fl" ‘H ‘V C) O r O - 7— —* v._ . #1 ‘ ~ - - - . NI. u VERSWY “5391"“ 1 4-.. i\\\‘\‘\\°\\l‘llllllllllllllll\‘5\\\l\\\\\ L if ~ , ,; 4 a, THESIS 3 ‘i‘zig‘ai ‘i0\612 6579 g . , léfifaxafi? ; tzxegezz £35ng vzzzders'z’ty This is to certify that the dissertation entitled HONEYLOCUST (Gleditsia triacanthos L.): GENETIC VARIATION AND POTENTIAL USE AS AN AGROFORESTRY SPECIES presented by Michael Alan Gold has been accepted towards fulfillment of the requirements for Ph . D . . FORESTRY degree in zit/pm 1/5/41444‘44/ / Major professor James W. Hanover AUGUST 7, i984 I)ate MSU is an Affirmative Action “Equal Opportunity Institution 0-12771 -C— --—- r1. I .I. '_| ,69‘0a t0. . {$5 I MSU LlBRARlES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. ‘ was; 71969:? in”??? 1005 ‘m . 1m, 3 c |.. ‘0 I HONEYLOCUST (Gleditsia triacanthos L.): GENETIC VARIATION AND POTENTIAL USE AS AN AGROFORESTRY SPECIES BY Michael Alan Gold A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1984 ABSTRACT HONEYLOCUST (Gleditsia triacanthos L.): GENETIC VARIATION AND POTENTIAL USE AS AN AGROFORESTRY SPECIES BY Michael Alan Gold Honeylocust, Gleditsia triacanthos L., has potential as a multi-purpose tree in agroforestry systems due to a combination of desirable traits including high wood specific gravity, abundant coppicing, a tap-rooted/profusely branched root system, drought tolerance, high carbohydrate pods, and high protein seeds and leaves. The potential for genetic and/or cultural improvement in the growth and form of honeylocust is unknown. An in- depth study of the geographic and genetic variation in honeylocust was initiated in 1979 with the establishment of a comprehensive rangewide provenance/progeny test at two locations in southern Michigan. At the end of the second growing season, significant differences among regions and half-sib families-within- regions were found in total height, caliper, thorniness, date of spring flushing, stem dieback, and fall growth cessation. Strong negative correlations were found between latitude of origin and stem dieback. The ranges of variation in spring flushing, fall growth cessation, leaf retention, and stem dieback appear to follow clinal patterns. Families of northern origin from the Lake States area are the best overall juvenile performers in terms of total height, stem caliper, cold-hardiness and degree of thorniness. Results of chemical analyses on pod sugars and seed and leaf proteins are reported. Total pod sugar content varied from 13.6 to 30.9 percent. Seed protein content varied from 16.6 to 27.8 percent. Leaf protein content ranged from 13.6 to 28.9 percent. The variation patterns in leaf protein, seed protein, and pod sugars are random with no particular provenance or region being especially high in any given trait. The use of yield components is discussed in relation to breeding strategies for maximizing sugar and protein yields. Results of two cultural studies on preemergent herbicides and spacings are reported. Ultra short rotation intensive culture systems for growing honeylocust can be succesfully accomplished by direct-seeding, followed immediately by application of the preemergent herbicide DCPA (dimethyltetrachloroterephthalate) with no harmful effects on seed germination. Planting direct-seeded honeylocust at three different spacings showed that a spacing of 10 x 15 cm gave the highest biomass yields in the first year after planting. ACKNOWLEDGMENTS I wish to express my sincere appreciation to Dr. James W. Hanover, Chairman, for his insights, support, and assistance which allowed this work to become a reality. I also wish to thank the other members of the Guidance Committee--Drs. R. Bandurski, M. Yokoyama, and J. Hancock-- for their assistance during the course of this study. Special thanks are due to my "little brother"--D. Van Arsdall--who gave me a tremendous amount of help in my field data collection efforts. Thanks are due to all the other members of MICHCOTIP who made this whole experience alot more bearable. Finally, my heart gives a big and sincere thank—you to my wife, Julie, for her unbelievable patience and encourage- ment during the long years of this project. ii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . LIST OF FIGURES . . . . . . . INTRODUCTION . . . . . . . . . Chapter I. GENETICS OF HONEYLOCUST Introduction . . . . Generic Relationship Phylogeny . . . . . . Sexual Reproduction . Seed Collection and Nu Asexual Reproduction Cytology and Mutation Genetics and Breeding Improvement Programs Conclusions and Recommend List of References . r 89 H CHOOOOKOOOO 000.comoogo DS 9).... (To... C II. HONEYLOCUST HALF-SIB PROVENANCE/PROGENY TWO YEAR RESULTS . . Abstract . . . . . . Introduction . . . . Materials and Methods Results and Discussion Conclusions . . . . . List of References . TEST: III. AGROFORESTRY SYSTEMS FOR THE TEMPERATE ZONE . Abstract . . . . . . Introduction . . . . Historical Development Managed Conifer Sawlog/grazing Systems Multicropping High Value Hardwoods With Agricultural Crops . Systems of Woody/woody Intercropping With N2 Fixing Woody Plants iii Page vii 10 12 18 20 24 25 32 39 41 46 46 47 49 59 74 75 77 77 78 81 88 94 98 Chapter Page Conclusions . . . . . . . . . . . . . . . . 104 List of References . . . . . . . . . . . . 105 IV. HONEYLOCUST (Gleditsia triacanthos L.): IMPORTANT CHEMICAL CHARACTERISTICS AND CULTURAL SYSTEMS FOR USE IN AGROFORESTRY SYSTEMS O O O O O O O O O O O O O O O O O O 112 AbStraCt O O O O O O O O O O O O O O O O O 112 IntIOdUCtion O O O O O O O O O O O O O O O 113 Silviculture and Genetics . . . . . . . . . 116 Historical Background . . . . . . . . . . . 117 Materials and Methods . . . . . . . . . . . 120 Results and Discussion . . . . . . . . . . 125 Conclusions and Recommendations . . . . . . 137 List of References . . . . . . . . . . . . 140 APPENDICES O O O O O O O O O O O O O O O O O O O O 0 14s A. CORRESPONDENCE AND COLLECTION FORMS . . . . . 146 B. ACCESSION RECORD FOR INITIAL HONEYLOCUST RANGEWIDE COLLECTION . . . . . . . . . . . 152 C. PLANTATION MAPS O O O O O O O O O O O O O O O 163 iv CHAPTER Table 1. 2. 3. CHAPTER Table 1. LIST OF TABLES I The genus Gleditsia . . . . . . . . . . . . Interspecific variation in the genus GleditSia O O O O O O O O O O O O O O O O Combined three year results of an experiment to determine the inheritance of the thorn trait in honeylocust .. . .. . .. . .. II Origin data for 57 half-sib families anal- yzed in 1983, and used in subsequent dis— CUSSionS O 0 O I O O O O O O O O O O O O 0 Expected mean squares in 3-level nested ANOVA used to partition out variance com- ponents and calculate heritabilities . . . Three-level nested ANOVA on 57 half-sib families from a rangewide provenance/pro- geny test at two locations in southern Michigan . . . . . . . . . . . . . . . . . Simple product-moment correlations on phen- ological, morphological and growth char- acteristics of honeylocust . . . . . . . . Variation among regions in traits analyzed . Origin, phenology, and survival data for 57 half-sib families used in analysis and diSCUSSion O O O O O O O O O C O O O O O O Page 31 52 56 57 58 61 62 Table 8. CHAPTER Table 1. Page Variation accounted for among regions, families within regions, and trees within families (expressed as a percent of total phenotypic variance), and heritability values . . . . . . . . . . . . . . . . . . 68 Performance of the 5 best honeylocust half-sib families at age two at two loca- tions in southern Michigan, based on height, caliper, and stem dieback . . . . 74 IV Variation in morphological and chemical traits of honeylocust parent trees from subsampled rangewide test . . . . . . . . 126 Means and range variation in honeylocust pod sugar content of seed sources among and within regions . . . . . . . . . . . . 127 Variation in leaf protein content among geographic regions . . . . . . . . . . . . 131 Preemergent herbicide effects on germina- tion, survival, and weed control of direct-seeded honeylocust . . . . . . . . 133 Biomass production in a l-year old, direct- seeded honeylocust USRIC system . . . . . 135 vi LIST OF FIGURES CHAPTER I Page Figure 1. Natural range of honeylocust . . . . . . . . 8 2. Gleditsia. a-j, g. triacanthos: a, tip of of branch with staminate inflorescences-- note once pinnately compound leaves on short shoots, X 1/2; b, staminate flower with 4 stamens, X 3; c, carpellate flower --note presence of aborted anthers, X 3; d, same in vertical section--note peri- gynous insertion of calyx lobes, petals, and stamens X 4; e, mature fruit-- northern latitude phenotype, with little pulp, X 1/4; f, seed, X 2; g, soaked seed in cross section, mucilaginous endosperm even-stippled, seed coat and cotyledons unshaded, X l; h, embryo from soaked seed, X l; i, supraxillary thorn from branch--note three-pronged thorn, X 1/4; j, adventitious thorn from tree trunk, can be up to 40mm in length, X 1/4 . .. . l4 CHAPTER II Figure l. Naturalized range of honeylocust and loca- tions of regions, seed trees, and test plantations O O O O O O O O O O O C O O O 55 vii INTRODUCTION The present solutions to many major global problems are destructive, inadequate and shortsighted. We face a future in which new ideas based on the concept of one ”global ecosystem" must come to the forefront. There is a need to think toward a long term future of cooperation, permanence, and stability. Included among the unending list of serious global problems facing the "less developed countries" are food shortages, deforestation, erosion, and a subsequent lack of fuelwood. In the "developed countries” there is a need to diversify the economic base, lessen dependence on fossil fuels and look toward renewable production systems. In all regions there is a need to make fuller, more productive use of marginal and hilly/steep lands. The solution to these major problems will certainly be a multi-faceted one which will include energy conservation, improved farming practices, and the development of alternative, less energy-intensive technologies. One important facet of the technological solution to these problems may lie in the new field of "agroforestry". This integrated system offers many opportunities and advantages in solving the energy, food, and soil erosion problems. In agroforestry, forestry is integrated with farming, animal husbandry, and horticulture to achieve both maximum output per hectare of land, and optimum conservation of the land resource on a permanent basis. Much of the literature discussing the value of agroforestry species and systems consists of speculation drawn from little concrete information. One of the species commonly mentioned as having potential for use as an agroforestry species is honeylocust, Gleditsia triacanthos L. This assertion is based on a very limited data base. In light of this situation, it was obvious that a long-term effort was needed to establish a firm foundation for future research. Gathering information on the extent of genetic variation in the species is among the first questions which need to be answered . Therefore, a major collection of germplasm was organized to address this question. This now exists as a three year old provenance/progeny test in two locations in southern Michigan. The main objectives of this study were to begin to discern the extent of genetic variation in honeylocust and demonstrate its potential for use as an agroforestry species. Other objectives included the organization of existing literature on the genetics and agroforestry of honeylocust into a coherent, useful form. And finally, to bring together the literature on agroforestry systems for the temperate zone, both to further the awareness of ongoing research, as well as describe some potential benefits derived from their use. CHAPTER I Genetics of Honeylocust INTRODUCTION Linnaeus erected the genus Gleditsia naming it in honor of Johann Gottlieb Gleditsch (1714-1786) (Sargent 1922). The genus Gleditsia, a member of the Leguminosae family, subfamily Caesalpinioideae, includes about fourteen species and one putative hybrid (Tables 1.1; 1.2). Honeylocust, Q. triacanthos L., grows naturally in the eastern half of the United States. It is a minor component in three forest associations: 1) Northern Red Oak - Mockernut Hickory - Sweetgum; 2) Sweetgum - Nuttall Oak - Willow Oak; 3) Sugarberry - American Elm - Green Ash. The first two cover types are edaphic climax associations (Putnam and Bull, 1923). Other common associates of honeylocust are elms, ashes, red maples, blackgum, persimmon, pecan, black walnut, box elder and Kentucky coffee tree. The wood is strong and durable, is used locally for fence posts and railroad ties, and also possesses many desirable qualities such as attractive figure and color, strength, and hardness (Panshin and De Zeeuw, 1970). Honeylocust was first cultivated by American colonists Table 1.1 The genus Gleditsia Family: Leguninosae Sibfamily: Caesalpinioideae Latin name Location References 5;. triacanthos L. Eastern United States 1 g. aquatica harsh. Southern United States 1 g. x £913.99. Sarg. Mississippi Valley, Texas 1 _G_. amorphoides (Gris.)Taub. Southern South America 1 _G. caspica Desf. Caspian Sea, Iran, USSR 1,2 _G_. assamica Bor. Northeast India 1 E. japonica Miq. Japan, Korea, China(PK:) 1,2 E. sinensis Lam. PRC 1,2,3 3. macracantha Desf. PRC l E. microphylla Gordon Central, Northern PRC l E. delavayi Franch. South central PK: 1,2 g. australis Hemsl. Southern PK: 1 E. Leg (Lour.)Merr. Southeastern PRC l _G_. rolfei Vid. Taiwan, Hainan, Viet Nam, 1 Philippines, Oelebes g. melanacantha Tang. & Wang. PK: 4 Taxonanic references: 1) 2) 3) 4) Gordon, D. 1965. A revision of the genus Gleditsia (Leguminosae). Ph.D. Dissertation. Indiana University, 115p. Paclt, J. 1982. On the repeatedly confused nomenclature of Chinese species of Gleditsia (Caesalpiniaceae) . Taxon 31:551. Onei School of Chinese Materia Medica, Szechuan. 1975. What is Gleditsia officinalis Helmsley? Acta. Phytotax. W(W Institute of Botany of the Chinese Acadauy of Sciences. 1972. Iconographia Cormophytun Sinicorum (2 vols.) Press. (Chinese and Latin). Science Table 1.2 Interspecific variation in the genus Gleditsia Growth habit Nunber of recognized species putative hybrids Chromosome nunber Nodules * Leaf morphology Trees or shrubs ; 3—50 meters in height 14 1 2N=28 absent pinnate or bipinnate; 1-8 pairs of pinnae; 6-46 leaflets; leaflets extranely variable in all species Flowers Snall and inconspicuous; polygamo-dioecious; lacking odor; insect pollinated; clustered to unbranched racanes Thorns present or absent; simple to multibranched; l-40cm in length; juvenile trait; arise from supra-axillary or adventitious buds Geographical North and South America; tanperate and distribution tropical Asia; Malay Archipelago i leaflets are extremely variable between and within all species, often in the same plant. Much of the variation is due to environ- mental conditions and the differences between juvenile and adult leaves. All attanpts to delimit taxa on the basis of leaf char- acters alone have failed. more recently has been widely planted as an ornamental replacement for American elm (Harlow and Harrar, 1968). Current interest in honeylocust is in its potential as a multi-purpose agroforestry crop tree for animal and chemical feedstocks. It has become naturalized east of the Appala- chian mountains from Georgia to New England in the East, and north to South Dakota in the West (Little, 1953) (Figure 1.1). Within the natural range of the honeylocust, a large amount of variation exists in both climatic and edaphic conditions. Average annual precipitation varies from 500 mm in South Dakota-Nebraska to 1800 mm in North Carolina. The frost-free period varies from a minimum of 140 days in the northern and northwestern portions of the range, to a maxi- mum of 340 days in the South Central States (USDA, 1941). Honeylocust is a shade intolerant tree with a strong taproot. It achieves its best growth on fertile, moist, alluvial floodplains and can attain a maximum size of 50 m in height and 1.8 m in diameter, but its normal size range is 18-24 m tall and 0.5-1.0 m in diameter (Harlow and Harrar, 1968). Honeylocust will also grow on soils of limestone origin, is resistant to both drought and salinity (Van Dersal, 1938), coppices vigorously when cut, and is hardy in the Great Plains. Little is known about the patterns or extent of genetic variation in honeylocust. Similarly, the potentials for genetic and/or cultural improvement in the growth, form and "f A, E l ‘l' I ' Q. aauafica + _ _7 _ ° g. x texana 'lln - O I. ~ _ - v v V l T Ju- " l \ Figure 1.1 Themathomlocnn. chemistry of honeylocust are unknown. GENERIC RELAT IONSHI P Gleditsia is most closely related to Gymnocladus. Both have polygamous flowers with polysepalous calyces. The tendency toward dioecism is much greater in Gleditsia than in Gymnocladus (Lee, 1976). In both genera a few pinnae of the leaves are often reduced to simple leaflets (Gordon, 1965). The seeds of Gleditsia and Gymnocladus are smooth, with a hard, impermeable testa and large amounts of endosperm surrounding the embryo. The endosperm is composed of thick walled cells filled with galactomannan gums, which are converted into mucilage upon the addition of water (Sayed and Beal, 1958). The presence of copious endosperm is generally regarded to be a primitive character and furthers the suggestion of a close relationship between the two genera (Lee, 1976). Both genera are known for their richness in fruit saponins (aglycones), which are triterpenoid in nature. Saponins in the fruit of both genera are commonly used in China as soap substitutes. The presence of structurally similar triterpenoid saponins in the two genera provides supporting chemical evidence of their close affinity, but are of little value in discerning among the species within each genus (Lee, 1976). Gleditsia and Gymnocladus are known to lack nodule formation in their roots (Allen and Allen, 1936; Grobbelaar, 1964). According to Burkart (1952), this condition is an 10 indication of the primitive character of both genera, particularly when one considers that both have disjunct geographic distributions and abundant fossil histories. PHYLOGENY The Caesalpinioideae is predominantly a tropical subfamily of the Leguminosae. Tropical Africa is regarded as the center of diversity for the subfamily, with tropical America regarded as a secondary center of diversity (Lee, 1976). More than 40 extinct Gleditsia or Gleditsia-like species have been reported in the literature. Fossils of Gleditsia have been located in places outside the present range of the genus (e.g., Europe and Western North America), indicating a much wider distribution of the genus in the past (Robertson and Lee, 1976). No representative of the genus Gleditsia is recognized with certainty in either the Upper Cretaceous or the Eocene. The genus Gleditsiophyllum from the Upper Cretaceous and the Eocene of North America may be the progenitor of the present day Gleditsia triacanthos (Berry, 1923). Berry (1923) reports on a species resembling Gleditsia found in the Oligocene of Europe, and on the existence of number of undoubted Miocene species. Several of these Miocene species have been described and they include a species (g. columbiana) from the state of Washington in the North- western United States (Prakash and Barghoon, 1961), a 11 species from Montana (g. montanese) (Prakash and Barghoon, 1962), a species from Japan (Hu and Chang, 1940), and other species from Europe where they range from Greece and Hungary to France (Berry, 1923). Hu and Chang (1940) recorded Pliocene species in Japan, Europe and Western North America. Berry (1923) recorded a species resembling G. triancanthos, from the early Pleistocene of Kentucky, and an extinct species was found from the interglacial deposits of the Don Valley in Ontario. The Eastern North American species of Gleditsia are more closely related to species occupying similar temperate areas in Eastern Asia, than to one another (Gordon,1965). Li (1952) has pointed out that these two areas, temperate Eastern North America and temperate Eastern Asia, are very similar ecologically. Both are quite old geologically and have remained relatively unchanged since the Paleozoic. They do not appear to have been submerged since the end of the Cretaceous. Li (1952) has interpreted the present isolated and disjunct floras of Eastern Asia and Eastern North America as remnants of a great mesophytic forest that extended over all the Arctic regions in the tertiary. Subsequent geological changes including glaciation have altered and destroyed the floras of many places. The mesophytic forest of the Tertiary in the Northern Hemisphere has survived principally in Eastern Asia and Eastern Northern America. Only scattered relics remain in Southeastern Europe, Western Asia 12 and Western North America. This interpretation fits in well with the present distribution of Gleditsia, especially if one considers the floristic relationships between the species from these two areas. SEXUAL REPRODUCTION The flowers of honeylocust are classified as polygamo- dioecious. Polygamy is defined as the condition of having staminate, carpellate, and hermaphrodite flowers on the same individual (Henderson and Henderson, 1963). In the flowers of Gleditsia, only one type of reproductive organ is usually functional, although a rudimentary or abortive organ of the opposite sex may be present. This phenomenon can be interpreted as incomplete dioecism (Lee, 1976). Although the occurrence of true polygamy is reported in the literature (O'Rourke, 1949; Grisyuk, 1958), individual trees are characterized, in general, as either staminate or pistillate, but perfect flowers are produced on occasion. Therefore, Gleditsia should be considered as a predominantly dioecious genus (Lee, 1976). Pistillate trees rarely produce pollen, but staminate trees frequently produce a few female flowers (Moore, 1948). When considering the entire range of the species, honey locust flowering phenology differs by as much as six weeks. The average flowering date for honeylocust in the southern limit of the range is May 10; in the north it is June 25 (Lamb, 1915). Flowers appear from the axils of the previous years growth when the leaves are nearly fully elongated 13 (Sargent, 1922). This lateness to flower helps prevent frost damage to the flowers and subsequent seed crop (Detwiler, 1947). Flowers are nonshowy and pale-yellow to greenish—yellow in color. The staminate inflorescence occurs in short many-flowered pubescent clustered racemes 2” - 2 1/2" in length. The pistillate infloresence occurs in few-flowered usually solitary racemes 2 1/2" - 3 1/2" in length (Figure 1.2) (Sargent, 1922). Honeylocust flowers are insect pollinated, although in contrast to flowers of the black locust (Robinia psuedo- acacia L.) with which they are sometimes confused, they are inconspicuous and somewhat less fragrant. Honey bees work freely gathering nectar, but usually not enough to create a surplus honey flow (Pellett, 1947). Pistillate flowers become receptive from the base of the raceme toward the tip. On individual trees, the period of maximum bloom (receptivity) lasts 7-10 days. Honeylocust tends toward an alternate bearing habit in which good pod/seed crops are produced every second or third year, with different trees producing good crops on alternate years. The average date of seed ripening varies according to latitude of origin ranging from mid-September to late October (Lamb, 1915). Mature pods begin to drop by mid- September and continue to drop throughout the winter. In good crop years, yield of sound seed per tree can be quite substantial. A good crop can easily exceed thirty b- “ . , '\\ g » vhf/fl %\\3 r} W Vl/fi/ / fii '- k , ,. / A’ \ ., v? // 4% ' \‘N .f/ %: P \ l— . \\ . W .A. x'; T» t r" 2,2 ~ - / . . . , r. \ it o' 2 4f ;l\.~. .‘./ It'ro,v 1“ 1‘ . §.w‘.v‘".v '1, r? i‘,-?:"“ I Figure 1.2 Gleditsia. a-j, g. triacanthos: a, tip of branch with staminate inflorescences--note once pinnately compound leaves on short shoots, X 1/2; b, staminate flower with 4 stamens, X 3; c, carpellate flower--note presence of aborted anthers, X 3; d, same in vertical section-~note perigynous insertion of calyx lobes, petals, and stamens X 4; e, mature fruit--northern latitude phenotype, with little pulp, X l/4; f, seed, X 2; g, soaked seed in cross section, mucilaginous endosperm evenéstippled, seed coat and cotyledons unshaded, X 1; h, embryo from soaked seed, X l; i, supraxillary thorn from branch--note three-pronged thorn, X 1/4; j, adventi- tious thorn from tree trunk, can be up to 40mm in length, X 1/4. 15 pounds of cleaned seed per tree (@ 2,800 seeds/1b.). Seed viability is over 90 percent. Planted from seed, trees begin to bear commercial quantities of seed by age 10 years, with an optimum age of 25 to 75 years (USDA, 1948). Clonally propagated trees from the Tennessee Valley Authority (TVA) tree crops project of the late 1930's- early 1940's such as the Millwood and Calhoun selections, began to bear at age three, bore significant crops by age five, and by age eight, the illwood clone bore a heavy crop, averaging 180 lbs. dry weight per tree (Moore, 1948). These trees also showed a definite tendency toward alternate bearing, although some pods were produced in the "off" years. Pollination Techniques Controlled pollination on a small scale has been successfully accomplished by many individuals (Detwiler, 1947; Gordon, 1965; Santamour, 1976). Standard crossing procedures, which include bagging of immature pistillate flowers and introducing pollen, have proven successful. Male flowers close to anthesis are removed from their inflorescences, brought indoors, kept at room temperature until pollen shedding and are then refrigerated at approximately 20 C. Because honeylocust flowers are insect pollinated, requiring a vector to transmit pollen to the female flowers, pistillate flowers chosen for artificial pollination need only be bagged with cheesecloth to prevent natural 16 pollination before they elongate and become receptive. If staminate flowers on nearby male trees are not yet shedding pollen, the cheesecloth can be temporarily removed and the pollen can be brushed onto a receptive stigma. If pollen shedding is occurring, pollen can be introduced through a syringe. The artificially pollinated flowers are then left bagged in cheesecloth until the staminate flowering has finished, usually within a week to 10 days. Hybridization Evidence of natural interspecific hybridization between Northern American species of Gleditsia, and the fact that the species from North America are more closely related to species occupying similar temperate areas in Eastern Asia than to one another, indicates that interspecific hybridization and crossability are not a barrier to genetic improvement in the honeylocust. Two recognized species of Gleditsia and one natural hybrid are found in the eastern United States. Honeylocust, E. triacanthos, is found throughout most of the eastern United States. Water locust, g. aguatica, is largely ecologically and geographically isolated from honeylocust occurring in swamps, low wet woodlands and the edges of bayous from South Carolina to eastern Texas and northward up the Mississippi and Ohio River valleys to southern Illinois and Indiana (Figure 1.1) (Robertson and Lee, 1976). 17 Considerable discussion regarding the validity of the hybrid g; x texana is found in Gordon (1965). He concludes that g; x texana is in fact a probable hybrid between the two species. The hybrid is located only in areas where the putative parents occur (Figure 1.1), and is intermediate in nearly every morphological character which separates the parental species. The blooming dates of the parents do occasionally overlap allowing cross pollination to occur. Support for the putative hybrid comes from Stebbins (1950), who points out that most morphological differences between species of plants depend on multiple factors rather than on single genes. These multiple factors show relatively little dominance and therefore a hybrid can be expected to be intermediate between parents in nearly every character. Further evidence in support of the putative hybrid comes from Santamour (1977), who reports that leaf flavonoids from mature trees and open pollinated seedlings of G; x texana show that g; triacanthos is definitely involved in parentage of the hybrid. Controlled Pollination and Crossability In 1938 and 1939, the Tennessee Valley Authority (TVA) made controlled pollinations among phenotypically superior, high pod sugar content, honeylocust selections. Attempted crosses were successful and female-male incompatibility was not found (Scanlon 1980). Gordon (1965) artificially crossed g; aquatica x g. triacanthos, and sound seed was produced from the crosses. Interspecific crosses between a 18 Chinese locust, g; melanacantha, and two ornamental staminate cultivars of 9; triacanthos yielded a total of over 100 hybrid pods with an average of 14 sound seeds per pod (Santamour, 1976). Santamour also reported that seed size and seed weight could be significantly influenced by the male parent. Data from Cold and Hanover (1984c) show that 13.5 seeds per pod is an average yield from pods of wild, open pollinated trees, indicating that the controlled pollination caused no reduction in seed set due to compatibility or crossability problems. SEED HANDLING AND NURSERY PRACTICE Collection and Extraction Mature pods can be collected from the ground soon after they drop, by hitting the branches to jar the pods loose, or by clipping the pods from the branches. Seed extraction, storage, germination and nursery practices are covered in the woody Plant Seed Manual (USDA, 1948) and the Hardwood Nurseryman's Guide (USDA, 1976). A few additional notes should be added to enhance the ease and success of handling pods and seeds. 0 After harvest, pods should be stored at or below 0 C. This will prevent fermentation of the pods and, if bruchid seed weevils (Amblycerus robiniae) are present in the pods, it will prevent them from continuing to develop, breed, and spread within the pods. 19 To prepare pods for extraction, place them on trays in a convection/seed drying oven for at least 2 hours at 350 C. This will dry out the succulent pulp surrounding the seed chambers which can clog a hammermill. All seed extraction should be done in a well ventilated area because the dust from the ground pods is very irritating to the sinuses. Dried pods can be put through a hammermill. The resultant pod kibble is placed in a seed tumbler with screens sufficiently large to permit the seed to drop through. Extracted seed can be cleaned in an air blower. Storage and Sowing Honeylocust seed will remain viable for several years if stored dry at 1-40 C. To obtain successful germination of honeylocust seed, they must be scarified and forced to break seed coat dormancy. This can be accomplished with sulfuric acid, hot water, or by mechanical means. To scarify honeylocust seeds with concentrated sulfuric acid, place the seeds into the acid for not less than 60 (maximum 120) minutes. After scarification, rinse the seeds thoroughly. Heit (1967) cites the advantages of sulfuric acid treatment as: 1) Little special or expensive equipment is required; 2) Relatively low cost; 3) It is highly effective; 4) Because there is a great deal of seedlot variation in optimal scarification time, preliminary tests can give the ideal length of acid treatment for each individual seedlot; 5) Following scarification, seeds can be dried, refrigerated, and held for several months without 20 loss of viability; and 6) Seeds can be sown with mechanical seeders because of their unswollen condition as opposed to seeds treated with hot water. To determine if seed weevil larvae are present, place the rinsed, acid treated seeds in a refrigerator overnight. Seeds which have minute entry holes due to weevil larvae will imbibe and should be discarded. The remaining should be viable and either dried and stored or sown immediately. Germination of sound seed should be in the range of 75-90 percent. Seeds should be sown 3/8 - 1/2" deep and if properly scarified, complete germination will occur within 21 days of sowing. ASEXUAL REPRODUCTION In its natural state, honeylocust tends to be an objectionable, thorny nuisance. Interest in honeylocust as a multipurpose pasture/fodder/fuel tree and as an ornamental has led to the development of asexual propagation of thornless genotypes to circumvent the thorn problem. Seed from thornless trees produce 60-80 percent thornless progeny (Chase, 1947; Soutemeyer £5 21., 1944). In lieu of thorns, thornless trees produce short vestigial shoots which are semi-persistent but not objectionable (Stoutemeyer 35 31., 1944). Thorniness exists as a juvenile trait. Trees which will ultimately be very thorny will show this trait in the first season, while trees which will be lightly thorned may 21 not be detected until the second year. As honeylocust trees increase in age thorn production diminishes and ultimately ceases in the upper crown. Typically, trees 10 years or older show a definite thornless region in the upper and outer areas of shoot growth (Chase, 1947). When thorny trees are used for scionwood, the collection of thornless scionwood from branches which have definitely ceased thorn production will produce thornless trees. As expected, scionwood from thornless trees produce only thornless trees (Stoutemeyer 31 31., 1944). Budding and Grafting Many common nursery techniques for budding and grafting are successful in propagation of honeylocust. Due to the geometry of the honeylocust bud, the inverted T-bud technique is used to achieve rapid, simple, vegetative propagation of scionwood. Successful budding can be accomplished with bud from dormant wood collected in the spring or from mature buds in late July or August. Inverted T-budding can result in trees up to 5 feet high in the first year (Stoutemeyer 33 31., 1944). Modified cleft grafts as well as whip and tongue bench grafts will yield up to 100 percent take. However, grafted honeylocust tend to sucker profusely, can be time consuming, and are best suited to small scale operations. Marcavillaca and Garcia (1971) have successfully grafted scionwood from the South American species, 9. macracantha Desf., on to honeylocust stock and achieved over 22 50% take within 45 days of grafting. This interspecific grafting success may prove to be important if varieties of honeylocust are desired which are disease or insect resistant, and adapted to a wide variety of soil types. Also, the development of dwarf varieties would open up new ornamental potentials for the honeylocust (O'Rourke, 1949). Rooting 31 Cuttings Currently, root cuttings are the best method of propagating honeylocust in large quantities and at reason- able costs. The biggest advantage of root cuttings is that large numbers can be obtained from prunings taken from the roots of young nursery trees. For maximum success, root cuttings over 8 cm in length and in excess of 12 mm in diameter should be used. Cuttings should be taken in early spring and planted directly into the greenhouse or nursery. Root cuttings from mature trees sprout less vigorously and should be avoided when possible (Stoutemeyer 31 31., 1944). Marcavillaca and Garcia (1971) reported success in rooting greenwood cuttings taken from a 13-year—old honey- locust tree. Subapical cuttings 30 cm in length, taken 30 cm from the tip of the branch, gave the best results. Dipping the cuttings in 10,000 ppm napthalene-7-acetic acid (NAA) or 10,000 ppm indole-3-butyric acid (IBA) adsorbed on talc powder as a carrier for the hormone gave the best results, with 77 percent and 87 percent rooting success, respectively. Soaking the cuttings for 24 hours in an 23 aqueous solution of 50-100 ppm NAA gave 67% rooting success. Concentrations above 100 ppm appear to be phytotoxic. Stoutemeyer et al (1944) observed that the best material to use for rooting greenwood cuttings originated from stump sprouts or from shoots grown from root cuttings. Hardwood cuttings can also be rooted in the greenhouse but are not useful for large scale production. To propagate "own-rooted" grafted or budded stock it is best to root greenwood cuttings. When the cuttings have well established root systems, root cuttings can be used to propagate new stock. Tissue Culture With significant advances in clonal micropropagation of woody plants in the past decade, tissue culture techniques hold the greatest promise for commercial scale asexual propagation. Micropropagation is essentially an extension of conventional propagation techniques using aseptic culture. The method of organogenesis consists of the micropropagation of explants from a variety of tissues including leaves, shoots, buds, reproductive structures, etc. (Brown and Somer, 1982). Past research has shown that specific tissue culture techniques need to be developed for each individual species. Honeylocust has been successfully cloned via organogenesis using stem tissues and regenerative callus culture (Rogozinska, 1968), and more recently regenerated plantlets have been obtained from shoot tips of honeylocust 24 seedlings (Brown, 1980). This approach may offer the greatest near term potential for commercial production of genetically improved strains of honeylocust. CYTOLOGY AND MUTAT ION Honeylocust is a diploid species, 2N=28, and evidence indicates that it is cytologically stable. Chromosome counts have been made on 6 of the 14 known species of Gleditsia and all were undisputed as 2N=28 (Gordon 1965). Atchison (1949) states that little variation in chromosome size and morphology can be noted among the species. No polyploidy has been detected within the genus. Mutation resulting in morphological variants are rare. Counts of half-sib seedling progeny from 400 families of honeylocust indicate that only seven families (1.75 percent of the population) showed any visible mutations. Gold and Hanover (unpublished) found three types of mutations among the seven families. The first type, total albinism in both cotyledons and true leaves, occurred in four families. Three of these four families had approximately four percent albino progeny. One exceptional family had 26 percent albino progeny indicating that a single recessive gene controlled the inheritance of the mutation. A second type of mutation, found in two families, had green cotyledons and albino true leaves. These mutations occurred in eight percent of the progeny. All of the albino-type mutations died within 90 days of germination. A third type of 25 mutation, showing extremely abnormal growth rate and morphology, survived for two years in the greenhouse. Two families, both originating in Iowa from the same general location, had this type of abnormal progeny. The seedlings had varying intensities of red coloration in the stem and leaves, dark green cotyledons, grew at an extremely slow rate (less than 10 cm. in two years), and had a long, unbranched, nonfibrous root system. In addition, the leaflets failed to open outward but instead folded in upon themselves. GENETICS AND BREEDING Provenance/progeny Testing Provenance tests are used to evaluate the performance of germplasm from many different locations in a common environment, allowing the inherent genetic differences to be observed and measured. Progeny testing allows for a further degree of refinement, including the understanding of the variation and heritability of any given trait within and between families. The size of a species natural range is a principal factor influencing the amount of geographic variability within a species (Wright, 1976). Honeylocust, with a large natural range and generally continuous distribution, has a great deal of genetic diversity and a continuous pattern of genetic variation (Gold and Hanover, 1984a). Significant differences among half—sib families exist in stem dieback, spring flushing, fall growth cessation, 26 thorniness, 2-year stem caliper, and 2-year height growth. The differences were all significant at the 1% level of probability. Additionally, the region x location and family-within-region x location interaction was nonsig- nificant for all traits indicating consistency of results over locations in southern Michigan. Variation 13 Date 3; Spring Flushing Honeylocust follows commonly cited patterns of variation in spring growth initiation in which families of northern origin (IA, NB, SD, IL, etc.) flush first. Families of more southerly origin were intermediate in their flushing date, and those families which were last to flush were all from southern origins (LA, GA, TX, MS, etc.). The range of variation follows a clinal pattern. According to Mather (1953), clinal variation develops after an initial disruptive selection in a base population migrating from the center of origin of a species. This is followed by stabilizing selection and gene exchange among adjacent populations over the species range (Haldane, 1948; Fisher, 1950). Continental glaciation may have caused a major disruption in the natural selection process in much of the flora of eastern North America, including many species of trees which have clinal patterns of leaf flush. The means of the earliest and latest flushing families differed by over 11 days, while 21 days separated the earliest and latest flushing individuals. Negative 27 correlations are found between flushing and latitude of origin (r3 -0.74). The correlation between flushing and frost free days is positive (rs 0.73). These correlations suggest that budbreak is influenced by the temperature distribution patterns at the location of origin. Narrow sense heritability for flushing is h2= .37 (Gold and Hanover, 1984a). Variation 13 Growth Cessation Photoperiod is thought to be the major factor in the control of growth cessation (Nienstadt, 1974). Families of southern origin which have evolved and adapted to mild climates are the last to stop growth in the fall. Northern sources are capable of responding to decreasing summer day- lengths much earlier than southern sources. Differences in sensitivity to a preset critical daylength, i.e. that day- length which triggers the cessation of growth and onset of dormancy, enables northern sources to go dormant at the proper time when grown in northern locations. For sources of southern origin grown in the north, the critical day- length threshold is not reached until late in the summer. This allows them to grow late into the fall causing frost damage to succulent tissues. Two year results of Gold and Hanover (1984) show that those families which originate furthest south (below 35°) from the test sites had the lowest percentage of progeny which had ceased growth by mid-October and the greatest subsequent degree of stem dieback. Sources from inter- 28 o o mediate latitudes (35 -37 ) had a higher percentage of pro- geny which had stopped growing and a less severe amount of stem dieback. Sources originating north of 370 N. latitude had the fewest number of progeny growing late in the fall (none in most cases) and suffered the least severe degree of stem dieback. Growth cessation is very highly negatively correlated with stem dieback (r= -0.88) and frost free days (r= -0.78), and is positively correlated with latitude (r= 0.79). Variation 13 Winter Hardiness (Stem Dieback) Southern Michigan is at the extreme northern limit of the natural range of honeylocust (Figure 1.1). After two seasons in the field, variation in winter hardiness in the form of stem dieback and death from winter injury were strongly evident. Intraspecific differences in winter hardiness of woody plants have been related to climate of geographic origin, latitude of origin, and elevation of origin (Flint, 1974). Differences in cold acclimation of provenances within a species is closely related to phenolo- gical differences within the species (Nienstadt, 1974). With few exceptions, honeylocust families whose source of origin is south of 37° N. latitude suffered at least 10 percent stem dieback. Families originating south of 350 N. latitude suffered between 30 and 80 percent stem dieback. Two families of Southern origin, from Georgia and Louisiana, proved to be outliers and did not suffer significant stem dieback. Early empirical evidence points 29 to other cold tolerant individuals within cold sensitive families. If this holds true, there will be good potential for within family selection for improved cold tolerance in southern sources. Selected individuals would be used for incorporating desired southern traits such as pod size into northern sources. According to Kriebel and Gabriel (1969), one possible explanation for the performance of these families is that relict populations from the Deep South and Mississippi Valley may retain a genetic capacity for winter hardiness normally found solely in trees with northern genotypes. Honeylocust has a sympodial growth habit and does not set a true terminal bud. Rather, that portion of the stem above the false terminal bud will die back before the onset of growth initiation in the spring. This amounts to 0-40mm natural dieback and is not viewed as the result of a lack of cold hardiness. Stem dieback is very highly correlated with active shoot growth late into the fall (r= 0.89), with frost free days at the point of origin (r= 0.77), with cooling degree days (r= 0.79) and with date of spring flushing (r= 0.88). The high positive correlation between stem dieback and fall growth suggests that photoperiod (in an indirect sense) may play a role in the ability of individual sources to go dormant, to attain a sufficient degree of cold hardiness and to resist subsequent tissue damage. Stem dieback is negatively correlated with latitude of 30 origin (r= -0.78) and with freeze days (r= -0.78). Stem dieback is also negatively correlated with 2-year seedling height (r= -0.40) and 2-year seedling caliper (r= -0.49). Narrow sense family heritability in stem dieback based on variance components is h = .96. This high degree of heritability will be of importance in future breeding prog- rams to incorporate some of the important economic traits inherent in Southern sources into cold hardy, Northern sources 0 Variation 13 Thorniness A common feature of wild, open pollinated honeylocust, is the presence of many sharp, 3-branched thorns occurring singly or in clusters. Thorns are considered to be abortive branches which arise from supra—axillary buds on the branches and from adventitious buds on the trunk (Blaser, 1956). Thorns complete development and lignification in one year and range in size from 2-40cm. Thorniness in honeylocust is thought to have arisen as an evolutionary adaptation to exposure to arid environments. Thorn shoots are thought to curtail transpiration loss (Grisyuk, 1959). Thorniness is a juvenile trait and the upper branches of thorned trees 10 years and older can be used as thornless scionwood to create "thornless" cultivars. However, the progeny of these grafted "thornless" cultivars will contain thorny seedlings, and this is highly undesirable. Thornless, open pollinated trees of the variety "inermis", produce 60-80 percent thornless progeny. There is a one to 31 two year lag time between sowing in the nursery and rogueing out the thorny seedlings. If only thornless progeny could be assured, seedling production costs would be reduced. The ability to eliminate the thorn trait through selection and breeding will expedite the widespread use of honeylocust. Additionally, the introduction of genetically thornless honeylocust into areas where it is not found locally would totally eliminate the thorn problem. Addressing this situation, Grisyuk (1959) reports on three years of controlled pollination experiments between thorny and thornless honeylocusts. Crosses were made in all combinations. Results indicate that crossing thornless females with thornless males will produce only thornless progeny (Table 1.3). Table 1.3 Combined three year results of an experiment to determine the inheritance of the thorn trait in honey locust (Grisyuk, 1959). * Hybrid combination Number 31 progeny Thorny Thornless Ratio T-less female x T-less male 0 303 --- T-less female x Thorny male 25 123 1:5 Thorny female x T-less male 136 . 147 1:1 Thorny female x thorny male 75 15 5:1 it All seedlings were examined for thorns for two years. 32 Testing of F progeny from controlled crosses of the F generation willzgive conclusive results on a breeder's agility to produce genetically thornless trees which "breed true". This is because the F generation will segregate out all of the genotype combinatigns. If all F progeny are thornless, a major hurdle will have been crissed in the practical use of the honeylocust. Although data from a rangewide provenance/progeny test show that the degree of thorniness has a moderate negative correlation with latitude (r= -0.57) indicating a general decrease in thorniness from south to north, the goal is absolute thornlessness. IMPROVEMENT PROGRAMS Economics Honeylocust is in an excellent position to benefit from tree improvement efforts. It is rather unique in that its present economic value is derived from its widespread use as an ornamental street tree. An individual ornamental honeylocust cultivar is worth an estimated $10 per foot for a six foot high sapling, or $60 per individual (Levenson, 1984). While recognized as possessing many desirable qualities (Panshin and de Zeeuw, 1970), present use and economic value of honeylocust wood is minor, and very local. Further, its potential use as a multi-purpose agroforestry species shows good promise, but it has no economic value whatever at the present time. Results of an economic analysis of tree improvement research in Michigan indicate that the potential economic 33 return derived from the development of new ornamental cultivars is much greater than the return gained from the development of genetically improved seed for timber or pulp and paper. This is because the unit value of an ornamental seedling is much greater than the unit value of a tree seedling to be planted for timber or pulp and paper (Levenson, 1984). A multi-faceted improvement program can work toward many goals simultaneously. If new cultivars are developed, the returns to tree improvement research will provide economic justification on this basis alone, allowing for other tree improvement efforts to move forward. The initial strategy of assembling a rangewide provenance/progeny test, determining basic patterns of genetic variation, heritibility, general and specific combining ability, progeny evaluation and selection, will apply to all categories of improvement. Because honeylocust has light, lacy foliage, the potential exists for intercropping of forage or vegetable crops in plantations or seed orchards to enhance the economic viability of the early portion of the testing phase by recovering a significant portion of the initial establishment costs. To date, honeylocust has not been included in any long term systematic tree improvement research program. Over 70 "chance" selections of honeylocust have been patented as cultivars (Santamour and McArdle, 1983). The original cultivars were selected 50 years ago for high pod sugar 34 content and for use as stock feed. More recently, selection has been directed toward the development of new ornamental cultivars. Two current tree improvement research efforts, both established in the field in 1982, include a comprehensive rangewide provenance/progeny test at Michigan State University, and a more limited regional provenance test at the University of Nebraska (Walt Bagley pers. comm., 1981). Under natural conditions honeylocust rarely grows in well stocked stands, but rather occurs in a scattered distribution pattern. Plantations of honeylocust, even- aged and regularly spaced, offer the best approach to improvement. Honeylocust are insect pollinated and predominantly dioecious. This combination of factors will tend to insure natural outcrossing and prevent excessive inbreeding. While the breeding cycle for honeylocust is not expected to be particularly lengthy compared with many other tree species, an expected 5-10 year generation time demands that initial selection and gain be derived from the vegetative propagation of superior individuals based on the results from the provenance/progeny testing. Following identification of superior individuals or families in a provenance/progeny test, an improvement program should consist of a controlled breeding program, crossing the most promising individuals. This method, which is preferred to one of allowing open pollination of superior progeny, gives the breeder fullest control of sub- 35 sequently produced hybrids, and maximizes opportunity for genetic gain. After the first round of controlled pollina- tions, the first concrete step forward in the overall genetic improvement of honeylocust will be in those traits, such as thornlessness or winter hardiness, which can be selected for at an early age in the nursery. Based on the particular breeding objective and long term goals, both intra and interspecific hybridization may be necessary. Objectives 33 Honeylocust Improvement Program Due to the very different end-use goals and objectives for which honeylocust is being considered, a breeding program will be multi-faceted in scope. Major objectives include: 1) Development and selection of new cultivars of ornamental honeylocust; and 2) Development of selections for use in agroforestry systems in temperate and highland tropic areas of the world. A minor objective is the selection of fast growing, straight-stemmed trees for sawlogs and veneer. Breeding for Ornamental Cultivars A tree improvement program for honeylocust as an ornamental should concentrate on the following traits: l) Insect and disease resistance, especially resistance to mimosa webworm, an insect which causes severe defoliation in many parts of the range; 2) Faster growth rate; 3) Straight stem form; 4) Thornlessness; 5) Cultivars which are 100% staminate and will not produce pods; 6) Development of additional cultivars with reddish color foliage; 36 and 7) Development of new dwarf varieties. In all cases attention must be paid to winter hardiness. It should be noted that many of these traits have been selected for and improved upon in the existing "named" cultivars. These same traits will continue to be important baseline criteria in any ornamental improvement program. The most serious problem facing honeylocust is the mimosa webworm, Homadaula albizzae Clarke. Since its discovery in the 0.8. in 1942, it has become the most serious insect pest of honeylocust. In a study to identify webworm phagostimulants in honeylocust foliage, Peacock (1967) found that the alkaloid triacanthine, isolated in largest concentrations from immature leaves, deters larval feeding. Resistance in 3. triacanthos to mimosa webworm has not been reported in the literature. The screening of native sources for webworm resistance, should be the first step in a search for resistant geno- types. Should resistance be discovered, an identification of the feeding deterrent would be desirable. Another ap- proach suggested by Santamour (1977) is to screen other species within the genus Gleditsia for resistance. If resistance is located, either in Q. triacanthos or another species, a program of backcrossing and recurrent parent selection would be needed to incorporate resistant genes into 3. triacanthos. As previously mentioned, initial selection for traits such as thornlessness, reddish foliage color, and winter 37 hardiness can be accomplished in the nursery as these traits will express themselves within the first two years. Further development of genetically thornless F selections will be possible when the thornless F progenyzare sexually mature. Selection for stem form and growth rate can begin in the nursery, however progeny testing in plantations for longer periods of time will be needed to accurately select for these traits. Narrow sense heritability for growth rate is estimated to be moderately high, h2= 0.63. Improvement in stem form has already been reported by members of the nursery industry who have released cultivars (i.e. "Green Glory","Shademaster") which are reported to maintain a central leader and are single stemmed (Pirone, 1978; Santamour and McArdle, 1983). Any improvement in growth rate and stem form would also benefit the use of honeylocust as a timber or veneer species. As individuals begin to flower, selection for staminate trees will be necessary. In all cases, due to the high commercial potential of honeylocust, outstanding or unusual individuals will be asexually propagated through root cuttings or T-budding for further testing and evaluation. A long term objective is to make interspecific crosses with exotic species of Gleditsia which are more shrub-like, with the ultimate goal of developing genetically dwarf varieties. Another approach to dwarfism would be through the use of dwarfing rootstocks as is the case in commercial varieties of apple trees. 38 Breeding for Agroforestry Systems While the concept of agroforestry is considered to be a very old idea, the "science" of agroforestry is very new, particularly in the temperate zone, and the development and genetic improvement of multi-purpose species has been rare. Honeylocust has been discussed as an ideal multi-purpose tree since the early days of the century (Smith, 1914). Research conducted between 1934-1947 by the Tennessee Valley Authority and other research stations led to the selection and clonal propagation of "plus trees" with high pod sugar content and sweet taste. As presently envisioned, honeylocust may be of value for the production of an array of chemical and animal feedstocks including ethanol, stock feed and industrial gums (Gold and Hanover, 1984c). Specific traits which need development and breeding to maximize the value of honeylocust in agroforesty systems include precocious flowering, annual bearing, high pod yields and pod carbohydrates, high seed sets and high levels of seed protein and galactomannan gums, cold hardiness, and resistance to bruchid seed weevils. If used in the highland tropics of less-developed- countries as sources of fodder, fuelwood and erosion control, additional traits to select and breed for would include high levels of leaf protein, excellent coppice ability, and well developed root systems. 39 Following the initial establishment of rangewide provenance/progeny testing, the objectives of the two major use categories will be almost 1800 opposed to one another. Selection and improvement for a few traits such as insect and disease resistance, cold hardiness and thornlessness will be desired for any use of the honeylocust (other than as living fences). However, based on past experience with nut trees, selection for precocious flowering, annual bearing, high pod yields, etc. are likely to favor trees with broad, branching crowns and poor overall stem form. The advantage in this two-pronged approach is that the poorest ornamental or timber ideotypes may be superior as agroforestry species. Southern sources of honeylocust are the ones with the largest, heaviest pods, and largest amount of total sugars (Gold and Hanover, 1984c). It may be necessary to hybridize sources of northern and southern origin to capture this type of pod morphology while maintaining winter hardiness. Breeding for resistance to seed weevils will require an approach similar to that pro- posed for mimosa webworm. CONCLUSIONS AND RECOMMENDATIONS Controlled breeding in honeylocust is feasible and will be used in improvement programs to upgrade and combine traits from superior selections. The two seed source studies established to date in Michigan and Nebraska are limited by either the range of sources under test and/or by the number of locations in which sources are being 40 evaluated. More tests are needed across the range of the species and there is an equally strong need for establishment of comprehensive provenance/progeny tests across the southern range of the species. Within each of these areas outplantings should be made at several locations to more accurately assess the effects of provenance on growth, form, cold tolerance, etc. Concurrently, there is a need for the establishment of a germplasm bank of all of the known species of Gleditsia for future use in interspecific hybridization. 41 LIST OF REFERENCES Allen, O.N. and E.K. Allen. 1936. Plants in the subfamily Caesalpinioideae observed to be lacking nodules. Soil Sci. 42:87-91. Atchison, E. 1949. Studies in the Leguminosae. IV. Chromosome numbers and geographical relationships of miscellaneous Leguminosae. J. Elisha Mitchell Sci. Soc. 65:118-122. Atkins, A.O. 1942. Yield and sugar content of selected thornless honeylocusts. Ala. Polytech. Inst. Ag. Expt. Sta. 53rd Ann. Rpt. pp. 25-26. Berry, E.W. 1923. Tree ancestors. Williams and Williams Co. Baltimore. 270 pp. Blaser, H.W. 1956. Morphology of determinate thorn-shoots of Gleditsia. Am. J. Bot. 43:22-28. Brown, C.L. 1980. Application of tissue culture technology to production of woody biomass. Proc. Int'l. Energy Agency. Brighten, England. Oct. 30 - Nov. 1, 1980. Brown, C.L. and H.E. Somer. 1982. Vegetative propagation of dicotyledonous trees. Chapter 5. In: Tissue Culture in Forestry. J.M. Bonga and D.J. Durzan, eds. Martinus Nijhoff, The Hague. Burkhart, A. 1952. Las Leguminosas Argentinas sylvestres y cultivadas. Ed. 2. XV + 569 pp. .Acme Agency. S.R.L., Buenos Aires. (In Spanish). Chase, 8.8. 1947. Propagation of thornless honeylocust. J. Forestry. 45:715-722. Deam, C.C. 1932. Trees of Indiana. Ed. 2, rev. Indiana Dept. Conserv., Pub. 13. 326 pp. 42 Detwiler, S.B. 1947. Notes on honey locust. Soil Conserv. Serv., U.S. Dept. Agr. 197 pp. Fisher, R.A. 1950. Gene frequencies in a cline determined by selection and diffusion. Biometrics. 6:353-361. Flint, H.L. 1974. Phenology and genecology of woody plants. IN: Phenology and Seasonality Modeling Volume 8:83-97. Fowells, H.A. 1965. Silvics of forest trees of the United States. U.S.D.A. Forest Serv. Handb. No. 271. 762 pp. Funk, D.T. 1957. Silvical characteristics of the honey locust. U.S. Forest Serv. Central States Forest Expt. Sta. Misc. Release 23. 14 pp. Gold, M.A. and J.W. Hanover. 1984a. Genetic variation in honeylocust (Gleditsia triacanthos L.): 2-year results. (In preparation). Gold, M.A. and J.W. Hanover. 1984b. Agroforestry for the temperate zone. (In preparation). Gold, M.A. and J.W. Hanover. 1984c. Honeylocust (Gleditsia triacanthos L.): Important chemical characteristics and cultural systems for use in agroforestry systems. (In preparation). Gordon, D.A. 1965. A revision of the genus Gleditsia (Leguminosae). Ph.D. thesis, Indiana University. 114 pp. Grisyuk, N.M. 1958. Polygamy and monoeciousness in Gleditsia triacanthos L. (In Russian). Bot. Zhur. 43(10):l488-1490. Grisyuk, N.M. 1959. The inheritance of thorn formation in honeylocust. (Translation from Russian) Moskovskoe Obshchestvo Ispytatelelg Prirody—Otdel Biologischeskii Byulleten 64(2): 117-122. Grobbelaar, N., M.C. Van Beijma and S. Saubert. 1964. Additions to the list of nodule-bearing legume species. S. Afr. J. Agric. Sci. 7:265-270. Haldane, J.B.S. 1948. The theory of a cline. J. Genetics. 48:277-284. Harlow, W.H. and E.S. Harrar. 1968. Textbook of dendrology. Ed. 5. 512 pp. illus. McGraw-Hill Co. New York. 43 Heit, C.E. 1967. Part 6: Hardseededness - a critical factor. Amer. Nurseryman. 125:10—12, 88-96. Henderson, I.F. and W.D. Henderson. 1963. A dictionary of biological terms. Ed. 8. Van Nostrand, Princeton. 640 pp. Hu, H.H. and R. Chang. 1940. A miocene flora from Shantung Province China. Palaentol. Seneca. N.S.A. 1-141. Kramer, P.J. and T.T. Kozlowski. 1979. Physiology of woody plants. Academic Press. New York. 811 pp. Kriebel, H.E. and W.J. Gabriel. 1969. Genetics of Sugar Maple. U.S.D.A. Forest Service Res. Pap. No. 7. 17 p. Lamb, G.N. 1915. A calendar of the leafing, flowering, and seeding of the common trees of the eastern United States. 0.8. Monthly Weather Rev., Sup. 2, Pt. 1. 19 pp. Lee, Y.T. 1976. The genus Gymnocladus and its tropical affinity. J. Arn. Arb. 57:91-112. Levenson, B.E. 1984. Economic analysis of tree improvement in Michigan. Ph.D. Dissertation, Michigan State University. 180 pp. Li, H.L. 1952. Floristic relationships between Eastern Asia and Eastern North America. Trans. Amer. Phil. Li, H.L. 1974. The origin and cultivation of shade and ornamental trees. University of Pennsylvania Press. 282 pp. Little, E.L. 1953. Check list of native and naturalized trees of the United States (including Alaska). 0.8. For. Serv. Agr. Handb. No. 41. 472 pp. Marcavillaca, M.C. and A.L. Garcia. 1971. Vegetative propagation of a leguminous tree (Gleditsia macracantha) Desf. (In Spanish). Rev. Invest. Agro. Serie 2. Biol. y Prod. Veg. 8(5):211-222. Mather, K. 1953. The genetical structure of populations. Symp. Soc. Exp. Biol. 7:66-95. Mathwig, J.E. 1971. Relationships between bruchid beetles (Amblycerus robiniae) and honey locust trees (Gleditsia triacanEEOS). Ph.D. Dissertation, University of_wis- consin. 144 pp. 44 Moore, J.C. 1948. The present outlook for honey locust in the South. Northern Nut Growers Assn. Ann. Rept. 19:104-110. Nienstadt, H. 1974. Genetic variation in some phenological characteristics of forest trees. In: Phenology and Seasonality Modeling. Volume 8:389-400. O'Rourke, F.L. 1949. Honey locust as a shade and lawn tree. Amer. Nurseryman 90(10):24-29. Panshin, A.J. and C. de Zeeuw. 1970. Textbook of wood technology. Ed. 3. 705 pp. Illus. McGraw-Hill Co., New York. Peacock, J.W. 1967. An investigation of the chemical constituents of honey locust, Gleditsia triacanthos L., as phagostimulants for larvae of the mimosa webworm, Homadaula albizzae Clarke. Ph.D. Dissertation, Ohio State University. 88 pp. Pellett, E.C. 1976. American honey plants. Ed. 5. Dadant and Sons. Hamilton, 111. 467 pp. Pirone, P.P. 1978. Tree Maintenance. ed. 5. Oxford University Press. New York. 587 pp. Prakash, U. and E.S. Barghoon. 1961. Miocene fossil woods from the Columbia basalts of central Washington. J. Arn. Arb. 42(2):165-203. Prakash, U. and E.S. Barghoon. 1962. Fossil wood of Robinia and Gleditsia from the tertiary of Montana. Putnam, J.A. and H. Bull. 1932. The trees of the bottomlands of the Mississippi River Delta Region. U.S. Forest Serv. South. Forest Expt. Sta. Occas. Paper 27. 207 pp. Robertson, K.R. and Y.T. Lee. 1976. The genera of Caesalpinioideae (Leguminosae) in the southeastern United States. J. Arn. Arb. 57(1):1-53. Rogozinska, J.H. 1968. The influence of growth substances on the organogenesis of honeylocust shoots. (In Polish). Acta. Soc. Bot. Pol., 37:485. Santamour, E.S., Jr. 1976. Metaxenia in interspecific honey locust crosses. J. Heredity. 67(3):185-186. Santamour, E.S., Jr. 1977. Flavonoid distribution in Gleditsia. J. Arboriculture. 3(1):14-18. 45 Santamour, E.S., Jr. and A.J. McArdle. 1983. Checklist of cultivars of honeylocust (Gleditsia triacanthos L.). J. of Arboriculture 9(9):248-252. Sargent, C.S. 1922. Manual of the trees of North America (exclusive of Mexico). Houghton Miffin Co. Boston and New York. 910 pp., illus. Sayed, M.D. and J.L. Beal. 1958. A histological study of some mucilaginous seeds. J. Am. Pharm. Assoc. Sci. Ed. 47(8):544-547 Scanlon, D.H., III. 1980. A case study of honey locust in the Tennessee Valley Region. In: U.S. Dept. of Energy. Solar Energy Res. Inst. Tree crops for energy co-production on farms. Nov. 12-14, 1980. Estes Park, Co. Smith, J.R. 1914. Soil erosion and its remedy by terracing and tree planting. Science 39:858-862. Stebbins, G.L. 1950. Variation and evolution in plants. Columbia Univ. Press, New York. Stoutemeyer, V.T., E.L. O'Rourke and W.W. Steiner. 1944. Some observations on the vegetative propagation of honey locust. J. Forestry. 42:32-36. U.S.D.A. 1941. Climate and man. Yearbook of Agriculture. 1248 pp., illus. U.S.D.A. Forest Service. 1948. Woody plant seed manual. U.S.D.A. Misc. Publ. No. 654. 416 pp., illus. U.S.D.A. Forest Service. 1976. Hardwood nurseryman's guide. U.S.D.A. Agr. Handb. No. 473. 78 pp. Van Dersal, W.R. 1938. Native woody plants of the United States, their erosion - control and wildlife values. Wright, J.W. 1976. Introduction to Forest Genetics. Academic Press. New York. xvi and 455 pp. CHAPTER II Honeylocust half-sib provenance/progeny tests: two-year results ABSTRACT Honeylocust, Gleditsia triacanthos L., is a minor associate in many natural forest cover types. Due to a combination of desirable traits including high wood specific gravity, abundant coppicing, a tap-rooted/profusely branched root system, drought tolerance, high carbohydrate pods, and high protein seeds and leaves, honeylocust appears to have potential for use as a multi-purpose tree in agroforestry systems. The potential for genetic and/or cultural improvement in the growth and form of honeylocust is unknown. An in- depth study of the geographic and genetic variation in honeylocust was initiated in 1979 with the establishment of a comprehensive rangewide provenance/progeny test at two locations in southern Michigan. At the end of the second growing season, significant differences among regions and half-sib families-within- regions were found in total height, caliper, thorniness, date of spring flushing, stem dieback, and fall growth 46 47 cessation. Strong negative correlations were found between latitude of origin and stem dieback. The ranges of variation in spring flushing, fall growth cessation, leaf retention, and stem dieback appear to follow clinal patterns. Families of northern origin from the Lake States area are the best overall juvenile performers in terms of total height, stem caliper, cold-hardiness and minimal degree of thorniness. INTRODUCTION Honeylocust, Gleditsia triacanthos L., is a minor component in many forest associations. The wood is strong and durable and is used locally for fence posts and railroad ties, and also possesses other desirable qualities such as attractive figure and color, strength, and hardness (Panshin and De Zeeuw, 1970). Honeylocust has been widely planted as an ornamental replacement for American elm (Harlow and Harrar, 1968). Its ease of production and culture, fairly rapid growth, and hardiness are among the commendable characters that make it popular for planting in parks, along highways, and in yards and gardens (Li, 1974). Current worldwide interest in honeylocust is based on its potential as a multi-purpose agroforestry crop tree for a variety of animal and chemical feedstocks. It has become naturalized east of the Appalachian mountains from Georgia to New England in the East, and from central Texas north to South Dakota in the West (Little, 1953). 48 Within the natural range of the honeylocust, a large amount of variation exists in both climatic and edaphic conditions. Average annual precipitation varies from 500 mm in South Dakota-Nebraska to 1800 mm in North Carolina. The frost-free period varies from a minimum of 140 days in the northern and northwestern portions of the range, to a maximum of 340 days in the South Central States (USDA, 1941). Honeylocust is a shade intolerant tree with a strong taproot. It achieves its best growth on fertile, moist, alluvial floodplains and can attain a maximum size of 50 m in height and 2.5 m in diameter, but its normal size range is 25-32 m tall and 0.8-1.2 m in diameter (Harlow and Harrar, 1968). Honeylocust will also grow on soils of limestone origin, is resistant to both drought and salinity (Van Dersal, 1938), coppices vigorously when cut, and is hardy in the Great Plains where it has been grown successfully in shelterbelts. Little is known about the patterns or extent of geographic and genetic variation in honeylocust. Similarly, the potential for genetic and/or cultural improvement in the growth and form of honeylocust is unknown. The only provenance test of honeylocust, other than the test being reported on here, is a regional provenance test currently underway at the University of Nebraska (Pers. comm. Walt Bagley, 1981). 49 An in-depth study of the genetic variation of honeylocust was initiated in 1979. The growth, morphology, ontogeny, phenology, physiology, and chemistry are being studied. This paper will report on the emerging patterns of geographic and genetic variation in honeylocust based on results at the end of the second season in the field. MATERIALS AND METHODS Honeylocust seed collection requests were sent out in July, 1979. Three groups, including members of the Society of American Foresters Tree Improvement Working Group, State Departments of Natural Resounces or their equivalent, and members of the Northern Nutgrowers Association (NNGA), were contacted by mail (Appendix A). Between August 1979 and February 1980, collections were obtained from 467 individual trees in 26 different states covering a majority of the natural and naturalized range of honeylocust. Upon receipt, each collection was assigned an accession number (Appendix B). This number consists of a genus and species code developed by the Michigan State Cooperative Tree Improvement Program (MICHCOTIP) to standardize record keeping for all genera and species used in the breeding program. 0 All accessed collections were kept refrigerated at l - o 4 C until extracted and measured. From March to June 1980, seed from all pods were extracted by hand, in order to obtain a maximum number of undamaged seed. During, and subsequent to seed extraction, morphological and chemical measurements were recorded on both pods and seeds. Results 50 are reported elsewhere (Gold and Hanover, 1984). Following extraction, seed from individual family seedlots were kept refrigerated until sown. In December 1980, 391 seedlots were sown in a greenhouse at the Michigan State University Tree Research Center. Prior to sowing, all seedlots were scarified in concentrated sulfuric acid to facilitate germination. Seeds were sown in 30.0 cm x 30.0 cm plastic containers, into which 36 paper plant bands (cells) were inserted and filled with a 1:1:1 soil mixture of peat:per- 1ite:vermiculite. Each cell consisted of a plant band 5.0 x 5.0 x 30.0 cm. A randomized block design with six replications and six seeds per replication was used. The containerized seedlings were grown under 16-hour day lengths with artificial lighting. Seedlings were removed from the greenhouse in their containers in late July 1981, placed in an outdoor over- wintering shelter and hardened off for subsequent spring planting. In mid-April 1982, the half-sib families were outplanted as 1-0 seedlings in a randomized block design at 2 locations in southern Michigan. At each location, three blocks were planted at 7'x 8' spacing, in four-tree linear plots. Location one in East Lansing, Mi., covers 5.1 acres with an average slope of 0-3% and sandy-loam to loam soils (Appendix C). Location two, near Battle Creek, Mi., covers 3.9 acres with slopes from 0-10% and has a sandy-loam soil type (Appendix C). 51 In the fall of 1982, total height was measured at the end of the first growing season. Data collection in 1983 was limited to the half-sib families which survived in all blocks at both locations. Fifty-seven of these families, representing a cross-section of the entire range, were chosen for data collection and analysis (Table 2.1). Data were collected on the three tallest trees in all plots at both locations. In the spring of 1983, survival, stem dieback, and date of leaf flush were recorded. In the fall of 1983, thorniness, height, caliper, fall growth cessation and leaf retention were recorded. Dieback was determined by measuring to the highest green portion of the stem. This juncture was clearly de- 1ineated after spring flushing commenced. Trees in all 57 families at both locations were observed every three days beginning May 27, 1983 (day one) to determine the date of leaf flush. This was defined as the day when an estimated 50 per cent of the "buds" had expanded to the point at which small leaflets were distinguishable. In the fall of 1983, the degree of thorniness was scored on a scale from 0, for total absence of thorns, to 4, for very heavy thorniness from base to shoot tips. Growth measurements included total seedling height (cm), and seedling stem caliper (cm) which was measured at 10 cm above the ground. Preliminary observations among half-sib families in the fall of 1982 indicated that growth cessation occurred over a period of more than 60 days. Therefore, in 52 oon oonn nn onno on non oonn nn onno on onn nooo nn oooo on non onoo nn onoo on non onno nn onoo on ooo nonn on moon oo nnn oono on oono on ono onon oo nnon no ono onon oo nnon no ono onon oo nnon no nnn nonn on nono nn nnn nonn on nono nn nnn nonn on nono nn noo noon oo ooon no ooo nnnn no ooon no ooo nnon no ooon no ooo nnon no ooon no ooo noon no ooon no ooo nnnn no oonn no ooo nnnn no oonn no ooo oooo on nono on ooo oooo on nono on ono ooon on nnnn nn ooo noon on onnn nn noo oonn on nooo no ooo nnon no onon on ooo nnnn no onon on ooo nnnn no onon on ooo nnnn no onon on ooo nnnn no onon on .nonn 333 \m moon 5:8. 3638 \m none 58 noon .oz mouooc ocnnoou a com: mwuowo ocnnomo nooooo O cNN common 83 umouu no 593 no oo oo no oz onn no oo oo no oz onn oo oo no no on oon nn oo no no on non nn oo no no on non on on no on n> oon oo no on oo zn nnn on no nn no on oon on no nn nn on non on no nn nn on non on no nn on on oon on no no on on non nn on no nn ox oon oo no on nn um onn on no no on no onn on no no on on onn on no on on no oon on no oo on on non oo oo on nn oz non oo oo nn on m: oon on on on oo on non on on no oo oo non on oo on on no non on on nn on go onn no oo nn on no noo nn no on nn no moo nn no no nn no ooo on no nn on no nno no no on on no nno on no on nn no ono .8 .93 3.3 33m annemn .8330va can nunsmou ucosvwmnnfi on mom: can .33 an Bahama—o mmnnnsflu nnmtunm: hm now mummy :nmnuo H.N 032. 53 .m mo Bonn 339895» 0933 m5 «0 monounmmoc o>nunmom no So \m o .m mo Baum 33869.9 moon—96 on... no cough—amp gnu—wows no So \m. 0 non momo nn oooo no oon oo oo mo no n: ooo non momo nn oooo no oon oo oo mo no n2 noo oon onoo nn oooo no non oo oo oo no on noo oon nooo nn oonn oo omn mo on mo no oz ooo oon onmo on oooo no oon nn oo nn no :0 ono oon onmo on oooo no oon oo oo oo no :0 nno oon onmo on oooo no oon nn oo nn no mo ono onn onoo on mono oo non oo no oo no on ono oon onno on ooon on non no oo oo oo om noo oon oomo on onoo on non on oon no no no ooo onn nnno nn onno on oon on oon on oo om noo nnn onoo on onom oo oon oo no nn no :0 ooo nno oomn no onno mo oon no no mo mo 02 ooo nno moon no onno mo oon no no mo mo oz ooo non oonn no oooo no oon om oo no no oz ooo oon oooo nn oooo mo oon oo oo on no on ooo oon oooo nn oonn oo oon no on on no n2 noo oon oooo nn oonn oo oon no on on no oz noo noo ooon mo oooo oo _ oon oo no om mm on ono oon onoo on nnoo oo oon no no nn no :0 ono oon oooo on onon no oon oo oo no no on noo oon oooo on onon no oon oo oo no no on ooo oon oooo on onon no oon oo oo no no on ooo ooo _ noon oo nnoo om oon om oo nn oo E. ooo non oonn no oooo no oon mm no om no oz noo non oonn no oooo no oon no no om no oz ooo noo nnon no onno oo oon on no no oo mo nno .mono zonon xm.nnmp nucoE umoonoo \m.mnmv .umm common oouu .oc moon .oz 333 ocnnooo .n. com: $38 0:38: nuance amouu mo 50:3 .83 Jon Baum nnnEmon O n.fl.u:oov n.N wnnma 54 the fall of 1983 the phenology of growth cessation was studied by recording the number of individual half-sib progeny in each family which were either actively growing or had ceased growth as of October 3, 1983. In effect, a "snapshot" of growth was recorded. Similarly, leaf retention/leaf drop was recorded on October 27, 1983. Individual progeny which still held more than 25 percent of their leaves were scored as retaining, while individuals with less than 25 percent of their leaves were scored as dropped. For purposes of analysis, the natural range was divided into six regions: Southeast, East-central, Lake States, Northwest, West-central, and Southwest portions of the range (Figure 2.1). A three-level nested ANOVA (Model II) using the combined data set from both plantations was run on 1983 field measurements to test for differences among regidns, families-within-regions, and trees-within-families (half-sib progeny) using individual trees as items. For fall growth cessation and thorniness, a two-level nested ANOVA was used using plot means as items (Table 2.2; Table 2.3). Simple product-moment correlations were calculated between each of fourteen variables, six relating to site of origin, and eight relating to field measurements made in the two plantations (Table 2.4). Family means were used in all correlation analyses. Percent data were transformed using the arcsine square root transformation, and ranked 55 "-5i?‘ *Plantations ( oSeed trees Figure 2.1 Naturalized range of honeylocust and locations of regions, seed trees, and test plantations. 56 Table 2.2 Expected Mean Squares in 3-1evel nested ADDVA used to partition out variance canponents and calculate heritabilities. Source of Degrees of variation freedan Site (8) S-1 Block-within-site B(S) B-l (S) Region (R) R-l Site x Region (S-1)(R-l) B(S) x Region B-l (S) (R-l) Family-within— F-l (R) Region F(R) Site(S) x F(R) (S-l)F-1(R) B(S) x F(R) B-1(S)F-1(R) Tree(N)-within-plot N—l (B) (S) (F-l) Variance components from expected mean squares 1/ Ve+NVb ( s) r+NBVsr+NRVb (s) +NBRVs Ve+NVb (s) r+NBVsr+NRVb (s) Ve+NVb(s) r+NBVsr+NBSVr Ve+NVb (s) r+NBVsr ‘ Ve+NVb(s)r Ve+NVb(s) f(r) +NBst(r) +NBSVf(r) Ve+NVb(s) f(r) +NBst(r) Ve+NVb(s) f(r) Ve 1/ R,F,S,B,N represent the regions, families-within-region, sites, blocks-per-site, and trees-within-plot, respectively. Ve, Vb(s)f (r) , st (r) , Vf (r) are variances due to tree-within-plot, fanily at block- within-site, family x site, and family respectively. S7 .ucmOnMncmnmncoz m: .nm>mn on on» no ncounnnconm nnnoonnonuoom on .no>wn no on» no nonunMncmnm nnannumnumum a. a- oonnn oooo.o .. ocon ocoon ooo oannosoo oo.o oonn on.o oo.o on noo oon n “coon ocoo.o ocnonn moon.o mono.o moon ocono no noonowHVEmo o oon o.no.o oonoon .oon.o .oon.o «ooo «.nonn no noonomnonnnEoo oo.o oonn . on.o no.o nn nnon .\ on n nouno ocno.o oconn mann.o mano.o ocnn ocnonn o conomo x ooo o.nn.n oooooo o.no.n o.on.n «oono «onoono \ o oonomo no.o oonon nn.n nn.o nno noon o Aconuooonoomo o:nn.o .«ooooo ocno.o moon.o .oonon .oooo n conuooon Inuuuuuntulnunullonmsvm camel:IuIIIIIIIIIIIIIIIII unonmn nomnnmo £u3ono noon“ xomQOno Eooownw conuMnnm> mmocncuons Hmonnn ummnln nnoo moon Eoum mo moonmoo mo ounsom .cmanOnz anonusom an moonumOOn 03» um umou acomonm\oocmco>onm onnsmmcmn m Bonn omnnnEmm nnmumnmn no so <>oz¢ poumoc nm>0nuomunn o.n onnms 58 .mmOn no o no an a Eoencns a nun: nnoo mo Hones: coo: .m mo sown M oomuo>m on» no mmuouummoo o>nunmom mo Eom \W .Cnmnuw no venom no common moquumOnu on» uo Samson oomuo>¢ \W .ooucomoum mum mucononmuooo conuMnouuoo cmEumomm \W .nm>on on onu no ucmonunconm mum omuonn mconuMnouuoo nno W” m on.o III III III III III III :nzono nnoo III oo.o no.o III III III III nomnnmo > no.o nn.o oo.oI no.oI oo.o III III noonu wannmo oo.o oo.o oo.oI oo.oI no.o oo.o III xumnono Eoum oo.o oo.o IIa III no.o oo.o oo.o conumunmnOOum 8.? 3.? II- {- oo.o- on.o.- on.o- Ionoo guano on.o nn.o oo.oI no.oI no.o on.o on.o \monmn omwoooIoCnnooU on.o on.o .i- III oo.o nn.o nn.o Ioooo mono-noon... III III no.oI no.oI III III III \0 mosunocoo on.oI on.oI III III no.oI on.oI on.oI ocsununn I I I I I I I I I I I I I I I I u I I I I I I I I I I I I I I I “mama “war... “may momma --M..-- drum mum“ ------- \m moon \mnnmm HmonIn umonIn nomencuosa mannam Eoum onuonuouomumno \M .uoooonnocon mo munuonuouomumco nusouo can nooannonmuoe .nmonmonococm co oconumnonuoo ucoEoEIuoocOum onmeno o.n onnma 59 data were subjected to a scalar transformation. (Little and Hills, 1978). RESULTS AND DI SCUSS ION Patterns of geographic variation can be derived from the results of the ANOVA and correlation analyses. Results of the 3-leve1 nested ANOVA show that trees from different regions differed significantly in all traits analyzed (Table 2.3). At the end of the second growing season in the field, the mean height and caliper at the East Lansing site were 104.5cm and 1.01cm, respectively. At the Battle Creek site, mean height and caliper were 90.8cm and 0.96am, respective- ly. For the 57 families used in the analyses, survival was equal at both locations, averaging 87 percent. Variation in Phenological traits Honeylocust follows the spring growth, fall dormancy patterns common to many other species with large natural ranges (Kriebel, 1957; Sluder, 1960; Bey, 1972). Trees from southern origins, which are adapted to mild climates, are the last to set buds in the fall. Low temperatures are often a major factor in limiting plant distribution because sources native to warm regions cannot often be successfully grown in colder regions. They do not harden off fast enough to survive early cold weather, never develop sufficient hardiness to cold temperatures, lose their hardiness 60 prematurely and thus are usually damaged or killed by sub- freezing temperatures (Kramer and Kozlowski, 1979). Studies by Campbell and Sorenson (1973) on phenology and frost damage in Douglas-fir showed that southern sources generally set buds later than northern sources. For provenances that set bud in the same week, southern sources were more frost sensitive than northern sources, with the proportion damaged increased by four percent for each degree of latitude. Each additional week which bud set preceeded frost, the proportion of frost damaged seedlings decreased by approximately 25 per cent. Southern Michigan is at the northern extreme of the natural range of honeylocust (Figure 2.1), and after two seasons stem dieback and death from winter injury are strongly evident. Stem dieback is very highly correlated with active shoot growth late into the fall (r= 0.89), with frost free days at the point of origin (r= 0.77), with cooling degree days (r= 0.79) and with date of spring flushing (r= 0.88). The high positive correlation between stem dieback and fall growth suggests that photoperiod may play a role in the ability of individual sources to enter dormancy (Nienstadt, 1974), to attain a sufficient degree of cold hardiness and to resist subsequent tissue damage (Table 2.4). Stem dieback is negatively correlated with latitude of origin (r= -0.78),with freeze days (r= -0.78), and with the 0 mean number of days with a minimum temperature of 32 F. or less (r= -0.78). 61 It is also negatively correlated with 2- year seedling height (r= -0.40) and 2-year seedling caliper (r: -0049) (Table 2.4). Based upon the above observations, perhaps the most critical trait to consider in the initial stages of provenance/progeny testing of honeylocust in northern areas, is the degree of winter-hardiness. The presence or absence of winter hardiness is reflected in the varying degrees of severity of stem dieback and mortality. As expected, families originating in northern areas above 40.5 latitude suffered little or no dieback. north The region showing the least overall dieback is the Northwest region, covering the Northern Plains states which have the greatest tempera- ture extremes and harshest climate (Table 2.5). Table 2.5 Variation among regions in traits analyzed. indicates the location of each region. Regions‘l/ Trait SE EC LS NW WC SW Stem dieback (%) 27.9 14.2 4.2 3.0 5.5 51.1 Leaf flush (days) 8.6 7.3 4.4 4.8 4.8 10.0 Fall growth (%) 2/ 42.8 20.9 2.6 0.0 3.9 60.4 Caliper (cm) 1.09 1.03 1.02 0.98 0.97 0.76 Height (cm) 102.9 103.1 102.2 96.0 93.4 82.8 Thorns (%) 3/ 83.4 62.5 32.0 44.9 66.7 67.5 1/ Values reported are means for each region. Figure 2.1 2/ Fall growth (%) represents the number of progeny in any given region which were actively flushing. 3/ Thorns (%) represents the number of progeny in any given region which were thorny. 62 Families from the Lake States region also proved to be hardy. Additionally, the West-central region encompassing the central Plains States between 350-37o north latitudes, was the only region in which families originating south of 370 north latitude proved winter-hardy under Michigan conditions. This is likely due to frequent exposure in this region to severe weather extremes in spite of a long growing season. With the exception of the West-central region, honeylocust families whose source of origin is south of 370 N. latitude suffered at least 10 percent stem dieback. The least winter-hardy region proved to be the Southwest portion of the range. Families from this region suffered an average dieback in excess of 50 percent of total height. In general, families originating south of 350 N. latitude suf- fered between 30 and 80 percent stem dieback (Table 2.6). Two families of Southern origin, from Georgia and Louisiana, proved to be outliers and did not suffer signifi- cant stem dieback. Field observations point to other cold tolerant individuals within cold sensitive families. If this pattern continues into the future, there will be good potential for within family selection for improved cold tolerance in southern sources. Selected individuals would be used for incorporation of desired southern traits, such as high levels of pod carbohydrates, into northern sources. According to Kriebel and Gabriel (1969), one possible ex- planation for the performance of these families is that relict populations from the Deep South and Mississippi 63 oon oo.n oo.o on.o on.o non oo oo nn oo zn nnn no oo.on oo.o oo.o on.o oon on oon nn oo oo non no no.o oo.o nn.o nn.o oon on oo oo no no ooo no no.nn oo.o oo.o no.o oon on on on no oz ono oon nn.on oo.o oo.o no.o oon on oo oo on on non oo oo.on no.o nn.o oo.o oon oo no oo on o: non oo no.on oo.o no.o on.o oon on oo oo no no noo oon oo.on no.o oo.n nn.o oon oo oo no no on oon no no.on oo.o oo.o oo.o non on no oo no on noo oo no.o oo.o oo.o no.o oon nn oo on no mo ono oo no.nn oo.o oo.o no.o oon no no on no mo oon no nn.on oo.o no.o oo.n onn on no nn no on oon no oo.o oo.o oo.n oo.n oon no no no no on non oo no.o oo.o on.n nn.n oon nn on nn no oz non oon no.no oo.o oo.o on.n onn nn no on nn no ooo oo oo.nn oo.o on.n on.n oon nn oo on no oo ono oo no.nn oo.o oo.o oo.n non nn oo no no on non oo no.on on.o on.o oo.n oon on oo on no mo nno oo no.n oo.o no.n on.n oon nn on nn no oz non oon no.nn oo.o no.n no.n oon no oo oo no oz onn oo no.o oo.o on.o oo.n oon nn no no no on non no oo.on oo.o nn.n oo.n non nn oo no no on non no no.on oo.o on.n oo.n oon no oo nn on on noo oon no.nn oo.o on.o no.n oon oo no oo on oz onn oon no.n oo.o no.o oo.n non no oo on oo oo noo oo no.n oo.o on.o oo.n nnn on oon no oo oo oon oo oo.nn oo.o no.o on.n oon nn oo nn on on oon oon no.on oo.o oo.o on.n oon no oo oo no oz onn oon no.o no.nn on.o on.n non oo no nn no mo oon oon oo.on oo.o on.n on.n oon on oo on no on oon oon no.on oo.o on.n oo.n oon nn no no no on oon o o oonncmomu n ouzouo guano nooowno moon 9: oz .oc .3333 HMS nnmon on name gm mouwlunouh .chn $9.533 33m. onnno-non _\m \m. \fnm \m- .conmflnomnc can mnmnnmcm on 8m: 8:25 nno-nno: no noo 38 335.6 Bo $8386 .528 on onnon. 64 .ooon .nn nonouuo so 853 nnon» wo ”.anqu ouoe no on 00530... no?! onnoao nno-nno: :08 on ofioonn no none... none... on» B n38 on n oonncflwn no.8 \m .noon.n nmoouuo co ocnnoono onm>nnno nnnaou anuImnm: nnon-.0 an newmonm uo ”585: n33 m5 3 nmsvm on a 530nm nnoo \m . amnion: 3 ummzon ooo-nu .xoononc Baum “.50qu nnon... mo nnomn o5 co noxcon mum mmnnneflm \M .82 on R5828 58 Bonn noon-n soon 28... nnonmoo 8o 8»ng 38 \o no no.no oo.on oo.nn on.no onn oo no nn nn on non no no.no oo.on on.n nn.nn oon on no no on no onn oo nn.oo on.on no.nn nn.oo onn on oo on on no non no no.no oo.on on.nn nn.no onn oo oo nn nn oz onn no on.oo on.oo on.o oo.oo onn oo no nn nn on oon oo no.no no.oo oo.on no.oo onn oo no nn no on nno nn nn.oo nn.on oo.on nn.oo oon on no no on no nnn oo no.no no.oo on.on oo.on onn oo oo nn nn o: nnn no no.no nn.oo oo.o no.on onn nn no no nn on ooo oon no.no oo.oo oo.nn oo.on onn no no on nn no nno oon no.no nn.no nn.o oo.on onn on no on nn on ono oon oo.no no.oo on.nn no.on onn on no no on no onn no oo.no nn.on oo.o on.nn nnn oo no on nn no onn oo nn.oo no.oo no.o no.on onn no no oo on nz oon oo no.no nn.oo nn.o no.on non on on no oo on non on no.on oo.on nn.o nn.on onn nn no no on no onn no no.no no.on oo.o oo.nn oon oo no nn on zn oon oon no.on no.no on.o on.on onn no no oo on nz oon oon on.nn nn.on no.o no.on onn on no oo on no non oon oo.no nn.on oo.o no.nn onn oo no nn on no oon oo no.no nn.o on.o nn.nn oon on no oo on on nnn oo no.on on.on oo.o oo.on oon nn oo oo nn no oon oo no.on oo.o oo.o on.on non oo nn nn oo on non no no.no nn.o oo.o no.o onn on nn no on o> oon oon no.on oo.n oo.o nn.o non oo no on no on ono no no.on no.o oo.o oo.n oon no oo no no oz oon oon no.on oo.n oo.o oo.o nnn on no nn on no onn o o connomnmn o no36no guano noonmno mono oz oz .0: noonzoo I .33 non B moon 23o mono-omen .ooon «Bunyan 83o nnnfio \o \m. oon no.nn-.03 o.~ Onnnma 65 Valley may retain a genetic capacity for winter-hardiness normally found solely in trees with northern genotypes. Narrow sense family heritability for stem dieback based on variance components is h = .96. This high degree of heritability will be of important in the future breeding program which will try to incorporate some of the important economic traits inherent in Southern sources into cold hardy, Northern sources. Similar patterns of regional variation are found when comparing the fall growth cessation patterns with stem dieback patterns (Table 2.6). Very few progeny from families originating in winter-hardy regions were actively growing when scored in October, 1983. In fact, there were no individuals in any family from the Northwest region which were still actively growing by that date. In contrast, 60 percent of the progeny in families from the cold intolerant Southwest portion of the range were actively growing in chober, 1983. In general, families originating south of 35 north latitude, with frost-free seasons at their point of origin in excess of 220 days, proved to be the least cold tolerant (Table 2.6). In the fall, honeylocust drops its leaves over a very short interval. Leaf retention shows a large amount of variation within regions and this may indicate quantitative gene regulation of leaf fall. In general, a definite trend exists toward more rapid leaf fall in families from northern regions. Leaf retention is positively correlated with fall 66 growth cessation (r= 0.76) and stem dieback (r= 0.83), and negatively correlated with latitude (r= -0.78) and freeze days (r= -0.80). Leaf retention was more highly correlated with precipitation at point of origin than any other trait (r= 0.63). An interesting feature of leaf retention is that one of the two southern "outliers", family number 055 from Georgia, which performs like northern families in all other traits, showed a high degree of leaf retention similar to other families from the same latitude. Based on the regional data, it would appear that selection pressure for or against leaf retention is very small, allowing for a great deal of variation within large regional areas. Honeylocust also follows common patterns of variation in spring growth initiation. Families of northern origin (IA, NB, SD, IL, etc.) flushed first, families of more southerly origin were intermediate in their flushing date, and families which were last to flush were all from southern origins (LA, GA, TX, MS, etc.). The range of variation in spring flushing appears to follow a clinal pattern (Table 2.6). The earliest and latest flushing families differed by over 11 days, while the earliest and latest individual progeny differed by 21 days. Negative correlations were found between flushing and latitude (r= -0.74), and the correlation between flushing and frost free days was positive (r= 0.73) (Table 2.4). These correlations suggest 67 that budbreak is influenced by the temperature distribution patterns at the location of origin. These data agree with the conclusions of Burley (1966a; 1966b) in his study of Sitka spruce. Burley found that when spring flushing is viewed in relation to the nature of the spring temperature distribution at the point of seed origin, a systematic pattern of flushing can be observed among seed sources. Stem dieback is highly correlated with spring flushing (r= 0.85) and the relative performance of these two char- acteristics are highly dependent on latitude and temperature of origin. Regional Variation The amount of variation accounted for by regional differences varied widely among the traits analyzed (Table 2.7). The variation in phenological traits accounted for by regional differences is larger than the variation accounted for by families-within-regions. This is because the strong effects due to site of origin and genetic background of the various families have already clearly expressed themselves in stem dieback, leaf flushing, and fall growth/dormancy. Effects of origin and genetic back- ground have not become fully evident in growth traits by age two. Regional variation patterns in height and stem caliper at age two do not follow the distinct patterns shown for phenological traits. Families from the East-central region were tallest and had the largest stem caliper, followed very 68 Table 2.7 Variation accounted for among regions, families within regions, and trees within families (expressed as percent of total phenotypic variance), and heritability values 3/. Trait Regions Families Trees Heritability family single tree 2/ Stem dieback 36.5 16.0 27.1 .96(.17) .82(.04) Leaf flush 16.9 11.0 25.1 .35(.15) .34(.O3) Fall growth 42.4 25.4 -- -— -- Caliper 7.3 14.8 30.9 .45(.18) .32(.07) Height 5.0 20.3 24.4 .63(.l9) .65(.ll) Thorns 19.5 55.8 -- -- -- a/ Heritability values calculated by method of Wright (1976). Variance components are derived from the expected mean squares in the analysis of variance. 2/ Numbers in parenthesis represent standard error values. Vf Family heritability = ----------------------------------- v /NBS + v /ss + v /s + v e fb fs f 4(Vf) Single tree h2 = ----------------------------------- V+V +V +V e fb fs f . N,B,S = The number of trees-within-plot, blocks, and sites, respectively. Ve,Vfb,st,Vf = Within-plot variance, error variance, family x site variance, and family variance, respectively. 69 closely by the Southeast and Lake States regions, respectively (Table 2.5). Considering the data on stem dieback, these early growth patterns are not expected to continue into the future. Personal field observation reveals that families from the East-central and Southeast regions regrew very vigorously from the point of dieback and therefore were able to equal or exceed the northern families by the end of the second growing season. However, as the trees continue to age, the ability of trees from less cold tolerant regions to "catch up" to northern families each year may diminish over time. Cold-hardy families from the Lake States region and even the slower growing West-central and Northwest regions will add incrementally to their total height each year and eventually surpass families of southern origin, particularly those originating below 370 N. latitude. At age two, families from the Southwest region were the only ones to exhibit a marked decrease in growth due to lack of cold-hardiness. Many of the progeny in these families died back to the ground level. Negative correlations between stem dieback and height (r= -0.40), and dieback and caliper (r= -0.49) are already in evidence. These correlations are expected to increase as dieback causes families of southern origin to fall further behind Northern families over time. A common feature of wild, open pollinated honeylocust, is the presence of many sharp, 3-branched thorns occurring 7O singly or in clusters. Thorns are considered to be abortive branches which arise from supra-axillary buds on the branches and from adventitious buds on the trunk (Blaser, 1956). Thorns complete development and lignification in one year and become extremely hard, range in size from 2-40 cm, and can be dangerous and difficult to work with (Harlow and Harrar, 1968). Thorniness in honeylocust is thought to have arisen as an evolutionary adaptation to exposure to arid environments. Thorn shoots are thought to curtail transpiration loss (Grisyuk, 1959). One goal of the honeylocust project is to determine the degree and heritability of the thorn trait, and to develop thornless selections. Thorniness is a juvenile trait and the upper branches of even the thorniest trees 10 years and older can be used as scionwood to create "thornless" cultivers (Chase, 1947). The progeny of these grafted "thornless" cultivars are genetically "thorny“ and will contain thorny seedlings, which is highly undesirable. Open pollinated progeny from the thornless "inermis" cultivar produce 60-80 percent thornless progeny. The ability to eliminate the thorn trait through breeding will expedite the widespread use of honeylocust. Additionally, the intro- duction of genetically thornless honeylocust into areas where it is not found locally would totally eliminate the thorn problem. Grisyuk (1959) reports on three years of controlled pollination experiments between thorny and thornless 71 honeylocusts. Crosses were made in all combinations. Results indicate that crossing thornless females with thornless males will produce only thornless progeny. Testing of F progeny from controlled crosses of the F generatiog should provide the basis for producing genetically thornless trees. If all F progeny are thornless, a major hurdle will have been crossed in the practical use of honeylocust. Regional differences in thorniness are present, with the Southeast region showing the highest percent of progeny which were thorny, over 80 percent (Table 2.6). A general decrease in number of thorny progeny per region is evident from south to north, with the Lake States region having the least thorny progeny, 32 percent. The negative correlation between latitude and thorniness is moderately high (r= -0.57). Family Variation Family-within-region differences were highly significant for all traits analyzed (Table 2.3). For stem caliper, the family-within-region component accounted for twice as much of the variation as the region component, 14.8 vs. 7.3 percent, respectively (Table 2.7). In terms of total height growth, the family-within-region component accounted for over 20 percent of the variation, four times greater than the region component. The narrow-sense family heritability for height at age two is estimated to be 2 2 h a .63, while family heritability for caliper is h = .45. 72 Two-year height and caliper were highly correlated (r= 0.82). Both characteristics also show a negative corre- lation with longitude, (r= -0.43) for caliper and (r= -0.42) for height. The effect of a relatively high heritability for height growth, coupled with a much larger component of variation among family-within-region than among regions, points to a good potential for capturing genetic gain through selection at the family level. Coefficients of correlation between l-year and 2-year heights are high (r= 0.78), and between 1-year height and 2-year caliper are (r= 0.71). Calculated single tree heritabilities are h2= .65 for height and h2= .32 for caliper. At the tree-within-family level, the individual half- sib progeny accounted for a large percentage of the variation in both height and caliper, 31.0 and 24.5 percent respectively. If this large amount of tree-within-family variation maintains itself over time, it would permit further genetic gain at the within-family level. The high percentage of variation due to tree-within-family may be due in large part to residual planting effects, initial seedling size differences, age effects, and strong early influence of microsite differences. The magnitude of these differences are expected to decrease as the age of the trees increase. This will reflect a truer picture of the tree-within-family contribution to the overall variation. 73 The plantation x region interaction, as well as plantation x family-within-region interaction, were nonsignificant for all traits (Table 2.4). This indicates consistency of results over locations in southern Michigan. However, if the rangewide test had been planted at more diverse locations throughout the range, significant location x region and location x family-within-region interactions would likely occur as families would be expected to perform quite differently in more southerly locations. Significant differences were found between the two plantations for height, days-to-flush, and stem dieback. No significant differences were found between plantations for caliper, fall growth and thorniness (Table 2.3). The mean 2-year height at the East Lansing plantation was 115 percent greater than at the Battle Creek plantation. Based on 1983 measurements and subsequent data analy- sis, the most promising families in terms of winter- hardiness, height growth, and caliper growth were compared to the best local family (Table 2.8). Because the data are based on results at age two in the field, some of the families which suffered significant levels of dieback were able to resprout with enough vigor to show up in the top 5 family rankings. As previously mentioned, it is expected that these less hardy families will fall behind the rest of the more winter-hardy families in overall height, diameter and survival as plantation age 74 increases. Therefore, families suffering more than 10 percent stem dieback were excluded from the top five ranking. Based on the data, selection on the basis of Table 2.8 Performance of the 5 best honeylocust half-sib fam11ies at age two at two locations in southern Michigan, based on height, caliper, and stem dieback g/. Family State Survival Height Caliper no. % % of best local source mean b/ 420 IL 100 119 (131)“ 104 (1.22) 300 VA 92 100 (110) 104 (1.22) 055 GA 100 102 (113) 99 (1.16) 448 M1 92 100 (110) 100 (1.17) 272 IN 100 110 (121) 97 (1.13) Overall mean values (098) (0.99) g/ Excludes families in which average stem dieback exceeded 10 percent of total seedling height. 2/ Numbers in parenthesis are actual family means. Units shown are in centimeters (cm). latitude or provenance alone may prove to be an inadequate measure of performance. The data also indicate the presence of variation for hardiness within region. CONCLUSIONS eneral recommendations for the selection of superior half-sib families would be to choose families originating north of 40.50 N. latitude, and generally favor the central portion of the range. Due to the great deal of geographic and genetic diversity found in honeylocust, opportunities for genetic improvement are excellent. Empirical observations on stem form, coupled with the measured 7S variation in thorniness, ultimately will lead to the use of progeny from specific individuals which exhibit an array of positive attributes including high survival, height, diameter, stem form and genetically derived thornlessness. 76 LIST OF REFERENCES Bey, C.F. 1972. Leaf flush in black walnut at several midwest locations. Proc. 19th N.E. Forest Tree Improvement Conf. pp. 47-51. Blaser, H.W. 1956. Morphology of the determinte thorn- shoots of Gleditsia. Amer. J. Bot. 43:22-28. Burley, J. 1966a. Genetic variation in seedling development of Sitka spruce, Picea sitchensis (Bong.)Carr. Forestry 39:68-94. . 1966b. Provenance variation in growth of seed- ling apices of Sitka spruce. Forest Sci. 12:170-175. Campbell, R.K. and F.C. Sorenson. 1973. Cold-acclimation in seedling Douglas-fir related to phenology and provenance. Ecology 54:1148-1151. Chase, S.B. 1947. Propagation of thornless honeylocust. J. of Forestry 45:715-722. Gold, M.A. and J.W. Hanover. 1984. Honeylocust (Gleditsia triacanthos L.): A multi-purpose tree for agrofbrestry systems. (In preparation). Grisyuk, N.M. 1959. The inheritance of thorn formation in honeylocust. (Translation from Russian) Moskovskoe Obshchestvo Ispytatelelg Prirody-Otdel Biologischeskii Byulleten 64(2):117-122. Harlow, W.H., and E.S. Harrar. 1968. Textbook of dendro- logy. Ed. 5, 512pp., illus. (McGraw-Hill Co. New York. Kramer, P.J. and T.T. Kozlowski. 1979. Physiology of woody plants. Academic Press. New York. 811 pp. Kreibel, H.E. 1957. Patterns of genetic variation in sugar maple. Ohio Agr. Expt. Sta. Res. Bull. No. 791. Spp. 77 Kreibel, H.E. and W.J. Gabriel. 1969. Genetics of Sugar Maple. U.S.D.A. Forest Service Res. Pap. No. 7. 17p. Li, H.L. 1974. The origin and cultivation of shade and ornamental trees. University of Pennsylvania Press. 282pp. Little, E.L. 1953. Check list of native and naturalized trees of the United States (including Alaska). U.S. For. Serv. Agr. Handbk. No. 41. 472pp. Little, T.M. 1978. Agricultural experimentation: Design and analysis. John Wiley and Sons, New York. 350pp. Nienstadt, H. 1974. Genetic variation in some phenological characteristics of forest trees. In: Phenology and Seasonality Modeling. Volume 8:389-400. Panshin, A.J., and C. de Zeeuw. 1970. Textbook of wood technology. Ed. 3. 705pp., illus. McGraw-Hill. New York. Sluder, E.R. 1960. Early results from a geographic seed source study of yellow poplar. USDA For. Serv. S.B. For. Expt. Sta. Res. Note No. 150. 2pp. U.S.D.A. 1941. Climate and man. Yearbook of Agriculture. 1248pp., illus. Van Dersal, W.R. 1938. Native woody plants of the United States, their erosion-control and wildlife values. U.S. Dept. Agr. Misc. Pub. No. 303. 362pp., illus. Wright, J.W. 1976. Introduction to Forest Genetics. Academic Press. New York. xvi + 455pp. Chapter III AGROFORESTRY SYSTEMS FOR THE TEMPERATE ZONE ABSTRACT The term "agroforestry" refers to land management systems which involve trees, agricultural crops, and domestic animals in any or all combinations. The combina- tions may be either simultaneous or staggered in both time and space. The historical development of a permanent agriculture system based on the use of agroforestry in the temperate zone is traced. The reasons for a renewed interest in agroforestry include the end of cheap, subsidized fossil fuels; increased concern about soil erosion and marginal land use; an international awakening as to the dangers of indiscriminant use of pesticides, herbicides and other chemicals; and a need to continuously increase food production to meet growing population demands. Three agroforestry management systems are reviewed which currently appear feasible for implementation in many of the industrialized countries of the temperate zone. These three systems include: 1) Animal grazing and intercropping under managed coniferous forests or plantations; 2) Multi-cropping of agricultural crops under 77d 78 intensively managed, high value hardwood plantations; and 3) Woody/woody intercropping involving nitrogen fixing woody plants. INTRODUCTION Two of the more pressing problems facing mankind in the 1980's are shortages of food and energy. From a production standpoint, our attempts to feed the worldhs population have been successful. Modern agriculture now produces more food per acre than at any other period in history. To meet the food shortage challenge, a highly mechanized, fossil fuel dependent, centralized production system has been developed. Farmers rely on fossil fuels for planting their crops, for pesticides, herbicides, harvesting, processing and transpor- ting of our food. Major problems have arisen in this food production scheme in the past decade. The cost of fossil fuel, the backbone of our modern food production system, has increased from under $2.00 bbl. to over $30.00 bbl. Many farmers are finding it difficult to afford the direct and indirect costs of the fossil fuel needed to maintain this large, energy intensive system. Our governmental, industrial, and acade- mic institutions are now looking to the food itself, in the form of ethanol, as an alternate source of energy to replace fossils fuels. Concurrent with the reality of expensive fossil fuel energy, farmers are becoming totally dependent on the food export trade to sell their crops. The grain embargo of the 79 U.S.S.R in 1979, after the invasion of Afganistan, led to the development of a massive farm surplus and depressed commodity prices. The multi-billion payment-in-kind (PIK) program of the federal government resulted from a need to decrease farm output and raise crop prices. A third major problem in the agriculture sector is that of soil erosion. In 1976, the average annual loss of topsoil from agricultural land in the U.S. was approximately 12 tons per acre. The annual fertilizer (N-P-K) losses amount to more than 50 million tons, worth about $7 billion (Pimentel st 21., 1976). The solution to these major problems will certainly be multi-faceted and will include the development of alterna- tive, less energy-intensive technologies, improved soil conservation practices, and more efficient, diversified farming systems. One important facet of the technological solution to these problems may lie in the field of "agroforestry”. Known variously as agri-silviculture, farm forestry, forest farming, tree crops, 3-D forestry, and taungya, this integrated farming system offers many new opportunities and advantages in solving the energy, food, and soil erosion problems. Agroforestry is an interdisciplinary approach to systems of land use, different from the sum of its two major components, agriculture and forestry. It refers to land management systems involving many interdependent components including trees, agricultural crops, and domestic animals in 80 any or all combinations. The combinations may be either simultaneous or staggered in both time and space (Lundgren, 1982). Agroforestry might be considered as the meeting point for a confluence of disciplines, both applied and basic in nature. Within its broadest scope it draws on the accumulated knowledge of many separate disciplines. It draws on forestry, agronomy, animal husbandry and horticulture for its major inputs, with necessary additional inputs coming from soil science, microbiology, ecology, plant breeding, chemistry, economics, sociology, agriculture engineering, and others. Implicit within the concept of agroforestry systems is the idea of using trees in nonconventional ways. This will demand a rethinking of the design, architecture and role which trees will play. The development of different agroforestry systems will be required for each individual locality based on existing biological, economic and political constraints. A review of the literature reveals numerous agro- forestry systems currently being researched. Some of these systems will entail only slight modifications of current practices, while others will require a more radical change. It is outside the scope of this paper to present a detailed review of agroforestry research in the less developed coun- tries (LDC's), however it should be noted that the bulk of the current research into agroforestry systems is being 81 conducted in the LDC's (Bene st 31., 1977; Huxley, 1983; Anon., 1983; MacDonald, 1981). While many potential agroforestry systems have been proposed, this review will be limited to three systems which appear close to practical application and implementation. A list of the temperate zone agroforestry systems reviewed in this paper include: 1) Systems of animal grazing and intercropping under managed coniferous forests or plantations. 2) Multi-cropping of agricultural crops under intensively managed, high value hardwood plantations. 3) Systems of woody/woody intercropping, involving nitrogen fixing woody plants. Another system, using multipurpose trees for energy fuels, chemicals, fiber, animal feed, and soil stabilization, is also considered to have a great deal of potential merit. An in-depth review of this topic is in preparation. The purpose of this review is to introduce the concept of agroforestry to foresters and hopefully to expand conven- tional thinking on the uses and potential uses of trees. Some ongoing work in the industrialized countries of the temperate zone will be highlighted. Historical development The idea of an agriculture based on trees was first outlined in the Uni by J. Russell Smith, an economic geog- rapher at Columbia University (Smith, 1909; Smith, 1911). 82 In a lifetime of travel and scientific observation, Smith documented the destructive results of erosion following cultivation of hilly, marginal lands. In his travels to the Mediterranean, Smith observed many examples of a permanent, tree-based agriculture, on steep rocky terrain (Smith, 1950). In the Mediterranean agriculture, chestnuts (Castanea spp.), oaks (Quercus spp.), carob (Ceratonia siliqua), olive (Olea europa), and figs (Ficus spp.) all provided a variety of agricultural and economic products to the people of that region. Smith proposed North American counterparts to each of the crop trees including nut trees (Carya spp., Juglans spp.), oaks (gercus spp.), persimmons (Diospyros spp.), mulberries (Morus spp.), mesquites (Prosopis spp.) and honeylocust (Gleditsia triacanthos) (Smith, 1914; Smith, 1950). As early as 1914, Smith was advocating the ideas which were "rediscovered" in the 1970's under the guise of agro- forestry. His ideas included interplanting crop trees with woody legumes coupled with animal grazing to gain maximum benefits from a given site and expand the area of useable lands. He advocated the use of tree crops for human and animal food, economic gain, for improving soil stabilization and increasing soil fertility, and for microclimate amelioration. He encouraged the search for additional candidate cropping trees, and the subsequent breeding of cropping trees to maximize their potential for producing food and wood (Smith, 1914). 83 Smith was also aware of the resistance of our political and agricultural leaders to a rethinking of our food produc- tion system. Therefore, to accomplish this, he proposed the establishment of a privately funded tree crop/hillculture research center for long term, uninterrupted research in to a permanent agriculture based on tree crops (Smith, 1950). Concurrent with the early writings of Smith, prelimin- ary research was underway on potential tree species suitable for agroforestry. Forbes (1895) and Garcia (1916) documen- ted the feed value of the mesquite (Prosopis juliflora) growing in the arid southwest U;S. Walton (1923) documented the chemistry and feed value of honeylocust and mesquite. Like Smith, these early tree crop researchers felt that the native vegetation merited serious consideration as multi- purpose, well adapted crop trees. The onset of the Great Depression in the 1930's brought along massive unemployment and the "Dust Bowl" of the Great Plains. This motivated the UQS.Igovernment into a temporary rethinking of our agricultural policies. Along with the large scale shelterbelt planting in the Plains States, a series of hillculture/tree crop projects were established in the eastern U.S. The focal point of the research was the Tennessee Valley Authority (TVA). The TVA began tree crops research in 1934 to provide for reforestation and proper use of marginal lands in the Tennessee Valley. This area of the U.S. was characterized by extensive areas of eroded and poorly utilized land, much 84 of it steep and unsuited for conventional agricultural use. A system of land management was proposed which would simul- taneously protect the soil, prevent soil erosion, and yield a cash crop. The TVA research efforts concentrated on the black walnut (Juglans nigra), but also included Chinese chestnut (Castanea mollisima), filbert (Corylus), hicko- ries, persimmon, and honeylocust (Hershey, 1935; Zarger, 1956). Other projects were located at Virginia Polytechnic Institute, the Alabama Polytechnic Institute (Zarger, 1956), Auburn University (Moore, 1948), and in Ohio (Smith, 1942). During the 1940's, the tree crops idea surfaced in other areas of the world. Eardley (1945) discussed the suitability of carob, mesquite, and honeylocust as supple- mentary fodder for livestock in southern Australia. Loock (1947) and Jurriaanse (1973) describe many fodder trees useful as stock feed to farmers in South Africa, where large areas were often stricken with drought. In 1947, a publica- tion by the Imperial Agricultural Bureaux detailed the uses and misuses of trees and shrubs as fodder throughout the British Commonwealth. Chemical composition and digestibili- ties were listed for over 800 species. Additionally, the concept of "protein pastures", in which trees and shrubs are used as fodder, windbreaks, shade trees and soil condi- tioners, is suggested (Anon., 1947). Schreiner (1959) and Huguet (1979) mention a polyculture system in Italy, "coltura promiscua", in which pollarded poplars are used as vine supports, and also provide fuel and lumber. The 85 fertile soils support a 3- and occasionally 4-story culture of poplars, grapes, dwarf fruit trees, with an annual crop or forage species at the base. The early 1950's heralded a post-war economic boom. Cheap fossil-fuels, herbicides, fertilizers, new farm machi— nery and economic prosperity dominated the thinking of American agriculturalists. One result of this economic climate was that the tree crops projects died a sudden death. According to Zarger (1956), the tree crops projects "had to be abondoned because the cooperating institutions needed the land to meet building program needs." Moore (1948) states, "Hillxnalture went under in June of 1947, and the Horticulture Department took this work over, and they thought they could not support the honeylocust pasture program in Hillculture, and the plot, of course, was pulled out and planted in peaches." Jurriaanse (pers. com., 1979) states "The main reason why I dropped the work (fodder tree research) in 1951 was because of lack of support”. This was largely due to the fact that there existed a controversy between Agriculture and Forestry as to which department should assume responsibility for this work, neither of them really being interested because no quick results were expected." The tree crops idea was all but forgotten in the 20-25 year period from the late l940's-early 1950's until the late 1960's-ear1y 1970's. Four major factors played a role in the renewal of interest in the "tree crops" concept. First, 86 the environmental and ecological concerns of the late 1960's resulted in the banning of numerous herbicides and pesti- cides widely used in our agriculture and forestry sectors (Eckholm, 1976). Second, the extremely high productivity achieved by agriculture in the post-war industrialized world was based on the abundant supply and heavy use of subsidized, low-cost energy (Hirst, 1974; Pimentel 35 21., 1973). 'The oil embargo of 1973 forced a re-evaluation of input-output costs of our farming systems. The ratio of energy used for each food calorie produced doubled in thirty years (Steinhart and Steinhart, 1974). Third, great concern was again surfacing by the mid-1970's on the effects of continued soil erosion, and the possible dire consequences it held for the U.S. food production capabilities (Carter 1977, Pimental, 1976). Fourth, the awareness of the ever- increasing size of the worldls population meant that world food producers would have to continue to increase their output. These events created a search for alternatives to fossil fuels for chemical feedstocks. This precipitated a renewed look at the potential role of trees as one component in the overall solution. In 1971, a prescient paper was written in which species of the genus Alggg, nonleguminous fixers of atmospheric nitrogen, were recognized as having the potential for use in forest management systems in a similar way to that of legumes in agriculture (Tarrant and Trappe, 1971). The first industrialized country to seriously test the idea of a 87 combined forestry/agriculture system of management, ie. agroforestry, was New Zealand. In New Zealand, conflict over land usage between agriculture and forestry, a need to revitalize the rural economy, and the desire to increase timber exports generated an intense interest among researchers to find a way to solve these diverse problems. The idea of "farm forestry" was put forward as a potential solution to many of these problems (Knowles, 1972; Barr, 1973; Olsen, 1974; Farnsworth and Male, 1975). It was determined that widely spaced radiata pine (Pinus radiata), planted at wide initial spacings for maximum growth, could be harvested for sawlogs on 20 year rotations. This lead to the further idea of grazing animals in between the trees, fully utilizing the resultant lush understory. Thus, the idea of purposely managing and integrating foresty and agriculture in a two-tier system was re-introduced to western industrialized countries. Up to this time, grazing in forested lands and forest plantations was always considered to be in conflict with or at best an ancillary benefit to the timber crop, but was not a part of a rigorous management systems (Adams, 1975). A combination of the ideas coming out of New Zealand and the energy/environmental/population/food crises of the mid-1970's led many others to a rediscovery of the ideas of J. Russell Smith. Douglas and Hart (1976) published a book restating and expanding on many of Smith's ideas. Following that, articles advocating the concept of agroforestry in one 88 form or another were written by Cumberland (1976), Farmer (1976), Hills (1977), MacDaniels and Lieberman (1979), Gordon and Dawson (1979), Felker and Bandurski (1979), Borough (1979), Tustin 33 31. (1979), Williams (1980), Spurgeon (1980), and others. All of these reviews articulate a need to rethink our agriculture and forestry systems to provide solutions and help alleviate many important local, regional, national, and worldwide problems. Managed conifer sawlqg/grazing systems In New Zealand and Australia forest managers are combining pasture management with open pine plantations. The purpose behind this integration is to diversify from animal husbandry to the more profitable forestry for wood export without losing the advantage of regular income potential provided by animal husbandry. These systems allow high light levels to reach the forest floor, resulting in heavy understory growth which can reduce stand access and increase fire hazard. The concept of deliberately managing this understory for profit by the grazing of animals has been developed into the full integration of agriculture and forestry. Both animals and pastures may derive benefit from the presence of trees. Animals are able to maintain their body temperature with less energy loss in the modified climate associated with the open tree stands (Farnsworth, 1975). Through their deep root systems and litter fall, trees can tap moisture and cycle nutrients not available to surface- 89 rooted grasses. Herbaceous plants capable of fixing atmospheric nitrogen can be used to enhance soil fertility and increase the combined productivity of forest and grazing lands. The combination of the nitrogen fixation abilities of pasture legumes with the phosphate releasing powers of tree mycorrhizae may also benefit both trees and pastures (Knowles _e__t_ 31., 1973). The initial interest in the agroforestry concept came from the New Zealand timber industries who quickly appreciated the advantages of early financial returns from agriculture, easier stand access, reduced fire risk, and simpler (though more closely planned and monitored) stand management. Conventional forestry in New Zealand is known to be a profitable alternative to agriculture but is often unattractive to farmers. Agriculturalists in New Zealand now acknowledge the role of tree crops in diversifying farm production, reducing market and biological risk factors, promoting soil stability, ameliorating microclimate, and making fuller use of farm labor during slack periods, all while maintaining acceptable stock carrying capacity (Tustin, 35 31., 1979). A simulation model has been developed, SILMOD, to compare volumes and present net worth (PNW) of radiata pine at final stocking rates of 100, 200, and 400 stems per hectare. ‘While a final stocking rate of 100 stems per hectare gave lower timber yields, the overall profitibility was highest at this stocking rate. Results from the model 90 indicate a high degree of compatibility between agriculture and forestry in New Zealand hill country (Knowles and Percival, 1983). In Australia, most of the agroforestry research is being conducted in the western part of the country (Borough, 1979). Additional research is ongoing at New South Wales, Victoria, Queensland, and Tasmania. Most of the research has involved plantations of radiata pine underplanted with subterranean clover (Trifolium subterraneum). Several of the trials include species of Eucalyptus (Batini, Anderson and Moore, 1983). The advantages of agroforestry in Australia are similar to those found in New Zealand. However, in western Australia the greatest potential for agroforestry is viewed as the development of efficient systems for the control of stream salinity in catchment areas, and the control of soil salinity levels on farms (Anonymous, 1978). In the upland areas of Great Britain there is a need to diversify production in order to improve farm vitality. While agroforestry has not been attempted in the uplands area, the National Farmers' Union of Scotland has recommended the development of a New Zealand type of system. The advantages of agroforestry over conventional forestry are considered to include intermediate returns, shorter rotations, and simpler management, and an end-product (timber) which can be sold when convenient to the farmer (MacBrayne, 1982). 91 Suggested species for agroforestry in Britain include Scotch pine (Pinus sylvestris), inland (U.S.A.) provenances of lodgepole pine (Pinus contorta) and western larch (Larix occidentalis). Light crowns, adaptation to dry, firm soils, and a deep rooting habit should enable species such as these to withstand stock trampling and resultant root damage. Conservatism is thought to be the biggest obstacle to the inception of agroforestry in Great Britain (MacBrayne, 1982). In the southern U.S.A. the potential for combined production of timber, livestock, and wildlife is unequaled compared to any other region of equal size in the U.S. The region contains over 80 million hectares of forest land with roughly half considered useable as forest range for live- stock (Shiflet, 1980). To date, few examples of successful integrated management exist in the southern Udi, but many situations involving damage from uncontrolled numbers of livestock with little or no management can be found (Pearson, 1983; Adams, 1975). The key to success in multiple—use management is the maintenance of a careful balance between forage and animals. Grazed firebreaks, nine meters or more wide, are one practical way to integrate forest trees and improved pastures (Halls 35 31,, 1960). Pearson (1982) reported on twenty years of research into cattle grazing, slash pine regeneration and growth, and economics. He found that southern pines appear highly resistant to grazing damage, 92 that cool season exotic grasses grown under pine stands can provide a source of green forage during the winter when native grasses are dormant, and that pine regeneration can be successful within grazed subclover pastures. The sum of the economic returns from multiple products such as food, fiber, and wildlife are expected to be larger than from a single output. Equally important, the increase in land management flexibility is a key factor in the survival of poor markets for any single output. Major challenges affecting forest grazing management in the South include the careful planning and development of intensive systems of grazing management compatible with pine regeneration. Other challenges include the design of economical livestock supplemental feeding regimes including the use of improved forages for winter grazing, and the creation of an atmosphere of information exchange to attain social acceptance of multiple-use management (Pearson, 1983). The objective for integrated land use in the interior regions of the northwest UkS. and southern BritiSh Columbia is to increase the sum total of production from all resources on each hectare of land. The Douglas-fir and ponderosa pine zones comprise the main multiple-use areas in this region. Livestock also graze other forest types within the region including grand fir (Abies grandis), western white pine (Pinus monticola), subalpine fir (Abies lasiocarpa),Iwestern larch and lodgepole pine. Grazing 93 management studies support the use of the interior Douglas- fir and Englemann spruce/subalpine fir zones for producing both trees and grass (Mclean, 1983). For integrated management in this region the most critical factors in determining tree—cattle compatibility are careful monitoring of the degree of forage utilization, and the length of time and season in which forage is utilized. Results of a study using sheep as a silvicutural management tool in the coast range of Oregon, suggest that both brush suppression and acceptable levels of animal production are obtainable. This can be accomplished through the use of a grazing system of light to moderate utilization of clearcuts in the spring, followed by heavier use in areas targeted for brush reduction in the summer and fall. Under this system damage to Douglas-fir is expected to be minimal (Sharrow and Leininger, 1983). Multicropping high value hardwoods with agricultural crops The deliberate intercropping of agricultural crops with high value tree crops is a practice which can be traced back over 100 years to Burma. In a system which became known as "taungya", agricultural crops such as sweet potatoes, and cotton were interplanted with teak (Tectonia grandis). The function of the agricultural crop was to enable the local population to farm a piece of land and get the benefit of the crop in return for weeding and tending the teak in the critical early years of the rotation (Blanford, 1958). 94 The advantages of dual cropping include: 1. More intensive use is made of the land. The area between the tree crop, formerly kept free of competition by cultivation or herbicides, is now made use of by an agricultural crop. 2. More acres of high quality land, often closer to processing facilities and markets, can be brought into fiber production. 3. Early returns from the agricultural crop will offset all or part of the establishment costs for the tree crop, greatly improving the return on the investment. 4. The benefits to the agricultural crop derived from tillage, fertilizer and weeding also benefit the tree crop. The two genera which have received the majority of attention in this type of system are Populus spp., valuable for rapid growth in short rotations and used mainly for pulp, and Juglans nigra, the high value black walnut, grown on long rotations for sawlogs and veneer. In Italy, a system of tillage and intercropping during the first four years after the establishment of the tree crop, is commonly practiced in conjunction with poplar stands planted at a 6 m x 6 m spacing. Maize is produced the first year, and legumes and grain crops are grown for the next three years. Poplar stands with both tillage and intercropping yield a higher economic return than stands with tillage alone (Sekawin and Prevosto, 1973). In Australia, a sequential combination of vegetable cropping and grazing is beingzused in conjunction with widely spaced poplars (6 m x 6 m). In the first two years vine crops such as melon and squash are planted. The vines provide a quick crop and cover the ground to restrict weed 95 growth. At the end of the second year, permanent pasture is sown beneath the poplars, and cattle are then grazed within the plantation. The prunings from the poplar are used for cattle feed (Anonymous, 1978). The Crown Zellerbach Corporation is also experimenting with intercropping in their cottonwood (Populus deltoides) plantations on the Mississippi delta. Soybeans and cotton are interplanted between rows of cottonwood and are heavily fertilized. The cottonwood is harvested for pulp on 10 year rotations averaging 23 cm in diameter (dbh) and 25 m in height (Pers. comm. P. Weber, 1983). In northern Alabama, tree growth in a two-year-old sycamore (Platanus occidentalis) plantation significantly increased during four years in which clover and vetch were grown within the trees (Haines, Haines, and White, 1978). Dual cropping with Populus is also being practiced in Ontario. Corn (Zea mays) and soybeans (Glycine soja) are being cropped between rows of planted hybrid poplars on Indian reserves (Mergen and Lai, 1982). In eastern Ontario, corn and potatoes (Solanum tuberosum) have been grown successfully between 3 m x 6 m spaced poplar during the first three years of the rotation. Potato yields of 12,000 kg/ha have been obtained in the second year of the rotation. Many other crops have also been successfully grown in the first three years of the project (Raitanen, 1978). Researchers in the United States Forest Service at Carbondale, Illinois, are currently hmvolved in attempts to 96 find the proper cover crops to interplant with black walnut in order to enhance the growth of the walnut, utilize the cover crop for hay or fodder, and reduce the use of herbi- cides and other cultural methods used to control weed compe- tition around the trees. .A silvicultural-economic model constructed by Kincaid 33 31. (1982) indicates that the degree of profitability from an investment in the production of black walnut is directly related to the level of manage- ment intensity. The model considered five different manage- ment regimes ranging from walnut timber alone, to a multi- crop management system of timber, nuts, soybeans, winter wheat, fescue and grazing. Intercropping with field crops yielded the highest economic returns on the highest quality site (SI-80), while management regimes using forage crops yielded the highest economic returns on sites of inter- mediate quality (SI-65). Early returns from agricultural production offset the higher initial cost of walnut estab- lishment and yielded a substantial increase in profit. The analysis concluded that multi-crop management offers the‘ greatest returns due to more intensive land use. Roth and Mitchell (1982) studied the effects of selected cover crops on the growth of black walnut. It was determined that clean cultivated black walnut was significantly larger at age six than walnut growing in any mixed planting system. However, they concluded that the energy necessary to maintain the clean cultivation coupled with the increased potential for soil erosion makes the 97 clean cultivation system less desirable for long term management. Underplanting black walnut with leguminous or grass covers can effect positive changes on the phenology of the walnut. Compared with clean cultivation, the maintenance of leguminous or grass covers in walnut plantations may delay bud break 6-12 days, thereby decreasing possible frost damage. Also, underplanting walnut with leguminous winter annuals was found to accelerate the onset of dormancy (Van Sambeek and Rink, 1982). Van Sambeek and Rietveld (1982) co-established plots of black walnut with leguminous cover crops. Their results show that the seeding of plantations with cool season legumes, both with and without chemical weed control around the walnut seedlings, can accelerate tree establishment and tree growth in intensively managed plantations. This indicates that the planned establishment of leguminous cover crops may be superior to allowing plantations to revegetate naturally. Systems pf woody/woody intercropping with N2 fixing woody ‘xplants Another agroforestry system which has recently been receiving attention consists of intercropping between woody species to maximize overall yields from a given site. In all cases, one of the intercropped species will be a symbiotic nitrogen fixer. With the exception of black locust (Robinia psuedoacacia), tree lupine (Lupinus 98 arboreus), and a handful of other temperate zone species, most leguminous trees and shrubs are located in the tropics (Allen, Gregory and Allen, 1955). The group of nitrogen fixers considered to have the greatest potential in the temperate zone are non-leguminous actinorhizal woody perennials. Most of these plants occur in temperate regions or in the highland tropics (Dawson, 1983). Intercropping systems may be either simultaneous or sequential (rotated) cropping systems. The concept of "intercropping vigor", in which dry matter production of mixed cultures exceeds that of pure plantings is now receiving serious attention in the U.S. Good reviews can be found in Tarrant and Trappe (1971), Gordon and Dawson (1979), and Dawson (1983). The major benefits derived from the use of woody nitrogen fixers include the realization of optimum biomass yields per unit of land area, a reduction or elimination of the need for applied nitrogen fertilizer, improvement in soil fertility and soil physical properites, suppression of soil pathogens, and the improved growth of associated species in mixed cropping systems (Tarrant and Trappe, 1971). The awareness that intercropping with nitrogen fixing‘/ trees will benefit the associated woody crop is not a new one. The beneficial effects of black locust on the growth of associated tree species has been observed on many occasions. Ferguson (1922) and McIntyre and Jeffries 99 (1932) presented evidence that catalpa (Catalpa speciosa) growing in association with the black locust showed increased diameter and height growth. Chapman (1935) reported that the heights and diameters of certain trees decreased significantly with increased distance from black locust stands. Chapman and Lane (1951) made a study of the growth and survival of hardwood trees growing in association with black locust, shortleaf pine (Pinus echinata) and sassafras (Sassafras albidum). The best survival and growth rates occurred in the association with black locust. Finn (1953) found that yellow poplar (Liriodendron tulipifera), black walnut and black cherry (Prunus serotina) showed a significant increase in both height growth and total foliar nitrogen when interplanted in black locust stands. Ten year results of a study using European black alder (Alnus glutinosa) as a nurse crop on mine spoils in Kentucky showed that height and diameter growth of hardwoods and pines are accelerated when interplanted at appropiate spacings with European black alder. Nitrogen fixation by European black alder increased foliar nitrogen of the interplanted species (Plass, 1977). Long rotation mixed cropping systems using nitrogen fixing trees as nurse plants usually require that the nurse crop must be harvested, poisoned, or removed before the final harvest of the timber crop. However, the benefits may outweigh this inconvenience. Funk 23 $1,, (1979) found that mixed plantings of the nitrogen fixing tree autumn 100 olive (Elaegnus umbellata) stimulated the growth of black walnut. After 10 years, walnut trees grown with autumn olive were 80% taller and 104% larger in diameter than those grown alone. Additionally, the mixed plots were higher in soil nitrogen, lower in soil moisture, and had lower soil and air temperatures. In a study of a 27 year old Douglas-fir/red alder (Alnus rubra) admixture, Tarrant (1961) found that in addition to increased height growth, the form of the Douglas-fir trees was improved. Total wood volume in the mixed planatation was more than twice that of pure Douglas- fir plantations. Atkinson £5 31; (1979) examined the feasibility of using red alder as a rotation crop with Douglas-fir. Four sequential cropping systems were analyzed for comparison to a system of continuous cropping of pure Douglas-fir. A net-worth analysis indicated that all the systems are profitable, though systems involving red alder were not as promising as those involving only Douglas-fir. He concluded that expanded markets for red alder, increased efficiency of small tree harvesting, or higher costs of nitrogen fertilizer could tip the balance in favor of alternate cropping systems. Binkley (1983) studied pure and mixed natural stands of Douglas-fir and Douglas-fir/red alder on sites of high and low fertility. Compared to the pure stand, the presence of red alder on the low fertility site (SI-25 m) increased the average diameter of Douglas-fir. Inclusion of alder biomass 101 increased the total stand basal area and basal area growth 2.5 fold. Total ecosystem biomass doubled and net primary production tripled when alder biomass was included. In contrast, on the high fertility site (SI-45 m) total ecosystem values were identical between pure and mixed stands, and Douglas-fir biomass and net primary production decreased. He concluded that admixtures of red alder and Douglas-fir have great potential for increasing Douglas-fir growth and ecosystem production on infertile, nitrogen deficient, marginal sites, but have little value on fertile, nitrogen rich sites. Another intercropping system with great potential is one in which nitrogen fixing shrubs are used as nurse crop plants in the early years of a fiber or timber rotation. Perceived silvicultural advantages to the use of nitrogen fixing shrubs include: 1) Elimination of the problem of nurse crops overtopping the main crop; 2) Elimination of competition for the same area of the canopy (photosynthetic surface); 3) Shrubs never need to be removed, harvested or poisoned; and 4) They may be suitable as fodder for grazing animals. Harrington and Deal (1982) have recently advocated the use of Sitka alder (Alnus sinuata) as a nitrogen fixing shrub for use on sites of low nitrogen or organic matter content. The early slowdown in height growth, coupled with its low profile, makes the Sitka alder suitable for use with Douglas-fir in mixed stands. 102 A prime candidate for biological forest fertilization in the southeastern Coastal Plain (USA) is wax myrtle (Myrica cerifera). This naturally occurring shrub is capable of growing well on acid soils in the understory of pine flatwoods. In a study of nitrogen fixation in slash pine plantations, wax myrtle was shown to fix substantial amounts of nitrogen. It is believed that the use of a substantial wax myrtle understory could contribute a significant amount of additional nitrogen to semi-mature slash pine stands (Premar and Fisher, 1983). Another study compared the growth of pitch (Pings rigida) and Japanese black (Pinus thunbergii) pines in association with clumps of the nitrogen fixing shrub bayberry (Myrica pennsylvanicah. Significantly greater height growth within bayberry patches occurred only in the young pitch pines (Tiffney and Barrera, 1979). Marrs gt al=_(1982), studied tree lupine, (Lupinus arboreus), as a nurse crop. Their results indicate that tree lupine could be a very valuable nurse crop for amenity plantings or on marginal lands where the nitrogen status of the soil is low. Advantages of using tree lupine as a nurse crop include rapid establishment and growth, and its natural tendency to die back after 5-7 years, thereby eliminating long term site competition and overtopping problems. The final intercropping system to be considered consists of short rotation,-intensive culture systems for energy, chemical feedstocks, or animal feedstocks, in which 103 nitrogen fixing trees may be used as an equal component of the final harvest. Mixed plantations of alders and poplars that take advantage of nitrogen-fixing trees are among the most intensively studied silvicultural systems. Both alders and poplars have wood properties which have proven to be acceptable for chip and fiber products (Dawson, 1983). Hansen and Dawson (1982) demonstrated that the height of 3-year-old hybrid poplar grown in short rotation intensive culture increased significantly with increasing alder (Alnus glutinosa) in the mixtures. Hybrid poplar heights in short rotation intensive culture mixtures containing the highest percentages of alder, were found to be comparable to those obtained from optimal rates of ammonium nitrate fertilization tested on an adjacent plot of pure hybrid poplar. De Bell and Radwan (1979) found that annual dry matter production in mixed plantings of 2-year-old coppiced black cottonwood (Populus trichcocarpa)/red alder was higher than production in pure cultures of cottonwood and alder. CONCLUSIONS The main benefits which can be derived from the use of agroforestry systems includes: 1) Socio-economic benefits from revitalization of rural areas; 2) Diversification of income sources through risk spreading; 3) Full, productive use of marginal lands; and 4) High quality lands can be brought to their maximum productive capacity. 104 Agroforestry is a complex applied science requiring knowledge of the environment, agriculture, forestry, horticulture, animal husbandry, and local socio-economic and cultural conditions. Although much is known about the com- ponents individually, relatively little is known about the interaction between them. There is a need for basic infor- mation on all aspects of agroforestry technologies. This includes a systematic compilation of knowledge on agro- forestry systems as well as the development of objective methods to evaluate the systems (Lundgren, 1982). 105 L IST OF REFERENCES Adams, SAL 1975. Sheep and cattle grazing in forests: A review. J. Applied Ecology 12:143-152. Allen, E.K., K.F. Gregory and O.N. Allen. 1955. Morphological development of nodules onICaragana arborescens Lam. Can. J. Bot. 33:139-148. Anonymous. 1947. The use and misuse of shrubs and trees as fodder. Imperial Forestry Bureau Joint Publ No. 10. Anonymous. 1978. Agroforestry - A new kind of farming? Rural Res. 99 CSIRO Anonymous. 1983. Agroforestry in the West African Sahel. National Academy Press, Waqshington, ELC. 83 pp. Barr, ILA. 1973. Forestry and farming in harness. Farm Forestry 15(1):1-12. Batini, F.E., G.W. Anderson and R. Moore. 1983. The practice of agroforestry in Australia. In: LLB. Hannaway, ed. Proc. International Hill Lands Symposium. Foothills for food and forests, Corvallis, OR pp. 233-246. Bene, J.G., H.W. Beall and A. Cote. 1977. Trees, food and people. Land management in the tropics. International Development Research Center. Ottawa, Canada. 52 pp. Binkley, D. 1983. Ecosystem production in Douglas-Fir plantations: Interaction of red alder and site fertility. For. Ecol. and Manage. 5:215-227. Blanford, PLR. 1958. Highlights of one hundred years of forestry in Burma. Emp. For. Rev. 37:33-42. Borough, CLJ. 1979. Agroforestry in New Zealand - the current situation. Aust. Forester 42(1):23-29. Carter, IuJ. 1977. Soil erosion: The problem persists despite the billions spent on it. Science 196:409-411. 106 Chapman, A.G. 1935. The effects of black locust on associated species with special reference to forest trees. Ecol. Monog. 5:37-60. Chapman, A.G. and R.D. Lane. 1951. Effects of some cover types on interplanted forest tree species. Central States For. Expt. Sta. Tech. Paper No. 125. Cumberland, K.B. 1976. Three-tier farming in New Zealand's economic future. Farm Forestry 18(2):37-52. Dawson, J.O. 1983. Dinitrogen fixing plant symbioses for combined timber and livestock production. In: D.B. Hannaway, ed. Proc. International Hill Lands Symposium. Foothills for food and forests, Corvallis, OR pp. 95-112. DeBelle, D.S. and M.A. Radwan. 1979. Growth and nitrogen relations of coppiced black cottonwood and red alder in pure and mixed plantings. Bot. Gaz. l40(Supp1.):S97- SlOl. Detwiler, S.B. 1947. Notes on honeylocust. U.S.D.A., Soil Conservation Service. 197 pp. Douglas, J.S. and R.A. de J. Hart. 1976. Forest farming: Towards a solutions to problems of world hunger and conservation. Watkins, London. 199 pp. Eardley, C.M. 1945. Tree legumes for fodder. J. Agric. S. Australia 48:342-345. Eckholm, E.P. 1976. Losing ground: Environmental stress and world food prospects. W.W. Norton, New York. 223 pp. Farmer, R.E., Jr. 1976. Tree crops and the back-to-the- land movement. Northern Nutgrowers Assn. Ann. Rep 67:33-39. Farnsworth, M.C. and A.J. Male. 1975. Is forest farming the answer to marginal lands problems in Northland. Farm Forestry 17(1):10-18. Farnsworth, M.C. 1975. The forest farmer and his physical environment. Farm Forestry l7(4):91-95. Felker, P. and R.S. Bandurski. 1979. Uses and potential uses of leguminous trees for minimal energy input agriculture. Econ. Bot. 33(2):172-184. Ferguson, J.A. 1922. Influence of locust on the growth of catalpa. J. For. 20:318-319. 107 Finn, R.F. 1953. Foliar nitrogen and growth of certain mixed and pure forest plantings. J. For. 51:31-33. Forbes, RJL 1895. The mesquite tree: its products and uses. Arizona Ag. Exp. Sta. Bull. No. 13, 26 pp. Funk, D.T., R.C. Schlesinger and F. Ponder, Jr. 1979. Autumn-olive as a nurse plant for black walnut. Bot. Gaz l40(supp1.):SllO-Sll4. Garcia, F.N. 1916. Mesquite beans for pig feeding. New Mexico Ag. Exp. Sta. 28 h Ann. Rep. pp 77-82. Gold, M.A. and J.W. Hanover. 1984. Honeylocust (Gleditsia triacanthos L.): Important chemical characteristics and cultural systems for use in agroforestry systems. (In preparation). Gordon, J.C. and J.O. Dawson. 1979. Potential uses of nitrogen-fixing trees and shrubs in commercial forestry. Bot. Gaz. l40(Suppl):S88-S90. Haines, S.G., L.W. Haines and G. White. 1978. Leguminous plants increase sycamore growth in northern Alabama. J. Soil Sci. Amer. 42:130-132. Halls, L.K., R.H. Hughes and F.A. Peevy. 1960. Grazed firebreaks in southern forests. U.S.D.A. Information Bulln No.226. ' Hanson, E.A. and J.O. Dawson. 1982. Effect of Alma glutinosa on hybrid Populus height growth in a short- rotatfon intensively cultured plantation . For. Sci. 28(1): 49-59. Harrington, C.A. and R.L. Deal. 1982. Sitka alder, a candidate for mixed stands. Can. J. For. Res. 12:108- 111. Hershey, J.W. 1935. Tree crops and their part in the Tennessee Valley. TVA Dept. of Forestry Relations Rept. 8 p. Huguet, L. 1979. Symbiosis of agriculture and forestry. Unasylva 31:25-29. Huxley, EhA., ed. 1983. Plant research and agroforestry. International Council for Research in Agroforestry. Nairobi. 617 pp. Hills, L.D. 1977. Farming without fields. The Ecologist 7:100-105. 108 Hirst, E. 1974. Food-related energy requirements. Science 184:134-138. Jurriaanse, A. 1973. Are they fodder trees? Pamphlet No. 116. Dept. of Forestry. Pretoria, South Africa. 32 pp. Kincaid, W.H., W.B. Kurtz and H.E. Garret. 1982. A silvicultural-economic model for black walnut. In: Black Walnut for the Future. U.S.D.A. For. Serv. N.C. For. Expt. Stn. Gen. Tech. Rept. Nc-74. pp. 122-127. Knowles, RJL 1972. Farming with forestry: multiple land use. Farm Forestry l4(3):61-70. Knowles, R.L., B.K. Klomp and A. Gillingham. 1973. Trees and grass - an opportunity for the hill country farmer. N.Z. Farmer 94(17):48-52. Knowles, R.L. and N.S. Percival. 1983. Combinations of g. radiata and pastoral agriculture on New Zealand hill country: Forest productivity and economics. In: ELB. Hannaway, ed. Proc. International Hill Lands Symposium. Foothills for food and forests. Corvallis, OR pp. 203-218. Loock, E.E.M. 1947. Three useful leguminous fodder trees. Farming S. Africa. 22:7-12,24. Lundgren, B. 1982. What is agroforestry? Agroforestry Systems l(l):7-12. MacBrayne, (LG. 1982. Agroforestry for upland farms. Scott. For. 36:195-206. MacDonald, IuH., ed. 1982. Agroforestry in the African humid tropics. Proc. of a workshop, Ibadan, Nigeria. April 27 - May 1. U.N. University. 163 pp. Marrs, R.H., L.D.C. Owen, R.D. Roberts and A.D. Bradshaw. 1982. Tree lupin (Lupinus arboreus Sims): an ideal nurse crop for land restoration and amenity plantings. Arboric Journal 6:161-174. McDaniels, [nH. and Ans. Lieberman. 1979. Tree Crops: A neglected source of food and forage from marginal lands. Bioscience 29(3):173-175. McIntyre, A.C. and c.0. Jeffries. 1932. The effect of black locust on soil nitrogen and growth of catalpa. 109 McLean, A. 1983. Producing forage for livestock on forest ranges. In: D.B. Hannaway, ed. Proc. International Hill Lands Symposium. Foothills for food and forests, Corvallis, OR pp. 175-183. Mergen, F. and C.K. Lai. 1982. Professional education in agroforestry in North America. In: International Workshop on professional education in agroforestry. ICRAF. Dec. 6-10, 1982. Nairobi, Kenya. 36 pp. Moore, J.C. 1948. The present outlook for honeylocust in the South. N.N.G.A. Ann. Rept. 19:104-110. Olsen, P.F. 1974. Forestry and livestock farming. Farm Forestry 16(3):61-76. Pearson, H.A. 1982. Economic analysis of forest grazing. In: T. Clason, ed. Proc. of Symp. Multiple use land management for nonindustrial pine forest land owners. Ruston, LA. pp. 77-88. Pearson, H.A. 1983. Forest grazing in the U.S. In: D.B. Hannaway, ed. Proc. International Hill Lands Symposium. Foothills for food and forests. Corvallis, OR pp. 247-260. Pimentel, D., E.L. Hurd, A.C. Bellotti, M.J. Forster, I.N. Oka, O.D. Sholes and R.J. Whitman. 1973. Food production and the energy crisis. Science 182:443-449. Pimental, D., E.C. Terhune, R. Dyson-Hudson, S. Rochereau, R. Samis, E.A. Smith, D. Denman, D. Reifschneider and M. Shepard. 1976. Land degradation: Effects on food and energy resources. Science 194:149-155. Plass, W.T. 1977. Growth and survival of hardwoods and pine interplanted with European Black Alder. U.S.D.A. For. Serv. Res. Paper NE-376. 10 pp. Premar, T.A. and R.F. Fisher. 1983. N2 Fixation and accretion by wax myrtle (Myrica cerifera) in slash pine (Pinus elliottii) plantations. For. Ecol. and Manage. 5:39-46. Raitanen, W.E. 1978. Energy, fibre and food; Agriforestry in eastern Ontario. Eighth World Forestry Congress. Jakarta, Indonesia. October 16-28. 13 pp. Roth, P.L. and R.J. Mitchell. 1982. Effects of selected cover crops on the growth of black walnut. In: Black Walnut for the Future. U.S.D.A. For. Service. N.C. For. Expt. Sta. Gen. Tech. Rep. NC—74 pp. 110-113. 110 Schreiner, EJL 1959. Production of poplar timber in Europe and its significance and application in the U.S. U.S.D.A. For. Service Ag. Handbk. No. 150. 124 pp. Sekawin, M. and Prevosto. 1978. (Technical and economic analysis of the influence of management system and intercropping in a poplar plantation located in Piacentro). (in Italian). Cellulosa e Carta. No. 8 18 pp. Sharrow, S.H. and W.C. Leininger. 1983. Sheep as a silvicultural tool in coastal Douglas-Fir forests. In: ILB. Hannaway, ed. Proc. International Hill Lands Symposium. Foothills for food and forests. Corvallis, OR pp. 214-231. Shiflet, TuN. 1980. ‘What is the resource? Proc. Southern For. Range and Pasture Resource Sympu, New Orleans, LA March 13-14, 1980. pp. 17-28. Smith, JJL 1909. Elimination of the gullied hillside through tree breeding. Amer. Breeders Assn. Ann. Rep. 5:265-268. . 1911. Breeding and the use of tree crops. Amer. Breeders Assn. Ann. Rep. 6:50-56. . 1914. Soil erosion and its remedy by terracing and tree planting. Science 39:858-862. . 1950. Tree Crops: A permanent agriculture. 1978 reprint of the 1950 edition. Harper and Row. New York. 408 pp. Smith, R.M. 1942. Some effects of black locusts and black walnuts on southeast Ohio pastures. Soil Sci. 53(5):385-398. Spurgeon, D. 1980. The promise of agroforestry. American Forests 86(10):20-23, 63-67. Steinhart, C.E. and J.S. Steinhart. 1974. Energy use in the U.S. Food system. Science 184:307-316. Tarrant, RJL 1961. Stand development and soil fertility in a Douglas-Fir/red alder plantation. Forest Science 7:238-246. Tarrant, R.F. and J.M. Trappe. 1971. The role of Alnus in improving the forest environment. Plant and S311 (Special volume):33S-348. 111 Tiffany, W.N. and J.F. Barrera. 1979. Comparative growth of pitch and Japanese black pine in clumps of the N - fixing shrub, bayberry. Bot. Gaz. l40(Supp1.):8108- $109. Tustin, J.R. and R.L. Knowles. 1975. Integrated farm forestry. New Zealand J. Forestry 20(1):83-88. Tustin, J.R., R.L. Knowles and B.K. Klomp. 1979. Forest farming: A multiple land-use production systems in New Zealand. For. Ecol. and Manage. 2:169-189. Van Sambeek, J.W. and G. Rink. 1982. Physiology and silvicuture of black walnut for combined timber and nut production. In: Black pp. 47-51. Walton, G.P. 1923. A chemical and structural study of mesquite, carob, and honeylocust beans. U.S. Dept. of Agriculture, Bull. No. 1194. 19 pp. Williams, G. 1980. Tree Crops for energy production in Appalachia. In: Tree Crops for Energy Co-production on Farms. U.S. D.O.E. Solar Energy Res. Inst. Symposium. Nov. 12-14, 1980. Estes Park, Co. pp. 7- 20. Zarger, T.G. 1956. Status of tree crops investigations in the Tennessee Valley region. N.N.G.A. Ann. Rept. 47:57-68. Chapter IV Honeylocust (Gleditsia triacanthos LJ: Important Chemical Characteristics and Cultural Systems for Use in Agroforestry Systems ABSTRACT The historical development of honeylocust from an unimportant, minor forest associate to a potentially valuable multi-purpose tree crop for agroforestry systems is reviewed. Various management scenarios for its use are suggested, both for industrialized countries, as well as third world nations. Proposed uses include; 1) As a component in multi-purpose shelterbelt systems; 2) As a perennial crop tree for marginal lands; 3) For use in watershed management systems and for erosion control; 4) In two-tier multi-cropping systems; and 5) In ultra short rotation intensive culture systems. Results of chemical analyses on pod sugars and seed and leaf proteins are reported. Total pod sugar content varied from 13.6 to 30.9 percent. Seed protein content varied from 16.6 to 27.8 percent. Leaf protein content ranged from 13.6 to 28.9 percent. The variation patterns in leaf protein, seed protein, and pod sugars are random with no particular provenance or region being especially high in any given 112 113 trait. The use of yield components is discussed in relation to breeding strategies for maximizing sugar and protein yields. Two general categories of natural pod phenotypes are described. Pods from northern regions can be characterized as having a papery pericarp fraction, minimal amounts of carbohydrate pulp, and consistently high seed sets. Pods from southern regions have a pulp-filled pericarp fraction, very poor seed sets, with seed chambers often filled with carbohydrate pulp. Results of two cultural studies on preemergent herbicides and spacings are reported. Ultra short rotation intensive culture systems for growing honeylocust can be succesfully accomplished by direct-seeding, followed immediately by application of the preemergent herbicide DCPA (dimethyltetrachloroterephthalate) with no harmful effects on germination of the seeds. Planting direct-seeded honeylocust at three different spacings showed that a spacing of 10 x 15 cm gave the highest biomass yields in the first year after planting. INTRODUCTION Honeylocust, Gleditsia triacanthos L., is a multi- purpose tree which has potential for use in numerous management scenarios and in many diverse locations throughout the world. A closer look at its potential uses points to the significant role which multi-purpose trees may have as components of agroforestry systems. These uses 114 include a variety of chemical and animal feedstocks from the pods, high protein food supplements and industrial gums from the seeds, animal fodder/green manure from the leaves, and high caloric value fuelwood. The added values of multi- tiered cropping systems, watershed management and erosion control, and fuller marginal land use must also be considered. Honeylocust has been advocated as a multiple-purpose crop tree for shelterbelts in the Great Plains (Bagley, 1976), and is suggested for similar use in the province of Heilongjiang in north-eastern China (Pers. comm. Jeff Gritzner, 1983). Because it can provide a source of fodder, protein, energy, and erosion control, honeylocust appears to be the most promising candidate for use as a staple per- ennial crop tree for marginal land in southern Appalachia (Williams, 1982). For these same reasons and because of its apparent high value leaf fodder, additional interest in testing the honeylocust exists in the Himalayan foothills of India (Pers. comm. P.K. Khosla, 1983) and other areas of the highland tropics (N.A.S., 1983). As a component in a multiple-use integrated farm system it may have value in much of the eastern U.S. (MacDaniels and Lieberman, 1979; Bagley, 1981), New Zealand (Davies and MacFarlane, 1979) and Australia. In Australia honeylocust is being promoted and marketed as a fodder tree for livestock, windbreaks, shade, erosion control and fence posts (Anonymous, 1982). 115 Other multi-purpose trees which have potential for use in integrated systems include black walnut (Juglans nigra) (Kincaid, 22.21}! 1982), hybrid poplars (Populus sppJ (Lora, and Wayman, 1979; Raitenan, 1978), and a host of other leguminous and non-leguminous nitrogen fixing trees such as the alders (_A_l_n_us_ spp.) (Tarrant and Trappe, 1971; Gordon and Dawson, 1979), black locust (Robinia psuedoacacia) (Keresztesi, 1983), carob (Ceratonia siliqua) (Coit, 1951; Merwin, 1980), and the mesquites (Prosopis sppJ (Parker, 1982). One would be remiss if mention was not made of the genus which has received more attention than all of the others combined, Leucaena. A recent bibliography compiled by the USDA contains over 2,000 citations on Leucaena (Oakes, 1982; Oakes, 1983) covering every imaginable topic from adaptation, to livestock, and utilization. Additionally, the National Academy of Sciences (Anon., 1977) devoted an entire publication to Leucaena, and there are now two journals, Leucaena Research Reports and Leucaena Forum, specifically dedicated to publishing results of Leucaena research. Two different cultural systems of use are currently envisioned for honeylocust. One system entails the use of a widely spaced, two-tiered orchard with a variety of forage, vegetable or woody crops grown beneath the trees. In addition to the use of the pods for ethanol and stillage for animal feed, and use of the seeds as protein supplements and industrial gums, the return from the annual crops can be 116 used to increase the overall economic stability and viability of the enterprise. A second system involves growing direct-seeded honeylocust on ultra-short rotations, and at very close spacings, for annual harvest(s) as a chemical or animal feedstock. SILVICULTURE AND GENETICS Within the natural range of the honeylocust a large amount of variation exists in both climatic and edaphic conditions. The native range extends from central Pennsylvania west to southeastern South Dakota-Northcentral Nebraska, south to central Texas, east along the Gulf to Georgia, and north to Pennsylvania (USDA, 1965). The average annual precipitation within the natural range varies from 500 mm in S. Dakota-Nebraska to over 1800 mm in North Carolina. The frost free period varies from a low of 140 days in the north western extremes to a maximum of over 340 days in southern Louisiana (USDA, 1941). The honeylocust achieves its best growth on fertile, moist, alluvial floodplains, but will also grow on soils of limestone origin and is resistant to both drought and salinity (Howell, 1939). Results of a study on the genetic variation in growth, phenology and winter-hardiness indicate that a large amount of genetic variation is present among and within regions for all traits analyzed (Gold and Hanover, 1984). In fact, in many of the traits studied, the range of variation is so large that it has proven to be a mixed blessing. Use of 117 unselected material has given honeylocust an undeservedly bad reputation in some areas (Mostert and Donaldson, 1960), while use of selected sources has led to high expectations in others (Moore, 1948). To maximize the potential benefits of growing honeylocust several factors must be taken into consideration. These include the close matching of ecological requirements to various locations throughout the world, the utilization of the most appropriate ideotypes for each intended use, and careful selection and breeding to develop superior varieties. HISTORICAL BACKGROUND By early 1900 honeylocust was being touted as a perennial forage tree for the eastern Udi (Smith, 1914). A detailed chemical analysis of honeylocust pods was first conducted by Walton (1923). The agroforestry potential of honeylocust received broader recognition after Smith (1929), illustrated the potential use of numerous tree crop species in developing a permanent agricultural system for marginal, hilly, and eroded lands. Honeylocust was included in the tree crops/hillculture projects at the Tennessee'Valley Authority (TVA), Virginia Polytechnic Institute, and Alabama Polytechnic Institute, which were initiated in 1934 (Hershey, 1936; Zarger, 1956). The honeylocust projects lasted only 14 years, from 1934-1947. Within that brief span of time great strides were made in converting honeylocust from a minor forest component of no commercial 118 value, into a potentially valuable multi-purpose cropping tree. A brief review of the TVA'S accomplishments include the location of wild selections of honeylocust with a total pod sugar content exceeding 35 per cent (Detwiler, 1947); development of a technique for propagating thornless trees by the careful selection of scionwood from thorny parent trees (Chase, 1947); vegetative propagation of superior clones followed by the establishment of grafted orchards of these clones (Stoutemeyer 35 31., 1944; Moore, 1948; Zarger, intercropped pasture tree (Moore, 1948; Zarger 1956); and a determination of the feed value of honeylocust pods through chemical analysis and animal feeding trials (Atkins, 1942; Moore, 1948). An abrupt change in national priorities following World War II terminated all research efforts on honeylocust except for the maintenance of an archive of superior clones at Norris, Tenn. (Moore, 1948; Zarger, 1956; Scanlon, 1980). A summary of correspondence, general information and specific research results involving honeylocust up to the mid-1940's is available (Detwiler, 1947). An excellent review of all honeylocust research conducted at the TVA can be found in Scanlon (1980). By the late 1940's, interest in honeylocust was also evident as far away as Australia (Eardley, 1945), South Africa (Loock, 1947; Jurriaanse, 1973), and Malawi (Douglas, 1967). 119 In depth studies on the identification of chemical constituents have provided basic information on the overall chemical composition of honeylocust (Wealth, 1956; Watt and Breyer-Brandwick, 1962). Felker and Bandurski (1976), and Becker (Pers. comm., 1982) analyzed the seed protein and amino acid content. Baertsche (1980) studied the animal feedstock value of intensively cultured seedlings. ‘Walton (1923), National Academy of Sciences (Anon., 1971) and Scanlon (1980) reported on the sugar content of the pods. Each of these studies have focused on a thorough analysis of one or a few individual trees. Building on these previous studies, and using materials obtained by a rangewide collection of honeylocust germplasm, the approach taken in this study was to identify the range of variation present in useful chemical constituents, namely seed protein content, leaf protein content, and total pod sugar content. Morphological measurements of pods and seeds were determined as a basis for future study of yield components, and for selection and breeding towards development of diverse ideotypes. Finally, results of two cultural studies are presented which will help to lay the groundwork for further research on closely spaced short rotation intensive culture (SRIC) honeylocust for animal and chemical feedstocks. 120 MATERIALS AND METHODS A rangewide collection of honeylocust germplasm was undertaken in the fall of 1979. By February 1980, over 450 accessions had been received, covering a majority of the natural and naturalized range. Collection details and results of a study on the genetic variation in 2-year old honeylocust seedlings are reported elsewhere (Gold and Hanover, 1984). Upon receipt, each collection was stored at 1-4°C. Ten pods from each accession were chosen for further morphological measurements and analysis. Morphological data Morphological measurements on the pods included length, width, thickness and weight (oven dry). Pod pubescence was scored on a scale from 0, representing total absence, to 4, representing heavy pubescence. Pubescence was scored to test for possible resistance to Amblycerus robiniae, the bruchid seed weevil, which can severely damage a seed crop. The total number of seeds per pod was recorded for each source, and these were divided into the number of sound seed and number of insect damaged seed. When possible, seed weight was determined as an average weight of 100 seeds per accession (fresh weight). Seed length, width, and thickness were determined as an average of 10 seeds using a dial guage accurate to 0.0005 millimeter. Seed volume was determined by water displacement. 121 Seed ppptein Following seed extraction and measurement, seed from 200 sources were analyzed for total protein nitrogen (N) on a whole seed basis. Ten seeds per source were used for N determinations. Prior to analysis, seeds were dried for 48 hours at 50°C. in a convection oven. Each group of 10 seeds was run in two lots of 5 and analyzed in duplicate. Samples were analyzed for total protein N according to the method of Wall and Gehrke (1975L. A.40 sample block digestor and Technicon Autoanalyzer were used for determination of ammonia nitrogen. Protein was calculated as N x 6.25. Pod sugars Subsequent to completion of morphological measurements on pods and seeds, a subset of 79 sources were frozen at -38°C prior to sugar analyses. Each sample was a composite of 10 pods per source. Pod samples were dried for 24 hours at 50°C in a convection oven and then passed through a Wiley mill with a 20 mesh screen. Reducing and non-reducing sugars were analyzed according to the method of Nelson (1944). Leaf proteins Leaf samples, consisting of recently matured leaves from the upper crown, were collected from the field at the East Lansing test site on August 1St and 2nd 1982, using the sampling method of Jones, Large, Pfleiveder and Klosky (1971). Eighty four sources, representing the entire range 122 of honeylocust, were chosen for analysis. Two plots per source were sampled and each sample consisted of a bulk of 4 trees within each plot. Immediately following harvest, samples were dried for 24 hours in a convection oven at 65° C and then ground in a Wiley mill through 20 mesh screen. Each sample was analyzed for total protein N employing the same technique used in seed protein analysis (see above). Protein was calculated as N x 6.25. Herbicide study A weed-free seedbed was prepared at the Michigan State University Tree Research Center, located at E. Lansing, Mi. A 1.5 percent solution of the postemergent herbicide glyphosate (Roundup)*was applied to kill existing vegetation. Seven days later the test plot was tilled. Honeylocust seeds were scarified for 60 minutes in concentrated sulfuric acid (H2804). Scarified seed were planted in a sandy-loam soil at a depth of 142 centimeters. Treatments consisted of 25 seeds sown linearly at 20 mm intervals, with three replications of each treatment in a randomized block design. Two preemergent herbicide treatments were tested, DCPA (dimethyltetrachloroterephtha- late or Dacthol) at a rate of 9.0 kg active ingredient (ai) per hectare (ha), EPTC (5-ethyldipropylthiocarbamate or Eptam) at a rate of 6.5 kg a.i. per ha., against a control of no herbicide application. 123 Selection of herbicides and rates of application were chosen based on availability and for comparison with results reported by Warmund, Long, and Geyer (1980), who tested the effects of 10 preemergent herbicides on the germination of honeylocust, black locust, and Kentucky Coffee tree seeds (Gymnocladus dioicus) grown in nursery containers. Herbicide treatments were applied 24 hours after the seeds were sown. In order to prevent volatilization of EPTC, the plots were irrigated immediately following herbicide application, and at regular intervals for the next 75 days to provide optimal moisture conditions. Germination data were recorded 21 days after herbicide application. Seedling survival and weed control were monitored for 75 days. Biomass yield study A direct-seed, spacing study was initiated at the MALU. Tree Research Center nursery on May 13, 1982. The 18 m x 17 m test site was treated with glyphosate (Roundup), a postemergent herbicide, for removal of existing vegetation. A week later, the test site was tilled. Soil conditions in the nursery at the time of planting were: sandy-loam, pH 6.5, 4% OM, 168 kg available P per ha., and 78 kg available K per ha. Based on results of soil analyses, the test plot was fertilized with one pound of granular 12-12-12, and then raked into the soil. Annual precipitation at the site averages 800mm. 124 Seed were scarified (as above) on May 20. Following scarification the seeds were soaked in water overnight and the imbibed seeds were sown the following day. A randomized block design, with 3 blocks and 3 treatments per block, was used to test the effects of different interrow spacings on the biomass yield of direct- seeded honeylocust. Each block was 5.0 m x 5.0 m in size and contained 3 spacing treatments 1 m x 4.5 m in size. Spacing treatment I consisted of 6 rows, spaced at 15 cm intervals, treatment II contained 3 rows spaced 30 cm apart, and treatment III had 2 rows, spaced 45 cm apart. Seeds were spaced at 10 cm intervals within rows in all treatments. Unfortunately, allowing the seed to imbibe before planting proved disasterous as the seed rotted in the soil and less than 1% germination occurred within the next 21 days. The experiment was repeated beginning on June 16 when the plots were retilled without further herbicide application. Another set of seeds was scarified on June 17, and then sown immediately. Lack of seed caused the elimination of one block on the replanting. Due to the late sowing date, harvest of the seedlings was postponed until the following summer. Hand cultivation was used to control weeds from the time of germination until the plots were harvested. On August 1, 1983 the 13 month old seedlings were mechanically harvested with a forage harvester. Fresh 125 weights were recorded for all treatments and 1000 9 samples were taken from each treatment for determination of moisture content. RESULTS AND DISCUSSION Multi-cropping Total sugar content of the honeylocust pods ranged from 13.6 to 30.9 percent (Table 4.1). Compared with selections located by the TVA, there were no exceptionally high in- dividual tree values for pod sugar content. The sources chosen for total pod sugar analyses represented a cross section of the natural range. Total sugar content is nega- tively correlated with latitude (r= -0.52), the percent pericarp (non-seed) fraction is negatively correlated with latitude (r= -0.46), and grams of total sugar per pod (peri- carp fraction only) is also negatively correlated with lati- tude (r= -0.4l). These correlations indicate that southern sources contain higher levels of total sugar as well as higher percentages of pericarp fraction. Based on these data, regional selection for sources with the highest total sugar content (expressed in grams of sugar per pod) would initially focus on southern sources. ‘Within the southern region, selection would have to be on an individual tree basis due to the very large amount of within-region variation (Table 4.2) . 126 Table 4.1 Variation in morphological arri chemical traits of honeylocust parent trees from subsampled rangewide test. Trait Mean Min. Max. s.d. _ _ - - pod .. - .. .. Morphological Total pod might (g) 9.3 2.1 22.1 3.5 1Weight pericarp fraction only (9) 7.3 3.0 21.4 3.1 2Pericarp fraction (%) 73.6 45.4 97.6 11.4 Pod length (cm) 29.0 17.9 41.3 5.0 Pod width (an) 2.9 1.9 4.7 0.60 Pod thickness (cm) 0.3 0.08 0.65 0.13 Chemical Total sugars (%) 22.4 13.6 30.9 3.7 Nonreducing sugars (%) 19.8 12.5 28.2 3.3 Reducing sugars (%) 2.7 0.68 5.0 1.1 3Total sugars perpod (g) 1.6 0.41 6.6 0.26 3Nonreducing sugars per pod (g) 1.4 0.38 6.0 0.23 Reducing sugars per pod (g) 0.2 0.02 1.1 0.03 - - - - Seed - - - - Morphological Total number of seeds per pod 13.5 1.5 30.0 5.4 Weight seed fraction (9) 0.19 0.07 0.30 0.04 Total seed might per pod (g) 2.5 0.26 5.3 1.1 4Seed fraction (%) 26.4 2.4 54.6 11.4 Sound seed per pod 10.3 0.00 25.8 5.0 Damaged seed per pod 3.2 0.00 16.3 3.1 5Sound seed (%) 58.2 0.00 100.0 29.7 Seedlength (mm) 4.0 3.3 5.1 0.32 Seed width (mm) 2.5 2.0 3.2 0.21 Seed thickness (mm) 1.5 1.1 2.1 0.17 Seed volume (ml) 0.14 0.60 0.20 0.03 Seeds per pound 2,389 6,486 1,513 -- Chemical 6Seed protein (%) 21.6 16.6 27.8 2.0 —Seed proteinjer pod (g) 0.53 0.05 1.05 0.23 l-Pericarp fraction (9) = total pod weight - total seed might. 2-Pericarp fraction (%) = (pod might only/total pod might) * 100. 3-Total sugars per pod (g) = (total sugars (%)/100) * weight pericarp fraction. 3-The same calculation was used in the determination of grams of reducing and nonreducirg sugars. 4-Seed weight (%) = (total seed might/total pod weight) * 100. 5-Sound seed (%) is a measure of damage by seed weevils. 6-Seed protein per pod = (seed protein (%)/100) * seed might. 127 Table 4.2 Means and range variation in honeylocust pod sugar content of seed sources among and within regions l/. ....... g - - - - - _ - Region 2/ Mean Minimum Maximum SE 23.9 17.7 30.9 SW 24.2 21.9 27.8 pg 24.4 17.7 30.6 SOUTHERN + EC 24.2 LS 19.1 13.6 24.4 NW 20.1 16.0 24.9 ‘wg 19.9 14.5 22.5 NORTHERN + WC 19.7 l/Regions are discussed in more detail {a Chapter 2 (G613 and Hanover, 1984). _2_/ SE,SW,EC,LS,NW,WC represent southeast, southwest, east- central, Lake States, northwest, and west-central regions, respectively. Differences in sugar content have been documented for trees from the same clone grown in different locations (Scanlon, 1980). Pods from the "Millwood" clone, a selection located by the TVA during its involvement in tree crops research, contained 36.8 percent total sugar when grown in Alabama and 21 percent total sugar when grown in Maryland (Detwiler, 1947). This points to the necessity for testing honeylocust in many diverse locations in order to accurately estimate its performance. Two general categories of natural pod phenotypes can be described. Pods from northern regions can be characterized as having a papery pericarp fraction, with minimal amounts of carbohydrate pulp, accounting for approximately 50 percent of the total pod weight, and consistently high seed sets. Pods from southern regions have a pulp-filled 128 pericarp fraction accounting for 75 to 95 percent of total pod weight, and frequently have very poor seed sets. The seed chambers are often filled with carbohydrate pulp. Crop yields of honeylocust pods, and maximum amounts of extractable carbohydrates on a per tree basis, are two of the important factors to be considered when assessing its cropping potential for chemical and animal feedstocks in two-tier integrated farming systems. The actual pod sugar content (fermentable carbohydrates) per pod is secondary in importance. Although reliable crop data from plantations of honeylocust are scarce, the best available data comes from a cooperative TVA-Auburn University study. A grafted plantation of the superior "Millwood" selection yielded an average of 33 kgs. (dry wt.) per tree between the ages of five and nine. However, due to the biennial bearing habit this average obscures that fact that 9 year old trees produced an average of 82 kgs. (dry vma) per tree (Moore, 1948). Based on these yield figures, at a 40' x 40' spacing (28 trees to the acre) honeylocust produced over two-and-a- half tons of pods per acre. The average crop of 33 kgs. would yield almost one megagram per acre (dry th. Seed proteins Results of analyses for seed protein content among 200 honeylocust sources show a wide range of variation (Table 4.1). Seed protein levels vary from 16.6 to 27.8 percent, with a mean of 21.6. Protein was determined on a 129 whole seed basis (calculated as N x 6.25). When trying to select for sources which contain maximum overall levels of seed protein, many factors must be considered. Total pod yield per tree, number of seeds per pod, seed weight, and seed size are all potentially important yield components which contribute to maximal seed protein yields per tree. In order to maximize seed protein production, the most important yield component is the total pod yield per tree. Because most of the collections used in this rangewide study were obtained by mail, an accurate assessment of this component will not be possible until field planted sources begin to flower and fruit. Other yield components were studied which affect total seed protein content on a per pod basis. Total seed protein yield per pod is determined by the total number of seeds per pod, the total weight of seeds per pod, individual seed size, seed weight and seed protein content. Little relationship was found when correlations between seed weight, seed size, or seed volume were run with seed protein content. Thus, the size, shape, and weight of the seed appears to be totally independent of seed protein content. Also, no relationship was found between protein content per seed, and total seed protein content per pod. The seed fraction of the pod is negatively correlated with pod weight (r= -0.30). Pods from northern latitudes have a higher ratio of seed fraction to pericarp fraction than pods from trees in southern latitudes. In northern 130 sources, the pericarp consists mainly of a papery exocarp with a minimal amount of carbohydrate filled mesocarp. In light of these data, variation in total seed protein content per pod appears to be due to the remaining components, i.e. the total number of seed per pod, and individual seed weights. Correlation analyses show that the total number of seeds per pod accounts for 67 percent of the variation in total seed protein content. Individual seed weight accounts for an additional 17 percent of the variation. Based on the results to date, high seed yields are more important to total seed protein yields than individual seed protein content. In the longer term, simultaneous selection for a combination of yield components will be desired to maximize pod yields, seed yields, seed protein content and seed protein quality. Due to the hard, impermeable seedcoat of the honeylocust seed (Heit, 1942) it has been reported that the unbroken seed would pass directly through the digestive tract of animals such as sheep in which case the protein value of the seeds would be lost (LeRoux, 1959; Mostert and Donaldson, 1960). In a recent study sheep were fed broken, uncrushed pods, with seeds intact (Small, 1983). Results indicated that sheep can digest the whole seed whether consumed alone or in the pods, and that at least 75-90% of the whole seeds were digested. Data from the M.A.S. (Anon., 1971) indicate that 66% of the protein in ground seeds and pods is digestible. 131 Leaf proteins Results of crude leaf protein analysis of leaves from 84 different half-sib families, revealed a large amount of variation (Table 4.3). Baertsche (1980) reported a "late harvest" figure of 20.23 percent crude protein for greenhouse grown honeylocust seedlings. Field results of the rangewide analysis of crude protein show that Baertsche's value falls very close to the overall mean for crude leaf protein content, 20.05 percent. Correlations between leaf protein content and latitude or longitude were nonsignificant. Variation within regions is so large that selection will only be effective at the family and within-family levels (Table 4;”. Based on these preliminary field results, selection for significantly higher levels of leaf protein may be feasible and current data indicate improvements up to 45 percent over the population mean through family selection alone. Table 4.3 Variation in leaf protein content among geographic regions 1/. Region Mean Minimum Maximum SE 20.4 13.6 27.4 EC 20.5 14.0 26.7 LS 20.3 14.7 28.9 NW 20.6 14.1 24.5 WC 18.1 15.0 19.7 SW’ 20.1 14.8 24.5 Overall mean 20.05 13.6 28.9 1/ Regions are more fully'described in Chapter 2 (Gold and Hanover, 1984). 132 Results of pod sugar and seed/leaf protein analyses indicate that significant improvements in the chemical properties of honeylocust are attainable through selection of appropriate families and/or individuals. Cultural systems will also require further development in order to make SRIC systems economically viable»for animal and/or chemical feedstock production. Cultural systems Results of the ANOVA support the findings of Warmund gt ‘gl. (1980) that herbicide application one day after planting is a possible alternative to hand-weeding in nurseries. (Table 4.4). Germination of honeylocust seed in DCPA treated plots was not significantly different than the control. Application of EPTC, recommended for use in alfalfa (Medicagg sativa) (Tesar, 1980). Proved toxic to honeylocust seed germination. After 21 days germination averaged 89.5 percent in the control plots, 77.9 percent in the DCPA treated plots, and EPTC plots showed 100 percent mortality. At the end of 75 days, weed control in both herbicide treatments was 100 percent, while in the control plot it was only 60 percent (Table 4.4). 133 Table 4.4 Preemergent herbicide effects on germination, survival, and weed control of direct-seeded honeylocust. Herbicide Treatment Germination | Survival Weed Control rate 21 days after 60 days after treatment treatment Kg a.i./ha. -------- % ---------- DCPA 8.9 77.9al/ 77.9a 100a EPTC 6.5 00.0b 00.0b 100a Control --- 89.5a 89.5a 60b l/ Means for each category followed by the same letter are not significantly different at the 1% level based on Duncan's multiple range test. A combination of factors may have caused the toxic effects of EPTC. First, immediately following application, the test plots were thoroughly watered to incorporate the EPTC into the soil and prevent its volatilization. As a result, the "effective" application rate may have been too high. Second, the study by Warmund £3 31. (1980) indicated that EPTC had a detrimental effect on honeylocust seed germination in nursery containers. Results show that use of the preemergent herbicide DCPA at rates near 10 kg ai/ha will give thorough weed control in the early stages of germination and growth, enabling the seedlings to fully occupy the site. A second study by Warmund _t _l, (1983) indicates that at least five other preemergent herbicides may also be used for estab- lishment of direct-seeded, field planted SRIC plantations. 134 Direct-seed biomass study Results of a direct-seed, spacing study testing the effects of different spacings on biomass yield of honeylocust, indicate that this technique is worthy of further research. Excellent stand establishment was obtained at all spacings. Significant differences were found between all spacing treatments (Table 4Jfl. Harvest data collected at the end of one year from seed indicate that the closest spacing - 15 cm x 10 cm - gave the highest yields. While the overall yields were not very high, it Should be emphasized that coppice yields are expected to be much higher than growth from seed alone. Coppice regrowth is an important part of the USRIC concept. The advantages of coppicing include; 1) the avoidance of extensive site preparation and replanting after each harvest; and 2) regrowth from established root systems is often faster than seedling growth. Geyer (1981) reports that coppice yields in 2-year old cottonwood (Populus deltoides) and silver maple (Acer saccharinum) were about 62 percent higher than seedling yields. 135 Table 4.5 Biomass production in a l-year old, direct-seeded honeylocust USRIC system l/. Interrow g/ kg/plot % FLC. Mg Forage yield/ha. spacing (cm) _3/ (12% M.C.) 4/ 15 2.5 58.2 1.19a 30 1.6 57.8 0.73b 45 1.1 59.6 0.47c 1/ Treatment means are presented. ;/ Spacing within all rows is 10 cm. 3/ Plot size for all treatments is 15 square meters. 4/ Means followed by a different letter are significantly different at the 1% level using Duncan‘s test. When selecting genotypes for use in USRIC systems, stem form becomes irrelevant, while the ability to coppice vigorously takes on an important role. In honeylocust, fast growing sources from southern latitudes of origin will be useful in USRIC systems as they tend to grow vigorously late into the fall (Gold and Hanover, 1984), allowing for fuller use of the growing season and a later fall harvest than would be possible with locally adapted sources. Field observations indicate that sources do not need to be completely winter hardy to be useful in USRIC systems. This is because the stems will be harvested close to ground level on an annual basis. Based on this line of reasoning, the ideotype of a species used in USRIC systems will have a somewhat different set of selection criteria than the ideotype of a species selected for use in a two—tier multi— crop system or other more traditional uses. Depending on the length of the growing season, one to three harvests per year are anticipated. Planting and 136 establishment costs should be lower, and the use of forage harvesting equipment is possible. Efficient harvesting and processing should be feasible because of great uniformity in the USRIC material derived from control over the genetic, environmental and cultural systems utilized. In combination with effective use of pre-emergent herbicides, systems of USRIC seem to be a viable alternative for the production of chemical and animal feedstocks. Industrial gums and other specialty chemicals In order to fully utilize the economic potential of the honeylocust, one must examine all the constituent components of the pods. In addition to sugars and protein, the component which may eventually have the greatest market potential is the galactomannan fraction in the seed endosperm. (As mentioned previously; honeylocust pods from sources originating in the northern part of the native range have high seed sets and little carbohydrate pulp in the pericarp fraction. This favors the development of products which are derived from the seed. Mucilage polysaccharides which swell to a gel in water or gum polysaccharides which dissolve in water, are often associated with legume seed endosperm as a vitreous layer on the inside of the seedcoat. Several, for example those from guar (Cyamopsis tetragonoloba) and carob (Ceratonia siliqua), are of industrial importance. The endosperm of honeylocust seeds is almost pure galactomannan gums. This gum fraction comprises over 30 per cent of the total seed 137 (Mazzini and Cerezo, 1979). Galactomannan gums are used in the food, bakery, textile, oil drilling, pharmaceutical, cosmetic, and paper industries (Whistler and BeMiller, 1973). Although no assays for specialty chemicals such as gums were done in the present study, they are a logical next phase in our investigations. CONCLUSIONS AND RECOMMENDATIONS Initial genetic gains from provenance/progeny testing are expected to markedly improve each trait compared to average, unselected sources. Results of pod sugar and seed/leaf protein analyses indicate that significant improvements in the chemical properties of honeylocust are attainable through selection of appropriate families and/or individuals. The variation patterns in leaf protein, seed protein, and pod sugars are random with no particular provenance or region being especially high in any given trait. Also, because these traits are yield traits and are composed of many different components, their patterns of inheritance are likey to be complex and as a result, their rate of improvement will be slower. In the short term, it may be a wise idea to take advantage of natural morphological differences inherent in honeylocust growing in northern vs southern regions of the country. The use of honeylocust in different regions for 138 different purposes will result in its most efficient use. In the west-central and northern regions, concentrate on seed protein and seed gum production, while in southern regions, take advantageeof pod sugar/ethanol/stillage production. In the longer term, simultaneous selection for a combination of yield components will be desired to maximize pod yields, seed yields, seed protein content and seed protein quality. When developing ideotypes for use in agroforestry systems, selection for maximum pod production will be the key factor, although attention to total sugar content and other important chemical and morphological traits should not be ignored. The ideotype of a honeylocust used in USRIC systems will have a somewhat different set of selection criteria than the ideotype selected for use in two-tier orchard systems. In addition, many cultural techniques are in need of further development in order to make multi- purpose agroforestry systems economically viable as animal/ chemical feedstock production systems. The Last Word When attempting to promote new ways of looking at the potentials of trees, one tends to accentuate the accomplishments and highlight the potential benefits, rather than dwelling on the ever-present and inevitable problems inherent in the development of new systems and technologies. However, many questions need to be answered, and many problems remain to be worked out. 139 A problem shared in common with all forestry research is that the development of multi-purpose trees and agroforestry systems will require long-term research commitments and project continuity, rare commodities indeed. Specific problems include the need for testing honeylocust in many diverse locations, determination of optimum spacings, harvesting and processing systems development, intercrop trials, and the development of precise management regimes. Animal damage, insect and disease problems etc.,*will all have to be overcome. In contrast with many scientists who work with trees in third world countries, a majority of foresters, agronomists, and others in industrialized countries have been trained to keep their disciplines separate unto themselves and conceive of only a limited role for the use of trees. Multiple-use trees and agroforestry systems such as those described in this paper, are not generally given much thought or credibility. This psychological barrier will have to be removed before these ideas, systems, and methodologies for using trees in new ways are accepted as viable and valid. 140 LIST OF REFERENCES Anonymous. 1971. Atlas of nutritional data on United States and Canada, National Academy of Sciences, Washington,lLC. Anonymous. 1977. Leucaena: Promising forage and tree crop for the tropics. National Academy of Science, Washington, D.C. 115 pp. Anonymous. 1982. Australian Forest Grower. Vol. 5. No. l. p. 21, 30. Anonymous. 1983. Firewood Crops. Shrub and tree species for energy production. National Academy of Sciences, Washington, D.C. Vol. No. 2. 87 pp. Atkins, AJL 1942. Yield and sugar content of selected thornless honeylocusts. Ala. Polytech. Inst., Agric. Expt. Stn. 53rd Ann. Rpt. pp. 25-26. Baertsche, SAL 1980. The potential utilization of short rotation biomass produced trees as a feed source for ruminants. Ph.D. Dissertation, Michigan State University, 125pp. Bagley, W.T. 1976. Multipurpose tree plantations. In: Shelterbelts on the Great Plains. Great Plains Ag. Council Publ. No. 78. Proc. of the Symp. Denver, Co. April 20-22. pp. 125-128. Bagley, W.T. 1981. Honeylocust - A potential farm crop. Chase, ELB. 1947. Propagation of thornless honeylocust. J. Forestry. 45:715-722. Coit, J.F. 1951. Carob or St. John's Bread. Econ. Bot. 5:82-96. Davies, D.J.G. and R.P. MacFarlance. 1979. Multiple- purpose trees for pastoral farming in New Zealand with emphasis on tree legumes. N.Z. Agric. Sci. l3(4):l77- 186. 141 Detwiler, S.B. 1947. Notes on honeylocust. U.S.D.A., Soil Cons. Service. 197 pp. Douglas, J.S. 1967. 3-D Forestry. World Crops. 19:20-24. Eardley, C.M. 1945. Tree legumes for fodder. J. Agric. S. Australia 48:342-345. Fowells, H.A. 1965. Silvics of forest trees of the United States. U.S.D.A. Forest Serv. Handb. No. 271. 762 pp. Geyer, W.A. 1981. Growth, yield, and woody biomass characteristics of seven short-rotation hardwoods. Wood Sci. 13(4):209-215. Gold, M.A. and J.W. Hanover. 1984. Genetic variation in honeylocust (Gleditsia triacanthos L.): 2-year results. (In preparation) Gordon, J.C. and J.C. Dawson. 1979. Potential uses of nitrogen-fixing trees and shrubs in commercial forestry. Bot. Gaz. l40(Supp1):588-590. Heit, C.E. 1942. Acid treatment of honeylocust. N.Y. Conserv. Dept. Notes on Forest Invest., No. 42, n.p. Hershey, J.W. 1935. Tree crops and their part in the Tennessee Valley. TVA Dept. of Forestry Relations Rept. 8 p. Howell, J., Jr. 1939. Tree and shrub species information. U.S. Dept. Agric., Soil Cons. Serv., Bull. No. 53, Woodland Ser. 7, 51 pp. Jones, J.B., Jr., R.L. Large, D.B. Pfleiderer, and H.S. Klosky. 1971. How to properly sample for a plant analysis. Crops & Soils 23(8):15-18. Jurriaanse, A. 1973. Are they fodder trees? Pamphlet No. 116. Dept. of Forestry. Pretoria, S. Africa. 32 pp. Keresztesi, B. 1983. Breeding and cultivation of black locust, Robinia psuedoacacia in Hungary. Forest Ecol. and Management 6:217—244. Kincaid, W.H., W.B. Kurtz and H.E. Garret. 1982. A silvicultural-economic model for black walnut. In: Black Walnut for the Future. U.S.D.A. For. Serv. N.C. For. Expt. Stn. Gen. Tech. Rept. Nc-74. pp. 122-127. LeRoux, P.L. 1959. Red Indians used honeylocust tree as source of sugar. Farming in S. Africa 35(3):40-42. 142 Loock, E.E.M. 1947. Three useful leguminous fodder trees. Farming S. Africa. 22:7-12,24. Lora, J;H. and M. Wayman. 1979. Fast-growing poplar: A renewable source of chemicals, energy and food. Rept. No. 27 In: Poplar Research, Management and Utilization in Canada. D.C.F. Fayle, L. Zsuffa and H.W. Anderson eds. 8 pp. Mazzini, PLN. and Ass. Cerezo. 1979. The carbohydrate and protein composition of the endosperm, embryo and testa of the seed of Gleditsia triacanthos. J. Sci. Food Agric. 30:881-891. McDaniels, IuH. and Ass. Lieberman. 1979. Tree Crops: A neglected source of food and forage from marginal lands. Bioscience 29(3):173-l75. Merwin, MJL 1980. The culture of Carob (Ceratonia siliqua I“) for food, fodder and fuel in semi-arid environments. International Tree Crops Instit. U.S.A., Inc. Winters, California 17 pp. Moore, JJL 1948. The present outlook for honeylocust in the South. N.N.G.A. Ann. Rept. 19:104-110. Mostert, J.W.C. and C.H. Donaldson. 1960. Value of honeylocust as fodder is neglible. Farming in S. Africa 36(1):40. Nelson, N. 1944. A photometric adaptation of the Somogyi method for the determination of glucose. J. Biol. Chem. 153:375-379. Oakes, A.J. 1982. Leucaena bibliography. U.S. Dept. of Agric. 1308 citations. —————. 1983. Leucaena bibliography; (LS. Dept. of Agric. 692 citations. Ogden, RA; 1983. Attempt to ferment sugars in the locust bean tree. Agroforestry Review 3(3 & 4):5. Parker, ILW., ed. 1982. Mesquite utilization. Symposium on mesquite utilization. Texas Tech. University, Lubbock, Tx. Oct. 29-30. Raitanen, VLF. 1978. Energy, fibre and food; Agriforestry in eastern Ontario. Eighth World Forestry Congress. Jakarta, Indonesia. October 16-28. 13 pp. 143 Scanlon, DJL, III. 1980. A case study of honeylocust in the Tennessee Valley region. In: Tree crops for energy co-produciton on farms. U.S. Dept. of Energy Solar Energy Res. Inst. Symposium. Nov. 12-14, 1980. Estes Park, Co. pp. 21-31. Small, BL 1983. Honeylocust pods and the digestion of protein by sheep. Agroforestry Review 4(2):6-7. Smith, J.R. 1914. Soil erosion and its remedy by terracing and tree planting. Science 39:858-862. Smith, J.R. 1950. Tree Crops: A permanent agriculture. Reprint of 1950 Edition. Harper and Row. New York. 408 pp. Stoutemeyer, V.T., F.L. O'Rourke and W.W. Steiner. 1944. Some observations on the vegetative propagation of honeylocust. J. Forestry. 42:32-36. Tarrant, R.F. and J.M. Trappe. 1971. The role of Alnus in improving the forest environment. Plant and Soil (Special ‘volume):335-348. Tesar, DLB. 1980. Clear seeding of alfalfa. Mich. St. Univ. C.E.S. Extn. Bull. E-961. 4 pp. U.S.D.A. 1941. Climate and man. Yearbook of Agriculture. 1248 pp., illus. Walton, G.P. 1923. A chemical and structural study of mesquite, carob, and honeylocust beans. U.S. Dept. of Agric., Bull. No. 1194, 19pp. Warmund, M.R., C.E. Long and W.A. Geyer. 1983. Preemergent herbicides for direct seeding Kentucky Coffeetree, honeylocust, and black locust. Tree Planters} Notes Vol. 34(3):24-27. Watt, J;M. and PLG. Breyer-Brandwijk. 1962. The medicinal and poisonous plants of southern and eastern Africa. Second edition. E. & S. Livingstone Ltd. Edinburgh and London. Wealth of India. 1956. Gleditsia Linn. (Leguminosae). Publications and Information Directorate CSIR New Delhi, India, Vol. 4:135-136. Whistler, R.L. and J.N. BeMiller, eds. 1973. Industrial gums: Polysaccarides and their derivatives. Second edition. Academic Press, New York. 144 Williams, G. 1982. Energy conserving perennial agriculture for marginal land in southern Appalachia. Final Tech. Report to Dept. of Energy Appropriate Technology Small Grants Program 37 pp. Zarger, 15G. 1956. Status of tree crops investigations in the Tennessee Valley region. N.N.G.A. Ann. Rept. 47:57-68. APPENDICES 145 APPENDIX A CORRESPONDENCE AND COLLECTION FORMS 146 147 At Michigan State University we are undertaking a rangewide Honeylocust (Gleditsia triacanthos) seed collection to begin a culprehensive evalua— tion of the genetic variation which exists within the natural range of the species. This is the first step of a long range project for the genetic inprovanent of Honeylocust. Our objectives in this work are to study the potential use of the species for fiber and feedstock production on marginal agricultural lands in the Lake States. We would like this rangewide study to be a cooperative one consisting of replicated plantations at various locations throughout the range of the species. Would it be possible for you to collect or coordinate a collect- ing are given on the attached material. All information obtained fran the study will be shared with those who are interested and progress reports will be issued periodically to keep you abreast of the research results. We will certainly appreciate any assistance you can provide in this effort. Sincerely, Michael Gold Graduate Student MSU mzjs 148 Dear Fellow Nut Growers: I recently had the pleasure of attending the 70th Annual N.N.G.A. meetings in Wooster, Ohio. Amng the various topics discussed was an interesting talk on Agri-silviculture (Forest farming) based on the tree crop idea of Dr. J. Russell Smith. One of the species unst often mentioned in the context of tree crops is the Honeylocust. I am aware of the N.N.G.A. interest in locating the "superior" selections of nut trees in the hopes of finding improved varieties for the northern areas. I would, therefore, like to ask you to contribute a bit of yom: time to a similar research project which we are undertaking at Miclfigan State University. A further explanation of the research is enclosed. Hope to hear from you. Sincerely, Michael Gold Dunner N.N.G.A. Graduate Student NBU Mkjs Ehc. 149 muons For Collecting Pods Pods cmbeharvestedinthefallchrringtheperiodofnatm'al ripening and will usually yield viable seed. Harvest of the current season's crop is preferred. Ground collections from directly under a tree are acceptable. FranS to 250runrepodsmaybecollectedfromeachof5 to 10 trees in my location. This is merely a guide: my umber of pods, trees, and of course, locations will be useful to us. Pods fran individual trees should be kept separate if possible, thm labeled, packaged, and mailed by air or regular mail to the following address: Prof. J. W. Hanover Dept. of Forestry mchigm State University East Lansing, Michigm 48824 Ship C.C.D. if you wish or indicate shipping charges and we will be happy to reinburse you. Please couplete the enclosed collection information form for each location and mail it with the pods. 'Ihanka very much for your cooperation! 150 HONEYLOCUST RANGEWIDE STUDY Collection Information Collector's flame and Address Collection Date: Date collected / I County Location where collections made (landmarks, roads, etc.) Section Township Range Natural or planted Seed crop:Light Medium Heavy Thorns:Present Absent Site Description:Drainage Slope Soil Tree No. Approx. Height D.B.H. Approx. age (feet) 1. 2. 3. 4. S. 6. 7. 9. 1°. Associated Species Additional Information? 151 Thanks again for your cooperation in the Honeylocust rmgewide collection project. The magnitude of the response is rapidly approaching our goal of 200 sources, broad enough to conduct a thorough evaluation of the genetic variation within Gleditsia triacmthos . Research results will be issued periodically in the form of Progress Reports and plmt materials will be made available as they are developed. We certainly appreciate your assistmce in this effort. Sincerely, Michael Gold Graduate Student PSU M3:js APPENDIX B ACCESSION RECORD FOR INITIAL HONEYLOCUST RANGEWIDE COLLECTION 152 153 Appendix B. Accession record for initial honeylocust rangewide collection. HONEYLOCUST RANGEWIDE STUDY Collection Information Species: Gleditsia triacanthos Genus code: 43 Species code: 33 Accession State of County of Latitude Longitude no.* origin origin 003 GA CHATTOOGA 34o.18'N 850.10'W 004 GA CHATTOOGA 34o.18'N 850.10'W 005 GA CHATTOOGA 34o.18'N 850.10'W 006 GA CHATTOOGA 34o.18'N 850.10'W 007 GA CHATTOOGA 34o.18'N 850.10'W 008 GA MONROE 330.02'N 83o.58'W 009 GA MONROE 33o.02'N 83o.58'W 010 GA MONROE 33o.02'N 83o.58'W 011 GA MONROE 330.02'N 830.58'W 012 GA MONROE 330.02'N 83o.58'W 013 GA MONROE 330.02'N 830.58'W 014 GA OGLETHORPE 330.31'N 84o.41'W 015 GA OGLETHORPE 330.31'N 840.41'W 016 GA OGLETHORPE 330.31'N 840.41'W 017 GA CLARKE 330.57'N 83o.24'W 018 GA OCONEE 330.51'N 830.26'W 019 GA MORGAN 33o.36'N 830.38'W 020 GA MORGAN 33o.36'N 830.38'W 021 GA MORGAN 33o.36'N 830.38'W 022 GA WALTON 330.47'N 830.43'W 023 GA MARION 320.18'N 840.32'W 024 GA MARION 320.18'N 840.32'W 025 GA TAYLOR 320.33'N 84o.16'W 026 GA BALDWIN 33o.04'N 830.13'W 027 GA PUTNAM 33o.20'N 83o.24'W 028 GA PUTNAM 33o.20'N 83o.24'W 029 GA PUTNAM 33o.20'N 83o.24'W 030 GA JASPER 330.19'N 830.41‘W 031 GA JASPER 330.19'N 830.41'W 032 GA HANCOCK 330.17'N 820.58'W 033 GA HANCOCK 33o.l7'N 820.58'W 034 GA HANCOCK 330.17'N 820.58'W 035 GA HANCOCK 330.17'N 820.58'W 036 GA STEPHENS 340.34'N 83o.21'W 037 GA STEPHENS 340.34'N 83o.21'W Appendix B 038 039 040 041 042 043 044 045 046 047 048 049 050 051 052 053 054 055 056 057 058 059 060 061 062 063 064 065 066 067 068 069 070 071 072 073 074 075 076 077 078 079 080 081 082 083 084 085 086 087 (Cont'd.) GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA KY KY KY 154 STEPHENS STEPHENS HABERSHAM JASPER PUTNAM PUTNAM JASPER PICKENS BARTON BARTON JONES JONES JONES JONES JONES WARREN WARREN MCDUFFIE MCDUFFIE MCDUFFIE BEN HILL BULLOCH BULLOCH BULLOCH COWETTA MERIWEATHER MERIWEATHER HEARD NEWTON NEWTON NEWTON NEWTON NEWTON NEWTON NEWTON NEWTON NEWTON HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY PENDLETON SCOTT OWEN o 34 .34'N 34o.34'N 34o.36'N 33o.l9'N 33o.20'N 33o.20'N 33o.l9'N 340.28'N 340.22'N 340.22'N 330.01'N 33o.01'N 33o.01'N 330.01'N 330.01'N 33o.23'N 33o.23'N 330.28'N 33o.28'N 33o.28‘N 310.43'N 320.28'N 320.28‘N 320.28'N 33o.23'N 33o.01'N 330.01'N 330.13'N 33o.35'N 330.35'N 330.35'N 330.35'N 33o.35'N 33o.35'N 33o.35'N 33o.35'N 33o.35'N 33o.22'N 33o.22'N 33o.22'N 330.22'N 33o.22'N 33o.22'N 330.22'N 33o.22'N 330.22'N 33o.22'N 380.48'N 380.20‘N 380.27'N o 83 .Zl'W 83o.21'W 83o.32'W 83o.4l'W 830.24'W 83o.24'W 830.41'W 84o.27'W 840.42'W 840.42'W 83o.33'W 830.33'W 830.33'W 83o.33'W 83o.33'W 820.40'W 820.40'W 820.31'W 820.31'W 820.31'W 83o.l6'W 810.47'W 810.47'W 810.47'W 840.48'W 84o.50'W 84o.50'W 84o.50'W 830.52'W 830.52'W 830.52'W 830.52'W 830.52'W 830.52'W 830.52'W 830.52'W 830.52'W 84o.06'W 84o.06'W 84o.06'W 84o.06'W 84o.06'W 84o.06'W 84o.06'W 84o.06'W 84o.06'W 84o.06'W 84o.23'W 840.49'W 840.49‘W Appendix B 088 089 090 091 092 093 094 095 096 097 098 099 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 (Cont'd.) KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY KY OK OK OK OK OK OK OK OK OK OK OK OK OK TX TX TX TX 155 OWEN FRANKLIN FRANKLIN FRANKLIN FRANKLIN FRANKLIN GARRARD GARRARD GARRARD GARRARD JESSAMINE JESSAMINE JESSAMINE WARREN HARDIN LYON LYON LYON LYON HICKMAN HICKMAN HICKMAN HICKMAN HICKMAN BATH BATH FLEMING FLEMING FLEMING FLEMING FLEMING GREENUP WASHITA WASHITA GARFIELD GARFIELD GARFIELD GARFIELD MCCURTAIN MCCURTAIN MCCURTAIN MCCURTAIN MCCURTAIN MCCURTAIN MCCURTAIN POLK POLK POLK POLK o 38 .27'N 380.11'N 380.11'N 380.11'N 380.11'N 380.11'N 37o.42'N 37o.42'N 37o.42'N 37o.42'N 370.52'N 370.52'N 370.52'N 360.95'N 370.35'N 360.57'N 36o.57'N 360.57'N 36o.57'N 36o.45'N 36o.45'N 360.45'N 360.45'N 36o.45'N 380.04'N 380.04'N 380.20'N 38o.20'N 380.20'N 380.20'N 380.20'N 380.34'N 350.21'N 350.21'N 360.24'N 36o.24'N 360.24'N‘ 360.24'N 330.54'N 330.54'N 330.54'N 330.54'N 330.54'N 330.54'N 330.54'N 300.42'N 300.42'N 300.42'N 300.42'N o 84 .49'W 84o.53'W 84o.53'W 84o.53'W 84o.53'W 84o.53'W 84o.34'W 84o.34'W 84o.34'W 84o.34'W 84o.34'W 84o.34'W 84o.34'W 860.25'W 850.49'W 87o.56'W 87o.56'W 87o.56'W 87o.56'W 89o.06'W 89o.06'W 89o.06'W 89o.06'W 89o.06'W 830.43'W 830.43'W 830.39'W 83o.39'W 83o.39'W 83o.39'W 830.39'W 820.52'W 98o.39'W 98o.39'W 97o.54'W 970.54'W 970.54'W 970.54'W 94o.50‘W 94o.50'W 94o.50'W 94o.50'W 94o.50'W 94o.50'W 94o.50'W 94o.58'W 94o.58'W 94o.58'W 94o.58'W Appendix B 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 (Cont'd.) TX TX TX TX TX TX TX TX TX TX TX TX TX TX TX TX TX TX TX TX DC DC DC DC PA PA PA PA PA PA PA PA LA LA LA LA LA LA LA MS MS MS MS MS MS MS MS MS MS 156 CHERROKE CHERROKE ANDERSON ANDERSON ANDERSON ANDERSON CHERROKE CHERROKE CHERROKE CHERROKE HENDERSON WASHINGTON WASHINGTON WASHINGTON WASHINGTON WASHINGTON WASHINGTON WASHINGTON WASHINGTON WASHINGTON LANCASTER HUNTINGTON HUNTINGTON HUNTINGTON HUNTINGTON HUNTINGTON HUNTINGTON HUNTINGTON EAST EAST EAST EAST EAST EAST EAST CARROLL CARROLL CARROLL CARROLL CARROLL CARROLL CARROLL OKTIBBEHA OKTIBBEHA OKTIBBEHA WINSTON WINSTON OKTIBBEHA OKTIBBEHA OKTIBBEHA NOXUBEE OKTIBBEHA o 31 .59'N 310.59'N 310.45'N 310.45'N 310.45'N 310.45'N 320.59'N 320.59'N 320.59'N 320.59'N 320.12'N 300.39'N 300.39'N 300.39'N 300.39'N 300.39'N 300.39'N 300.12'N 300.12'N 300.12'N 380.55'N 380.55'N 380.55'N 380.55'N 400.01'N 400.23'N 400.23'N 400.23'N 400.23'N 400.23'N 400.23'N 400.23'N 320.48'N 320.48'N 320.48'N 320.48'N 320.48'N 320.48'N 320.48'N 330.27'N 33o.27'N 330.27'N 330.30'N 33o.30'N 330.30'N 33o.30'N 33o.30'N 33o.30'N 330.30'N o 95 .l9'W 950.19'W 950.39'W 950.39'W 950.39'W 950.39'W 950.55'W 950.55‘W 950.55'W 950.55'W 9So.Sl'W 960.24'W 960.24'W 960.24'W 960.24'W 960.24'W 960.24'W 96o.37'W 960.37'W 96o.37'W 77o.00'W 77o.00'W 77o.00'W 77o.00'W 760.19'W 77o.54'W 77o.54'W 77o.54'W 77o.54'W 77o.54'W 77o.54'W 77o.54'W 9lo.10'W 9lo.10'W 9lo.10'W 9lo.10'W 9lo.10'W 9lo.10'W 9lo.10'W 880.50'W 88o.50'W 880.50'W 880.50'W 88o.50'W 88o.50'W 880.50'W 880.50'W 880.50'W 88o.50'W Appendix B 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 (Cont'd.) MS SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC SC NC NC SC SC SC SC SC SC SC SC SC SC SC SC SC SC 157 OKTIBBEHA UNION UNION YORK YORK YORK ABBEVILLE ABBEVILLE ABBEVILLE ABBEVILLE ABBEVILLE ABBEVILLE PICKENS PICKENS OCONEE OCONEE OCONEE OCONEE ANDERSON PICKENS ANDERSON ANDERSON PICKENS PICKENS FAIRFIELD FAIRFIELD FAIRFIELD FAIRFIELD FAIRFIELD CHESTER CHESTER CHESTER CHESTER GASTON GASTON YORK YORK YORK YORK YORK YORK YORK UNION UNION UNION UNION UNION AIKEN AIKEN o 33 .30'N 34o.42'N 34o.42'N 34o.59'N 34o.59'N 34o.59'N 34o.00'N 34o.00'N 34o.00'N 34o.00'N 34o.00'N 34o.00'N 34o.37'N 34o.37'N 34o.53'N 34o.53'N 34o.53‘N 34o.53'N 34o.53'N 34o.30'N 34o.53'N 34o.53'N 34o.37'N 34o.37'N 34o.00'N 34o.00'N 34o.00'N 34o.00'N 34o.00'N 34o.43'N 34o.43'N 34o.43'N 34o.43'N 350.14'N 350.14'N 34o.59'N 34o.59'N 34o.59'N 34o.59'N 34o.59'N 34o.59'N 34o.59'N 34o.42'N 34o.42'N 34o.42'N 34o.42'N 34o.42'N 330.34'N 330.34'N o 88 .50'W 810.37'W 810.37'W 810.14'W 810.14'W 810.14'W 820.14'W 820.14'W 820.14'W 820.14'W 820.14‘W 820.14'W 820.50'W 820.50'W 820.58'W 820.58'W 820.58'W 820.58'W 820.58'W 820.39'W 820.58'W 820.58'W 820.50'W 820.50'W 810.00'W 810.00'W 810.00'W 810.00'W 810.00'W 810.14'W 810.14'W 810.14'W 810.14'W 810.12'W 810.12'W 810.14'W 810.14'W 810.14'W 810.14'W 810.14'W 810.14'W 810.14'W 810.37'W 810.37'W 810.37'W 810.37'W 810.37'W 810.44'W 810.44'W Appendix B 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 (Cont'd.) SC SC SC SC SC SC SC SC SC KS KS KS KS KS KS KS FL LA LA LA LA LA LA LA LA LA LA LA WV IN IN IN IN IN IN IN IN IN AL AL OH OH OH OH OH OH OH OH OH 158 AIKEN AIKEN SALUDA LEXINGTON CHESTER CHESTER YORK YORK YORK CHAUTAUQUA CHAUTAUQUA FRANKLIN FRANKLIN FRANKLIN JEFFERSON JEFFERSON BAY BOSSIER BOSSIER BOSSIER LINCOLN LINCOLN BIENVILLE BIENVILLE BIENVILLE BIENVILLE BIENVILLE NATCHITOCHES MONONGALIA PERRY PERRY PERRY OWEN OWEN OWEN LAWRENCE LAWRENCE GRANT LAWRENCE BIBB VANWERT VANWERT WARREN WARREN DELAWARE DELAWARE DELAWARE DELAWARE DELAWARE o 33 .34'N 33o.34'N 33o.56'N 330.56'N 34o.43'N 34o.43'N 350.01'N 350.01'N 350.01'N 37o.05'N 37o.05'N 37o.45'N 37o.45'N 37o.45'N 39o.12'N 39o.12'N 300.10'N 320.31'N 320.31'N 320.31'N 320.32'N 320.32'N 320.33'N 320.33'N 320.33'N 320.33'N 320.33'N 310.52'N 39o.38'N 37o.56'N 37o.56'N 37o.56'N 39o.18'N 39o.18'N 39o.18'N 380.56'N 380.55'N 400.33'N 34o.28'N 320.57'N 400.43'N 400.43'N 39o.26'N 39o.26'N 400.23'N 400.23'N 400.23'N 400.23'N 400.23'N o 81 .44'W 810.44'W 810.33'W 810.30'W 810.14'W 810.14'W 810.18'W 810.18'W 810.18'W 960.31'W 96o.3l'W 950.10'W 950.10'W 950.10'W 950.33'W 950.33'W 850.41'W 930.44'W 930.44'W 930.44'W 920.39'W 920.39'W 920.56'W 920.56'W 920.56'W 920.56'W 920.56'W 930.12'W 790.57'W 860.46'W 860.46'W 860.46'W 860.46'W 86o.46'W 860.46'W 860.22'W 860.37'W 850.40'W 87o.l8'W 87o.ll'W 84o.06'W 84o.06'W 840.12'W 840.12'W 820.57'W 820.57'W 820.57'W 820.57'W 820.57'W Appendix B 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 (Cont'd.) OH WI VA VA VA VA VA VA VA VA VA VA VA VA VA VA VA IL IL IL IL IL IL IL IL IL IL IL IL IL IL NB NB NB NB NB NB IA AR AR AR AR AR AR AR AR AR AR MO 159 CHAMPAIGN DANE AUGUSTA AUGUSTA AUGUSTA AUGUSTA AUGUSTA AUGUSTA AUGUSTA GILES GILES GILES GILES GILES GILES GILES FAIRFAX OGLE OGLE OGLE OGLE OGLE OGLE DU PAGE DU PAGE COOK DU PAGE LOGAN LOGAN LOGAN LOGAN CASS CASS CASS CASS CASS CASS POTTAWATTAMIE MONROE MONROE MONROE ARKANSAS MONROE MONROE MONROE MONROE MONROE PRAIRIE TEXAS o 40 .04'N 430.04'N 380.10'N 380.10'N 380.10'N 38o.10'N 380.10'N 380.10'N 38o.10'N 37o.l9'N 37o.l9'N 37o.l9'N 37o.l9'N 37o.l9'N 37o.l9'N 37o.l9'N 38o.51'N 420.01'N 420.01'N 420.01'N 420.01'N 420.01'N 420.01'N 410.47'N 410.47'N 410.48'N 410.52'N 400.10'N 400.10'N 400.10'N 400.10'N 410.00'N 410.00'N 410.00'N 410.00'N 410.00'N 410.00'N 410.14'N 34o.4l'N 34o.4l'N 34o.4l'N 34o.4l'N 34o.4l'N 34o.4l'N 34o.4l'N 34o.4l'N 34o.4l'N 34o.4l'N 37o.31'N o 83 .34'W 89o.22'W 79o.95'W 79o.95'W 79o.95'W 79o.95'W 79o.95'W 79o.95'W 79o.95'W 800.39'W 800.39'W 800.39'W 800.39'W 800.39'W 800.39'W 800.39'W 77o.l9'W 89o.21'W 89o.21'W 89o.21'W 89o.21'W 89o.21'W 89o.21'W 880.00'W 88o.00'W 870.49'w 88o.00'W 89o.21'W 89o.21'W 89o.21'W 89o.21'W 950.52'W 950.52'W 950.52'W 950.52'W 950.52'W 950.52'W 950.54'w 9lo.19'W 9lo.19'W 9lo.19'W 9lo.19'W 9lo.19'W 9lo.19'W 9lo.19'W 9lo.19'W 9lo.19'W 9lo.19'W 910.51'W Appendix B 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 Y61 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 (Cont'd.) M0 M0 M0 M0 M0 M0 M0 M0 M0 CO TN TN TN TN TN TN TN TN TN TN TN TN TN TN TN GA IA IA IA IA IA IA IA IA IA IA IA OH KY KY KY KY WV NY NY NY NY NY NY 160 TEXAS TEXAS TEXAS TEXAS TEXAS PHELPS PHELPS PHELPS PHELPS LARIMER UNION UNION UNION UNION UNION UNION UNION UNION UNION UNION UNION WASHINGTON WASHINGTON WASHINGTON WASHINGTON NEWTON BOONE BOONE BOONE STORY STORY STORY STORY STORY STORY STORY STORY ERIE OHIO HOPKINS CHRISTIAN BUTLER MONONGALIA SARATOGA SARATOGA ALBANY ALBANY ALBANY TOMKINS o 37 .31'N 37o.31'N 37o.31'N 37o.31'N 37o.31'N 37o.56'N 37o.56'N 37o.56'N 37o.56'N 400.35'N 36o.12'N 360.12'N 36o.12'N 360.12'N 360.12'N 360.12'N 360.12'N 36o.12'N 360.12'N 360.12'N 36o.12'N 360.12'N 360.12'N 360.12'N 36o.12'N 330.35'N 420.02'N 420.02'N 420.02'N 420.02'N 420.02'N 420.02'N 420.02'N 420.02'N 420.02'N 420.02'N 420.02'N 410.27'N 37o.34'N 370.16'N 360.50'N 37o.09'N 39o.38'N 430.16'N 430.16'N 420.40'N 420.40'N 420.40'N 420.23'N o 91 .Sl'W 9lo.51'W 9lo.51'W 9lo.51'W 9lo.51'W 9lo.55'W 9lo.55'W 9lo.55'W 9lo.55'W 1050.05'W 830.50'W 830.50'W 830.50'W 830.50'W 830.50'W 830.50'W 830.50'W 830.50'W 830.50'W 830.50'W 830.50'W 820.42'W 820.42'W 820.42'W 820.42'W 830.52'W 93o.33'W 93o.33'W 93o.33'W 930.33'W 930.33'W 930.33'W 93o.33'W 930.33'W 93o.33'W 930.33'W 93o.33'W 820.42'W 860.30'W 87o.31'W 87o.30'W 860.54'W 79o.57'W 73o.36'W 73o.36'W 730.49'W 730.49'W 730.49'W 760.32'W Appendix B 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 431 (Cont'd.) NY NY IL MO MS NC NC NC OH CT SD SD SD SD SD SD SD SD SD SD GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA GA MD IL IL NM NM OH OH OH OH OH OH NY 161 TOMKINS TOMKINS ROCK ISLAND IRON SHARKEY BURKE BURKE BURKE LORAIN FAIRFIELD HUGHES HUGHES CORSON CORSON CORSON CORSON CODINGTON CODINGTON CODINGTON DUEUL STEWART STEWART STEWART STEWART STEWART RANDOLPH RANDOLPH HARRIS HARRIS HARRIS HARRIS HARRIS HARRIS HARRIS HARRIS HARRIS HARRIS PRINCE GEORGES COOK COOK BERNALILLO BERNALILLO DEFIANCE DEFIANCE DEFIANCE WILLIAMS HENRY DEFIANCE TOMKINS o 42 .23'N 420.23'N 4lo.25'N 37o.42'N 320.55'N 350.45'N 350.45'N 350.45'N 410.22'N 410.07‘N 440.23'N 440.23'N 450.31'N 450.31'N 450.31'N 450.31'N 440.54'N 440.54'N 440.54'N 44o.34'N 320.03'N 320.03'N 320.03'N 320.03'N 320.03'N 310.50'N 310.50'N 320.44'N 320.44'N 320.44'N 320.44'N 320.44'N 320.44'N 320.44'N 320.44'N 320.44'N 320.44'N 380.57'N 4lo.38'N 410.38'N 350.05'N 350.05'N 4lo.17'N 4lo.17'N 4lo.17'N 4lo.30'N 4lo.17'N 4lo.17'N 420.26'N o 76 .32‘W 760.32'W 900.34'W 900.53'W 900.54'W 810.47'W 810.47'W 810.47'W 820.06'W 730.25'W 1000.20'W 1000.20'W 1000.25'W 1000.25'W 1000.25'W 1000.25'W 97o.08'W 97o.08'W 97o.08'W 960.52'W 84o.49'W 84o.49'W 84o.49'W 84o.49'W 84o.49'W 84o.52'W 84o.52'W 84o.54'W 84o.54'W 84o.54'W 84o.54'W 84o.54'W 84o.54'W 84o.54'W 84o.54'W 84o.54'W 84o.54'W 760.56'W 87o.40'W 87o.40'W 1060.38'W 1060.38‘W 84o.21'W 84o.21'W 84o.21'W 84o.34'W 84o.21'W 84o.21'W 760.30'W 162 Appendix B (Cont'd.) o o 432 NY SENECA 42 .35'N 76 .35'W 433 NY SENECA 420.35'N 760.35'W 434 NY SENECA 420.35'N 760.35'W 435 NY SENECA 420.35'N 760.35'W 436 NY SENECA 420.35'N 760.35'W 437 IA POLK 4lo.44'N 93o.36'W 438 IA POLK 4lo.44'N 930.36'W 439 IA POLK 4lo.44'N 930.36'W 440 IA POLK 4lo.44'N 93o.36'W 441 IA POLK 4lo.44'N 93o.36'W 442 IA POLK 4lo.44'N 93o.36'W 443 IA POLK 4lo.44'N 930.36'W 444 IA POLK 4lo.44'N 930.36'W 445 IA POLK 4lo.44'N 930.36'W 446 IA POLK 4lo.44'N 93o.36'W 447 MI INGHAM 420.45'N 840.30'W 448 MI INGHAM 420.45'N 84o.30'W 449 NB DOUGLAS 4lo.15'N 96o.00'W 450 CO LARIMER 400.35'N 1050.05'W 451 WV MONONGALIA 39o.38'N 79o.57'W 452 WV MONONGALIA 39o.38'N 79o.57'W 453 WV MONONGALIA 39o.38'N 79o.57'W 454 WV MONONGALIA 39o.38'N 79o.57‘W 455 WV MONONGALIA 39o.38'N 79o.57'W 456 WV MONONGALIA 39o.38'N 79o.57'W 457 WV MONONGALIA 39o.38'N 79o.57'W 458 WV MONONGALIA 39o.38'N 79o.57'W 460 NY ST. LAWRENCE 440.40'N 750.01'W 461 IL PIATT 39o.48'N 880.37'W 462 IL PIATT 39o.48'N 880.37'W 463 IL PIATT 39o.48'N 880.37'W 464 IL PIATT 39o.48'N 880.37'W 465 IL PIATT 39o.48'N 880.37'W 466 IL PIATT 39o.48'N 880.37'W 467 IL PIATT 39o.48'N 880.37'W 468 IL PIATT 39o.48'N 880.37'W 469 IL PIATT 39o.48'N 880.37'W 470 IL PIATT 39o.48'N 880.37'W 7* Add 43,330,000 to all accession numbers APPENDIX C PLANTATION MAPS 163 MICHCOTIP Plantation No.-82 Rangewide provenance/half-sib progeny :ontaineri:ed seedling stock planted hare :ultiplanter. 7 x 8' spacing. Io: Randomi:ed Sleditsia triacanthos E. Lansing, H1. test, Water Quality research area. E. Lansing, Mi. root by Hose, Miller. Cold. et. al. on 4/12 - 4/15 1 3 block design with 3 replications, and 4-tree plots. 1 T24? 03d 4-0-3 2 xith flea :ccified ,- .:N, ilk, section 6) :ote: add 43,330.000 to all numbers to get complete accession number. Species code-4333. : COWS-IVS 1 2-3 6-9 10-13 14-17 13-21 22-25 26-29 30-33 34-37 38-41 42-45 46-49 50-53 54-57 55-61 62-65 66-69 ‘0-73 74.77 ‘S x 45“ 427 360 21: 314 3' 33" :03 :63 .232 2:3 :16 2:0 13 3' :10 x x 175 170 123 350 151 3‘0 233 32 398 257 73 439 542 29 358 150 x x 140 88, 206 298 8 369 268 321 54 194 306 355 297 455 59 106 x x 193 157 2: 30 168 105 33 389 424 255 469 373 27 161 378 15 x x 114 172 :95 406 52 365 109 159 221 368 214 229 290 434 334 :61 x x 163 384 64 78 181 186 334 326 447 333 452 117 348 31 398 222 x x 70 467 386 385 218 224 311 424 397 251 274 16 409 48 244 146 x x 401 178 276 449 309 377 7 72 451 376 393 307 82 374 466 416 x x 66 85 68 118 21 435 139 69 351 141 409 208 319 174 337 49 x x 148 280 303 434 65 108 310 304 41 89 187 184 179 237 225 162 x x 445 235 414 37 62 462 156 331 93 299 441 362 339 76 17 320 x x 396 292 67 86 238 262. 135 84 98 104 112 165 74 236 269 248 x x 275 394 361 143 259 34 301 167 2 134 296 92 335 291 58 71 x x 101 456 28 410 425 195 24 330 431 61 119 423 155 349 463 420 x x 322 154 26 382 395 300 325 446 381 180 324 341 421 177 450 120 x x 282 213 273 363 234 290 153 364 461 185 316 359 246 278 27 443 x x 457 437 283 412 152 436 442 465 302 419 312 260 338 303 429 468 x x 55 53 209 392 332 371 345 305 247 171 366 329 258 103 470 56 x x 111 182 77 147 27 323 339 427 107 17 23 169 340 122 308 27 x x 36 ,x x x x x x x x x x x x x x " x x 178 2 136 225 93 330 141 147 74 134 108 290 7 112 257 326 xxxx x x p 156 114 88 134 117 75 48 58 7 423 424 444 330 401 296 391 135 x x 276 437 393 54 85 139 184 425 . 34 70 157 342 384 58 470 x 66 174 123 108 307 69 2 244 69 383 282 364 449 71 224 x 140 112 86 414 320 25 106 311 383 29 268 310 66 261 60 x 248 91 401 351 451 98 61 163 393 247 115 37 61 - 86 x 7 115 53 257 152 386 380 298 174 186 305 136 427 118 209 x 24 403 416 445 68 406 376 457 184 382 194 304 283 55 249 x 440 303 160 23 331 89 224 233 234 456 361 229 441 394 339 x , 328 456 321 426 278 429 427 258 - 297 299 7 362 154 100 151 x 277 431 122 2 209 283 124 249 194 469 379 452 269 389 420 161 23 7 :81 x 37 326 247 7 285 107 391 S6 395 251 444 458 27 53 233 273 461 306 17 x 348 133 148 30 238 458 361 349 55 299 333 110 453 213 280 329 386 180 238 x 77 7 182 457 282 261 398 8 410 195 273 165 370 48 387 148 1‘2 28 146 x 422 443 425 63 84 74 382 302 186 90 465 107 308 455 2 147 246 103 465 x 49 332 17 384 59 291 49 7 250 325 338 312 168 119 443 302 262 450 368 x 162 214 17 295 464 32 262 311 322 165 345 333 365 439 396 316 468 109 409 x 419 280 193 7 297 362 441 468 296 310 52 371 193 380 159 176 :06 403 177 x 396 33 2: 157 :6 433 177 16 381 146 316 34 469 339 63 448 221 2 84 x. 143 208 273 34 2 370 329 292 404 104 462 120 260 214 464 349 17 301 447 x 411 306 167 387 120 339 208 17 100 159 450 412 185 62 181 322 421 179 303 x 434 206 259 235 323 234 180 237 421 435 433 33 27 334 64 369 117 292 323 x 223 244 333 2 7 366 110 :7 300 105 364 423 162 446 295 7 374 233 11: 133 x 2 383 390 154 169 158 69 385 161 185 363 435 77 182 42 352 379 354 49 x 354 73 229 436 153 213 172 153 111 467 37 377 463 406 30 381 12 52 :07 x 166 109 31 176 171 272 150 470 35 424 371 41 56 67 319 26 414 143 404 x :xxx xxxx xxxx xxxx xxxx xxxx :33 420 339 :5: 1:9“ - ' :51 125 :31 341 466 166 x x 439 41: 103 163 218 442 305 428 404 140 332 91 248 x x xxxx xxxx xxxx xxxx xxxx 47 ? 4 2 25 88 345 451 105 708 x x 309 85 16 393 274 12 312 445 x x 462 437 248 68 27 178 98 :25 x x 89 276 416 13 104 167 259 93 x x I 325 335 307 42 7 65 3:4 395 337 x x L32;419 91 - so 39 331 321_ x xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx x >~roau. on e.. es use a zone" 8 a. . ~-m m. m ‘9’ 3m 3 u m2 .3 .5 so. .mm .mm mu was as. use umm mo .mg a.» an» m. nee mm. aam on .em .ao um. sea .36 smm mm» mcm e. .m. us as am» a. am~ u.m moo ..e m. as. use .m~ mom a. . u . . : mom .aa I.a~m an. mes mam sue me: ~=u mo mom .2 35 a.» 3m mom a.m mam .q an .39 mum .3. .ma m3 mem .69 mm m. won me: ~em mom am» mm» mm: mm... 3.. .3 3 amm mom sea u.- amm map I .ea .30 amm you any .mm w .mm aid. .3 .3 am: am; ..e u». aam am: um: amq .m. ma. sme .u mm .uo .qm am: .u. me. .aq a.» .em mam .mo .mu noo .mm mum sum ~90 mam .em .so no .em .uq u.m um ea apnea eo.am on.on an xenon ace—nnn. axumoaxnm mean can: «re see seepnnee oc.»»u.eanon - a. neon v.6nu. na.m. nee. scones: my 22:2... 41.9338 nuance—ea snoeesuanexvepmaepve aqeaesw noun. venison «was anesenne. use am.mme.ooe no s.. sear-q. no men ones—ens nausea—ea seamen. nercxzm Mw.. 50.50 manna a. «a muco— aucom oo.oo 9°.qm a. «a an m— auuum no ”a nu— u_a au— 4 uses snares psaxnmaa 4 .o-.m .a-.q .m.u. nn-nm no-no me-um maumq mm-.. men—emu sense 4:». we «so canes; om «so a.eanen.esu .a 4:.a enema. «so . «one 6.5 >noun neon-.eenmuoe neon—_sm sneer 1.eanoe cone noon ex no.n. mos-ewe...e. man.a:a. eveensm. neaaes.uee ¢.enr eea.ae a.nr u «om.—n=n.eau. «.32: 8... . an. . 7— A! mm» N we. .9 «pa no. a.» 09 ~aa a». an. mu —_9 mam nu. was man .mu mun . u am. an u.m ma. mam now. as. was «an mum ”ea mum mom mum ”no use ~99 .9— ”an _mm mum mam mom ~em ”mm on mum me» .me an. mu me. amm ume u.m me. am. .e. uam .ee .em mm» .qm an; mam mo na. mew aum Nmm mam mm: _m_ am: am; a»: 6m mac ~.m use a a. 3:. 3 a .6. .qm/ a.. eu ~mm .mn _em n»: mam Nee use ~am mam ~mm .m as .44 mam Nam .m. as. as: mm. are Ram umm am ~m us. an; mam . a use amm .23. use mam we» a.e mom .66 mm .mm aao mum ... u. see «we ..u sq: sou mum mm. 3:. .mm was mam as; mm mm ~m. _sc