GEOGRAPHIC AND STAND VARIATION IN JACK PINE (PINUS BANKSIANA LAMB.) Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY DAVID S. CANAVERA 1969 “- IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 448 4757 This is to certify that the thesis entitled @805r‘apI1fc lam? Shun! Varaei'fon In 3“" “WNW 142mb.) presented by Dan/.215 3. Canave V‘CL has been accepted towards fulfillment of the requirements for M degree in My Major professor MW O~169 LIBRARY I Michigan 3 rate University Lean :. veg-7 £8121 1' “’93 I ; KW (I E ifa‘fiia’ - I n— 41"?"3 rm. R _L LU» IAJM 2i {be ... l. I... .IlilillltIIel .II': ABSTRACT GEOGRAPHIC AND STAND VARIATION IN JACK PINE (PINUS BANKSIANA LAMB.) by David S. Canavera The objectives of this study were: (1) to determine the geographic variation pattern in jack pine, (2) to determine the amount of variation that exists between - and within - stands of jack pine from Lower Michgan and to establish seed orchards for the production of genetically improved seed for use in Lower Michigan, (3) to determine the effective- ness of mass selection for height growth in natural stands, (4) to determine if there is natural resistance between 2-0 half-sib families from Lower Michigan to ovipositing by the jack pine budworm, and (5) to determine what effect different levels of root pruning has on height growth. Variation in growth characters was investigated in 95 range-wide sources. The characters were measured on seedlings grown for three years at the Tree Research Center, Michigan State university, East Lansing, Michigan and on 4-year old trees growing in two test planta- tions located in southwestern Michigan. Analysis of variance, the summation-of-differences technique, and correlations between progeny performance and latitude, length of growing season and growing degree days at the place of seed collection were the principal statistical tools used. David S. Canavera The overall variation pattern in jack pine is continuous. No sharp breaks are evident in the papulation. The southern sourcesgenerally grew taller, remained greener in winter, and produced more flowers and lemmas shoots than the northern sources. Height and l- and 2-year autumn coloration were highly correlated with the previously mentioned climatic variables at the place of seed collection. Individual-tree selections were made in the summers of 1965 and 1966 in natural stands of jack pine in the Lower Peninsula of Michigan. A total of 382 trees from 61 stands were collected. The trees were selected for height growth, stem straightness and presence of open cones. Average and below-average trees were also selected from each stand. The seedlings were grown in a replicated randomized complete block design at the Tree Research Center. Three-year nursery data for height growth and other characters shows that the variance within-stands is much larger than the variance among-stands. Progeny heights were only slightly correlated with parental heights and it was concluded that mass selec- tion for height growth in natural stands was ineffective. I Six half-sib progeny plantations were established in Lower Michigan with 2-0 stock from the Lower Michigan individual-tree selections in the springs of 1968 and 1969. They consist of replicated randomized complete blocks. The plantations will be converted to seed orchards by the removal of the poorest families and poorest individuals within families. The plantations will be measured at 5-year intervals and the poorest families will be removed after each measurement. The genetic quality of the seed orchards will increase with each thinning. Twenty-five sources of the individual-tree selections were tested for resistance to ovipositing by the jack pine budworm. Single tree David S. Canavera plots of 2-0 trees were planted on a one-foot square spacing at the Tree Research Center. Each source was replicated 12 times in three lOO-tree blocks. Larvae and pupae of the insect were collected in July and reared on jack pine foliage until pupation was complete. After the adult moths had meted, they were released on the trees. A 11 x 11 x 3 foot cage covered each block of trees. The female moths oviposited within a week and the egg masses were counted. No significant difference for the number of egg masses per tree was found between sources. The roots of three hundred and sixty 2-0 seedlings had none, one- quarter or one-half of their lengths removed before spring growth had started. They were transplanted on a one-foot square spacing at the Tree Research Center. First year growth increments did not differ significantly between root pruning levels. GEOGRAPHIC AND STAND VARIATION IN JACK PINE (PINUS BANKSIANA LAMB.) VI David SiCanavera A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1969 ACKNOWLEDGEMENTS The author is indebted to the members of the Guidance Committee -- Drs. M. w. Adams, J. E. Cantlon, J. W. Hanover, L. F. Wilson, 8. N. Stephenson, and J. v. Wright (Chairman) -- for their encouragement, inspiration and assistance throughout the course of this study. Thanks are also due to w. L. Nance for his assistance in analyzing the data. Finally I wish to thank my wife, Nancy, for her sacrifices and inspiration during the preparation of the thesis. The financial support for this study was provided by the Michigan Department of Natural Resources, Forestry Division. Use of the Michigan State university computing facilities was made possible through support, in part, from the National Science Foundation. ii TABLE OF CONTENTS Page ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . ii LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . viii CHAPTER I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 1 Description of Species . . . . . . . . . . . . Previous Genetic Studies in Jack Pine . . . . Provenance Research Individual-Tree Inheritance 9N II. DESIGN OF THE PROVENANCE EXPERIMENT . . . . . . . . . . 10 Materials and Methods . . . . . . . . . . . . . . . 10 Seed Procurement Handling in Nursery Outplanting Procedure Plantation Measurements Analysis III. ran csocurnrc wanna: par-rem . . .‘ . . . . . . . . 15 Individual Traits . . . . . . . . . . . . . . . . . 15 Height Autumn Color Bud Set Number of Crooks Lammas Growth Male Flower Production Female Flower Production .AOV's Compared to Correlations . . . . . . . . . . . 29 Summation-of-Differences Analysis . . . . . . . . . 31 iii CHAPTER IV. EVOLUTION.AND MIGRATORY HISTORY . . . V. DESIGN OF INDIVIDUAL-TREE INHERITANCE STUDY . . Materials and Methods . . . . . . Seed Procurement Handling in Nursery Measurements Analysis VI. VARIATION AMONG MICHIGAN HALF-SIB PROGENIES . Individual Traits . . . . . . . . Height Color Lemmas shoots and female flowers Between- and Within-Stand Variance Components Effectiveness of Selection for Height . . VII. JACK PINE BUDWCRM RESISTANCE STUDY . . . . . Life Cycle of the Jack Pine Budworm Materials and Methods . . . . . . . . . . . Results and Discussion . . . . . . . . . VIII. RESPONSE TO DIFFERENT LEVELS OF ROOT PRUNING Materials and Methods . . . . . . . . Results and Discussion . . . . . . . . . IX. POSSIBILITIES Fat FUTURE IMPROVEMENT . . . LITERATURE CITED . . . . . . . . . . . . . . . . . . . VITA O O 0 O O O O O O 0 O O O 0 O O O O O 0 APPENDIX . . . . . . . . . . . . . . . . . . . . . . . iv Page 37 39 39 46 46 49 S3 57 57 58 61 62 62 63 65 68 71 72 LIST OF TABLES TABLE Page 1. Origin and 3-year growth data for provenance test sown in 1964, summarized by stand progeny. . . . l6 2. Correlations (r) between nursery and plantation heights, based on 92 source means . . . . . . . . . . 21 3. Correlations (r) between height and autumn color and climatic data at place of seed collection, based on 95 source means. . . . . . . . . . . . . . . 22 4. Analyses of variance for 4-year growth data at Kellogg and Allegan . . . . . . . . . . . . . . . . . 25 5. Analyses of variance for flower production at hurlery, All-98‘“ ‘nd “11088. e e s e s s e e s s s' s 26 6. Correlations (r) between flower production and climatic data at place of seed collection, based on source means . . . . . . . . . . . . . . . . 28 7. Comparison of the amount of variance explained by correlation (r ) and analysis of variance . . . . . . 30 8. Summation-of-differences table for sources east of 80° longitude. . . . . . . . . . . . . . . . . . . 32 9. Summation-of-differences table for sources west Of 800 longitude O O O O O 0 O O O O O O 0 O O 0 O O 0 34 10. Analyses of variance for Michigan 1/2-sib progeny test sown in 1966 . . . . . . . . . . . . . . . . . . 47 ll. Analyses of variance for Michigan l/2-sib progeny test .wn in 1967 O C O O O O O O O O O O O O O O O O 48 TABLE 12. 13. 14. 15. l6. l7. 18. 19. 20. 21. 22. 23. 24. 25. Correlations (r) between growth characters and climate at place of seed collection fer Michigan 1/2 aib-progenies, summarized by stand . . . . . Analyses of variance for Michigan 1/2-sib progeny te.t' O O O O C I O O O O O O O C O O O O Within-stand, between-stand, and error variances expressed as a percent of the total variance . . Parental and 3-year progeny growth data for individual-tree selections within a randomly chosen stand . . . . . . . . .,. . . . . Four-year growth data at Allegan and Kellogg, pl‘nted 1966 O I O O O O O O O O O O O O I O O 0 Flower production at Tree Research Center, Allegan and Kellogg, summarized by source . . . . . . . . Analyses of variance for provenance test at the Tree Research Center, includes seedlot 97 which was later discarded . . . . . . . . . . . . Analyses of variance for provenance test planted at Kellogg, includes seedlot 97 which was later d1.c.rded O O O O O O O I O O O O O O O O 0 Analyses of variance for provenance test planted at Allegan, includes seedlot 97 which was later d1.carded O O O O 0 O O O O O O O O O O 0 Analyses of variance for provenance test arranged by regions. All measurements were made at the Tree Research Center except for the percent of trees with lemmas growth, which*was measured at Allegan . . . . . . . . . . . . . . . . . . . Correlations between characters in provenance study Tests of interaction for height between Allegan, Kellogg and the Tree Research Center in the provenance study . . . . . . . . . . . . . . . . Test of interaction for percent of trees with female flowers between Allegan, Kellogg and the Tree Research Center in the provenance test . . . Test of interaction for percent of trees with lammas growth between Allegan and Kellogg in the provenance test . . . . . . . . . . . . . vi Page 50 51 52 73 76 79 80 81 82 83 885 86 87 TABLE Page 26. Origin and growth data for 1966 sown Michigan l/2-sib progeny test, summarized by stand Of origin 0 O O O O O O O O O O O O O O O O O O O O O 88 27. Origin and growth data for 1967 sown Michigan 1/-sib progeny test, summarized by stand of origin . . . . . . . . . . . . . . . . . . . . . . 91 28. Analyses of variance for 1967 sown Michigan 1/2’81b Progeny te‘t s s s s s e e s s s s s e o s s 93 29. Analyses of variance for 1966 sown Michigan l/2-sib progeny test . . . . . . . . . . . . . . . . 94 30. Analyses of variance for between- and within-stand variances for Michigan l/2-sib progeny tests . . . . 95 31. Correlations for Michigan l/2-sib progeny cellection. O O I O O O O O O O V O O O O I O O O O O O 97 32. Data on budworm resistance study, summarized by .ource O O O O O O O O O O O O I O O O O O O O O O 99 33. Analysis for variance for number of egg masses per source in budworm resistance study . . . . . . . 100 34. Correlations in budworm resistance study . . . . . . . . 100 vii LIST OF FIGURES FIGURE Page 1. Natural distribution of jack pine in North America (shaded area from Critchfield and Little, 1966) and sources included in this experiment (numbers). . . . . . . . . . . . . . . . 5 2. Distribution of the parental stands (black dots) from which the half-sib progenies were obtained . . . . . . . . . . . . . . . . . . . 42 3. The cages pictured here are 2 feet high and 11 feet square. Each cage covers a block of 100 tree. I O O O O O O O O Q I O O O O O O O O O O 59 viii CHAPTER I INTRODUCTION Jack pine (Ping; banksiana Lamb.) is now the most prominent pine in the forests of Michigan's Upper Peninsula and the upper half of the Lower Peninsula. Originally it inhabitated a very limited area of the state. Pure stands of jack pine were found only on the poorest sandy soils. On better soils, it grew in mixture with red pine (g. resinosa Ait.) and.white pine (g, strobus L.). Jack pine has been planted exten- sively in the state and has also spread naturally into burned and de- ' forested areas. The jack pine type now ranks third in acreage among the forest types, occupying about one million acres of forest land. Jack pine was regarded as an undesirable species in Michigan 100 years ago. Since then, the forests have changed drastically and the demand for pulpwood has increased, so that its present day value is great. Jack pine has been a favorite fer reforestation due to its rapid juvenile growth, desirable pulping characteristics, ability to grow on poor sites, and low woods waste. Its least desirable character- istics are extremely poor stem form and coarse branches. 'The serotinous cones were advantageous when fires ravaged the state. However, the present day fire danger has been greatly reduced, and the early cone Opening is now needed to obtain natural regeneration. Many existing plantations have to be considered failures because of slow growth, poor stem form, coarse branches and lack of natural regeneration. Future 1 plantings must produce more value per acre than the present jack pine. Genetic improvement through selection and selective breeding has been achieved for a number of tree species. The goal of this study was to examine the possibility of obtaining improvement in jack pine. It was conducted as a cooperative effort between the Forestry Division of the Michigan Department of Natural Resources and Michigan State University. Description of Specie; The following information is taken from Silvics of Forest Trees of The united State; (P. O. Rudolph, in U. S. Forest Service 1966), Michigan Trees worth Knowing (Smith 1961), Silvicultural Research in Jack Pine (Cayford 1967), and Natural Variation in Jack Pine (giggs_banksiana Lambert) (Schoenike 1962). Jack pine matures at about age 60, attains heights of 60 to 70 feet and average stem diameters of 12 to 18 inches. The lower branches die rapidly but remain on the tree for many years and give it a ragged appearance. The bark is light brown and only slightly scaly on young trees with new growth, dark gray or black with loose scales on older branches and irregular, scaly ridges on old trunks. Needles are in 2's, dark yellow-green, frequently exceeding 3 inches in young seed- lings but only 3/4 to 1 1/2 inches long on mature trees. Staminate flowers are yellow; ovulate flowers are purple. Cones are 1 1/2 to 2 inches long, sessile, oblong-conic, frequently strongly incurved, pointing forward, both serotinous and non-serotinous. Seeds are tri- angular, black and roughened, 1/2 inch long and 1/3 inch wing. Jack pine grows under a wide range of habitat conditions. In general, the areas are characterized by warm-to-cool summers, extremely cold winters, moderate rainfall, light sandy soils, and rolling to level topography. The average July temperatures range from 55° to 72° farenheit. Within most of its distribution, precipitation is 15 to 35 inches, the extremes are 10 to 55 inches. Periods of 30 or more days without precipitation occur commonly from Michigan west. It is con- fined largely to soils of the podzol region: melanized sands, podzolic sands, sandy podzols, and the gley-podzolic sands. Maximum.development is reached on.well-drained loamy sands. Its altitudinal range is be- tween sea level and a little over 600 meters. Jack pine is very prolific in seed production and can produce female flowers at age 2. Open-grown trees begin seed production at 5 to 10 years. Jack pine is a pioneer species. It frequently inhabits burned areas where there is exposed mineral soil, and little plant competition. Extensive even-aged stands are common. In the absence of fire or other catastrophies, jack pine is replaced by more tolerant species. On the poorest, driest sites it forms and edaphic climax. A typical succession on better sandy soils (loamy sands or sandy loams) is from.jack pine to red pine, to eastern white pine, to a hardwood type composed of north- ern red oak (Quercus 322;; L.), basswood (gilig_americana L.) and sugar maple (Age; saccharum Marsh.). On loamy soils it is replaced by black spruce (3.1.9.22 mariana (Mill.) B.S.P.), white spruce (me; Mg; (Moench) Voss) and balsam fir (A213; balsamea (L.) Mill.). Other species commonly growing in mixture with jack pine include trembling aspen (Populus tremmloides‘Michx.), white birch (Betula papyrifera Marsh.), northern pin oak (Quercus ellipsoidalis E. J. Hill) and bur oak (Quercus macrocarpa Michx.). The geographical range of jack pine is shown in figure 1. It ' extends over 1,000 miles in its greatest north-south direction and about 2,600 miles in a southeast-northwest direction. From its northernmost point along the Mackenzie River in western Canada, it stretches south and east to the Lake States, northern New York and New England, and the Maritime Province of Nova Scotia. The ranges of jack pine and lodge- pole pine (g. contorta Dougl.) overlap in Alberta and MacKenzie and natural hybridization occurs. Previous Genetig_§tudigs in Jack Pine Provenance research.--The first geographic origin test of jack pine was made by the University of Minnesota in 1940 (Schantz-Hansen and Jensen 1952; Schoenike, T. D. Rudolph and Jensen 1962). It con- sisted of 32 seedlots collected throughout the range of jack pine, from.Maine to Alberta. The trees were grown for two years in the nursery and outplanted in 1943 at the Cloquet Forest Research Center in northern Minnesota. The heights varied from 2 to 8 feet at age 8, and from 6 to 24 feet at age 15. The tallest sources were from Michigan, Minnesota and Ontario. The source from Bar Harbor, Maine was prostrate. The majority of trees from all sources had poor form. Morphological measurements were made on 23 sources (Schoenike, T. D. Rudolph, and Schantz-Hansen 1959). Trees with closed cones appeared most frequently in sources from.northeastern Minnesota and adjacent parts of Ontario and Manitoba, the Upper Peninsu1a of Michigan, and most of Canada. Contrasted to this, trees with open cones were found with highest frequency in sources from Michigan's Lower Peninsula, southern and western Minnesota, New Brunswick and.Maine. Cones on Figure 1.--Natural distribution of jack pine in North America (shaded area from Critchfield and Little, 1966) and sources included in this experiment (nUmbers). The two measUred plantations are shown as A - Allegan and K - Kellogg. . _._ ea _ . __ A .. a 7: I l L llIl .. . .V nnnnnnnn II III .. . .. III I x . _ I/ .V r. _ 1.1/1. . — [IF . e ,. _ . — — — \ . . .... . (U. sources from the western portion of the range were more curved and had more acute angles (the angle produced between a line connecting the tip and point of attachment of the cone, and the branch upon which the cone was borne) than eastern and Great Lakes sources, which had straight cones and larger cone angles. Branches on trees from eastern sources had larger angles (the upper angle produced by the basal one-foot length of branch and the major axis of the tree) than the Great Lakes and western sources. During the winter of 1947-48, severe winter injUry occurred. There was no consistent geographical pattern in the degree of injury. The most severely damaged sources were from Baldwin and Manistee in the Lower Peninsula of Michigan. A few trees were killed, and in some plots as high as 93 percent of the trees showed injury. Eleven other scattered sources showed moderate to light injury. No injury was found in 12 sources. Natural stands of jack pine in the Cloquet area suffered little damage. The second large scale geographic origin test was started in 1951 by P. O. Rudolf of the Lake States Forest Experiment Station. It con- sists of 29 seedlots from the Lake States. Each seedlot is made up of seed from several dominant and codominant trees in a single native stand. Seventeen permanent outplantings were established in the Lake States with 2-0 stock. Each test plantation also had one "local" seed source supplied by a commercial nursery in the area of the plantation. Several reports on this experiment have been published (Stoeckler and P. O. Rudolf 1956; Jensen, Schantz-Hansen and P. O. Rudolf 1960; Batzer 1961; Arend et al. 1961; T. D. Rudolph 1964; King 1964; Alm and Jensen 1969). Height measurements were made in three Lower Michigan plantations at age 5; in 11 of 17 plantations scattered throughout the Lake States at age 10; and in one Minnesota plantation at age 13. The results were consistent. The group of sources from Michigan's Lower Peninsula performed best in the Upper and Lower Peninsula of Michigan and Wisconsin. Sources from north central Minnesota did best in the northern Minnesota plantation. Winter color differences were evident in 1-0 nursery stock, but disappeared with age. Trees from northern sources turned purple, where- as those from southern sources remained green. Height growth in northern conifers usually occurs in a single flush without any intermittent rest periods. At the end of the growth period, buds are set and primordia are laid down for next seasons growth. A current season's growth is largely determined by what was formed the previous year. Some trees or individual shoots on a tree may~have more than one flush in a single growing season, After initial growth in the spring, intermittent periods of rest and growth follow. The later flushes are developed from primordia formed the same year. These late growth flushes occur frequently in jack pine. They are called lemmas shoots when formed on the terminal shoots and prolepsis shoots if formed on the lateral shoots located at the base of the terminal shoot. In the 1951 Lake States study, lemmas growth and prolepsis varied significantly between seed sources in the four Minnesota and two Wisconsin plantations. Southern sources had a higher frequency of these late shoots than northern sources. Differences among sources in white pine weevil incidence occurred on two plantations in northern Minnesota and three plantations in Lower Michigan. Seed source differences also occurred in needle cast infec- tion in a southern Wisconsin plantation and a western Upper Michigan plantation. The most recent provenance test was started by Mark Holst of the Petawawa Forest Experiment Station, Chalk River, Ontario. ~The genetic information received from earlier provenance studies were limited because they contained a small number of seed lots. In the Petawawa experiment, 95 stands were sampled throughout the range of jack pine. The collections were later subdivided so that nursery and field tests could be performed in the United States, Canada, Denmark, Finland, Scotland, Czechoslovakia, Holland, New Zealand and Germany. I Yeatman, (1967) reported on growth changer, greenhouse and nursery performance. The average response over all of the environments tested provided the best discrimination between provenances. The interaction of provenance and photoperiod‘was the second best discriminator. Northern sources were hardier than southern. The degree of hardiness was positively correlated with amount of foliar sugar content, needle coloration and time of bud formation. Following germination, 76 percent of the variation in seedling size was explained equally by seed weight and growing degree-days at the place of seed origin. Temperature was more important than photoperiod in initiating spring growth. Individual-tree inheritance.--An open-pollinated progeny test, made up of individual-tree collections from 10 closed-cone and 18 open-cone trees, was started in 1939 in northern Minnesota. In 1957, 52 percent of the offspring of the open-coned mother trees produced cones which opened soon after ripening, contrasted to only 13 percent from the closed-cone mother trees (T. D. Rudolph, Schoenike, Schantz-Hansen, 1959). CHAPTER II DESIGN OF THE PROVENANCE EXPERIMENT Earlier jack pine provenance studies do not provide adequate infor-* mation on the performance of seed sources from outside of the Lake States when planted in Michigan. In experiments at Michigan State University, origins of eastern white pine and white spruce from.outside of Michigan grow 20 to 30 percent faster than native sources when planted in southern Michigan (Wright gl_gl, 1963). The Petawawa range- wide test, which I studied,‘was planted in Lower Michigan in order to study the performance of jack pine seed sources from the entire range of the species when planted in Lower Michigan. Materials and Methods Seed procurement.--Seeds from 95 origins of the Petawawa range- wide collections were used. Each seedlot was collected from 3 to 100 trees in a native stand. Seed collection areas are shown in figUre l, and origin data is presented in table 1. Handling in nurserz.--On October 30 and 31, 1963 the seeds were sown in a 4-replicated, randomized complete block design at the Tree Research Center, Michigan State university. Each replicate contained one 4-foot row of each seedlot. The rows were one foot apart. Seeds were sown at eight spots in each row, with three to ten seeds planted per spot. Germination in the spring of 1964 averaged 85 percent for 10 11 the experiment as a whole. The spots were thinned to two seedlings per spot. A fifth replicate was broadcast sown to provide stock for outplant- ing. Enough seed was sown to give a density of 40 seedlings per square foot. The seedlings in these beds grew very fast. In June of 1965, they were mowed back to a height of one foot in order to produce a more favorable root/shoot ratio for outplanting. All of the seedbeds were treated with methyl bromide prior to seeding. After germination, the seedbeds were watered, kept weed free, fertilized, and mulched*with one-half inch of sawdust in the autumn. Parallel strips running lengthwise in the seedbeds showed areas of poor and good growth-on July 14, 1964. Soil analysis showed a nutrient deficiency in the poor growth areas and adequate fertility in the good growth areas. Corrective measures were taken on July 17 by a foliar application of fertilizer to the poor growth areas. Broadcast fertil- izer application on the entire seedbeds were made on July 20 and August 3. The seedlings responded rapidly and no strips of differential growth were visible by August 13. No edge effects were observed either within-rows or over the seed- beds as a whole. There was little within-row competition. However, the position of a seedlot in a replicate did affect growth rate because‘ of differential watering. .Some areas in the seedbeds produced taller seedlings than other areas, irrespective of which seed sources were present. Because of this, I adjusted source heights according to their position in the replicates. I accomplished this by first dividing each replicate into eight parts and recording in which of the eight a source was located. Then I calculated the difference between the plot 12 means and the overall average for the same sources. Then the differ- ences for all sources in each of the eight parts were added and the average difference for each part was calculated. Then an amount equal to the average difference, but opposite in sign, was added to the mean for each plot in each of the eight parts. For example, if the average difference for an octile was -2, then +2 was added to the mean_of each plot. Nursery measurements.--A1together 22 complete sets of measurements were made on different characters or the same character measUred at various times. First year measurements were made by Fan Kung, and I made all subsequent measurements. Metric traits were measured to an accuracy of approximately 1/20 of the range between extremes. Nan- metric characters were evaluated in terms of the smallest recognizable difference between units. Color was defined in terms of live-tree standards, and was scored by comparing the average color of a plot with a certain color standard set in the observer's mind. Crookedness was measured along the main stem on the tallest tree in each row. A crook was defined as a bend in the main stem of the tree. The deviation of this bend was measured as the perpendicular distance between an imaginary straight line drawn along the main stem of the tree and the greatest point of departure of the bend. Outplanting procedure.--Experimental plantations were established in Michigan in the spring of 1966 at the Allegan State Ghme Area and W. K. Kellogg Forest (figure 1). Both plantations contain 92 sources and follow a randomized complete block design, with 4 trees per plot. There are ten replicates at Allegan, two at Kellogg. Spacing between trees is 8 feet. Survival was 82 percent at Allegan and 90 percent at Kellogg. 13 The Allegan plantation is situated on a uniform level site. Com- peting vegetation is sparse; no chemical weed control was used. The soil type is Plainfield sand. The plantation at Kellogg Forest is located on a southwest facing slape that has a 5 to 20 percent grade. The area had previously been planted with Scotch pine (P. sylvestris L.), which was removed in 1963 and 1964. The site was sprayed with mmino-triazole in the fall of 1965 to control the heavy sod and weeds that invaded after clearing. The soil type is Oshtemo sandy loam. Seventy-five percent of the top soil had eroded prior to 1940. Plantation Measurements.--Height, cone production and number of trees with lemmas shoots were measUred in July 1968 at Kellogg and August 1968 at Allegan. Analysis.--An analysis of variance was calculated for each charac- ter studied. Plot means were used as items. A typical analysis of variance table follows: Source of Variation Degrees of Freedom Expected Mean Squares 2 2 Source ‘S-1 0; + R0; Replication , R-l 0:2 + SUI}2 Replication x source (S-l)(R-1) 0:2 Total (S x R)-1 R and S are the number of replicates and sources respectively. Differences between sources were tested with replication x source mean square. 14 Least Significant Differences (L.S.D.) were computed for each character at the 1 percent and 5 percent levels. This technique is used to determine which sources differ from each other. I used Tukey's method for multiple comparisons. I divided the sources into natural regions using geographic features,climatic data and similarity of growth traits of the progeny and determined the amount of variation that exists between- and.within- regions. A typical analysis of variance table follows: Source of Variation Degrees of Freedom Expected Mean Squares Sources Between regions 4 032 + 40;,2 + 740i,2 Within regions 90 052 4092 Total 94 Simple correlations, using progeny means as items, were calculated to determine the amount of variation in height, autumn color and cone production that could be explained by length of growing season, growing- degree days and latitude at the place of seed collection. Correlations were also calculated between nursery and field heights and between other characters of possible biological significance. The correlations and analyses of variance were done on Michigan State university's CDC 3600 digital computer. CHAPTER III THE GEOGRAPHIC VARIATION PATTERN Results from.the provenance test show the major geographic trend in jack pine is north-south. The correlations between latitude, length of growing season and growing degree-days (sum of positive values of (dain maximum in °r + daily minimum in 6r) - 42) at the place of seed and the progenies' chiracteristics were very high. The origins from the western part of the range fit the climatic data better than those from the east. Over its entire range, jack pine has adapted to the climate where it is growing. The data were analyzed in a variety of ways in order to obtain the maximum amount of information about the geographic variation pattern. Traits were analyzed individually over‘the entire range, individually by regions, and collectively over the entire range. Individual Traits figigh§.--Origin and 3-year nursery growth data are given in table 1. Average 3-year height in the nursery was 101 cm. In the plantations, the average 4-year height was 63 cm. at Allegan and 86 cm. at Kellogg Forest. The source x plantation interaction was significant between the nursery and both plantations. It was not significant between 15 16 Table l.-- Origin and 3-year growth data for provenance test sown in 1964, summarized by stand progeny. ,‘ Crookedness 0nt.Num- .93 , ’3 Autumn Trees Crooks Size ber and 3 g 3 V Height Color With Buds Per of Place of North West 3 g :3 :1 Age Age Sept. 22 Tree Crooks Origin Lat. Long. (‘5 3’, g 3 3 l 1964 Age 3 Age 3 (1) (2) (3) (4) 15) (6) (7) (8) L9) 3200 hun- per 16 + 0 0 days dreds _c_m_._ grade trees 339‘ an, Sources west of 800 west longitude 99 NWT 63.2 123.5 136 17 46 1 3 0 0 98 NWT 61.8 121.3 136 18 48 2 3 1 6 96 NWT 60.1 112.0 136 14 62 2 1 1 6 92 Sea 57.1 102.0 150 15 66 1 3 0 0 91 Man 54.8 101.9 153 18 66 2 2 O 0 89 .Sas 54.2 105.0 156 20 64 2 2 1 6 95 Alb 56.6 111.9 156 20 70 3 4 l 13 85 Ont 53.0 93.3 147 18 79 3 4 1 13 94 Alb 55.2 111.9 164 21 79 3 l 1 13 90 Man 52.1 100.4 165 24 83 3 5 1 13 87 See 53.1 106.1 168 23 85 4 4 1 13 93 Alb 54.1 115.7 160 20 76 4 4 l 6 64 Ont 50.1 82.2 152 20 76 4 4 1 32 76 Ont 48.8 87.4 156 21 90 4 4 1 32 77 Ont 49.6 86.0 154 20 90 4 6 1 19 88 See 53.8 107.7 160 22 75 4 3 l 19 86 Ont 51.0 94.1 166 24 99 5 5 1 19 59 Mic 46.4 84.3 171 27 103 S 5 1 38 83 Man 49.5 95.8 172 28 114 5 6 1 25 84 Ont 49.8 93.3 168 25 105 7 5 1 25 82 Ont 49.8 94.5 171 27, 121 9 4 1 32 81 Ont 48.8 93.5 177 26 114 7 5 2 32“ 63 Ont 48.8 80.8 ,157 22 117 8 6 2 32 62 Ont 47.7 80.7 164 24 104 7 4 1 25 61 Ont 46.8 81.6 169 26 105 8 4 1 32 60 Ont 46.8 84.0 171 24 107 7 5 1 45 58 Ont 45.8 82.9 184 28 108 7 2 2 45 57 Ont 45.0 81.5 195 29 110 8 5 1 32 80 Min 47.3 94.6 189 30 118 8 5 1 45 79 Min 46.7 92.6 182 29 121 9 3 1 38 17 Table 1.--Continued ,‘ Crookedness Ont.Num— .03 , A Autumn Trees Crooks Size her and .5". g 3 SHeight Color With Buds Per of Place of North West 3 3 .23 3, Age Age Sept. 22 Tree Crooks Origin Lat. Long. :5 c7; 3 3 3 1 1964 Age 3 Age 3 L1) 12) Q) (4) 15) (6) L7) L81 (9) 3200 hun- per 16 + ° 0 days dreds _c_m,_ grade trees gg_._ ml_n._ 78 Min 46.3 94.2 192 31 134 8 5 2 32 70 Wis 45.6 89.9 188 28 127 8 4 3 76 55 Out 44.0 81 8 193 33 108 9 5 1 25 69 Wis 44.8 89.7 202 32 127 7 5 2 38 68 Wis 44.3 89.0 205 33 124 8 6 3 51 75 Mic 46.0 86.5 186 28 131 7 7 2 64 74 Mic 44.5 84.8 191 32 124 8 3 2 51 73 Mic 44.5 84.8 191 32 135 8 6 3 57 72 Mic 44.5 85.4 195 33 136 7 7 2 57 56 Out 44.5 80.0 199 31 137 8 7 2 70 65 Wis 43.6 90.2 217 37 128 8 7 2 32 66 Wis 43.8 89.8 207 36 132 7 5 2 45 67 Win 44.3 89.7 193 34 132 8 7 2 51 71 Mic 44.1 86.1 210 35 124 9 3 2 45 54 Ont 43.2 81.9 207 36 124 8 4 2 57 Sources east of 800 west longitude 53 Que- 54.0 76.5 111 10 50 3 4 0 0 55 NS 46.9 60.3 191 24 76 3 0 1 13 44 NS 46.8 60.4 191 24 80 6 1 0 0 38 Que 50.4 73.9 141 14 79 4 7 1 6 21 Que 49.7 67.3 159 16 79 7 4 0 0 20 Que 49.7 68.4 159 16 90 7 5 1 19 19 Que 49.3 69.9 152 17 73 6 2 1 13 3 NS 45.3 61.0 182 26 78 7 1 1 13 1 M81 44.3 68.1 208 27 85 8 2 1 19 15 Que 48.1 67.5 160 24 92 5 5 1 25 37 Que 49.6 72.2 154 18 90 5 5 1 19 32 Que 48.9 71.8 163 22 92 6 2 l 19 13 Mai 45.9 69.6 176 27 84 6 2 1 19 36 Que 49.4 74.0 147 18 95 6 3 1 6 35 Que 48.0 74.3 156 21 96 6 6 1 32 18 Table l.--Continued ,. .9522!2§EE££_. Ont .Num- «:3 . A Autumn Trees Crooks Size her and .5 8 35 Height Color With Buds Per of Place of North West 3 2 $3 3, Age Age Sept. 22 Tree Crooks Origin Lat. Long. 5 c)”: 3 3 3 1 1964 Age 3 Age 3 ' Ll 2 (3) 14) (5) (6) (7) L8) (9L 3200 hun- per 16 + ° ° days dreds gs; grade trees g2; 92‘ 34 Que 47.7 74.0 159 23 108 6 2 1 38 50 Que 47.9 77.4 166 23 99 6 4 1 25 29 Que 47.6 70.2 169 25 97 5 4 1 13 14 NB 47.4 66.5 166 26 100 6 6 1 32 6 NS 45.7 63.8 182 28 94 8 4 1 45 2 NS 43.8 65.4 205 27 93 8 2 1 25 16 Que 47.6 69.6 167 26 106 6 5 2 32 18 Que 47.7 69.7 170 25 100 6 5 2 45 30 Que 48.3 70.9 167 24 97 8 4 1 32 31 Que 48.7 72.0 163 23 105 7 4 1 13 51 Que 48.0 75.4 165 21 102 6 5 1 19 28 Que 46.9 71.4 176 29 105 7 3 1 19 22 Que 45.0 73.8 190 36 90 9 0 1 25 24 NH 43.9 71.3 180 27 96 7 2 1 32 33 Aue 47.3 73.9 163 25 107 7 3 1 32 49 Que 47.8 76.7 165 23 102 7 5 1 13 48 Que 46.8 76.1 175 25 111 8 4 1 32 9 NB 46.0 66.1 178 27 100 8 3 1 13 8 NB 46.0 65.0 182 29 125 8 7 2 51 7 PEI 46.6 63.9 183 28 107 8 2 1 32 11 NB 46.5 65.9 179 27 107 8 4 2 32 10 NB 46.7 65.6 180 27 105 7 4 1 19 12 NB 47.5 65.4 172 25 100 8 5 1 13 47 Que 46.4 76.2 184 27 113 8 4 2 45 25 NH 43.9 71.6 180 27 106 8 1 1 38 26 Mai 45.5 70.2 176 27 119 7 5 1 32 23 NY 44.3 73.8 187 36 119 8 3 2 51 27 Que 46.4 72.6 184 28 116 8 4 2 45 40 Ont. 44.6 77.0 198 30 109 7 4 1 32 39 Ont. 44.7 77.9 187 31 107 9 2 1 45 19 Table l.--Continued ,‘ Crookedness Ont.Num- e03 . 3 Autumn Trees Crooks Size her and 5 3 3 V Height Color um. Buds Per of Place of North West 8 3 :1 3. Age Age Sept. 22 Tree Crooks Origin Lat. Long. :3 c7: 3 3 3 1 1964 Age 3 “Age 3 (I) (2) :31 741 (5) (6) (7) 48L 19L 3200 hun per 16 + ° 0 days dreds gm; grade trees gg_’ my; 41 Ont 44.7 77.1 187 30 107 7 l 2 51 42 Ont 45.5 77.0 188 30 123 8 2 2 32 43 Ont 45.5 76.1 190 32 122 7 S 2 45 44 Que 45.9 76.7 192 29 127 9 5 2 45 46 Ont 45.8 77.4 183 29 133 7 4 2 38 Std. deviation of a source mean 5.05 0.80 1.23 0 37 10 9 L.S.D.(.05) of a source mean 30.48 4.8 7.4 2.2 66.0 F value 17.5*** 6.5*** 1.9*** 3 1*** 2.4*** Percent of variance due to source 81 58 18 34 25 Color grades: 0-3 purple, 4-6 intermediate, 7-10 green *** Significant at 0.1 percent level. (a) Number of days on which mean temperature exceeds 42°F. 0 o _ (b) Sum of positive values of‘iggill'm8X1mum F.-2daily minimum.in F;)r42 (c) Measured from a straight line along the stem for the tallest tree in each plot. 20 Allegan and Kellogg. Correlations between nursery and field heights are given in table 2. The correlation (r) between Allegan and Kellogg was .74. At all test locations, the fastest growing sources were fromlsouth- ern Quebec, and southern Ontario. Height growth did not differ signifi- cantly among these regions. The next fastest growing sources were from' eastern Quebec, New Brunswick, northern Maine, northern New York and southwestern Ontario. The fast growing sources from the eastern part of the range (26-Mai, ll-NB, 8-NS, lO-NB, 23-NY), rank much higher in the plantations than in the nursery. Sources 88-Sas and 93-A1t also rank higher in the plantations than in the nursery. These results explain at least part of the reason for the interaction between the nursery and plantations. Sources_from the Northwest Territory were the slowest growing at all test sites. Most of the genetic variation was explained by latitude, length of growing season and growing degree-days at the place of seed collection. Trees that grew fastest came from areas that were most favorable for growth. The highest correlations were with the number of degree-days and latitude. The lowest correlation was with length of growing season. Correlations were highest in the nursery and smallest at Allegan (table 3). From these results, it is apparent that growth is governed more by temperature than photoperiod. In the provenance test started by the university of Minnesota in 1940, the source from Bar Harbor, Maine was prostrate. The three sources from.Maine in this study are straight growing. Sources 59 and 75 from the Upper Peninsula of Michigan grew better than expected. Arend ££_31, (1961) reported that sources from this 21 Table 2.--Corre1ations (r) between nursery and plantation heights, based on 92 source means. Height in Nursery FoursXegrgggigh£_A£___ At Age Allegan Kellogg ............ r ------------ 1 .79*** .71*** 2 .82*** .77*** 3 .83*** .74*** *** Correlation significant at 0.1 percent level. 22 Table 3.--Correlations (r) between height and autumn color and climatic data at place of seed collection, based on 95 source means. Climatic Data Age When Length Grow- Degree- Place Character Measured Latitude ing Season(a) Days(b) ygggg, ---- ----------- r -------------- Nursery Height 3 -.77*** .73*** .80*** Nursery Autumn color 1 - .85*** .71*** .72*** A1 legan Height 4 - .s7*** .24*** .74*** Height 4 - .63*** 58*“ . 79*” Kellogg (a) Number of da s on which dail mean tem.erature exceeds 42°F. Y Y P (b) Sum of positive values of {daily maximum in °F. + daily minimum.in OF.) - 42. 2 'k m Correlation significant at 5 percent level. Correlation significant at 0.1 percent level. 23 area were much slower growing than those from the Lower Peninsula of Michigan; the two papulations could be separated into distinct groups. In my study, no significant difference was found between these two regions. In fact, source 75 was among the fastest growing sources at all three test sites. This source is from Gladstone, which is on Lake Michigan in the southwestern part of the Upper Peninsula and has a warmer climate than most of the Upper Peninsula. Autumn color.--The north-south trend was also evident in first and second year autumn coloration. Northern sources turned purple, whereas southern sources remained green. The differences were most pronounced the first year and disappeared with age. The trend was most clearly defined in the western part of the range and more variable in the Ieastern part. Sources 4-NS, 21-Que, 20-Que, 19-Que, 3-NS, and l-Mai were much greener than expected. The correlation (r) between latitude and first year autumn coloration was .85. There were no summer color differences. Bud set.--The number of buds set by September 22, 1966 was quite variable but significant among sources. The correlations between bud set with latitude and growing degree-days‘at the place of seed collection were both r - .22. No adaptive features are evident from these low correlations. In addition to this, no regional differences were found. Apparently, the time of bud formation in jack pine is not controlled by photoperiod or temperature. Number of crooks.—-The number of crooks for the tallest tree in each row was highly correlated with source heights (r - .72). Only the slow growing sources 99-NWT, 92-Sas, 91-Man, 53-Que, 4-NS, and 21-Que had straight central stems. None of the fast growing sources were straight stemmed. The tall sources also had the largest crooks (table 1). 24 Lemmas growth.--The number of trees per source with lemmas and prolepsis growth were counted at Allegan and Kellogg. Seventeen percent of the trees at both plantations had these late growth shoots. The source X plantation interaction was significant. This relationship is also apparent from the correlation (r - .46) between plantations. When planted at Allegan, the correlation (r) between trees per source with secondary growth and latitude was .57: and the same character correla- ted with both length of growing season and growing degree-days at the place of seed collection was .74. The correlation with height was (r - .66). The correlations at Kellogg for the same characters ranged between r - .36 for length of growing season, and 4 - .39 for height. At Allegan, sources 6-NS, 8-NS, 29-Que and 98-NWT produced more trees with secondary growth than expected. But, 72-Mich and 25-NH had fewer trees than expected with secondary growth. The following sources did not produce any trees with secondary growth at Kellogg: 54-Ont, 55-Ont, 59-Mic, 23-NY, 26-Mai, and 9-Mai. Sources 64-Ont and 84-Ont produced more trees than expected with secondary growth. The lammas growth data from Allegan discriminates seed sources better than the data from Kellogg. Thirty-nine percent of the variance at Allegan can be explained by sources, compared to only 27 percent at Kellogg. The data at both plantations shows a north-south trend. How- ever, the trend is more clearly defined at Allegan than Kellogg. Per- haps growing conditions make this so. The plantation at Kellogg is on a sandy 10mm, and Allegan is on a deep course sand. The plantation at Allegan grew only 73 percent as fast as the plantation at Kellogg. Male flower production.--Ma1e flowers were present in the nursery on 89 percent of the sources at age three (table 5). All of the sources 25 Table 4.--Analyses of variance for 4-year growth data at Kellogg and Allegan. Allegan Kellogg Trees With Trees With m Lamas Shoots Eight Lemmas Shag; 1968 1968 1968 1968 gm; gercent .c_m_,_ percent Range of source means 20 to 95 0 to 68 27 to 141 0 to 100 P value 14.9*** 7.5*** 3.,7*** 1.7*** Percent of variance source explains 58 39 58 27 ** m Significant at 1 percent level. Significant at 0.1 percent level. 26 Table 5.--Ana1yses of variance for flower production at nursery, Allegan and Kellogg. Nursery Allegan Kellogg Trees With Female Fls. Trees With Female Fls. Trees With Female Fls. Flowers For Trees Female For Trees Female For Trees Male Female With Fls. Flowers With Fls. Flowers With Fls. 1967 1968 1966 1968 1968 1968 1968 --percent--- number gercent number percent number Range of source means 0 to 2 to 68 73 1 to 5 0 to 34 0 to 9 0 to 88 0 to 8 F value 3.9*** 3.2*** 3.06*** 2.14*** 2.09*** 1.54* 1.44* Percent of variance source explains 43 35 34 i 11 10 21 18 * Significant at 5 percent level. *** Significant at 0.1 percent level. 27 that did not produce male flowers were from.the northern part of the range except for l3-Mai and 53-Que from the southern part of the range. Sources 58-Ont, 55-0nt, l9-Que, and 4-NS produced very few male flowers. The correlation (r) between male flower production and height was .55, and correspondingly with latitude at the place of seed collection -.56 (table 6). Somewhat lower were the correlations with length of growing season (r - .36), and growing degree-days (r - .35). Regional patterns of male flower production also existed. Sources east of 800 west longitude flowered more heavily than sources west of that line. The differences between these two regions was significant. Within the eastern region, sources that produced the most flowers were from.New Brunswick and central to southeastern Quebec (table 5). In the western region, sources from Wisconsin, Michigan and southeastern Ontario flowered heaviest. As a group, they differed significantly from the remaining northern and western sources. Female flower groduction.--Female flower production began in the nursery at age 2. At this time, all sources had flowers (table 5). Seventy-seven percent of the sources at Allegan and 83 percent at Kellogg had female flowers by age 4. Total flower production was heaviest in the nursery, where 26 percent of the trees had flowers, and each flowering tree produced an average of 2.8 flowers. Trees flowered more heavily at Kellogg than at Allegan. Twenty-six percent of the trees flowered at Kellogg, with 2.0 flowers produced per flowering tree. At Allegan, only eight percent of the trees flowered; each flowering tree produced an average of 2.8 flowers. From these results, it is apparent that growing conditions contributed to toal flower production. The growing conditions were best in the nursery and worst at Allegan. 28 Table 6.--Correlations (r) between flower production and climatic data at place of seed collection, based on source means. Flower Production Climatic Data Age When Length Grow- Degree- Character Place Measured Latitude ing Season(a) Days(b) 1535;, --------------- r -------------- No. female fls.‘ Nursery 2 -.43*** .39*** .40*** Allegan 4 -.41*** .3a*** .43*** .Kellogg a - 38*** .29** .39*** Percent of trees with female fls. Nursery 2 -.55*** .46*** .45*** Allegan 4 -.40*** .41*** .42*** Kellogg a -.22* .12 .11 No. male fls. Nursery 3 -.56*** 36*** .35*** (a) Number of days on which daily mean temperature exceeds 42°F. (b) Sum of positive values of (daily maximum in OF. + daily minimum in 01".) - 42. * fl Correlation significant at 5 percent level. ** Correlation significant at 1 percent level. *** Correlation significant at 0.1 percent level. 29 Female flower production was correlated with height at all test sites. The correlations were r - .62 at Allegan, r - .61 in the nursery and r - .39 at Kellogg. The correlations for female flower production between test sites were r - .40 between the nursery and Allegan, r - .26 between the nursery and Kellogg, and r - .37 between Allegan and Kellogg. Source X planta- tion interaction was significant between the nursery and both plantations and also between Allegan and Kellogg. The heaviest flowering sources were 44-Que, 79-Min, SS-Ont, 27-Que, 40-Ont, 46-Ont in the nursery; 734Mic, 54-Que, 26-Mai, 75-Mic, 79-Min, 50-Que, 69-Wis at Allegan; 75-Mic, 30-Que, 62-Ont, 73-Mic, and 50-Que at Kellogg. The least flower- ing sources at all test sites were usually from the Nerthwest Territory, Saskatchewan, Alberta, Manitoba and northern Quebec. With the exception of these sources, most of the remaining sources generally flowered heavily at one or more of the test sites. The correlations between female flower production and latitude, length of growing season, and growing degree-days at the place of seed collection are shown in table 6. The correlations with latitude were highest in the nursery and Allegan, r - -.56 and r - -.40 respectively, and lowest at Kellogg r - .22. This same general relationship between plantations holds true when female flower production is correlated with length of growing season and growing degree-days at the place of seed collection. AOV's Comgared to Correlations Variation patterns can also be recognized by properly comparing the results of analyses of variance and correlations between growth I characters and a feature of the habitat at the place of seed collection, 30 Table 7.-¥Comparison of the amount of variance explained by correlation (r2) and analysis of variance. Character Analysis of Variance Percent of Variance Between Regions Correlation Percent of Variance . Latitude Explains Height Autumn color Percent of trees ‘with male flowers Percent of trees with female flowers Number of female flowers Number of crooks Size of crooks Percent of trees with lemmas growth 55 44 17 10 11 28 24 24 31 30 19 49 45 32 31 such as latitude. The analysis of variance was performed on sources grouped by natural geographic regions and whose progeny had similar patterns of variation. It is designed to show differences between groups. The correlations calculated over the entire range, show the deviations from perfect correlation. The amount of variance between groups is then compared to the variance explained by the correlations. Provided there is an adequate sample size, it can be concluded that if more variance is explained by the correlations than by groups, the variation pattern is continuous. But if groups explain more variance than the correlations, the variation is discontinuous. I performed this type of analysis on growth characters that best differentiated seed sources. I divided the range of the species into nine regions that most logically grouped sources geographically, climatically and by similarity of growth traits. The results of the different analyses are presented in table'7. From this table, it is apparent that more of the variance is explained by straight line rela- tionships of the correlations than by regions. Consequently, it is best to consider the variation pattern continuous. §ummation~Of-Differences Analysis The third technique that I used to study the variation pattern was a summation-of-differences analysis. This is a multivariate analysis that combines data from several characters simultaneously. The eight characters used were: 2-year height, l-year autumn color, percent of trees with secondary growth, percent of trees with male flowers, percent of trees with female flowers, number of female flowers for trees with flowers, amount of deviation of a crook, and date of bud set. The formula used to calculate the summation-of-differences is as follows: Table 8.--Summation-of-differences table for sources east of 800 longitude. 32 total difference is s-ti. nits bsulsss sources listed below nd source a the lsft(s). Ths grsbsl here. is “are is illustrative of Small fists tstwsss tbs ssercss being csqstsd dustes similarity. let's sdsrs dissimilarity. place of clisel variatin. origin 0st. .- ber emd d N 3l° O|ee o unto-4 0 0 nose—nus n isséé en's: 2| NIO 0'00 2'000 8.0000 00000 00‘00 "0000 00"‘00 gonna-s 332:5 2:"sa '5 0'! XP 2'00 :‘000 3'0000 00000 00000 00000 00000 00000 00000 00000 0000" 0000'! HHNH‘ 33353 32232 :P 2P° 8'000 3'0000 00000 00000 00000 00000 00000 00000 00000 00000 00000 000N0 00000 0000‘ 000‘0 fluid—1N 00.0. 33 C" n sp 3'00 3'000 «10000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 0‘00" 0.400" 0-‘00-4 use-coon 00'40N «nun—see 'Oifltfllh 2 NS 16 Que 18 Que 30 Que 31 Que n 3P 2‘00 3.000 2'0000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 "0000 00000 “0000 00000 0‘00" "do—IN undo—0.4 red—nun NHHN—d nnvenN 05050 51 Que 28 Que 22 Que 26 MN 33 Que oloc m "coo 310000 00000 000-fl0 00000 00000 00000 00000 00000 000-40 000N0 00000 0"!0'00 0~0~0 000N0 0~0N0 00060 0~0¢u can-4'04 ONOQ—t on—sn—s use-son FOOCNH a N RIG 0 N|00 us 2'000 2.0000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 000~0 #0000 00050 0 3 £223. MONK“ NHHCN 0 M SP 0'00 N ”'000 N 0|0F‘00 N 0000‘ 00000 00"0"‘ 00000 00000 00000 00000 00"‘0‘ 00.40" 00‘0“ onus—so 00000 oaso—s—s F9000" 00‘0" 00‘0" 00"‘0'4 000-90 8 10 10 2 12 26 Mai 23 n 27 Que 40 Ont 39 Out 0 Q 4'0 ~e «5’00 ~s Nl000 ve ~|oooo C OOH-1'1 00N00 00-‘00 A0000 HNOOO u—sucn Nae—sou n—‘NON “ocean NdNOe-e Node-do“ 000400 nun-en nuoon fiNNO-i "NC-son o—o—s—sN NNNHN N's—sou Macon N‘NON unnnu nnnnn QCMNQ 11 13 11 13 16 16 10 41 (kit 42 (hit 43 Ont 44 Out 46 (int BM .05) {Mutual difference between sources - LSDLOSH (a) St—stioa unit - Table 9.--Summation-of-differences table for sources west of 80° longitude. 34 the sources hoist cqsred daeoes siuilerity. lee-e floss diseiuilsrity. tetel differeeco is s-stioo nits botueee sources listed N1. ad source so the left“). he eredusl increoo is o‘er-e is illustretive of hfll-‘ushuuu phaofchulnfluum at. .- 6er .d H .I :P 3'0 0 0‘0 0 0 0 :‘000'. 8382;? 8| 0 0 2P° ‘3'000 $|0000 00000 0000N 0000N 00"‘0N NN‘HO ”Sc 93 A16 as Out 96 Alt ”I. h ~| OF N :I°° ”.000 0 5000'! .I 00000 00000 0"‘000 00000 Oil's-60F. 00000 O‘HF‘H Hun—on I-OFINNO “.000 ”8‘ 93“. 66h: 76m: 77m: 3| 8P 3P0 moon mound 00000 00".00 000600 abduo conu— 0004‘" Gonna «mung OOhnn «noon III-6".“ .03.. 00:0. MON0O '60-. "can: .00 U .8213 33:2: 35 u ol NP 0 H 0 .I° H ’P00 8'0000 00000 00000 00000 00000 hands 00000 00000 ounun NNN00 vuvun {0000 u ounnc Qnouu I I 1 9 m 1I 16 13 12 12 11 U n u u 7 N U U U n 1 1I 12 12 11 16 11 10 10 ”an I1“: 63“: 62mt 61m: m 8P arc 3.000 3'0000 00000 00000 cow-sun 00000 00000 000—60 0000!! 00000-6 00000 NNnHO 0000" 00“". NHNU‘D. FIG-6N0“ “finch vene- 0'0“.“ OI. n n n u m m u n m 9 20 13 16 16 11 11 20 17 13 16 12 13 2I 23 23 22 1I 13 U u won SIN: 57m: mun nun 3| 3P aP° 2'000 :P000 0'4000 00000 00000 0d000 00000 ##000 FOO-0000 00000 0d000 00HNO 00H00 dance wanna 00'0“ NQHNN neuron—o nun-on “0'3“" 5.... 0055s 0300. D U m B u I 9 I 22 26 21 1I 23 20 1I 1I 16 12 11 26 20 19 19 16 17 16 26 1I 17 16 16 13 12 26 19 1I 19 16 N u nun 70 His SSNt 69 His uwu x|° N .P° «P00 h n'0000 5 00000 06°60 00°00 cocoa cocoa 0400004 00000 0000'. «coda uounu uonuu nuuun “000'. “00-60" Mod—IN" “000'. uouuu nuunn Goose ‘N.'. “"030“ sons. hnhh. I 7 I 1 9 22 21 20 17 13 11 11 1 9 22 20 20 20 19 1I 16 11 11 u u n 23 19 19 1I 16 16 13 27 1I 1I 20 1I 13 13 10 N N 27 22 22 22 19 1I 73Mc 76Mc 73Mc 12m: 56ht m 2P 2'00 3'000 3'0000 00000 00000 00000 00000 00000 00000 00000 00000 00000 NOH0~ «anon 0000" '60-‘08 H0000 «anon nouns Ndddfl Henn- «anon "anon anon. damn. Ndfldfl nonno ssh-o nnnno ones. 0500:: 00.0: 33:33:! N m 27 21 21 21 1I 19 31 23 23 23 22 20 36 27 27 27 26 22 2I 23 22 22 19 2I 22 23 22 19 ”an 66 His 67 His 71mc 36kt “DJ (.) ,_.-.“.- “-1: _ Z}(ectuel differeecs betueee sources - LSD(.03)) 36 summation-of-difference ,‘23 actual difference between sources - LSD§.03)). ' LSD(.05) . Four was used as a multiplier to eliminate the use of decimals. This analysis was done individually for each trait and then all of the summation units for each source were totaled. A total value of zero means that the sources being compared do not differ significantly in any of the eight traits. Increased values above zero indicate increased significant differences between sources for several traits. The tables give a visual presentation of the variation pattern. Sources from the same latitude in the eastern and western parts of the jack pine range differed. For this reason, I divided the sources into east (of 80° longitude) and west (of 80° longitude) sectors and prepared two separate summation-of-differences tables (tables 8 and 9). The sources for both tables were arranged by latitude, length of growing season and number of growing degree-days at the place of seed collection. The results of the summation-of-differences analysis show that the variation pattern is continuous. The summation units for all eight characters in both the east and west sectors increase gradually from north to south. No sharp breaks in the tables are evident. It must also be pointed out that there is no sharp break between the southwestern Ontario and southern Michigan sources. Yeatman (1967), working with seed from the same sources, also concluded from nursery, greenhouse and experimental plantations in Ontario, that the variation was continuous. There are two logical reasons for the continuous variation pattern. The first is that there are no sharp breaks in the range of jack pine and consequently there is a free interchange of genes. The second reason is that there is a gradual environmental gradient from south to north ‘wdthin the range of jack pine which has caused a gradual change in characters over the entire range. CHAPTER IV EVOLUTION AND HIGRATORY HISTORY Jack pine is now growing on an area that was almost completely ice covered during the Wisconsin glaciation of the Pleistocene. The Drift- less Area in southwestern Wisconsin.was the only part of the present range of jack pine not glaciated during this period. Prior to the Pleistocene, jack pine was generally believed to have been widely dis- tributed in Canada and the Great Lakes region (Mirov 1967, Yeatman (1967). Refuges for jack pine during the Wisconsin glaciation have been postulated in (l) a general area south of the Great Lakes, (2) the Yukon River valley, and (3) the Atlantic continental shelf (Halliday and Brown 1943). Schoenike (1962) studied morphological variation in natural stands of jack pine throughout its range. He concluded from his data, that the present population of jack pine probably came from one glacial remnant which was located south of the Great Lakes, However, Schoenike points out that an experiment such as his is not designed to determine how many refuges existed, and he did not exclude the possibility of there being several. Yeatman (1967) conducted a literature survey of geological and paleobotanical evidence, and concluded that during the Uisconsin glaciation jack pine survived at a single place in the .Appalachian mountains. He excluded the areas south and west of the 37 38 Great Lakes, and the Yukon River valley as possible refuges. He also thought that the continuous variation of his growth data was proof that there was only one center of origin for modern jack pine. The present study also indicates that the variation pattern is continuous. However, from this fact and the small amount of paleo- botanical evidence available, it cannot definitely be stated.whether the present population of jack pine came from one or several Pleistocene remnants. It is easiest to explain the variation pattern, however, if the present population is considered to have emerged from only one southern population. During its northward migration, the species would have adapted to the environment around it. Some important features of the environment would be length of photoperiod, temperature during the growing season, and length of growing season. Results from.my data indicate this to be true. On the other hand, it is also possible that distinct ecotypes in the present population of jack pine could have evolved, even if the entire present population came from trees that existed in only one refuge. More preglacial pollen samples, adequately dated and identified, are needed to resolve this question. CHAPTER V DESIGN OF INDIVIDUAL-TREE INEERITANCE STUDY Recently there has been a great amount of interest in seed orchard establishment and improvement through selective breeding. An individual- tree progeny test was conducted to determine whether there is sufficient genetic variation within jack pine from Lower Michigan to make selec- tive breeding worthwhile and also to develop seed orchards that will provide seed for future plantings in Michigan. This type of study per- mits an evaluation of the between- and within-stand components of variance. The selections were limited to jack pine from Lower Michigan because the 1951 Lake States provenance study and early results from the all-range provenance study showed that sources from Lower Michigan were fastest growing (Arend g; 9;. 1961). Materials and Methods Seed procurement.--I made individual-tree selections in the summers of 1965 and 1966 on land owned by the State of Michigan. Selections *were made in native stands that appeared to be growing on uniform sites. .Ages of individual trees within a stand.were determined by increment cores and found to vary by only five years above or below the average age of the stand. Most of the stands'were above average for height growth, stem straightness and number of trees with open cones. Several of the stands had been thinned by foresters. The trees they left were 39 40 tall, straight and many had open cones. I chose the best of these trees. The selection intensity in these areas was higher than my records indi- cate because it was bmpossible to determine how many trees were removed from the stand. For example, if a stand had 500 trees before thinning and 50 after thinning and I selected the one best tree remaining, 1 would say my selection intensity'was 1 in 50, when in actuality it was 1 in 500. Many of the stands were heavily attacked by the jack pine budworm (Choristoneura ping; Free.). Most of the trees with budworms did not produce any cones, so my selections were mainly limited to trees that were not attacked. I selected for height, stem straightness and presence of open cones. Trees with open cones were selected because fire danger has been greatly reduced in the state and cones that shed seeds freely are needed for natural regeneration. The feature of open versus closed cones appears to be under simple genetic control. An open-pollinated progeny test started in Minnesota in 1939 showed that 52 percent of the offspring collected from open-coned parent trees produced open cones, whereas, only 13 percent of the offspring collected from closed-coned parent trees produced open cones (T. D. Rudolph, Schoenike, Schantz- Hansen 1959). It appears to be a relatively easy task to obtain a large portion of trees with open cones in the first generation. The superiority of each tree was judged in relation to other trees growing within 200 feet. The selection intensity was such that a selected tree was best on about 2 acres for one, two or three traits. Each tree was evaluated separately for each trait. Average- and below- average trees were also chosen from each stand. Cones were collected by felling the trees. I collected seed from 199 trees in 40 stands in 41 the fall of 1965, and 183 trees from 21 stands in the fall of 1966. Seed collection areas are shown in figure 2. Data on stands is present- ed in tables 10 and 11. Individual parent tree data is available upon request. I Handling in nursery.--Seeds from both years of selection were planted at Michigan State University's Tree Research Center at East Lansing. The first year selections were sown on May 23 and 24, 1966, and those from the second year selections were sown on May 15 and 16, 1967. The design for both experiments consisted of 4 randomized comp plete blocks. Each block contained one 4-foot row of each seedlot. Twenty sound seeds were evenly spaced in a row. The rows were one foot apart. Germination in both years was over 90 percent. After germina- tion, the seedbeds were watered, kept weed free, fertilized and mulched with one-half inch of sawdust in the autumn. The seedbeds were relatively uniform except for fifteen feet on both ends. This area did not get watered adequately and consequently height was reduced. I applied a correction factor for height growth, since it was affected by the location of a seed source in the replicate. I accomplished this by first dividing each replicate into 20 equal parts and recording in.which of the 20 parts a source was located. Then I calculated the difference between the plot means and the overall average for the same sources. Then the differences for all sources in each of the 20 parts were added and the average difference for the sources in each of the 20 parts was calculated. Then an amount equal to the average difference, but apposite in sign, was added to the mean for each plot in each of the 20 parts. Seed weights were measured and not found to be significantly correlated with height, so no correction.was made for them. Figure 2.--Distribution of the parental stands (black dots) from which the half-sib progenies were obtained. County names are as follows: EMM - Emmet, CHE - Cheboygan, RIV- Presque Isle, OTS - Otsego, MOM - Montmorency, ALP - Alpena, BEN - Benzie, GT - Grand Traverse, KAL - Kalkaska, CRA - Crawford, OSC - Oscoda, HEX - Hexford, MIS - Missaukee, ROS - Roscommon, OGE - Ogemaw, IOS - Iosco and LAX - Lake. 42 43 --.;T---..---m--- H a u m 0 O u .3 u u m u e u m u m L" n u m. m .m 5: m a 1 u n u u u ......... e e e" u __ Q. .m . m 53 m m m_s. “ u e. n e “ :uuuuuumnuubminl rs uuuuuuu .“r uuuuuuuuuuuuu w:....: on n e u ...... O O" u n C e u __ e e u («u m iv. n be u 2% 0. 0 u. 0 u _ . . n m 1-17-11, 44 Measurements.--Eight sets of measurements were made on the seed- lings sown in 1966 and four sets on the seedlings sown in 1967. Metric traits were measured to an accuracy of approximately 1/20 of the range between extremes. Hon-metric traits were evaluated in grades of the smallest recognizable difference between units. Color was defined in terms of live-tree standards, and was scored by comparing the average color of a plot with-a certain color standard set in the observer's mind. Analysis.--An analysis of variance was calculated for each trait studied. A typical analysis of variance table, using progeny means as items, follows: Source of Variance Degrees of Freedom Expected Mean Squares Source S-1 032 +110;2 Replication R-l OLZ + 80;.2 Replication x source (S-1)(R-l) qu Total (8 x R)-1 R and S are the number of replicates and sources, respectively. Differences between sources were tested with the replication x source ‘mean square. The analysis of variance table for the within? and between-stand components of variance for the sources sown in 1966 follows: Source of Variance Degrees of Freedom Expected Mean Squares Sources 2 2 2 Between stands 39 O; + 405 + 2005 Within stands 159 0.32 + 4032 Total 198 45 A similar table, but with different degrees of freedom and exponents for the components of variance, was constructed for the sources sown in 1967. Simple correlations , using source means as items, were calculated between latitude and length of growing season at the place of seed collec- tion, and height, autumn coloration and flower production. Correlations were also calculated between characters of possible biological signifi- cance. The correlations and analyses of variance were done on Michigan State University's CDC 3600 digital computer. CHAPTER VI VARIATION AMONG MICHIGAN HALF-SIB PROGENIES Origin data and the average performance of progenies from each of the 40 stands sown in 1966 and 21 stands sown in 1967, is summarized in tables 10 and 11. The genetic variation pattern within the Lower Michigan population of jack pine is random. Individual Traits Trends which were quite evident in the range-wide study are not apparent in the Lower Michigan population. Correlations between progeny performance and climatic data at the place of seed collection are given in table 12. Several of the correlations are reversed from those in the rangedwide study. Many characters indicate different trends if correla- ted with latitude or length of growing season. One possible reason for this is that Lakes Michigan and Huron moderate the climate of areas around them. Consequently, within the natural range of jack pine in Lower Michigan, the areas most favorable for growth are found along the southern, western and eastern parts of the range. The area least favor- able for growth is centered in Roscommon, Kalkaska and Missaukee counties. flgigh£.-2Areas which produced the fastest growing progenies appear to be randomly located. The fastest growing sources from the selections sown in 1966 came from Oacoda, Iosco, Roscommon, Missaukee and Grand Traverse counties. Those from the sources sown in 1967 came from Alpena, 46 47 Table 10.--Analyses of variance for Michigan l/2-sib progeny test sown in 1966. Trees With Trees With Lemmas Female Female Height Shoots Flowers Flowers 1968 1967 1968 1968 percent of per 40 overall mean percent percent trees Range of source means 86 to 106 5 to 30 2 to 23 2 to 13 F value 0.990 2.061** 0.839 0.928 ** Significant at 1 percent level. 48 Table 11.--Analyses of variance for Michigan 1/2-sib progeny test sown in 1967. Trees With Lemmas Height Color Shoots 1968 1968 1968 percent of overall mean grade percent Range of source means 84 to 128 4 to 9 4 to 17 F value 1.638 1.007 2.739** ** Significant at 1 percent level. 49 Lake and Grand Traverse counties. Not all of the sources were sampled both years (tables 10 and 11). Cplgp,--Autumn color in the sources sown in 1967 was correlated with latitude. Sources from two bordering northeastern counties (Alpena and Montmorency) turned the deepest shade of purple, while sources from the southernmost county (Lake) remained greenest (tables 11 and 12). Lemmas shoots and female flowers.--Lammas shoots and female flowers were abundantly produced on sources from all areas. No area produced appreciably more or less amounts of either. The amount of variance explained by source for both characters was small (table 13). getween- and Within-Stand Variance Components The individual parent progeny data obtained in this study permits an assessment of the amount of variance that exists between- and within stands. This information is valuable because it gives the breeder a clue as how to proceed in his selection program to achieve maximum genetic gain. The method used for this analysis was given earlier in the 'Analysis' section. The results for the growth data from both years of seed collection are given in table 14. The parent-within-stand variance was significant for height, trees with female flowers, number of female flowers and autumn color. Improvement by family selection would be an efficient method of improvement for these traits. On the other hand, the variance between stands was appreciable for only trees with lemmas growth and autumn color. Stand selection would be a productive way of improving these traits. Autumn color was the only character that had significant amounts of both within- and between-stand variance components and a combination of stand and family selection could be used to improve it. 50 Table 12.--Correlations (r) between growth characters and climate at place of seed collection for Michigan l/2-sib progenies, summarized by stand. Tree Character Climatic Variable Was the Trend in the Year Tree Age When Frost-Free Same Direction as in Sown Character Measured Latitude Days the Rangewide Study? frost:free years ---r ------- r---- latitude _days 1966 Height 3 -.52 -.15 Yes 'no Trees with lammas growth 3 +.21 -.03 No No Trees with female fls. 3 -.l3 -.17 Yes No Female fls. per source 3 -.ll -.22 Yes No 1967 Height 2 -.30 +.50* Yes Yes Trees with lemmas growth 2 -.64** +.06 Yes No Color 2 -.51** -.33 Yes No Correlation significant at 5 percent level. ** Correlation significant at 1 percent level. l _ 51 Table l3.--Ana1yses of variance for Michigan 1/2-sib progeny tests. Standard Percent of Deviation Variance Date Unit of of F Explained Character Measured Measurement Source Mean Value By Source Sown in 1966 - Degrees of freedom for sourcezl98, replicate:3, error:594 (4) Height 9/68 Inches 6.150 1.825*** 17 (S) Lemmas shoots 6/67 1 of trees 0.087 1.319*’ 7 (6) l-year cones 6/68 1 of trees 0.093 1.249 6 (7) l-year cones 6/68 No.per tree 0.148 1.281 7 Sewn in 1967 - Degrees of freedom for source:182, replicate:3, error:546 (4) Height 10/68 Inches 2.886 3.304*** 37 (5) 0616: 10/68 Grade 1.254 1.854*** 18 (6) Lammas shoots 10/68 Z of trees 0.083 1.173 4 ** Significant at 1 percent level. *** Significant at 0.1 percent level. 52 Table 14.--Within-stand, between-stand, and error variances expressed as a percent of the total variance. Percent ongotal VariancegQue to Parent Year Age When Within Sown Character Measured Stand Stand Error years 1966 Height 3 17.1*** o 82.9 Trees with lammas growth 3 2.1 s.3** 92.6 Trees with female £16. 3 ' 5.3** 0 94.7 Female fls. ** per source 3 6.9 0 93.1 1967 Height 2 31.9*** 7.1 61.0 Trees with lemmas growth 2 3.9 0.5 95.6 Color 2 10.7*** 12.3** 77.0 ** Significant at 1 percent level. *** Significant at 0.1 percent level. 53 The amount of variation that exists within a typical stand can be seen in table 15. The stand is located on a level site in Crawford County and is growing on Rubicon sand. At the time of seed collection, the average height of the stand was 40 feet and 90 percent of the trees were attacked by the jack pine budworm. The progenies at age 3 varied in height from 19.6 to 26.1 inches. A breeder trying to obtain increased height growth in this population at this age could get substantial improvement by selecting family 4253. He‘would obtain an increase (over the stand average) of 17 percent. It was not possible to select a genotypically superior stand for height growth. For example, three similar-aged parent stands, growing on level sites on Rubicon sand in Presque Isle County were 39, 50 and 61 feet. The average 3-year heights of the progeny from each stand were respectively: 23.3, 22.6 and 22.9 inches. In this instance, the short- est parent stand produced the fastest growing offspring. Effectiveness of Selection for Height Parent trees were selected for height, stem.straightness and non- serotinous cones. Some indication as to the effectiveness of selection for height can be evaluated at this time. One method that I used to evaluate the effectiveness of selection for height was to divide parent trees into 'control' and 'select' groups, and then I compared progeny performance between the two groups. The 'control' group consisted of parent trees that were either below-average or average, when compared to the average height of trees in the stand. Parents in the 'select' group were at least 10 percent taller than the average height of trees in the stand. 54 Table 15.--Parental and 3-year progeny growth data for individual-tree selections within a randomly chosen stand. PrQfiSEX. Trees With Trees With Height Lemmas Female Female ‘ MSFG Parent Sept. Shoots Flowers Flowers. Height 1968 Oct. 1967 June 1968 June 1968 (1) (2) (3) (4) ' (5) percent of per 40 feet stand mean percent percent trees 4247 50 90 4 4 l 4248 44 108 23 ll 4 4249 48 103 32 16 6 4250 50 104 5 15 6 4251 44 89 5 3 2 4252 46 103 ll 0 0 4253 40 117 10 16 6 4254 52 88 11 13 S 55 The results show that selection for height growth in natural stands was only slightly effective in 1966 and nil in 1967. Progenies from 118 parents in the 'select' group in the sources sown in 1966 averaged 23.48 inches at age 3, whereas 81 progenies from the 'control' group averaged 23.38 inches. In the 1967 sown sources, 96 progenies from the 'control' group averaged 7.54 inches and 87 progenies from the 'control' group averaged 7.52 inches. Neither of the differences between the groups was statistically significant. The second method that I used to evaluate the effectiveness of selection for height was to correlate the height of parent trees with progeny height. For this analysis, I expressed parental heights as a percent of the average height of the stand, and progeny heights as a percent of the average overall progeny height. This permitted a direct comparison of the parental trees, of various ages growing on different sites with progeny performance. The results from these parent-progeny correlations agree with those received above from analysis of the 'select' and 'control' groups. The correlation between parental height and progeny height in the study sown in 1966 was (r - .16). Therefore (r2) or 2.6 percent of the pro- genies' height could be explained by parental height. Stated another way, the heritability (ability of a parent to transmit to its offspring) was 2.6 percent for height growth. In the sources sown in 1967, the 2 was negligible. parent-progeny correlation was 0.0025, and r Results from.the within- and between-stand analysis show there is sufficient variation within stands to make individual tree selection worthwhile. The difficulty lies in selecting a genetically superior tree for height growth. Results from the individual tree selections 56 show that mass selection for height growth in natural stands is rela- tively ineffective when the progenies are three years old. Jack pine is noted for its rapid juvenile growth and possibly the true genetic potential of the families will not be realized until the trees are older. Continued testing of the progenies will show whether or not this is true. CHAPTER VII JACK PINE BUDWORM RESISTANCE STUDY The jack pine budworm has caused much damage in Michigan. It attacks young and old jack pine in both natural stands and plantations. The larvae feed primarily on staminate flowers and newly emerging needles. Several successive years of defoliation can kill large num- bers of trees. Certain varieties of Scotch pine have been shown to be resistant to European pine sawfly (Neodiprion sertifer (Geoff.)) attack by providing unsuitable sites for ovipositing (Wright _£,pl, 1967). This study was undertaken to determine if there is resistance to ovi- positing by the jack pine budworm between widely diverse individual- tree selections of jack pine. Life Cycle of_the Jack Pine_§udwg£! Winter is spent as nonfeeding second-instar caterpillars in silken cases called hibernacula. These cases are spun in old staminate flower bracts, under bark scales on the trunk and larger limbs, or between needles. In spring, the caterpillars emerge and feed for about six weeks. During this time, the larvae go through five additional instars. Pupation begins from late June to late July, depending on locality and weather, and moths issue from early July to early August. A female can lay about 150 eggs. Hatching occurs about two weeks after the eggs are laid. A few days later, without feeding, the first-stage caterpillars 57 58 spin their hibernacula, molt to the next stage and hibernate (MacAloney and Drooz 1956). Materials and Methods The trees used were 2-0 seedlings from the Michigan individual-tree selections sown in 1966. Twenty-five sources were planted on a one-foot spacing at the Tree Research Center in April 1968. A randomized com- plete block design was used with single tree plots. The experiment contained three blocks. Each block contained four replicates of the 25 sources. The blocks were 10 feet square. A cage 2 feet high and 11 feet square covered each block (figure 3). Larvae and pupae were collected in the first week of July from three places near Grayling, Michigan. The larvae were reared on jack pine foliage until pupation was complete. The pupae were then separated by six. A total of 1,052 female pupae and 826 male pupae were obtained. The male and female pupae were then split into three equal groups. Then one of the three groups of male pupae was placed in a box with one of the three groups of female pupae. The three boxes were kept at 50°F. The moths soon emerged and mated. The males and gravid females were released in the center of each block of trees on July 27. One box of insects was released in each of the three blocks of trees. The moths dispersed readily throughout the cages. Egg laying started 2 days after the insects were released and was finished within a week. The easily visible egg masses were counted on each tree when egg laying was completed. Other measurements taken on each tree were height, fresh weight of the foliage, and.width of 10 needles. Figure 3.--The cages pictured here are 2 feet high and 11 feet square. Each cage covers a block of 100 trees. '1 59 60 61 Results and Discussion The female moths found the environment suitable for ovipositing and laid 1,030 egg masses, which is almost one egg mass for each released adult female. They laid eggs on 244 of the 300 trees. How- ever, the female moths found no sources which were unsuitable for ovipositing. The 'F' value for the number of egg masses laid on the different sources was 1.007, which was nonsignificant. I also performed correlations between the number of egg masses and height, amount of foliage, and needle width. All of the correlations were nonsignificant at the 10 percnet level. Height of the sources produced the highest correlation (r - .23), amount of foliage was next with (r - .18), and needle width was lowest (r - .08). From these low correlations, it seems that sources may have been chosen for ovipositing because they were tall and had a large amount of foliage. Parental trees of 22 of the 25 selections were not attacked by the jack pine budworm. Whether they were not attacked because of genetic reasons or because they were escapes is not known. The results from this study indicates that they were probably escapes. At this early age, however, it cannot be determined whether all of the sources will be susceptible when they reach maturity. CHAPTER VIII RESPONSE TO DIFFERENT LEVELS OF ROOT PRUNING Seedlings are frequently root pruned when transplanted or field planted in order to make planting easier and to produce a dense, fibrous root system. Reduced height growth results whenever seedlings are moved. The fine lateral roots are damaged when the seedlings are lifted and consequently the absorbing area of the roots is reduced. The pur- pose of this experiment was to determine if different levels of root pruning affect the height growth of transplanted seedlings. Materials and Methods The seedlings used were 2-0 stock from the Michigan individual-tree selections sown in 1966. At the end of their second year, the seedlings were about 9 inches tall and had well developed root systems. In March of the following year, a lifter was run under the seedlings that were to be transplanted and cut the roots at 8 inches. The lifter was not run under seedlings that were to remain in the nursery. I applied further root prunings to the seedlings that the lifter had run under. The levels of root pruning were: none, one-fourth, and one-half of the length of the existing roots. The lengths of the roots after the three levels of pruning were respectively: 8, 6 and 4 inches. Each level of root pruning was applied to 30 plots, with four trees per plot. The seedlings were planted on a one foot square spacing on April 15, at 62 the Tree planting I P differen The underCut leaSOn. and Vere r“PM!“ that ha: fast as First Ye quarter. Tvaed bet‘EQQ initial The differen 63 the Tree Research Center. Heights were measured immediately after planting and again in September. I performed an analysis of covariance to adjust for initial height differences. The form for the analysis of covariance follows: Source of Variation Degrees of Freedom Total 89 Between pruning levels 2 Within pruning levels 87 Reduction due to regression 1 Deviations from regression 86 Rppplts and Digcugpigp The seedlings that remained in their original seedbeds and not undercut by the lifter were 23 inches tall at the end of the growing” season. The seedlings that were transplanted had reduced height growth and were only 12 inches tall. However, there was no differential response to the three different levels of root pruning. The seedlings that had one-half of the length of their existing roots removed grew as fast as those that were not pruned beyond what the lifter had removed. First year growth increments for the seedlings that had none, one- quarter, or one-half of the length of their existing root systems removed were respectively: 2.66, 2.64 and 2.68 inches. The 'F' value between the different levels of root pruning, after adjusting for initial height, was 0.0046. There are several possible reasons why the seedlings did not have different heights as a result of the various levels of root pruning. Jack [aim the root The roots probably The to the a: length 01 absorbing the surf: of the 1. of the a] Anot between 1 were giv. tal “re and feta Ente in Sire“ c “Calved It of degree Ii“Went takes t 1 64 Jack pine is noted for its rapid juvenile growth and it is likely that the root systems grew rapidly and soon attained maximum absorption area. The roots were pruned before growth started in the spring, and they probably elongated very rapidly with initiation of growth. The roots were pruned according to their length and not actually to the amount of absorbing area. Thus to say that one-half of the length of the roots was removed does not mean that one-half of the absorbing area was removed. The roots were much more developed near the surface and tapered down to a point. Therefore, removing one-half of the length of the root system actually only removed about one-third of the absorbing area. Another possible reason why there was no difference in growth between the different levels of root pruning is that the seedlings were given good growing condtions. They were not subject to environmen- tal stress. In the nursery environment, they had ample room to grow and received no competition for water or nutrients. Perhaps a differ- ence in height growth would have resulted if they were grown under stress conditions of inadequate water and growing space, or if they received further competition from competing vegetation. It is probable that transplant 'shock' is not so much a function of degree of pruning or surface area, but simply the fact that reestab- lishment of root contact with soil and renewed absorption capabilities takes time and results in net growth loss. T} nance t Siva sc for hej Pet away. it Van Permane three 1 Perme Si smdy ‘ soul-CeI 11Shed CHAPTER IX POSSIBILITIES FOR FUTURE IMPROVEMENT The results obtained to date from the Petawawa range-wide prove- nance test and the Lower Michigan individual-tree inheritance study give some insight as how to proceed in genetically improving jack pine for height growth. The number and distribution of samples in the Petawawa provenance test was large enough to clearly show the geograph- ic variation pattern. No further provenance tests are needed. The permanent plantations of this study will continue to be measured every three to five years in order to calculate correlations between progeny performance at various ages. Six plantations of the Lower Michigan individual-tree inheritance study were established in the spring of 1968 with 2-0 stock from the sources selected in 1965 and the remaining three plantations were estab- lished in the spring of 1969 with 2-0 stock from the sources selected. in 1966. The plantations have replicated randomized complete blocks with 5 and 10 tree plots. There are 40 trees per family in each planta- tion. The spacing between trees is 6 by 6 feet and the rows are straight in both directions. No significant difference was found between stands for height growth in the nursery, so in the future each family will be evaluated on its own performance. The plantations will be measured.when flowering starts. 65 Additional me: branch diamett to disease am analyzed at 5- The plan! after each set selection by 1 seed quality. plantations . 10,000. The , families and 4 Will Sive a m “1088. there i ‘t ‘88! 10, 1] increase Vith Removal 1 c(Inversion of be Undertaken of cone Colle. include Comm. branches. The 1‘pa 'mr'l‘Senen ‘32 20 by LP; 66 Additional measurements taken at this time besides height growth include branch diameter, stem.straightness and other traits such as resistance to disease and insects. Later sets of measurements will be made and analyzed at S-year intervals. The plantings will be thinned by removal of the poorest families after each set of measurements. This type of thinning is termed 'family' selection by plant breeders and is the process which leads to improved seed quality. There are approximately 30,000 trees growing in the six plantations. After the first thinning, this number can be reduced to 10,000. The subsequent thinnings will continue to remove the poorest families and also the poorest individuals within the best families; this will give a combined 'family and mass' selection gain. After these thin- nings, there will be approximately 4,000, 2,500 and 1,000 trees remaining at ages 10, 15 and 20. The genetic quality of the seed produced.will increase with each thinning. Removal of the poorest families from the progeny tests marks the conversion of the plantations to seed orchards. Cultural treatments may be undertaken in the seed orchards to promote seed production and ease of cone collection. Some of the cultural treatments that can be applied . include complete weed control, fertilization, and pruning of the lower branches. The l-parent progeny test-seed orchards are the first step in a several-generation improvement program. They'will be outmoded at about age 20 by 2-parent progeny test-seed orchards capable of producing seed of higher genetic quality. A 2-parent progeny test is produced by arti- ficial pollination so that both parents are known. The identity of each tree x tree combination is maintained separately. The 2-parent progeny tests give twice as much gain as l-parent progeny tests of the same size. The c quate flow eatablishe crossing i possible 1 will be 1!: seeds) pe give maxi All lination identity will be tree se‘. in the : “7 Per “Oat of £0110“ Tl them c; obtaim exPeCt Senera 'I "ith c is the V111 l tion 67 The controlled pollinations will be made as soon as there is ade- quate flower production in the l-parent progeny test plantations already established. This will be when the trees are about 8 years old. The crossing scheme will consist of making as many biparental crosses as possible with the 5 best trees in the 25 best families. Each cross will be made in sufficient quantity to produce about 10 cones (300 + seeds) per tree x tree combination. This type of crossing scheme will give maximum gain. All control-pollinated cones will be collected 16 months after pol- lination and the seeds will be extracted immediately. Each seedlot's identity will be maintained by seed and pollen parent. Nursery handling will be as was previously described for the Lower‘uichigan individual- tree selections. Permanent progeny test-seed orchards will be established in the spring with 2-0 stock. The data from the l-parent progeny test may permit some culling of undesirable progenies in the nursery. However, most of the thinnings and subsequent conversion to seed orchards will follow the procedure previously described for the l-parent progeny test. The process of selecting superior families and making crosses between them can be continued into the third and fourth generations. The progress obtained in the first two generations will indicate what gain can be expected if the breeding program is continued into these later generations. The present information also shows that it is best to leave trees with open cones in areas that are to be thinned and natural regeneration is the only way to obtain reproduction. Most of the trees in the area will have Open cones if the procedure is carried on two or three rota- tion periods and fire will not be needed for seed dispersal. him, A. Arend, J Batzer, C3onrd, LITERATURE CITED Alm, A. A., and R. A. Jensen. 1969. Results of a Lake States jack pine seed source test in east-central Minnesota. Univ. Minn. Notes No. 200. 4 pp. Arend, J. L., N. F. Smith, S. H. Spurr, and J. W. Wright. 1961. Jack pine geographic variation--five-year results from Lower Michigan tests. Mich. Acad. Sci., Arts, and Letters Papers 46:219-238. Batzer, H. O. 1961. Jack pine from Lake States seed sources differ in susceptibility to attack by the white pine weevil. Lake States Forest Expt. Sta. Res. Note 595. 2pp. Cayford, J. H., 2. Chrosciewicz, and H. D. Sims. 1967. A review of silvicultural research in jack pine. Can. Dept. Forestry and Rural Deve10pment, Forestry Branch Departmental Publ. No. 1173. 255 pp. Critchfield, W. B., and E. L. Little. 1966. Geographic distribution of the pines of the world. U.S. Dept. Agric., Misc. Publ. No. 991. 97 pp. Durzan, D. J., and V. Chalupa. 1967. Free sugars, amino acids, and soluble proteins in the embryo and female gametophyte of jack pine as related to climate at the seed source. Can. J. Botany 46:417-428. Fowells, H. A. 1965. Silvics of forest trees of the United States. U.S. Dept. Agric. Handbook 271:338-354. Fowler, D. P. 1965. Natural self-fertilization in three jack pines and its implications in seed orchard management. For Sci. 11:55-58. Giertych, M. M., and J. L. Farrar. 1961. The effect of photoperiod and nitrogen on the growth and development of jack pine. Can. J. Botany 39:1247-1254. . 1962. A provenance study in jack pine seedlings. Silvae Genetica 11:111-114. Graham, S. A., 1935. The spruce budworm on Michigan pine. Mich. Univ. School of Forestry and Conservation. Bull. 6. 56 pp. 68 impor 301“, J. Jon hes}: lwakava , H. , 1967. proge Fore: Jensen, R. A jack Note: King. J. P. Proce HacAloney, H U.S. Kirov. N. T. Nienstaedt, 1 Proc« RUdolph, T. 1 0f o: and 2 pp Ru“101101. r. Lake SChant Z‘Hans on g SCthnike ’ R Char Minn SChoenike ’ R Char Minn SChoenike’ R Lamb Smith, N. F, Mich Stoeckler. J 0f .1 Aeit 69 Halliday, W. D., and A. A. Brown. 1943. The distribution of some important forest trees in Canada. Ecology 24:353-373. Holst, J. J., and C. W. Yeatman. 1961. A provenance test in Pinus banksiana Lamb. Recent Advances in Botany 1612-1616. Iwakawa, M., M. Watanabe, T. Sato, M. Susumu, M. Inuma, and S. Kida. 1967. Inheritance of some characters in open-pollinated progenies of Pinus densiflora Sieb. and Zucc. Bul. Gov. Forest Expt. Sta., Tokyo. No. 207:31-67. Jensen, R. A., T. Schantz-Hansen, and P. 0. Rudolf. 1960. A study of jack pine seed sources in the Lake States. Univ. Minn. For. Notes No. 88. 2 pp. King, J. P. 1964. Ten-year height growth in Lake States jack pine. Proceed. seventh Lake States tree improv. conf. 84-88. MacAloney, H. J. and A. T. Drooz. 1956. The jack pine budworm. U.S. Dept. Agric. forest pest leaflet No. 7. 4 pp. Mirov. N. T. 1967. The genus Pinus. Ronald Press Co. 602 pp. Nienstaedt, H. 1960. The ecotype concept and forest tree genetics. Proceed. fourth Lake States forest tree improv. conf. 14-25. Rudolph, T. D., R. E. Schoenike, and T. Schantz-Hansen. 1959. Results of one-parent progeny test relation to the inheritance of open and closed cones in jack pine. Uhiv. Min. For. Notes No. 78 2 pp- Rudolph, T. D. 1964. Lammas growth and prolepsis in jack pine in the Lake States. For Sci. Monograph No. 6. 70 pp. Schantz-Hansen, T., and R. A. Jensen. 1952.. The effect of seed source on growth of jack pine. J. For. 50:539-544. Schoenike, R. B., T. D. Rudolph, and T. Schantz-Hansen. 1959. Cone characteristics in a jack pine seed source plantation. Univ. Minn. For. Notes No. 76. 2 pp. Schoenike, R. B., T. D. Rudolph, and R. A. Jensen. 1962. Branch characteristics in a jack pine seed source plantation. Univ. Minn. For. Notes No. 113. 2 pp. Schoenike, R. B. 1962. Natural variation in jack pine (Pinus banksiana Lambert). Ph. D. Thesis. Uhiv. Minn. 211 pp. Smith, N. F. 1952. Michigan trees worth knowing. Forestry division, Mich. Dept. Cons. 60 pp. Stoeckler, J. H., and P. 0. Rudolf. 1956. Winter coloration and growth of jack pine in the nursery as affected by seed source. Aeitschr. f. Forstgenetik u. Forstpflanzenzuchtung 5:161-165. Wells, 0. G! Wilson, L b1 PIN- 5 Fright. J 1 Wright 1 , fi 1 Wright, Yeatman! 70 Wells, 0. 0. 1964. Geographic variation in ponderosa pine. Silvae Wilson, Wright. Wright, Wright, Genetica 13:89-103. L. F. 1963. Host preference for oviposition by the spruce budworm in the Lake States. J. Econ. Ent. 56:285-288. . 1964. Oviposition site of the spruce budworm, Choristoneura fumiferana, modified by light. Annals of the Ent. Soc. of. ‘Amer. 57:643-645. ., and Bean, J. L. 1964. Site of spruce budworm egg masses their preferred hosts in the Lake States. J. Econ. Ent. 56:574-578. J. W., and Bull, W..I. 1962. Geographic variation in European black pine--two-year results. For.Sci. 8:32-42. J. W. 1963. Genetic variation among 140 half-sib Scotch pine families derived from 9 stands. Silvae Genetica 12:83-89. ., W. L. Lemmien, and J. Bright. 1963. Geographic variation in eastern white pine--6-year results. Mich. Agric. Expt. Sta. Quart. Bul. 45(4):691-697. . 1966. Early flowering patterns in Scotch pine. Mich. Agric. Expt. Sta. Quart. Bul. 49(4):189-199. J. W., L. F. Wilson, and W. K. Randall. 1967. Differences among Scotch pine varieties in susceptibility to EurOpean pine sawfly. For. Sci. 13:175-181 Yeatman, C. W. 1967. Geographic variation in jack pine (Pinus banksiana Lamb.) seedlings. Ph. D. Thesis. Yale University. 283 pp. VITA David S. Canavera Candidate for the Degree of Doctor of Philosophy Final Examination: October 21, 1969 Guidance Committee: Drs. M. W. Adams, J. E. Cantlon, J. W. Hanover, L. F. Wilson, S. N. Stephenson, J. W. Wright (Chairman) Dissertation: Geographic and Stand Variation in jack pine (Pinus banksiana Lamb.) Biographical Items: Born September 4, 1943, Norway, Michigan Married Nancy Nelson, June 21, 1969 Education: Michigan Technological University, BSF, 1965 Michigan State University, MSF, 1967 Michigan State University, Ph.D., 1969 Experience: Forestry Aide, U. 8. Forest Service, Rapid River, Michigan, summers of 1963 and 1964 Conservation Aide, Michigan Department Natural Resources, Michigan State University, 1965 to present Organizations: Society of American Foresters Xi Sigma Pi ‘ - 71 APPENDIX 72 73 Table 16-Four-year growth data at Allegan and Kellogg, planted 1966. Ont. num- Allgggn Kgllggg bar and Height Trees with Height Trees with place of Age lemmas age lemmas origin 4 shoots 4 shoots 1968 .1968 .1968, .1968 3200 on. ~ meat on. patent + Sources west of 80° west longitude. 99 INT 2O 0 27 O 98 NWT 39 8 38 0 96 m 33 o 65 a 92 Sas 3h 0 47 0 91 Man #1 0 68 O 89 Sea 45 0 65 0 95 Alb 95 3 55 9 85 Ont 50 3 78 8 9L; All) 52 0 1+1 0 90 Man 59 6 59 0 87 See 54 0 82 0 93 Alb 55 3 88 8 6h Ont 65 10 87 33 76 Out 55 5 73 0 77 Ont 65 O 71 0 88 Sea 70 10 96 O 86 Ont 58 13 76 13 59 Mic 76 7 62 o 83 Man 63 20 93 13 84 Ont 56 10 9h 28 82 Ont 61 6 85 17 81 Ont 65 10 96 25 63 Ont 53 16t 79 13 62 Ont 59 8 109 17 61 Ont 72 10 99 0 60 out 72 8 1114 0 58 Ont 63 19 87 17 57 out 76 13 137 17 80 Min 76 8 109 8 79 Min 84 27 84 25 78 Min 75 16 109 25 70 Wis 76 55 123 28 55 Ont 7o 20 105 17 69 Wis 82 148 106 36 68 Wis 72 1+6 88 17 74 Table 16--Continued. Ont. num- Allegan Kellogg her and Height Trees with Height Trees with place of Age lemmas age lemmas origin 4 shoots 4 shoots 1968 1968 ‘1968 1968 3200 ea. 29.119111 on. 28.19.8111. + 75 Mic 84 53 128 13 74 Mic 8O 48 104 13 73 Mic 95 18 93 O 72 Mic 75 52 134 38 56 Ont 74 31 102 63 65 Wis 83 62 97 25 66 Wis 81 68 123 38 67 Wis 91 55 141 71 71 Mic 84 64 94 100 5a Ont 94 74 91 0 Sources east of 80° west longitude. S3 Que 41 O 47 0 5 NS 37 o 59 o 4 NS 43 3 33 O 38 Que 48 0 67 O 21 Que 46 5 70 o 20 Que 66 10 64 O 19 Que 48 0 4O 0 3 NS - - - - 1 NS - - - - 15 Que 48 3 7O 42 37 Que 59 0 79 25 32 Que 50 6 81 38 13 Mai 56 15 84 38 36 Que 56 o 74 o 35 Que 60 0 73 33 34 Que 57 0 59 0 so Que 59 o 100 13 29 Que 60 O 71 0 14 N8 61 0 78 25 6 NS 57 31 71 13 2 NS - - - - l6 Que 66 3 82 13 18 Que 63 3 84 13 30 Que 59 23 9o 25 31 Que 61 10 67 25 75 Table 16--Continued. c...- mw.nwr' Abomm Kalmm her and Height Trees with Height Trees with place of Age lemmas Age lemmas origin 4 shoots 4 shoots 1968 1968 1968 1968 3700 .912. percent. on. wet + 51 Que 7O 5 82 O ?8 Que 66 11 73 17 ?2 Que 64 24 84 17 24 1:11 69 34 96 75 33 Que 53 7 67 25 49 Que 61 3 91 13 43 Que 73 6 104 25 9 NB 64 8 94 o 8 NB 59 34 111 13 7 PEI 65 17. 78 17 11 NB 75 15 117 13 10 MB 62 21 112 13 12 NB 67 10 72 17 47 Que 61 13 108 25 25 NH 65 13 67 25 26 Mai 68 5 120 o 23 NY 73 23 114 O 27 Que 30 29 85 25 40 Ont 68 47 102 25 39 Out 63 48 93 33 41 Ont 73 63 91 58 42 Ont 7O 54 105 29 43 Ont 91 32 105 13 411 Ont 7O 33 100 0 46 Ont 75 23 75 25 L.S.D.(.O5) of a provenance mean 7.4 12.5 24.2 3.2 e value 14.9‘*’ 7.5m 3.7‘*‘ 1.7“ Percent of variance .536 to,sourc§. 58 39 58 27 7’ - significant at 1 percent level. mes _ significant at 0.1 percent level. 76 Table 17-Flowor production at the Nursery, Allegan and Kellogg summarized by source. Ont. num- Nursery Allegan. K 1 her and Trees Fem. fls. Trees Fem. fls. Trees Fem. fls. place of with for trees with for trees with for trees origin flowers with fls. female with fls. female with fls. male fem. flowers flowers 1967 1966 1966 1968 1968 1968 1968 3200 mm matte: 17.6mm: met) 1' 22:28:11 number 4. Sources west of 80° west longitude. 99 NWT 0 6 1 0 O O 0 98 NWT 0 2 1 8 3 O O 96 NWT 0 4 2 o o 13 2 92 Ses O 3 2 O O 13 1 91 Man O 9 2 O O O O ‘39 Sas 13 9 1 3 4 13 4 _5 Alb O 36 2 3 2 37 3 85 Out 0 9 2 O O 25 2 94 ,Alb 0 lo 3 O 0 0 0 90 Man O 31 2 5 3 13 l 87 See 6 25 2 O O 25 2 93 Alb 16 9 2 0 o 37 1 64 Ont 8 9 3 5 4 25 2 76 Ont 18 18 2 O 0 5O 3 77 Ont 38 15 2 8 2 13 3 88 See 6 20 2 5 3 25 2 86 Ont 27 19 2 3 1 13 1 5 Mic 4O 47 3 8 3 O O 83 Men 16 37 3 10 1 0 0 84 Ont 14 25 3 3 l 37 2 82 Ont 37 43 2 6 1 l3 1 81 Ont 37 20 2 5 3 25 2 63 Ont 12 43 3 0 o o o 62 Ont 33 46 4 5 5 63 .3 61 Ont 36 32 3 13 2 l3 1 60 Ont l9 5 2 15 3 25 3 58 Ont 9 36 3 o o 25 3 57 Ont 33 25 2 l 3 25 2 30 Min 18 46 2 18 3 13 3 79 Min 14 59 4 25 3 37 2 78 Min 13 34 2 3 l 13 8 70 Wis 48 40 5 21 3 37 2 55 Ont 4 5- 3 8 9 25 3 69 Wis 31 39 2 23 2 0 0 68 Wis 29 40 3 18 3 50 3 77 Table l7-Continued. Ont. num- Nursery Allegan Kellogg her and Trees Fem. fls. Trees Fem, f1s. Trees Fem. fls. place of with for trees with for trees with for trees origin flowers with fls. female with fls. female with fls. male fem. flowers flowers 1967 1966 1966 1968, 1968 21968 1968_ 3200 percent number percent .ngghgg percent number f 75 Mic 38 40 3 25 2 88 3 74 Mic 35 20 3 20 3 37 l 73 Mic 60 36 3 32 3 63 2 72 Mic 34 19 3 l3 3 37 4 56 Ont 55 36 2 lO 2 5O 3 65 Wis 25 18 2 O O 25 2 66 Wis 22 1o 2 o o o o 67 Wis 33 25 2 3 3 13 2 71 Mic 22 '44 3 8 3 13 1 54 Ont 66 41 3 33 2 l3 1 Sources east of 80° west longitude. 53 Que O 3 2 3 2 l3 1 5 NS 16 24 4 3 2 o o 4 NS 8 19 2 O 0 13 1 3R Que 38 13 2 6 3 SO 3 21 Que 14 17 2 O 0 0 0 20 Que 44 16 '1 12 2 50 2 19 Que 9 22 3 O O O 3 NS 19 30 4 - - - - 1 Mai 42 46 3 - - - - 15 Que 34 31 2 8 2 O 0 37 Que 46 28 2 0 O 13 1 32 Que 67 23 S 5 2 13 3 13 Mai 0 29 2 3 3 37 2 36 Que 46 23 2 6 9 25 2 35 Que 56 26 3 3 4 25 2 34 Que 46 47 4 3 2 37 2 5O Que 39 31 2 23 1 63 2 29 Que 15 21 3 5 2 O 0 14 NE 47 14 2 O O 13 2 6 NS 39 28 2 3 2 o o 2 NS 47 43 4 - - — - 16 Que 15 '30 1 11 6 13 4 l8 Que 39 27 3 3 3 25 3 3O Que 24 42 3 8 2 87 2 31 Que 39 2? 2 O 0 l3 2 78 Table 17--Continued. Ont. num- Nursery Allegan Keiloag her and Trees Fem. fls. Trees Fem. fls. Trees Fem. fls. clace of with for trees with for trees with for trees origin {lowers with fls. female with fls. female with fls. male fem. flowers flowers 196781968 1966 1963_2 1968 1968 1968 1200 percent number percent Ingggg; percent number + 51 Que 21 2h 2 12 2 11 1 28 Que 13 b0 b O O 25 2 22 Que 1b 10 2 8 2 l1 1 2b NH #6 28 2 5 1 25 3 2% Que 27 26 3 o o 25 2 mg Que 15 11 2 1 1 37 2 ha Que 38 27 2 10 u 17 1 9 NB 2? 37 1 18 1 25 2 8 NB b8 28 1 1 2 11 1 7 PEI 19 22 2 15 2 25 2 11 N3 67 32 2 15 1 17 2 10 TB 52 29 2 1 3 50 Ll 12 NB 51 no 1 1 1 O O h? Que 6h 31 3 6 2 17 1 25 NH 68 1+9 2 8 6 13 2 26 Mai h6 R1 3 1h 2 17 2 23 NY 37 29 h 11 3 17 2 27 Que 70 57 u 18 3 37 2 no Ont 63 51 2 7 1 l1 1 19 Ont 13 29 1 3 8 25 1 bl Ont 52 22 1 12 h 25 1 1&2 Ont 68 37 2 10 ‘3 17 1 h? Ont u5 10 1 22 1 25 2 no Que 1k 71 1 1 2 13 1 h6 Ont 68 52 h 9 5 37 2 Percent of sources with flowers 89 100 77 83 L.S.P.(.05) of a source mean 6.1 17.65 5.02 0.39 1.63 3.69 3.1a F value 3.9*.. 3.2... 3.1*¢. 2.1uttt 2.09Qt‘ 1.5u‘ 1.uufi Percent of variance source explains 43 15 39_ 11 10 21 18 ' - significant at 5 percent level. “‘ - significant at 0.1 percent level. 79 John." 0583 do an abdoaanwan a co. ma ...~mo.n Ho.a ammmwa.a mfimwea.we omneo~.wfi nae: .093 :5 53m 6: .323 :w ...Hmo.m nm.o Hmmmee.~ ooon~a.n~ aeaama.m ape: - :9: .3. 3:1 3: mm ...na~.m mo.~ momoa.:r Honnfim.eo Neomm.eoa coupon .uau non Sh $3 oaaohfins mm ...mmm.~ ma.o ~omame.o mmmmma.~ mmowae.a ”anon“ mm\oa conga -HponANaV am ...amo.m em.o mfioomm.o :amamw.m mQOOHu.H you 5 025 unoHHu» “on .9: mw\oH uncouoflaflv ma ..omuw.~ «N.H mnoeao.o nm:me.moa HammmN.HH «not» ma .39 .3” 33¢ .3585 mm ...emm.m om.o Noummm.o mmammm.H wemaum.m cease mo\0H Aav mm ...owa.o om.o nmmoom.~ onenec.a omamw.wa cease am\oH uoaoo Am. 8 2.33.2 84 8823 «$8.8m 92.128 .305 £3 E aw ossmo:.a m~.a wmnmm.~u HomaL.:Hoa Hwamm.mmw “no mm\~H Amy mm .sssma.e :o.n Neoao.um Hemme.mm~ nmmms.mm= .na m\H mm\m «Auden ”my 3.32 $29 : Jjggiajfiflm up coauaanuo a m ‘Iruudmmmiumar uoapumoa «cos 023.73» uncouom H Me 3.5 oven uoaoguno no uOHUmOm nogaocH .mmm .233 I new .8»: .m 353%.” .3 00.38 you aocoohm Mo macaw).— .vovuuoug .253" an! #3:) .uopaoo nourgom 3.5 on» an awe» oongobopm no“ oondapmp mo 035.:le Spun. 80 Job..." vacuum“... H.o as £33:an 43.3 28939 H an undefinafiau Job: 33.—on n «a 33:3»: 9 O. O. em tame; land 3390 mfimooé 033.0 A»? cook» 32 Auschwibénooom ma .mm:.a oam.a m:m:o.w Hadmw.m mommm.n condo» you .02 am nmmmé N36 «amid 93mm; amomaé 3.2. 30.5 .02 monoo uwohlono on :oaom.m .3de 303.0 $3de «033.0 condom you. .oh 9 tsogd 3N6 $53.0 «38.0 388.0 fit. «00.5 :5 women .80.»th mm .3856 so; $30.3 289:: 336: 38 S: inflow «mmmou «3%, N J «am a .mfl 383QO $03on anomI hp “Engage .m m z . m a: Iohznwoa mecupob accepom u no mo 93D hovosdno .mmH .393 I mm .3»: .H 3830?» .mm 8.38 no.“ 8300.5 mo moohmoa €3.30» .3 hoard we) 33:: no pogo: «3:53: .mwoaaom as 6353 can» oogobonm no.“ 0282c» no acmhmnddl. o." Oden. 81 .aoboa uncouom H.o an uncoamaqwan H .99 om ...mma.u omo.~ munan.~: maswm.o uoma~.mm nu“: moon» no onoouom nvyopwuhudcnooom Ha ...~aa.~ mmc.o mm~mo.o mamoa.o mmmoo.o avg: nook» no anoonom monoo nwomlono m ...nam.~ has.” mmmmm.om mmwwn.o ommom.m no“: mocha Mo “flockom 0H ...Hmo.~ Hum.o Hoamu.o ummom.a o:m~m.a coupon non nopnum monoo hd0h39)9 mm ...:mo.za mmm.a moamo.:H ammo~.mn aaoa.mmm anon 0H\H ugmdom Illnnnnquunauuuunluunauw . ca. ruumnu14QMMWIludmadaqum uoauoom anon hp voaadamuo h m I m «udqumlddaw abhflndoa condaucb uncouom .u Lu mo uflnpv uoaochuno .mmm Have» I mmm uouuo .m oamoaamou .~¢ coupe. you BovOOhm Mo nooumop .uoundouav uoada an: #03: no 90300» wogaog .3034 3 0093.3 amo» oogoboum you 023??» Mo 30331.10" 0.38. 82 ow 8 am m5 m4 m .3 fit. 893 $3 38% mo pcoouom mmeamn ww : an o.m m.v H.33 mononm mm\OH exam xooho mm 0 mm Wm m6 m6: :8 a“ cop» amoaawa you .02 mw\oH mkoono 3 mm 2 $2 3% do? 8.38 awn .oz mm\w .mfih oH¢th mm :« 0H AH mm Hca nu“: neon» mo uncouom ww\w .mHM cadaoh mm mm Na na m: wmm gag: moon» mo unconom nw\m .mam can: 3 3 3 S cm 0% 03.5 :33 .830 2 mm mm mm mm: $Nm. mafia H mg m 23 cm pepum IdedMMMMI wudamuu youum aadMMuMIIImndflmmMIlllldduallllIddudu Iwmpowudno manna: cmmzpom cane“: noo>pom Iandmmos Iwoa «waunum> ”mace Ha aquuuum lauddom awe: mo pans ovum .mmm neppo .mm occmpm saga“: .m acqum coospmp aocoouM mo moopwou I nmwmaa< p4 .mmm popuo .wm mucwum canvas .m mvnwpm noozaop Eouooum mo mooumou I nopnmo gondowom cone on» »< ohm) wunoaohdwroe Had .cmmmHH4 aw vansmwos mm: gown: .npaouw mmsewa nua3 momup Mo anachcn on» you pooomo hopnoo Aoudowmm make on» at ouws .mnofimou hp vmwc¢pna umop oonanobopm hem oocdfinmb Mo mmmzawn¢rIHN capae 83 Table 22.- Correlations between characters in nrovennnce study. Key to characters: 1 thru 3 were measured at place of seed collection 1. latitude . length of growing season . number of growing degree-days JN h thru 12 were measured at the Tree Research Center h. height at age 1. 5. height at age 2. 6. height at age 3. 7. autumn color at age 1. 8. number of male flowers for trees with flowers. 9. number of female flowers for trees with flowers. lO. nercent of trees with female flowers. ll. nercent of trees with buds set by September 22. 1966. 1?. number of crooks for the tallest tree in each nlot. 13 thru 16 were measured at Kellogg 1?. height ate age U. 1n. number of female flowers for trees with flowers. 1?. percent of trees with.fema1e flowers. l6. mercent of trees with lammas growth. 17 thru 20 were measured at Allegan 17. height at age u. 19. number of female flowers for trees with flowers. 19. percent of treeSIwith female flowers. 20. nercent of trees with lemmas growth. 84 o.H I mm. I I m. I I I I I I HH. NH. NN.I ma o.H I I I I I I I am. I I or. om. mr.I ea ma 3a ma wH HH ca 0 (O (x \O U \ .2? M (‘0 H .pmop cocacobonm cw mumporpmno awesome mcofiumHOppooII.mm nance 85 Table 23““Tests of interaction for height between all gun, Kelloqr an? the C " Tree 3 search ' .entcr in the {rovennnce study. Test between Allaxan and fiellozg. Degrees 0?? Sum of {can F Source of variance Freedom SIQarfi Square Value Plantations 1 f5f6.8 3586.5 Rep(plantations) JO h1h.0 h1.h Source )2 31135-3 339-5 Plantation X Sorrce 92 351.1 3.8 0.29? hot Significant. eror 920 3? 35.8 15.7 TLTnL 11197 Cjfl2h.0 Test between Allegan and Trie Research Center. Degrees of bum of been F Source of variance freedom Squares ggpare value Plantations l 779Pl.6 97351.6 Red; lanuticns) 12 11M: .1. 9S .7 Source )2 3Y7F1.b h23.7 kluntation x Source ?2 (191.9 75.9 S.O‘*' SignifiCant at 1 {arcent level. jrror llob l/f28.2 13.2 To?“ 1501 IZlVZl.L Test between Kellovg and Lurs~ry. Degrees Of Sum of Mean F Source of variance freedom ggnares Square value Plantations 1 1711908 1711? 08 Rep(p1antations) h 990.b 257.6 Source 92 32762oh 358.3 Plantation x Source 92 3381.8 36.8 1.8O**’Signifi0unt at . l {arc nt level. Error 368 7552.6 20.5 Tom. 55? 65007.0 86 Table zh-‘l‘est of int erection for percent of trees with fa;.:le flowers between Allegan, hellogz and t‘e Tree Recearch Center in the Test between allegan and Kellogg. DI‘CVCnfltT‘LC‘? 1.93:1. o Degrees of Sun. of lean F Source Of variance freedom annexes sluare value __ Plantations 1 1.2702 1.2702 Rep( plantations) 10 1.0002 .10002 Source 92 6.27h5 .06820 Plantation x Source 92 O.h757 .0051? 1.759%* Significant at 1 percent level. Error 920 27.0618 .0029b TUTAI 1113 35.082H Test between Allegan and Tree Re: earch Center. Degrees of Sum of Mean r Source of variance Heedom Squares Square value FlEintCtjons l llo’JéeO 1101462 Rep(plantations) l2 1.30h5 .10870 Source 92 9.71h0 ~10559 Elantation x Source 92 3.7613 .00088 1.337” Significant at 5 percent level. Error 1102; 33 . 75 13 .0305 7 Test between Kellogg and Nursery. Degrees of Sum of Mean 1" Source of variance freedom squares square _“_va1ue Plantations 1 3.14968 3.11968 Rep(planta tions) 14 0.3529 .08823 Source 92 3.15h5 .0382? Plantation x Source 92 5.8011 .06306 2.6204W Significant at. 1 percent level. Error 368 8 .8 1.1111 . 02 1403 __ TLTAI. 597 21.6h9h 87 Table 25 -Test of interaction for percent of trees with lammas growth between Allegan and Kellogg in the provenance test. Degrees of Turn of fl 8" earl 1‘ Source of variance freedom squares Square value Plantations 1 0.0011; .0011: Rep(plantations) 10 0.56115 .0565 Source 91 3502077 03869 Plantation x Source 91 6.14256 .0706 1.14014.” Significant at l perCent IEVGlo Error 910 15 . 8168 .05 03 It'll-3L 1103 88.0760 88 m «H 0H 00 N0 mw.:m n:.:: mom : 0N HH ea :H Noa 00 mr.:m 0:.:: 000 m 0H Ha :a HH :ca N0 00.:0 m:.:: mmo m «H 0 0 ma :0H 00 mm.:0 N:.:: 000 m NH 0 NH 0H mo 00 mN.:m m:.:: 00.0 m 0a 0 0H NH Hca 00H 0N.:m Nm.:: 000 m ma Ha 0N m 00 00 0:.:m 00.:: .00 : :H m «a 0N 00 00 0:. mm.m: mao 0 «H m «a. 0N 00H 00 m:.:0 ma.m: 000 0H NH m NH 0H :0H 0NH NN.m0 Nn.:: Hem : HH ea .N ea mca mm :0.:w mm.:: who : 0H : N 0H 50 00 em.:0 0N.:: who 0 0 : 0H ma m0 m0 N0.:0 mm.:: 6.5 m 0 0 5H 0N Noa mm N0.:m mm.:: who 0 N 0 0H Ha 00H 0NH nm.mm m:.:: moH : 0 0H mN ma moa 0Na m:.mm N:.:: 00H : m 5 NH 0N no 00 ma.:m mm.:: 000 m : HH NH :H 00H 00 N.:m mm.:: 000 m r ma mN ea 00 00 Nm.:m 00.:: on0 : N N HH ma Noa 00 N.:m 0n.:: 000 :H H axon Sancho mum” a moon» 0: pop unmouom, mmmmmmm. mo uncouom .mm. twee Imou mmofl meoa Nome, mwmau OGSh .poo mesa macron mpoonm whee Advaam«uo weave mposoam oflneom mecewa .paom ompm .wcon .ucq Mo nfi noon» cacaom nu“: moons sue: mocha pnmaom amoufi p003 ganoz nuance Mo .02 eccpm .nfiwauo mo vnwpm an woufiaxEESm .pwou mcowcac pawlm\a sewage“; :300 wwoa nah were Aviohm 0pc nwmprIlmN capes 89 m 0 NH mm 00H m0.mm 0N.m: Hm m 0: m 0H 0H 00 0NH NN.:0 Nm.m: Hm m we 0H :H on 00 mNH 00.:0 mN.m: Hm m me 0 mH :H 00 0NH mm.mm 0N.m: Hm 0 N0 HH mH 0H mo omH 00.:0 0N.:: Hm : 0p m 0H «H m mNH Nm.:m m:.:: 0:0 0 we 0 «H :H H0H 0NH m:.:m mm.:: 0:0 m :r m 0H 0N N0 00H main 3. .. gem N 9. 0 m nH m0 MNH mm.:m mm.:: 0:0 N Nm m 0H HH 00H 00H om.mw mm.:: 000 m H0 0 0H :H 00H mHH mm.mm m:.:: mom : 00 0 0H 0N 00H 00H mm.mm Hm.:: 00 m 0N N N 0 m0H 00H mm.mm NN.:: 90 m 0N m NN NN :0H 00H 00.:m Nm.:: mom m NN m :H HN 00H NoH mH.mm Nn.:: 0H: m 0N 0 0H mH 00 00H 0:.:m m:.:: mom 0 mN 0 0 HH 00H Hm 00.:0 mm.:: 000 N :N m HN 5H 00H 00H :0.:m 0H.:: mom m mN N NH :H 00 00 m0.:m n:.:: mom : NN m HH 0N 00 00 m:.:w NN.:: mom m HN H .H b much a «dump-nonlddd fledged pmoouom mo €00.39 .mdu twee Imod 000:, 0004 Nme, 71000H mafia. .000 0:95 mu030Hm upoozm mane AnvnfiwHuo 0:000 muwonM onEom messed .umom menu .chA .qu Mo nH woman oawaoh HHS) 0095. £33 0029 uanom $0.5 awe; H382 50300 Mo .02 dmflm .eosereoosuwN oHpee 90 .OHmH odvmchm I Hm .panm I Bea .n0Mhopono n one .ofiucwm n mom .owumbcpa wanna n aw .ooxncmmwz n max .coesoomom H mom .>¢Eomo n omo .owmmuo u mpo .nxmmmeM H Haw .0pom3duo u «no .oomoH I 00H .mwocmo omo ”mchHom mm.cna moeed nuance Adv .HopoH 0000000 H ex pawoamHemHm n .. mwm.o 0mm.c vano.N omm.o Qusp h mam .3393. moon» 0: non «madman. mummmmm mo unooamm mwma mmmH mmma wmma 0:00 .000 0:50 muoSOHm mucosa whosoam onBom mmaSwH .pgom oamaom nu“: mocha £0H3 mocha uswwom .Umficfiunoollmm manna 91 N 0 N0 00 8.:0 00.3 80 N mH 0H m 50H 00 mm.:m nm.dd 000 HH :H NH 0 00 N0 NN.:m N:.:: 80 N. 0H NH 0 :0 mm. 3.:0 0:5: 80 0H NH 0H 0 00H 00H 0:.mm N03: .8 0 HH 0H m mo mNH md.mm mw.:: so HH cH NH 0 N.0H mNH 3.3 $.:: 00 0 0 :H N. HNH 00H 00.? 00.? 03 m m 0H 0 :HH 00H 3.3 3.? 0:3 H N. mH 0 00H 00H 25% 9.3 03 0 0 HH m mo no HH.:m no.4: no: 0H m 0 N. N0 m0 8.:0 00$: 82 N. : :H m :m. no Nm.:m 3.3 no: N. m N. 0 N0 3 mN.:m 00A: 82 mH N : : mNH mNH $.00 NH.m: 0H4 : H nmmmlddmmumq mud”. «Maw Mummmum. udmuu. mo uncouom .aq. luau awed wwoa Iwme lwwma .900 mucosa wand AoVnHmHuo uc¢¢m udEEdH .poo .poo 00AM .maon .umq mo as moon» an3 mocha NOHoo vnwwom umouh peek apnoz Monaco no uopEdz ddnpm .nHwHuo Mo agape hp woNprEEdm .pmmp anemone pHmIN\H nwanon £300 NQOH how open nuroum 0:0 QHMHAOIIAN oHpna 92 .000000 u 000 .noaaoowom u mom .0uomzcno u duo .omompo u 000 .omuobwua wanna u an .3eEomo u omo .oqu H #09 .honouoEunoz N no: .mnoRH<.u 9H4 umonHom mm and nosed menace .aoouw CHIN .opmHvosaoch mI: .oHpaum mIc "mouwhm uoHou Hay .HoboH vaoouom H 00 pncOHchme I 00 aammN.N NNo.H mmw.H osHmb M NH 0 m0 00H mN.:m NN.:: mom HH HN 0H 0 N0 00 0:.:m mm.:: .00 NH ON 0 N mm mm Nm.:m Nm.:: 0.5 HH 0H NH N :0 mm 00.:0 mm.:: 0N0 0 0H 0 m :0 00 Nm.:m Nm.:: 000 m NH 0 N 00 00 m:.:m NH.m: 0:0 0H 0H H105 :8 >0 mil-mop moon Aduupmm mmwmu. mo unmouon .mm. Imod Inca 1w00H mon mmmH ofloo 000030 0380 HaVnHmHuo dunno mweema .000 .900 can» .wqoq .009 Mo qH moon» spHx mocha uoHoo panom awoam paw: xenon mpaseo no nopauz wanna .e000H00001INN oHpae 93 .Hm>oH 00.0000 H.0 00 unaoHNHanm .HoboH 080qu 0H 00 and Shaw? Iii : .mNH.H «00.0 HmH:NN0.0 000HN00.0 00mHN«0.0 :NH3 .0000 Mo annouom mm\oa mvoogm defidA 0H ...:mm.H :mN.H NmmmNNN.0 HNmH0.H0H :00000.HH oeaac 00N0H 00H00 Ne oecdor.r mwm.m HHHNHr.mr #:Homc.@m NN3NC.OHH wonocH mm\oa pgwwom uummau «flame : 000m H paw vwwunmmwm .7HdAdmow wmmmv. duhd aoaumwwnw. mp cocHnHmko h m: 000m K me oamddm dam: Ioudmmoa Imeoe ooanumb unoopom H.b mo pHQD open .HmN H0000 I mam hohuo .m opmoHHQmu .NmH oonScm you souooum no moohwoq .ummp unawaun pHmIN\H cmmHnon grow Nwoa you concHuwb mo mothnnfilI.QN wanna 94 .H0>0H ocoopoa H.0 00 paeonHanm .H0P0H uaoohop H 00 uqonmapwHo .HoboH 0000009 w 00 pqaoamHanm I}. {l N onm.H maH.o Hmmmeo.o ocoH:or.o :NomNHH.o 000» hog 909652 mw\w 00000 awohloqo 0 .0:N.H 000.0 NN00:00.0 ermNCH.0 m000m:0.0 00H: 00000 mo ucoouom mw\w mocoo u0mhlono N .00H0.H N00.0 0MH0000.0 N00N000.0 meom000.0 00H: 00000 M0 anoouom nw\w apoosm 005509 NH .00mN0.H 0mH.0 N000N.HmH :N0:N.00H 0mHmc.mNN 00:00H mw\m angom NN ...m:H.N mwm.N 00:n00.Nm 0:0NN0.HH 0000Nm.0N 00:00H N0N0 0:0H0m .300 .01ng : 000m finglflflddjgg hp cocHoagHo h WMIumMWIHEun .uumnmm a00x Ioudmmos I003 ooquum> oaooucm u M0 png 0009 \! .mON H000» I don uoauo .r 0000HHuoa .moH 000360 hog aouoopm mo 0009M0n .0000 mcowoun pHmIm\H awagon nsom wmma uom 00n0H00> mo m0thnn¢II.mN ompcp 95 Table 30,-0An81yses of variance for between- and within-stand variances for Michigan 1/2-sib progeny tests. T s so 66. Height age 3. .. Degrees Sum of Mean F S rc v r a ed m a .re value Between stands 39 10,683.55 273.9h 0.990 Within stands 159 “3,980.60 276.61 Between progenies 198 5h,664.15 Trees with laggas growth ass 3. Degrees Sum of Mean F Sggrgg of variance freedom squares _gggagg_ valug nut Between stands 39 2.63 0.?70 2.061 Within stands 159 5.21 0.131 SignifiCant at Between progenies 198 7.8M 1 percent level. Trees Wilh femalg {lowers age 3.. Degrees Sum of Mean F hWLms—W e m s ...snas re 13112.19. Between stands 39 1.h68 0.0377 0.835 Within stands 159 7.166 0.0851 Between nroeenie s 198 8.6314 Number of female flowers Der source egg 3, Degrees Sum of Mean §ggxg§ g: 33:18 neg frggggg ggggres ggm rg Iglngw Between stands 39 4.1331 0.1060 0.928 Within stands 159 18.1572 0.1192 Between nrogenies 198 22.2903 96 Table 30.-Continued. TQSt 39:22 in m. 39‘ a 2. Degrees Sum of Mean F 593:9: g: Igzigngg freedom_ squares sguare value Between stands 20 3,369.1 168.96 1.638 Within stands 162 16,66h.5 102.87 Between nrogenies 182 20.033.6 Trees with lemma growthegQ 2. - -. Degrees Sum of Mean F Sgnzcg of variance {reedog squares sgggrg valgg Between stands 20 0.686u 0.03u3 1.076 Within stands 162 5.1651 0.03.9 Between progenies 182 5.8515 Autumn color age 21 ___, __ .__ Degrees Sum of Mean F Source of variance_ freedom. gnuares ‘, gggggg_ value Between stands 20 516.69 26.835 2.719" Within stands 162 1587.07 9.799 Significant at Between nrogenies 182 2123.76 1 oercent level. 97 Table 31 p-Correlet ions for l'ichigan l/2-sib progeny collections. Ker to characters: 1966 sown collections . l and 2 were measwr d at the place of sesfi cellection. J. latitude. 2. length of :rowi ng 5 3:15 on. 3 thru 6 were measured at the Tree Research Center. 30 heizht 3t 82': 30 h. percent 0f trees with lemmas growth. 5. percent of trees with female flowers. 6. number of fa'r'ale flowers for trees with flowers. 1967 5 mm collections . l and 2 were measured at th place of Seed collection. 1. latitude. 2. length of grm-ing season. 3 thru 5 were measured at the Tree Research Center. 3. height at age 2. L. percent of trees With lammas growth. 5. autumn color at age 2. 98 Table 3).-~09rrg1ations figr Michigan lzg-sih nrogenx collections. 1966 sown collections. 1 .lsflfl 2 "" .l‘QQ 3 -.95 -.15 .1192 it .36 .28 -.08 1.9.9 5 -.11 -.17 .21 ---- 1.9.9 6 '109 '17.} .23 ---"° “- 1-00 1967 sown collection. 1 1.9.9. 7 “"' .llgfl 3 -.10 .50 .1129 h -.6M -.06 --- .ngg 5 -.51, ~11: --- --- 1.00 l 2 '3 1+ 5 99 Table 32.--Data on budworm resistance study. summarized by source. County of Egg Needle MSFG originLa) masses Heighte__ Weight Width (1) (2) _(3) (8) #000 £2812; lashes .ng. .nn. + 206 Osc 3.9 17.1 u.7 11.67 221 Osc u.7 17.1 3.3 11.68 238 Ios 9.0 15.9 3.0 11.69 290 Cra 3.8 14.7 9.0 11.69 251 Cra 2.6 19.6 3.2 11.52 257 Cra 3.0 16.6 3.2 11.58 27h Ots 2.6 16.1 3.3 11.68 280 Ots h.“ 17.5 h.h 11.61 288 Ots 3.6 16.8 3.9 11.57 299 Oge 2.6 15.9 3.0 ll.b9 302 Oge 3.8 17.1 u.0 11.76 310 Oge h.8 17.7 9.0 11.38 327 Ros 2.9 16.8 3.9 11.66 338 R08 “.9 16.7 3.4 11.57 357 Ron “.6 16.6 4.0 11.59 36h Ros 3.0 17.9 3.9 11.68 376 GT 1.7 16.2 2.9 11.60 379 Wex 3.5 16.5 3.6 11.68 380 Ben 3.0 17.3 3.0 11.67 38h Che 2.8 13.“ 2.h 11.h2 388 Emm 3.3 16.5 3.“ 11.63 391 Emm 2.7 15.5 3.1 11.66 #05 Che 3.3 15.6 2.9 11.67 807 PI 3.7 16.6 3.0 11.68 #27 PI 3.7 15.0 3.2 11.63 (a) County names are as follows: Osc Oscoda. Ios = Iosco. Cra 2 Crawford. Gt: 2 Otsego, Oge Ogemav. Ros 2 Roscommon, GT : Grand Traverse, We: 3 Wexford. Ben Z Benzie, Che = Cheboygan, Emm = Emmet. PI 2 Presque Isle. 100 Table 33.-qAnalysis of variance for number of egg masses per source in budworm resistance study. Degrees F Source 9; Eggiation freedom Sum of squares Mean ggugrg value Between sources 2“ 166.5000 6.938 1.007NS Replication 11 391.826? 35.621 Rene x Seedlot 26h 1819.3400 6.891 Total 299 2377.666? Table 34.-—Correlations in budworm resistance study. Key to characters: 1. Haighte 2. Volume of needles on tree. 3. Number of egg masses. h. Needle width. 7 -- 1.00 HICHIGRN STRTE UNIV. LIBRQRIES