers. U. laciniata (Trautv.) Mayr. U. laciniata is native to Japan and eastern China as far north as the Sakahalin peninsula. Seed germination for the species is very low. Komorov (1936) states that this Species is very similar to U. glabra. It seems to be a far eastern race replacing U. gZabra after considerable geographic gap. 28 Features typical of U. laciniata may be found in Ural Moun- tain populations of U. glabra. The Shape of mature speci- mens of U. laciniata is reported to resemble U. americana (Heybroek,l968). U. bergmcnniana Schneider U. bergmanniana is a farily common tree in western Hupeh and eastern Szechuan provinces of Central China. The species is cultivated in its native range (Schneider,19l7). Very few specimens of the species have been grown outside the native range. U. japonica Sarg. U. propinqua Koidz. U. davidiana var japonica Nakai U. japonica is native to northern China, Korea and Japan extending into the Transbaikal Region of the U-S-S-R- It is one of the hardiest of all elm (Weikoff, 1941). Komorov (1936) described five variations of the species based on combinations of leaf smoothness, twig corkiness and growth habit. He stated that these five forms are not strictly delimited geographically but they are clearly differentiated ecologically. The wood of the tree furnishes valuable lumber in Japan. Chun (1921) stated that the tree sometimes assumes 29 the characteristic vase shape of the American elm and should be useful as a street tree because of its rapid growth, ornamental character and hardiness. Elms of the Himalayas and Southeast Asia The elm flora of this part of the world is still not completely known. Most of the elm trees occur in re- mote areas that have not been fully explored botanically. Three species have been described from the region and at least one more is waiting to be described (Heybroek and Melville, in press). The destructive practice of lopping trees for livestock fodder has made it difficult to find naturally occurring elms except in parks or in very inaccessible regions. U. wallichiana Planch U. wallichiana is native in the western Himalayas from Nepal to Kashmir at altitudes of 3,500-10,000 feet. It is often planted in its native range and the branches lopped every other year for cattle fodder (Brandeis,l9ll). The practice of lopping all young branches for livestock fodder has made it very difficult to collect seed from this species except in a few protected areas (Heybroek, personal communication). 30 The Dutch have used this species extensively in their breeding program in recent years. It seems to cross easily with other members of the section Ulmus. U. 'chumiZa' This is a new Species from the western Himalayas. Heybroek (1968) used the name but a valid taxanomic des- cription of the Species has not been published. U. X brandisiana is reported by Heybroek to be a natural hybrid between this species and U. wallichiana. U. Zancifolia Roxb. U. Zancifolia is a tropical species native to southeast Asia. It is the only species of Ulmus with a native range that extends south of the equator. The dormant period of the species coincides with the seasonal rainfall pattern. The species usually blooms in midwinter and has been classified by authors as both spring and fall blooming. The species is closely related to the fall blooming U. parvifolia (Touw and Van Steenis, 1968). Hybrids between the two species have been reported from the island of Hainan (Heybroek, 1968). 31 U. villosa Brandeis U. villosa is native to the northwest Himalaya re- gion from the Indus to the Jumna rivers. The wood of this species is of no particular value but the branches are lopped for fodder. This species produces root-suckers to an extraor- dinary extent and probably reproduces itself in this way. The seed is reported to fall off the tree before it is ripe. Cytology The base chromosome number of Ulmus is x=l4 (Sax, 1933; Leliveld,l934; Krijthe,1939). All species of Ulmus in which the chromosomes have been counted are regularly diploid (2n=28) except U. americana which is a tetraploid (2n=56) (Sax 1933, Darlington and wylie,1955). Sax (1933) believed that the species was probably an autotetroploid. Staub (1968), after studying all the evidence available, concluded that more data is needed to determine the origin of the polyploid condition. Krijthe (1939) reported that U. pumila var. pinnato ramosa sometimes showed 28 chromosones in the haploid meta- phase stage of the pollen grain. The Swedish Forest Tree Breeding Association at Kalltorp produced triploid and tetraploid forms of U. gZabra. These forms proved to be 32 just as susceptible to Dutch Elm Disease as normal U. gZabra and it was considered useless to include them in a hybridiza- tion scheme (Went,l954). Ehrenberg (1949) studied asynapsis (decreased chrom- osome pairing at first metaphase) in U. gZabra. A high de- gree of sterility resulted when asynaptic X asynaptic elms were crossed. Only 19 seedlings were produced from 1636 seeds obtained in the crossing. Fifteen of these were tetraploid and the remaining four triploids. Asynaptic X normal showed a higher fertility level. Seedlings from this cross were all triploids. All seedlings obtained from normal X normal crosses were normal diploids. The triploid seedlings produced by Ehrenberg (1949) were slightly more vigorous than diploids while the tetraploids grew only about half as fast. Stomate size of the triploid seedlings was intermediate between the larger tetraploid and the diploids. Grudzinskja and Zakharyeva (1967) studied the karyo- types of seventeen forms of six species of elm native to the 0.5.8. R. Individual chromosomes were similar in all forms studied and they concluded that karotypes are of no diagnostic value in Ulmus taxonomy. Polyploids have been induced in normally diploid species of Ulmus by colchicine application (Heimburger, 1966; Derman and May,l966; Heybroek,l968). Most of this work has been aimed at overcoming the crossability barriers 33 between the tetraploid U. americana and diploid species (Heimburger,l966; Derman and May,l966). Heybroek (1968) reports that tetraploids have been made of some of the best clones produced in the Dutch breeding program. When this material reaches flowering age it will be crossed with diploid varieties in an attempt to produce vigorous trip- loidS. Lester (1968) attempted to produce polyhaploids (haploids doubled) in American elm. Flowers of American elm were pollinated with non-viable,irradiated American elm pollen and with pollen of the diploid species, U. pumila and U. rubra, which normally do not cross with Amer- ican elm. Over 300 filled seed and 168 seedlings were pro- duced from the crosses. All of the seedlings proved to be tetraploid. Lester stated that the pollination seemed to increase the amount of selfing in the American elm parents. Counting the chromosomes of all seedlings to de- termine the success of attempts to produce polyhaploids has been very time consuming. An attempt is now being made to develop a screening technique using marker genes that will make it possible to divide seedlings into groups of high and low probability of reduced chromosome number (Lester,l968). CHAPTER III HYBRIDIZATION IN ULMUS Natural Hybrids Natural hybrids have been reported many times in Ulmus. However, only about a third of the species have been involved and only a few of the possible combinations between these species are reported. Natural hybrids that have been reported in the literature are shown in Table 3. In almost all cases natural hybrids have been the result of man interfering with the native vegetation. The most classical example of this is in southern England where there is evidence that man has imported and planted elm for the past 2,000 years (Richens, 1955). The elms of southern England reproduce mainly by root suckers. However, about every ten years a crop of viable seed is produced (Richens, 1955). New genetic com- binations are formed in these occasional flurries of re- productivity and then propagated asexually. As a result of this, the population of elms of southern England has evolved into a group that have confused taxonomists for the past 300 years (Melville, 1946). 34 35 Table 3. Natural interspecific hybrids (=X) of Ulmus. Reported in the Literature. americana thomasii U. americana rubra U. thomasii rubra minor procera U. minor >< U. x x U. procera U. glabra x glabra japonica >< X pumila X X x U. pumila wallichiana U. japonica davidiana U. wallichiana >< U. chumila U. parvifolia lanceaefolia davidiana X U. parvifolia lanceaefolia x C: C'. C! C! C: C‘. C: C! CC} C: C C'. C'. canescens X X 36 Richens (1955, 1958, 1959, 1961a, 1961b, 1965, 1967) compiled a series of papers on the local elm flora of England based on historical and morphological data. Jeffers and Richens (1970) combined the morphological data from the earlier studies into a multivariate analysis. The results of these studies Show that four groups of elms occur in England. They correspond with U. procera, U. minor, U. glabra and U. minor X U. glabra hybrids. U. glabra is the only elm native to England. Hybridization Techniques Flowers of Ulmus are minute, only about one-eighth inch across. Each flower contains 4 to 8 stamens. The flowers in a single cluster do not develop simultaneously. As much as two weeks may elapse between the emergence of the first and last flowers. Pollination bags need to be removed daily during this period to inspect the flowers and remove any newly emerged stamens. Emasculation is time-consuming and limits the num- ber of pollinations that can be accomplished. Recent hy- bridization work in Ulmus has depended on self sterility to prevent selfing. A marked protogyny that prevents selfing has been reported in U. americana (Shattuck,l905), U. minor and U. glabra (Henryyl910). Went (1954) reported on ten years of selfing stud- ies involving 189 trees. Self fertility ranged from 0 to 37 100%. In addition to the large difference among individ- uals, she also found a large year to year variation within a tree. Of the 189 trees involved, 12% produced over 10% fertile seed. She concluded that emasculation was the only sure method of knowing the parents of a seedling produced from controlled pollination. Lester (1968) reported that the amount of selfing in U. americana was apparently increased by pollination with non-viable irradiated American elm pollen or pollen of diploid species of elm. He also found a wide variation in the amount of selfing among individual trees. Cut branches of elm placed in water can be forced to flower and produce fruit in the greenhouse. The his- torical development of this technique was reviewed by Pauley (1949). Wright (1949) studied the effect of time of collection, branch size and greenhouse environment on the production of elm seed on cut branches. One to three foot branches, three fourths of an inch in diameter, gave the best results. Branches collected shortly before the natural flowering time flowered more quickly and produced more fruit than earlier collections, unless flowervbuds had been damaged by cold weather. GreguSS (1970) Showed a tree to tree variation in the dormancy required before a tree could be forced to flower in U. minor and U. glabra. One U. minor tree produced normal pollen from branches cut immediately after natural defoliation. 38 Collins (1967), Ozolin (1958) and Went (1954) stated that production of seed on cut branches gives er- ratic results. The method is not used extensively in most elm breeding work. However, the forcing of cut branches to collect pollen is commonly used. Greguss (1970) reported that pollen collected in this way was more fertile than pol- len produced naturally on the tree in U. laevis, U. minor and U. glabra. Collins (1967) improved on the cut-branch technique by grafting scionwood from mature flowering trees onto potted seedlings. The potted grafts maintained their small size, flowered the first year after grafting, and those that survived were still flowering after eleven years. How- ever, the production of seed on this grafted material was erratic. Crossability Patterns of Ulmus Knowledge of the crossability pattern in Ulmus is far from complete. Most crosses reported in the literature have not been authenticated or are poorly authenticated. More detailed hybridization studies in the genus are needed, especially studies using rigorous rules for certification of hybridity. Differences in flowering time form an obvious crossability barrier between the spring and fall blooming 39 elms. No effort to overcome this barrier through pollen storage or other techniques has been reported. A strong crossability barrier seems to isolate U. americana from the rest of the genus. The nature of this incompatability is not known. Sax (1933) theorized that it might be due to unequal chromosome.pairing at‘ meiosis between the tetraploid U. americana and diploid species of elm. This theory is generally accepted by most elm workers and efforts to overcome the incompat- ability by inducing tetraploids of other species is in progress at several places. Tetraploid U. pumila pro- duced by the U.S.D.A. (Derman and May,l966) and growing - at the Shade Tree Research Center, Delaware, Ohio, has now reached flowering age and it should be known in a few years if the increase in ploidy level will overcome the compatability barrier. Heybroek (1968) suggested that part of the incom- patability may be due to genic differences between sec- tions. U. laevis, the other well known member of the section BZephocarpus, is difficult to cross with other Species. U. americana X U. laevis has been reported at least four times in the literature, but only one of the reports states that the progeny showed characters inter- mediate between the two parents (Collins,1967). Sax (1933) reported abundant seed set and high germination from this cross, but was unable to find any triploids in the progeny. 40 Graft compatability may parallel pollination com- patability. Collins (1967) presented data that suggests a compatability barrier to grafting may exist between the sections BZepharocarpus and Ulmus. He reported a high de- gree of success grafting scions of U. laevis on U. americana rootstocks. Complete failure resulted from grafting the same material on U. pumila rootstock. U. pumila, U. rubra, and U. pumila X U. rubra hybrids were all successfully grafted on U. pumila rootstocks. Attempts to graft this material on U. americana rootstocks showed little success. Weaker compatability barriers may be present be— tween the other sections of Ulmus. All crosses reported between sections have either not been authenticated, or resulted in poor seed set and low survival except U. laevis X U. brandisiana (a little known species from China) ree ported by Collins (1967). It is interesting to note that Went (1938) reported that U. brandisiana could only be grafted on U. laevis root- stock. Attempts to graft it on U. glabra and U. minor root- stocks were unsuccessful. The hybrid combinations reported in the literature are shown in Table 4. A list of the references in which elm hybrids have been reported is given in Appendix B. 41 Table 4. Ulmus hybrid Combinations produced by con- trolled pollinations. americana americana N U. I» U. laevis U. thomasii laevis alata U. rubra procera thomasii alata rubra U. U. minor procera minor 3 U. H U. glabra *‘ U. pumila U. U. bergmaniana glabra japonica pumila 4 3 4 3 1 2 1 [—1 japonica bergmaniana 2 wallichiana wilsonia U. wallichiana 1 laciniata U. wilsoniana l 1 X X N N U. laciniata O c c: c c: c c: c c: c c: c c: c c: c U. parvifolia parvifolia X crassifolia C'. 1. Easy cross 4. Very difficult cross 2. Intermediate in difficulty X. Cross has been re- ported in the litera- 3. Difficult cross ture but no indica- tion of the difficulty was given. CHAPTER IV THE DUTCH ELM DISEASE History The story of how the Dutch Elm Disease was intro- duced into the New World on elm veneer logs imported from Europe in the early 1930's is well known. These logs brought both the deadly fungus and its main insect carrier, the European elm bark beetle. On their way to sawmills in the midwest, these logs gave rise to several centers of in- fection. An eradication program launched by the U.S.D.A. was able to stamp out some centers, but others soon Spread out of control. Today the disease has covered almost the entire native range of the American elm. The origin of the disease in Europe has never been determined. Eastern Asia is generally believed to be the ultimate source of the disease. It is the only area of the world where a significant number of elm species Show at least a moderate degree of resistance to the disease. The disease has never been reported from eastern Asia, but most of the area where it might occur has never been in- vestigated by scientists interested in the disease. Another possible theory is that a harmless saprophytic fungus mutated into the deadly parasite we now know. 42 '43 The disease was first reported from Tilburg, Holland in 1919, hence its name. The first possible oc- currence of the disease dates back to 1917 in northern France and even 1912 in England. The history of the dis- covery of the disease and isolation of the causal orga- nism has been reviewed by Westenburg (1932) and Elgersma (1969). The disease Spread rapidly over Europe, reaching Italy and Yugoslavia in 1929 and Tashkent in Uzbekistan, U.S.S.R. beyond the Urals in 1939. This amounted to an average advance of 150 miles per year. The disease is now spread over most of Europe and western Asia (Heybroek, 1966a). Description of the Fungus and the DiSease Dutch elm disease is caused by the fungus Ceratocystis ulmi (Buis.) Moreau. The fungus develops in the living wood as a parasite and in dead elm wood as a saprophyte. In living trees the fungus occurs in the water conducting vessels of the xylem. Yeast-like spores are produced by the fungus by budding. These spores are carried through the tree by the flow of sap. Gum and tyloses produced by the fungus are deposited in the vessels interrupting the flow of water and causing wilting and death of the tree (Walter,et. alq.l943). 44 Production of toxin by the fungus in culture has been demonstrated by Zentmeyer (1942), Dimond (1947); and Santamour and French (1958). Toxin has not been isolated from diseased trees and the importance of it in the patho- genicity of the fungus is not known. Zentmeyer (1942) and Dimond (1947) believe that the toxin formed by the fungus is the main cause of abnormalities in the tree. They point out that it is unlikely that tylose and gums could com- pletely block a large trunk. The rapidity with which the fungus develops lends support to their concept (Tchernoff, 1965). Kerling (1955) injected cultured toxin into a tree. The tree developed symptoms similar to those produced by inoculation with a spore suspension. After its host dies, the fungus, growing as a sap- rophyte, produces spores under the loosened bark and in insect galleries formed between bark and wood. The spores are produced in sticky masses which can cling to the out- side of an insect and be carried from diseased to healthy trees. The main insect vector involved in the spread of the disease in America is the European elm bark beetle ScoZytus multistriatus. In Europe the larger ScoZytus scolytus is probably the main vector, but Scolytus multistriatus and Scolytus laevis are important in the spread of the fungus. The native American elm bark beetle (HyZuroyopinus rufipes) may also spread the disease in 45 America. The beetles feed in the small twig crotches of healthy elm trees, but they breed only on recently cut, dead or dying elms. In the spring the adult beetles emerge from infected wood and fly to nearby living elm trees to feed. When the feeding injuries penetrate through the bark of the wood, the disease organisms may be introduced into the vascular system of healthy trees. Adult beetles may fly several miles in search of suitable breeding places and, thus, transport the disease from one area to another. Elm trees killed by the disease provide breeding areas for the beetles. This results in an increased pop- ulation of beetles in the area which increases the spread of the disease. This vicious cycle continues until some- thing breaks the causal chain or until the population of elms available for feeding declines until it affects the beetle population. Development of the Disease After Artificial Inoculation Elm bark beetles were used in the first attempts to artificially inoculate elm trees with Ceratocystis uZmi fungus. Contaminated beetles were placed on the tree in cheese cloth cages and allowed to feed for five days to insure inoculation. Later investigations showed that in- fection could be obtained more easily by introducing a spore suspension directly into the xylem with a syringe (Went 1938). Tchernoff (1965) was able to obtain more 46 uniform infection by placing a drop of Spore suspension on the stem and then wounding the stem with a sharp pointed knife. Positive results could be obtained only when the xylem had been cut followed by active absorption of the sus- pension. The number of spores in the inoculum seems to have little affect on the success of inoculations. There was no marked difference in severity of symptoms between trees in- oculated with 200,000 spores per tree and trees inoculated with only 200 spores (Tchernoff,1965). Age of Spores does affect the success of inocula- tion. As spores get older they lose their virulence with no apparent change in morphology. Tchernoff (1965) recom— mends that spore suspensions less than 10 days old be used for comparative inoculations. Strains of C. uZmi which vary in virulence have been reported many times. Therefore, it is recommended that several isolates of C. uZmi be mixed in the Spore suspension to insure virulence of the inoculum. If pos- sible, the virulence of the inoculum should be checked on material of known susceptibility. The way in which test plants are raised may deter- mine their reaction to inoculation. Trees should be grow- ing vigorously when inoculated. If growth is inhibited, fewer symptoms will be shown. After transplanting, trees need at least one growing season to become established 47 before they will exhibit typical symptoms. Hot weather and direct sunlight after inoculation are most favorable for expression of disease symptoms. Watering plants before or directly after inoculation is favorable for obtaining best results (Tchernoff,l965). Environmental Factors Affecting the Disease The spread of Dutch elm disease depends upon a rather close c00peration between the fungus and its insect vector. The fungus is not very dangerous if the insect vector is not present. The population of beetles may fluc- tuate widely under the influence of weather and availability of dead elms for breeding places. Even the percentage of fungus-carrying beetles in a newly hatched population will fluctuate with weather conditions and the presence or ab- sence of mites in the galleries (Franzen,l939). The Netherlands forest service kept a record of all elms in the country that had to be removed because of the disease between 1930 and 1944. Not only was the total num- ber counted, but also, the number of diseased trees on dif- ferent soil types (Went,1954). Between 1932 and 1937, 34% of the elms on peat soil and 37% of those on sandy soil be- came diseased. During this same period, only 8% of the elm on clay soils became infected. Vazda (1952) found that centers of infection were in groups of elms growing on sites where soil moisture had 48 undergone a sudden unfavorable change in moisture content. The affect of soil moisture on expression of the disease was studied by Kais,et. al.,(l962). Soil type, moisture- holding capacity and water availability influenced both degree of infection and duration of susceptibility to the disease in Ulmus americana. Burger (1938) found that elms growing near the Dutch coast were relatively unaffected by the disease. The line drawn to denote the area where large elm planta- tions still occurred in 1938 shows a striking conformity with lines denoting an equal mean wind velocity of 4.5 meters per second during summer (Heybroek,l957). This suggests that wind has a strong influence on the severity of the disease. Went (1954) believed that this influence was primarily on the insect that carried the disease. Heybroek (1957) stated that the cause of this influence is not known. It may be that the wind decreases tree growth which makes them less susceptible. A high relative hu- midity would be maintained in this area by the wind which blows from the sea. This would avoid a vapor pressure deficit which may hasten the development of a wilt disease like the Dutch elm disease. Went (1938) stated that arti- ficially inoculated seedlings exhibit few symptoms in cool moist growing seasons. Although a high relative humidity during the grow- ing season seems to inhibit the expression of symptoms in 49 diseased trees, there is ample evidence that a high rela- tive humidity is necessary at the time of inoculation. Tyler, Parker and Pope (1940) studied the affect of rela- tive humidity on young elm trees in the greenhouse. Plants grown in a moist chamber for a time and then transferred to the greenhouse developed more symptoms than trees grown in the greenhouse for the entire period. Kais, Smalley and Riker (1962) found that inoculation of susceptible elm seedlings with dry spores resulted in infection only when moisture loss was prevented by covering the wounds with masking tape. Under field conditions they found that in spite of a high relative humidity during the entire sus- ceptible period, the incidence of infection in uncovered inoculated wounds was consistently lower than in similar taped wounds. A suspension of spores in water gave a higher incidence of infection than dry spores. The highest rates of infection were obtained using a spore suspension in com- bination with taping the wounds. Temperature also affects the expression of the disease. Tyler (1945) reported that a temperature of 15 to 29°C. favored disease development. He found that trees ex- posed to 26°-29°C. wilted in three to five days after inoc- ulation. Temperatures of 9.5° to 12°C. and 32° to 37°C. inhibited wilt expression. Kais, et. a1.,(l962) studied the interaction of soil and air temperature on the development of Dutch elm disease. Maximum infection occurred at air 50 temperatures of 24°C. with soil temperature of 26°C. Soil temperatures seemed to affect disease development more than air temperature. Physiological Factors Affecting Dutch Elm Disease Vigorously growing trees are more susceptible to the disease than slow growing ones (Went,l954; Heybroek, 1957; Peace,l960). Tchernoff (1965) stated that any fac- tor which tends to increase the vigor of a susceptible tree will enhance the development of disease symptoms after in- oculation. Transplanted elm generally do not Show symptoms when inoculated until they have had at least one growing season to recover (Went,l954; Heybroek,1957; Smalley,l963; Tchernoffil965). Grafted material also needs at least one year to become established before it can be successfully inoculated. Vigorous trees have a better chance of recovering from an infection than slow growing trees (Peace,l960). The fungus is generally confined to the large vessels of springwood. It seldom penetrates the denser summerwood. If the tree can survive the initial attack long enough to lay down a new layer of wood it has a good chance of re- covery if not reinfected (Banfield, et. a1,,l947). Fungus was isolated from recovered elms in Great Britain after a period of up to 7 years (Peace,l960). In recovered trees 51 the fungus becomes confined to vessels invaded during the year of symptom expression and is buried by subsequent rings of healthy tissue (May,l935; Pope,1943; Banfield, 1968). Peace (1960) found that spraying trees to prevent insects from reinoculating them greatly increased their chance of recovery. He also found a difference between varieties in ability to recover. Manka (1953) reported that U. laevis was better able to recover from the disease than U. minor. Smalley and Kais (1966) found a large difference between and within species in the time and duration of maximum susceptibility during the growing season. The dissolution of the end walls of the earlywood vessels has been closely correlated with initiation of the period of susceptibility in American elm. Pomerleau (1965) observed a close relationship between early cambial activity and the beginning of susceptibility. The increase in resistance during the growing season is closely related to the begin- ning of latewood production (Banfield,l968). Branched seedlings generally develop fewer symptoms than plants with a single stem (Arisumi and Higgens,1961). Smalley (1962) believed this difference to be due to the Shortened period of elongation of multi-branched seedlings. The period of maximum susceptibility can be extended by the application of nitrogen fertilizers (Smalley,l962). 52 Heybroek (1957) stated that 2-year-old seedlings are not susceptible to the disease. Three-year-old seed- lings develop symptoms of the disease upon inoculation only if they are extremely well grown. He observed that when a group of seedlings are inoculated year after year additional individuals become infected. He interpreted this to mean that susceptibility ot the disease increases with age. In most clones a jump in susceptibility has been observed at about the same time as a tree begins to flower. Smalley (1962) studied the susceptibility of young American elm seedlings in the greenhouse. He found that seedlings of American elm were susceptible during their first growing season from the time the first true leaves emerge until the cessation of terminal growth. There was no evidence of juvenile resistance. Tchernoff (1965) stud— ied the effect of seedling age on susceptibility. Age of seedling did not affect susceptibility. In another experi- ment he compared two groups of plants of the same age, be- longing to the same clone, one group in the juvenile and the other in the adult stage of development. This is pos- sible in elms because stem cuttings maintain the adult character while callus cuttings from the same tree are in the juvenile stage (Tchernoff,1963). No difference in de- gree of susceptibility was observed between the two groups. 53 Smalley and Kais (1966) showed that the duration of susceptibility in American elm decreases with age if trees are inoculated in the crown. However, age did not affect susceptibility if all trees were inoculated in the trunk. Internal Mechanismsgf Resistance to Dutch Elm DiSease McNabb, et. a1. (1971) hypothesized two possible mechanisms of resistance. First, resistant hosts limit establishment of the pathogen and prevent the systemic development of the disease. Second, external symptoms of disease are limited but systemic establishment of the pathogen takes place. The host—parasite relationship is more neutral than pathogenic. Experimental evidence suggests that localizing of pathogen establishment is the primary mechanism involved in most cases. McNabb, et. a1. (1971) proposed four pos- sible host factors that could regulate this general mech- anism of resistance. They are: (l) vessel size and group- ing in the functional xylem; (2) relative time required for tylose growth response in parenchyma cells adjacent to in- vaded vessels; (3) gross nutrient of xylary fluid for fun- gus growth, reproduction and toxin production, and (4) in- hibition or promotion of fungus growth reproduction and toxin production by hormone like host metabolites. These factors may work singly or in combination. 54 Pope (1943) compared the vascular anatomy of U. pumila and U. americana and suggested a possible anatomical mechanism. He thought that early initiation and subsequent production of latewood isolated the fungus in the portion of the wood where inoculation occurred and shortened the period of susceptibility. However, Buisman (1936) stated that wood structure had no affect on the movement of spores through either resistant or susceptible elms. Dimond, 2E; ‘21. (1949) found no difference in vessel length and vessel diameter between U. americana and U. pumila. Elgarsma (1969) did an extensive study of factors affecting the resistance of elm to C. uZmi. He found that vessel diameter strongly influenced water movement through the xylem in the material studied. He concluded that ves- sel diameter is a primary factor in determining suscepti- bility or resistance in elms. McNabb, et. a1. (1971) measured the vessel diameter and area of contiguous vessel-groups (product of average vessel diameter and average number of contiguous vessels) in twenty-three different selections or clones of European and American elms of known resistance. They found that vessel-group size was more highly correlated with suscepti- bility than individual vessel size. Sinclair, et. a1. (1970) found resistance in U. americana was closely correlated to vessel Size diameter. Vessel diameter is now used as one of the criteria of 55 selection in their elm breeding program at Cornell Univer- sity. New candidates for selection are included in the program for further study if the average vessel diameter is less than 60 microns. Smalley (1962) was able to increase the resistance of susceptible American elms for up to two years by treat- ment with trichlorophenyl acetic acid (T.C.P.A.). Forma- tion of tyloses in the vessles of treated trees seemed to be the main factor involved in the increased resistance. The plugging of vessels by the growth of tyloses (Banfield 1968) and accumulation of gums (Elgarsma,l967) has been suggested as a mechanism of localizing the Spread of the fungus. However, McNabb, et. a1. (1971) found that many vessels in a group are not adjacent to parenchya cells. They suggest that these vessels do not become plugged with tyloses. Elgarsma (1967) could find no Significant differ- ence between resistant or susceptible elms in the sugar content of the sap. He did report a higher concentration of free amino acids in the sap of resistant elms. However, after a more extensive study (Elgarsma,1969) he concluded that there was no direct relationship between amino acid content and resistance. CHAPTER V THE GENETIC IMPROVEMENT OF ULMUS Improved varieties of elm have been selected and propagated by tree planters for several hundred years. Green (1964) in his "Registration of Cultivar Names in Ulmus" lists over 400 named cultivars of elm. Undoubtedly many other cultivars have been grown that were never named. Many of these were never accepted by tree planters and were planted only on a very limited scale. However, some were very successful and were planted extensively over a wide range. Almost all elm cultivars have resulted from in- formal selection by nurserymen of chance seedlings or mu- tations. Only a few have been produced as a result of for- mal selection and breeding studies. The first such program was begun in Holland in 1928. Any new selections that are released to the public will need to incorporate resistance to Dutch elm disease with other desirable characteristics. Most progress will be made through formal selection and breeding programs. The following chapter summarizes the progress of formal studies aimed at the genetic improvement of Ulmus. 56 57 Europe Netherlands Elm was the most important tree grown in the Netherlands before Dutch elm disease invaded the country (Heybroek,l957). Over 95% of the elms planted in the country belonged to a Single clone (U X hoZZandica c1. belica). This clone proved to be highly susceptible to the disease and the entire elm population of the country was threatened. An elm-disease committee was established to study the problem and make recommendations. They con- cluded that no other tree species could replace elm in many sections of the country, and the most practical method of combating the disease was to develop new genetically re- sistant varieties of elm. As a result of this decision, an elm breeding program was established in 1928. The progress of this program has been reviewed by Went (1938, 1954) and Heybroek (1957, 1966b). The initial phase of this program consisted of screening Ulmus material for possible sources of resistance. Three groups of elms were included in the screening program. The first group consisted of cultivated varieties of Ulmus which had been developed in Europe. Both asexually propa- gated clones and seedlings from these clones were collected and tested. About 10,000;1 or 2 year-old seedlings of Euro- pean species of Ulmus made up the second group. These seed- lings were obtained from nurseries throughout Europe. The 58 third group was made up of seed collections of Ulmus spe- cies native to North America and Asia. Most of this seed was obtained from arboretums and botanical gardens. Differences in susceptibility were exhibited be- tween clones included in the study. However, none of the cultivars contained enough resistance to be recommended for general planting (Went,l938). U x hoZZandica cl. 'vegeta' was one of the clones that Showed some resistance to the disease. Later, this clone was considered to be acceptable for planting in areas of the country where the disease had not been very severe (Heybroek,l957). Screening of the seedlings was more successful. A number of resistant individuals were obtained in this way. However, the percentage of resistant individuals in a population was very low. Only seven resistant individ- uals were selected out of 5,700 U. minor seedlings tested in the first 10 years of the program. Six of these seed- lings were selected from a group of elms from Spain. The other selected seedling originated in France. One selected Spanish seedling was released to growers in 1936 under the name 'Christine Buisman'. Al- though this clone is highly resistant to Dutch elm disease, it proved to be very susceptible to another disease caused by the fungus Nectria cinnabarina (Tode) Fr. Because of this susceptibility, its planting was soon stopped in Holland. The clone has been grown successfully in Italy 59 and the United States where Nectria disease is not a pro- blem (although it is present in both areas). In 1946 a selected seedling from France was re- leased under the name 'Bea Schwartz'. This selection was resistant to both Dutch elm disease and Nectria cinnabarina, but it had a poor growth form and was never accepted very well by tree planters. All eastern North American species of elm screened in the program were very susceptible to the disease. The Asiatic material ranged from very susceptible to highly resistant. In 1937 hybridization was started between the var- ious Species and varieties that remained from the original collections after screening. After the disappointing re- sults of their first two introductions, the Dutch developed a set of eight criteria. to be used in selection. They are as follows: . Resistance to Dutch elm disease . Resistance to Nectria cinnabarina l 2 3. Resistance to frost 4. Resistance to wind 5. Fast growth 6. Good form 7. No epicomic shoots or suckers 8. Valuable timber 60 If any of these characteristics were missing they consid- ered the tree to be of little value (Heybroek,1957). As a result of the hybridization and selection program, the clone 'Commelin' was released in 1961. This clone has good form but only moderate resistance to Dutch elm disease (Heybroek,l96l, 1963). The clone 'Groenveld' was released in 1963. It is highly resistant but rather slow growing. More new hybrid varieties are now being tested for possible future release. Future elm breeding work in Holland will concen- trate on incorporating Himalayan species of Ulmus into the breeding program. An attempt to produce fast growing triploid varieties is also being made (Heybroek,l968). Several organizations in the United States have been interested in importing the resistant Dutch varieties and growing them in this country. Heybroek (1966a) pointed out that elms in Holland are grown primarily for lumber, ornament and for protection from the wind. The less Shade a tree produces the better, Since the sun does not shine very much in Holland. Trees selected for Dutch conditions would probably not be accepted in this country where elm is grown primarily for shade. The elm disease committee of Holland has made both seed and scionwood of their most promising material avail- able to research workers in this country. This material is presently being tested at several locations in the United States and Canada. 61 England The British policy toward the control of Dutch elm disease is to do nothing about it except occasionally re- move dead trees (Heybroek,l966). This official policy was adopted as early as 1929 and has proved to be quite effec- tive. Peace (1960) found that the diseaSe, though wide- spread in England, has killed only 10 to 20% of the British elm population in the past thirty years. After the peak years of 1931 and 1936-1937 the disease has exhibited a gradual decrease in the country. The reason the disease has been less serious in England than on the continent or in America is not com— pletely understood. The elms of England are generally more resistant to the disease than the elms of the rest of EurOpe or America. However, this difference alone is not sufficient to account for the difference in severity of the disease. The maritime climate of Britain may limit develop- ment of the disease. Heybroek (1957) stated that the dis- ease has not been as serious in areas of Holland near the seacoast as further inland. In some cases a clone that is highly susceptible in a more continental climate has not been attacked in these areas. Recovery of diseased trees is commen in England (Peace,l960). Apparently, it is unpredictable whether a 62 diseased tree will be healthy or diseased again next year. Trees with three-fourths of their crowns dead from the dis- ease have been known to recover (Heybroek,1966a). One factor that is well known is that far fewer elm trees would grace the English countryside today if the early policy of removing trees as soon as they showed symptoms of the disease had been continued (Peace,l960). Selection and testing clones of U. procera resistant to Dutch elm disease has been carried out for several years by the Forestry Commission, Alice Holt Lodge, Wrecclesham England (Jobling,1969). Methods of asexual propagation of selected clones are also being studied. Italy The growing of grapes in Italy has depended upon the use of elm trees to serve as props since ancient times. U. minor was the species used for this purpose before Dutch elm disease invaded the country about 1929. An elm breeding program to develop varieties resis- tant to the disease was reported by Goidanich (1938) and Goidanich and Azzarolli (1938, 1939). Both artificial in- oculation and hybridization were used in the program. As a result of this research U. pumila var. pinnato-ramosa largely replaced U. minor for use in the grape growing in- dustry. Castellani (1966) reported that new resistant strains of U. minor from Spain have now largely replaced 63 the less desirable U. pumila var. pinnato-ramosa. Italian research is now concentrated on the selection of U. minor and U. minor X U. pumila hybrids that combine disease re- sistance with useful agricultural characteristics. U.S.S.R. A Russian program to develop elms resistant to Dutch elm disease began in 1946. Experiments were conduc- ted in 22 locations in southwestern Siberia. In these areas almost all of the elm trees had been affected by the disease (Ozolin 1958, 1959). The first step in the program was to locate surviving trees in areas heavily infected by the disease. Attempts to inoculate these trees in the field to determine resistance were not successful (Ozolin,l959). Later experiments concentrated on seedlings from the ser lected trees. Selections were in open-pollinated and con- trol-pollinated seed lots. Results of studies on Species resistance to the disease reported by Ozolin differ from the results reported from other countries. The largest difference is that the Russians rated U. pumila as highly susceptible to the dis- ease. Elm researchers in other countries usually regard U. pumila as the most resistant species. The U. pumila material used in the Russian study came from Kharobovsk, in the Transbaika1.a region probably not sampled in other elm research projects. The source difference may account 64 for the unusual results. U. pumila var. pinnato-ramosa, which is native to the region where the study was conducted, was rated as highly resistant to the disease. In addition to species resistance, the Russians also found individual resistance in otherwise susceptible species. Controlled pollination was carried out on mature trees in the field and on cut branches in the greenhouse. Blossoms were emasculated in the control pollinations. Studies of the inheritance of disease resistance, tree form, and seedling vigor were reported by Ozolin (1958). In addition to disease resistance, vegetative pro- pagation and seed storage was studied. U. laevis seed stored in a paper bag two years under room conditions still germinated but germination time was doubled. Grafting gave better results than cuttings. Studies on the biochemical nature of resistance included amylase activity and carbohydrate concentration. Ozolin (1959) stated that results of these studies were highly contradictory and do not allow for conclusions to be drawn. Two thousand trees resistant to the disease were distributed to nurseries in 1958. No information on the success of this material is available. 65 Portugal Selection of U. glabra for resistance to Dutch elm disease has been conducted for the past eleven years at the Estacao De Biologia Florestal, Oeiras, Portugal. Five clones which showed resistance to five different Portuguese strains of the fungus have resulted from this work. The resistance of foreign clones to Portuguese strains of the fungus is also being investigated. Resis- tant clones produced by the Netherlands elm breeding program are being utilized in this study (Natalina F. dos Santos de Azevedo, personal correspondence). United States University of Wisconsin A program to obtain and screen worlddwide seed col- lections of elm for resistance to Dutch elm disease was be- gun by the Department of Plant Pathology, University of Wisconsin in 1957. To date almost 500 seed collections and clones from all over the world have been screened. (Dr. Eugene Smalley, personal communication.) These collections include species of other genera in the UZmaceae family as well as Ulmus (Smalley and Riker,l962). As a result of this testing over 1,000 resistant individuals representing eight species and several hybrids of elm are now growing in the elm arboretum near Madison, Wisconsin. 66 In 1966 an elm genetics and beeding program was developed jointly between the Department of Forestry and Plant Pathology to determine patterns of inheritance for disease resistance and ornamental traits (Lester,1969). The material now being studied at Wisconsin can be divided into two groups. One group is made up of U. americana and the other includes the three diploid species, U. pumila, U. japonica and U. rubra. The objective of the Wisconsin work includes clarification of crossability pat- terns in Ulmus, estimation of self compatability and deter- mination of the extent to which desirable characters can be transferred among species and individuals. Screening of established and newly obtained progenies for resistance to Dutch elm disease is being continued. Anatomical and bio- chemical studies of resistant and susceptible individuals and selection of resistant individuals for desirable orna- mental characteristics are also being conducted (Lester, 1969). Agricultural Research Service of U.S.D.A. The U.S. Department of Agriculture began studying the possibility of selecting American elms resistant to Dutch elm disease shortly after the disease was discovered in Ohio in 1930. It has been a major project Since 1937 (Smucker,l944). In the initial phase, 35,000 elm seedlings collected from throughout the natural range of the species 67 were grown in test nurseries and inoculated with the fungus. Only two seedlings remained after three consecutive seasons of inoculations. These later proved susceptible to new strains of fungus (Clapper,l952). In the 1930's a U.S. financed field laboratory was established in England to compare resistance of American elm with European species. American elm was the most sus- ceptible of all species tested. The English-American pro- gram.was closed out when it became threatened by World War II (L. R. Schreider, personal correspondence). One of the original objectives of the U.S.D.A. pro- gram was to combine the resistance of Dutch elm disease and phloem necrosis of U. pumila, the high quality lumber of U. thomasii and the beauty of U. americana all into one variety. Compatability barriers stopped this program in the first stages (Smucker71944). Several thousand seedlings were collected by elm breeders in the Department of Agriculture from mixed plant- ings of U. pumila, U. rubra and U. americana in the Great Plains Region. They were looking for natural hybrids that might be resistant to the disease (Smucker51944). All U. pumila hybrids proved to be too susceptible to Dutch elm disease for recommended use (Swingle,l961). Swingle (1961) reported that over 200,000 American elm trees have been tested for resistance to the disease, of which 19 were moderately resistant. Most of the 68 selections were carriers of one or more viruses that make them difficult to propagate. The viruses may have influ— enced their reaction to inoculation with Dutch elm disease fungus. Tests for resistance to phloem necrosis were more encouraging. Individuals that had survived in areas heavily infected by the disease were selected. Resistance tests on seedlings and clones of these survivors showed positive results (Swingle,l945). Clapper (1952) stated that 2,000 American elm had been selected for their resis- tance to phloem necrosis. In later research, seedlings were first screened for their resistance to Dutch elm disease. If they showed resistance, they were subjected to phloem necrosis. The results of this work have not been published. Present research at Delaware, Ohio, and at the U.S. National Arboretum is devoted to elm hybridization and further selection in American elm for resistance to Dutch elm disease (Egolf,1968). Cornell University Selection of American elm for resistance to Dutch elm disease was begun by Cornell University in 1934 in cooperation with Boyce Thompson Institute for Plant Re- search. Elm seed and seedlings were collected from 69 throughout the northern half of the range of American elm (Welch,1949). Fifty-five collections of material involving over 21,000 individual trees were included in the study. In the early stages of the program, all seedlings that exhibited external symptoms after artificial inocula- tion were discarded. This standard was later relaxed and modified as the number of trees diminished and it became apparent that American elm displays a variable syndrome after inoculation. The criteria of selection used in later stages of the program allowed retention of a tree if (a) it survived infection and recovered from external symptoms, (b) and was not successfully attacked by elm bark beetles, (c) and displayed potential for growth rate and form char- acteristic of American elm (Sinclair, et. al., 1968). The first inoculations confirmed the low level of resistance of American elm; 94% of the seedlings wilted so badly that they were discarded. Only in 9 of the 55 col- lections were more than 10% of the trees alive after the first inoculation. By 1967, 17 survivors had been selected by one or more of the criteria listed earlier. Of these, six survived infection but were very slow growing trees; and, one did not have the typical form of U. americana. Only one was considered promising and it later developed elm leaf mosaic, a virus disease and was withdrawn from possible release (Sinclair,1970). 70 Selection was not as rigid in the Cornell program as it was in other American programs where all trees that showed symptoms were discarded (Oullett and Pomerleau,1965; Smucker,l944). Sinclair suggested that if the stern se- lection criteria used in other programs had been employed at Cornell, they might not have recognized any natural re- sistance in U. americana. Whether the apparent resistance exhibited by the selected trees is a strongly heritable character remains to be determined. Present work at Cornell involves selection of new material from areas infected by the disease for a long per- iod of time. New selection techniques are being developed. University of Massachusetts Several years of research on the resistance of elms to Dutch elm disease have been conducted by the Shade Tree Laboratory of the University of Massachusetts. Attempts to induce mutation of American elm seed in hopes of finding resistance to the disease has been a major part of the pro- gram. During the past ten years over six million elm seeds have been irradiated. From this material 110 trees survived two or three inoculations. From 4,573 trees inoculated in 1970, 68 were apparently wilt free. Reciprocal crosses of resistant elm trees are being attempted (McKenzie,l970). 71 South Dakota State University A serious weakness of Siberian elm in the northern Great Plains has been its lack of winter hardiness. In 1944, workers at South Dakota State University began a pro- gram to develop hardier strains of the species. Two sources of seed, from Manchuria and one of unknown origin from a nursery in Washington, were included in the study (Maxon, 1951). A source from Harbin, Manchuria, proved to be the best adapted to South Dakota conditions. The seedlings of this source became dormant early in the fall and were not subject to damage by sudden freezes. In 1952 the strain was named "Chinkota" and released to the public. The seedlot used to develop "Chinkota" elm was also used to develop the "Harbin," "Manchu," and "Dropmore" strains in Canada and many people consider them all to be the same. The one major difference is that the research workers in South Dakota realized the importance of maintain- ing a source of high quality seed and developed a seed or- chard to produce it. Since it was first introduced "Chinkota" elm has been produced under certification stan- dards of the Seed Stocks division of South Dakota State University and only certified seedling stock has been sold to growers (Collins,l955). Collins (1967) did an extensive study of Ulmus hy- bridization and hybridization techniques. He also did an intensive study of U. pumila X U. rubra hybrids. Backcross 72 progenies showed characteristics intermediate between the F1 and the recurrent parent. No studies on resistance to Dutch elm disease were included in the project. Syracuse University In 1957 the genus Ulmus was added to existing forest tree improvement programs at the State University College of Forestry at Syracuse University. The objective of this project was to: 1. perfect the hybridization methods using cut branches in the greenhouses 2. determine the crossability pattern within available Ulmus species 3. determine the best possible grafting method for each species 4. provide hybrid seed and test material resistant to Dutch elm disease Four American elms from the city of Syracuse, that had been infected with the disease and apparently recovered, and eight resistant clones from the Cornell University pro— ject were included in the study. Selection of other Ulmus species was based on assumed resistance or on known success- ful crosses (Anonv 1962). The results of the hybridization study were published by Wineski (1960) and Britwum (1961). No attempt to certify the authenticity of the hybrids produced in the study has been published. 73 Canada The Canada Department of Agriculture began an elm breeding program at the L'Assomption Experiment Station, L'Assomption, Quebec, in 1948. Three years of crossing work with U. americana and U. pumila resulted in one trip- loid seedling (the only authenticated case of a triploid seedling resulting from a controlled cross between the two species) (Heimburger, 1966). The triploid seedling resisted artificial inoculation of Ceratocystis uZmi for several years, but finally succumbed. Screening of U. americana for resistance to Dutch elm disease was begun in 1952. From 1952 to 1956, 43,000 seedlings were inoculated. After four annual inoculations only two seedlings remained. Irradiation of elm seed to induce mutation as a possible method of finding resistance was done in 1955. In 1956 thermal neutrons were used for the same purpose. Since 1955, 20,000 seedlings, from seed subjected to in- duced mutation, have been inoculated. Three of these plants have escaped infection after three inoculations. According to a newspaper release, the 'Quebec' elm, a slow growing but highly resistant clone of American elm, was released to the public in 1970. CHAPTER VI RATIONAL AND CONDUCT OF THE MICHIGAN STATE UNIVERSITY ELM BREEDING PROJECT The elm breeding project at Michigan State Univer- sity is part of a comprehensive Dutch elm disease research program developed jointly by the Departments of Entomology, Plant Pathology and Forestry. The heavy loss of native Michigan elm to the Dutch elm disease was the primary rea- son for initiating this program. The biology and control of the insect vector and the potential value of systemic fungicides for controlling the disease in existing elm trees are being investigated by the Departments of Ento- mology and Plant Pathology respectively. The selection and breeding of elms for resistance to Dutch elm disease was undertaken by the Department of Forestry with assis- tance from the campus Parks and Planning office. A number of tree species from other genera have been suggested as possible replacements of the American elm. However, none of these species contain all of the desirable characteristics of American elm. There is still a need for elm genetically resistant to Dutch elm disease. 74 75 Developing varieties of elm with the desirable char- acteristics of American elm combined with an acceptable level of resistance to Dutch elm disease is the objective of the elm breeding program. After considering the possible alternatives, it was decided that these objectives could be most easily obtained by working with the foreign species of Ulmus American col- lections of Ulmus are limited. Therefore, it was decided that the most efficient approach to the problem would be to obtain as much foreign material as possible and screen it for disease resistance and other ornamental traits. Development of varieties of elm that have a broad genetic base and can be produced from seed is the goal of the Michigan State University elm breeding program. This is the main difference between the M.S.U. program and other elm breeding programs. Materials and Methods Seed Procurement In the spring of 1968, Dr. George Parmalee, Campus Curator of Woody Plants, and Dr. J. W. Wright, Professor of Forestry, wrote personal letters to cooperators in Europe and Asia asking for collections of seed from elm trees growing in their vicinity. They received 576 76 individual tree collections of Ulmus seed in response to these requests; 516 arrived during the spring and summer of 1968 and the remainder during the following year. Nine species of Ulmus, representing three sections of the genus are included in the collection. Seven of these species are from the section Ulmus. They include: U. elliptica of the series Fulvae; U. glabra, and U. laciniata of the series Glabrae; U. wallichiana, U. minor and U. japonica of the series Nitentes, and U. pumila of the series Pumilae. The other two species are U. laevis of the section BZepharocarpus and U. parvifolia from the section Microptelea. Open pollinated seed collected from the most prom- ising hybrid clones developed in the elm breeding program at Wageningen, Netherlands are also included in the collec- tion. The number of seed collections of each species is summarized by country of origin in Table 5. As soon as it was received, each seedlot was given an accession number (MSFG, Michigan State Forest Genetics number) and a description of the seedlot was entered in the permanent accession record maintained by the Department of Forestry. Cultural Practices Elm seed is reported to lose its viability rapidly in storage. Therefore, all seed was planted as soon as possible after it arrived. To insure against loss, a 77 mmv hm mHH and h HmH m m mmumum copes: .mH ma 4 s s mfl>mamomss .NH HHH Ne cm 5 me .m.m.m.D .HH 0H m h a mwcmaom .oa mm mm 0H m condom .m cm mm m H nonmanonumz .m ms ma km samuH .k km ma m «H summcsm .G m m oomoum .m Hm m ma OH scaguwo .4 mm m om m camamcm .m mm ha mm ha oflxm>oamocoouu .m mm m ma n mauumsd .H Hobos moflunhm mH>mmH .D munoam .D moaumflaam .D MOSHE .D camfluo souoo mmmmmmm mo mnucsoo (I .chHHo mo muucsoo an pouflumfifidm coauomaaoo .D.m.z map .cfl cmucomoummn mssNb mo moflowmm commouom Hsom men mo mGOfiuooHHoo mo Hmbeoz .mm manna 78 mNH h we mH 0H mw mH m m mmumum meHcD .m mH mH .m.m.m.o .5 H H cmumemm .m h m H mocmHHonumz .m h v m MOHOM .v we v OH m we comma .m HH HH 33H . m H H 983% . H Hmuoe mcmwaoHHHmz mHHEdm MHHOMH>HMQ MDMHcHomH MUHCOQMn chHHo .D .D .D .D .3 mo wuucsou g .chHHo mo auucsoo an omuHumEESm SOHuooHHoo .D.m.z map CH omucmmmumou meHb mo mmHommm GMHm¢ w>Hm may no mcoHuuwHHoo mo Hmnfidz .bm pome 79 portion of each seedlot was planted in the greenhouse. The remaining seed was planted directly in the nursery at the Tree Research Center on the Michigan State University Campus. Seedlings germinated in the greenhouse were potted in two inch clay pots when they were 2-3 weeks old and transplanted to outdoor beds 2-3 weeks later when they were 4-6 inches tall. In the spring of 1969 the nursery-grown seedlings were transplanted to a nursery study consisting of four replications of 10 tree plots. A modified randomized com- plete block design was used. All accessions of a species were randomized and grouped together. These species groups were randomly distributed through each of the four repli- cations. This made it easier to detect differences both within and between species. Not all accessions were included in all replica- tions. An accession was included if it contained at least 20 seedlings, enough for two replications. In a few cases where a Species was represented by only a few accessions, it was included in the test if only 10 seedlings were available. In the spring of 1970 every seedling in the repli- cated test was inoculated with a suspension of Ceratocystis uZmi: (Dutch elm disease) spores using the method described by Tchernoff (1965). A drop of spore suspension was placed 80 on the stem of each seedling and the stem.was wounded, al- lowing the spore suspension to penetrate into the stem. Dr. John Hart of the Department of Plant Pathology prepared the spore suspensions and supervised the inocula- tions. All 7,800 seedlings in the replicated test were inoculated within 48 hours. To test the effectiveness of the spore suspension 10 young American elm trees 10 to 15 feet tall were also inoculated. All of these trees devel- oped symptoms and three died completely. Measurement of Characters The choice of characters used in this study was based on three things: review of the literature, study of the material itself, and ease and rapidity of measure— ment. For convenience, the study was conducted in three parts. The first part was the study of a 4-samara sample which had been selected from each seedlot before it was planted. Study of a lO-leaf sample collected from each accession in the nursery at the end of the first growing season was the second part. Characteristics of seedlings in the nursery were studied in the third part. Some collectors removed the wings from the seed before it was shipped. In other cases, the samaras were too badly damaged for measuring so that all accessions were not included in the samara study. Leaf measurements were made only on accessions that produced at least 5 seedlings. 81 Accessions that were received during the second year or late in the first growing season were not included. Not as many accessions were included in the leaf study as in the samara study. The nursery study included only those accessions that produced enough seedlings to be included in the replicated nursery test. The number of accessions in- cluded in each part of the study are summarized in Table 6. Table 6. The number of accessions included in each part of the study, summarized by species. Number of accessions included in: Species Samara Study Leaf Study Nursery Study U. minor 132 92 65 U. elliptica 7 l 0 U. glabra 140 48 12 U. japonica 39 16 10 U. laciniata 10 2 l U. laevis 112 ' 90 78 U. parvifolia 18 l 1 U. pumila 38 33 29 U. wallichiana 6 4 4 Total 519 303— 233 seedlot and mounted for further study. Samara Characters Four undamaged samaras were selected from each Twenty-six different 82 characters were measured, estimated or computed on these samaras. Characters that were present on every samara of one species and not shown by any other were not used. These include the ciliate samara margins of U.Zaevis and pubes- cence over the locule of U. elliptica. Eight different measurements of Size were made on each samara. Since these measurements are dependent on overall samara size, they are also expressed as ratios of total length or width. Nine other characters of size, shape or color were scored on a scale of one to five. Each character was scored by choosing five samaras to represent the five grades of the character and comparing all others to these samples. A com- plete list and description of all characters studied is given in Appendix C. Leaf Characters A leaf collection was made in the fall of 1968 by collecting the largest leaf from five different trees in each accession. A Similar collection was made in 1969 by collecting the third leaf from the terminal in each tree. If the third leaf was damaged, the next lower undamaged leaf was taken. Seventeen different characters were measured, es- timated or computed for each sample. Thirteen of these 83 were variables that depended on overall size and were also expressed as a ratio of leaf length making a total of 29 leaf characters. The same characters were measured on 60 accessions in 1969. There was a high correlation between years (r= .70 to .90). Since this material did not seem to add any new information, the measurements were discontinued. A complete list and description of all leaf characters stud— ied is given in Appendix C. Nursery Characters In the fall of 1968 four different characters were scored on each accession in the unreplicated nursery plant- ing. Sixteen additional characters were studied during the next 2 years in the replicated planting. A list and de- scription of the twenty nursery characters studied is given in Appendix C. Susceptibility to Dutch elm disease was recorded as the number of trees in each plot that showed branch die- back from the disease (few died completely). In August of 1970, there was a large difference exhibited among Species in damage by a leaf-eating insect. Although the insect was not identified, damage was scored as the number of trees attacked in each plot. CHAPTER VII RESULTS Germination and Growth Although this study was not designed to test dif- ferences in germination, several factors were observed that may be of value to future studies of this type. For example, there was no significant difference in the number of seedlings produced from seedlots shipped moist and those shipped dry. The best evidence of this comes from 12 seed- lots representing three species from Austria. Each seedlot was divided into two equal parts by the collector. One half was shipped moist and the other half dry. The method of shipping had no significant effect on the number of seedlings produced. Comparison of other seedlots shipped moist and dry also failed to show any significant difference. However, the seed shipped dry was the easiest to handle. The time of year that the seed was planted seemed to be one of the major factors affecting germination and growth. In almost all cases seeds planted early in the growing season produced the most seedlings. Ulmus pumila and U. laevis were the only species that showed a significant amount of germin- ation when the seed was planted after the middle of July. 84 85 The average number of seedlings per seedlot and the percent of seedlots that produced some seedlings are summarized by species in Table 7. Five seedlots of U. glabra from Riga, Latvia, were collected June 20, 1968 and shipped by surface mail. They arrived in East Lansing, September 17th and were planted September 19th. The seed did not germinate until the fol- lowing spring, but the percentage of germination was the highest of any U. glabra in the study. One seedlot of U. pumila var. 'Chinkota' from South Dakota had been stored for three years by the collector. This seedlot also showed a high percentage of germination. These results indicate it may not be as necessary to plant elm seed as soon as possible after maturation as is generally believed. In a study of this type there are obvious advantages to storing the seed and planting it all at the same time. In spite of the southern origin of many of the accessions, there was very little winter injury to the nur- sery grown seedlings. Almost all injury was confined to accessions that had been planted late in the growing season and did not have time to harden off before winter. The only exception was U. wallichiana, a species native to the Himalayan region, which was killed to the ground both of the first two winters. 86 mm on om mm mcHHnsm noose mm mm mm vm mcoHSOHHHms .2 me am mm mm «HHSSE .3 mm mm me he mHHomH>umm .2 mm mm mm mm mH>mmH .D o OH mm m oumHSHUMH .3 OH ow om m moHaOEmn .o m mH mm m MHQMHm .D o o mm m MUHUQHHHm .3 mm on mm mm HocHE .D mocprmom om mmcHHUmwm oH mcHHpmmm H uOHcmmm mom mmcHHpomm mo mmHoomw pmooooum umcu muoprmm mo mmmucmonmm Hones: mmmum>¢ mnoHpmmm mo mmmucmonmm map can HOHpomm mom mmcHHpmmm mo Hones: mmmuo>m one .moHoomm Mb omuHHmEESm .mmcHHpmmm om Ho oH .H ummmH um mcHosooum .h OHQMB 87 The potted seedlings produced in the greenhouse did not grow as well after transplanting as did the nursery grown stock. Data from this material was not used in any of the analyses. The roots of the potted stock were curled in the soil ball transplanted with them. They were slow to grow into the coarse textured nursery soil. As a result, these seedlings were subject to frost heaving during their first winter and suffered unusually high mortality. These results indicate that the work involved in germinating and transplanting greenhouse seedlings was not worthwhile. All seedlings transplanted in 1969 showed high sur- vival except U. wallichiana. The poor survival of this species was probably due to winter injury. U. laevis seed- lings were severely attacked by a leaf spot disease (Gnomium uZmi) which affected their growth. All other seed- lings grew well after transplanting. Variation Between Species There were highly significant (.01 level) differ- ences between species for every character studied. The high level of significance obtained (less than .0005 for all characters but two) indicates that the differences are real in spite of the differences in sample size. The percent of total variance accounted for by differences between species varied greatly among characters. Samara characters were easiest to recognize, but samara, 88 leaf and nursery characters were of almost equal value statistically in distinguishing between species, as is shown in the following tabulation. Variance accounted for by differences among species Character Range Average Samara characters 11 to 93% 55% Leaf characters 17 to 75% 47% Nursery characters 11 to 86% 56% Species were divided into three groups (high, medium, or low) for each character depending on the spe- cies mean. Tukey's W. procedure was used to determine which species belonged in each group. There was slight overlap between groups, but the groups differed signif— icantly (.01 level). This arrangement makes it possible to see which species could be distinguished from one another for each character. The species groups are shown in Table 8a, 8b, and 8c for samara, leaf and nursery char- acters respectively. The use of a factor analysis or other multivariate technique to determine which characters were of most diag- nostic significance was considered. However, Jeffers and Richens (1969) tried several of these techniques in a study of the elms of southern England. They obtained the same results as earlier studies using simplier techniques. The 89 cumcmH MAMEMm m3.oH.Mm.HE sm.mm >H.Hm.Ho mm \omMQ ou ucHom umosz .vH m3 numcoH om.>H on.HE .mm.0H.Hm.Hm he mum88m\nupH3 mnoEmw .MH m3 aumsoH sm.>H on.HE .mm.oH.Hm.Hw Hm mnmfiom\£umsmH nouoz .NH m3.sm 0H.Hm.Ho .mm.>H.Mn.HE mm cumcmH mumeom\£ouoc comm .HH mm.>H m3.9m.mn.HE 0H.Hm.Ho mm sumsoH museum\£umcmH comm .oH on.HE m3.0H.Hm.Hm sm.ma.>H aw camcoH mum58m\£umcmH ommm .m Hm.Hm.HE sm.>H.0H m3.mm.8n eh aumcmH mnmfimm .m sm.Hm.Hm.HE 83.0H mm.>H.Mn mm cutH3 poem .5 Hm.Ho.HE 83.5m.oH.mm mm.>H on once ou ucHom ummpwz .m m3 Hm.Hm 0H.HE .Sm.om.>H.mm Nb buoH3 mumfiom .m om.Hm.Hm.HE >H.mn m3.mm.0H HH numcmH couoz .e 83.5m 6H.Hm.Hm .mm.>H.mn.HS em nouocuwmmm .m Hm.Ho.HE sm.>H o3.mm.0H.mm mm sumcmH comm .N Hm.Hm.HE 83.0H.Mn sm.om.>H vw numcoH ommm .H Hmuou m0 w cmHm ESHUoz 30H wosoHHo> mews use moHoQO .oc Hmuooumeo mo momma nuH3 mmHowmm somzumm .mmHoomm macaw moocmHmMMHo we no“ omucsooom mOGMHHm> Hmuou mo usmoumm on» new Hmuomnmno mnmfimm comm How smHn Ho EdHUwE .30H owumu osome moHowmm .om mHan 90 .onE:& .buom .owNoamosom .aumm «Hozwsuwfi «cowtomoh .bflMn «osocmm .buHm «oowummNNmuHm mum «owomoN .oroweowHNcs .bum3 .bu>H «ouowSHSUH .SHOH mGOHumconop moHommm« HOHOO mcH3 can MOHoo 83 oh Hm.Ho.HE .sm.mm.>H.oH me comm cmozuon ummuucou .mm >H .0H.Mn.Hm.Hm.HE 5m m3.mm Hm mowamo unnumHmHmm .mN 83.0H.Mn.Hm.Hm om.mm.>H.HE he museum mo mmmcomuu .vm mm.>H.mfl.Hm.Hm 0H.HE 83.5m we HoHoo comm .mm mm.>H 0H.Hm.Hm m3.sm.on.HE em HoHoo mcHz .NN m3.mm.>H.oH on.Hm.Ho.HE um um women loum .Hm sm.m«.Hm.Ho >H.0H.HE o3.om me numSMH onuw .om m3.>H.Mn sm.HE mm.oH.Hm.Hm mm nouoc mo mmoccmmo .mH m3.sm.>H 80.0H.Hm.HE om.Hm mm mango mmom .mH m3.sm ofl.Hm.Hm.HE mm.>H.oH mm cameo awe .hH m3 cupH3 mm >H.0H.Mn .sm.Hm.Hm.HE mm mumEMm\£uUH3 comm .mH sumcmH so.mm.>H m3.oH.Hs an.Hm.Hm He mum58m\suoH3 comm .mH Hmuou «0 w cmHm EoHooz 30H mocoHnm> memo was mmHommm .oc Houomumsu mo momma £UH3 mmHommm cowsumm A.o.ucoov .Mm mHQmB 91 cm mOMMHSm mmmH 0H.Hm 83.Hm .ma.>H.on.HE mm Homes mo wocmommnsm .ow m3.>H.Hm mm.HE om.0H.om.Hm mm cumcoH oHOHumm .mm >H.mH Hm mm.mH.Hm.Hs om suHmum>no .mm 0H.Hm m3.>H.Mn.Ho om.mm.HE we GOHuMHHmm Mommaluommz .hm >H.0H.Hm no.8m.Ho m3.sm.HE Ho SOHumnumm HocHEIHommz .om >H.UH.Hm on.Hm m3.sm.mm.HE vm nutH3 GOHumuumm .mm >H.0H.Hm m3.Mn.Hm sm.mm.HE em numsmH SOHumnuom .vm 83.>H.0H.Hm on sm.m&.Hw.HE om sumoo SOHumHHom .mm oH.Hm m3.>H.Hm sm.mm.mH.He on EH» op ucHom ummcHz .mm mcoHu 0H.Hm 83.8n.Ho.HE sm.mm.>H mm Imuuom mo Hones: Hance .Hm sm.mm mGOHu m3.0H.Hm >H.Mn.Hm.HE mm Imuuom Momma mo Hobeoz .om 83.0H.Hm >H.tn.Ho 5m.om.HE No mch> mo Hmnfidz .mm oH.Hm m3.>H.mH sm.mm.Hm.Hs Hm auqu mmmH .mm oH.Hm m3.>H.Ho SQ.MQ.Mn.HE Nb numGoH moon .hm Hmuou m0 w anm EoHooz 30H mocoHHo> memo pom moHommm .oc umuomnmno mo momma SUHB moHoomm cmmsuom .mmHoomm macaw mmocmHoMMHp he now pmucnoooo mUSMHHm> Hmuou mo unmouom on» com Hmuomuono mmoH zoom How anA no EdHooE .30H pouch memos mmHoomm .bm meoB 92 m3 numcmH mm.>H mfl.Hm.Hs .sm.oH.Hm mm HmmH\numcmH :oHumuumm .om . om camcoH m3.mm.>H.HE mm .0H.Hm.Hm mm momH\cumwt coHumuumm .mv m3.5m.>H camcoH mmoH mm.Hw oH .mfl.Hm.Hs SH \mHu op HcHoo ammon .mv >H numsmH mooH\mc0Hu om.HE wB.om.ma .oH.Hm.Ho on Imuuwm mo Hogans Hobos .hv m3.>H cumcmH momH\mc0Hu sm.om.HE oh .0H.Hm.Hw hm Imunmm Momma mo Hmofioz .mv sumcmH SQ.MQ.HE m3.>H.Mn.Ho 0H.Hm mm mmoH\mch> mo Honfisz .mv mm.>H.mH.Hs m3.oH.Hm sm.Hm mm bumqu mmmeauuHs mama .4. mSOHumHHom mo Hones: Hmuou\mc0Hu om.Hm.HE 83.0H.Ho sm.mm.>H mm Imunmm Momma mo Hmnfisz .mv m3 chHumuumm Momma mo mm.oH.Ho SQ.Mn.Ho.HE >H Hm HmnEdc\mch> mo Hobaoz .Nw mUMMHSm momH @3.Mm 0H.Hm.Hm.HE sm.mm.>H mm Hoon mo mocmomnsm .Hv Hobo» m0 w smHm SSHpmz 30H mocmwum> 8...... amazofi mo momma SHH3 mowowmm comzuom u no A.U.ucoov .nm OHQMB 93 sumamH m3.ma.>H.HE on.Hm 59.0H.Hm om mooH\£umcmH oHOHumm .mm m3 numcmH sm.>H ma.mH.Hs .oH.Hm.Hm o. mmmH\HuHmum>no ..m on numcmH mooH\c0Hu .>H.0H.Mn.HE 5m m3.Hm.Hm mm Imuumm nonmaluonmz .mm m3.sm numsoH momH\c0Hu mm.>H on.HE .0H.Hm.Hm om Imnnmm HOGHEIHonmz .Nm 63.5m camcoH «moH mo.mm >H.Hs .oH.Hm.Hm Hm \auqu aoHumuumm .Hm Hmuou mo cmHm SSHomz 30H mocoHHm> mama cam moHommm .oc Hmuomnocu mo mcmmE zqu moHoomm comzumm 1.8.ucoov .nm mHnme 94 M3 >H.0H.Mn.Hm HE 9m.mm vm uoHoo mHSB .hm sm.om 0H.Mn.Hm sm.om.HE mo HoHoo Eoum .mw sm.mm 0H.Hm.HE o3.>H.Mn om mo>moH somuo .mm o3.>H.oH.Hw 5m.0n.HE mm mm huoncH umoum .vm uoosm dm.mm 0H.mn.HE m3.>H.Hm Nb unosmluooam mcoq .mm 5m.Mn HE m3.mm.oH.Hm mm HoHoo HHom .mw cm on m3.HE .mm.>H.0H.Hm on mmHsu mxuou .Hm am.m3 mmmH .hummcm >H .mm.mfl.Hm.HE om «EN: Sawsotw .om >H.Mm.HE Hm m3.mm.0H.Hm we muoncH umonm .mm >H >H.mH.He Hm .oH.mfl.Hm.Hs HH uoHoo HHmH 8mm .mm m3.sm.mm.oH momH .musmcH >H Hm .mH.Hm.Hs.Hm o. HEN: 53.202S .km mo>mmH om.>H.on.HE Ho m3.om.0H.Hm hm muHmommo unmouom .mm Hmuou m0 w amHm EsHomz 30H mUQMHHm> mass can mmHowmm .oc nouomumsu mo msmofi cuH3 moHoomm somzumm .mmHommm macaw moocmquMHo ma non poucooooo oocoHHm> Hobo» mo unmouwm map can Hmuomumso wnmmnsc some now nmflz Ho EsHomE .3OH pouch momma mmHommm .Om mema 95 HoHoo mHsu can Eoum mm.0H.Mn >H.Hm.HE m3.5m Hv somzvmn ammuucou .mn 5m mm.0H.Hm.HE 83.>H.mn hm xomuum DoomcH .vh >H on.Hm.HE m3.sm.ma.oH om muHHHnHummomsm .mn 0H.Mm.Hm m3.sm.HE mm.>H mm ucmHmc mmmuo>¢ .Nn sm.mm.>H on.Hm m3.oH.HE we use mchomH .Hn mm SQ.>H.0H.HE mh.Hm mH mHmsm nosmum .on sm.mm.8n HE m3.>H.0H.Hm mm mcHsocmub whopcoomm .mw 5m.mm >H.Mn.HE m3.>H.Hm mm mocosoub mo Honesz .mo Houou m0 w cmHm EsHpmz 30H oocmHHm> mam: cam mmHommm .oc kuomumnu mo momma suHB moHoQO comsumm A.U.ucoov .Om OHQMB 96 percent of variance due to between Species differences given in Table 8 is a good indicator of the discriminating ability of various measured traits. Results of the artificial inoculations are summar- ized in Table 9. These results are very similar to results obtained at the University of Wisconsin (E. B. Smalley, personal communication) in a similar study. The Dutch hybrid seedlings were grown from open pollinated seed from clones of known resistance produced in the elm breeding pro- gram at Wageningen, Netherlands. The results obtained from artificial inoculation of these seedlings was closely corre- lated to the resistance rating of the parent. Table 9. Results of artificial inoculation with Ceratooystis uZmi. Percent of Total num- seedlings ber of No of Showing Range in seedlings accessions Species symptoms percent tested tested U. minor 50 6 to 90 2934 93 U. glabra 51 32 to 70 226 11 U. japonica 24 0 to 78 122 11 U. laciniata 14‘ 14 7 1 U. laevis 86' 40 to 100 2629 80 U. parvifolia 3, 3 4O 1 U. pumila 8 0 to 40 1016 37 U. wallichiana 6. 0 to 8 34 4 Dutch hybrids 12 0 to 44 388 15 97 Species means for each character were converted to units of standard error to make comparison between species. The average differences between species was then computed as the sum of the differences between characters divided by the number of characters. The average differences are presented in Table 10 to Show the relationship between spe- cies. Table 10. The average difference per trait between nine Eurasian species of Ulmus. l. 1. U. elliptica — a. 2. U. glabra 0 3. 3. U. laciniata 3 2 i. 4. U. wallichiana 3 3 3 ‘5. 5. U. minor 3 3 4 4 ‘6. 6. U. japonica 5 3 5 3 l l. 7. U. pumila 9 4 7 6 3 5 8. 8. U. laevis 8 5 6 4 3 5 7 9. U. parvifolia 10 7 8 7 4 6 7 4 One unit of difference = .66 standard error. Standard error= Square root Seedlot within species mean square average number of seedlots per species Since size measurements of samaras and leaves are highly correlated with overall size, the length and width 98 of samaras and leaves were the only size measurements in- cluded in the calculating of these differences. Fifty- five characters were used in computing the differences ex- cept for comparisons involving U. elliptica. Since this species was not included in the replicated nursery study, only 39 characters were used in computing the average dif- ference between it and other species. The species are arranged in Table 10 according to taxonomic grouping. The seven species of the section Ulmus are grouped by series. U. laevis of the section BZepharo- carpus and U. parvifolia of Microptelea are separated from the other species. Although the relationships in this table are not clear cut, there is generally more similarity between spe- cies within a group than between groups. The relation- ships agree with the taxonomic arrangement even though only one of the characters, the distance between the seed and notch of the samara, used by taxonomists to recognize these groups was included in the study. Ulmus parvifolia, the only fall blooming elm included in the study, shows the greatest difference from the other species. Variation Within Species U. minor The number of accessions of U. minor included in each part of the study is listed by geographic region in 99 Table 11. In addition to the material listed, 12 acces- sions from outside the native range of the species were studied--3 from England and 9 from Cass County, Michigan. These accessions were not included in the analysis of var- iance but they were compared with the other material to see if their original source could be determined from this data. A complete list of the accessions studied is given in Appendix E. Table 11. The number of accessions of U. minor included in each part of the study. Number of Accessions in Samara Leaf Nursery Region study study study West Central Europe 10 3 1 East Central Europe 37 24 22 Italy 27 18 16 Balkans 16 15 12 Central European U.S.S.R. 11 ll 3 Southern European U.S.S.R. _31_ _31_ 11 TOTAL 132 92 65 The percent of variation due to differences between region and the regional averages are given in Table 12 for each character that was significantly different between re- gions. The regions in Table 12 are grouped in north-south 100 m H m m s H «.mm HumcmH mmmH \msoHumnumm .flos mo .02 we 2 z z m H H .mm HumamH mmmH\HuwH3 HmmH a. H H H z 2 m «5H oOMMHSm momma mo .osm ow z 2 H m H m 44mm SOHumuumm .mmfil.nmz mm H 2 H 2 2 m «emu COHumnumm .GHEI.flmz mm 2 H H m H m ««Hm euoH3 coHumnumm mm H z 2 m z H «.mm HumcmH HOHumuumm em 2 H H s H m .«me SH» ou ucHom ammoHs mm H 2 H m z m «pH mcoHumuuom .oc Hopes Hm H H H m H m «ewe mSOHuouumm Momma .02 on 2 H H z m m «45m mcHo> mo .02 mm H 2 H m z m .«Hm HumcmH mmmH km muouomumau mmmH m H z z m 2 .MH HpmcmH ummmxapuHs comm GH 2 m z s z m .mH HumamH mumSMm\mmmmu.m.z HH m 2 H z E H aeH sumcoH MAMEMm\£umcmH comm OH H z m z 2 H .HH HumcmH museum m 2 H m m 2 H «tom sch3 poem 5 H z m 2 m 2 .HH HuwcmH Houoz H H m m m s z .oH HumcmH mmmm H a muouooumno muoeom Hmuou m0 w ammo Mmmo .mem .ousm meuH .ousm oUSMHHm> mews .ouom .ousm .ucmu .usmo conoH pom Hones: Houomumsu Hgsom .ucmo ummm. umoz comsnmm wmmmum>m HmGOHmmm .wHw>HuoQOoH 30H com EchmE .smHs "H .2 .m .kuomnmso m>HuocHume some now mommum>m HMGOHmmH can mGOHmmn cooSDmH moocmumMMHp an How couscouom Sores .S SHHHH3 mocoHum> Hmuou mo ucmoumm one .NH mHHma 101 .mHm>Huommmon mHm>oH Ho. can mo. um ucmonHcmHm mamas .. .. H H m H m H «.km uoHoo mHsu mufim Ewum GOOSHGQ HmMHuGOU mm. H H m H m m ..m. uHmHmm N. H 2 2 H H m aom mmzocoun mo .02 mm 2 H H z m 2 ..¢m am\~H\~H mm>mmH ammuo mm s S H m m H «.mm Hoon uuonmuuooem HHOH mm H H z m H H eaem mm>mmH muHmommo w mm mnmuomumnulwummnsz Hmuou mo w MmmD MmmD . vZHmm . OHHHm \meuH . OHDm GUGMflHMKV TENS .ousm .ousm .ucmu .ucmu SOHmmH cam Hones: Hmuomnmso Husom .ucmo ummm umOS cwmsumm mommnm>m Hmcowmom 1.8.ucouc «H mHHma 102 pairs and arranged in a west to east order across the table. This makes it easier to detect any geographic trends that may be present. Four of the seven samara characters that were significantly different between regions were Size mea- surements and three were ratios. The overall variation in U. minor samara characters seemed to be random. Nine of the twelve leaf characters that were significantly different between regions were size char- acters. Only two were ratios. This indicates that the U. minor leaves studied vary more in size than shape. Leaf size decreased from west to east and from north to south. Time of planting varied for seedlots. At least part of the difference in leaf size may be due to differ- ences in planting time. In the western and central regions seedlots from southern origins were planted before north- ern ones. All material from Russia was planted within two weeks. The fact that the northern origins produced larger leaves than southern origins in the western and central regions indicates that the true differences may have been underestimated. The difference in leaf size between the two Russian regions should not have been affected by plant- ing time. The decrease in leaf size from north to south seems to be true differences. However, the west to east trend is questionable. 103 The Six significant nursery characters do not show any particular patterns except that the eastern regions were consistently low. The relationship between seed sources was shown by computing the average difference between each source. These differences are shown in Table 13. The English material seemed to be more closely re- lated to the sources from the southern part of the West Central Europe region. However, both the English and Amer- ican populations showed a closer relationship to all other sources than any of the material from within the native range of the species. The seed for these two sources may have been obtained from an area where trees from several different parts of the native range had been brought to- gether and allowed to hybridize with each other. Southern England, where U. minor was first introduced before Roman times, seems to be just such an area. There is evidence that the U. minor of southern England has resulted from several very early introductions (Richens, 1955). The American material might easily have originated from south- ern England. There was no significant difference in susceptie bility to Dutch elm disease, either between regions, or locality within regions. There were, however, differences among the offspring of different individual trees. Simple correlations were computed between susceptibility to Dutch Table 13. U. localities which are defined in Appendix D. 104 The average differences among 26 geographic origins of Zaezris The underlined numbers refer to specific England 29 Eng. 30 West central Europe 1, 40 Ger. 1 F2- Ger. 1 11:1 Ger. 1 Italy A 12 lta. \l —-NN \nOOU'IO _. N N Czhtrai Euro I35 Rus. 0 Southern Eur 130 Rus. 0 T35 Rus. 0 T38 Rus. l Rus. 0 '1- L2 2 0 —-w—-—-—-—wwN N—‘NNNO—"UO-‘d O O (D NHNN_m_.m_._._. N—‘NN—IDO d p ddON‘dNNch—fl J:- _. _._-o_.._._.NN—-3_._._a —n 3 C owe—o a ‘0 \l _. ('D U. —'0 —'\l N —'O—‘—'O _._._..o_._.NN_. _._._.N—. ‘03de —-—'—‘—‘—‘NNWN . O O \l U») N—l—ld—ldeNl U) U, --23NNN 27 . —'—'—'N—" O N—‘WNN wNNwN _.a NN-‘dN—‘NN NNN _a ‘0 _. —‘U'l 0‘ WWW WNNNN—‘—‘ N N _n ._.a dNNNN www NH—‘NNN N—‘de god—go \l N O N——a._._. _.a N \J'I _.....o_. MN— —I _a_a_._.—- —l—l—l—l—a -'O—'OO _- _. —_‘_._._. N NNNWN —ld—l—l—l 105 elm disease and each of the other 74 characters studied to see if any relationships existed. Only three characters, seed length, samara width and samara tip shape, were signif- icantly correlated with susceptibility. Since you would ex- pect 3 out of 74 characters to be correlated by chance alone at the 5% level, these correlations were considered meaning- less. U. laevis The numbers of accessions of U. laevis included in each part of the study are summarized by region of origin in Table 14. Table 14. The number of accessions of U. laevis included in each part of the study. Samara Leaf Nursery Region study study study West Central Europe 21 19 19 East Central EurOpe 46 37 32 Balkans 4 2 2 Northern European U.S.S.R. 13 8 5 Central EurOpean U.S.S.R. 7 5 3 Southern EurOpean U.S.S.R. 5 4 4 Eastern European U.S.S.R. 16 15 12 TOTAL 112 90 ’ 78 106 In addition to the material listed, two accessions from England, which is outside the native range of the spe- cies, were compared to the other material to see if their original source could be determined. The number of characters in each part of the study that were significantly different between regions or source within regions are shown in Table 15. Table 15. The number of characters in each part of the U. laevis study that were significantly different between regions or source within region. Region Locality within region .05 level .01 level .05 level .01 level Samara study 4 10 8 3 Leaf study 6 6 6 1 Nursery study 3 5 2 7 TOTAL 1T E? I? I1" The percent of total variance accounted for by dif- ference between regions and the rating of the region aver ages for each character that was significantly different be- tween regions is shown in Table 16. The percent of variance accounted for by differences in region and sources within region is given for each character in Appendix E. The regions in Table 16 are grouped into North-South groups and arranged in an East-West order across the table. 107 A A m m A m A «tmN numcmA cowpmuuwm um A z m z z m 3 «.mm map ou ucflom ammoflz mm A z z m z m m «5A mcowumuumm .oc .uoa Am A 2 z m E m 2 «ma mcflw> mo Hmnfisz mm A A m m A m m «cam £u6w3 mmmA mm A A m 2 A z 2 «ma numcmA mmmA hm mumwomumco mmmA m z m A A .m m *mA mmmccmmum unnamm vm A A z m m A A «NN Hvoo comm mm A m m m m 2 m «kmm mmmnm mAaum AN m .. z A m z z z ««om souoc mo mmmccmmo mA A m A A m 2 m «tow mmmnm mmmm mA A 2 m A m A 2 «gmm madam mas 5A m A z m A m 2 «emu nuwfiz mumfimm\£uow3 comm mA A m A m 2 S A «saw numcmA mumfimm\numcmA nuboz NA m m A A m A A agqm cumcma mumfimm m m 2 E m m z A «ma nuow3 boom 5 m m S 2 m 2 A «tom nuwwz MHmEmm m 2 m A m z 2 A wave numcma couoz v m m z z m z A «5A sumcma comm N m m m A z A 2 *«mm numcma mmmm A mumuomumnu mumamm Amuou m0 m mmmD MmmD MmmD mmmD .xamm .Ondm .OHDM wOGMHHm> GEMS .ousm .onsm .onsm .onsm .ucmu .ucmu scammu paw Hogans Hmuomumnu ummm Ausom .ucmo nuuoz. ummm ummz awmzumm mommum>m AMGOAmmm .wAm>HuommmmH 30A can Esflwma .nmws HA .2 .m .Hmuomumno m>fluocflwmww zoom How mmmmum>m AMGOHmmH can macamou cmmzumn mmoamumMMHo an wow cmucsooom wwamoN .b casuw3 mocmwnm> Amuou mo unmoumm was .mA magma 108 .maw>AuommmmH mam>mA He. can mo. um unmoflmwcmflm mammfi «« .« A A m A m z z ..¢m ummuuaoo nvoo mAzu scum me A m A z z m m *«Hw usmwwn mmmuw>¢ mm A z A A m m 2 «mm chnocmun .ocoomm mm A m z z m m m «*Nm monocmnn mo .oz mm m m m m A m m ««Am Hoaoo Emum ow m 2 z m A A 2 ««Av mmmA .wusncfl ASA: Ezwsoxb cm A A z m z z m «an mmmA .musnaA ASA: szwsozu am mumuomumsu mummusz A A z m A z A ««~¢ numqu .AA\.uumm .nmsu.nmz mm 2 A A 2 A z A «mA ApmnmA AmmA\Au6A3 .uumm Am m z A z z A 2 «5A cpmcmA mmmA\.cmA .Huom om m m A m m z x ««mm Apmcwa .MA\.HHmmt.oc Amuoa 5v A A z z A m A ««m~ .Humm mo .0: Amuou\.nu0m .flmE mo .02 mw z z m m z z m «mA ApmcmA mAOAumm mm Amuou m0 w mmmo Ammo mmmo mmmo .xAmm .ousm .ounm mocmAum> mam: .ousm .ousm .ousm .onsm .ucmu .pcmu GOAmmH ocm Hones: Hmuomumnu ummm Ausom .ucwu cuuoz ummm ummz cmwzumm mwmmum>m Amcowmmm A.©.ucouv ma manna 109 No strong geographic patterns were exhibited by any of the significant characters. The pattern of variation seems to be random. The average differences between sources was computed to show the relationship between seed sources. These dif- ferences are shown in Table 17. The northern European population showed the greatest resemblance to the English material. However, there was enough difference to make it seem unlikely that they origi- nated in this area. The U. laevis population of northern France and Germany, a likely source of the trees growing in England, was not sampled in the study. Simple correlations were computed between suscepti- bility to Dutch elm disease and each of the other 74 char- acters. Two samara characters, greenness of samara and con- trast between seed color and wing color, showed a signifi- cant positive correlation to susceptibility. No biological explanation of these results was available. Leaf length, leaf width, number of veins, number of major serrations, total number of serrations and the distance from the widest point of the leaf to the tip were also pos- itively correlated to susceptibility. These characters all measure leaf size and they are all positively correlated with each other. These six correlations show that leaf size is correlated with susceptibility. 110 Table I7. The average differences among geographic origins oft/Ininozn Underlined numbers refer to Specific localities which are defined in Appendix E. England .22. Eng. 1Q West central Europe flgcer. l ’40 !-_I_l_Ger. 1+ 1711 East central Europ lQAus. 20 Cze. lI§:Pol. 111_Pol. Ll§_Pol. _l_l_5_ Pol. Balkans _6_9_Hun. giHun. LQRom. fiiYug. l§Z_Yug. Northern MRus. mRus. Central 132;RUS. ELMS. mRus. Southern Eur 1&3_Rus. h Eastern EurOp lfl;_Rus. 3 1&3_Rus. 5 -‘G O N '0 on \I —00000 0‘ NN—-—-N—- NNNwoo —-ooo- --ooo 00:) oo :1 \n 0‘ 0 4:10 U1 N N —'NN-‘N row-no NN WW U1 —‘U'1 U1 \1 e WN-‘xw—‘W-‘NN-‘d NNN WWN w—g N—n N—d J-‘w l'fi m .b-U'iNCNNc rrwww 1 O 3 3 WN—CNHCNWNN— “Nd \H-Pw w—' --w WWW :ww WN wwd I.Jx: #' p an U.S.S.R. l l l l S.S.R. l 3 3 2 l 2 3 2 N m N N W .P .b' \D a U 4? M 3 3 L. 2 l a 3 L, ea 5 L. z. e 3 n 5 3 1 O WNCD-‘OWN-‘UN—‘UWWN—O MN. 0 NW J:- W NW NW Jr'N new [0' 0 indicates no significant difference between localities. 5 indicates the greatest difference. 111 Two nursery characters, Gnomium uZmi injury and average tree height, had significantly positive correlations with susceptibility to Dutch elm disease. The average height was significantly correlated with each of the leaf size characters. Leaf size and seedling height both indicate seedling vigor. Tchernoff (1965) re- ported that vigorous seedlings are more susceptible to Dutch elm disease than slow growing ones. The correlation between susceptibility and Gnomium uZmi injury may have been due to mistakes in scoring Dutch elm disease symptoms. Seedlings that had been severely attacked by Gnomium uZmi were almost defoliated, which made it difficult to recognize Dutch elm disease. U. glabra Out of 143 seedlots of U. glabra, only 48 produced enough seedlings to be included in the leaf study and only 12 in the nursery study. The numbers of accessions included in each phase of the study. Three of the leaf characters; widest point to the tip, serration depth and ratio of the number of veins to leaf length, were significantly different (.05 level) between regions. The English, East Central Europe and Balkan had the highest values in those three traits West Central European and Italian trees were the lowest values. 112 Because of the small number of accessions in the nursery part of the study no statistical analysis was at- tempted. Six of the twelve accessions studied were from Judenberg, Austria. The amount of variation in these six accessions was almost as great as the amount of variation in all 12 accessions. U. pumila Suwon, Korea and Alma Ata in central Asiatic U.S.S.R. are the only localities within the native range of U. pumila represented in the M.S.U. collection. Because of this, the U. pumila accessions were divided into two groups, an eastern group represented by four seedlots each from Japan and Korea and a western group consisting of 18 seedlots from European and western Asiatic U.S.S.R. The 11 seedlots from Italy showed evidence of hy- bridization with both U. minor and U. glabra and were not included in the analysis. Five seedlots from the United States were also excluded from the analysis. Only two samara characters, ratio of widest point to base to samara length and persistant calyxes, were sig- nificantly different between regions. The highest values for both of these traits was from the western region. The number of veins, total number of serrations and distance from widest point to leaf were all signifi- cantly greater in the western region. ll3 Accessions from the western region showed signifi- cantly less injury from autumn frost but were slower grow- ing than the eastern material. U. japonica All but 5 of the 49 collections of U. japonica seed came from the island of Hakkaido, Japan. These other five seedlots were from central Honshu island. Only 11 seedlots, all from two localities on Hokkaido, produced enough seedlings to be included in the leaf and nursery parts of the study. None of the samara characters showed any signifi- cant differences between islands. In fact, there was more variability within the Hokkaido material than between Honshu and Hokkaido. The two Hokkaido locations tested in the leaf and nursery parts of the study did not differ significantly in any character. U. parvifolia Eighteen seedlots of U. parvifolia, the only fall blooming elm studied, are included in the M.S.U. collection. One seedlot, MSFG 6323, was collected in the early spring of 1968 from seed persisting on the tree. The other 17 ac- cessions were not received until the fall of 1968 or the winter of 1968-1969. The seedlot received in the spring of 114 1968 was the only one included in the leaf and nursery parts of the study. Therefore, the only within species comparisons that could be made are on samara character. Nine of the eighteen seedlots were from the island of Honshu, Japan, one from the island of Kyushu, Japan, two from Suwon, Korea and six were collected from trees growing on the M.S.U. campus. The Korean accessions had narrower seeds and a lower seed width/samara width ratio than the other accessions. These were the only characters that showed a significant difference (.05 level) between sources. U. laciniata Eight of the ten seedlots of U. laciniata in the study came from Japan. Six of these seedlots were from Hokkaido, one from northern Honshu and one from central Honshu. The other two seedlots came from England and the Netherlands. They were the only seedlots that produced any seedlings. No analysis of the variation within species was at- tempted, but a comparison was made of the samara data from the five different sources. The Japanese sources resembled each other more than they did either of the other two sources, but there was more similarity between the samaras from England and the Netherlands than between any of the Japanese sources . 115 U. elliptica Four of the seven seedlots of U. elliptica were from Leningrad, which is outside the native range of the species. The other three seedlots were from the Caucasus Mountain region of the U.S.S.R.--two from Yerevan and one from Tbilisi. Two of the seedlots from Leningrad produced enough seedlings to be included in the leaf study, but neither produced enough for the nursery study. No analysis of the within species variation was at- tempted because of the small amount of material. However, a comparison of samara characters was made between the Leningrad material and the other two sources. The same similarity was shown between the Leningrad and Tbilisi, and Tbilisi and Yerevan. A comparison of Leningrad and Yerevan material showed almost twice as much difference as the other two comparisons. U. wallichiana Six of the seven accessions of U. wallichiana were obtained from the Netherlands elm breeding program. The wings had been removed from seeds from Pakistan, the only seedlot from within the native range of the species. None of the Pakistan seed germinated, so there was no material available for making comparisons within species. 116 Summary of Within-Species Variability Five species--U. minor, U. laevis, U. glabra, U. pumila and U. japonica--were represented by enough material to make studies of within-species variation of samaras pos- sible. Two of these species--U. laevis, and U. minor showed significant differences among regions in samara size char- acters. U. laevis was the only species that showed signifi- cant differences in samara shape and color characters. Within species variation of leaf characters was stud- ied in four species--U. glabra, U. laevis, U. minor and U. pumila. Significant differences between regions were shown in nine U. minor, two U. glabra, seven U. laevis and two U. pumila leaf characters. Petiole length and pubscence of upper leaf surface was significantly different between re- gions in U. laevis and U. minor respectively. Only three species contained enough material for a study of within-species variation of nursery characters. None of the characters showed significant between-region dif- ferences in all three species. The number of branches per seedling and the number of seedlings with secondary branching were significantly different between regions in U. laevis and U. pumila. Average height was significantly different be- tween regions in U. minor and U. pumila. Six characters did not show significant between region differences in any of the species. Susceptibility to Dutch elm disease was one of these characters. 117 The only character that showed any recognizable pattern of geographic variation in any of these species was leaf size in U. minor. The lack of geographic varia- tion within species is surprising. There had to have been enough isolation and natural selection at one time to al- low speciation to occur. Part of the uniformity within species may have been due to man's activities in planting. However, it seems un- likely that there would have been enough movement of mater- ial between regions within the native range of the species to have affected the overall genetic pattern. It seems likely that there was not a great amount of geographic var- iability within species before man started moving trees around. CHAPTER VIII THE POSSIBILITIES OF BREEDING ELMS RESISTANT TO DUTCH ELM DISEASE Ulmus americana has proved to be the most suscep- tible to Dutch elm disease in every test that has compared it with other species. Smalley and Kais (1966) found that seedlings from localities where the disease had been pre- sent for many years contained a higher level of resistance than seedlings from areas recently invaded by the disease. This indicates that some natural selection for disease re- sistance has taken place. The importance of this factor in a tree breeding program is obvious. However, it is doubtful if the disease has been present in Michigan long enough for natural selection to have had a significant ef- fect on the American elm population. The American elm is also susceptible to phloem necrosis, a virus disease. This disease is reported to have killed more American elms in some parts of the species range than Dutch elm disease. Phloem necrosis has not been reported in Michigan, but it is present in Ohio and Illinois and it would be unwise for an elm breeder to ignore the disease. Incorporating resistance to this disease into 118 119 varieties resistant to Dutch elm disease would take at least one additional generation of breeding and testing, which would add 15 to 20 years to a breeding program. Ulmus americana and the closely related species U. laevis are the only species of elm susceptible to phloem necrosis. Both of them are also susceptible to Gnomium uZmi. The degree of susceptibility of U. americana to G. uZmi has not been reported. However, the high degree of susceptibility of U. laevis observed in this study would limit its value in future breeding work. These factors indicate that it would be much easier to develop desir- able cultivars of elm from other species of Ulmus than from either U. americana or U. laevis. Ulmus pumila is highly resistant to Dutch elm disease but has other undesirable characteristics. To utilize this resistance, an extensive interspecific hy- bridization program would be necessary to combine it with desirable characteristics of other species. Ulmus wallichiana possesses many desirable char- acteristics in addition to Dutch elm resistance. However, it probably does not contain sufficient winter hardiness to be grown in Michigan. Ulmus parvifolia is highly resistant to Dutch elm disease but the species develops into a small tree which would not be acceptable as a substitute for American elm. 120 U. laciniata and U. glabra both possess the desir- able characteristics of leaf size, growth rate and form. There is also evidence that selections containing an accept- able level of resistance to Dutch elm disease could be ob- tained from these species. However, unless the poor germ- ination, exhibited by material in this study, can be over- come, future work in either of these species will be limited. Based on the information obtained in this study, U. minor and U. japonica have the most potential value for fu- ture breeding work. Seed from either of these species can be readily obtained from within their native ranges. The seed of U. minor is more easily germinated than U. japonica, but the latter species shows a higher degree of resistance to Dutch elm disease. Selection in either of these species would need to be on a family basis. An eight acre planting of U. minor at the Fred Russ Experimental Forest in southwestern Michigan is a potentially valuable source of material. This planting, which is owned by Michigan State University, is about 25 years old. Few of the trees in the planting have been lost to Dutch elm disease even though mortality of American elm in the vicinity is nearly complete. Progeny of nine trees from the planting in- cluded in this study ranged from 40 to 70% susceptible to Dutch elm disease. However, more resistant families may be discovered with further screening work. 121 If parent trees that produce a high percentage of resistant progeny could be located in this planting, it could be easily developed into a seed orchard. A source of resistant material could be available within a few years. Open pollinated seed from the best Dutch clones showed a low level of susceptibility in this study. The degree of susceptibility exhibited by the seedlings was closely correlated to the rating received by the parent clone in the tests in the Netherlands. Further screening work with this material would probably be rewarding. The native ranges of potentially valuable Chinese species, not included in the present study, are isolated either politically or geographically. It would be very difficult to obtain seed from them at the present time. Every effort should be made to obtain this material if the political situation becomes more favorable in the fu- ture. All of the material in the Michigan State Univer- sity program has been reinoculated in the 1971 growing season. The data obtained in this study and the results of the 1971 inoculation will be used to select the most promising families. This material will be planted in a planting designed to insure a maximum amount of crossing between families. When these trees reach flowering age, progeny tests will be used to evaluate possible parents. If these results 122 are encouraging, the planting may be developed into a seed orchard. Additional seed of some of the most promising material will be requested from the original collector. An attempt will also be made to obtain additional collec- tions of seed from species not well represented in the pre- sent study. LIST OF REFERENCES LIST OF REFERENCES Literature Cited Albensky, A. V. 1951. Results of hybridization of larch maple, elm and ash trees. Trud. Inst. Lesa 8:88-94, 205. P.B.A. 25:558. (Russian original not seen). Albensky, A.V. 1969. Selection and seed production for protection plantings. Second World Consultation on Tree Breeding, Washington D.C., pp. 1041-1048. Anon. 1947. The use and misuse of shrubs and trees as fodder. Imp. Agric. Bur. Joint Publ. 10. Anon. 1954. Forest tree breeding in Canada. J. of Forest. 52:682-685. Anon. 1962. Forest tree improvement progress report. State University College of Forestry, Syracuse, New York. Arisumi and D. J. Higgins. 1961. Effect of Dutch elm disease on seedling elm. Phytopath. 51(12): 847- 850. Banfield, M. W. 1968. Dutch elm disease recurrence and recovery in American elm. Phytopath. 62: 21-60. Banfield, M. W.; Rex, E. G.; and May, C. 1947. Recur- rence of Dutch elm disease in American elms in relation to tree structure. Phytopath. 37: 1-2. Bean, W. J. 1951. Trees and shrubs hardy in the British Isles. 7th ed. Oliver and Boyd Ltd., London ' Bechtel, A. R. 1921. The floral anatomy of the Urticales. Am. Jour. Bot. 8: 386-410. Berry, E. W. 1923. Tree ancestors. Williams and Wilkins Co., Baltimore. 270 p. Britwum, S. P. K. 1961. Artificial hybridization in the genus Ulmus. 8th Northeastern Forest Tree Impr. Conf. Proc., pp. 43-47. 123 124 Buisman, Christine J. 1936. Results of inoculation with the elm disease at the Willie Commelin Scholten in Baarn during 1935. Tijds. Plziekt. 42:21-44. (Phytopath. Transl. 30:1-38). Burger, F. W. 1938. Iepensterfte in Nederland. Tidjs. Plziekt. 44:177-207. (Reviewed in Heybroek 1967). Castellani, Ettore. 1966. Summary of research concerning disease resistance in southern EurOpe. In Gerhold et. a1. Breeding pest resistant trees. Pergamon Press, Oxford, pp. 11-16. Cheng, W. C. 1963. Species novae et nomines emendata arborum utilium Chinae. Scientia Silvae 8(1):1-14. Chun, Woon Young. 1921. Chinese economic trees. Commer- cial Press, Ltd., Shanghi. 309 p. Clapper, R. B. 1952. Breeding and establishing new trees resistant to disease. Econ. Bot. 6:271-293. Collins, P. E. 1955. Chinkota elm. South Dakota Farm and Home Research 7:14-16, 27. Collins, P. E. 1967. Hybridization studies in the genus Ulmus. Unpublished Ph.D. Dis., University of Minn. 118 p. Derman, H., and May, C. 1966. Colchiploidy of Ulmus pumila and its possible hybridization with U. americana. Forest Sci. 12:140-146. Dimond, A. E. 1947. Symptoms of Dutch elm disease repro- duced by toxins of Graphium uZmi in culture. Phytopath. 37:7. Egolf, Donald R. 1968. Current developments in the breed- ing of woody ornamentals. Hort. Sci. 3:4:1068-1072. Ehrenberg, C. E. 1949. Studies on asynapsis in the elm Ulmus glabra Huds. Hereditas 35:1-26. Ehrenberg, C. E. 1954. Elm disease history and measures for the production of elms resistant to the disease. Svenska Skogsvforoen. Tidskr. 52:35-41. P. B. A. 14:3384. (Swedish original not seen). Elgersma, D. M. 1967. Factors determining resistance of elms to Ceratocystis uZmi. Phytopath. 57:641-642. 125 Elgersma, D. M. 1969. Resistance mechanisms of elms to Ceratocystis ulmi. Medel. Phytopathology lab. Willie Commelin Scholten. No. 77, Wageningen, Netherlands. Elias, Thomas S. 1970. The genera of UZmaceae in the southeastern United States. J. of the Arnold Arboretum 51:18-40. Engstrom, Harold E., and Matthew, Lewis S. 1942. The effect of the 1940 Armistice Day freeze on Siberian elm in the Plains country. J. of Forest. 40:704- 707. Fernald, M. L. 1945. The indigenous variations of Ulmus americana. Rhodora 47:132-133. Fowells, A. A. 1965. Silvics of forest trees of the United States. U.S. Dept. Agr. Handbook 271. Wash- ington D.C. Fransen, J. J. 1939. Elm disease, elm bark beetles and their control. Dissertatie, Wageningen. Mededeling Comite Bestrijding lepeziekte, Nr. 32. 118 p. (Reviewed in Heybroek, 1967). George, E. J. 1937. Storage and dewinging of American elm seed. J. of Forest. 35:769-772. Goidanich, G. 1938. Notes on research in selection of elms resistant to elm disease. Ital. Agr. 75:69-74. P. B. A. 8:1347. (Italian original not seen). Goidanich, G. and Azzaroli, F. 1938. Rept. on experi- ments of selection of elms resistant to elm disease and of artificial inoculation of G. uZmi made in 1937. Boll. Staz. Pat. Veg. Roma 18:149-178. P.B.A. 9:864. (Italian original not seen). Graves, A. H. 1948. Forest tree breeding. Econ. Bot. 2:284—305. Green, P. S. 1964. Registration of cultivar names in Ulmus. Arnoldia 24:6-8, 41-80. Greguss, Ladislav. 1970. Temperature response of pollen mother cells in Ulmus. I.U.F.R.O., Section 22. Working group meeting on the sexual reproduction of forest trees. Varparanta, Finland, April 28- May 6. 126 Grudzinskja, I. A. 1956. Notes on Ulmus pumila. Bot. Mat. Gerbar. Bot. Inst. Komarov Akad Nauk S.S.S.R. 21:113- 124. Grudzinskja, I. A. 1966. Inflorescences of the species of Ulmus L. their formation, structure and certain prob- lems of their evolution. Bot. Zhur. Moskova 51:15-27. (In Russian, English summary). Grudzinskja, I. A., and Zakharyeva, O. I. 1967. On the diagnostic significance of cytological characters in the taxonomy of certain tree species (taking the genus Ulmus as an example). Bot. Zhur., Moskova 52:641-650. (In Russian, English summary). Harlow, W. M., and Harrar, E. S. 1958. Textbook of den- drology. 4th ed. McGraw-Hill Book Co., New York, New York. Heimburger, C. 1966. Genetic improvement for disease and insect resistance of forest trees in Canada. In Gerhold, et. al. Breeding pest resistant trees. Pergamon Press, Oxford, pp. 49-52. Henry, A. 1910. On elm-seedlings showing Mendelian re- sults. Linn. Soc. Bot. Jour. 39:290-300. Hess, Lloyd W., and Dunn, David B. 1967. Evidence of ecological isolations between Ulmus thomasii Sarg. and Ulmus rubra Muhl. Transactions of Missouri Academy of Science 1:31-36. Heybroek, H. M. 1957. Elm breeding in the Netherlands. Silvae Genetica. Vol. 6, No. 3 & 4, 1957. Heybroek, H. M. 1963b. Diseases and lopping for fodder as possible causes of prehistoric decline of Ulmus. Acta Bot. Neerl. 12(1):l-ll. Heybroek, H. M. 1966b. Aims and criteria in elm breeding in the Netherlands. In Gerhold, et. a1. Breeding pest resistant trees. Pergamon Press, Oxford, pp. 387-389. Heybroek, H. M. 1968. Taxonomy crossability and breeding of elms. In Proc. Int. Symposium on Dutch elm disease. H. S. McNabb, [ed.] Iowa State Univ. Press, (in press). Heybroek, H. M. 1969. The future of elms in North America. Proc. Midwestern Chapt.-Internat. Shade Tree Conf. 24:43-49. 127 Jeffers, J. N. R., and Richens., R. H. 1970. Multivariate analysis of the English elm population. Silvae Genetica 19:31—37. Jobling, J., and Low, A. J. 1969. Elms. In Forestry commission report on forest research for the year ended 1969, pp. 56-67. Kais, A. G.; Smalley, E. B.; and Riker, A. J. 1962. En- vironment and development of Dutch elm disease. Phytopath. 52:1191-1196. Kerling, L. C. P. 1955. Reactions of elm wood to attacks of Ophiostoma uZmi (Buism.) Nannf. Acta. Bot. Nerrl. 4:398-403. Komorov, V. L. (ed.) 1936. Flora of the U.S.S.R. Akademiya Nauk S.S.S.R. Vol. 5:284-294. Translated 1970 by Israel Program for Scientific Translations. Krijthe, Neetje. 1939. Report on the activities for the elm disease committee carried out at the genetic laboratory in 1938. Tijds. Plziekt. 45:63-70. Kurz, H., and Godfrey, R. K. 1962. Trees of northern Florida. Gainsville, Florida, 331 p. Leliveld, J. Adolph. 1934. Cytological studies in the genus Ulmus L. Genetica 15:425-432. Lester, D. T. 1968. Genetic approaches to increased orna— mental value of elms. Univ. of Wise. Forest. Res. Note No. 139. Lester, D. T. 1969. Research in forest genetics and tree breeding at the University of Wisconsin. Ninth Lake States Forest Tree Impr. Conf. Proc., pp. 12-16. Manka, K. 1953. The progress of Dutch elm disease (Ceratostmella uZmi (Schw.) Buisman) in the area of Poznan, Poland. Acta. Soc. Bot. Pol. 22(2):355-378. Maxon, Marcus A. 1951. Siberian elm. South Dakota Farm and Home Res. 2:28-30. McDaniel, S., and Swift, C. 1967. Ulmus crassifolia (UZmaoeae) in Florida. Sida. 3:115-116. McKenzie, M. A. 1970. Pathology of wilt diseases. Univ. of Mass. Shade Tree Labs. Xerox Statement. 128 McNabb, Harold 5., Jr.; Heybroek, Hans; and MacDonald, William L. 1971. Anatomical factors in resistance to Dutch elm disease. Phytopath. Zeits. (In press). Melville, R. 1938. Is Ulmus campestris L. a nomen ambigum? J. Bot., pp. 261-265. Melville, R. 1946. Typification and variation of the smooth leaved elm, Ulmus carpinifolia Gleditsch. J. Linn. Soc. Bot. 53:83-90. Melville, R. 1957. Ulmus oanesoens: an eastern Mediter- ranean elm. Kew Bull. 12:499-502. Muller,, C. H. 1936. New and noteworthy trees in Texas and Mexico. Bull. Torr. Bot. Club 63:147-155. Muller, C. H. 1937. Studies in Mexican and Central American plants. Amer. Midl. Nat. 18:842-855. Muller, C. H. 1942. Notes on the American flora, chiefly Mexican. Amer. Midl. Nat. 27:470-490. Otis, C. H. 1931. Michigan trees. Univ. of Mich. Press, Ann Arbor, Michigan. Ouellett, C., and Pomerleau, R. 1965. Recherches sur la resistance de 1'orme d'Amerique au Ceratocystis uZmi. Can. J. Bot. 43:85-97. Ozolin, G. P. 1958. Breeding of UZmaoeae for resistance to the Dutch disease. Uzbek Academy of Agricultural Science. Bull. 4. Translated by C. C. Nikiforoff. Ozolin, G. P. 1959. Experience in selecting elm species resistant to Dutch elm disease. Vestnik Selskoh. Nauk, Moskva 12:139-142. Phytopath. Trans. No. 65. 1963. Pauley, S. S. 1949. Forest-tree genetics research, Econ. Bot. 3:299-330. Peace, T. R. 1960. The status and development of elm disease in Britain. Bull. For. Comm. Lond. No. 33. Pomerleau, R., and Bard, J. 1968. Hardiness and resis- tance to Ceratocystis uZmi (Buisman) C. Moreau of hybrids and clones of European and American elms. Can. Dept. For. and Rural Dev. Bimonthly Res. Note. 24:26. 129 Pomerleau, R. 1965. The period of susceptibility of Ulmus americana to Certocystis uZmi under conditions pre- vailing in Quebec, Canada. Jour. Bot. 43:787-792. Pope, S. A. 1943. Some studies on the Dutch elm disease and the causal organism. Unpublished Ph.d. thesis, Cornell University Library, Ithaca, New York. (Reviewed in McNabb, et. al., 1971). Rehder, A. 1940. Manual of cultivated trees and shrubs. 2nd ed. Macmillan Co., New York. Rehder, A. 1949. Bibliography of cultivated trees and shrubs hardy in cooler temporate regions of the world. The Arnold Arboretum of Harvard Univ., Jamaica Plains, Mass. Richens, R. H. 1955. Studies in Ulmus: I. The range and variation of east anglian elms. Watsonia 3:138- 153. Richens, R. H. 1956. Elms. New Biology 20:7-29. Richens, R. H. 1958. Studies in Ulmus: II. The village elms of southern Cambridgeshire. For. 31:132-146. Richens, R. H. 1959. Studies in Ulmus: III. The village elms of Hentfordshire. For. 32:138-154. Richens, R. H. 1961a. Studies in Ulmus: IV. The village elms of Huntingdonshire and a new method for explor- ing taxonomic discontinuity. For. 34:47-64. Richens, R. H. 1961b. Studies of Ulmus: V. The village elms of Bedfordshire. For. 34:181-200. Richens, R. H. 1963. Monophage analysis of elm popula- tions. World Consultation on Forest Genetics and Tree Breeding. Stockholm, Sweden. Sect. 3/16. Richens, R. H. 1965. Studies on Ulmus: VI. Fenland elms. For. 38:225. Richens, R. H. 1967. Studies of Ulmus: VII. Essex elm. For. 40(2):185-206. Rovskii, V. M.; Ozolin, G. P.; and Solovjiva, A. L. 1950. Breeding elms for resistance to Dutch elm disease. Lesnoc Hoz. 4:43-45. P.B.A. 22:3017. (Russian original not seen). 130 Sax, K. 1933. Chromosome numbers in Ulmus and related genera. J. Arnold Arboretum 14:82-84. Schneider, C. K. 1917. Ulmaceae. In R. Sargent, Plantae Wilsoniana 3:238-265. . Schreiber, L. R. 1968. Developing resistance in American elm. In Proc. Int. Symp. on Dutch elm disease. H. S. McNabb, (ed.) Iowa State Univ. Press, (In press). Shattuck, C. H. 1905. A morphological study of Ulmus americana. Bot. Gaz. 40:209-223. Sinclair, W. A. 1970. Pathology of wilt diseases of trees. C.S.R.S. Progress Report. Sinclair, W. A.: Welch, D. S.; Parker, K. G.; and Tyler, L. G. 1968. Selection within Ulmus americana for resistance to Ceratocystis uZmi. In Proc. Int. Symp. on Dutch elm disease. H. S. McNabb, (ed.) Iowa State Univ. Press, (In press). Smalley, E. B. 1963. Seasonal fluctuations in suscepti- bility of young elm seedlings to Dutch elm disease. Phytopath. 53:(7):846-853. Smalley, E. B., and Kais, A. G. 1966. Seasonal variations in the resistance of various elm species to Dutch elm disease. In Gerhold, et. al., Breeding pest resistant trees. Pergamon Press, Oxford, pp. 279-287. Smalley, E. B., and Riker, A. S. 1962. Tropical members of the Ulmaceae resistant to Dutch elm disease. Wisc. Forest. Res. Notes 77, 4p. Smith, E. Chalmers, and Nichols, Charles, Jr. 1941. Spe- cies hybrids in forest trees. J. of the Arnold Arboretum XXII:443-454. Smucker, G. J. 1944. Rebuilding the American elm. Amer. Forests 50:104-107, 137-138. Smucker, S. J. 1940. Apparent recovery of American elms inoculated with CeratostemeZZa uZmi. Phytopath. 30:1052-1054. Sokolov, S. J. (ed.) 1958. Ulmus in Darevja Kustarniki S.S.S.R. Moskova, Leningrad Akad. Nauk S.S.S.R. 2:494-510. 131 Standley, P. C. 1922. Trees and Shrubs of Mexico. Contr. U.S. Nat. Herb. 23:198. Standley, P. C. 1936. Studies of American plants VII. Chicago Field Museum of Botany 17:155-284. Standley, P. C., and Steyermark, Julian A. 1946. Flora of Guatemals. Fieldiana Botany. Vol. 24, Part IV. Chicago Nat. Hist. Museum. Staub, R. 1967. Some variation in the fruit seed and seedling characteristics in the American elm, Ulmus americana L. in relation to geography of seed source. Unpublished Ph.d. thesis. Univ. of Minnesota. Swingle, R. V. 1945. Another enemy flanks the elm. Am. For. 51:334-335. Tchernoff, V. 1963. Vegetative propagation of elm by means of shoots cut from callused roots. ACTA Bot. Neerl. 12:40-50. Tchernoff, V. 1965. Methods for screening and for the rapid selection of elms for resistance to Dutch elm disease. ACTA Bot. Neerl. 14:409-452. Tippo, O. 1938. Comparative anatomy of the Moraceae and their presumed allies. Bot. Gaz. 100:1-99. Touw, A., and Van Steenis, C.G.G.J. 1968. Note on Ulmus in Malaysia (Ulmaceae). Blumea 16:84. Troels-Smith, J. 1960. Ivy, mistletoe and elm, climatic indicators-fodder plants. Danmarks Geol. Unders. IV. Raekke, Bd. 4. 429p. (Reviewed in Heybroek 1963b). Tutin, T. G. 1964. Ulmus. In: Tutin, et. al., Flora EurOpaea, Cambridge Univ. Press. 1:65 Tyler, L. J. 1945. The influence of temperature on the Dutch elm disease in potted American elm seedlings. Phytopath. 35:302-304. Tyler, L. J.; Parker, K. G.; and Pope, S. 1940. Relation of wounds to infection of American elm by Ceratostomella uZmi, and occurrence of spores in rain water. Phyto- path. 30:29-41. 132 Vazda, Z. 1952. Causes of mass deterioration of elms. Glasn. Sumske Pojuse. 10:105-197. F.A. 14:309. (Polish original not seen). Walter, J. W.; May, Curtis; and Collins, C. W. 1943. Dutch elm disease and its control. U.S.D.A. Circular No. 677. Webb, E. Walter. 1948. A report of Ulmus pumila in the Great Plains region of the United States. J. of Forest. 46(4):274-278. Weikoff, A. D. 1941. What can the Manchurian flora as well as the flora of neighboring countries give to the gardens of Manchuria itself and other countries with cold climates. Harbin, Manchuria. Welch, D. S. 1949. Research on disease resistance in American elm in the Cornell plantation. The Cornell Plantation 15:6-8. Went, J. C. 1938. Compilation of the Investigations on the susceptibility of different elms to CeratostemeZZa uZmi (Buisman) in the Netherlands. Phytopath. Ztschr. 9:181-201. Went, J. C. 1954. The Dutch elm disease: summary of fifteen years hybridization and selection work (1937-1952). Fijds. Plziekt. 60:109-127. Westenburg, J. 1932. On the history of the Graphium disease of elm trees. Tijds. Plziekt. 38:61-66. Winieski, J. A. 1960. Artificial hybridization and grafting methods with Ulmus americana. Northeast. For. Tree Improv. Conf. Proc. 7:48-51. Wright, J. W. 1949. Producing elm seeds on cut branches. J. Forest. 47:210-214. wyman, D. 1965. Trees for American gardens. The Mac- millan Co., New York. 502p. Zentmeyer, G. A. 1942. Toxin formation and chemotherapy in relation to Dutch elm disease. (Abstr.) Phyto- path. 32:20. 133 Other Sources Anon. 1948. Woody-plants seed manual. U.S. Dept. Agr.- U.S. Forest Service. Misc. Pub. No. 654. Bancroft, H. 1935. The elm problem. Quart. J. of Forest. 29:102-106. Barton, L. 1939. Storage of elm seeds. Contrib. Boyce Thomp. Inst. 10:221-233. Bates, C. G. 1942. Comment. J. of Forest. 40:704-707. Bogdonov, P. L. 1931. On the photoperiodism of tree spe- cies. Trudy po lesnomu opytnomy delu. 10:54-55. State Research Institute for Forestry and Forest Industry. Leningrad, U.S. Forest Service Transla- tion No. 137. 1955. Bor, N. L. 1953. Manual of Indiana Forest Botany. Ox- ford University Press, London, pp. 480-481. Boyce, J. S. 1941. Exotic trees and diseases. Jour. of For. 39:907-913. Brandis, Dietrich. 1911. Indian trees. Constable and Company, London. Burton, S. S. 1929. Chinese elm (Ulmus pumila). Journal of Forestry 27:303-305. Caroselli, Nestor E., and Feldman, A. W. 1951. Dutch elm disease in young elm seedlings. Phytopath. 41:46-51. Carter, J. C. 1946. Inheritance of foliage variegation in variegated English elm (Ulmus procera argenteo- variegata). Trans. Illinois Acad. Sci. 39:43-46. Chevalier, A. 1942. Les ormes de France. Revue de Bot. Appl. 22:429-459. Clinton, G. P., and McCormick, F. A. 1936. Dutch elm disease: Graphium uZmi. Conn. Agr. Exp. Stat. Bull. 389. Clapham, Tutin, and Warburg. 1962. Flora of the British Isles. 2nd ed., Cambridge Univ. Press. ‘ Darlington, C. D., and Wylie, F. D. 1955. Chromosome atlas of flowering plants. Allen and Unwin Ltd., London. 134 Dimond, A. E.; Plumb, G. H.; Stoddard, E. M.; and Horsfall, J. G. 1949. An evaluation of chemotherapy and vec- tor control by insecticides for combating Dutch elm disease. Conn. Agr. Exp. Stat. Bull. 531. Doorenbos, S. G. A. 1938. Kruising sproven met lepen te is Gravenhage (Crossing experiments with elms at the Hague). Tijds. Plziekt. 44:161-164. East, E. M. 1940. The distribution of self sterility in the flowering plants. Amer. Phil. Soc. 82:449- 518. Endtmann, J. 1967. Taxonomy of central European species groups of Ulmus. Arch. Forstw. 16(6/9)667-672. Feldman, A. W.; Caroselli, NestorfE.: and Howard, F. L. 1950. Physiology of toxin production by Ceratostomella uZmi. Phytopath. 40:341-354. Gerhold, H. D.: Schreiner, E. J.; McDermott, R. E.: and Wineski, J. A. 1966. Breeding pest-resistant trees. Pergamon Press, Oxford, England. 502p. Grudzinskja, I. A. 1956. The importance of Ulmaceae to afforestation in the Step. Trud. Inst. Lesa. P.B.A. 27:3477. (Russian original not seen). Hara, Hiroshi. 1966. The flora of eastern Himalaya. The University of Tokyo Press, Tokyo, Japan. Harrar, E. S., and Harrar, J. G. 1946. Guide to southern trees. Whitlesey House, New York. Henry, A. 1904. The artificial production of vigorous trees. Jour. Dept. Agr. Tech. Instr. Ireland 15:1. Heybroek, H. M. 1961. De 1ep Commelin. Ned. Bosb. Tijdschr 33 (11):325-328. Heybroek, H. M. 1963 a. De 1ep Groenveld. Ned. Bosb. Tijdschr 3S(9):370-374. Heybroek, H. M. 1967. The Dutch elm disease in the Old World. XIV IUFRO-KONGRESS, Section 24, Munchen. Holmes, F. W. 1965. Virulence in Ceratocystis uZmi. Neth. J. Plant Path. 71:97-112. Hooker, J. D. 1885. The flora of British India. L. Reeve Co., Ltd., Ashford, Kent England. 3 vol. 135 Johnson, L. P. V. 1939. A descriptive list of natural and artificial interspecific hybrids in North America forest tree genera. Canad. J. Res. 17 (Sect. C) 411-444. Johnson, L. P. V. 1946. Fertilization in Ulmus with spe- cial reference to hybridization procedure. Canad. J. of Res. 24:1-3. Johnson, L. P. V., and Heimburger, C. 1946. Preliminary report on interspecific hybridization in forest trees. Canad. J. of Res. 24 (Sect. C) 308-312. Kurz, S. 1877. The forest flora of British Burma. Govern- ment of India, Calcutta. Leach, J. G. 1953. Bacteria, fungi and insects. In Plant diseases. U.S.D.A. Yearbook, 1953. Lester, D. T., and Smalley, E. B. 1968. Prospects for elm breeding in Wisconsin. Sixth Central States Forest Tree Impr. Conf. Proc., pp. 37-41. Li, Hu-Lin. 1963. Woody flora of Taiwan. Livingston Publishing Co., Narbeth, Pennsylvania. May, C. 1935. The Dutch elm disease from a research stand- point. National Shade Tree Conf. Proc. 11:122-127. Melville, R. 1937. On elm seedlings. Quart. J. Forest. Oct., pp. 1-9. Melville, R. 1939. The application of biometrical methods to the study of elms. The Linnean Society of London, Proc. pp. 152-159. Melville, R. 1944. The British elm flora. Nature 153: 198-199. Merrill, Elmer Drew. 1934. An enumeration of plants col- lected in Sumatra by W. N. and C. M. Bangham. Con- tributions from the Arnold Arboretum of Harvard Univ. Jamaica Plains, Mass. Miranda, F. 1952. La vegetacion de chiapas tuxtla. Gutierrez, Mexico. 410 p. Nakai, Takenoshin. 1932. Flora sylvatica Koreana, pars XIX UZmaoeae et Moraceae. The Forest Experiment Sta- tion. Government General of Chosen, Keijyo, Japan. 136 Neely, P., and Carter, J. C. 1965. Species of elm on the University of Illinois campus resistant to Dutch elm disease. Plant Dis. Reptr. 49:552. Ohwi, Jisaburo. 1965. Flora of Japan. National Science Museum, Tokyo, Japan. Pearson, P. S., and Brown, H. P. 1932. The commercial timbers of India. Govt. of India Central Pub. Branch, Vol. 2. Perring, F. H. and Walters, S. M. 1963. Atlas of the British flora. Thomas Nelson and Sons, London. Planchon, J. E. 1873. Ulmaceae. In de Candolles "Prodromus" 17:151-210. Pomerleau, R. 1956. Summary report to committee on forest tree breeding. 5th meeting of the Committee on For- est Tree Breeding in Canada Proc. Pridham, A. M. S. 1965. Propogation of American elm from cuttings. Int. Plant Prop. Soc. Proc. 14:86-88. Rehder, A. 1923. The ligneous plants of Northern China. Jour. of the Arnold Arboretum 4:163-168. Rehder, A. 1924. Enumeration of the ligneous plants of Northern China II. Jour. of the Arnold Arboretum 5:137-224. Richens, R. H. 1945. Forest tree breeding and genetics. Imp. Bur. Plant Breed. and Genetics Jt. Pub. 8. Rovskii, V. M. 1951. Breeding forest trees in Uzbekistan. Trad. Inst. Lesa. l. P.B.A. 25:563. (Russian orig- inal not seen). Rovski, V. M. 1956. On the radical method of fighting Dutch elm disease. Bot. Zurn. 41:1478-14781. (In Russian, English summary). Rudolph, Paul O., and Patton, Robert F. 1966. Genetic improvement of forest trees for disease and insect resistance in the lake states. In Gerhold, et. a1. Breeding pest resistant trees. Pergamon Press, Ox- ford. Santamour, Frank S., and French. David W. 1958. Toxin in relation to resistance to Dutch elm disease. Minn. Forest. Notes. No. 65. 137 Santamour, Frank S. 1970. A natural hybrid between Ameri- can and Siberian elms. Forest Sci. 16:149-153. Schreiber, Annelis. 1957. Ulmaceae. In Hegi, Illustrierte flora von mitteleuropa 3:245-263. Seymour, F. D. 1952. Notes on Moraceae and Ulmaceae. Am. Midland Nat. 18:149-150. Singh, D., and Smalley, E. B. 1966. Nitrogenous compounds in the xylary sap of UZmaoeae species varying in re- sistance to Dutch elm disease. Phytopath. 56:901. Smucker, S. J. 1941. Comparison of susceptibility of American elm and several exotic elms to CeratostomeZZa uZmi. Phytopath. 31:758-759. Sokal, Robert R., and Rohlf, F. James. 1969. Biometry, the principles and practices of statistics in biolog- ical research. W. H. Freeman and Co., San Francisco. Sokal, R. R., and Sneath, P.H.A. 1963. Principles of numerical taxonomy. W. H. Freeman and Co., San Francisco. 359 p. Sokolov, S. J., and Cvjazeva, O. A. 1965. Geographic distribution of woody plants in the U.S.S.R. V. L. Komarov Botanical Institute, Leningrad. Stevenson, Frederick J., and James, Henry A. 1953. Some sources of resistance in crop plants. In Plant diseases. U.S.D.A. Yearbook, 1953. Swingle, R. V. 1949. Dutch elm disease. U.S.D.A. Year- book, 1949, pp. 451-452. Swingle, R. V. 1950. Report on research on the Dutch elm disease. Nat. Shade Tree Lab. Proc., Delaware, Ohio. Teed; Griffith, H., and Turrill, W. B. 1932. On flora of the near east, the elms of western Thrace. Kew Bull. 8:232. Teng, S. C. 1947. Silvaculture of Kansu trees. Bot. Bull. Acad. Sci. 1:232-233. Troup, R. S. 1921. Silvaculture of Indian trees. Clarendon Press, Oxford. Vol. III. 138 Tucker, C. M. 1945. Phloem necrosis a destructive disease of the American elm. Missouri Exp. Sta. Leaflet Circ. 305, 15 p. Vaartaja, O. 1959. Evidence of photoperiodic ecotypes in American trees. Ecol. Mong. 29:91-111. Walter, H. M. 1939. Effects of Ceratostomella uZmi on Ulmus americana and some types of European elms. Phytopath. 29:23. Wang, Chi-Wu. 1961. The forests of China. Maria Moors Cabot Foundation Publ. 5. Harvard Univ. Press, Cam- bridge, Mass. 313p. Whitten, R. R., and Swingle, R. V. 1948. The status of research on two epidemic elm diseases. Nat. Shade Tree Conf. Proc. 24:113-119. Wyman, Donald. 1951. Elm grown in America. Arnoldia 11:79-93. Zohary, Michael, and Feinbrun, Naomi. 1966. Flora Palastina Jerusalem. The Israel Academy of Science and Humanities. Zudlin, V. A. 1969. Studies on the resistance of elms to the Dutch elm disease. Les. Hoz. 3:62-64. 139 Vita Final Examination--July 12, 1971 Guidance Committee--J. H. Hanover, M. W. Adams, J. H. Beaman, J. W. Wright (Chairman) Dissertation--Variation Between and Within Nine Eurasian Species of Ulmus Biographical Data--Born May 4, 1940, Edwards County, Kansas Married Elaine Miriam Stickney, December 24, 1968 Education--Dodge City College A.A. 1960 Kansas State University B.S. 1963 University of Nebraska M.S. 1965 Michigan State University Ph.D. 1971 Experience--Graduate Assistant, University of Nebraska 1963-1965 Instructor, University of Nebraska 1965-1968 Graduate Assistant, Michigan State University 1968-1970 APPENDICES 141 Appendix A Species of Ulmus Bean 1951 Table A1. Authoratative references for the Species of Ulmus Authoratative referende . Species Botanical description ‘Range U. laevis Bean 1951 Sokolov and Komorov 1936 Cvajezeva 1965, Rehder 1940 Schreiber 1957 U. americana »Rehder 1940 'Fowells 1965 Harrar & Harrar 1946 U. mexicana Standley & Steyermark Standley & _ 1946 Steyermark 1946, Staub 1967 , Miranda 1952' . thomasii Rehder'1940 Fowells 1965 U. alata Rehder 1940 Fowells 1965 Harrar & Harrar 1946 a U, rubra Rehder 1940 'Fowells 1965 _‘ Harrar & Harrar 1946 U. elliptica Komorov 1936 ' Sokolov & _ Bean 1951 Cvajezeva 1957 U. glabra Bean 1951 Sokolov & Rehder 1940 Cvajezeva 1965 Komorov 1936 ‘ Tutin et a1. 1967 U. laciniata Komorov 1940 Sokolov & ‘Cvajezeva 1965 Ohwi 1965 Nakai 1932 Table A1. a. U. U. bergmaniana uyematsui wallichiana . minor japonica . wilsoniana procera pumila parvifolia) lanceaefolia crassifolia monterreyensis (Con't.) 14a Rehder 1940 Schneider 1917 Li 1963 Hooker Brandis Rehder 1940 Bean 1951 Komprov Nakai Komorov Bean 1951 Schneider 1917 Bean 1951 Rehder 1940 Komorov 1936 Rehder 1940 O-Kurz 1877 Rehder 1940 Harrar & Harrar 1946 Muller 1942 Schneider 1917 Li 1963 Hara 1966 ‘ ' Pearson & Brown 1932 Schreiber 1957 Sokolov & Cvajezeva 1965 ; Rehder 1923,1924 Sokolov & Cvajezeva 1965 Nakai 1932 Ohwi 1965 Schneider 1917 Rehder 1923,1924 Bean 1951 Sokolov & Cvajezeva 1965 Rehder 1923,1924 Wang 1961 Nakai 1932 -°Wang 1961 Li 1963 Ohwi 1965 Pearson & Brown 1932 Mbrrill 1934 Touw & VanSteenis 1968 McDaniel & Swift 1967 ‘Mu11er 1942 Table A1. U. serotina . kunmingeasis davidiana . macrocarpa villosa lesuerri . divaricata multinervosa (Con't.) Rehder 1940 Harrar & Harrar 1946 Cheng 1963 Nakai 1932 Komorov 1936 Nakai 1932 Bor 1953 Hooker 1885 Standley 1936 Muller 1937 Muller 1936 Staub 1967‘ Cheng 1963 Rehder 1923,1924 .Wang 1961 Sokolov & Cvajezeva 1965 _Wang 1961 Schneider 1917 Bar 1953 Hara 1966 Standley 1936 'Mu11er 1937 Muller 1936 Talus too too 300 400 W 75° 7°. The range of Ulmus americana 145 146 Ulmus alata The range of 147 mubzu wseAm mo omen» may. 148 . mucounooo pmuonwfl ux «AAomAmmmuo mseAD mo mwcmu one mm 00_ Hb/i. ; s x A e 4—o ... _ .. J . l J . . a . v _ , J A Efixxxxxxxfii_ 0. xxx a 4 v \\ ngmxfi‘ \\ \\ ‘\ t ,\ x “§§-::. ‘~~K , ‘ ,et \\xx\ . v o. l x xxxxxxxxxx (-.-..- -...i..-_.v-. .-.__ it \\ \sni J filtrf! w J . . I m >|\\l‘.(/ .u _ ,r I) C l \ a '7. . . .\ .I.|..|. xx 4.]. . .\ \ .3. 1 . p. .9, ....... \imdfi. x A Ii — I" "'-~ ‘ " ' " " " -'"' now- -- , , i ‘ 5 '- 6 x . ‘ l ._ n ‘ . l - .p . IOO 290 30° 4'90 MILES l L . l . Our-O I . r - l 200 400 600 KILOMETERS The range of Ulmus serotina o 7*. ..- 150 mcmoAXmE mseAm mo mwcmu one N mm 00. MEUFUIOJS. ”mi! 000 000 OON O. <—-....-.—.. 152 \l| ... x\\§,\x\ §§x \\ “\\xW\st\x.“.M\\x x\\- \Nwwwwkmwx ISOO MILES A ' I 2000 -xIL0METéns The range of Ulmus minor .I .III mcoommcwo msEAD mo mwcmu m5. 156 157 .wmAOMA>umm mosab mo owcmu may on. Guru’s! 4|“ 4 ad: 8» Il|||||.. Sendai/mp mos—AD mo owamu o5. . «uneconomic 3.2.5 «o swam.» one The. range of Ulmus wilsoniana 500 In.” . Km“! RS The range of Ulmus bergmaniana v I I 0 500 590 : IOOO l - I ,Lg‘ IOOO ISOO 2000 L ‘\:‘\ 1' J$€s ; 90 IOO ‘ :\ ..- ISOO MILES KILOMETER The range of Ulmus wallichiana 164 The range of I 'illosa Ulmus \ ___._ The range of Ulmus lanceaefolia = isolated occurence X Appendix B Table B1 Species hybrids reported in Ulmus 'Species . Authority 1 U. americana *- X U. laevis 6,7,17,18,22928‘ X U. rubra 18 ” "””’ X U. pumila 4,14,22,27,3O X U. bergmaniana 5,6 U. laevis ' 6 X U. thomasii 6,18 X U. minor 2,17,22 X U. pumila 1,2,5 X U. bergmaniana 6 X U. wilsoniana 6 U. thomasii X U. rubra 9,15 X.U. pumila 30 U. alata X U. pumila 6 U. rubra ' ' X U. pumila 3,6,12,17,31 U. procera X U. pumila 5 18 X U. minor 23 U. minor X U. canescens 20 . X U. glabra ‘ 16,17,21, , X U. pumila 2,6,8,10,13,16,24 X U. japonica’ 16, 34 X U. wallichiana 16,34 X U. laciniata 16 U. canescens 7X U. glabra 20 U. glabra , X U. pumila 11,17,21,22, X U. ja nica 29 X U. wa lichiana' 16,34 X U. laciniata 16 167 Table B1 (Con't.) U. pumila X U. japonica 16,19,34 X U. wallichiana 16 X U. wilsoniana 5 U..japonica' X U. wallichiana 7,16,22,33 X U. wilsoniana 12,29,32 U. wallichiana . X U. chumila 16 XU. laciniata 34 U. parvifolia- X U. crassifolia 26 B U. crassifolia . X U. lanceaefolia 16 1 authorities are listed by number below ' - 1. Albensky 1951 18. Johnson & Heimburger 1946 2. Albensky 1969 ' 19. Lester 1968 3. Anon. 1950 20. Melville 1957 4. Anon. 1954 21. Rehder 1940 5. Britwum 1961 (22. Richens 1945 6. Collins 1967 23. Richens 1961a 7. Doorenbos 1938 ‘ 324. Rovskii 1951 8. Ehrenberg 1954 25 ‘Rovskii.et a1. 1950 9. Fowells 1965 26 Santamour 1969*‘ 10. Goidanisch & Azzerolli 1938 27. Santamour 1970 11. Goidanisch & Azzarolli 1939 28. Sax 1933 12. Graves 1948 29. Smith & Nicholos 1941 13. Grudzinskaja 1961 30. Smucker 1944 - 14. Heimburger 1966 31. Swing1e19‘}5 15. Hess & Dunn 1967 32. weikoff 1941 16. Heybroek 1968 33. went 1938 17. Johnson 1939 34. went 1954 *Unpublished report to the Ii.C.R. 32 committee on vascular wilt diseases Morgontown,‘Jest Virginia 0013 027,1909 163 . Appendix 0 Characters Studied Tableicl. lame and Description of the Eight Samara Size Characters Measurednand the Ratios Computed. 'gfiggzgter ‘ Character Name Measured From 1; ' Base length Base of samara to base of seed 2. Seed length Base to tip of seed 3. Seed-notch- , Tip of seed to base of notch 4. Netch length . Base of notch to furthest point on samara tip 5. Samara Width Widest point on samara 6. . Widest point-base Widest point on samara to base _- of samara ~ ‘ 7. _ Seed width, , Widest point on seed 8. ‘ Samara length , .Base to tip of samara. Ratios Computed 9. Base length/samara length 10 ' Seed length/ Samara length 113 J Seed-notCh/'samara length _ 12. Netch length/ samara length 413. Samara width/samara length. 14. Widest point-base/samara length 15. Seed width/samara length i '16. Seed width/samara width l , . *‘I All samara characters were measured to the nearest thirty-second of an inch and recorded as the total number of thirty seconds of an inch for four samaras.» w r! 169 \ Table C2. Name, Range and Limits of the Eight Samara Shape and Color Characters. Character Character Range lowest No. Highest No. _Number Name Assigned to Assigned to 17. Tip Shape l-S f'pointed' 2-cleft 18. Base shape 1-5 straight most outcarved ‘ 19. Openness or .'1A5 closed openest - notch -' 20; Style length 1-5 shortest longest 21. Style shape 1—5 straight most ~ ‘ curved 22. Wing color 1-5 lightest darkest .23. Seed color 1-5 lightest darkest 24. Greedpss of 1-5' no green greenest samara . 25. Persistent ‘O-4il none all calyxes persisting persisting 26. Contrast a) same color largest between seed - J. contrast‘ . color and wing color 17’ PerSiStent calyx was scored by counting the number still attached to the samara. 2;) The contrast in color was computed by subtracting character 22 from 23. darker than wing color. Seed color was.always as dark or 170 Table 03. Name and Description of the Leaf Characters and the Ratios Computed. —r Character Character Description of Character Number Name 27. I Leaf lengthlj’ ~Measured from the tip of the . leaf to the lowest point on the base. 28. Leaf width Measured at the widest point 29. Number of veins Counted on the smallest half of the leaf. 50. Number of major Counted on the smallest half ' serrations of the leaf. 31. Total number of Counted on the smallest half serrations of the leaf. 32. I Widest point . ‘ Measured-from the widest point to tip ‘ on the leaf to tip of leaf. 2 33. .Serration depth ’ Measured from the tip of ' ' serration to base of next higher . serration 34. Serration lengthgl Measured from tip of serration to base of next lower serration 55. . ‘Serration widthgj deasured across the base of : serration from base of next lower to base of next higher. 36. Major—minor . Measured from the tip of major serration serration to the tip of first lower serration. 37. - Major-major Measured from the tip of major serration serration to the tip of the next major serration. 38. Obversity The difference between the ‘distances from the tip of the leaf to the furthest point on the short side and the long side of the leaf TableC3 . 39. 40. ' 41. t 171 Leaf Characters (continued) Petiole length Length of the petiole Pubescence of Scored on a scale of upper leaf surface 1 (glaborous) to 5 (densely pubescent). Pubescence of Scored on a scale of lower leaf surface 1 (glaborous) to 5 (densely pubescent). Ratios:Computed . 42. 45. 44.-.‘ ' 45. _ 46. 47. ”48. 49. so. 51. '52, 53. 54. , 55. Number of veins/number of major serrations Number of major serrations/total number of serrations ' Leaf width/leaf length Number of veins/leaf length. Number of major serrations/leaf length Total number of serrations/leaf length Widest point to tip/leaf length -Serration~depth/leaf length Serration length/leaf length Serration width/leaf length Majoreminor serration/leaf length.- Major-major serration/leaf length Obversity/leaf length Petiole length/leaf length ;>/ All size measurements were made to the nearest m.m. and scored as the total number of m.m. for five leaves. _§/’ All serration measurements were made on the largest serration which is usually located just above the middle of the leaf. 172 | Table C4 . Name and Description ee the Seedling Characters Measured in the Hursery. Char. No. Character Date Recorded Range Description 56. 57. 58. 59. ‘ 60. 61. 52. 63. 64. Percent Opposite leaves ”Gnomium ulmigj Red fall color Frost injury Gnomium ulmi injury Corky twigs l l Fall.color Long shoot- short shoot Frost injury Sept.68;l Oct.68 Oct.68 Nov.68 Aug.6921 C Sept.69 Oct.69 0ct.69 NOV. 69 0-100 I 0-4 0-4 0-10 0-10 0-4 0-10 0-5 Percentage of seedl: in each accession w: opposite leaves. Percent of trees in each accession show: symptoms of this 1;;_ ‘spot disease; to ALL: 25%. Percent of seedling: in each accession that have some red leaves;to nearest 2: Percent of leaves damaged by frost; to nearest 10%. (Relative amount of injury in each plot from none to severoe No.0f seedlings in each.plot with cork; outgrowths on twogs. 'Relative amount of fall color in each 3. Some seedlingshad t types of leaves, In: leaves on the main : and small ones on branches, rated as * number of seedlings in each plot showing this character. Relative amount of injury from slight t severe. Table C4 . 65. 66. 67. 68. 59. 70. 71. 72. 73. 74. 75. Green.leaves Stem color Twig color Number of .branches Secondary branching Branch angle Leafing out Average height .Susceptibility to Dutch Elm Disease Insect attack ' Contrast between stem and twig color 173 Nursery Characters (continued) Dec. 69 Mar.70 Liar . 7O iar.7O Apr.7o Apr. 70 ' Apr.7o Sept.69 Aug.70 Aue.7o | 0-10 0—10 0-10 No. of seedlings in each plot that still retained gree leaves On D00 12 0 Rated from the lightest to the darkest. Rated on the same scale as Charo 660 .“.3 Estimates as the sq. 0-10 0-100 0-100 root of the average no. of branches per seedling No. of seedlings in each source that were branched more than once. Angle of the branches with the stem from 0-15 degrees to 45-90 . degrees. Scored as the average progress of leafing meach lot from 0 (dormant? to 5 (leaves fully expanded). Measured to the nearest inch. Percentage of seedlings -in each source showing symptoms 60 days after inoculation. Percentage of seedlings in each source attacked by an unidentified leaf eating insect. Computed by subtracting char.66 from char.67. .lJ' All 68 measurements were made in original unreplicated nursery beds. it Gnomium ulmi disease identified by Dr. John Hart , t. of slant Pathology. ..2/ A11 69 and 70 measurements made in replicated transplant bod? 174 Appendix D Ulmus minor Table D1. Source of U, minor accessions Locality no. and place Lat. N.. Long. E. MSFG No. Austria 0 10. Vienna 48 11' 16°23' 6351-54,6357 . 6360-61 - Czechoslovakia o ' 20.* - Brno 49°10' 16 35' 6381, 6385, 2933's? -1 25. Prague 50°10' 14°30' 6498, 6834, 2323’ 2323’ 26. Stanvico 48952' 19°03' 6580182 England 0 ' 30. Kew - 51 29 0°02' 6408, 6481 Germany , 4o. Freiburg 48°05' .7035' 6433-38 41. ‘Munich 48°10' 11°30' 6464 42. Hannover-MMnden 51°24' 9040' 6519-20, 6522 Greece 50. Thessalonika 40°35' 23°02' 6341-6349 Hungray . 6 . Asvanyaro 47305: 18300: 6395-6400 66. Szeged 46 15 20 05 6988-92 Italy ‘ . o 7 . Florence 43 40' 11°10' 2325-35 , 533-36 72. Padova 45°24' 11°50' 6324 Poland 115. Krakow 50°00' 20°00' 6418 116. Kornicke 52°17' 17°06' 6475 117. wroclaw 51°10' 17°00' 6488, 6491 Table D1. (Con't.) U.S.S.R. 130. Baku 134. Krasnodar 135. Nikitiski 138. Tbilisi 139. Yerevan 146. Makhachkala Yugoslavia 155. Skopje 157. Zagreb 48330' 38 30' 37015' 45°00' ' 44°30' 56°10' 21330' 15 58' 6556-61 6711-13 6562-71 6643-44 6697-98, 6716-20 6795-6803 6336, 6338-40 6632 Table D2. 176 The percent of variance in samara traits of U, minor accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality and error. Character no. Percent of variance accounted for by and name Region Locality/region 1. Base length 10* 32** 2. Seed length 3 19* 3. Seed-notch 6 9 4. Notch length 11* 25** 5. Samara width 4 30** 6. Widest point-base 5 34** 7. Seed width 20** 20** 8. Samara length 11* 39** ~9. Base length/samara length 3 19** 10. Seed length/samara length 14 0 11. Seed notch/samara length 7 9 12. Notch length/samara length 7 0 l3. Samara width/samara length 13* 0 14. Widest point-base/ 0 17* samara-length 15. Seed width/samara length 7 11 ’16. Seed width/Samara width 13* 11 17. Tip shape. 5 7 18. Base shape 0 27** . 19. openness of notch 0 30** 20. Style length 0 31** 177 Table 02.;Con't.) 21. Style shape 22. Wing color 23. Seed color 24. Greeness of-samara -b b- h) h: L0 25. Persistant calyxes 26. Seed-wing color contrast 1 *, ** significantly different at .05 and .01 level reSpectively. 14 25** 27** 37** 24** 19* Table D3. 178 The percent of variance in leaf traits of Q, minor accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality and error. Character no. Percent of variance accounted for by and name Region Locality/region 27. Leaf length 31** 29* '28. Leaf width 7 14 4 29. Number of veins 27** 22* 30. Number of major serrations 44** 0 31. Total number of serrations' 17* . 49** 32. Widest point to tip 48**1 21 33. Serration depth . 14 53** . 34. Serration length 35** 2 35. Serration width 31** 19 36 . Maj or -minor serration 43H? 0 37. Major-major serration 35** 18 38. Obversity 9.- 40** 39. Petiole length 4 11 40. Pubscence of upper leaf 17* 20 surface ‘ 41. Pubscence of lower leaf 0 22* surface 42. Number of veins/number of major serrations l 3 43. Number of major serrations/ , total number of serrations ‘.0 0 179 Table D3. (Can't ) 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. Leaf width/leaf length Number of veins/leaf length Number of major serrations/ leaf length Total number of serrations/ leaf length Widest point to tip/ leaf length Serration depth/leaf length Serration length/leaf length Serration width/leaf length ‘Major-minor serrationI' leaf length ‘Major-major serration/ leaf length Obversity/leaf length Petiole length/leaf length 23* 29** O *, ** significantly different at .05 and reSpectively. 70** 33** 0 17 11 50** 33** 11* .01 level 180 Table D4. The percent of variance in nursery traits of g, minor accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality error. Character no 1 Percent of variance accounted for by and name - Region _ Locality/region 56. Percent opposite leaves . 24 36** . 57. Gnomium 21mg, injury 1968 O O 58. Red fall color 0 ' 12 59. Frost injury, 1968 ' 12 9 60. Gnomium 21m; injury, 1969 . O ' 0 61. Corky twigs - ' .0 9 62. Fall color 8 20* 63. Long shoot-short shoot . 39** 0 64. Frost injury. 1969’ 0 17 65. Green leaves, 12/12/69 60** 2 66. Stem color 1 I 27 67. Twig color 11 . ' 4 66. Number of branches 20 0 69. Secondary branching 0 18 70. Branch angle 15 > O .71. Leafing out ‘ 7 O 72. Average height , 45** 10 73. Susceptability to Dutch 2 10 elm disease . 74. Insect attack 0 0 75. Contrast between stem 1 37** 5 and twig color *, ** significantly different at the .05 and .01 level vaamnP-i\ra1 u 131 Appendix E Ulmus laevis 'Source of Q. laevis accessions Table El. Locality no. ‘J-ua-l .- .— MSFG N0. 22°oo' and place Lat. N. Long. E. 'Austria - _ .10. Vienna 48°ll' 16017' 6355, 6357-58 Czechoslovakia~ _ of ' V 20. Brno 49°10' 16 35' 6377-80,6386, - 6391, 6509-11, 0‘ o 6516 25. Prague 50 10' 14 30' . 6502, 6829-31, 6838-41 England 0 . 30 Kew 51 29' 0°02' '6404-05 Germany 0 ' o 40. Freiburg 48 05 7 35' 6440-44 41. Munich 48°10' 11°30' 6468-69 Hungary ‘ o 60. Asvanyaro 47°05' 18 00' 6392-94,6401-03, ' - - 0 6983-87 66. Szeged 46 15' 20°05' . 6368-69 Netherlands A o 100. Wageningen 51 55' 5 30' 6621-22 Poland ~ 0 115. Krakow 50°00' 20°00' 6413-16, 6420-21 116. Kornicke 52 17' 17°06' 6425-26 ‘117. wroclaw ~ 51°10' 17°00' 6489-90, 6493-94 118. Rogow 51°00' 20°00' 6583-87 Romania 0 122. Oradea 47 00' 6763-64 Table El (Con't.) U.S.S.R. 132. 133. 134. 137. 139. 140. 143. 144. Minsk Leningrad Krasnodar Riga Yerevan Voronezh Alma Ata Kiev Yugoslavia 155. 157. Skopje Zagreb 182 53352' 59 58' 45°00' 56°58' 40°08' 51°42' 43°14' 50°24' 42°00' 45°49' 27°32' 30°18' 38°30' 24°04' 44°28' '39002' 76°56' 30°27' 21828' 15 29' 6722 6685, 6724-28 6708, 6710 6683-84, 6686-90 6691-6695 6709, 6781 ' 6772-76, 6785-89 6777 6337 6637-39 Table E2. The percent of variance in samara traits of Q. laevis accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality and error. Character no. Percent of variance accounted for by and name Region Locality/region 1. Base length 28* 7 2. Seed length 17 14* 3. Seed-notch .5 15* 44 Notch length 44** 15* 5. Samara width 30* . 0 6. Widest point-base 5 25** 7. Seed width 16 0 8. Samara length 24* 43** 9. Base length/samara length 0 10 10. Seed length/samara length 13 0 11. Seed notch/samara length 0 o 12. Notch length/samara length 41** 10 13. Samara width/samara length ‘0' 13 14. Widest point-base/ . 0 9 samara length ' . 15. Seed width/samara length 0 0 l6. Seed width/samara width 25* o 17. Tip shape 25* 13 18. Base shape 40** 17*. 19. Openness of notch 30* 0 20. Style length 0 23* 195 Table 02. (Con't.) 21. Style shape 22. Wing color 23. Seed color 24. Greeness of samara 25. Persistant calyxes 26. Seed-wing color contrast *, ** significantly different at reSpectively. O 82** 0 84** 0 89** 0 88** 55* 24 3 25 .05 and .01 level !.:' '- l" 185 Table E3. The percent of variance in leaf traits of g, laevis accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality and error. Character no. I Percent of variance accounted for by ,and name Region Locality/region 27. Leaf length ' 18* 10 28. Leaf width 29** >10 29. Number of veins 1 19* 14* 30. Number of major serrations 7 J13 31. Total number of serrations 17* ' 18* 32. Widest point to tip 25** I 19* 33. Serration depth a 7* 34. Serration length :25** ~ 26* 35. Serration width ' ' 10 7 36. Major-minor serration 6 , 8 37. Major-major serration 9 8 38. Obversity 0 . 18* 39. Petiole length 19* ' ll 40. Pubscence of upper 0 0 leaf surface 41. Pubscence of lower * ‘0 0 leaf surface 42. Number of veins/number of 0 . 9 major serrations 43. Number of major serrationS/ . 23** 6 total number of serrations Table G3. (Con't.) 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.. 54. 55. *, ** significantly different at .05 and .01 level Leaf width/leaf length Number of veins/leaf length Number of major serrations/ leaf length Total number of serrations/ leaf length Widest point to tip/ 1eaf_length Serration depth/leaf length Serration length/leaf length Serration width/leaf length ‘Major-minor serration/ leaf length Major-major serration/ leaf length Obversity/leaf length Petiole length/leaf length reSpectively. 15 35* 32* 50%"): ' 0 0 44** 0 41** 25* 33* 48** 33* 49** 21 63** 'The percent of variance in nursery traits of g. laevis accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality and error. Table E4. 4 Character no. Percent of variance accounted for by -and name Region Locality/region 56. Percent opposite leaves 3 11 57. Gnomium glgi_injury, 1968 32* 18 58. Red fall color 0 0 S9. Frost injury, 1968 17 21* 60. Gnomium 21Eg_injury, 1969 41** 25* 61. Corky twigs 0 0 62. Fall color 0 o '63. Long shoot-short shoot 0 0 64. Frost injury, 1969‘ ' 13 33** 65. Green leaves, 12/12/69 0 32** 66. Stem color 61** 0 67. Twig color 12 25** 68. Number of branches 52** 15 69. Secondary branching 33* 13 70. Branch angle 1 13 13 71. Leafing out 0 29** 72. Average height 2 -47** 73. Susceptibility to Dutch 13 43** elm disease 74. Insect attack 29* 2 -75. Contrast between stem 64** 0 and twig color , *' ** a!an4F4nnnf1u.A{FFn-ranr af- m; an"! n1 1aua1 racmnfivplv 1...! 0) Appendix F Ulmus glabra Table F1. Source of g, glabra accessions 'Locality no. and place, Lat N Long. MSFG No. Austria 10. Vienna 48°11' 16°18' E. 6350, 6359 11. Innsbruck 47314: 11329: E. 6411-12, 6766. 12. Judenburg 47 10 14 40 E. 6458-63, 6592 Czechoslovakia 20. Brno 49°10' 16°35' E. 6372-76, 6390 6517-18 :25. Prague 50°10' 14°30' E. 6495-97, - 2502-06’3 36 50 8 5- 26. Stiavnia 49°00' 19°00' E. 6572180 England . 30. Kew -51°29' 0°02' w. 6406-07 31. Oxford 51°50' 1°10' w. gggg-gg, 33%. Banburg 33:18: 1838: W. €473, 6481 . Towcester l W. 48 34. Warlingham 51925' 0000' 6623-24 35. Long Ashton 51101' 2040' w. 6659-64 36. wrecclesham 51°05' 1006' W. 6659-64 Germany 0 40. Freiburg 48°58: 7333: E. gzég-gg '41. Munich . 4 1 11 E. - 42. Hannover-Mbnden 51°24' 9040' E. 6521, 6523 43. Gottingen 51°00' 9055' E. 6455-57 Hungary ' 65. SOpron 47°41' 16°30' E. 6367, 6370 Italy . o 71. Florenc 43 40' 11°10' E. 6445-52, 6537-44 Table F1. Poland 115. Krakow 116. Kornik 117. wroclaw .118. Rogow Romania 121. Bucurest 122. Oradea 123. Cluj U.S.S.R. 132. Minsk . 133. Leningrad 134. Krasnodar 137. Riga 138. Tbilisi Yugoslavia' 155. Skopje 157. Zagreb Scotland 160. Dingwall (Con't.) 50°04' 52°17' 51°07' 51°49' 44°30' 47°00' 46°40' 53°52? 59°58' 45300' 569-58. 41°40' 42°00' 45949' 57°32' [11131111 ["1 F1 mm mm 1:1th mm 6417, 6419, 6423, 6427 6476 6492 6588-91 6453-54 6761 6832-33, 6894 6721 ,6677-78, 6729-34,6805 6714 6679-80, 6843-47 6641, 6715 6382-84 6633-36 6627-31 190 Table F2. The percent of variance in samara traits of Q, glabra accounted for by differences among regions and localities within regions. The -remaining variance not accounted for is due to parent within locality and error. Character no. Percent of variance accounted for by and name Region -Loca1ity/region 1. Base length 1 35** ‘ 2. Seed length . 4 24** 3. Seed-notch 2 47** 4. Notch length 5 0 5 . Samara width 2 21* 6. Widest point-base 2 26** 7. Seed width 1 37** 8. Samara length 4 22* 2 9. Base length/samara length 2 29** 10. Seed length/samara length _6 , 33** 11. Seed notch/samara length 3 37** 12. Notch length/samara length 3 0 13. Samara width/samara length 6 3 14. Widest point-base/ 0 7 samara length - 15. Seed width/samara length -10* 17* 16.'Seed width/samara width 0 5 17. Tip shape 4 0 18. Base shape 1 22* 19. Openness of notch 4 22* 20. Style length 1 0 36** 191 Table F2. (Con't.) 21. Style shape 22. Wing color 23. Seed color 24. Greeness of samara 25. Persistant calyxes OOOH'OO‘ 26. Seed-wing color contrast *, ** significantly different at .05 and .01 level reSpectively. 22** 47** 43** 27** 17** 32** 192 Table F3. The percent of variance in leaf traits of Q. glabra accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due . to parent within locality and error. Character no. Percent of variance accounted for by and name Region Locality/region 27. Leaf length 12 6 28. Leaf width 15 7 29. Number of veins 0 31 30. Number of major serrations 32* 4 13 31. Total number of serrations 35* j 0 32. Widest point to tip 15 5 33. Serration depth 9 36 ‘ 34. Serration length 25 5 . 35. Serration width _ ‘ 16 5 36. Major-minor serration 21 26 37. Major-major serration 9 25 38. Obversity 10 0 39. Petiole length 0 . ' o 40. Pubscence of upper leaf 1 33 surface 41. Pubscence of lower leaf 0 75 surface 42. Number of veins/number of 30* ' 25 major serrations 43. Number of major serrations/ 0 18 total number of serrations 193 Table F3. (Con't.) 44. Leaf width/leaf length 0 45. Number of veins/leaf length 0 46. Number of major serrations/ 0 leaf length 47. Total number of serrations/ 0 leaf length 48. Widest :point to tip/ 12 leaf length 49. Serration depth/leaf length 5 50. Serration length/leaf length 0 51. Serration width/leaf length 4 52. Major-minor serration/ 21 leaf length 53. Major-major serration/ 0 leaf length » 54. ObVersity/leaf length 17 55. Petiole length/leaf length 0 *, ** significantly different at .05 and .01 level reSpectively. 23 30 26 11. 21 48 ‘ HF 194 Appendix G Ulmus pumila Table 61. Source of g, pumila accessions Locality no; Dakota and place Lat. N. Long. E. MSFG No. Italy 0 71. Florence 46 10' 13°15' 6545-55 Japan 0 84. Nabeyashiki 39 30' 141°00' 6504-07 Korea ' 90. Suwon 37°20' 127°00' 6656-57, 6742-43 Romania Cluj 6895 . U.S.S.R. - . 131. Khabarovsk 47°50' 135°10' 6675 133. Leningrad 59°58' 30°18' 6645 137. Riga 56°58' 24°04{ 6676 139. Yerevan 40°00' 44°30' 6699-6703 140. Voronezh, 51°42' 39°02' 6782 143. Alma Ata 43°14' 76°56' 6767-71, , 6790-94 144. Kiev 50°24' 30°27' 6778-79 145. Askania Nova 33°58' 46°27' 6780 147. Krasno yarsk 52°10' 92°24' 6972 U.S.A. , M.S.U. Campus 3331-69, Brookings, South 6892 Table G2. 195 The percent of variance in samara traits of Q. pumila accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality and error. Character no. Percent of variance accounted for by and name ‘~ Region Locality/region 9' 1. BaSe length 0 69** 2. Seed length 0 8 3. Seed-notch A 0 8 4. Notch length 5 17 5. Samara width 14 6 6. Widest point-base 3 17 7. Seed width 0 32* 8. Samara length . 0 14 9. Base length/samara length 0 '0 10. Seedlength/samara length 0 42* 11. Seed notch/samara length 0 .0 12. Notch length/samara length 0. 41* , 13. Samara width/samara length 16 41* 14. Widest point-base/ 44* o samara length 15. Seed width/samara length 0 26 16. Seed width/samara width 0 47* 17. Tip shape 0 0 18. Base shape 0 17 19. Openness of notch 0 83** 20 . ' .0 75** Style length 199 Table G4. The percent of variance in nursery characters of Q, pumila accounted for by differences among regions and localities within regions. The remaining variance not accounted for is due to parent within locality and error. Character no. Percent of variance accounted for by and name Region Locality/region 56. Percent opposite leaves 4 16 57. Gnomium 2121 injury, 1968 ' . 0 - o 58. Red fall color 3 15 59. Frost injury, 1968 _ 39* 20 60. Gnomium 91mg_injury, 1969 0 ' 0 61. Corky twigs ' 0 ' 0 62. Fall co1or ‘ 0 I 0 63. Long shoot-short shoot 0 ‘ 12 64. Frost injury, 1969- . , 24 14 65. Green leaves, 12/12/69 ‘I 4 6 66._Stem color 4 1 0 ' O 67. Twig color 0 . 17 68. Number of branches 17 '19 69. Secondary branching 5 23* 70. Branch angle, 0 . 0 71. Leafing out 15 19 72. Average height , 32* . 36* 73. Susceptibility to Dutch 18 31* elm disease ‘ 74. Insect attack I w 19 24* 75. Contrast beteen stem and 0 12 twig color *, ** significantly different at .05 and .01 level respectively. 3m .1. 200 Appendix H Ulmus japonica Table H1. Sources of Q, japonica accessions H”‘ Locality no. and place ' Lat. N. _Long. E. MSFG No. f Japan . 80. Sopporo 43°00' 141°00' 2819-28, . 973-82 1. 81. Ashoro Machi 43°18' 142°00' 6646-55 2; 82. Ebetsu 43°04' 141°03' 6744-48 - 83. Kuri ama 43°OO' ' 142°00' 6665-74 86. Chic ibu ' 36°00' 139°10' 6819-28, 6973-82 201 Appendix I Ulmus parvifolia Table 11. Sources of g, parvifolia accessions Locality no. and place Lat. N. Long. E. MSFG No. Japan 85. Fugi-Yoshido 35330' 139000' 6323 87. Nagasaki 32 50' 129°40' 6858 88. Nagoya 35°10' 136°55' 6859-66 Korea V 90. Suwon 37°20' 127°00' 6658 90. Suwon 37°20' 127°00' 6873-74 U.S.A. ' 171. M.S.U. campus 6848-51, 6870-71 202 Appendix J 91322 1aCiaiata Table J1. Sources of g, laciniata accessions Locality no. E7 and place Lat. N. Long. E. MSFG No. f 5 England ' ' 30. Kew 51°29' 0°02' 6410 . .L Netherlands ' “ 100. Wageningen 51°55' 5030' 6620 Japan 0 84. Morioka 39°30' 141 00' 6508 86. Chichibu . 36°00' 139010' 6807 89. Yamabe - 43°00' 142°30' 6749-54 [“3 C) \A Appendix K Ulmus wallichiana Table Kl. Sources of 11,. wallichiang accessions Locality no. .and place Lat. N. Long. E. 'MSFG. No. West Pakistan ' o 180. Peshawar 34°10' 71 50' 7000 Netherlands- . 100. Wageningen _ 51°55' 5°30" 6614-19 204. Appendix L Q1QE§Ielliptica Table Ll. Sources of g, ellipita accessions U.S.S.R. 139. Yerevan 40°08' 44°28' 6696-97 138. Tbilisi 41°40' 44°57' 6642 133. Leningrad 59°58' 30°18' 6681-82 133. Leningrad 59°58' 30°18' 6741 4