NEON {N EMA-NEE 9-YEAR STUBY ED STATES GEOGRAPNC VAR LARCHH RESULTS OF A m NORYH CENTRAL UNET Thesis for the Degree of M. S. MECH‘GAN STATE UNNERSHY DAN HENRY FARNSWORTH 1 9 B 8 "' i: LIBRAR 1’ was“! Michigan State University "i '5' BINDING IV H0“ & SBNS’ ABSTRACT GEOGRAPHIC VARIATION IN JAPANESE LARCH-- RESULTS OF A 9-YEAR STUDY IN NORTH CENTRAL UNITED STATES by Dan Henry Farnsworth A replicated, 7-origin provenance test of Japanese larch, Larix leptolepis, was planted in 1958 in the Michigan State University nursery. A second replicated test in- cluding those 7 and 15 additional origins was planted in 1959. Seventeen plantations were established in five north central states in 1960 and 1961 using a randomized complete block design with 5 to 12 replicates. An analysis of variance was used to determine if differences occurred among origins, and simple correlations were used to relate characters to each other and to the parental environment. One source each of Larix decidua, E. gmelini, and E. laricina were included with the 7-origin test. E- def cidua grew as fast as E. leptolepis, a. laricina grew slower, and E. gmelini grew even slower. E; laricina was much more frost hardy than E, leptolepis. Japanese larch can grow in north central United States on sites with well-drained loam soils, light weed competition, and freedom from growing-season frosts. Dan Henry Farnsworth On such sites growth rate was rapid, height averaging 4.1 meters at age 10. However, growth was poor if these con— ditions were not met. There was no geographic variation pattern in height, time of leaf growth, or flower production. A geographic variation pattern existed in spring frost resistance and in preparation for winter dormancy. Preparation for winter dormancy was measured as the date when buds were set, leaves changed color, and the amount of injury resulting from early fall frosts and winter temperatures. The order in which sources prepared for winter dormancy was similar in North Central United States plantations. Small differ- ences in foliage color occurred among sources during the growing season in the nursery, but the differences in the 1958 sowing were not duplicated in the 1959 sowing. GEOGRAPHIC VARIATION IN JAPANESE LARCH-- RESULTS OF A 9-YEAR STUDY IN NORTH CENTRAL UNITED STATES BY Dan Henry Farnsworth A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Forestry 1968 ACKNOWLEDGMENTS I wish to express my deepest appreciation to those individuals who have helped me in the completion of this thesis. I am very grateful to Dr. Jonathan Wright, who started the study and gave many hours of his time in guidance, previous measurements, instruction, and helpful comments. I also wish to thank Dr. James Hanover for his many helpful suggestions and personal guidance. I wish to thank my committee for their many helpful com- ments. I am also indebted to the Michigan State staff members and previous students who spent their time mea- suring the Japanese larch plantations, and to the N.-C.51 committee members who so kindly sent measurements from their plantations. I must also thank my typist who put up with a "messy" draft. Last but not least I am extremely grate- ful to my wife who patiently typed all my rough drafts,. and worked as a full time nurse while I attended school. I am also grateful for the research assistantship granted me by Michigan State University which made it possible for me to continue my education for a Master's degree. ii TABLE OF CONTENTS Page ACKNOWLEDGWNTS O O O O O O O O O O O O O O O O 11 LIST OF TELES O O O O O O O O O O O O O O O O O iv LI ST OF FIGURES O O O O O O O O O O O O O O O O v INTRODUCTION O O O O O O O O O O O O O O O O O O l Genetics of larch . . . . . . . . . . . . . 5 OBJECTIVES O O O O O O O O O O O O O O O O O O O 8 MATERIALS AND METHODS . . . . . . . . . . . . . 9 Seed procurement . . . . . . . . . . . . . . 9 Nursery methods . . . . . . . . . . . . 9 Plantation establishment . . . . . . . . . . 12 Measurement methods . . . . . . . . . . . . 16 Statistical methods . . . . . . . . . . . . l6 PERFORMANCE IN NORTH CENTRAL UNITED STATES . . . 19 mortality O O O O O O O O O O O O O O O O O 19 Height O O O O O O O O O O O O O O O O O O O 21 Form O O O O O O O O O O O O O O O O O O O 26 Cone and flower production . . . . . . . . . 28 Leaf growtll O O O O O O O O O O O O O O O O 31 Spring frost damage . . . . . . . . . . . . 32 Growing season foliage color . . . . . . . . 35 Preparation for winter dormancy . . . . . . 37 Disease and insect resistance . . . . . . . 41 REFERENCES CITED O O O O O O O O O O O O O O O O 43 APPENDIX A. Data collected from plantations given as source totals and analysis of variance . . . . . . . 45 APPENDIX B. Analysis of variance involving more than one plantation . . . . . 88 APPENDIX C. Simple correlation coefficient . . 102 iii LIST OF TABLES Table Page 1. Location, climate, and characteristics of 22 parental stands of Larix leptolepis . . . 10 2. Location and mortality of Japanese larch test plantations . . . . . . . . . . . 15 3. Height at age 10 of seven geographic origins of Japanese larch and comparison with seedlots of other species as measured in the 1960 plantations . . . . . . 22 4. Height at age 9, bole form, and flowering of Japanese larch origins planted in 1961 at Kellogg and Russ Forests in southwestern Michigan . . . . . . 23 5. Numbers of cones produced - in 1960 plantations . . . . . . . . . . . . 29 6. Order of growth initiation among 7 sources of Japanese larch sources planted at different plantations . . . . . . 33 7. Damage to new growth by spring frost in 1960 plantations . . . . . . . . . 34 8. Damage to new growth by spring frost in 1961 plantations . . . . . . . . . 36 9. The similarity of the geographic variation pattern of Japanese larch for earliness or lateness of bud set, leaf color change, and amount of winter injury at 1-61 Kellogg, 7—61 Russ Forest, and Michigan State University nursery . . . . . . . . . . . . . 38 10. The amount of damage done by fall frosts to trees with growing buds in Wisconsin . . 40 11. The resistance of Japanese larch to larch sawfly attack and susceptibility to Valsa sp. canker attack in comparison to .tamarack . . . . . . . . . . . . . . . . . . 42 iv LIST OF FIGURES Figure Page 1. Map of the central portion of Honshu Island, Japan, showing the range of Japanese larch and location of seed sources used in this experiment . . . . . . 3 2. Map of North Central United States showing location of test plantations . . . . 14 INTRODUCTION Japanese larch, Larix leptolepis (Sieb. and Zucc.) Gord. is an important forest species. In Japan it consti- tuted 28 percent of the planting stock from 1952 to 1954 (Anon., 1957). It is also important as an exotic in England (6 percent of the conifers planted) and other nothern European countries (MacDonald gt_§1., 1957; Wright, 1963). Larch are seldom planted in the United States. The state of New York plants the most; four percent of its planting stock is larch, mostly Japanese (Eliason, 1965). More foresters, particularly in northeastern United States, are becoming interested in Japanese larch principly because of rapid juvenile growth. Tests by Cook (1948) showed that it outgrew five other conifers in New York. Chandler (1967) found that Japanese larch outgrew EurOpean larch (Larix decidua Mill), Tamarack (E. laricina Duroi, K. Koch), and Western larch (E. occidentalis Nutt.). Japanese larch has been planted only in arboretums and test planta- tions in north central United States. The natural range of Japanese larch is about 200 km. square on the central portion of the Japanese island of Honshu (Figure 1). There it is found on mountains at Figure 1. Map of the central portion of Honshu Island, Japan showing the range of Japanese larch and location of seed sources used in this experiment. 1+ i0 02‘06 01M, 1‘" 0’ "Of" . . ,1 :au elevations from 900 to 2,500 meters. It usually grows in- terspersed among other species or in pure stands where the original vegetation has been destroyed through land slip- page or fire. The most characteristic features of Japanese larch are its light, green, flattened leaves with two white bands below, cones ranging from 1.5 to 3.5 cm. with recurved tips, and reddish brown branchlets. It grows to a maximum height of 35 m. and a maximum diameter of 1.5 m. European larch, tamarack, £° gmelini Rupr., and Japanese larch were present in the 1960 test plantations. Tamarack could be identified by its light brown branchlets, small cones about 1.5 cm. long with 12 to 15 lustrous, con- cave cone scales. European larch was easy to distinguish because of its dark green foliage and dull cones with straight scales. The distinguishing characteristics of E. gmelini were its slow growth and lustrous cones with straight scales. In the native range of Japanese larch, mean annual precipitation ranges from 1330 to 2840 mm. and the mean annual temperature ranges from 3.2° to 6.5°C. In the United States equivalent temperatures would be found in Philadel- phia, Pennsylvania and similar amounts of precipitationFWOuld be found on the Olympia Peninsula in Washington. Japanese larch is a very intolerant tree which at— tains its best growth on moderately drained, limestone soils (Anonn 1957). Aird and Stone (1953) found that sur- vival and growth were very poor on dry or poorly drained sites. Japanese larch is an excellent source of poles and pilings because its wood is very strong, dense, hard, and decay resistant, but it is not quite as desirable for lum— ber due to its tendency to split when nailed. It is as satisfactory as Douglas-Fir for pulp production. Genetics of 1arch.--European larch is the only larch species in which substantial provenance tests have been con- ducted. The Japanese larch provenance tests conducted be- fore Langner's test have been small and unreplicated. The major range wide provenance test is the 1944 IUFRO Eur0pean larch provenance study. The sources are 51 European larch, 2 Siberian larch, 1 Japanese larch, and 1 European x Japanese larch hybrid. Seventeen plantations were established in 9 countries. The major drawback of this test is that the plantations lacked replication. Genys (1960) reported results from two plantations growing in the United States, one in New Hampshire, and one in New York. He found that Polish, Sudeten, and Slovakian origins grew the fastest; low elevation sources from Austria had a medium growth rate; and high alpine sources grew the slowest with the best form. The Japanese larch source and the European x Japanese larch source grew as well as or bet- ter than the best European larch. Schober (1958) summarized the results of six 1944 IUFRO plantations and nine other species and provenance trials of European larch in Europe. He concluded that Japanese larch and Polish, Sudeten, and Slovakian sources of European larch grew the fastest. Also Japanese larch was resistant to larch canker, Dasyscypha willkommii (Har- tig) Rehm. Another range-wide European larch provenance test was initiated in 1958 by Schober to substantiate the find- ings of previous tests. This test includes 69 provenances most of which are European larch, but also includes a few Japanese and European x Japanese larch hybrids. Also prog- enies from a total of 264 individual trees in 29 localities will be tested separately. Twenty plantations are located in 9 countries in northern Europe, and North and South Amer- ica. Two plantations are located near Ann Arbor, Michigan. Early results from a German nursery show that Euro— pean larch provenances from Slovakia and Sudeten are the tallest, and that high alpine sources are the shortest (Schober, 1961). Plus tree selection does not hold much promise in improving larch. Matthews (1960) conducted a diallele cross of nine Japanese larch and three European larch clones. The progenies from Selfed plants expressed strong inbreeding depression, the heritability was low, and the interspecies hybrids were much superior to either parent. Hybridization appears to be the most promising method of improving Japanese larch. The following hybrids have been made (Wright, 1963): 1. Natural hybrids occurring where ranges overlap E. lyallii x E. occidentalis E. potaninii x E. mastersiana E- gmelini x E. sibirica. 2. Natural hybrids from cultivated plants 9. decidua x E, leptolepis L. leptolepis x E. sibirica g. laricina x E. decidua. 3. Artificial hybrids E° decidua x £° occidentalis It" . decidua x . leptolepis IL" It" . decidua x It" gmelini . sibirica It" . decidua x lb H." . leptolepis x E. gmelini. The other combinations have not been tried. The most promising hybrids are those produced by crossing Japanese and European larch. These are the par- ents of the famed Dunkeld larch (E. decidua x L. leptolepis). Hybrids involving Japanese and European larch have been pro- duced by 27 different individuals. The hybrids of Japanese and European larch always perform better than the parent species. OBJECTIVES The objectives of this experiment are as follows: Determine where Japanese larch will grow in north central United States. Determine the amount of genetic variability in growth characters. Determine possible relations of those characters to the parental habitats. Determine the relationship of these characters to each other. Determine if the same results are obtained when the same materials are planted in different locations and in different years. MATERIALS AND METHODS Seed procurement.--The Japanese Forest Service pro- cured seed for Dr. Wolfgang Langner at the Institute of Forest Genetics, Schmalenbeck, Germany. The seeds were collected from 10 or more trees in each of 25 native stands (Figure 1, Table 1). Two sets of seed were received by Michigan State University. The first consisted of 7 origins shipped directly from Japan. The second shipment of 22 sources, including the original 7, was received from Ger- many through State University of New York, College of Forestry. Nursery methods.--The 7-origin test was sown in the Michigan State University nursery May 2, 1958, and the 22- origin test was sown May 14, 1959. The nursery eXperimental design consisted of five randomized replicates in the 1958 sowing and 4 randomized replicates in the 1959 sowing. Plots within the replicates consisted of 124 cm.-long rows apced 15 cm. apart. The remaining seed of each origin was sown in a large rectan— gular block with the blocks being separated by 50-cm. strips. The trees received intensive nursery care. They were kept weed free by hand weeding and watered as needed. The average height of the 2-0 stock was 45 cm. The trees in the large rectangular blocks in the 1958 sowing were not 9 10 as ma ms ommm m.m oosa m.mma m.mm Amma .acamsmw hm hm cc chem c.c cmva «.mma c.mm Avmacmmnm «V cm cc cmmm m.m ccma v.cma c.0m Ammavhcnm amucmz .us av cm cma cvcm c.¢ cccm a.cma «.mm Amma .mvcmcm am ca mm cmna m.c ccma ~.cma m.mm Aamavmmhm mmcmm camucdoz armamxc cm am mm cmma a.m cmba m.mma m.mm Acmavemnm mv cm mm ccma m.c cmva . m.cma m.mm Amaa .cvmmhm mm on cc ccha w.m ccha m.cma c.mm Amaavmchm we mm In cmva a.m ccca m.cma a.cm Abaavamnm me am on ccva v.m cmna v.mma c.cm Amaavcmnm av ma ms cmma a.c cmha v.cma c.mm Amaavcchm mmcmm camucsoz columns» cm ca cw ccma m.c ccma >.cma c.cm Aeaa .mccccm omsnd .uz am. ma mm cmha c.m coca n.mma ¢.mm maavccnm mm ha me cmca N.c cmma >.cma v.mm Aaaavmchm ansm .uz A.EEV 2.00v . A.Eov A.Ev A.mnmv .ownm .Qfimu Amuwumfiv .mco%mmwum0wwma A oz.WMm£v .EMao .um mod amsccm assess .>mam cum: smog “mom suuoz Momncmamecom m.mamvaonmma xanma mo mocmum amucmnmm mm mcu mo moauaHODOMHmco was .mumEaao .coaumooall.a magma 11 m>aumc m scum cmuomaaoo Aacaoauma.m~ aa cam: on techno: some cmuomaaoo Amaamaoumwa .qv oa ohms .1 .camcoomaa .Hmocwamcanm Ham: pcmum «mHOaHH aaaammm aam..mamaa .EdumuomH¢ .qv m ommz .Amscaomc .ac c Ohm: mums mcoaumucMam ummu 050m ca pmosaoca muoacmmm amcoauacocm cm ma cma ccmm m.m cmma h.cma a.mm Ammavmccm ms nu cc cmam m.m ccma c.hma c.mm Avma .hvcccm mm ma cm ccma m.m ccma c.hma «.mm Amma .wvcccm mmcmm camucsoz spam cm cm cma ccmm m.m cmma c.5ma c.mm Ammavnccm mmumaom .uz mm ma cm chma m.w coca m.cma v.cm Ammavmccm mm mm cc ccwa m.c cmva c.cma v.cm Acmavmccm mm va cm cmma N.m ccma m.cma v.cm Acmavacmm madam .uz cw am on ccma m.v cmna m.cma c.mm Amma.mvccmm mammacm .uz A.EEV 1.0 c . A.Eov A.Ev A.mumv .omnm .mEmn AmeumEV .mc0%mmmummmwma A oz.MMm2V .Emaa .um mmé amsccm amsccm .>mam new: new: ummm nunoz Mowncmamficom .GOSGaucooll.a manna 12 as good planting stock as the 1959 material due to the ef- fects of overcrowded beds. The 22 origins included in the 1959 sowing were also grown near Hamburg, Germany, in a design similar to the one used in Michigan. Growth data for those trees, as published by Langner and Stern (1965), have been used where necessary in this paper. I have also used nursery and plantation data for 20 origins identical with mine as tested in New York (Stairs, 1965). Plantation establishment.--The 7-origin test was outplanted in 1960, and the 22-origin test in 1961. 2-0 seedlings were used in both studies. The trees which died the first year were replaced with 3-0 stock in Mich— igan, while 2-1 transplants were planted by the out-of- state co-operators. Plantations were established through- out north central United States (Figure 2, Table 2). The plantation design was a randomized complete block with each plantation consisting of 5 to 12 replicates of 4-tree plots. The trees were planted using a 2.44 meter spacing. Border rows were planted on all sides of the plantation. Plantation care was not very intensive. In most plantations, trees were planted in furrows and then left to grow. Intensive care occurred at the following plantations: (l) at Russ trees bent by snow were straightened, (2) 2,4,5-T was applied to control brush at 2-60 Kellogg 13 Figure 2. Map of north central United States showing the location of the test plantations used in this experiment. l4 no. on o 0. .1 \ .... _. 0.9:: 3.: lilac! o a \ rloflfilo ‘ a cfi (ii _ 4 , , :T x _ , , l I.” _ gag co mm. _ coon.“ T. I «to. .33.) l f y ‘ on? _ . p o——. 0‘. p on. ’00 h — oh—O oO—O 15 Table 2.--Location and mortality of Japanese larch test plantations. . Percent Plggggzion mOrtality in Name County, State 1961 and Y8ar °f or 1967 establishment 1962 10 1960 Dunbar Forest Chippewa, Michigan 11 49 7 1960 Higgins Lake Crawford, Michigan 23 90 11 1960 Manistee Manistee, Michigan 4 8 24 1960 Wexford ‘WeXford, Michigan -- 94 13 1960 Allegan Allegan, Michigan 23 90 6 1960 Rose Lake Shiawassee, Michigan 34 47 2 1960 Kellogg Forest Kalamazoo, Michigan 48 62 l 1961 Kellogg Forest Kalamazoo, Michigan 12 26 4 1960 Russ Forest Cass, Michigan 38 63 6 1961 Russ Forest Cass, Michigan 55 74 7 1961 Russ Forest Cass, Michigan 24 40 21 1960 Flambeau Sawyer, Wisconsin 48 90 22 1960 Oneida Oneida, Wisconsin 17 42 36 1961 Oneida Oneida, Wisconsin 10 66 15 1960 Paint Creek Allamakee, Iowa 8 16 49 1961 Soaper Creek Davis, Iowa 30 30 20 1960 Ohio Wayne, Ohio 48 50 14 1960 Nebraska Cass, Nebraska 38 16 (3) Amitrol-T was applied for weed control at 1-61 Kellogg and 6-61 and 7-61 Russ, (4) at Rose Lake 3 oz. of an equal mixture of 40 percent urea and 5-20-20 fertilizer was applied around each tree plus rototilling was done around each tree. Measurement methods.--I measured the southern Michigan plantations in 1966, 1967, and 1968. Dr. Wright and Michigan State staff made the previous measurements. The plantations outside of Michigan were measured by N.C.-51 co-operators in those states. The Kellogg, Russ, and Rose Lake planta- tions were measured much more intensely due to their prox- imity to the university. Metric traits were measured as the means of 4-tree plots. Presence-absence traits were measured as the number of occurrences per plot. Leaf color change and leaf fall was measured as the percent of the needles remaining on the tree that had changed color plus the percent of fallen needles. Statistical methods.--I used plot means as items when calculating analysis of variance, for which the degrees of freedom and expected mean squares are as fol- lows: 17 Source of Degree of Expected mean variance freedom squares 2 2 2 Source S-1 0 + Ro sxp + RPoS . 2 2 A 2 Plantation P-l o + R0 + RSo sxp p . 2 2 Source x plantation (S-l)(P-l) c + R0 sxp Replicate (within plantation) (R-l)(P) o + S°r(p) Source x replicate (within plantation) (S-l)(R-l)(P) c Total SXPXR-l I used the source x plantation mean square to test signifi- cance of the source mean square. Six mountains were represented by two or more sources and were tested for between-mountain differences. I used the within-mountain mean square to test significance of dif- ferences among mountains. Degrees of freedom and expected mean squares were as follows: Source of Degrees of Expected mean variation freedom squares Sources 18 2 2 2 2 Between mountains 5 ge + Ropxs + Row + Rch . . . 2 2 2 Within mountains 13 0e + Rapxs + Row 18 I used simple correlations or rank correlations when so specified to test whether characters were related to each other and to factors of the parental environment. Seedlot means were used as items. PERFORMANCE IN NORTH CENTRAL UNITED STATES Mortality.--The site requirements of Japanese larch must be closely met or high mortality will result. Those plantations planted on sites with sandy loam to clay loam soils, with light weed competition, and mild frost damage had low mortality, but mortality rapidly increased whenever even one of these conditions was not met (Table 2). Mortality in the 1960 plantations was sometimes high because of the use of spindly stock grown in dense stands in the nursery. Less mortality was obtained in an European larch plantation than in the Japanese larch plan- tations using l-l stock instead of 2—0 stock. At Rose Lake the stock dried out and some dead trees were planted. Mor- tality was almost total in plantations on sites with sandy soils, heavy weed competition, and repeated frosts, or a combination of these factors- At Dunbar, 89 percent of the Japanese larch died which were planted on a poorly drained 19 20 site and suffered a secondary canker attack by Valsa sp. Woodchucks caused much mortality in Ohio by nipping off the tree stems. Sites which would be suitable to plant to pine are not always suitable for planting to Japanese larch. Scotch pine planted near the Japanese larch plantations at Allegan and Higgins Lake have survived and grown well while most of the Japanese larch have died. The climax vegetation of a site may predict the performance of Japanese larch. In Michigan plantations on sites growing northern hardwoods or a hardwood mixture containing red, black, and white oaks, beech, and sugar maple had low mortality, while almost all, the Japanese larch died on scrub oak, jack pine, or tamarack sites. Another important part of the mortality story in Japanese larch is that trees will continue to die after the plantation appears to be established and is growing well (Table 2). Comparison of mortality in plantations containing 280 Japanese larch and 40 tamarack showed that more tamarack remained than Japanese larch in plantations on sandy soils 21 with repeated frost damage and on the poorly drained site. The average mortality was 55 percent for Japanese larch and 79 percent for tamarack at Russ, Kellogg, and Rose Lake plantations. Height~--Japanese larch has exhibited rapid juve- nile growth in the test plantations (Tables 3 and 4). At Kellogg Forest Scotch pine, spruce and Japanese larch provenance test plantations of the same age are adjacent to each other. The tallest Japanese larch trees are about 25 percent taller than the tallest Scotch pine and almost twice as tall as the tallest spruce. Japanese larch has an indeterminate growth pattern in that trees begin growth in May and continue to grow throughout the summer until buds are set in September. Height growth can be strongly influenced by the hot, dry summers that are common in north central United States. Differences in summer cli- mate may explain why 9-year-old plantations in north cen- tral United States are the same height as a 7-year-old one in New York or why 5-year-old trees are only half as tall in Michigan as in Germany. The average plantation height at 10 years of age was 5.9 meters in Nebraska and 4.1 meters at Manistee, 22 Table 3.--Height at age 10 of seven geographic origins of Japanese larch and comparison with seedlots of other species, as measured in the 1960 planta- tions. Height in meters at Origin Number Kellogg Rose Manis- Dun- , , and MiCh . Lake tee bar Neb. WISC. Ohio Mean Speciesa MiCho MiCh. MiCho 2-60 6-60 11-60 10-60 14-60 22-60 20-60 2788 JAP 3.4 4.1 4.1 3.2 6.1 3.2 3.7 4.1 2793 JAP 3.6 4.0 4.3 3.3 6.3 3.5 4.0 4.0 2796 JAP 3.7 3.6 4.6 3.4 6.3 3.9 3.5 4.1 2799 JAP 3.4 4.0 4.0 3.4 5.6 4.2 3.1 4.2 2800 JAP 3.9 4.2 4.6 4.1 5.0 -- 2.4 4.3 2806 JAP 3.9 3.1 3.4 2.6 6.2 3.2 3.3 3.8 2808 JAP 4.0 3.3 3.9 2.9 6.0 3.3 3.3 3.9 Average 3.7 3.8 4.1 3.9 5.9 3.5 3.3 4.1 8 EUR 4.2 —- 4.7 3.9 -- -- -- 4.3 9 GME 1.7 -- 1.7 -- -— -- -- 1.7 10 JAP -- -- -- 4.2 -— -- -- 4.2 11 TAM 2.0 -- 2.7 3.2 -- -- -- 2.6 F values showing significance of between-origin differences for seven Japanese larch sources. .5 1.7 . 2.6* 1.1 25.0** 4.2** 4.6**l.l Including sources 8-11. All species 7.5** -- 16.7** 3.7** -— -- _- 16.7.. aJAPanese larch, EURopean larch, L. GMElini, TAMarack. *Significant at .05 level. **Significant at .01 level. 23 Table 4.--Height at age 9, bole form, and flowering of Japanese larch origins planted in 1961 at Kellogg and Russ Forests in southwestern Michigan. MountainOri in Mean Mean number Trees with flowersa of n ger height :fmcfoogsr female male - - umb s e ea e origin (meters) (per tree) (percent) Fuji 2785 3.2 3.1 2.8 10 18 Fuji 2786 3.3 2.7 2.2 13 24 Azusa 2788 3.1 3.1 2.2 1 l Yatsu-ga 2789 3.3 3.5 2.3 3 9 Yatsu-ga 2790 3.2 3.7 2.5 3 4 Yatsu-ga 2791 3.2 3.5 2.6 1 1 Yatsu-ga 2792 2.7 2.9 2.3 4 5 Yatsu-ga 2793 3.0 3.6 2.1 2 4 Yatsu-ga 2794 3.0 3.3 2.2 l 1 Akaishi 2795 2.9 3.0 2.3 10 16 Akaishi 2796 2.6 3.6 2.2 0 1 Nantai 2797 3.0 3.1 2.3 2 6 Nantai 2798 3.1 2.6 2.1 3 4 Nantai 2799 3.3 2.5 2.4 5 7 Shirane 2800 2.6 2.5 2.1 0 5 Asama 2801 3.0 3.4 2.2 3 8 Asama 2802 2.6 3.4 2.0 19 27 Asama 2803 2.2 2.9 1.7 l2 l3 Koma-ga 2807 2.8 3.6 2.4 O 4 Hida 2806 3.0 2.5 2.6 0 l Hida 2808 3.2 3.5 2.5 0 3 Hida 2809 3.2 4.0 2.6 2 5 F value 1.00 1.08 2.20* 3.64** 1966 at 1.66 aProduced in 1966, 1967, and 1968 at Kellogg and RUSS. *Significant at .05 level. **Significant at .01 level. 24 Michigan. These plantations are on sites with loamy soils and light weed competition. Another factor at Manistee is the strong moderating influence of Lake Michigan on the climate. The average height for all plantations except those with very high mortality was 4.1 meters at 10 years of age. Those plantations with high mortality planted on sandy soils and repeated frost damage averaged 2 meters tall at 10 years of age. Trees grown from seed sown in 1958 averaged 11, 43, and 137 cm. tall at ages 1, 2, and 3; seed sown in 1959 produced trees 9, 48, and 109 cm. tall at the same ages. These differences in height reflect differences in spacing. The 1958-sown trees grew in very dense, uniform rows whereas the 1959-sown trees grew in sparsely stocked and variable rows. Evidently full stocking induced better growth. There,were significant between-source differences in both the 1958 and 1959 nursery experiments. The differ— ences were relatively small and not correlated with latitude, longitude or altitude of origin. The biological meaning of these early differences is very obscure because: 1. 'Results of the 1958 and 1959 experiments were not correlated with each other. 2. Results obtained with the same 22 origins in Mich- igan and German nursery tests were not correlated. l 25 3. Nursery height was not correlated with height in the permanent plantations established with either the 1958 or 1959 material. Therefore the early height data have very limited genetic significance. Within plantations there was a 25 percent difference in height between the tallest and shortest origins. These differences varied from nonsignificant at the .05 level in three plantations to significant.¢at the .001 level in Ne— braska (Table 3 and 4). However, the origins which were tallest in one plantation were not the tallest in another plantation (Table 3). When all plantations were considered as one experiment, differences among seedlots or groups of seedlots from the same mountain were not significant, but plantation x seedlot interaction was (1 percent level) in both the 1960 and 1961 series of plantations. Correlation calculations substantiated this conclusion from analysis of variance. The correlations were only r = .20 and r = .17 between height of an origin in Michigan (age 10) and Iowa (age 7) or Michigan and New York (age 6). The height of an origin in the permanent plantations was not related to altitude, precipitation, mean annual tem- perature, or growth characteristics of the parent stand. Nor did simple correlation indicate that it was related to other measured characteristics such as the amount of spring and fall frost damage or winter injury. 26 The height variability found in Japanese larch is comparable to the differences found in an area of similar size in a species with a larger range, or even to differ- ences found in a single stand of a more variable species. The significant differences at individual plantations which were not repeatable at other plantations bear the following implication to testing the small differences found in plus tree selections of other species; to test the validity of small differences between trees, the plantations should be repeated in different locations and years to check if the same differences will be repeated. Three other species of larch were present in the Michigan plantations established in 1960 (Table 3). Euro— pean larch grew as fast as the Japanese larch. Tamarack grew slower; in relation to Japanese larch it grew slowest in southern Michigan, and fastest farther north. At Dunbar, in Michigan's Upper Peninsula, it outgrew some Japanese larch origins. L. gmelini was slowest growing of all. Form.--I paid special attention to measurement of form because most literature indicated there was a differ- ence among origins, and because Japanese larch is considered to have poorer form than European larch. I measured form as the number of times the stem of the tree departed from a straight line, and recorded separately the number of crooks in the leader and the older portions of the stem. Crooks were counted whether due to bending of the stem or loss of leader. 27 Differences in form among origins were not detectable in the field because within a 4-tree plot there was often one very crooked and three straight trees, giving an appear- ance of more variation within a plot than between plots. An analysis of variance showed no significant dif- ferences among origins in the plantations with 22 origins (Table 4). In the plantations with 7 origins, source 2800 was significantly straighter. It was the second straightest origin in the 22-origin Kellogg plantation, but only average at Russ. The correlation between number of crooks in the main stem and in the leader was r = .28 for the 7-origin plantations and r = .24 for the 22-origin plantations. Therefore the crookedness or straightness of the stem was not due to some genetic factor controlling the growth of the leader, but due to some environmental factor occurring after the completion of growth. Differences in form occurred among the four species present at the Manistee plantation. The European larch source averaged 5.8 crooks per tree to 3.5 for Japanese larch. Tamarack and L. gmelini had a similar number of crooks in the portion of the stem, one year or older, to Japanese larch. L. gmelini averaged only .25 crooks per leader compared to 3.0 crooks per leader in Japanese larch. This difference is probably due to the short leader of L. gmelini,.averaging .2 meters, compared to the Japanese larch average of 1 meter. 28 Cone and flower production.--Tamarack produced cones at an earlier age (8 years at Dunbar) than the other three larch species in Michigan plantations 2-60 (Kellogg), 10-60 (Dunbar), and 11-60 (Manistee). Tamarack continued 'to grow heavier cone crops, producing four times as many cones as all other species combined (Table 5). At Kellogg r3 Forest 2. gmelini first produced one cone in 1966 (age 9), ‘ one more cone in 1967, and none in 1968. EurOpean larch produced one cone at age 8 at Dunbar and no other cones i until age 10. The first cones appeared on Japanese larch j at age 6 in Nebraska, but Japanese larch was the last spe- cies to yield cones in plantations with other species present. Japanese larch flowers appear before leaf develop- ment starts. They are easily visible until the leaves are almost fully grown. Spring frosts often kill the flowers but the dead flowers remain visible on the tree. In Japa- nese larch up to age 11, more trees had male than female flowers, and if a tree had female flowers it also had male flowers. The‘number of female flowers which matured into cones was often low. For example, 134 trees produced flowers in plantations 1-61 (Kellogg) and 7-61 (Russ), but yielded only 31 cones. Flower production varied from year to year. The number of trees with female flowers declined at Rose Lake from 3 in 1967 to 1 in 1968, at 2-60 (Kellogg) from 16 in 1967 to 2 in 1968, and at 1-61 (Kellogg) from 92 in 1967 to 9 in 1968. 29 Table 5.--Numbers of cones produced in 1960 plantations. Source Number Kellogg Rose and Forest Lake Dunbar Manistee Neb. Iowa Wis. Total Speciesa 2788 JAP 0 0 2 0 2 10 o 14 2793 JAP o 0 126 36 1 15 Highb 178 2796 JAP 0 0 0 4 0 0 6 2799 JAP 0 0 21 0 0 2 0 23 2800 JAP 4 15 575 o 978 -- 1574 2806 JAP 0 9 o 0 15 0 24 2808 JAP 0 15 0 3 2 0 26 8 EUR 3 -- 11 124 -- -- -- 138 9 GME 2 —- -- 0 -- -- -- 2 10 JAP -- -- 376 -- -- -- -- 376 11 TAM 135 -- 1700 400 -- ‘-- —- 2235 Years of 1966 1967 1965 1967 1963 1964 1967 recorded to to 1967 1965 cone 1968 1968 to production 1967 aJAPanese larch, EURopean larch, L. GMElini, TAMarack. b24 trees were estimated to produce 5 cones per tree. 30 Heavy cone production occurred in plantations at Dunbar Forest, Michigan and Nebraska for unknown reasons (Table 5). The Dunbar plantation is the slowest growing, wettest, and farthest north; the Nebraska plantation is the fastest growing and most southerly and is planted on a silt loam soil. Of the seven Japanese larch sources in 1960 planta— tions, source 2800 produced 84 percent and source 2793 10 percent of the cones (Table 5). Each of those sources yielded cones in five different plantations. The chance of that occurrence is less than 1 in 100. Only in Iowa did another source fruit more heavily than one of these. Heavy cone production by a particular source was repeated year after year in the same plantation. In Ne- braska 25 to 50 percent of the trees of source 2800 pro- duced over 200 cones each year after 1965 but all trees of other sources produced only 6 cones altogether. The number of cones a source produced was not cor- related with height, time of bud set, or other characters measured in the plantations. For example, source 2800, which flowered most heavily, was shortest in Nebraska and second tallest at Dunbar. In the 22-origin plantations established in 1961 trees of five sources flowered heavily (Table 4). However, sources which produced many female flowers in 1-61 Kellogg did not do so at 7-61 Russ. For example, 10, 10, and 19 31 percent of the trees in sources 2785, 2795, and 2802 pro- duced female flowers at 1—61 Kellogg whereas no trees in these sources produced female flowers at 7-61 Russ. Two sources (2793 and 2800) produced most of the female flowers in the plantations established in 1960, but those same two flowered lightly or not at all in the $1 plantations established in 1961. Such an interaction is L} difficult to explain. Results of this provenance test have an implication for Japanese larch seed orchard establishment. To guaran- E1 tee large quantities of seed one would have to plant many different sources at many different places and hope. Leaf Growth.--Japanese larch is one of the earliest species to begin leaf growth. Near East Lansing, leaves appear three weeks earlier than in most tree species, and up to seven weeks earlier than in some. In a Japanese larch plantation there is a 5-day difference between the day the first tree develops leaves and the day the last tree begins leaf growth. The order in which trees initiated leaves was measured as the amount of leaf development on one date or on certain dates. Leaf growth was measured in two differ- ent years in three plantations. The order in which sources initiated leaves was similar from year to year in the same plantation. 32 Small differences in order of leaf initiation existed between sources at individual plantations (Table 6). However, the order in which sources initiated leaves at one plantation was not significantly correlated with the order in which sources developed leaves at another planta- tion. For example, the correlation for amount of leaf growth between l-61 and 2-60 Kellogg on April 22, 1968 was r = -.45. The order in which species develOped leaves-at Kellogg Forest and L. gmelini, European larch, Japanese larch, and tamarack. L. gmelini developed leaves one week earlier than did tamarack. Spring frost damage.-—Spring frost damage occurred once in the 10 years of the study in southern Michigan and in each year that spring measurements were made in Wisconsin. Japanese larch are much more susceptible to spring frost damage than tamarack. Significant differences in the amount of spring frost damage occurred among sources of Japanese larch. However, the order with which sources suffered damage was not constant from plantation to planta- tion (Table 7). ' The only 22-origin plantations in which spring frost damage was recorded were 6-61 and"7-61. In these plantations, the sources showed a distinct geographic variation pattern. Grouping the origins by mountain of 33 Table 6.--Order of growth initiation among 7 sources of Japanese larch sources planted at different plantations. Earliness (=1) or lateness (=7) of leafygrowth of trees 21233.4 33333: rrrerr 2331132313833 333 2333:: Rank 2808 1 2 l 2 3 4 2788 2 4 2 l 2 3 2799 4 3 6 3 5 l 2796 4 l 7 5 7 2 2793 4 6 3.5 4 6 5 2806 6 7 3.5 7 l 6 2800 7 5 5 6 4 7 34 Table 7.--Damage to new growth by Spring frost in 1960 plantations. Rose Lake Russ Wis. 5/25/63 5/25/63 1964 Species gagggg Rank 1 =, large, 8 = small JAP 2808 l l 2 JAP 2806 3 2 3 JAP 2788 6 3 1 JAP 2796 2 5 4 JAP 2793 4 4 6 JAP 2799 5 6 5 JAP 2800 7 7 - TAM ll 8 8 - F value 2.85* 1.22 26.3** JAPanese larch, TAMarack. *Significant .05 level. **Significant .01 level. 35 parental stand explained 72 percent of the variation due source (Table 8). Spring frost damage was not related to the earliness or lateness of leaf initiation at Rose Lake and 6-61 and 7-61 Russ. The only plantation in which both spring and fall frost damage occurred was Wisconsin. Those sources which were least damaged by spring frost were also least damaged by fall frosts. Growing season foliage color.--Spring and summer foliage colors of Japanese larch were measured in the nursery. The differences in color among origins were so slight they could be distinguished only under the best light conditions and by a person with excellent color per- ception. The person grading color differences stood 6 meters from the plots and quickly read the color grades to the person recording. Colors measured by a different per- son ten minutes later agreed with the first color measure- ments. Five color grades were distinguishable in spring color, grade-l being the yellowest. Differences in spring color among origins were significant at the .01 level in both the 7-origin and 22-origin tests. The yellow origins at age 2 were not the yellowest origins at age 3 in the 7-origin test, nor were the yellowest origins in the 7- origin test the yellowest origins in the 22-origin test. 36 Table 8.—-Damage to new growth by spring frosts in 1961 plantations. .. . ‘6-61 and 7-61 Russ 5/25/63 Mountain EEEEEE. Percent '---- sources '---- Akaishi 2 72 Yatsu-ga 6 63 Azusa l 62 Fuji 2 60 Asama 3 57 Shirane l 56 Hida 3 56 Koma-ga l 55 Nantai 3 39 F value 3.4* *Significant .05 level. 37 Spring color was related to autumn color and winter dieback in the 7-origin test but no similar relationship was found in the 22-origin test. Two summer color grades were measured at age 1 in the 22-origin experiment with grade-l trees having a very slight brown tinge and grade-2 trees lacking a brown tinge. There was no relationship between summer color and any other character measured. Preparation for winter dormancy.-—The earliness with which trees set buds or changed leaf color, the amount of damage done by fall frosts, and winter injury were used as indicators of preparation for winter dormancy. These characters were highly related to each other in all planta- tions (usually at the .01 level), and appeared to be under uniform genetic control which was not strongly influenced by the local environment. Those origins which entered winter dormancy early in the Michigan State University nursery also did so in a nursery in northern Germany (Langner, 1965). When the trees in the Michigan State University were outplanted, the order in which sources prepared for dormancy was constant from nursery to planta- tion, plantation to plantation, and from year to year in the same plantation or between plantations. The geographic variation pattern for preparation for winter dormancy is shown in Table 9. 38 Table 9.--The similarity of the geographic variation pat- tern of Japanese larch for earliness or lateness of bud set, leaf color change, and amount of winter injury at 1-61 Kellogg, 7-61 Russ Forest, and Michigan State University nursery. Preparation for winter dormancy Source Mountain (1=early, 22=13te> Leaf color Winter Bud set change injury Rank 2785 Fuji 21 22 10 2786 Fjui 17 21 14 2788 Azusa 20 17 21.5 2789 Yatsu-ga 10 6 19 2790 Yatsu-ga 19 13 11 2791 Yatsu-ga 11.5 11 9 2792 Yatsu-ga 18 l4 19 2793 Yatsu-ga 11.5 10 14 2794 Yatsu-ga 15 19 19 2795 Akaishi 22 12 14 2796 Akaishi 13 18 7 2797 Nantai 4 4 8 2798 Nantai 6 3 3 2799 Nantai 3 5 6 2800 Shirane 1.5 9 1.5 2801 Asama 1.5 l 1.5 2802 Asama 7 16 14 2803 Asama 9 20 4 2807 Koma-ga 5 7 17 2806 Hida 14 15 14 2808 Hida l6 8 21.5 2809 Hida 8 2 5 39 Bud set is the first phase of dormancy. In Michigan trees start to set buds in the first Week of September, with the last buds being set the last week of September. The only inconsistency in order of bud set occurred in Wisconsin where source 2796 set its buds early but did not do so in plantations at Kellogg, Rose Lake, and Ohio. The amount of fall frost damage was recorded only in the 22-60 Wisconsin plantation where fall frost damage was observed in three out of four years of measurement. Fall frost damage was least on trees which set buds early (Table 10). In Michigan the leaves of Japanese larch began to turn yellow in the middle of September. Leaves nearest the trunk turned yellow first. All leaves turned yellow by the middle of November and dropped by early December. Leaf color change was measured as the combined percentages of the leaves that had changed color and those that had fallen. Trees from Mt. Fuji averaged 25 to 50 percent more green leaves than other sources during the first six weeks in which leaves changed color. Winter damage was recorded in the nursery and at Dunbar. Sources 2799 and 2800 were least injured at Dun— . bar and were also least damaged in the 7-origin test in the nursery. It should also be noted that although the sources from Mt. Fuji set their buds, and their leaves 40 Table 10.--The amount of damage done by fall frosts to trees with growing buds in Wisconsin. Schmalenbeck Number of trees with Number of trees with Number growing buds fall frost damage Percent 2788 35 85 2793 2 8 2796 48 75 2799 6 6 2800 -- -- 2806 17 77 2808 12 56 F value 4.83** 12.7** **Significant at .01 level. 41 changed color much later than other sources, they did not suffer as much winter injury as would be expected (Table 9). Disease and insect resistance.—-Larch sawfly, Pristiphora erichsonii (Htg.), invading from surrounding tamarack stands, attacked the plantation at Dunbar in 1964. The sawfly attacked tamarack much more than the Japanese or European larch (Table 11). All Japanese larch sources were similar in resistance to larch sawfly damage. A Zalga sp. canker also attacked the Dunbar plan- tation in 1965 with new cankers appearing each year since 1965. The canker attack is apparently secondary, attack- ing the trees growing in a very poorly drained portion of the plantation. Japanese and European larch are much more suscep- tible to damage by Kalga sp. canker than tamarack (Table 11). 42 Table ll.--The resistance of Japanese larch to larch sawfly attack and susceptibility to Valsa sp. canker attack in comparison to tamarack. Trees with 3338:: Species Yélégygg}lgggker sang$bzglgfiies Percent Per 40 trees 2788 JAP 70 0 2800 JAP 65 7 2793 JAP 55 l 2799 JAP 55 4 8 EUR 52 l 2806 JAP 50 1 2808 JAP 50 3 2796 JAP 35 1 ll TAM 15 107 F value 23.8** 3.58** JAPanese larch, EURopean larch, and TAMarack. **Significant at .01 level. REFERENCES CITED Aird, P. L., and E. L. Stone. 1955. Soil characteris- tics and the growth of European and Japanese larch in New York. J. For. 53:425-429. Anon. Forestry Agency Ministry of Agriculture and For- estry. 1957. Forestry of Japan. 2nd edition. Toyko, Japan. 216 pp. Chandler, Clyde. 1967. A progress report on the larch improvement program at Boyce Thompson Institute. Contrib. Boyce Thompson Inst. 23(9):319-26. Cook, Dave B., and D. L. Schierbaum. 1948. Planting an adverse site in New York. J. For. 46(5):377. Eliason, E. J. 1965. New York State forest tree improve- ment program. Northeast For. Tree Improve. Conf., 30. 66 pp. Genys, John. 1960 Geographic variation in European larch. New Hampshire Forestry and Recreation Comm. Bul. 13. 100 pp. 43 44 Langner, W., and K. Stern. 1965. Untersuchunqen der geo- graphischen variation und kovariation einiger merkmale in einem herkunftsversuch mit Japanisher larche (Larix leptolepis Gord.). Zeitschrift fur pflanzenzfichtung. 54(2):154-68. MacDonald, James, R. F. Wood, M. V. Edwards, and J. R. i} Aldhous (eds.) . 1957. Exotic forest trees in 1 Great Britain. Forestry Comm. Bul. 30. 216 pp. 41‘ Mathews, J. D., A. F. Mitchell, and R. Howell. 1960. The E analysis of a diallel cross in larch. Proc. 5th World Forestry Congress. Seattle, Wash. pp. 818 Schober, R. 1958. Ergebnisse von larchen art and provenenzversuchen. Silvae Genetica. 7:137-154. Schober, R. 1961. The international larch provenance trial 1958-59. First observations on provenances. Forsch. Landes Nordrgein-Westfalen. 4:96-104. Stairs, Gerald. 1965. Geographic variation in Japanese larch. Northeast. For. Tree Improve. Conf. 13: 12-14 Wright, Jonathan W. 1962. Genetics of Forest Tree Im- provement. FAO, Rome. 399 pp. VITA Dan Henry Farnsworth Guidance Committee: V. J. Rudolf, J. W. Hanover, D. P. White, J. W. Wright (Major Professor). Thesis: Geographic Variation in Japanese larch--Results of a 9-Year Study in North Central United States. Education: Graduated Hayward High, 1961 (Diploma) Attended Greenville College, 1961-1963 Attended Michigan Technological University, BSF, 1966 Attended Michigan State University, MSF, 1968. Biography: Born February 11, 1943, Ladysmith, Wisconsin. Resided in Hayward, Wisconsin until 24 years old. Married, June 25, 1966. Son born, April 8, 1968. Employment: Graduate Research Assistant Michigan State University, 1966-1968. Michigan Conservation Department, 1968, Assistant Area Forester, Crystal Falls, Michigan. AAAAAAA “N1111111133111Es ""1\1\\\\\\wfi\im 3193