a WWNW“WNWWlWWWW! 604 _{ I IUD 4 LIBRARY Michigan State University This is to certify that the thesis entitled GENETIC VARIATION OF STEM DIAMETER IN RED PINE (Pinus resinosa Ait.) IN MICHIGAN presented by EKO BHAKTI HARDIYANTO has been accepted towards fulfillment of the requirements for MS degree in _F_QRELSIRI 962% W Dr . DANIEL E.KEATHLEY Major professor Date W 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution lllllllllljllllllllllllllllllllllfillll 93 10744 99 4 hViESI.) RETURNING MATERIALS: Place in book drop to LIBRARIES remove this checkout from 4—3—— your record. FINES will be charged 1f book is returned after the date stamped below. GENETIC VARIATION OF STEM DIAMETER IN RED PINE (Pinus resinosa Ait.) IN MICHIGAN By: Eko Bhakti Hardiyanto A THESIS Submitted to Michigan State University in partial fulfillment of the requirement for the degree of MASTER OF SCIENCE Department of Forestry 1986 K ’(o '1'. TQLSYO ABSTRACT GENETIC VARIATION OF STEM DIAMETER IN RED PINE (Pinus resinosa Ait.) IN MICHIGAN By: Eko Bhakti Hardiyanto This study was undertaken to determine the genetic components of variation and to examine the potential for obtaining genetic gain in stem diameter in red pine (Pinus resinosa AitJ Red pine seeds were collected from 272 unselected trees, 58 from the Upper Peninsula and 214 from the Lower Peninsula of Michigan. In 1964, three-year-old stocks were used to establish four permanent plantations in Michigan. The plantations were arranged as randomized complete block design with four-tree row plots in one to eight replications. . The spacing was 8 x 8 feet (2.4 x 2.4 m). In 1984, trees at the Allegan and Crawford plantations were measured for stem diameter. The results of this study indicated that there were significant differences between plantations. There were also significant differences between the two seed collection regions. Seeds from the Lower Peninsula grew 6 to 9 76 faster than those from the Upper Peninsula. However, there were no significant differences among stands within regions. Differences among families within regions were significant. No genotype—plantation interaction was detected. The component of the total genetic variation attributable to regions, stands within regions and families within stands were 51.29, 16.01 and 32.69 %, respectively. Narrow-sense heritability of family means was found to be 0.227 i 0.031. Immediate gains in diameter growth rate could be realized by using seed from the best region. More genetic gains could be realized by selection the best families and the best individuals within the best families. The progeny test could be converted into a seedling seed orchard using selected families or these selected families could be used for grafted orchards and clonal forestry. ACKNOWLEDGEMENTS I wish to extend my sincere gratitude to my major professor, Dr. Daniel E. Keathley whose guidance and assistance were essential to the successful completion of this study. I wish also to express my appreciation to the other members of my guidance committee Dr. James W. Hanover and Dr. Thomas G. Isleib for their essential contributions. I owe a special debt of appreciation to my friend John Davis, Boen Purnama and R. Wasito for help in the data collection. I also wish to thank to Dr. Raymond Miller for help in providing computer programs for the statistical analysis. Finally, I am indebted to my parents for their support and encouragement throughout my academic career. ii TABLE OF CONTENTS Page List of Tables ..... ...... ............. ....... . ...... iv List of Figures ..... ........ . ............. . ...... .. v INTRODUCTION ............................... ....... . 1 REVIEW OF LITERATURE ..... .......................... 4 MATERIALS AND METHODS Seed Procurement ..... ...... . ...... . ...... ..... 13 Nursery Practice .... ........ .................. 13 Plantation Establishment ...................... 14 RESULTS AND DISCUSSION Test of Significance ....... . ..... ........ ..... 2O iiitfiiifil’t‘ugigweii. is??? TERRIER? iii. . .. 23 Components of Variance ........................ 24 Heritability and Genetic Gain Estimation ....... 26 Age—Age Correlations ........... .. ............. 28 Conclusions ....... ... ......................... 29 LIST OF REFERENCES ................................... 31 APPENDICES ........................................ 35 iii Table LIST OF TABLES Site conditions in two test plantations ......... Form of analysis of Variance ..................... Relative 20-year stem diameter of red pine grown from seed collected in 19 Michigan Counties at two location tests .................. ...... . ..... Analysis of variance of 20 year stem diameter data for red pine grown at two plantations ...... Variance component estimates of stem diameter in red pine 0.. ......... 00............OOOOOOOOOOO... Narrow-sense family heritability estimates in stem diameter from other conifers ......... ...... Expected genetic gains in stem diameter growth in red pine for different selection intensities Age-age correlations in stem diameter reported from other pines . ............ ... ........ .. ...... iv 25 26 27 LIST OF FIGURES Figure Page 1. Location of seed collection areas and of the four red pine plantations in Michigan ..... 15 INTRODUCTION Red pine (Pinus resinosa Ait.) is the primary conifer planted for reforestation in the north central United States. Its natural range extends from the northeastern coast to the Great Lakes region in the United States and across portions of southern Ontario, Quebec and New Brunswick in Canada. Twenty to twenty-five million red pine seedlings are planted annually in the north central United States for reforestation (Ager gt 3;}, 1985). In the state of Michigan about 10 million red pine seedlings were produced for reforestation in 1981. It is projected that over 15 million seedlings will be produced in 1986 (Levenson and Hanover, 1985). The popularity of red pine among foresters is due to its good form and rapid growth on well drained loam and loamy soils. The principal uses of red pine are for the the production of lumber and pulpwood (James 3.1.7.31” 1982) Red pine has little or no genetic variation with respect to a wide array of characters including growth rate, morphology, and wood density (Flower and Lester, 1970). Provenance test studies on red pine indicate that little to no significant differences exist in survival rate (Rudolf, 1947), growth rate (Wright gt a;., 1965; Lester and 2 Barr 1965), Phenology (Rudolf, 1954; Rehfeldt and Lester, 1966), photoperiodic response (Vaartaja, 1962), wood quality (Bees and Brown, 1954; Peterson, 1966; Ager et al., 1985), frequency of lammas growth (Lester and Rehfeldt, 1967), and foliage polyphenols (Thielges, 1972). In a progeny test with red pine in Michigan there were significant differences in height growth (1 % level) among families (Ya°.§£.§lr 1971). However, a similar study in Wisconsin by Ager gt §g3(1983) indicated that, although still significant at the 5 % level, the differences among families for height growth were smaller than reported by Yao EE.§l° (1971), and no significant differences in wood density could be detected. Studies based on inbred population (Flower, 1964 b, 1965) and the examination of allozyme variation (Flower and Moris, 1977) have confirmed that red pine is less variable genetically than other pines. Potential for genetic improvement is dependent upon the and type of genetic variation that is present in the species. The fact that red pine shows less genetic variation than most forest trees has lessened the effort, but not the need for a tree improvement program focused on red pine. The extensive red pine planting programs in the north central United States have led to the initiation of a tree improvement program for this species in that region. This study reports the results of individual measurements from two plantations established by Michigan State University in 1964 in Michigan. The objectives of this study were: (1) to analyze the genetic components stem diameter stem diameter variation in red pine and (2) to examine the potential for obtaining significant improvement in stem diameter in red pine. REVIEWOFIJTERATURE Work aimed at assessing the potential for improving growth rates and other characteristics in red pine through a breeding program was initiated by the Lake State Forest Experimental Station (now, North Central Forest Experimental Station) in 1928. This work began with the collection of seed from 37 different locations throughout the Lake States and New England. Seedlings from this seed were used to establish plantations in the Superior, Chippewa and Huron National Forests in 1951. In 1935 three additional plantations were established in the same areas using trees from 144 seed sources. The Allegheny Forest Experimental Station conducted a provenance test containing 50 different seed sources in the Kane Experimental Forest in 1957 (Hough, 1957; Rudolf, 1955). A plantation was established at Cass Lake, Minnesota in 1937 using 48 seed sources that wereleft over after the establishment of the 1951 and 1933 plantations (Rudolf, 1955; Buckman and Buchman, 1962). Due to a combination of drought, fire and other problems, among the 1951 and 1935 plantations, only those plantations in the Superior National Forest have survived (Rudolf, 1964; Hough, 1967). In 1947, Rudolf (1947) reported on the surviving 1931 5 red pine plantation in the Superior National Forest. At 16 years, his results indicated that seedlots from northern Minnesota and northern Wisconsin had the highest rate of survival, and had more rapid height and diameter growth rates. Seedlots from central Wisconsin, Michigan and New England had poor growth. These results were only based upon the relative performance of the seedlots, no statistical analysis was performed. Wright gt a}. (1958) reanalyzed the same experimental data statistically, and found no significant differences among the Lake States origins growing in the Superior National Forest. When analyzed at 25 years from seed, the 1953 red pine plantation yielded results that were similar to those of 1951 plantation reported by Rudolf (1947). When the seed sources were ranked based on cubic feet of volume per 100 trees planted, all of the sources in the top 205%were from localities in Minnesota, northern Wisconsin and the Upper Peninsula of Michigan. None of the sources from central Wisconsin, Michigan's Lower Peninsula or the northern states was included in this category (Rudolf, 1964). Based upon these results, Rudolf (1964) concluded that the local and near local sources performed best at this age, and those from a farther distance performed more poorly. Hough (1957) reported on the results of the 1937 plantation from the Kane Experimental Forest. At five years of age, survival and height growth did not differ significantly among seed sources. Differences in height 6 growth among seed sources were, however, found to be significant at 10 years. Hough (1967) reanalyzed this plantation at 20 and 25 years. In both cases his results indicated that there were small but significant differences for height growth among seed sources. Seedlots grown from seed collected from southern latitudes had better height growth than those from northern latitudes. The best seedlots were from the Lower Peninsula of Michigan and eastern Wisconsin. Peterson (1966) analyzed the 1957 plantation in the Kane Experimental Forest at 27 years of age. He analyzed 10 of the 50 sources, and found that differences among seed sources were highly significant for increment width of stem diameter and wood specific gravity. Rees and Brown (1954) measured and analyzed the 1937 plantation in the Cass Lake Forest, Minnesota at 17 years. They analyzed 19 of the 48 seed sources. The results showed that the following traits were not significantly affected by seed source: percentage of summer wood, average diameter inside the bark at 82 inches (2.10 m) above ground, height growth and volume index. The same experiment was reanalyzed by Buckman and Buchman (1962) at 27 years of age. They found that there were no significant differences in average ‘ tree height between the eight regional groupings. They concluded that red pine exhibited less racial variation in height growth than did other pines. In 1949, the University of Wisconsin in cooperation 7 with the Wisconsin Conservation Department initiated a tree improvement program for red pine. The program was started by collecting seed from 72 individual trees in 10 locations in Wisconsin and Canada. Seedlings from this collection were used to establish two plantations in 1952 and four plantations in 1954 using lattice designs (Lester and Barr, 1965). The 1952 and 1954 plantations were measured and analyzed by Lester and Barr (1965) at the ages of nine and 11 years, respectively. The results showed significant family effects for the following traits: height growth, stem—diameter, volume, and and mortality; They indicated that it would be possible to attain genetic gains for growth rate, but selection would have to be based upon a progeny test with very high precision. In 1957, a provenance test with red pine was established in Wisconsin. This test contained 18 seedlots collected from Canada. At eight years of age, shoot elongation was measured andanalyzed. The results showed that there were significant differences in the following characteristics associated with shoot elongation: total height, total elongation, bud length and the termination, duration and growth rate of elongation. Differences in the date of the initiation of elongation, however, were not significant (Rehfeldt and Lester, 1966) Wright gt §l° (1963) published a provenance test study on red pine using nursery data involving 77 different seed 8 sources. At three years from seed, the results for height growth indicated the presence of significant differences among progenies, both within and between regions. They also reported that there were no significant differences among progenies forother traits such as foliage color, foliage length, bud type and hardiness. These differences in height growth, however, were relatively small in comparison to those found in other pines. In the provenance test of red pine studied by Wright gt 3}. (1963), the tallest seedlot grew two times as fast as the slowest one. In comparison, the corresponding figure at the same age in scotch pine (Pinus sylvestris) was six times (Wright and Bull, 1963), and three times in jack pine (Pinus banksiana) (Canavera, 1969)- The above planting stocks studied by Wright gt El' (1963) with additional 14 seed sources were tested in eight locations in the north central states in 1963. The results of this experiment were reported by Wright et_ag. (1972). For height growth, seedlots from Michigan's Lower Peninsula were found to grow the fastest at all sites except one. On average they were 8 % the all-plantation average at 11 years of age. Seedlots from New Brunswick, Manitoba, and western Ontario grew the slowest at all sites. Their average growth was 8 % less than the all—plantation average. Wright 33 El. (1972) also observed that there were significant differences in height growth among the regions of seed collection. In 1964, Michigan State University established 8 an open—pollinated progeny test at four locations in Michigan. This test contains 272 seedlots from unselected individual trees from the Upper and Lower Peninsula of Michigan. Yao gt El' (1971) reported statistically significant differences among the offspring of different stands in the same peninsula and between the progeny of trees in the same stand for height growth. The experiment also showed the presence of genotype-enviroment interactions for height growth. In the Lower Peninsula plantations, trees grown from seed collected in the Lower Peninsula were 10 % taller than trees grown from seed collected in the Upper Peninsula. In the Upper Peninsula plantations, however, trees grown from seed collected in the Lower Peninsula were only 5 % taller than trees grown from seed collected in the Upper Peninsula. The narrow-sense heritability of family means for height growth was found to be 0.204 for the Lower Peninsula data, while the corresponding heritability, calculated from the Upper Peninsula data was CL124. Based upon the retaining of 25 tallest families in the last thinning, Yao gt El' (1971) estimated that the genetic gains for height growth were 3.6 and 2.5 % for the Upper Penisula and Lower Peninsula plantations, respectively: They recommended that seed should be used where it was produced. Steiner (1979) studied the Kellogg Forest plantation of the provenance test described by Wright 23.2lf (1972) for ‘ bud—bursttiming. He found that there were no significant lO differences among seed sources. Steiner (1979) also analyzed similar studies for seven other north-temperate pines at the same location test. All seven species showed significant differences in bud-burst timing among seed sources. In 1970, the University of Wisconsin conducted a similar open-pollinated progeny test with red pine using 310 seedlots from natural stands throughout Wisconsin. The experiment was established at three loCations in Wisconsin. Height growth and wood density were measured and analyzed in these plantations at 13 years of age by Ager gt El' (1983). The analysis of the height growth data indicated that significant differences were present among families and test locations. Significant family-plantation interactions were detected. Differences among families within stands accounted for 88 % of the total genetic variation, while stands within regions and regions accounted for 12 and 0 %, respectively. Narrow-sense heritability of family means for height growth were calculated to be between 0.40 and 0.50. The genetic gains were predicted to be 5 to 4 9% for height growth and 9 to 11 % for stem volume. Attempts have been made to increase the amount of genetic variation in red pine through interspecific hybridization. Thus far the results have not been promising due to strong interspecific barriers. Wright and his coworkers made 55 species crosses of the hard pines, series Sylvestres during the period 1948 to 1956. Thirty-one of ll the species tested failed to cross to other species. One of those unsuccessful species was red pine (Wright and Gabriel, 1958)- Many attempts have been made to cross red pine with Austrian pine, since the latter species has been successfully crossed with several other pines. For example, Wright and his coworkers pollinated more than 300 female strobili in an attempt to make this cross. Both species were used as the female parent. All the crosses failed (Wright and Gabriel, 1958). At the Institute of Forest Genetics in California, the cross between red pine and Austrian pine was attempted using more than 500 strobili from 50 different trees. The results have not been successful so far (Critchfield, 1965). In Canada, Flower (1964) reported a number of crosses between red pine and Austrian pine using more than 700 female strobili from 24 different trees. None of those crosses was successful. Out all of the attempts at hybridization, only one cross, Pinus nigra var. austriaca x P; resinosa) made at the Institute of Forest Genetics in 1955 yielded interspecific hybrids (Critchfield, 1963; Flower and Lester, 1970). The hybrids are intermediate between their parents in most characteristics, such as size of conelet and cone, flowering time, leaf dimension and leaf anatomy but they exceed either Parents in height growths Morris 33 El‘ (1980) used isozyme variation to analyze the genotype of the putative hybrids at the Institute of 12 Forest Genetics. The results of this study indicated that they were not red pine hybrids, but rather hybrids between Austrian pine and other unidentified Species. Moulalis gt Ei' (1976) obtained 25 putative hybrids of Pinus nigra x P; resinosa and 21 of P; heldreichii x P; resinosa. However, the authenticity of these hybrids has not been verified yet. MATERIALS AND METHODS Seed Procurement.- Seed was collected in 1960 from natural stands of red pine in Michigan. The collection effort was coordinated by JQW. Wright and W.I. Bull of Michigan State University. The collections were made by personnel of the United States Forest Service and the Division of Forestry, Michigan Department of Natural Resources. Seeds were collected from and maintained in individual tree seedlots. Seed collections were made from unselected 272 trees, 58 from the Upper Peninsula and 214 from the Lower Peninsula of Michigan (Figure 1). Seeds were accompanied by data on the location, relative height and stem diameter, and stem form of the parent trees. Nursery Practice.— All seeds were sown in the Michigan State University Experimental Nursery in East Lansing, Michigan in 1961. The seedlots were sown in a randomized complete block design with each seedlot replicated four times. Nursery plots were four feet long and six inches apart. There was an average density of 50 seedlings per square foot in the plots. Two years after sowing the seedlings were transplanted using the same design. The average density of the transplants was 10 seedlings per square foot. l4 Plantation Establishment.— In 1964, four permanent plantations (Figure 1) were established using 2-1 planting stocks. The experiment used a randomized complete block design with four-tree row plots and spacing of 8 x 8 feet (2.4 x 2.4 m). (Due to the low number of seedlings in some of the seedlots, seedlots are not represented in all locations or in the full number of blocks in each plantationr Further details concerning the establishment of the individual plantations are as follows: MSFGP-1/2/3/-64: Planted 4/15/64 at Allegan County : eight replicates, randomized complete block designs, four-tree row plots; site level, sandy soil, sparce weed cover; no herbicide treatement before planting. MSFGP-5/6/7-64: Planted 5/6/64 at Crawford County: eight replicates, randomized complete block design, four-tree plots; site nearly level, a loamy sand with a dense quack sod; plowed and disked before planting, treated with a simazin and amino triazole spray after planting. MSFGP-8-64: Planted 5/11/64 at Delta County: three replicates, randomized complete block design, four-tree row plots; sandy soil with dense sod; plowed disked, and treated with aldrin before planting. Figure 1. 15 Location of seed collection areas (circles) and of the four red pine plantations in Michigan. A— Allegan C— Crawford, D— Delta, G- Gogebic. l—vw-T‘zf-t'!‘ H O‘ MSFGP-9—64: Planted 5/13/64 at Gogebic County: one replicate, randomized complete block design, four—tree row plots; site rough; furrowed prior to planting. This study only deals with the Allegan and Crawford County plantations. Further details about site conditions in those two plantations are depicted in Table 1. Table 1. Site Conditions in two test plantations 1) Name of Mean of temperature Ann.prec. Soil plantation Jan. July Ann. 0 o o ( C) ( C) ( C) (M) Allegan — 5.32 21.78 8.79 942 Sandy Crawford - 8.68 19.60 6.27 741 Sandy loam 1) Climate data: thirty years average (1954 — 1983), National Climatic Center, United States Department of Commerce. Stem diameter measurements were taken in July and August 1984 for the Allegan and Crawford plantations, respectivelyuThe measurements were made on individual trees at breast height to the nearest millimeter using a diameter tape. Four replicates were measured for each plantation. An analysis of variance was conducted according to the model presented in Table 2. Plantations, replicates, stands within regions and families within stands were considered to be random. Regions was considered to be fixed. Plot mean 17 Table 2. Form of analysis of variance Source of DF MS EMS variation Rep.(Plant.) (bp - p) MSB ___________ 2 2 2 Plantations (p — 1) MSP‘ aE+bOFP+meT<—— 2 2 2 Families (m - 1) . MSF 10E+quP+prF 2 2 2 Regions (r - 1) MSR OE+bOF(R)P+bgORP+ 2 2 +prF(R)+bpgoR 2 2 2 Fam.(Reg.) (m — r) MSF(R) OE+bOF(R)P+prF(R)G—1 2 2 2 Stands(Reg.) (n - r) MSS(R) OE+bOF(S)P+beS(R)P+ 2 2 +prF(S)+bpfOS(R) 2 2 2 Fam.(Stands) (m — n) MSF(S) OE+bOF(S)P+prF(S)fi 2 2 Fam.x Plant. (m—1)(p—1) MSFP OE+bOFP 2 2 2 Res-x Plant- (r-1)(p-1) MSRP aE+bOF(R)P+bg‘JRP€ l 2 2 Fam.(Reg.)x (m-r)(p-1) MSF(R)P OE+bOF(R)P‘af » Plant. 2 2 2 Stands(Reg.)x (n—r)(p-1) MSS(R)P aE+bOF(S)P+beS(R)Pi1{-l Plant. 2 2 L Fam.(Stands)x (m-n)(p-1) MSF(S)P OE+bOF(S)P4 Plant. 2 ‘— Error by subtraction MSE OE Total (mbp—1-missing plots) where number of replicates number of plantations total number of families tested total number of stands tested total number of regions tested harmonic mean number of families/stand harmonic mean number of families/region Obi—bdDBWO‘ HIIH uu "1| I V 18 values were used as entries in the analysis of variance. Hewever, attempts have been made to use individual measurements as entries in the data analysis. Due to the difficulty in finding an appropriate statistical package program which is able to handle the available data, the efforts have not succeeded so far. F tests were performed as indicated by arrows in Table 2, except that synthetic F tests and degrees of freedom were computed by the method of Cochran (1951) to test the effect of stands within regions and regions. The estimate of heritability on family mean basis was calculated according to Wright (1976): 2 2 OF hf = 2 2 2 OF + OFP/p + GE/bp 2 2 where hf = narrow—sense family heritability, OF 2 variance component of family) OFP = variance component of family— plantation interactions, CE = variance component of error, p: numberof plantations,and b: numberof replicates per plantation. The standard error of the family heritability was calculated as 40t from the intraclass correlation equation according to Becker (1984) : 2 2 2 2(n.- (1 — t) [1 + (k —1)t] k (n.— S)(S — 1) S—1 no 2 where t: 1/4l1 for half—sib families, S==number of families, n = number of plots for ith family and n.= total number of plots. Estimates of genetic gains in stem diameter expected from thinning were calculated using the formula : Gs : h: i Op where Gs = genetic gain, hf = heritability, i = selection intensity, and Up = phenotypic standard deviation of family means. The association between stem diameter of stand means and north latitudes was determined using simple linear regression. Finally, correlations between ages of 15 and 20 years for stem diameter of family means were calculated. RESULTS AND DISCUSSION Test gt Significance Based upon the grouping of families according to stands and origins (Table 3), an analysis of variance was conducted. A test of homogeneity indicated that the variance among families across the two regions was homogeneous. The analysis of variance was then performed as shown in Table 4. The results indicated that there were highly significant +ifferences between the Allegan and Crawford plantations. Trees in the Allegan plantations grew more slowly than those in the Crawford plantation (Table 3). This was due to the lower soil fertility at the Allegan site, since other site conditions were more favorable for growth than bhose found in the Crawford plantation (Table 1). The differences due to regions of seed collection were significant (Table 4). Trees from seed sources in the Lower Peninsula had stem diameters that were 6 to 9 96 larger than those from seed sources in the Upper Peninsula (Table 3). Similar results for height growth were reported by Yao gt El: (1971) from the same experiment. The red pine forest of the Lower Peninsula is separated from that of forest in the Upper Peninsula by the Straits of Mackinac which form a 'natural barrier to crossing so that natural selection could 2O 21 Table 5. Relative 20-year stem diameter of red pine grown from seed collected in 19 Michigan Counties at two location tests. County of North Relative stem diameter when planted origin Lat. Allegan Crawford (degree) (percent of plantation mean) Upper Peninsula Schoolcraft 45.9 91 96 Iron 46.0 95 95 Luce 45.7 105 98 Chippewa 46.3 85 89 Average 95 95 Lower Peninsula Grand Traverse 44.5 102 101 Alpena 44.2 105 98 Otsego 45.0 102 98 Cheboygan 45.5 106 102 Cheboygan 45.2 102 103 Cheyboygan 45.5 101 107 Ogemaw 44.6 105 103 Crawford 44-6 95 95 Crawford 44.6 105 107 Crawford 44.6 97 102 Alcona 44.6 105 100 Newagyo 43.7 102 101 Oscoda 44.6 99 101 Maniste 44.5 101 101 Wexford 44.3 100 100 Average 'TO2 TOT Actual mean stem diameter (mm) 149 166 22 Table 4. Analysis of variance of 20-year stem diameter data for red pine grown at two plantations Source of df MS F , EMS 1/ variation value Reps(Plant.) 6 5595.84 2 ' 2 2 Plantations 1 82253.086 168.90** OE+4OFP+6520P Families 162 751.75 2 2 2 Regions 1 15808.09 36.85* OE+4OF(R)P+4gORP+ 2 2 80F(R)+8gOR 2 2 2 Fam.(Reg.) 161 638.11 1.31* OE+4OF(R)P+80F(R) 2 2 2 StandS(Reg.) 17 1165.19 1.95ns OE+4OF(S)P+4fOS(R)P+ 2 2 80F(S)+8fOS(R) 2 2 2 Earn. (Stands) 144 576.12 1.19ns OE+4OF(S)P+80F(S) Fam.xPlant. 162 576.97 2 2 2 Reg.xPlant. 1 279.12 0.57ns OE+4OF(R)P+4gORP 2 2 Fam.(Reg.)x 161 488.26 0.95ns OE+4OF(R)P Plant. 2 2 2 StandS(Reg.)x 17 506.095 1.04ns OE+4OF(S)P+4fOS(R)P Plant. ' 2 2 Fam.(Stands)x 144 486.150 0.95ns OE+4OF(S)P Plant. 2 Error 849 511.60 ,OE Total 1181 *, ** = significant at 5 and 1 percent level, respectively 1/ f, g = harmonic mean number of familes/stand, families/ region were 12.87, and 108.91, respectively. 25 result in the development of distinct races (Wright 23.2l” 1972). In contrast, there were no significant differences among progenies of trees from different stands within regions or among progenies of trees from different families within stands. .Differences among families were significant, when the effect of stands was confounded into the family effect. This indicates that differences among families within regions exist. No genotype-environment interaction was detected. Trees grown from seed collected from different regions, stands or families grew at the same relative rates in in the Allegan and Crawford plantations. Somewhat different results in height growth were reported by Yao gt gt. (1971) from the same experiment. In that study, differences among families within stands, among stands within regions, as well as family-plantation interactions in height growth were significant. The genetic variation in stem diameter reported here was in agreement with that observed by Lester and Barr (1965) from a similar study in Wisconsin. However, they were able to detect differences among families when the data were analyzed in lattice designs. When the same data were analyzed in a randomized complete block design the differences among families in stem diameter growth were not significant. Association between Stem Diameter Growth and North Latitudes Correlations between stem diameter growth and north 24 latitudes were calculated for each plantation using the data in Table 3. The results indicated that the association between stem diameter growth and north latitudes was significant at the 1 % level for both plantations. The coefficients of correlation (r) were calculated to be - 0.57 and -CL59 for the Allegan and Crawford plantations, respectively. Trees grown from seed collected at more southern latitudes had greater stem diameter growth than those from northern latitudes. These results were in accordance with those reported by Hough (1967) from a progeny test with red pine in Wisconsin, but he did not mention the magnitude of the coefficient of correlation. Components gt Variance The amount of genetic variation in stem diameter that can be attributed to regions, stands within regions, and families within stands was estimated using components of variance derived from the analysis of variance in Table 4. Variance components are expressed as a percentage of the total genetic variation (Table 5). Knowledge of the amount of variation associated within each component indicates in which level selection will achieve the greatest genetic gain per generation. The total genetic variation attributable to regions represented the biggest portion of the total genetic variation in stem diameter growth (51.29 %). iFamilies within stands and stands within regions accounted for 52.69 and 16.01 % of the total genetic variation, respectively. 25 Table 5. Variance component estimates of stem diameter growth in red pine Component % of Sources estimate total Regions 17.65 51.29 Stands (Regions) 5.51 16JN Families (Stands) 11.25 32.69 Ager _e_t_ g. (1983) reported different results for height growth of red pine in Wisconsin. Regions of seed collection did not contribute to the total genetic variation, while the components of variance attributable to families and stands were 88.5 and 11.7 %, respectively. In that study the absence of region contribution to the total genetic variation was apparently due to the limited sample of regions. The distribution of the genetic variation among regions, stands, and families has an important implication from a practical standpoint. This information can be valuable indirecting further improvement work for stem diameter with red pine in Michigan. Over one-half of the potential genetic gain in stem diameter growth could be realized by selection from the best region. The Lower Peninsula is the best seed source for collecting seeds used for reforestation in this peninsula. Further gains could be realized by selection from the best families within the best regions. From an economic vieWpoint selection between 26 regions is the most desirable because the expected gain can be realized immediately without waiting for the more expensive and time-consuming family selection. Heritability and Genetic Gain Estimation Narrow—sense heritability of family means in stem diameter were calculated using components of variance derived from the analysis of variance in Table 4. The family heritability was found to be 0.227 i 0.051. This family heritability is low in comparison to values for other conifers (Table 6). This heritability was similar to that value for height growth reported in red pine from the same plantations (Yao gt 223' 1971). Since there have been no other studies of this type reported in red pine, comparison within this species cannot be done. Table 6. Narrow—sense family heritability estimates in stem diameter from other conifers Age 2 Species (yr.) hf Reference E. white pine 8 0.85 desBordes and Thor (Pinus strobus LJ (1979) E. white pine 9 0.84 Mullins (1985) Caribbean pine 8 0.65 Ledig and Whitmore (Pinus caribaea Morelet) (1981) Jack pine 12 0.31 Ernst gt a_l. (1983) (Pinus banksiana Lamb.) White spruce 20 0.55 Merrill and Mohn (Picea glauca (Moench)Voss) (1985) 27 Estimates of genetic gains were calculated for several selection intensities expected from thinning among families (Table 7). The expected genetic gains in stem diameter were found to be small in comparison to those reported by Yao gt a}; (1971) for height growth from the same experiment. A genetic gain of 3u29 % would be expected if 1 % of the original families were retained. These expected genetic gains would likely be higher when within family selection was practiced. They might also have been higher, if a different experimental design had been used. Lester and Barr (1965) reported from a study on a red pine test that lattice designs had 111 to 126 % better precision than randomized complete block designs in detecting differences among families for stem diameter growth. However, they did not calculate the expected genetic gains for this trait in their study. Table 7. Genetic gains expected from thinning in red pine for different selection intensities Sg%ggtig§ % Gain in Volume (cubic meter/ha) stem diameter Before After Increase 25 % retained 1.66 185.63 194.86 9.23 10 % retained 2.30 185.63 198.47 12.80 1 % retained 3.29 185.65 204.26 18.63 2 1/ Cubic feet per tree = [(dbh./2)-1)] and assuming there are 1500 trees per hectar. 28 The concern for the effect of relatedness due to imposing high selectiOn intensities might be negligible, since inbreeding has little or no loss in seed production or progeny vigor in red red pine (Flower, 1965). gge-Age Correlations The only data on stem diameter available from the previous measurements were the 1979 data. These 1979 and 1984 data were used to calculate coefficients of correlation between ages for each plantation. The associations were highly significant. The phenotypic coefficients of correlation (r) were found to be 0.865 and 0.872 for the Allegan and Crawford plantation, respectively. Thus far, no other studies of this type have been done in red pine. Several studies, however, have been conducted in other species. For comparison, Table 8 presents age-age correlations in stem diameter from other species. 29 Table 8. Age-age correlations in stem diameter reported from other pines (Wakely, 1971) Measurement Species ages correlated r Slash pine 10, 30 0.75 0.95 (Pinus elliottii Engelm. var. elliotii) 15. 30 0.88 0.96 20, 30 0.97 0.98 Longleaf pine 10, 30 0.70 0.90 (Pinus palustris Mill.) 15. 30 0.80 0.96 20, 30 0.81 0.98 Loblolly pine 10, 50 0.74 0.88 (Pinus taeda L.) 15. 30 0.88 - 0.96 20, 30 0096 - 0097 Shortleaf pine 10, 50 0.67 - 0.76 (Pinus echinata Mill.) 15! 30 0087 - 0°90 Conclusions This study indicates that a significant amount of genetic variation in stem diameter exists in red pine in Michigan. The major components of this variation were found to be whether the seed was collected in the Upper or Lower Peninsula, the stand within that region where it was collected, and the family within the stand. The regions accounted for the biggest portion followed by families within stands and stands within regions. The Lower Peninsula of Michigan is the better region for seed 50 collection for use in the Lower Peninsula. Trees grown from seeds collected in more southern latitudes had greater stem diameter growth when tested in the Lower Peninsula of Michigan. The heritability estimate and genetic gains were low compared to those reported from other pines. More efficient experimental designs should be considered to increase the amount of genetic gain in progeny tests with red pine. Although the expected genetic gain is small, this gain will have a considerable impact in increasing wood production in Michigan, since this state has an extensive red pine planting program. More than 10 million red pine seedlings are planted annually in Michigan. The present results indicate that collecting seed from the best region would give immediate genetic gains in diameter growth. The combined family and within-family selection was probably the most promising approach to obtain more genetic gains for diameter growth in red pine. The progeny test then could be converted into a seedling seed orchard using the best families and the best individuals within the best families. However, it might be desirable to use the selected families for grafted orchards and clonal forestry. It was not possible to ascertain non-additive genetic variance from the present data, therefore the possibility to exploit the existance of non-additive variance should be considered in the improvement program with red pine in the future. LIST OF REFERENCES LIST OF REFERENCES Ager, A.,IL Guries, and C.IL Lee. 1983. Genetic gains from red pine seedling seed orchards. Proc. of the Twenty-eighth Northeast For. Tree Impr. Conf., Durham: 175-194. Becker, W. A. 1984. Manual of quantitative genetics. Academic Enterprises, Pulman, Washington. 4th ed. 186p. Buckman, R. E. and R. G. Buchman. 1962. Red pine plantation with 48 sources of seed shows little variation in total height at 27 years of age. USDA For. Serv., Lake States For. Expt. Sta. Tech. Note 616. Canavera, D. S. 1969. Geographic and stand variation in jack pine (Pinus banksiana Lamb.). PhD dissertation, Michigan State University, East Lansing. 100 p. Cochran, WkG. 1951. Testing a linear relation among variances. Biometrics 7: 17 - 52. Critchfield, W. B. 1963. The Austrian x red pine hybrid. Silvae Genetica 12: 187 — 192. desBordes, K. and E. Thor. 1979. Estimates of heritabilities gains from open-pollinated progeny test of eastern white pine. Proc. of the First North Central Tree Improvement Conference, Madison, Wisconsin: 44 - 53. Ernst, S. G., G. Howe, J. W. Hanover, and D. E. Keathley. 1985. Genetic variation and gain of specific gravity and woddy biomass in a jack pine half-sib progeny test in Michigan. Proc. of the Third North Central Tree Improvement Conference, Wooster, Ohio: 111 - 122. Flower, D. P. 1964 a. Hard pine breeding at the Southern Research Station, Maple, Ontario, 1962 and 1965. Proc. of the Ninth Meeting Comm. For. Tree Breeding in Canada, II: 53 - 37. ' 1964 b. Effects of inbreeding in red pine, Pinus resinosa Ait. Silvae Genetica 13: 170 - 177. 32 1965. Effect of inbreeding in red pine, Pinus resinosa Ait., II. Pollination studies. Silvae Genetica 14:.12 - 23. Flower, D. P. and D. T. Lester. 1970. Genetics of red pine USDA Forest Service Res. Pap. W0-8. 15 p. Flower, IMF. and R. W. Morris. 1977. Genetic diversity in red pine: evidence for low genic heterozygosity. Can. J. For. Res. 7: 543 — 347. Hough, A. F. 1957. The red pine provenance test on the Kane Experimental Forest. Proc. of the Fourth Northeastern Tree Impr. Conf.: 5 - 5. 1967. Twenty-five-year results of a red pine provenance study. Forest Science 15: 156 - 166. James, L. M., S. E. Heinen, D. D. Olson, and D.E. Chappelle. 1982. Timber products economy of Michigan. Mich. State Univ. Agric. Expt. Sta. Res. Rep. No. 446. 20 p. Ledig, F. T., and J. C. Whithmore. 1981. Heritability and genetic correlations of caribbean pine in Puorto Rico. Silvae Genetica 30: 88 - 92. Lester, D.EL and G.IL Barr. 1965. Provenance and progeny test in red pine. Forest Science 111: 327 — 540. Lester, D. T. and G. E. Rehfeldt. 1967. Frequency of lammas growth in a provenance test of red pine. Can. J. Bot. 45: 853 — 838. Levenson, B. T. and J. W. Hanover. 1985. Statictical survey of the Michigan tree seedling industry. Mich. State Univ. Agric. Expt. Sta. Res. Rep. No. 455. 20 p. Merrill, R. E. and C. A. Mohn. 1985. Heritability and genetic correlations for stem diameter and branch characteristics white spruce. Can. J. For. Res. 15: 494 - 497. Morris, R. W., W. B. Critchfield, and D. P. Flower. 1980. The putative Austrian x red pine hybrid: a test of paternity based on allelic variation at enzyme- specifying loci. Silvae Genetica 29: 95 - 100. Moulalis, IL, C. Bassiotis, and D. Mitsopoulos. 1976. Controlled pollinations among pine species in Greece. Silvae Genetica 25: 95 - 107. 22 Mullins, J. A. 1983. Heritabilities and gains in volume, diameter and height for open-pollinated progeny of eastern white pine from Cumberland Mountains. Proc. of the Third North Central Tree Improv. Conf., Wooster, Ohio. Peterson, T. A. 1966. Variation in radial growth patterns and specific gravity of red pine (Pinus resinosa Ait.). PhD. dissertation. University of Wisconsin, Madison, Wisconsin. 191 p. Rees, L. W. and R. M. Brown. 1954. Wood density and seed source in young plantation red pine. J. of For. 52: 662 -665. Rehfeldt, G. E. and D. T. Lester. 1966. Variation in shoot elongation of Pinus resinosa Ait. Can. J. Bot. 44: 1457 - 1469- Rudolf, P. O. 1947. Importance of red pine seed source. Proc. Soc. Amer. Forester 1947: 384 - 398. 1953. Forest genetics work at the Lake States Forest Experiment Station. Proc. Lake States For. Genetics Conference. Lake States For. Expt. Sta. USDA For. Serv. Misc. Rept. 22: 1-— 10. 1954. Seed source and earliness of shoot growth in young red pine seedlings. USDA For. Serv., Lake States For. Expt. Sta. Tech. Note 425. 1 p. 1964. Some evidence of racial variation in red pine (Pinus resinosa Ait.). Proc. of the Ninth Meeting Comm. Forest Tree breeding in Canada. II: 143 -149. Steiner, K. C. 1979. Pattern of variation in bud-burst timing among populations in several pine species. Silvae Genetica 28: 185 - 194. Thielges, B. A. 1972. Intraspecific variation in foliage polyphenols of pines (subsection Sylvestres). Silvae Genetica 21: 114 — 119. Vaartaja, O. 1962. Ecotypic variation in photoperiodism of trees with special reference to Pinus resinosa and Thuja occidentalis. Can. J. Bot. 40: 849 - 856. Wakeley, P. C. 1971. Relation of thirtieth-year to earlier dimensions of southern pines. Forest Science 17: 200 - 209. 34 Wright, J. W. and W. J. Gabriel. 1958. Species hybridization in the hard pines, Series Sylvestres. Silvae Genetica 7:109 - 114. Wright, J. W., R. T. Bingham, and K. W. Dorman. 1958. Genetic variation within geographic ecotypes of forest trees and its role in tree imorovement. J. of For. 56: 505 - 508. Wright, J. W. and W. I. Bull. 1965. Geographic variation in scotch pine. Results of a three-year Michigan study. Silvae Genetica 12: 1 - 15. Wright, J. W., W. I. Bull, and G. Mitschelen. 1963. Geographic variation in red pine. Three—year results. Quart. Bull. Mich. State. Univ. Agr. Expt. Sta. 45: 622 - 630. Wright, J. W., R. A. Read, D. T. Lester, C. Merritt, and C. Mohn. 1972. Geographic variation in red pine. Silvae Genetica 21: 205 - 209. Wright, J. W. 1976. Introduction to forest genetics. Academic Press, New York. 465p. Yao, Y. N., J. A. Pitcher, J. W. Wright, and P. C. Kuo. 1971. Improved red pine for Michigan. Mich. Agr. Expt. Sta. Res. Rep. 146: 1 — 12. APPENDICES 35 Appendix A. Stem diameter means for regions, stands within regions and families within stands Stem diameter mean(mm) Region Stand Family Crawford Allegan L.Peninsula 168 152 Grand Traverse 167 152 301 177 143 302 185 142 303 169 143 304 158 144 306 173 143 307 166 140 308 168 158 309 178 159 310 148 143 511 167 154 312 175 158 313 174 159 314 158 155 315 179 160 316 160 157 317 177 147 521 170 151 322 156 153 323 178 149 324 170 154 325 159 211 326 161 136 327 147 148 328 157 156 329 184 148 330 175 147 Alpena 162 156 551 175 161 352 157 147 354 170 151 358 144 165 Otsego 162 152 399 164 146 401 164 159 402 171 157 405 167 154 407 167 146 408 169 159 409 154 151 410 158 152 L. Peninsula Cheboygan 170 158 428 163 135 450 192 153 431 175 201 432 182 140 36 Appendix A (continued) Stem diameter mean (mmT Region Stand Family Crawford Allegan L. Peninsula Cheboygan 171 152 433 176 156 435 171 148 436 175 150 437 161 153 Cheboygan 171 151 438 188 148 439 184 159 440 154 153 445 183 144 447 176 151 Ogemaw 171 153 460 183 142 461 171 162 464 177 169 465 153 147 466 154 156 467 161 138 468 182 154 Crawford 157 142 ' 473 155 151 474 164 141 475 174 151 480 166 137 481 126 152 482 157 136 484 162 149 Alcona ' 166 156 514 182 160 516 155 161 518 153 151 520 180 158 521 173 155 523 145 123 524 163 170 527 186 163 528 173 154 530 146 154 L. Peninsula Newagyo 168 152 531 167 151 532 171 194 534 154 158 535 186 154 536 170 157 537 178 153 538 184 141 539 196 146 540 159 143 Appendix A (continued) ’Stem diameter mean (mm) Region Stand Family Crawford Allegan Newagyo 542 173 148 543 144 144 544 183 147 545 157 150 546 157 157 547 149 139 548 183 141 549 170 150 552 166 158 555 151 146 Oscoda 167 148 557 172 151 559 153 146 560 177 155 562 169 140 Crawford 177 157 563 173 160 564 179 159 565 180 152 Crawford 169 145 566 158 142 567 176 139 569 178 131 570 172 156 571 182 139 572 159 143 573 145 158 574 179 166 575 162 152 576 161 145 577 189 144 578 157 145 579 175 . 141 581 179 138 586 161 153 587 157 157 L. Peninsula Maniste 167 151 594 167 144 595 177 145 597 157 151 598 181 147 600 154 166 Wexford 166 149 608 173 140 612 181 150 614 155 154 616 154 145 618 168 155 588 178 142 Appendix A (continued) 38 Stem diameter mean (mm) Region Stand Family Crawford Allegan U. Peninsula 157 158 Schoolcraft 159 135 563 154 125 364 165 144 365 152 131 566 159 140 367 151 140 369 158 128 370 159 147 371 160 126 375 173 130 377 163 140 378 153 131 379 162 129 Iron 158 141 380 156 149 581 152 150 382 157 147 383 158 132 384 150 134 387 149 136 388 161 140 390 145 145 393 176 158 394 178 141 Luce 163 154 448 156 161 450 167 211 451 165 150 452 150 149 453 175 149 454 121 138 455 170 147 456 164 146 457 172 142 459 182 143 Chippewa 148 125 487 136 126 491 174 123 492 134 119 39 Appendix B. Familiy ranks in stem diameter growth in two plantations Crawford Allegan Family Stem diameter Family Stem diameter number (% of plant. mean) number (% of plant.1nean) 539 118.1 450 141.6 577 113.8 525 141.6 438 113.3 431 134.9 535 112.0 552 130.2 527 112.0 524 114.4 502 111.4 464 113.4 538 110.8 600 111.4 329 110.8 574 111.4 439 110.8 358 110.7 445 110.2 527 109.4 460 110.2 461 108.7 544 110.2 448 108.1 548 110.2 351 108.1 514 109.6 516 108.1 468 109.6 563 107.4 432 109.6 315 107.4 571 109-6 514 107-4 459 109.6 564 106.7 598 109.0 313 106.7 612 109.0 439 106.7 565 108.4 309 106.7 520 108.4 408 10637 564 107-8 401 106.7 581 107.8 520 106.0 574 107-8 308 106.0 315 107.8 312 106.0 309 107.2 552 106.0 537 107-2 534 106.0 394 107.2 587 105.4 523 107.2 402 105.4 588 107.2 536 105.4 569 107-2 546 105.4 595 106.6 516 105.4 464 106.6 466 104.7 317 106.6 433 104.7 560 106.6 570 104.7 301 106.6 328 104.7 567 106.0 521 104.0 433 106.0 560 104.0 393 106.0 618 104.0 447 106.0 314 104-0 579 105.4 614 103.4 351 105.4 405 103.4 453 105.4 528 103.4 330 105.4 524 103.4 40 Appendix B (continued) Crawford AIlegan Family Stem diameter Family Stem diameter number (5 of plant. mean) number (% of plant. mean) 436 105.4 311 103.4 312 105.4 530 103.4 431 105.4 535 103-4 475 104.8 468 103.4 313 104-8 537 102.7 491 104.8 522 102.7 528 104.2 586 102.7 608 104.2 437 102.7 521 104.2 440 102.7 563 104.2 450 102.7 542 104.2 565 102.0 375 104.2 410 102.0 306 104-2 575 102.0 457 103.6 321 101.3 570 105.6 531 101.3 557 105.6 475 101-3 461 105.0 557 101.5 455 105.0 447 101.3 402 103.0 518 101.3 532 103.0 409 101.3 354 102.4 597 101.3 556 102.4 473 101.3 455 102.4 554 101.5 549 102.4 549 100.7 321 102.4 436 100.7 324 102.4 612 100.7 408 101.2 581 100.7 562 101.2 545 100.0 303 101.2 451 100.0 618 101.2 453 100.0 308 101.2 - 523 100.0 594 100.6 452 100.0 407 100.6 484 100.0 450 100.6 380 100.0 311 100.6 435 99.3 405 100.6 329 99-3 480 100.0 458 99.3 552 100.0 327 99.5 307 100.0 317 98.7 451 99.4 465 98.7 364 99.4 382 98.7 399 98.8 352 98-7 401 98.8 544 98.7 456 98-8 370 98.7 ‘i—w—— 41 Appendix B (continued) Crawfard Allegan Family Stem diameter Family Stem diameter number (5 of plant.mean) number (% of plant. mean) 524 98.2 598 98.7 428 98.2 455 98.7 377 98.2 399 98.0 575 97.6 555 98-0 379 97.6 539 98.0 484 97.6 456 98.0 586 97.0 559 98.0 467 97.0 407 98.0 388 97.0 390 97.3 437 97.0 616 97.0 326 97.0 578 97.0 576 97.0 595 97.0 371 96.4 445 96.6 316 96.4 543 96.6 370 96.8 364 96.0 540 95.8 594 96.0 325 95.8 304 96.0 572 95.8 306 96.0 366 95.8 540 96.0 304 95.2 572 96.0 369 95.2 459 96.0 314 95.2 576 96.0 410 95.2 303 96.0 566 95.2 577 96.0 383 95.2 310 96.0 546 94.6 301 95.3 597 94.6 588 95.3 328 94.6 457 95.3 587 94.6 566 95.3 545 94.6 302 95.3 482 94.6 460 95.3 352 94.6 579 94.6 382 94.6 474 94.6 578 94.6 548 94.6 380 94.0 538 94.6 322 94.0 394 94.6 448 94.0 562 94.0 516 93.4 307 94.0 614 93.4 367 94.0 473 93.4 608 94.0 466 92-8 377 94.0 600 92.8 366 94.0 616 92~8 432 94.0 363 92.8 388 94.0 409 92.8 547 93.3 440 92.8 571 93.2 "11111141111111111“