GENmC “MW-{OE 1‘13 5311113 fiO‘lSWRE CGNTEN‘F m RATE OF MOISTURE LOSS MORE “YEMEN RACES- OF DOUGLAS-FIR The“: for ”is Dayna of M. S. MICHIGAN STA“. UNIVERSE“ Danald H. DeHayes i974 IIIIIIIIIIIZIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 0525 7160 ABSTRACT GENETIC VARIATION IN FOLIAR MOISTURE CONTENT AND RATES OF MOISTURE LOSS AMONG INTERIOR RACES OF DOUGLAS-FIR BY Donald DeHayes Douglas-fir is a valuable Christmas tree and ornamental throughout much of western and northern United States. Through the use of provenance tests, the genetic variability of the species has been determined. These tests have yielded useful information as to the fastest growing and most suitable races for specific planting purposes. My objectives were to deter- mine the genetic variability in foliar moisture content and foliage drying rates among interior races of Douglas—fir and to learn the evolutionary significance of such variation. Seeds were collected in 1961 from various parts of the species' natural range and were sown in Michigan State University's experimental nursery. Four permanent test plantations containing 68 seedlots were established between 1965 and 1967 in Michigan. During the summer of 1973, I collected current-growing and year-old foliage from trees growing at the 4 provenance plantations in Michigan. I determined foliar dry content for each 4-tree plot. Analyses of variance were performed to determine the significance of differences in dry matter contents. - Donald DeHayes In a supplementary study, I compared rates of moisture loss from Arizona and Inland Empire seedlots. Branches were removed from 2 seedlots of these races and were dried under normal room conditions for 27 days. Moisture loss was deter~ mined by periodically weighing the branches. The significance of differences were tested by an analysis of variance and t— test. The largest differences in moisture content were found between current—growing and year—old foliage. The year-old foliage had about 10% higher dry matter content. Although current and year—old foliage differed in moisture content, the relative trends among the races were similar. One seedlot from slightly east of the White Pass in central Washington had the lowest foliar dry matter content, 29.6% and 38.7% for current and year—old foliage reSpectively. This seedlot probably represents an area of overlap between the coastal and interior Douglas—fir varieties. It owes its high moisture content to its affiliation with the coastal variety. The Arizona Douglas—fir seedlots had the next lowest dry matter content. They averaged 32.2% and 42.5% dry matter content for current growing and year-old foliage respectively. Trees with the highest dry matter content were from central Washington and the Inland Empire. They averaged 3-4% less moisture than the Arizona trees. A study in Oregon also showed that Arizona seedlots had the highest moisture content of interior Douglas-fir seedlots. -Donald DeHayes The Arizona trees also lost moisture at a significantly slower rate than the Inland Empire. The rates and differences in moisture loss were greatest during the first 2 days of drying. At this time, Arizona branches had lost 17% of their fresh weight, while the Inland Empire branches lost 30% of their fresh weight. Differences in rate of drying were clearly reflected in foliage appearance. Inland Empire branches turned brown more rapidly. Differences in drying rates may be related to waxes which coat the needles of Arizona Douglas-fir trees. My results indicate that the waxy surface of the tree leaves may be of selective value in retarding moisture loss from the Arizona trees. Natural selection may also have played a role in promoting racial differences in foliar moisture content. Selection has probably resulted in two modes of adaptation. In the hot dry southern Rockies, selection has resulted in Arizona trees with high moisture content as well as slow rates of moisture loss. In the colder northern Rockies, selection for high dry matter content may be an adaptation to colder temperatures at place of origin. GENETIC VARIATION IN FOLIAR MOISTURE CONTENT AND RATE OF MOISTURE LOSS AMONG INTERIOR RACES OF DOUGLAS~FIR BY Donald H. DeHayes A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Forestry 1974 \A‘} {9 ACKNOWLEDGEMENTS I wish to extend my sincere gratitude to my major professor Dr. J. W. Wright, whose guidance and assistance were essential to the successful completion of this study. I also wish to thank my colleagues, especially Kim Steiner and Nicholas Wheeler, for their suggestions and constructive criticism. Finally, I am indebted to my wife, Annmarie, and my parents for their patience and encouragement throughout my entire academic career. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES CHAPTER 1. INTRODUCTION PHD OBJECTIVES Objectives 2. SPECIES DESCRIPTION 3. REVIEW OF PAST LITERATURE Genetic Variation in dry matter content of Scotch Pine Variation in Dry Matter Content Within Species Variation Diurnal Variation Seasonal Variation Genetic Variation in Douglas—fir 4. MATERIALS AND METHODS Seed Procurement Nursery Practice Outplanting Procedure Sampling Technique Adequacy of Techniques Rate of Moisture Loss Growth Rate and Foliage Color iii 10 10 I6 21 25 29 31 32 CHAPTER Analysis of Variance and Correlations 5. RESULTS Geographic Trends in Dry Matter Content Current Years Foliage Year—old Foliage Dry Matter Content of Current vs. Year—old Foliage Genetic Variation in Rates of Moisture Loss Effect of Moisture Loss on Foliage Appearance Variation in Percent Dry Weight Due to Time of Collection 5. EVOLUTIONARY CONSIDERATIONS Isolation Douglas-fir Rangegaps Natural Selection Selection and Moisture Content Drought Resistance Cold Resistance 7. PRACTICAL IMPORTANCE AND APPLICATION Importance Application LITERATURE CITED APPENDIX iv Page 33 36 36 51 58 59 60 60 64 75 75 78 81 86 TABLE 1. 1A. 2A. LIST OF TABLES Page Dry matter contents (% fresh weight) of Scotch Pine foliage from different parts of its natural range (Langlet, 1936). . 12 Seasonal variation in foliar sugar and dry matter content for northern and southern seed- lots of Scotch pine. l4 Differences among geographic races of Douglas- fir in the dry matter content of fresh foliage. 38 Racial differences in susceptibility to frost damage and time of leafing out of Douglas-fir growing on plantation MSFGP 16-65/67 (Kellogg Forest) in Michigan. 39 Analysis of variance for dry matter content data of current-years foliage of Douglas-fir at plantations MSFGP 16-65/67 (Kellogg Forest) and MSFGP 5-65 (Camp Kett). 41 Analysis of variance for rate-of-drying data for ARINEM vs. INEMP Douglas-fir races growing in Michigan. Data analyzed after 2 days of drying. 54 Racial differences in height, foliage color, and frost damage in Douglas-fir growing at plantations MSFGP l6-65/67 (Kellogg Forest) and 5-66 (Camp Kett) in Michigan. Height data for trees 12 years old from seed. 79 Analysis of variance of current—years and year-old foliage moisture content data from plantations MSFGP 16-65/67 (Kellogg Forest), MSFGP 5-65 (Camp Kett), MSFGP 9-66 (Kellogg Forest) and MSFGP 4-67 (East Lansing). Levels of significance indicated by asterisks. 92 Analysis of variance of height and foliage color data from plantation MSFGP 16-65/67 (KellOgg Forest). 94 TABLE 3A. 4A. 5A. 6A. Analysis of variance showing significance of year of planting at the Kellogg Forest (MSFGP 16-65/67) on current-growing and year-old foliage moisture content data and on height and foliage color. Differences in year of planting significantly affected height growth. Analysis of variance showing the effect of time Of collection (within and between replicates) on current-growing and year—old foliage moisture content at plantation MSFGP 16-65/67 (Kellogg Forest). Error mean squares for moisture content data from Douglas-fir races at the Kellogg Forest plantation (MSFGP 16-65/67) (both current- years and year-old foliage). High error mean squares for NOCOL and SOCOL (current-years foliage) are due to variability caused by frost damage. Climatic variation throughout the Douglas—fir range. Various climate regimes for the environments occupied by each interior Douglas~ fir race are given. vi Page 95 96 97 98 FIGURE 1. LIST OF FIGURES Natural range of Douglas-fir and location of stands from which seed was collected and grown. Douglas—fir races as recognized by Wright 33.21. (1970) are illustrated. Variation in foliar dry weight percent of Scotch pine in relation to temperature at place of origin. Trees grown from seed collected in regions with few warm days generally had higher dry weight percents (Langlet, 1936). Differences in current-years foliage dry matter content between and within races of Douglas-fir at Kellogg Forest (plantation MSFGP 16-65/67) in southern Michigan. Differences in current—years foliage dry matter content between and within races of Douglas-fir growing at Camp Kett (plantation MSFGP 5-65) in southern Michigan. Differences in year—old foliage dry matter content between and within races of Douglas-fir growing at the Kellogg Forest (plantation MSFGP 16-65/67) in southern Michigan. Dots represent. Differences in rates of moisture loss of cut branches of ARINEM and INEMP Douglas-fir races growing at Kellogg Forest (plantation MSFGP 16-65/67) in southern Michigan. A schematic illustration of Levitt's sulfhydryl- disulfide hypothesis of frost injury and resistance in plants. vii Page 8 ll 44 45 47 52 87 CHAPTER 1 INTRODUCTION AND OBJECTIVES Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) is one of the most widely distributed conifers in western United States and has great potential for planting. It can be planted successfully in large portions of western and northern United States. The extreme variation in foliage color and growth rate, obtainable from several geographical seed sources, allows the nurseryman, Christmas tree grower, landscape designer, or homeowner to select the apprOpriate type for his personal preferences or intended use. Douglas-fir is recognized as one of the most important timber trees in the world. It comprises about 50% of the standing timber of western forests, and more than 25% of the standing timber of the entire nation. In sawtimber volume there is more Douglas-fir standing today than in any other species. Of the 51 billion board feet of softwood cut for the market in 1973 more than eight billion was Douglas—fir (Wharton, 1974). Douglas—fir is also an excellent Christmas tree and in western United States is the most commonly used tree for that purpose. Recently, geographic origin studies on interior Douglas-fir have yielded useful information on its Christmas tree possibilities in the East. The tree's soft, graceful foliage, its excellent needlemholding capacity, and its natural pyramidal habit are important factors contributing to its large potential in the Christmas tree market. Variation within a Species creates Opportunities for locating better types, and raises problems in avoiding un- desirable types. Foresters have found that it may be just as detrimental to use the wrong race of a species as to use the wrong Species. The term provenance, as used in forestry, refers to the particular place where trees are naturally growing or to the place of origin of seeds or trees. Forest geneticists, by testing seed of different provenances, have been able to study genetic variability within species. The results of such tests have been the identification of genet- ically distinct types which are suited specifically to various environmental situations. Further study of specific proven- ances within the distinct types will, hopefully, furnish the evidence necessary to identify the most genetically superior seed sources. Forest geneticists have recently established provenance tests to study the genetic variability among interior origins of Douglas—fir. These tests have largely been responsible for the identification of fast growing and hardy ecotypes, thus making it a feasible Christmas tree and ornamental possibility in much of the nation. The information obtained up until this point has been derived from genetic studies concentrated on growth rate, foliage color, and frost damage. These results have been of tremendous benefit to the growers of southern Michigan; they have already been able to cut the rotation of Christmas trees in half. However, as more in- formation on growth rate is gathered, it has become apparent that data on other tree characteristics were needed. A sufficient supply of water is one of the most important factors enabling a plant to carry out its life processes and more than 70% of the protoplasm of actively functioning plant cells is made up of water. A severe lack of water may result in death or extreme reduction in protoplasmic activity. In addition, the cell sap and cell walls of most living cells contain large amounts of water. In brief, water functions in a plant as a: (l) raw material for food manufacture; (2) solvent of gases, mineral nutrients, and food; (3) medium of tran5portation of raw materials and food; (4) a medium for maintaining turgor in living cells. OBJECTIVES The primary objectives of this study were to determine the amount of genetic variability in foliar moisture content of Douglas—fir and to learn the evolutionary significance of such differences. A secondary objective was to learn differences in drying rate of cut branches from two geogra- phically and morphologically extreme Douglas-fir races. As in most genetic studies, the ultimate goal was to learn the practical significance of existing differences. The effects of foliar moisture content on moisture loss, foliage appearance, and fire susceptibility are practical considerations important to Christmas tree growers and the public alike. CHAPTER 2 SPECIES DESCRIPTION "Its somber shape, its serrated crowns and sharp lancepoint tips and long swaying boughs become printed like a lasting eidolon on all our memories of the Pacific Northwest. And even in the desert states of the Southwest we meet it again, on high peaks, with gratitude for its dim, cool groves, after the glare and heat of the rocky wastes below... With experience one comes to recognize a Douglas-tree in the field from almost as far as it can be seen... The dense, compact crowns, the lusterless, dark-blue green of the foliage; the darkly, deeply furrowed old holes, the mast—like stems, and the grand down sweeping of the boughs, all go to make the character of this Species. But one feature there is which is peculiarly distinctive, and that is the way that numberless long slender twigs clothed on a spiral of needles, hang vertically from the branches. Though the general habit of the tree is not what gardeners call weeping, these long pendants have a sort of sorrowful grace. When the summer winds blow lightly through the forest they stir this shawl-like fringe in an idle, ferny way; when winter rains come driving through the forest, level and endless from the storm bound Pacific, then these long pennants lie out waving upon the gale in a way that gives the whole tree a wild and streaming look." A description of Douglas-fir by Donald Culross Peattie (1956). Douglas—fir was first discovered on the Pacific Coast by Archibald Menzies in 1792. It received its common name in the honor of the Scotch botanist, David Douglas, who first sent seed to England. After being classified as a Pings, Abies, Taxus, $3333! and Pigea, it was finally separated into a new genus in 1867. It was given the scientific name Pseudotsuga meaning "false hemlock". The genus is a member of the Pinaceae, the largest family of conifers. Six Species are recognized:: four are native to various regions of China, Formosa, and Japan and the other two are native to Western North America. Of these six, Douglas-fir is the most familiar and economically important. Douglas-fir is mighty in stature. It towers up to heights as great as 400 ft and may reach 17 ft in diameter. Except for the Sequoias, it is the tallest tree in North America. It is generally a long lived species as well. Ages in excess of 500 years are common and many have exceeded 1000 years. Douglas-fir typically grow into tall trees, having thick, deeply furrowed, dark brown bark. Buds are long and narrow, sharply pointed, and reddish-brown in color. The needle-like leaves are one inch long and are flattened and flexible with short petioles. The leaves range in color from yellow-green to bluish-green and are persistent up to 8 years. Male strobili are borne along the twigs, while ovulate strobili are terminal on short twigs. Cones are 3-5 inches long and easily identified by the distinctive three-pointed bracts protruding beyond the rounded cone scales. The narrowly shaped cones hang down from the tree, mature in one season, and persist on the tree into the following year. Seeds are long-winged and light reddish—brown in color and are often produced as early as 10 years. The number of cotyledons varies from six to twelve. Douglas-fir consists of two widely recognized forms. The coastal form grows on the islands and mainland of the Pacific Coast, and the interior form is distributed through- out the Rocky Mountains. The north-south range of Douglas- fir extends over 3,000 miles from the southern half of British Columbia, through the Pacific Northwest and Rocky Mountains, and well into the higher elevations of Mexico (Figure 1). Its east—west range extends from eastern Montana and Colorado to the Pacific Coast. Typically, Douglas—fir is found in mild humid regions with annual temperatures ranging from 400 to 65°F with extremes of --300 to llOoF. Annual precipitations varies from 15 to 100 inches but the summers are usually dry. Douglas-fir have their best deve10pments on soils of Figure 1. Natural range of Douglas-fir (shaded area) and location of stands (dots) from which seed was collected and grown. Douglas—fir races as reCOgnized by Wright at 31. (1970) are illu- strated. l. B . u a? N. l .. ..\\ .Q \ .\\\\ «xx \ \ ss\\ RV\\..\\“..\\.\5\\<\\ \\\\\m x \..\\\\C\ \\... x. \\\ \ \\ I‘ . \\\\\\\\\\M\\\\\\\\\\\\L\\\\\\.\\\\ \ \ \\\.\ \\\\ P \\\\. .\ . \\ . \\\.C\\\\\x\. \ a. \..\“\\ o I . \. \ $§\\\\\\. \ \\ .. Sees Lie. \\\\m\\\\\r . . \\\\\\ .e. \ .s § I’m \- \1 1‘ \s! fol. KICK \s \x NOROC . I 'I' I, 'I’ I, s & \ m . \ _ a .\ s . \ ‘ .\ \. \. \\\\. \\ W...\ I 3:. \\\\\\ .\ \\\\\\\\\\\\\\\\\\\x\\\\x\\ . 1. sedimentary and volcanic origin. They will not thrive on poorly drained soils or soils with an impervious layer near the surface. It grows best on soils with pH 5 to 5.5. The tree is generally rated as a prolific-seeder, and on good sites, a large number of the viable seed will germinate. However, about 75% of the seedlings usually die during the first two years because of heat injury, drought, competition, frost, insect, root rot, dampening off, and rodents. Fire has played a varied role in the regeneration of Douglas-fir. In the coastal varieties, the Species wide- spread occurrence in even—aged stands is largely attributed to stand destruction by fire, clear cutting, or insect attack. However, it is believed, that if protected from these disturbances, Douglas—fir would gradually be replaced by the more tolerant hemlock, cedar, and true firs. On the other hand, since the interior trees regenerate best on litter and moist soil, soils blackened by fire are unfavor- able because of high surface temperatures. The unfavor- ability of the burned over soil is believed to have caused the replacement by lodgepole pine of all but 5% of the original Douglas-fir in Colorado and Wyoming. CHAPTER 3 REVIEW OF PAST LITERATURE GENETIC VARIATION IN DRY MATTER CONTENT OF SCOTCH PINE: A REVIEW OF LANGLET (1936). The first and most extensive study on the genetic variation in foliar dry matter content within a species was undertaken by Langlet (1936). Langlet spent four years (1929-1933) determining dry weights for 582 Swedish seedlots of Scotch pine (Pinus sylvestris), and for several other seedlots obtained throughout the natural range of the species. Langlet's work deserves special recognition because of its relevancy to my Douglas-fir study, and because of the physio- logical implications which his data suggest. Langlet's data, in general, suggest a strong latitudinal trend in foliar dry matter content. Trees from northern Sweden approached 40% dry weight, while southern Swedish seedlots ranged from 31-32% dry weight (Figure 2). Northern regions with few warm days provided trees with the highest dry weights, i.e. the lowest moisture contents. Data from the rangemwide study follow the same patterns. Seedlots from northern parts of Norway and Finland had 10-12% higher dry weight percents than seedlots from Italy and Spain (Table l). Figure 2. 11 Variation in foliar dry weight percent of Scotch pine in relation to temperature at place of origin. Number of 60 (C) days is defined as the number of days per year with mean temperatures greater than 60 (C). Trees grown from seed collected in regions with few warm days gener- ally had higher dry weight percents. (Langlet, 1936). °/o DRY WEIGHT 39 37' o o 35’ o o o o o o 33' o J l l J l l L 100 120 I40 160 ISO NUMBER OF 5°(C1 DAYS 12 Table 1. Dry matter contents (% fresh weight) of Scotch pine feliage from different parts of its natural range (Langlet, 1936). Country of # of Seedlots Dry Matter Content Origin % fresh weight Finland 25 37.6 Norway 26 34.9 Russia 16 34.2 Scotland 3 30.1 Lithuania 7 32.2 Germany 3 31.7 Ukraine 7 30.3 Poland 11 29.7 France - 3 29.3 Central France 1 29.0 Italy 1 29.1 Spain 2 27.0 13 Langlet also studied seasonal variation in dry matter content. His seasonal data showed the same relative latitu- dinal trend. Dry weights increased toward the end of the growing season, drOpped during December, and reached maximums during early spring. In an effort to explain the seasonal variation in dry matter content, Langlet determined the seasonal variation in foliar sugar (glucose) content of several Scotch pine seed- lots. The seasonal pattern in sugar content paralleled closely the pattern for dry matter content. From fall to winter, the sugar content increased for both northern and southern seedlots. The consistent winter increase in sugar suggests a relation between sugar storage and adaptation to cold conditions. To better document the variation in sugar content, Langlet periodically determined sugar content and dry weight from September to June for four northern and two southern seedlots. Table 2 contains the results of this study. In fall and early winter, trees from northern regions accumulate more sugar than trees from southern regions. A possible explanation for this phenomenon could be the early occurrence of frost in northern areas. The accumulation of sugar may either aid in a rapid acclimation to cold conditions or provide a store of energy which may be required for the functioning of another cold resistance mechanism. During December, sugar and dry matter content dropped for trees from both northern and southern regions. This drop may be 14 Table 2. Seasonal variation in foliar sugar and dry matter content for northern and southern seedlots of Scotch pine. Month Northern seedlots Southern seedlots (Northern Sweden) (Hungary and Germany) Dry matter sugar Dry matter sugar --------------- % fresh weight-------------------- September 35.3 7.8 26.8 5.0 November 34.5 17.1 26.4 16.2 December 33.8 14.5 26.7 13.7 February 34.9 17.9 27.8 22.8 April 36.5 15.3 28.2 20.2 June 36.2 10.8 35.3 9.4 15 related to the completion of the plant's acclimation to early winter cold, and an initiation of enzymatic activity once again preparing the plant for the extreme cold of winter. The rise in sugar content with the onset of winter was much greater in trees from the southern regions. This may be related to the fact that temperature decreases from December to February are greater in the southern regions than in the already cold northern regions. Finally, the decline in sugar content arrives with the onset of spring. The low sugar con- tents during Spring in the northern trees are probably related to their earlier break from dormancy than trees from southern regions. 16 VARIATION IN DRY MATTER CONTENT Within-species variation: Few studies have been done on genetic variation in dry matter content since Langlet's (1936) Scotch pine study. One of the more relevant studies Since then was undertaken by Pharis and Ferrell (1966). They determined foliar dry matter contents of 5 and 16 month old Douglas~fir seedlings growing under well—watered conditions in a greenhouse at Corvallis, Oregon. The study compared 4 inland and 3 coastal seedlots. For seedlings of both ages, coastal and Arizona seedlots had significantly higher moisture content than seedlots from Montana, northeastern Washington, and Utah. Whether these differences in moisture content indicate adaptations which are physiologically significant to the plant in c0ping with their environment or whether seedlings with low moisture content merely have smaller cells or thicker cell walls is not known. In 1963, J. W. Hanover compared foliage dry matter con- tent of western white pine (Pinus monticola) trees which were resistant and susceptible to infection by the blister rust fungus. He found that dry matter content varies with age of needles and groups of trees growing in contrasting environments. However, no significant differences were found in the dry matter content of susceptible and resistant strains. The slightly higher dry weight found in the susceptible tissue was believed to be due to the effects of rust infection on susceptible trees and was unrelated to resistance. Schfitt and Hoff (1969) pursued this study and also found no 17 differences in dry matter content between the different strains. They realized, however, that foliage dry matter comparisons of different western white pine trees are com— plicated by a large within—tree variability for which light is the controlling factor. Needles from the sunny portion of the crown had significantly less dry matter content than shaded needles. A study by Philpot (1963) compared dry matter content of ponderosa pine (Pinus ponderosa) trees growing at various elevations in the central Sierra Nevada Mountains. His results showed no significant differences in dry matter content of trees growing at different elevations. He did discover a relationship between dry matter content and part of the crown sampled, Dry matter content of the foliage increased from the tip of the crown to the bottom. Saetersdal (1963) compared drying rates of excised plants of various provenances of Norway spruce and coastal Douglas- fir. In general, Douglas-fir seedlots from Bella Coola, British Columbia had the lowest moisture content and the Slowest drying rate. Seedlots from the mountains of Washington had the next slowest drying rate, while the low- land provenances from the Olympic Peninsula dried the fastest. In Norway spruce moisture content and drying rate were highly correlated with latitude of origin. The southernmost German seedlots had a significantly higher moisture content and slower drying rate than the northern Norwegian seedlots. Saetersdal suggests that Slow cuticular drying rates may 18 result in resistance to desiccation during late winter when the ground is frozen and plants are subjected to sunny con— ditions. Diurnal variation: Because of the daily lag of absorp- tion behind transpiration, a daily cycle in water content occurs in plant tissues (Kramer, 1949). Kozlowski and Peterson (1960) suggested that internal moisture distribution and absorption of water from the atmosphere are additional factors initiating diurnal changes in moisture content. The leaves of most trees show a daily cycle in water content, whether growing on a dry or moist soil. In general, minimum water content occurs in the early afternoon or about the time of maximum transpiration; maximum water content usually occurs near midnight and decreases toward morning. (Kramer.and Kozlowski, 1960). In keeping with the patterns of tranSpiration, one would eXpect maximum water content to occur just before sunrise. The speculated reason for the early morning decrease is‘a redistribution of carbohydrates in the plant during the night which initiates translocation of water from the leaves to other organs (Kramer and Kozlowski, 1960). Although the water content of leaves varies with Species, age, and season, a similar diurnal cycle has been reported by several investigators (Ackley, 1954; Kozlowski and Peterson, 1960; Zaerr, J. B., 1971; Gibbs, 1935; Jones, 1972; Cleary and Waring, 1965). Studies refuting a diurnal variation pattern in foliar moisture content have also been reported. Working with l9 Engekmann spruce (Picea engelmanni) in northern New Mexico, Gary (1971) found no significant diurmal differences in moisture content of needles up to 4 years of age. The small differences existing were attributed to variation within crowns. Jameson (1966) collected foliage from 10 juniper (Juniperus spp;) and 10 pinyon pine (Pinus edulis) trees at 2-hour intervals from daylight to darkness in the summer and winter. There was a significant midday depression in moisture content of the juniper trees. The pinyon pine results re- vealed no diurnal differences in moisture content during the summer. In the winter, however, pinyon pine foliage had slight increases in moisture content in the morning, with a significant afternoon drOp. Jameson speculated that pinyon pine leaves may possess some inherent mechanism to retard moisture loss that would normally occur during summer. Seasonal variation: Seasonal changes in foliar moisture content also occur, the water content of trees usually de- creases toward the end of the growing season. Often a slight rise in moisture content is noted during early winter, again followed by a steady decrease until the onset of the new growing season. Studies by Ackley (1954) have shown that the seasonal reduction in moisture content is caused by increased leaf dry weights rather than a decrease in moisture content. Another theory concerning seasonal variation has been prOposed by Jameson (1966). He suggests that seasonal variation occurs because of differential behavior of the 20 stomates; i.e., stomates of some species may close during dry periods in order to retard transpiration and allow moisture content to increase. Evidence supporting Ackley's theory is offered by several authors (Gary, 1971; Kramer and Kozlowski, 1960; Pharis, 1967). Studies by Kozlowski and Clausen (1965) have indicated similar seasonal patterns for various angiOSperm and gymno~ Sperm species. However, different degrees of seasonal change were noted between angiosperms and gymnOSperms. Dry weight changes in gymnosperms are believed to result from rapid carbohydrate translocation out of older needles and into growing needles. Thus, there is a decrease in dry weight in the older needles early in the growing season. 21 GENETIC VARIATION IN DOUGLAS-FIR Provenance research in Douglas—fir began in 1912 when T. T. Munger of the United States Forest Service started an experiment in northwestern Oregon and western Washington. Seed were collected from 14 localities and two years later were outplanted at 4 plantations, using the same arrangement at each site. Munger and Morris (1936 and 1942), reported on this study and found a consistent superiority in height growth in trees grown from seed collected in Granite Falls and Darrington, Washington. Morris (1957), working on two of the previously established test sites, discovered a strong genetic control over time of bud bursting in Douglas-fir. Trees from localities with warm spring days and nights started growth earlier than trees from localities with warm days but cool spring nights. The results of this early provenance test encouraged many others in the Pacific Northwest to pursue studies on the genetic trends in Douglas-fir. Irgens-Moller (1957) collected Douglas—fir seed from 7 localities at various elevations in western Oregon and sowed them in a greenhouse under controlled conditions. PhotOperiod studies revealed that trees grown from seed collected at high altitudes were the latest to start growth under natural daylength conditions and were significantly affected by long day treatments. Long day treatments hastened bud burst. In 1958, Irgens-Moller studied genotypic variation in time of growth cessation of trees grown from seed collected 22 in British Columbia, Washington, and Oregon. Trees native to the coast stopped growth one to four weeks later than trees from interior British Columbia, and higher coaStal elevations. A positive correlation was found between growth cessation and height of trees; those stepping growth earliest attained less growth. The differences in time of growth cessation were ascribed to different responses to the natural decrease in day length during the summer. The study indicated that sensitivity to daylength changes is of selective advantage to plants native to areas where the climatic changes from summer to winter are severe and abrupt. Ching and Bever (1960) started a provenance study in Corvallis, Oregon. They collected seed from trees growing in 14 locations ranging from southern Oregon to Vancouver Island. Nursery performance of the seedlots was evaluated with respect to variation in time of bud burst, bud set, frost damage, height and needle length. Seedlings from Vancouver Island, British Columbia, and the Shelton Area of Washington were generally faster growing and had the longest needles. Southern seedlots burst bud the earliest, set buds the latest, and suffered the greatest amount of frost damage. They suggested that differences in selective pressures from locality to locality have led to the evolution of genetically distinct types with regard to those traits studied. Improvement of Douglas-fir was also undertaken abroad. A provenance test was started in Rotorua, New Zealand, 23 containing 30 seedlots of Douglas—fir from west of the Cascade Range and Sierra Nevada Mountains. Sweet (1965) reported that the fastest growing seedlots were from Santa Cruz, California. He suggested that Douglas-fir from this area had probably developed superior growth as a result of a warmer and longer growing season. In the eastern United States, early studies indicated poor survival of coastal Douglas—fir seedlots. Interior seedlots, however, were relatively hardy and varied widely in performance. One of the first provenance tests of interior seedlots was undertaken by Baldwin and Murphy (1956) in New Hampshire. They sowed seed collected from New Mexico, Idaho, Washington and Montana. Trees from New Mexico had the least mortality, greatest height, and longest shoots. Further- more, all four Douglas~fir seedlots grew faster and had less mortality than native balsam fir (Abies balsamea) of the same age on the same site. Byrnes (1958) grew several Douglas—fir seedlots at 24 test sites in Pennsylvania. He categorized the seedlots as coastal, southern interior (southern and central Rocky Mountains), and northern interior (northern Rockies and British Columbia). Southern interior trees were superior in survival and growth. Gerhold (1965) planted several of the same seedlots of Douglas-fir that I studied in Pennsylvania. His results showed coastal sources to be the tallest but the most vari— able in survival; Arizona and New Mexico seedlots were the 24 tallest of the Rocky Mountain origins and had good survival; trees grown from seed collected in northern Idaho and north- western Montana had good height,growth and excellent survival. A New York provenance test was established in 1968 by Heit. He found that seedlings of Arizona and New Mexico origins had the fastest growth rate. His studies also showed that seeds from the coast, northern Rockies, and British Columbia require a prechilling treatment to break dormancy. A genetic study on interior populations of Douglas-fir was established by J. W. Wright and W. I. Bull of Michigan State University in 1962. The test contained 128 seedlots which were planted in four plantations in Michigan and Nebraska. Wright eg.al. (1970 and 1971) reported that seed- lots from Arizona-New Mexico were the fastest growing and had the bluest foliage.‘ Northern varieties tended to be slower growing and very hardy. Trees grown from seed col- lected in the Inland Empire were vigorous and very hardy. Wright £3.31: suggests that by careful seed source selection and prOper cultural methods Douglas—fir could be grown as a Christmas tree in Michigan on rotations of 6-10 years. CHAPTER 4 MATERIALS AND METHODS PLANTATION MATERIAL Seed procurement: In 1961, J. W. Wright and W. I. Bull of Michigan State University obtained seed from 128 natural stands of Douglas—fir (Figure 1). Two years later, H. D. Gerhold of Pennsylvania State University obtained seed from an additional 41 natural stands and shared the seed with .Michigan State University. Seed collection was promoted through the courtesy of the United State Forest Service and various local agencies within the sampled regions. In both cases, seeds from any given locality were col- lected from each of about 10 average trees in a native stand of at least 10 acres in size. Cones collected within a 5 mile radius were lumped together. Both series of collections covered the entire natural range of the species. Seeds were accompanied by data on latitude, longitude, elevation, soil type, aSpect, slope, collection date, etc. Nurseryppractice: Seeds from the 1961 collection were sown in the spring of 1962; seeds collected in 1963 were sown in the spring of 1964. The seeds were sown in the Michigan State University experimental nursery in East Lansing. In each case the seedlots were replicated and a randomized 25 26 complete block design was used. Each nursery plot consisted of a single 4 foot row in which 20 seeds were sown. The rows were spaced 1 foot apart. Seedlings from the 1961 collection were transplanted once or twice while in the nursery and were field planted as either 1-2 or 1-2-2 stock. Seedlings from the 1963 collection were not transplanted and were field planted as 2-0 stock. Outplanting procedure: In 1965 and 1966, four permanent test plantations were established in southern Michigan. In all plantations a 6x6 ft. spacing and 4 tree plots were used qoexcept MSFGP 4-67. Weed control was obtained by applying amino-triazole (1 gal./acre) in 2-foot strips the autumn prior to planting. After planting, simazine was applied at the rate of 4 lbs. per acre. All Sites were covered with dense grass and alfalfa sod before planting. Of the 128 seedlots collected in 1961, 68 were included in test plantations. Of the 41 seedlots collected in 1963, 28 were included in test plantations. Most of the seedlots which were not planted were of West Coast origins and suf- fered high mortality in the nursery from winter cold. From a comparison standpoint, it is desirable that every seedlot be represented in every replicate in all plantations. However, nursery stock was too limited to permit this in all cases, or in every plantation. Therefore, some seedlots are represented in 1-2 plantations only or in less than the full number of blocks in a plantation. Further details for individual plantations are as follows: 27 MS FGP 16-65/67 Planted 4/15/65 and 4/15/67 at W. K. Kellogg Forest, Kalamazoo Co., 65 miles SW of East Lansing:' 67 seedlots; 340 plots; 5 replicates; randomized complete block design; mortality 9%; average height 7 ft. in 1973. Site a level hilltOp; sandy loam soil of the Bellefontaine series; weed control with simazine resumed in 1968 and continued each subsequent year. MS FGP 6-65 Planted 4/27/65 at Camp Kett, Osceola, 150 miles NW of East Lansing; 36 seedlots; 153 plots; 5 replicates; randomized complete block design; mortality 25%; average height 5.5 ft. in 1973. Site rolling with 5-20% north SlOpes; sandy loam soil of the Montcalm series. MS FGP 9-66 Planted 4/19/66-4/20-66 at W. K. Kellogg Forest; 22 seedlots; 76 plots; 4 replicates; randomized complete block design; mortality 60%; average height 4.0 ft. in 1973. Site level: sandy loam soil of the Bellefontaine series. MS FGP 4-67 Planted 4/67 at Michigan State University's experi- mental nursery in East Lansing, Ingham County; 12 seedlots; 12 plots; 25-tree plots; 2 ft. spacing; not replicated; mortality 3%; average height 4.5 ft. 28 in 1973. Site Open and level; sandy loam soil of the Miami series; weed control continued annually. The reason that plantation MS FGP 16-65/67 was planted in 2 separate years was related to the size of the planting stock. In 1965 there were large differences in growth rate among seedlots. In 1965 the tallest seedlots averaged 5-12 in. tall and were ready for field planting. Others averaged 2—5 in. tall and were not ready for field planting. The Slow growing seedlots were lined out on the East Lansing nursery and field planted as 1-2-2 stock in alternate rows two years later. Further details are contained in a publication by Wright g: 3;. (1970) . 29 SAMPLING TECHNIQUE I collected foliage samples during the summer of 1973. Samples consisted of either current year's foliage or year- old foliage. In all plantations I collected foliage from all living trees in each plot. All material collected from one plot was combined and placed in a labelled, tared paper envelope and then sealed in a tared plastic bag to prevent moisture loss. Current year's foliage was collected from plantations MS FGP 16-65/67 at the Kellogg Forest and 5—65 at Camp Kett. The samples consisted of lO-45 gm. of needles and attached twigs. The time lag between removal and sealing in the plastic bags was approximately one minute. Year-old foliage was collected from plantations MS FGP 16-65/67 at Kellogg Forest, 9—66 at the Kellogg Forest, and 4-67 at the East Lansing nursery in the same manner as de- scribed above. The time lag between removal and sealing in the plastic bag was about two minutes. The dates on which the samples were collected from each plantation and the number of replicates sampled on each day are listed in the following tabulation: Current-year's foliage Year-old foliage Rep. # 16-65/67 5-65 16-65/67 vw9-66 4-67 1 June 26 Aug. 23 1 July 13 Aug. 17 Aug.13 2 June 26 Aug. 23 Aug. 3 Aug. 17 ~- 3 July 9 Aug. 23 Aug. 3 Aug. 17 ~- 4 July 9 Aug. 23 Aug. 6 Aug. 17 ~- 5 July 13 Aug. 23 Aug. 6 Aug. 17 ~- 30 AS shown above, samples were collected on different days, but all samples from one replicate were collected the same day. An analysis of variance Showed that differences in moisture content between replicates were significant (1% level), either because of site variation or day-to-day variation. Since an entire replicate was sampled in one day, sampling time did not affect the genetic results. To standardize collection procedures, each sample was removed from the middle portion of the crown and from the south side of a tree. In order to account for daily and diurnal fluctuations in moisture, accurate records of samp- ling time were recorded to the nearest minute. A general description of the weather conditions on each sampling day were also compiled. All samples were returned to the laboratory for weigh- ing within 24 hours after they had been sealed in the plastic bags. The envelopes containing the plant tissue were then placed in an oven at 70°C, and remained until they reached constant dry weight (approximately 36 hours). Again I weighed the samples and their moisture content was com- puted as a percent of their fresh weight according to the following formula: . . d wei ht Dry weight in % = figsh nggHEx 100. Adequacy of techniques: Within about 5 minutes after I collected samples, noticeable quantities of moisture had accumulated inside the plastic bags. This led me to believe that small quantities of moisture were probably lost during 31 the l or 2 minute lag time before sealing the plastic bags. This moisture escape could not be prevented or measured. I estimated that the amount of moisture lost was only .l-.2% of the total sample weight. Once samples were sealed in the plastic bags moisture loss did not occur. By a series of weighings on a tOp loading balance, it was determined that no measurable quantity of moisture could escape following the procedures used. RATE OF MOISTURE LOSS A separate study was designed to determine the rate of moisture loss between two races (ARINEM and INEMP), and the effect of inherent moisture content on rate of moisture loss. This study involved trees growing at Kellogg Forest (planta- tion MS FGP 16-65/67) only. One 2-3 ft. branch was removed from each tree in two seedlots of each race. The 30 branches (18 from ARINEM trees and 12 from INEMP trees) were collected from the middle portion of the south side of the crowns of vigorous trees in two replicates. The branches were sealed in tared plastic bags and brought to the laboratory where their fresh weights were determined. They were then placed on a laboratory table and allowed to dry at normal room temperatures (28°C). The I samples were weighed daily for the first 4 days, and perio— dically thereafter until 27 days after cutting. Also re- corded was the effect of moisture loss on foliage appearance. 32 Dates were recorded on which 50% and 100% of the foliage on each branch turned brown. A t-test was used to determine the significance of differences in moisture loss. It compared the differences between mean percent moisture loss of the two races studied. The degrees of freedom were 1 and 28 for race and within race, reSpectively. The data on the change of foliage appearance was analyzed by the Chi~square test. Each day's data were arranged in a 2 by 2 table with 1 degree of free- dom for each day. GROWTH RATE AND FOLIAGE COLOR Growth rate and foliage color are also important charac- teristics and were therefore measured. Height measurements were made at Camp Kett (MS FGP 16-65/67) and Kellogg Forest (MS FGP 5-65) at the end of the 1973 growing season (age 12 from seed). Foliage color was measured at Kellogg Forest (MS FGP 16-65/67) on September 4. In scoring color I used live tree standards with O as the greenest and 20 as the bluest foliage. In previous work color data taken with live tree standards proved more reproducible than that with paper standards. 33 ANALYSIS OF VARIANCE AND CORRELATIONS ANALYSIS OF VARIANCE Variation due toyyear of_planting: The Kellogg Forest plantation (MS FGP 16—65/67) was planted in two different years. During the interim, the trees were subjected to different environmental conditions. Ten seedlots were planted in each of both years, and were thus a means for comparing the performance of the trees planted in different years. Analyses of variance were done to determine the possible significance of differences due to year of planting. In these analyses, the degrees of freedom were 9, l, and 9 for seedlot, year, and seedlot X year interaction, respec- tively. The analysis of variance showed that year of planting did not result in differences in dry matter content or foliage color. Accordingly, year of planting was ignored in subsequent analyses of the data for these traits. The results of the analyses showed, however, that those trees planted in 1965 were significantly taller (1% level) than those of the same seedlot planted in 1967 (F value = 12.1). The seedlots planted in 1965 averaged 1.07 feet taller than those planted in 1967. As a result, 1.07 ft. was added to the height of each tree planted in 1967 before cal- culating racial means and the significance of racial differ- ences. Genetic variation: An analysis of variance was performed 34 for each trait (dry matter content, foliage color, and growth rate) using plot means as items. In seedlots which were re— presented in, at least, 2 but not in all 5 replicates, re- placement values were computed for the missing plots. This was done by using the seedlots mean, adjusted upward or downward as necessary, to compensate for differences among replicates. By using this method, daily variation as well as between replicate variation was considered. One degree of freedom was subtracted from the degrees of freedom for error for each missing plot. The Kellogg Forest plantation (MS FGP 16-65/67) contains 67 seedlots (of which 10 were planted in both years) belong- ing to 9 races as recognized by Wright £3 a1. (1970), 5 replicates (blocks), and 340 total plots. Degrees of free- dom were as follows: block-~4, seedlot-~66 (= 8 for race + 58 for seedlot within race), seedlot X replicate (= error)-- 268, year of p1anting—-l, total-~339. The levels of signif- icance were determined by testing the seedlot and seedlot within-race mean squares against the seedlot X replicate mean squares, and the race mean Square against the seedlot within-race mean square. The Camp Kett (MS FGP 5-65) and other Kellogg Forest plantation (MS FGP 9—66) contained 36 and 22 seedlots, respectively, belonging to 7 races. Similar analyses were done for data collected at these plantations. Of course, the degrees of freedom differed and it was not necessary to account for differences in year of planting. 35 An analysis of variance was used to detect possible significant differences in rate of drying. Thirty branches belonging to 2 races were collected. The degrees of freedom were as follows: race-—l, seedlot within race-~2, plot within seedlot——4, tree within plot (= error)--22, total-~29. Variation with time: Additional analyses were calculated to determine whether differences in sampling time within a replicate were important. Plot means were used as items. Spearman's Rank Correlation Coefficient Rank Correlations were run to compare moisture content results for different studies and plantations. The various sets of data were arranged in order of increasing moisture content, and the Spearman's coefficient (R) was computed directly from the differences between the ranks of two paired variables. The variables tested were moisture content data between: current—growing and year-old foliage, current foliage at different plantations, and means for Arizona seed- lots from all plantations. Degrees of significance were determined by comparing the computed coefficient (R) for the number of listed values (N), with the significance levels listed in the rank correlation table. CHAPTER 5 erases GEOGRAPHIC TRENDS IN DRY MATTER CONTENT For the purpose of comparison and clear presentation the sampled origins were divided into 9 geographical races (Figure l) and given code names according to the classifica- tion of Wright et 31. (1971). Their grouping was based on similarities and differences in several characteristics. My data supported their grouping in general, but I found it desirable to split their CWASH and ARINEM races. The races and their distribution are as follows: ALB ------------ 1 .lberta CMON ——————————— central Montana, Wyoming NOROC ---------- northern Rockies, west central Montana INEMP ---------- Inland Empire, northern Idaho and northwestern Montana CWASH Northern ————— north-central Washington Tieton Road——south—central Washington, Yakima Co. COAST ---------- Western lepe of the Cascade mountains of Oregon, Washington, and British Columbia NOCOL ---------- northern Colorado and adjacent Utah 36 37 SOCOL ---------- southern Colorado and adjacent Utah and Arizona ARINEM New Mexico———central and southern New Mexico Arizona ------- central and southern Arizona CURRENT YEAR'S FOLIAGE Differences between races: The racial differences in dry matter content of current years foliage are shown in Table 3. A single seedlot from Tieton Road in Yakima Co., Washington had the most succulent foliage, i.e., the lowest dry matter content. Its exceptional performance will be discussed in length later. At Kellogg Forest, Colorado—Utah trees (NOCOL and SOCOL races) had the next lowest dry matter content. This was probably related to a very damaging frost (ZSOF) at Kellogg Forest on May 18-19. When sampling this plantation in late June and July, it was quite apparent that the foliage collected from the NOCOL and SOCOL trees was fresher and more succulent than foliage of other races. Studies by Kim Steiner showed that the southern races leafed out earliest and were most severely damaged by the frost (Table 4). As a result, the foliage which I sampled was on shoots which had arisen from adventitious buds. These shoots were only 1-2 in. long and bore needles only half as long as those produced in previous years. Studies by Pharis (1967) have shown that new foliage of Douglas-fir contains much less dry matter than foliage 38 Table 3. Differences among geographic races of Douglas-fir in the dry matter content of fresh foliage. Region C Dry matter content of j urrent—year Year-old Foliage of foliage foliage collected collected at collected at at Corvallis, or(B) Origin Kellogg Camp Kett Kellogg and East S-mo-old l6-mo-old 16-65/67 5—65 Lansing(A) Foliage Foliage ---------------- % of fresh weight--------------------- ALB 33.1 -- 43.0 -- -- CMON 33.7 34.3 43.5 43.2 45.2 NOROC 34.0 -- 43.2 -- -- INEMP 34.7 35.1 43.3 43.2 47.0 CWASH Northern 34.0 35.5 43.7 -- -- Tieton Rd. 29.6 -- 38.7 -- -- COAST -- -- -- 37.7 41.6 NOCOL 30.6 35.2 44.2 -- -- SOCOL 30.8 34.4 43.5 41.8 -- ARINEM New Mexico 33.2 34.5 43.4 -- -- Arizona 32.1 32.5 42.7 38.2 41.7 (A) tions MSFGP 16-65/67, 9-66, 4-67. (B)Data from Pharis and Ferrell (1966). Data for year-old foliage determined as means from planta- 39 Table 4. Racial differences in susceptibility to frost damage and time of leafing out of Douglas-fir grow- ing on plantation MS FGP 16-65/67 (Kellogg Forest) in Michigan.(c) Region Frost damage (6/73) Time of leafing out (spring,73) of % buds killed 1 = early Origion 10 = late ALB 71 5.1 CMON 59 6.2 NOROC 35 8.4 INEMP 29 8.1 CWASH 57 4.8 NOCOL 88 3.0 SOCOL 74 3.3 ARINEM 59 2.9 (C)Data courtesy of Kim Steiner. 40 which is only a few weeks old. As a result, low dry matter content of NOCOL and SOCOL foliage may have been a consequence of leafage rather than genotype of the trees. A further effect of the frost damage was to increase the variability of dry matter content of foliage from damaged seedlots. The frost damage resulted in error mean squares 2-10 times greater for NOCOL and SOCOL races than for others. Other seedlots at Kellogg Forest were also damaged by the frost. However, due to the lesser damage and the larger tree size, shoots arising from adventitious buds were not collected in these seedlots. In the Camp Kett plantation, where frost damage was much less noticeable, Colorado-Utah trees had normal dry matter contents. This would probably have been the case at Kellogg Forest, also, in a normal year. The Camp Kett data shows that Arizona seedlots had the lowest dry matter content, having 3% greater moisture content than the more northern INEMP group. At the Kellogg Forest, the Arizona trees also had low dry matter contents. An analysis of variance showed that racial differences in foliar dry matter content for both plantations were highly significant (Table 5). Tieton Road vs. Northern CWASH: The Tieton Road seed- lot had previously been placed in the CWASH race by Wright 35 al.(197l). It was included because morphological studies showed it was somewhat similar to others in this race and because it had been postulated that the crest of the Cascade Range was the dividing line separating CWASH from the COAST 41 Table 5. Analysis of variance for dry matter content data of current years foliage of Douglas-fir at planta- tions MS FGP 16~65/67 (Kellogg Forest) and MS FGP 5~65 (Camp Kett). ., . Kellogg Forest Camp Kett Source of Variation df ms f df ms f *** Seedlot 66 25534 2.63 35 679.4 1.64* *** *** between region 8 10479 7.18 6 1978.5 4.82 * within region 58 1460 1.51 29 410.7 (1 *** *** Replicate 4 18113 18.68 4 2230.0 5.37 Year of planting l 11.77 1 Total error 268 1003 113 593.5 Total 339 152 * significant at 5% level *** significant at 1% level or better 42 race. As a result of its exceptionally low dry matter content, further studies were made of its area of origin. The parental stand was located a few miles east of White Pass (elevation 4500 ft.), which is located at the crest of the Cascade range. Travellers through White Pass report Douglas-fir to be grow- ing there, so the Tieton Road seedlot may represent an area of overlap between the COAST and CWASH races. This is further indicated by data on other characteristics, as shown in the following tabulation. All differences between the Tieton Road and other CWASH seedlots indicate the Tieton Road trees tend to be more similar to the COAST race as described by Wright 33 31., 1970: Gerhold, 1966: Byrnes 33 31., 1958. Seedlot Height Color % buds Time Dry_matter origin ’ killed by of Current old frost leafing foliage foliage damage (3 yr.av.) ft. 0=green % l=ear1y % of fresh 20=blue 10=1ate wt. CWASH 6.9 11 26 4.8 34.0 43.7 Tieton Rd. 7.2 8 37 1.9 29.6 38.7 Arizona seedlots: Excluding trees from the two highly damaged races, Arizona seedlots had the highest moisture content of all interior races. That agrees with the results of a study done at Corvallis, Oregon by Pharis and Ferrell (1966). They compared foliar dry matter content of 5 and 16 43 month old seedlings of five Douglas—fir races, each repre— sented by 1-3 seedlots. For seedlings of both ages, the Arizona and COAST races had the lowest dry matter content _(Table l). A rank correlation was run comparing the dry matter contents of the Arizona seedlots growing at both Kellogg Forest and the Camp Kett plantation. The results Showed that the data from the two plantations were independent. The lack of correlation could be attributed to plantation x seed- lot interaction or differences in date of foliage sampling. Variation within races: Figure 3 illustrates the varia- tion in dry matter content within the between races,based upon Kellogg Forest data (plantation MS FGP 16-65/67). The seedlot means ranged from 25% to 36%. The races with the largest variability were NOCOL (range=8%) and SOCOL (range= 8%) and SOCOL (range=6%). An analysis of variance showed that differences in moisture content within races were signif- icant at the 5% level (Table 3). The significance was prob- ably largely attributed to variation within the highly damaged NOCOL and SOCOL races. The within race data from the Camp Kett plantation (MS FGP 5-65) was somewhat less variable than the Kellogg Forest data. The range was from 30% to 37%, as shown in Figure 4. The NOCOL and SOCOL races were not severely damaged by frost at Camp Kett and thus showed comparatively little variation. An analysis of variance showed that this variation was not significant. 44 Figure 3. Differences in current—years foliage dry matter content between and within races of Douglas-fir growing at Kellogg Forest (plantation MS FGP 16-65/67) in southern Michigan. Dots represent seedlot means. bio—m3 :mumn. “5.x. mm dII-dl! - d u - d S a 22:4 Suzi 3.5: .2 Image dfimz oomoe v 222“ 3a em :23. 52:32 000 00 0 one 060:0 coo 00300008 0 o owLoomwooo 0 0000000 00 45 Figure 4. Differences in current~years foliage dry matter content between and within races of Douglas-fir growing at Camp Kett (plantation MS FGP 5-65) in southern Michigan. Dots represent seedlot means . him—m3 Immm... “.0 .x. . . oc . . . . MmI. A 23:4 o o EMZE< 00.xo: .Z ofoo o 1.000% o oooooo 1.0002 o o o o Aim; ooooor o o Imdgo o o o 2020 o o 00 46 YEAR-OLD FOLIAGE The purpose of sampling year—old foliage was to compare dry matter content of current and year-old foliage, and to test the effect of frost damage on the dry matter content of the NOCOL and SOCOL races. The results as Shown in Table 1, reveal higher dry matter content for the year-old foliage in general, and considerably higher dry matter content for the NOCOL and SOCOL origins in comparison with others. Since the year—old foliage was not damaged by 1973 frosts, the relatively high dry weights for these races supports the previously mentioned belief that current foliage data for NOCOL and SOCOL races was not representative of their normal dry weights. Differences between races: The racial variation in dry matter content of year—old foliage was considerably less than that in the current year's foliage. Data from three plantations (MS FGP 16—65/67, 9-66, 4—67) were averaged and are presented in Table l. The Tieton Road seedlot had the foliage with the lowest dry matter content. The Arizona seed— lots were also low in dry matter content. An analysis of variance showed that trees from these races were significantly (2.5% level) higher in moisture content than all other races growing at the Kellogg Forest (plantation MS FGP 16-65/67). These results concur with the data from the current-year's foliage. Differences within races: The plotted seedlot means in Figure 5 shows that dry matter content of year—old foliage 47 Figure 5. Differences in year—old foliage dry matter content between and within races of Douglas—fir growing at Kellogg Forest (plantation MS FGP 16-65/67) in southern Michigan. Dots represent seedlot means . .5..ng Imam“. m0 o\o no a . . 00 . . . mm a a . 10m... 88 8 83:4 ace 6 00.x ,2 .z 2sz o owomuwo 40004 3 oeooowonww 1.0002 o .3... 88:. o oo 595.62 Iwx ooowmmgo afimz oooowooo oomo L o oo o o 2020 ac an.“ m:- 1" Fl) . I 48 was less variable than that in current—year's foliage. With the exception of the Tieton Road seedlot, the range in dry weight was from 41% to 46%. The dry matter content differ— ences within races were not significant. Rank correlations were run comparing dry matter con- tent trends between year—old (plantation MS FGP 16—65/67) and current-year's foliage (plantations MS FGP 16-65/67 and 5—65) of Arizona seedlots. In various combinations of testing all sets of data, the correlation coefficient was not significant. As a result, the value of selecting for high moisture content among different stands within Arizona is not, at this stage of research, a valuable asset. 49 DRY MATTER CONTENT OF CURRENT VS. YEAR-OLD FOLIAGE In comparing all moisture content data, the largest differences were between current and 1 year old foliage. It is important to emphasize that these differences are not necessarily associated with differences in actual amounts of moisture. Kramer and Kozlowski (1960) explain that as plant tissues become more mature, their cell walls thicken and the amount of dry matter increases. Therefore, a decrease in moisture content (expressed as a % dry weight) will occur without appreciable loss in water content. Evidence to support this is provided by Korstian (1933) who reported leaf moisture contents ranging from 52% (% fresh weight) in one year old leaves to 78% in newly formed leaves of ever- green Magnolia (Magnoliaflgrandiflora). Ackley (1954) found water content of Bartlett pear leaves to decrease from 73% (% fresh weight) in May to 59% in August, while at the same time actual moisture content increased. More recently, Kozlowski and Clausen (1965), found the same general trends in the leaves and buds of several species, including red pine (Pinus resinosa) and eastern white pine (Pinus strobus). Another relation noted was the decreasing variability with increasing age of the foliage. The results reveal that the moisture content of the current growing foliage collected at Kellogg in early summer was considerably more variable than it was a few months later at Camp Kett. In the same manner, the 1 year-old foliage sampled in the middle of the summer was less variable than either set of current foliage 50 data. These differences are probably associated with meta— bolic changes within the plant that occur early in the grow- ing season. As a result, Spring was probably a poor time to sample trees for moisture content. I ran rank correlations to compare racial trends in dry matter content for current and year-old foliage. The current foliage data from the Camp Kett plantation showed a moderate correlation (r=.63) with the trends in year—old foliage at Kellogg Forest. Similarly, a moderate correlation (r=.69) was found between current and year—old foliage at the Kellogg plantation(MS FGP 16—65/67). The primary consistency corro- borated by all sets of data was that trees grown from seed collected in Tieton Road (Washington) and Arizona contained the highest moisture content. 51 GENETIC VARIATION IN RATES OF MOISTURE LOSS Rates of moisture loss were compared for branches re- moved from trees of ARINEM and INEMP races, which represent extremes in foliar moisture content and have the greatest economic potential in southern Michigan. The samples were composed of branches from 18 trees belonging to 2 seedlots from Arizona and of branches from 12 trees belonging to 2 seedlots from northern Idaho. The patterns of moisture loss are graphically illustrated in Figure 6. The rates of moisture loss and the differences in moisture loss between the two races were greatest during the first two days. At the end of the second day, the branches from the ARINEM race had lost 17% of their fresh weight, while the branches from the INEMP race had lost 30% of their fresh weight. During the following four days, both races lost moisture at about the same rate. After 6 days of drying, branches from the INEMP race had a slight reduction in moisture loss. At the same time, the branches from the ARINEM race continued to lose substantial amounts of water. As the number of days increased, the curves for the two races began to approach each other. After 20 days of drying, the INEMP branches reached a constant weight (any loss after this was probably due to needle drOp), while the ARINEM branches continued to lose moisture. The experiment was stOpped after 27 days and the ARINEM branches had not yet reached a constant weight. Size of branch did not affect rates of moisture loss. 52 Figure 6. Differences in rates of moisture loss of cut branches of ARINEM and INEMP Douglas-fir races growing at Kellogg Forest (Plantation MS FGP 16-65/67) in southern Michigan. ox. or FRESH WEIGHT I00 90' 80 7O 60 50" 4O 30- 20' 10 I 1 3 6 9 I2 IS 18 T2! NUMBER OF DRYING DAYS 24 2“! 53 The smallest and largest branches from each race lost similar percentages of their fresh weights. Even the smallest ARINEM branches had not reached a constant weight on the last day of measurement. An analysis of variance was performed on the second days data. Differences between races were significant at better than the tests were used between the two days of drying, differences (1% 1% level (Table 6). For other days data t- to compare differences in moisture loss races. This test was used after 2, 6, and 9 respectively, and indicated significant or 2% levels) for each day. 54 Table 6. Analysis of variance for rate-of-drying data for ARINEM vs. INEMP Douglas-fir races growing in Michigan. Data analyzed after 2 days of drying. Source of Variation df ms f Race 1 586.7 10.04*** Seedlot within race 2 41.1 .7 Plot within seedlot 4 61.20 1.05 Error (tree/plot) 22 58.41 Total 29 *** Better than 1% level. 55 DISCUSSION The evidence indicates that trees from the ARINEM race have evolved some mechanism for conserving water. Whether the water conservation mechanism is related to the 3% lower dry weight in the ARINEM race is not known. One might have expected more rapid rates of moisture loss in the ARINEM race due to the higher moisture content. On the other hand, the retarded moisture loss might be the factor responsible for the higher moisture content. A study by Palpant (1973), compared rates of moisture loss of Douglas—fir of Colorado origins with central Montana, Idaho and British Columbia origins. The study was made under two humidity regimes and, in both cases, the northern Rocky Mountain origins dryed faster than the Colorado origins. This evidence agrees with my results that genetic variation in rate of moisture loss exists and southern Douglas-fir origins lose moisture more slowly. It is important at this point to compare the respective foliage color of the two seed sources examined in this study. Results of several color analyses, including one in this study, concur that ARINEM Douglas-fir are significantly bluer than other races. INEMP seedlots are among the most green of all Douglas-fir types. The color differences are attrib- uted to waxes, which coat the needles of many evergreens. The blue glaucous appearance of the ARINEM race is due to the light reflecting characteristics of the waxy foliage jsurface. The structure, arrangement and density of wax 56 largely determine the degree of claucousness (Hanover and Recovsky, 1971). It is no coincidence that the ARINEM races of Douglas- fir are bluer than the more northern races. This same ten- dency is found in several other species whose range extends into the southern Rocky Mountains (Kung and Wright, 1972). The southern Rockies have almost twice as many hours of sunshine per year as the northern Rockies. This region is also characterized by a low precipitation—evaporation ratio. The waxy surface of the tree leaves has probably been of selective value in several ways. It may act to reduce water loss and gaseous exchange from the plant surface. Also, photo-destruction of auxin is rapid under blue light. There- fore, reflection of blue light would be an effective auxin con- trol mechanism in areas of intense solar radiation (Kung and Wright, 1972). Results of this study suggest that waxy foliage surfaces may, indeed, play a role in modification of moisture loss. It must be emphasized, however, that trees compared in this study are of different genotypes, as well as different foli- age color. As a result, the possibility that other genetic- ally fixed mechanisms may be involved in modifying rates of moisture loss can not be eliminated. The relation between glaucousness and moisture retention has been questioned in a study by Recovsky (1973). He found no significant differences in moisture loss of glaucous and nonglaucous foliage of blue spruce (Picea pungens). Recovsky 57 subjected blue Spruce foliage to three treatments of various temperatures and humidity. His results for all treatments showed that glaucous foliage lost Slightly more water early than the non—glaucous foliage. After time the pattern re- versed. His results were essentially the Opposite of those fOund in this study. Recovsky selection of trees was based solely on foliage color with no regard to genotype. In conclusion, it can only be stated that differences in the rates of moisture loss between the ARINEM and INEMP races growing at the Kellogg Forest are considerable. The reason for these differences, if not related to the prescence of waxes, could be related to any of several anatomical or environmental adaptations. The need for further research on this matter is evident. 58 EFFECT OF MOISTURE LOSS ON FOLIAGE APPEARANCE While branches from the ARINEM and INEMP races were drying, I kept records on the effects of moisture loss on the appearance of foliage. The differences in drying rates were clearly reflected in the rate of foliage browning. The tabulation below lists the percentage of branches from each race which had at least 50% of their foliage turn brown after drying for several days. ARINEM branches retained a richer color for a considerably longer time than the INEMP branches. The most pronounced differences were found in the first 8 days of drying. At this time, all of the ARINEM branches had maintained a rich color while 42% of the INEMP branches had turned brown. % of branches turning brown after drying for: Race 3 8 ll 20 27 days days days days days ARINEM O 0 ll 39 88 INEMP 8 42 50 75 100 A Chi-Square analysis was used to test the significance of these results. The analysis was done after each measured day of drying. The analysis indicated Significant differences in foliage appearance between the two races for branches that were drying from 6-11 days. 59 VARIATION IN PERCENT DRY WEIGHT DUE TO TIME OF COLLECTION The leaves of most tree species Show a diurnal cycle in moisture content, whether growing on a dry or a moist soil. The chief cause of diurnal variations is the lag of water absorption behind water loss (Kramer and Kozlowski, 1960). Variation within replicates: Due to diurnal variation in moisture content found in many Species, it was necessary to perform a test to detect if time of collection affected moisture content. In order to facilitate such an analysis, time of collection was recorded as each plot was sampled. Each replicate was then divided into 5 one-half hour time periods according to the order in which the samples were collected. By subtracting the percent dry weight for each plot from the seedlot mean, a pattern of positive and nega- tive deviations vuns established. The deviations for each one-half hour time period were then compared in an analysis of variance. The analysis was done for current foliage and 1 year old foliage moisture content data at Kellogg Forest. The results showed that time of collectioncid not significantly influence the moisture content of samples from the same replicate. Detected differences were therefore attributed to error. The value of the analysis was in assuring that the moisture content results were due to genetic variation, and not related to transpirational differences associated with time of collection. CHAPTER 6 EVOLUT IONA RY CONS IDERAT I ONS I S OLAT I ON The Douglas-fir provenance study was designed to yield practical and evolutionary information about racial variation within the species. Populations occupying climatically different regions tend to adapt in such a way as to become genetically differ- ent. However, a prerequiste to such adaptation is isolation so as to prevent migration of genes between them. Absolute range gaps (such as mountain ranges, deserts, etc...) are the most effective barriers. As a result, racial differen- tiation is most apparent in areas such as the western United States and southern EurOpe where the ranges are disrupted. Distance can also act as an effective barrier to the migra- tion of genes within a continuous population. Several experiments have shown that pollen travels relatively short distances in quantity. Thus, isolation may be essentially complete if two populations are separated by a distance of 20 miles and there is no consistent one—way air or water movement (Wright, 1962). Population density is related to the distance needed to effectively isolate a population. If population density is 60 61 low there are fewer paths by which genes can migrate and genetic differentiation becomes more pronounced. Small equal sized pOpulations separated by 2 or 3 miles may often be considered as isolated. Douglas-fir range gapg: In Douglas—fir the most effective range gap is the desert—grassland region stretch- ing from Mexico to Canada and separating the West Coast from the interior forests. The Cascade and Sierra Nevada Mountains effectively cut off precipitation from the interior. As a result, the intermountain area has prevented migration ever since its existence. The boundaries of the northern interior range are out- lined by the Salmon River on the south, the Flathead river on the east, and the eastern Washington desert in the north. Trees lying within these boundaries grow at lower elevations than in the surrounding areas. These differences in eleva- tions are seemingly sufficient to effectively separate the Inland Empire region from central Montana and other northern Rocky Mountain pOpulations. The boundary separating the northern and central Rockies is the desert area along the Snake River of southern Idaho and the treeless area in the Wyoming basin. Within the cen- tral Rockies, the racial distinctions are not great. The trees are separated by small elevational differences, which follow a pattern of increasing latitude and decreasing ele- vation. A pattern which tends.to compensate itself with regards to temperature and length of growing season. The 62 races in the central Rockies are possibly portions of clines, rather than distinct races in themselves. The approximate boundaries separating the races of the southern Rockies are the Painted Desert and the treeless region south of Zion Canyon (Kung and Wright, 1972). A racial boundary seems to exist along the Colorado-New Mexico border even though it cuts through forested regions. NATURAL SELECTION Most racial differences are associated with climate at the place of origin, which suggests that natural selection is a major factor promoting racial differentiation. In most cases, the elevational zone occupied by a species varies so that similar climatic conditions occur throughout the entire range of a Species. This phenomena does not hold for the entire range of Douglas—fir. At the same elevations, the climate of the coastal Douglas-fir region is warmer than that of the interior; it is warmer and moister in the Inland Empire than it is in surrounding areas; and, it is warmer and dryer in Arizona and New Mexico than in the northern parts of the range. Furthermore, Douglas-fir grows at surprisingly low elevations in northern Idaho and Arizona and New Mexico, which further accentuates the climatic contrasts. The result of these distinct climatic regimes has been the evolution of specific adaptations. Natural selection favors the fastest growing trees in a habitat, providing that those trees are fit in other respects. 63 The lower elevation and more southerly location produce a much warmer climate for Douglas—fir of the ARINEM race. Natural selection has resulted in adaptation to these con- ditions, resulting in a type which grows one month longer and taller in Michigan than other types. In the north, the 10wer elevation of the Inland Empire results in a warmer and moister climate there than is found farther south or east. Adaptation to this climate has resulted in INEMP trees with faster growth than trees from the northern Rockies and cen- tral Montana (Wright, 1971). 64 SELECTION AND MOISTURE CONTENT Natural selection has probably played an important role in determining the moisture content of Douglas-fir races. The differences are believed to be related to the varied climatic conditions throughout the interior Douglas-fir range. In the hot, dry southern Rockies selection for drought resistance would be a valuable asset, whereas in the colder northern Rockies cold resistance may be of primary importance. Both of these factors are believed to have played a role in selection for foliar moisture content. DROUGHT RESISTANCE The effect of foliar moisture content on drought resist- ance has been debated for several years. Early views sug- gested that drought resistant plants have low rates of water loss, but this was discarded when it was discovered that many xerOphytes transpire rapidly when supplied with water. Emphasis then shifted to the idea thatability to withstand dehydration is the primary factor involved in drought resist- ance. At present, it is suggested that drought resistant plants must have protOplasm with some capacity to endure dehydration combined with the morphological and anatomical characters which decrease the rate of water loss and post- pone the development of critical internal water deficits. Drought resistance can be differentiated into two specific types as suggested by Levitt (1956), that is, those plants which endure drought and those which avoid 65 drought. Those plants which endure drought are able to with- stand a dry internal environment resulting in severe tissue dehydration. Drought avoidance, on the other hand, may in- volve some aspect of the plant's physiology or morphology which enables it to avoid a dry internal environment. Re- sistance often involves both ability to endure drought and avoidance of drought. Many non-succulent, drought-resistant species are able to endure drought. That is, they lose exceptionally large quantities of water so that their protOplasm is subjected to very negative potentials, and yet they are not killed. These plants are true xerOphytes and exhibit the prOperty of hardiness to drought. True xerophytes are capable of maintaining protOplasmic elasticity to high levels of desiccation. Succulent species, such as cacti, are drought avoiders. They avoid drought by storing water in their succulent tissue. Enough water is stored in their tissue, and its rate of loss is so extremely low that they exist for long periods of time without added moisture. Many non-succulent species avoid drought by various anatomical modifications such as a very deep root system, reduction in leaf size, sunken stomates, or heavy pubescence on the leaf surface. Avoidance of drought has often been explained on the basis of water conservation. Water savers are able to restrict transpiration long before wilting occurs, and they may succeed under conditions of more extreme drought than plants 66 which are water Spenders. Many of the more moderate xero- phytes and even some meSOphytes are water savers. As a result, the superior drought resistance of a species or a variety has been found to be partly or even solely due to water conservation (Kramer and Kozlowski, 1960). This adaptation for water conservation has accounted for the differences in habitats of many conifers. StudieS' by Johnson Parker (1951 and 1954) compared relative rates of moisture content and moisture loss on various species of northern Rocky Mountain conifers. His studies Showed that the leaves of dry site ponderosa pine (Pinus ponderosa)con- tained greater moisture and lost water less rapidly than moist and wet sites species of Douglas-fir and western arborVitae(T§ujayplica33). Similar results have also been found within Species that occupy large and varied environments. Meuli and Shirley (1937) showed genetic variation in drought resistance of green ash (Fraxinus Pengsylvanica). Drought resistance in- creased significantly from south to north in the prairie plains states. Trees showing the greatest resistance to drought were consistently associated with regions of driest climate and longest drought periods. Similar evidence was accumulated by Kriebel (1957 and 1963) in studying the genetic variation in sugar maple (Acer saccharum). Trees from the cool, damp climate of the northern Appalachian forests produce progeny more susceptible to drought condi- tions than trees from the hot and dry climates of southern 67 and midwest habitats. In 1966, a study was undertaken by Ferrell and Woodward to determine the effects of seed origin on drought resistance of Douglas-fir. Trees from six interior and coastal seed sources were subjected to various drought conditions, and their relative resistance was measured in terms of a "time to death" index. In all cases, interior seed Sources were more drought resistant than coastal sources. In two of three drought experiments, Arizona seedlings lived significantly longer than seedlings from other origins. In a third test, the seedlot from northeastern Washington survived longer than the Arizona seedlings, but the differences were not significant. In seeking further evidence on drought resistance in Douglas-fir, this study was pursued by Pharis and Ferrell (1966). Their first study compared the foliar moisture content of 6 interior and coastal seedlots. The results showed that the Arizona and coastal seedlings (Vancouver Island and Corvallis, Oregon) had significantly higher moisture content than other interior seedlings. For the Arizona seedlings this can be postulated as an adaptation to a relatively dry habitat. The coastal region, however, is well watered, (except in summer). These trees may, therefore, not require a moisture retention mechanism to survive. This lack of adaptation was again clearly shown in the "time to death" and "lethal soil-moisture content" indices of drought resistance. The results of these tests l _ 68 showed that the interior seedlings were Significantly more drought resistant than the coastal seedlings. In the time to death test, the northeast Washington seedlot proved slightly hardier than the Arizona seedlots. In the soil moisture lethal point study, it was discovered that the Arizona seedlings were able to survive at the lowest soil moisture content of all seedlots. It is this adaptation for drought resistance among dry habitat plants that is believed to be, at least, partly responsible for the high moisture content of the Arizona Douglas-fir seed source. The results of my study verify that the southern ARINEM race of Douglas-fir have an inherently higher moisture content and Slower rate of moisture loss than more northern sources. It is probable that in dry habitats selection favored types that are able to utilize and conserve limited moisture. Similarly, experimental evidence has Shown that southern Arizona seed sources have a significantly Slower rate of moisture loss than a more northern race. It is also interesting to note that surface waxes, evident on the foliage of southern seedlots only, are believed by many to be an adaptation to water conservation. This point, however, has not been satisfactorily proven in all cases. Trees from Arizona seedlots have probably adapted a mechanism for re- tarding transpiration and thus conserving moisture. Further evidence supporting drought resistance in southern seed sources was discovered in Douglas-fir provenance 69 tests in Michigan and Nebraska. In all Michigan plantations, the ARINEM and INEMP races were the fastest growing (Wright 33 31., 1970). However, on the dry prarie soils of Nebraska only the ARINEM seedlots grew fast. INEMP seedlots decreased in vigor with time. The evidence thus far accumulated suggests that natural selection has responded to the dry climates of the Arizona region by providing Douglas-fir from that area with adapt- ations for high moisture content and moisture conservation. Based on the results of Pharis and Ferrell, it appears that natural selection has not played the same role in the northern and coastal Douglas-fir races, as in the southern races. Their results present two inconsistencies: (l) coastal seed sources (Corvallis and Valsetz, Oregon and Vancouver Island) although comparatively non-drought resist- ant, contain an equal or at times a greater amount of foliar moisture than the dry habitat Arizona seed sources, and (2) in some cases, north-eastern Washington seedlots were equal in drought resistance, but contained less foliar moisture than Arizona seed sources. It is feasible to postulate that these inconsistencies may be a result of any or all of the following circumstances: 1) The coastal seed sources are clearly a separate taxonomic variety, and are probably subjected to unique genetic, as well as, environmental modifications. 2) Although the coast receives an abundant supply of moisture on an annual basis, a very small amount (2-4%) 70 occurs during the summer; the time in which the studies were performed. .As a result, the higher summer moisture content among coastal sources may represent an adaptation to the relatively xeric conditions present during those months. 3) Another factor, not drought related, may be wholly or partially responsible for moisture content adaptation. The first two explanations are clear possibilities, which can only be substantiated by further experimental evidence. The third explanation is subjective, and there- fore allows for further selective interpretation. COLD RESISTANCE Another climatic factor which may be related to varia- tions in percent dry weight is temperature. Literature explaining the physiological or environmental relationship between the two variables is comparatively scarce; never- theless, data Showing correlations between high dry matter content and low temperature is available. Langlet's (1936) study of variation in dry matter con- tent of Scotch pine showed significant correlations between dry weight and winter temperature at place of origin (Figure 1). Trees from cold northern latitudes had the highest per- cent dry weights. This relation held true for the Swedish provenances, as well as range wide provenances. The differ- ences among Swedish seedlots were large even though precipi- tation is fairly constant throughout the country. Langlet and others have studied the seasonal variations 71 in percent dry weight as well. Although the results for different species vary slightly, all the data indicate an increase in dry weight with the onset of colder seasOns. In other studies, Langlet Showed that the seasonal increase in dry weight was at least partly due to increases in sugar (glucose) content in the plant. Further investigations by Langlet showed positive correlations between high dry matter content and high sugar content during the winter. The trees from the colder northern origins had higher dry weights and higher sugar content. It is, therefore, feasible that selec- tion for high dry matter content is an adaptation which enables trees to endure winter cold. A relationship between temperature at place of origin and dry matter content is also evident in Douglas-fir. It has been found that trees from cold northern parts of the Douglaséfir range contain significantly higher dry weights than trees from warm southern and coastal Sources. The dry weight data presented by Pharis and Ferrell agree with this relationship. Such evidence suggests that selection for dry matter content might be a response to temperature at place of origin, rather than precipitation as previously suggested. If this is the case, an adaptation for high dry matter content in colder climates would result in the advantage of cold resistance found in the northern parts of the Douglas-fir range. At the same time, low dry matter content, found in warmer southern and coastal races, occurs at the expense of 72 cold resistance. This suggestion is supported by the fact that coastal origins of Douglas-fir suffered severely from winter cold when planted in Michigan (Wright, 1971)." Assuming this theory to be valid, then the low dry matter content of coastal sources and the high dry matter content of northeastern Washington seedlot (Pharis and Ferrell, 1966) are explainable with respect to variation with temperature at these respective origins. 73 CONCLUSION Attempts to document variations in moisture relations within and between species have often led to sttdies of drought resistance. AS a result, evidence linking moisture relations and drought resistance is generally overwhelming. However, a closer look into the exceptions to this evidence suggest that another factora namely temperature, is equally involved. It is likely that both temperature and drought hardiness play a varying role in the evolution of dry matter content in different geographical areas. The results of the foregoing discussion point out that selection processes Operate to develOp different adaptations in various environments in nature. In general, Douglas-fir originating from the southern Rocky Mountains contains more moisture and have developed a means of retarding moisture loss. These adaptations are probably a response to the droughty conditions and warm temperatures. The northern interior Douglas-fir races possess higher dry weights, which is likely an adaptation to the cooler climates and high moisture availability. Coastal Douglas-fir is subjected to a completely different set of environmental conditions than the interior variety. It is probable that high moisture content in trees of this region represents a lack of adapta- tion to cold and a storage mechanism for hot dry summers. Selection and isolation have combined to segregate Douglas-fir into several races. Because Douglas-fir is a wide ranging species, its components are subjected to 74 variable environments and a vast array of competing and associated plant species. Such variation requires flexible adaptive adjustment for survival insurance. Presumably, the basic mechanism for COping with the physical and biolo- gical environment is genetic differentiation or adaptation. This, in turn, facilitates selective adaptation and the pro— motion of genetic strains within species. CHAPTER 7 PRACTICAL IMPORTANCE AND APPLICATION IMPORTANCE As of 1972, Douglas-fir comprised only 4% of the total Christmas trees sold in Michigan. It also demanded the highest price (approximately $1.25 per foot wholesale). The reason for the relatively low production is because, until recently, eastern growers have found Douglas-fir to be a slow grower. Recent genetic studies have offered information which irefutes the slow growing reputation Douglas-fir has attained for the past several years. The results of these studies have been the identification of races with optimum growth, foliage color and hardiness characteristics when planted in southern Michigan (Wright, 33 31., 1970 and 1971). Correct use of these published results have resulted in rotation ages half as long as had previously been experienced. One of the aims of studying the variation in dry matter content was to bring a better and safer Christmas tree to growers and the public. The effect of moisture content on the longevity and appearance of Christmas trees has been drastically overlooked in the past, despite the complaints of the public and salesman of dried out trees. In 1972, the 75 76 Christmas tree harvest in Michigan began as early as September 11 on some plantations. In these cases, the trees either lied around in lots or in houses for 3 months. The result being undesirable and unsaleable trees in many in- stances. In an effort to reduce the drying problem, many growers delayed cutting until October 9, but still many Christmas trees had to be sprayed with pigment to be sale- able. The results of this study offer evidence which could minimize drying problems in Douglas—fir Christmas trees. The 3-4% higher moisture content of Arizona seedlots is of tremendous significance and will greatly enhance the longe- vity of trees from that origin. This, of course, is of practical benefit to those who store or ship trees. Of practical benefit to growers and the public is the effect of dry matter content on the appearance of the tree. The studies have Shown that trees of high dry weight tend to dry faster and lose their color more rapidly than low dry weight trees. This would be a detriment to the seller and produce a displeasing tree for the public. In addition, the association of high moisture content with fire resistance is an apparent one which is of extreme interest to the public. It is this problem of fire susceptibility which has gradually leaned the public towards fireproof artificial trees. An additional value of moisture content data might be found in connection with site requirements. Douglas-fir is known to be a demanding species with reSpect to moisture and 77 soil requirements. Due to the fact that soils of Michigan are sandy to sandy loam at best and have a relatively low moisture holding capacity, it would appear that the high moisture varieties would be better adapted to handle such conditions. It is important to emphasize that in addition to racial selection, care must be taken to avoid dessication in cut trees. Palpant (1973) has shown that by storing Norway spruce (Picea abies)in a cool refrigerated environment, moisture content of the tree could be maintained. This is of particular importance in shipping trees. In trees left outside, studies have shown that a cool, moist, and shaded site with light winds had the greatest effect in maintaining moisture content. The use of anti-desiccants did not im- prove moisture retention in Norway spruce (Palpant, 1973). An important implication of drying rate may be found during late winter when complete cessation of water uptake occurs. This occurs when plant roots are confined to the frozen stratum of the earth, and plants are subjected to sunny conditions and dry winds. During such conditions, Saetersdal (1963) has found that slow drying seedlots of Norway spruce (southern German seedlots) are more resistant to winter drought damage than faster drying northern seed- lots. Winter drought damage occurs when plants transpire, and are unable to withdraw water from frozen soil. .It results in browning of foliage and may possibly lead to death of the tree. Under conditions of winter drought, the 78 differences between Douglas—fir seedlots in drying rate may be of importance. Based on drying differences, ARINEM seed- lots may be expected to survive better than INEMP seedlots under such conditions. APPLICATION Moisture content data can only be of value to the Christmas tree industry if other growth characteristics are also suitable. Growth rate and hardiness are probably the most important considerations in choice of seed source for \ ornamental or Christmas tree use. The value of moisture content data comes in choosing between fast growing and hardy types, and in improving the quality of the tree. Table 7 represents a compilation of growth data on two Douglas-fir plantations in Michigan. Differences between races in all characteristics listed were statistically significant. Arizona and New Mexico origins are fastest growing, have the bluest foliage, and low dry matter content, but they do suffer considerably from late spring frosts. Nevertheless, they have been very successful in Michigan and command tOp price, despite their reputation for frost damage. SOCOL trees have the same desirable foliage as the ARINEM race, but they are slower growing and suffer a great deal more frost damage. At very young ages the SOCOL trees were among the fastest growing, but they have since fallen off the pace due to repeated yearly frost damage. The northern origins are more hardy than the southern 79 Table 7. Racial differences in height, foliage color, and frost damage in Douglas-fir growing at plantations MS FGP l6-65/67 (Kellogg Forest) and 5-66 (Camp Kett) in Michigan. Height data for trees 12 years old from seed. it a! Region Height Foliage color frost damage of (ft.) 20= blue % buds origin 0= 1t. green, killed J ' Kellogg Camp Kett Kellogg Kellogg Camp Kett (3_yr.av.)(2 yr.av.) ALB 6.5 --- 14.1 36.5 --- CMON 6.3 4.0 10.0 29.0 15.0 CWASH Tieton Rd. 7.2 --- 8.0 37.0 --- No. CWASH 6.8 5.2 11.0 26.0 22.0 NOROC 8.0 --- 6.5 19.0 --- INEMP 8.8 5.6 7.3 18.0 12.0 NOCOL 5.2 4.1 12.1 53.3 24.0 SOCOL 7.7 4.9 15.4 44.0 32.0 ARINEM Arizona 10.6 5.9 '16.4 35.0 33.0 New Mex. 10.4 7.4 17.3 37.5 24.0 80 ones, but seem to be lacking in growth rate, have a less desirable foliage color, and a high dry matter content. Of these, only the INEMP race is fast enough growing for large scale planting. Although hardier than the ARINEM race, the INEMP race is slightly slower growing, has greener foliage, and dries out much faster than the Arizona trees. It is important to remind growers, that no one type of seed will be best for all conditions or all uses of the tree. The grower must define his planned use of the tree and evaluate local climatic conditions before choosing a seed source. Site selection and care are also important factors. Even the best seed sources will not flourish in frost pockets or very dry sites. It is believed that growers in southern Michigan can achieve maximum results with a mixture of ARINEM and INEMP Douglas-fir seed sources, in combination with good cultural practices. The ARINEM race seems to offer the best quality tree, but the hardiness of the INEMP trees provides the insurance which is essential to the success of any large scale Operation. LITERATURE CITED Ackley, W. B., 1954. Seasonal and diurnal changes i the water content and water deficits of Bartlett pear leaves. Plant Physiol. 29:445v448. Baldwin, H. I., and D. Murphy, 1956. Rocky Mountains Douglas— fir succeeds in New Hampshire. Fox Forest Note 67: pr. Byrnes, W. R., H. D. Gerhold, and W. C. Bramble, 1958. Douglas«fir varietal tests for Christmas tree plantations in Pennsylvania. Progra. Rep. Penn. Agr. Exp. Sta. No. 198. 6 pp. Ching, K. K., and D. Bever, 1960. Provenance study of Douglas-fir in the Pacific Northwest region. 1. Nursery performance. Silvae Genetica 9:11—17. Cleary, B. D., and R. H. Waring, 1965. Seasonal and diurnal variation in the international moisture stress of mature Douglas-fir in the Siskiyou Mountains. Proc. of Western For. Gen. Associ., (3rd session) abstract. Ferrell, W. K., and E. S. Woodward, 1966. Effects of seed origin on drought resistance of Douglas-fir. Ecology 47:499—503. Fuchigami, L. H., D. R. evert, and C. J. Weiser, 1971. A translocatable hardiness promoter. Plant Physiol. 47: 164-167. Gary, H. L., 1971. Seasonal and diurnal changes in moisture content and water deficits of Engelmann spruce needles. Bot. Gaz. 132:327—333. Gerhold, H. D., 1965. Douglas—fir growth and winter injury demonstration. Thirteenth No. East For. Tree Impr. Conf., Am. Christ. Tree Growers Jour.:Nov. fl_, 1966. Growth and winter injury in Douglas-fir in a 3 year provenance demonstration. Northeast For. Tree Impr. Conf. Proc. 13:50-52. Gibbs, R. D., 1958. Patterns in the seasonal water content of trees. In K. V. Thimann (ed.), The Physiology of Forest Trees, Ronald Press, New York: 43-69. 81 82 Hanover, J. W., 1963. Comparative biochemistry and physiology of western white pine resistant and susceptible to in- fection by the blister rust fungus, Washington State University, Ph.D. thesis: 173 pp. , and D. A. Reicosky, 1971. Surface wax deposits on foliage of Picea pungens and other conifers. Amer. Jour. of Botany 58:681-687. Heit, C. E., 1968. Prepagation from seed, part 17. Testing ' and growing Douglas—fir from different sources. Amer. Nurseryman, Nov. 15 issue. Irgens-Moller, Helge, 1957. Ecotypic response to temperature and photo-period in Douglas-fir. For. Sci. 3:78-83. , 1958. Genotypic variation in the time of cessation of height growth in Douglas—fir. For. Sci. 4: 325-330. Jameson, D. A., 1966. Diurnal and seasonal fluctuations in moisture content of pinjon pine and juniper. U. S. For. Ser., Rocky Mtn. For. Range Exp. Sta. Res. Note RM-67: 7 pp. Jones, J. R., 1972. Moisture stresses in Arizona mixed conifer seedlings. U.S.D.A. For. Serv. Res. Paper RM-86: 10 pp. Korstian, C. F., 1933. Physiochemical prOperties of leaves and sap as indices of the water relations of forest trees. Union Intern. Inst. Recherles For., Congress of Nancy, Comptes Rendus: 312-325. Kozlowski, T. T. and A. E. Peterson, 1960. Variations in moisture content of dormant buds, For. Sci. 6. and T. J. Clausen, 1965. Changes in moisture content and dry weights of buds and leaves of forest trees. Bot. Gaz. 126:20-26. Kramer, P. J., 1949. Plant and Soil Relationships. McGraw- Hill Publ. Co., New York: 274-278. , and T. T. Kozlowski, 1960. Physiology of Trees. McGraw-Hill Publ. Co., New York: 342-367. Kriebel, H. B., 1957. Patterns of genetic variation in sugar maple. Ohio Agr. Exp. Sta. Res. Bull. # 791:17-22. , 1963. Selection for drought resistance in sugar maple. WOrld Consultation on For. Gen. and Tree Impr., Doc. # 319:5 pp. 83 Kung, F. H. and J. W. Wright, 1972. Parallel and divergent evolution in Rocky Mountain Trees. Silvae Genetica 21: 65-72. Langlet, 0., 1936. Studier overtallens fysiologiska variabilitet och dess samband med klimatet. Ett bidrag till kannedomen om tallens ekotyper. Stockholm: Statens Skogsforsoksanstatt Meddelanden, no. 29:219-420. Levitt, J., 1956. The Hardiness of Plants. Academic Press, Inc., New York: 278 pp. , 1962. "A sulfhydryl-disulfide hypothesis of frost injury and resistance in plants." Jour. of Theoretical Biology 3:355-391. Meuli, L. F. and H. L. Shirley, 1937. The effect of seed origin on drought resistance of green ash in the prarie plains states. Journ. of For. # 35:1060-1062. Morris, W. G., R. R. Silen and H. Irgens-Moller, 1957. Consistency of bud bursting in Douglas-fir. Jour. of For. 55:208-219. Munger, T. T. and W. G. Morris, 1936. Growth of Douglas-fir trees of known seed source. U.S.D.A., Tech. Bull. 537: 40 pp. , 1942. Further data on the growth of Douglas- fir trees of known seed source. Pac. Northwest For. Exp. Sta., April 10:12 pp. Olien, C. R., 1967. "Freezing stresses and survival." Annual Rev. of Plant Physiol. 18:387-408. - Parker, J., 1951. Moisture retention in leaves of conifers of the northern Rocky Mountains. Bot. Gaz. 113:210-216. , 1954. Available water in stems of some Rocky Mountain conifers. Bot. Gaz. 115:380-385. Palpant, E. H., 1973. Christmas tree desiccation - Is there a solution? Am. Christmas Tree Grow. Jour. 37:11-12. Pharis, R. P. and W. K. Ferrell, 1966. Differences in drought resistance between coastal and inland sources of Douglas- fir. Canad. Jour. of Bot. 44:1651-1659. Pharis, R. P., 1967. Seasonal fluctuations in the foliage moisture content of well-watered conifers.‘ Bot. Gaz. 128:179-185. 84 Philpot, C. W., 1963. The moisture content of Ponderosa pine and whiteleaf manzanita foliage in the central Sierra Nevada's. U.S. For. Ser. Res. Note PSW-39, Pacific 50. West For. and Range Exp. Sta.: 7 pp. Reicosky, D. A., 1973. Developmental and physiological as- pects of surface waxes of blue spruce. Michigan State University, Ph.D. thesis: 84 pp. Saetersdal, L. S., 1963. The rate of drying of young excised plants of various provenances of Norway spruce and Douglas—fir. Meddelser fra Vestlandets Forstlige Forseksstasjon # 38: 88 pp. Schfitt, P. , and R. J. Hoff, 1969. Foliage dry matter of Pinus monticola its variability with environment and blister rust resistance. U. S. D. A. For. Ser. Res. Note INT-102:6 pp. Shirley, H. L. and Meuli, L. J., 1939. Influence of moisture supply on drought resistance of conifers. Jour. of Agr. Res. 59: 1-21. Silen, R. R., 1965. A 50 year racial study of Douglas-fir in Western Oregon and Washington. Proceed. of West. For. Gen. Assoc., (lst session). Sweet, G. B., 1965. Provenance differences in Pacific Coast Douglas-fir. Silvae Genet. 14:46-56. United States Forest Service, 1965. Silvics of forest trees of the United States. U.S. Dept. of Agr. Handbook 271: 782 pp. Weiser, C. J., 1970. Cold resistance and acclimation in woody plants. Science, 169:1269-1278. Wharton, D., 1974. Douglas-fir: Our most valuable tree. Am. Forests, 80:30-34. Wright, J. W., F. H. Kung, R. A. Read, W. A. Lemmien, and J. N. Bright, 1971. Genetic variation in Rocky Mountain Douglas-fir. Silvae Genetica 20: 54- 60. Wright, J. W., 1962. Genetics of forest tree improvement. FAO Forestry and forest products study # 16:399 pp. Wright, J. W., R. A. Read, F. H. Kung, J. N. Bright, and W. A. Lemmien, 1970. Better Douglas-fir for Eastern United States. Am. Christmas Tree Gro. Jour., 14:23-28. Zaerr, J. B., 1971. Moisture stress and stem diameter in young Douglas-fir. For. Sci., 17. 85 Zavithovski, J. and W. K. Ferrell, upon rates of photosynthesis, 1968. Effect of drought respiration, and trans- piration of seedlings of two ecotypes of Douglas-fir. Bot. Gaz. 129:346-350. APPENDIX APPENDIX Al SUGAR AND COLD RESISTANCE During freezing water is removed from plant cells, resulting in high water stress in the plant. Levitt (1959) has prOposed that high water stress has an effect upon the sulfhydryl groups of protein. As the layer of water around the protein molecule becomes thinner, the sulfhydryl groups begin to contact each other in adjacent proteins (Figure 7). Upon oxidation the hydrogen is removed and the disulfide linkages are formed. After rehydration, the disulfide link- ages hold proteins tOgether in such a way that the developing water layer results in distortions of the protein molecules. Resistance to cold damage is related to an inhibition of these intermolecular disulfide linkages which form between the proteins and ultimately distort them. Heber and Ernst have found that in cold resistant herbaceous plants sugar replace water in forming a protective shell around the protein molecules. This, if it were true for trees, could account for the increase in sugar content at the expense of water during the colder months of the year. Charles Olien (1967) studied the interference of freezing caused by large water soluble polysaccharides polymers ex- tracted from the cell walls of hardened plants. These sub- stances, which contained large amounts of sugar (xylose and RE 87 Figure 7. A schematic illustration of Levitt's sulfhydryl- disulfide hypothesis of frost injury and resistance in plants. (Based upon J. Levitt, 1962). Unhozen prolmn WAN?! helix. between proteins SH HS '0‘ Wale! Removed by Freezing: SE; Wine! Returned at Thawmg; ES 8:0) {05 .‘O'J 88 arabinose), interfered with freezing by competing with water molecules for sites in the ice lattice at the liquid inter- face. The result was that they tended to stop crystal growth, causing an imperfect ice mass to form. Only polymers from cold hardy plants or plant varieties resulted in imperfect crystal formation in the plant. In attempting to explain cold resistance much has been 1 fl learned by studying the plants acclimation to cold. Weiser :9 (1970) has found that the leaves are the site of perception of short days which initiates the first stage of acclimation. He proposes that short day induced leaves are the source of a translocatable hardiness promoting factor which moves from the leaves to the stems. The substance of the hardiness promoter is not known; however, it is believed to be either a growth inhibitor, a sugar, or a regulatory hormone. In work done by Fuchigami at 31. (1970), it was discovered that abscissic acid (a growth inhibitor) and gibberelin (a growth hormone) did not induce or enhance acclimation of different races of red-osier dogwood (Cornus stolonifera). Weiser (1970) has compiled considerable circumstantial evidence which suggests that some basic level of sugar is probably required for acclimation: (1) leaf discs of cabbage and leaves of gardenia increased in frost resistance when infused with sugar, (2) plants cannot acclimate when they are depleted in photosynthetic substrate, (3) sugar protects the enzyme systems associated with oxidative phosphorylation in isolated spinach chloroplasts subjected to freezing, (4) considerable data 89 indicates that starch is converted to sugar in plant tissues exposed to low temperatures. Although this evidence seems considerable, no experimental evidence shows that feeding sugar to woody plant tissue has increased their resistance. Attempts to do this with red-osier dogwood have been un- successful. Langlet's (1936) dry weight and sugar content data offer relevant information to the problem of cold resistance in plants. His data show significant correlations indicating that sugar is in some way related to reinforcing the plant for the cold of winter. APPENDIX A2 STATISTICAL FORMULAS 1. Analysis of Variance (Machine Formulas): Source Degrees of freedom Sums of Squares Mean Squares Seedlot n-l SS/d.f. Replicate r-l SS/d.f. Error (n-l)(r—1) SS/d.f. Total nr-l Seedlot # Replicate # gum n = # of seedlots |t-‘ N [w H II # of replicates 501 A A A B A = plot means 502 A A A B B = seedlot sum 503 A A A B C = replicate sum 504 A A A B D = grand total 505 sea a C1C2C3 D Sums of Squares (SS): 2 Seedlot SS = l/r B2 - 12L— nr 2 Total SS = A2 - 12L— nr 2 . _ 1 c2 _ (D) Replicate SS - H Ti?— Error SS = Total SS - (Seedlot SS + Replicate SS) 90 91 2. Least Significant Difference (LSD): v = error d.f. LSD = todV) 2 (error mean square) n = # of plots per race “ t = t value from table o<= deSired level of significance 3. T-test: for even samples Y = sample mean t = (Y1 - Y2) - (ul - “2) u = parametric mean (0) ‘Jl/n (812 + S22 52= variance of sample mean n = number of observations for uneven samples 1 - Y2) - (ul - uz) (Y t: \J (n - l)s 2 + (n - l) s 2 n + n l l 2 2 1 2 nl + n2 ’ 2 nlnz d.f.= n + n - 2 4. Spearman's coefficient of rank correlation: ranks of the variables n 2 - 6 : (R1 R2) R n n ( number of observations - 1) n 5. Chi-square test (two by two table): 2 (Observed frequencyl - expected frequencyl) x: Expected frequency + (observed frequency2 - expected frequencyz) Expected frequency degrees of freedom = l APPENDIX A3 ANALYSIS OF VARIANCE Table 1A. Analysis of variance of current—years and year—old foliage moisture content data from plantations MSFGP—16—65107 (Kellogg Forest), MSFGP-5-65 (Camp Kett), MSFGP-9—66 (Kellogg Forest), and MSFGP-4—67 (East Lansing). Levels of significance indicated by asterisks.A Kellogg Plantation - MSFGP 16-65167 Current—Growing Foliage Yr-old Foliage Source of Variation d: MS F MS F Seedlot 66 2553.4 2.63** 467.50 2.76*** between region 8 10,479 7.18*** 1007 . 2.56** within region 58 1,460 1.51** 393 2.32*** Replicate 4 18,113 18.68*** 10,049 59.37*** Error 268 969.7 169.25 Total 339 339 Camp Kett Plantation - MSFGP 5-65 Current-Growing Foliage Source of Variation gf MS E Seedlot 35 679.4 1.64* between race 6 1978.5 4.82*** within race 29 410.7 .99 C)? 93 Camp Kett Plantation (continued) 21.: 131.8. a Replicate 4 2230 5.37*f Error 113 414.9 Total 152 Kellogg Plantation - MSFGP-9-66 & 4-67 (combined) Year-Old Foliage Source of Variation gf MS F Seedlot 21 517.7 2.06* between race 6 83.19 .12 within race 15 691.57 2.35** Replicate 4 1381.75 5.51** Error 59 251.44 Total 85 * significant at 5% level ** significant at 1% level *** better than 1% level Table 2A. 94 Analysis of variance of height and foliage color data from plantation MSFGP 16—65/67. Levels of significance indicated by asterisks. Height Foliage Color Source of Variation d: MS_ 3' MS 5 Seedlot 68 263.79 8.38*** 71.43 6.64*** between race 7 1902 24.95*** 594 34.62*** within race 61 75.8 2.41*** 16.05 1.49** Replicate 4 34.75 1.10 1.18 .11 Error 235 31.48 10.75 Total 307 95 Table 3A. Analysis of variance showing significance of year of planting at the Kellogg Forest (MSFGP 16-65/67) on current-growing and year-old foliage moisture content data and on height and foliage color. Differences in year of planting significantly affected height growth. Source of Variation Moisture Content Currenthears Foliage Year-Old Foliage 9.1: 14.5. F is F Seedlot 9 36.64 5.22* 2.69 1.12 Year of Planting l 11.77 1.68 .17 .07 Error 9 7.02 2.31 Total 19 Height and Color Height ___ Color Source of Variation df MS S MS S Seedlot 11 278.63 5.27** 63.18 ' 6.26** Year of planting 1 637.07 12.06** 30.0 2.97 Error 11 52.84 10.1 Total 23 96 Table 4A. Analysis of variance showing the effect of time of collection (within and between replicates) on current growing and year-old foliage moisture con- tent at plantation MSFGP 16—65/67 (Kellogg Forest). Current-Years Foliagg_ Year-Old Foliage Source of Variation SS MS F MS F Seedlot 76 Time effect between replicates 4 142.79 14.05** 93.76 53.58** within replicates 20 9.94 .94 2.43 1.36 Error 242 10.58 1.78 Total 342 97 Table 5A. Error mean squares for moisture content data from each Douglas-fir race at the Kellogg Forest plan- tation (MSFGP 16-65/67). (Both current-growing and year old foliage). High error mean squares for NOCOL and SOCOL (current growing foliage) is due to variability caused by frost damage. Error Mean Sggg SS Current Foliage Year-Old Foliage: ALB 6 1314 461 CMON 22 1036 417 NOROC 40 279 159 INEMP 70 427 239 CWASH 17 1592 175 NOCOL 37 2649 572 SOCOL 43 2262 361 ARINEM 33 1163 331 Total 268 98 oonm mm Hg ma ma ov 5H mm mm mm m I oomn szHmd oomm ma av mm ma on mm mm mm am am: oomn qouom ooom ma ov mm Ha mp ma mm .mm gm Hm: oomn A0002 ooom mm we ma ma oma ma Ho om ooH mm: oowm 2020 ooom on ma om gm oma om mo mm hm mm: oomm AszH ooom Hm ma mm mm oha mm mo mm moa Hm: oomm Oomoz oovm mm n ma we oom mm mm mm mm man oomm mmmzo ||.mueus llllllll amuou mo w uuuuuu Innn.ce llllllllllllllllll mo IIIIIIIIII Inn.umlul mcflsmcsm .>oz .msd has .Dwm assess .ummm mono .umz .oma.m30cm .meomum maso .smo mmonm>< mcfluso Hammsflmm amazed B omwuo>< mez Baez coeum>mam moom .Aemm-mmmnefl madman loco: .Hoom .mmumum omUACD cumumms map mo moumeflao seepssoz .vgma ..m .umxmmv .cm>em mum moon newsmmamsoa HowumDCfl some we poemsooo mucofisoufl>sm on» How mmeflmwu oepmseao msoflnm> .wmsmu Hemummamsoo map usonmsounu COAuoHHm> oeumaflao .so manme mzmmefidm UHBQSHHU vd Nanmmmd HICHIGRN STQTE UNIV. LIBRRRIES U lllll H II II | 25 31293105 7160