<%BANNER%>

Genetic Effects of Rooting Ability and Early Growth Traits in Loblolly Pine Clones


PAGE 1

GENETIC EFFECTS OF ROOTING ABILITY AND EARLY GROWTH TRAITS IN LOBLOLLY PINE CLONES By BRIAN STEPHEN BALTUNIS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Brian Stephen Baltunis

PAGE 3

iii ACKNOWLEDGMENTS I would like to thank Drs. Timothy White Dudley Huber, Barry Goldfarb, Hank Stelzer, John Davis, and Rongling Wu for serving on my advisory committee. Special thanks go to Tim White and Dudley Hube r for their guidance, advice, and support throughout my entire program. I also would like to thank the Forest Bi ology Research Cooperative for financial support for this research, the members of th e FBRC for establishing field trials, and especially International Paper Company for providing the facilities for the propagation of the rooted cuttings for this project. Many pe ople were involved w ith bringing this study together, and I would like to express my gra titude to all the faculty, staff, workers, and fellow graduate students who helped. My a ppreciation especially goes to Brian Roth; without his effort, this proj ect could never have happened. Finally, I would like to thank my wi fe, Jacqueline, for her patience and encouragement these last five years.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ..x CHAPTERS 1 INTRODUCTION........................................................................................................1 2 GENETIC EFFECTS OF ROOTING LO BLOLLY PINE STEM CUTTINGS FROM A PARTIAL DIALLEL MATING DESIGN...................................................6 Introduction...................................................................................................................6 Materials and Methods.................................................................................................8 Population..............................................................................................................8 Experimental Design.............................................................................................9 Statistical Analyses..............................................................................................10 Results and Discussion...............................................................................................16 Average Rooting..................................................................................................16 Observed Variance Components.........................................................................18 Causal Variance Components..............................................................................21 Heritability Estimates..........................................................................................22 Type B Genetic Correlations...............................................................................26 Selection for Rooting...........................................................................................27 Conclusion..................................................................................................................28 3 GENETIC ANALYSIS OF EARLY FIELD GROWTH OF LOBLOLLY PINE CLONES AND SEEDLINGS FROM THE SAME FULL-SIB FAMILIES.............30 Introduction.................................................................................................................30 Materials and Methods...............................................................................................33 Population............................................................................................................33 Propagation..........................................................................................................34 Field Design.........................................................................................................35 Statistical Analyses..............................................................................................36

PAGE 5

v Results and Discussion...............................................................................................43 Overall Growth of Clones and Seedlings............................................................43 Genetic Components of Variance........................................................................44 Heritability Estimates..........................................................................................49 Type B Genetic Correlations Between Propagule Types....................................50 Genotype x Environment Interaction..................................................................52 Conclusion..................................................................................................................54 4 GENETIC GAIN FROM SELECTION FO R ROOTING ABILITY AND EARLY GROWTH IN VEGETATIVELY PROP AGATED CLONES OF LOBLOLLY PINE...........................................................................................................................55 Introduction.................................................................................................................55 Materials and Methods...............................................................................................57 Population............................................................................................................57 Rooting and Field Trials......................................................................................57 Statistical Analyses..............................................................................................58 Causal Components of Variance.........................................................................60 Heritability Estimates..........................................................................................61 Type B Genetic Correlations...............................................................................62 Genetic Gain........................................................................................................63 Results and Discussion...............................................................................................65 Causal Components of Variance.........................................................................65 Type B Genetic Correlations...............................................................................66 Genetic Gain........................................................................................................68 Conclusion..................................................................................................................75 5 CONCLUSION...........................................................................................................78 APPENDICES A LOBLOLLY PINE PARTIAL DIALLEL MATING DESIGN. THIRTY-TWO PARENTS WERE CROSSED TO GENE RATE 70 FULL-SIB FAMILIES............81 B VARIANCE COMPONENT ESTIMATES FOR ROOTING ABILITY..................83 C VARIANCE COMPONENT ESTIMATES FOR EARLY GROWTH TRAITS OF LOBLOLLY PINE CLONES AND SEEDLINGS FROM THE SAME FULLSIB FAMILIES...........................................................................................................86 D VARIANCE COMPONENT ESTIMATES FROM THE BIVARIATE ANALYSES OF ROOTING AND 2ND YEAR HEIGHT..........................................88 LIST OF REFERENCES...................................................................................................94 BIOGRAPHICAL SKETCH...........................................................................................101

PAGE 6

vi LIST OF TABLES Table page 2-1. Experimental design for 5 rooting trials of loblolly pine stem cuttings. All trials were established in randomized complete block designs with 4 to 6 blocks and 4 to 9 ramets per clone in a row plot within each block................................................9 2-2. Summary of rooting from 5 loblolly pine trials set over two years and three seasons......................................................................................................................17 2-3. Genetic parameter estimates (standard e rror) for rooting of loblolly pine stem cuttings across 5 trials..............................................................................................23 2-4. Type B additive and dominance va riance correlations among pairs of rooting trials for loblolly pine stem cuttings (above and below diagonal, respectively). Standard errors are given in parentheses..................................................................26 3-1. Location of six field trials, establishmen t date and total number of test trees for each test....................................................................................................................34 3-2. Total number of clones, full-sib families, half-sib families, average number of clones per full-sib and half-sib family, and average number of seedlings per fullsib and half-sib family establis hed at the six field trials..........................................36 3-3. Mean 1st year height, 2nd year height, height increment, and crown width by propagule type for each of the six field trials. Although means are expressed in meters, analyses were conducted using measured traits in centimeters...................44 3-4. Genetic parameter estimates for 1st year height, 2nd year height, height increment, and crown width by propagule type across all six trials. St andard errors are given in parentheses.................................................................................................46 3-5. Genetic correlations between propagule types for 1st year height, 2nd year height, height increment, and crow n width at the parental (propGCABr ) and full-sib family (propFSBr ) levels...........................................................................................................51 4-1. Means, variance component estimates, heritabilities, and genetic correlations from the bivariate analysis of rooting ability and 2nd year height. Standard errors are given in parentheses...........................................................................................67

PAGE 7

vii 4-2. The total genetic correlation (TGBr ) between rooting ability and 2nd year height from analyses of a single rooting trial and field trial. Shaded values indicate the rooting trial in which cuttings originated from for their respective field trials. Standard errors are given in parentheses..................................................................70 4-3. The number of full-sib families (half-sib families) and average number of clones per full-sib family (half-sib family) sele cted from selecting 10% or 1% of the top clones using the combined selection index........................................................77 B-1. Observed variance component estimat es for rooting of loblolly pine stem cuttings from single-trial analyses............................................................................83 B-2. Observed variance component estimates for rooting of loblolly pine stem cuttings from pair-wise-trial analyses......................................................................84 B-3. Observed variance components for roo ting ability from the across-trial analysis using all five rooting trials.......................................................................................85 C-1. Observed variance components for l oblolly pine clones from the across-trial analyses of 1st year height, 2nd year height, height incr ement, and crown width. A separate error variance was modeled for each trial..............................................86 C-2. Observed variance components for lobl olly pine seedlings from the across-trial analyses of 1st year height, 2nd year height, height incr ement, and crown width. A separate error variance was modeled for each trial..............................................87 D-1. Observed variance component estimates from the bivariate analyses of rooting ability from Spring01 and 2nd year height from each of the field trials...................88 D-2. Observed variance component estimat es from the bivariate analyses of rooting ability from Summer01 and 2nd year height from each of the field trials................89 D-3. Observed variance component estimates from the bivariate analyses of rooting ability from Winter02 and 2nd year height from each of the field trials...................90 D-4. Observed variance component estimates from the bivariate analyses of rooting ability from Spring02 and 2nd year height from each of the field trials...................91 D-5. Observed variance component estimates from the bivariate analyses of rooting ability from Summer02 and 2nd year height from each of the field trials................92 D-6. Observed variance component estimates from the bivariate analysis of rooting ability using all five rooting trials and 2nd year height using all six field trials.......93

PAGE 8

viii LIST OF FIGURES Figure page 2-1. The proportion of the additive (2 1 0h), dominance (2 1 0d), and epistasis (2 1 0i) genetic variances on the observed binary sc ale for rooting of loblolly pine stem cuttings from each of the five separate tr ials (biased heritabilities) and from the combined analysis of all five trials. Standard error bars for broad-sense heritability estimat es are included............................................................................19 2-2. Narrow-sense (2Nh) and broad-sense (2NH) heritability estimates for rooting of loblolly pine stem cuttings transformed to the underlying normal scale using the threshold model of Equation 2-11. Standard error bars are included......................25 2-3. Full-sib family mean ( 2FSH) and clonal mean ( 2CLH) heritability estimates for rooting success of 2,200 clones from 70 full-sib families of loblolly pine. Standard error bars are included...............................................................................25 3-1. The proportion of additive (add ) and nonadditive (na) genetic variance components for clones and seedlings across the six field trials, where 2h = add and 2 H = add + na: a. 1st year height, b. 2nd year height, c. Height increment, and d. Crown width. Standard error bars for broad-sense heritability estimates are included..............................................................................................................48 3-2. Rank-rank plots show ing type B genetic correla tions between clones and seedlings from Test B based on: a. Pare ntal BLUP values, b. Full-sib family BLUP values, where full-sib BLUP values are equal to the sum of the predicted general combining ability for each of the two parents plus the predicted specific combining ability of the cross..................................................................................53 4-1. The genetic gain in rooting ability (%) over the population mean for deployment of the best half-sib family, full-sib family, best clone from the best ten full-sib families, and the single best clone when selecting for rooting ability or 2nd year height........................................................................................................................71 4-2. The genetic gain in 2nd year height (%) over the population mean for deployment of the best half-sib family, full-sib family, best clone from the best ten full-sib families, and the single best clone when selecting for rooting ability or 2nd year height........................................................................................................................72

PAGE 9

ix 4-3. Responses to selection in rooting ability and 2nd year height with various selection indices for two cl onal deployment options: 10% of clones selected and 1% of clones selected...............................................................................................75

PAGE 10

x Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy GENETIC EFFECTS OF ROOTING ABILITY AND EARLY GROWTH TRAITS IN LOBLOLLY PINE CLONES By Brian Stephen Baltunis December 2005 Chair: Timothy L. White Major Department: Forest Resources and Conservation Loblolly pine is the most important commercial tree species in the southern United States with over 1.1 billion seedlings plan ted annually. With elite genotypes becoming available, several forest industry companie s in the southeastern United States are developing rooted cutting and somatic embryogenesis programs aiming towards deployment of tested clones or families. However, before clones can be deployed, sufficient data need to be co llected on the population in orde r to have reliable information about the clones for deployment decisions. This dissertation reports on the genetic eff ects of rooting abilit y and early growth traits in nearly 2,200 clones of loblolly pine from 70-full-sib families. More than 239,000 stem cuttings were set in five rooti ng trials over two years. Overall rooting success across the five trials was 43%, and si gnificant seasonal effects were observed. Heritability of rooting ability was estimated both on the observed bina ry scale and on the transformed underlying normal scale.

PAGE 11

xi Rooted cuttings from these trials alon g with seedlings from the same full-sib families were established at several sites, and early growth traits through age two were compared between propagule types. All grow th traits demonstrated genetic variation, and parental and full-sib family rankings were similar for both propagule types. However, estimates of dominance genetic va riance in the seedling population appear to be inflated at the expense of additive effect s due to a lack of randomization of seedlings prior to field establishment. Little ge notype x environment interaction was observed across sites for all traits. A successful clonal forestry program for l oblolly pine based on rooted cutting technology needs to consider selection for bot h rooting ability and subsequent growth. There was a positive genetic correlat ion between rooting ability and 2nd year height at the parental, full-sib family, and clonal levels indi cating that selection for one trait will also lead to improvement of the other. Th e genetic gains in rooting ability and 2nd year height associated with several selection and deploymen t strategies are discussed. Moderate to high family and clonal mean he ritabilities, moderate to high type B correlations, and substantial among-family and among-clone gene tic variation indicate the potential for increasing rooting efficien cy and improving growth.

PAGE 12

1 CHAPTER 1 INTRODUCTION Loblolly pine (Pinus taeda L.) is the most important comm ercial tree species in the United States with over one billion seedlings planted annually (McKeand et al. 2003). Most commercially important tree species rema in relatively undomesticated, and loblolly pine is no exception. Genetic improvement of loblolly pine has been occurring since the 1950s in several tree improvement program s. There are three cooperative tree improvement programs in the southern United States that focus on improvement of southern pine species including loblolly pine : the Cooperative Forest Genetics Research Program (CFGRP), North Carolina St ate University-Industry Cooperative Tree Improvement Program (NCSUITIP), and the Western Gulf Forest Tree Improvement Program (WGFTIP). Loblolly pine tree improvement programs in the South are beginning their 3rd generation of breeding with gain s in volume per unit area up to 30% over unimproved loblolly pine (McKeand et al. 2003). Long-term tree improvement programs ai m to increase the population mean breeding value of a few key traits such as stem volume, disease resistance, and wood properties through breeding and se lection of superior genot ypes. Loblolly pine tree improvement programs are based on recurrent selection for general combining ability, which captures only the additive portion of the genetic variance. However, the nonadditive portion of genetic variation, dominance and epistasis, may be important components of variation for traits.

PAGE 13

2 Both additive and nonadditive genetic vari ation can be captured by deploying fullsib families or clones. However, deployment decisions should be based on reliable information. Field trials established with fu ll-sib seedlings allow the genetic variation to be partitioned into additive and nonadditive components. These trials not only provide ranks of parents or individuals for selection, but also of full-sib families in order to provide information for ma king deployment decisions. In any given generation of breeding, maxi mum genetic gains can be achieved in the deployment population by capturing all of th e genetic variation through operational propagation and deployment of selected clones. However, clonal forestry is not a breeding method to develop better genotypes. Clonal forestry is a method to massproduce well-tested genotypes. Short-term genetic gains may be maximized through deployment of well-tested clones, but longterm gains need to involve both clonal selection and recurrent selecti on for additive genetic variat ion through repeated selection and breeding. Clonal tests derived from full-sib familie s do provide an opportunity to estimate additive and nonadditive components of variance associated with a part icular trait or set of traits. However, tests shoul d be designed with a sufficient genetic structure in order to precisely quantify the genetic variation. For example, Frampton and Huber (1995) reported that they had low power in partitioni ng the genetic variation because of the lack of a mating design among the parents of the fu ll-sib crosses in a loblolly pine clonal study. In another study comparing clones from 30 full-sib families derived from two disconnected 4x4 factorials, Paul et al. (1997) concluded that futu re clonal studies should include more parents in the mating design. Although Isik et al. (2003) estimated the

PAGE 14

3 genetic variances from both clones and seedlings from the same nine full-sib families of loblolly pine, they identified a weakness of their study in that there were a limited number of parent trees used in the mating design. Finally, Frampton and Foster (1993) warned that interpretation of the results may be difficult for studies that only include seedlings and cuttings from a common checklot to be compared to the clonal propagules from select parents and families. In this case, any differences in the field performance because of propagule type may be confounded with the differences in genetic improvement (Frampton and Foster 1993). Clonally replicated progeny trials have been suggested as part of a tree improvement strategy for radiata pine (Jay awickrama and Carson 2000) and for loblolly pine (Foster and Shaw 1987; Isik et al. 2004; Byram et al. 2004) for a number of reasons. First, field trials established with clon ally replicated progeny allow for further partitioning of the genetic variation into th e additive, dominance, and epistatic genetic variation (Foster and Shaw 1988). Second, clonally propagated seedlings can provide genetic information more efficiently and w ith greater precision than zygotic seedling progeny (Burdon and Shelbourne 1974; Isik et al. 2004). Finally, clonal testing and selection strategies can provide greater ga in than seedling options (Shaw and Hood 1985; Mullin and Park 1994; Isik et al. 2004). Based on current technologies, several forest industries in the southeastern United States are pursuing clonal forest ry programs with loblolly pine (Weber and Stelzer 2002). In the initial stages of these clonal forestry programs, forest managers needed assurance that the clonal propagules gr owth corresponded to that of se edlings. Therefore, most of the earlier studies were designed to test wh ether cuttings grew similarly to seedlings.

PAGE 15

4 Based on those results, it is generally accepted that cuttings rooted from juvenile stock plants grow and perform comparably to seedlings. For example, Foster et al. (1987) reported that loblolly pine r ooted cuttings should perform co mparably to seedlings when the cuttings come from vigorous juven ile stock plants. In addition, McRae et al. (1993) concluded that for loblolly pine there were no significant differences between seedlings and rooted cutting propagules from common ch ecklots through five years of growth. Similar results were obtained by Frampton et al. (2000) where they reported no significant differences between the means of rooted cuttings and seedlings for height, diameter at breast height, and volum e through six years in the field. Trials established with clones and seedlings from the same families provide an opportunity for comparing both half-sib a nd full-sib family performances for both propagules. Genetic correlati ons between propagule types ca n provide further assurance that selections made through traditional tree improvement activities for recurrent selection for general combining ability can also be used successfully in breeding families to test in a clonal forestry program. Alt hough a number of studies have been reported comparing rooted cutting and seedlings, very few have been designed to estimate the genetic correlation for a tra it between propagule types. Two main criteria need to be met prior to operational deployment of loblolly pine clones. First, loblolly pi ne clones must perform well, e.g., meet the selection criteria for the desired traits. This involves the accumu lation of reliable data for the clones from greenhouse screening, field trials etc. Second, the selected clones have to be propagated in large enough numbers for deployment. For a rooted cutting based clonal program, this involves bulking up the number of hedges (ram ets) of a particular clone or group of

PAGE 16

5 selected clones through serial propagati on and then producing reforestation stock efficiently from the bulked-up clones. Only those clones that can be propagated easily and in sufficient numbers will be economically feasible for deployment. A clone that grows well in the field but roots poorly may not be economically feasible to include in a clonal program based on r ooted cutting technology. This dissertation is unique in that a comp lex genetic structure was utilized in order to increase the power in quantif ying the genetic variation associated with several traits in loblolly pine clones and seedlings. In Chapte r 2 rooting ability was assessed for nearly 2,200 clones of loblolly pine from 70 fullsib families derived from a partial diallel mating design in order to estimate genetic parame ters associated with rooting. More than 1,200 of these clones along with zygotic seed lings from the same full-sib families were established together on multiple sites across the southeastern United States. Genetic parameter estimates are compared between pr opagule types for early growth traits in Chapter 3. Finally, selection for both rooti ng ability and field growth is addressed in Chapter 4.

PAGE 17

6 CHAPTER 2 GENETIC EFFECTS OF ROOTING LOBLO LLY PINE STEM CUTTINGS FROM A PARTIAL DIALLEL MATING DESIGN Introduction Loblolly pine is the most important commercial tree species in the southern United States with over 1.1 billion s eedlings planted annually (McKeand et al. 2003). There are three cooperative tree improvement programs in the southern United States that focus on improvement of southern pine species includ ing loblolly pine: the Cooperative Forest Genetics Research Program (CFGRP), No rth Carolina State University-Industry Cooperative Tree Improvement Program (NCSUI TIP), and the Western Gulf Forest Tree Improvement Program (WGFTIP). Loblolly pine tree improvement programs in the South are beginning their 3rd generation of breeding with gains in volume per unit area up to 30% over unimproved loblolly pine (McKeand et al. 2003). Long-term tree improvement programs ai m to increase the population mean breeding value of a few key traits through breedi ng and selection of superior genotypes. These programs are based on recurrent select ion for general combining ability which captures only the additive portion of the genetic variance. In any given generation of breeding, maximum gains can be achieved in the deployment popul ation by capturing all of the genetic variation (additive and nonadditive components) through operational propagation of selected clones. With elite genotypes becoming available, several forest industries in the southeastern United States are developing rooted cutting programs for

PAGE 18

7 loblolly pine aiming towards deployment of tested clones or families (Weber and Stelzer 2002). Clonal tests derived from full-sib fam ilies provide an opportunity to estimate additive and nonadditive genetic components of variance associated with a particular trait or set of traits (Isik et al. 2003; Isik et al. 2004). In clonal field tr ials traits of interest may include height, volume, wood quality, and disease resistance. However, in order to establish clonal field trials the clones must fi rst be propagated. Th erefore, clonal rooting trials are important for estimating genetic va riance components associated with rooting. Previous rooting studies of loblolly pine have been relatively small in size ranging from several hundred (Goldfarb et al. 1998; Foster et al. 2000) to several thousand cuttings (Foster 1990; Anderson et al. 1999). Many studies contai ned a small number of families from factorial mating designs (Goldfarb et al. 1998; Anderson et al. 1999; Cooney and Goldfarb 1999; Frampton et al. 1999) and few have been designed to estimate genetic parameters associated with r ooting in loblolly pine (Foster 1978; Foster 1990; Anderson et al. 1999). The current study is unique in that a large number of cuttings were set in each trial (> 34,000), and rooting was assessed on nearly 2,200 clones of loblolly pine from 70 full-sib families derived from a partial diallel mating design in order to estimate genetic parameters associated with rooting. The objectives of the study were to (i) evaluate the rooting ability of stem cu ttings from nearly 2,200 loblolly pine clones, (ii) determine the causal components of va riance in rooting of stem cuttings, (iii) assess heritability estimates for rooting from fi ve trials both on the observed binary scale and the underlying normal scale, and (iv) determine the Type B genetic correlations for

PAGE 19

8 both additive and dominance gene tic effects to measure th e correspondence in rooting performance across five setting dates. Materials and Methods Population The parental population for this study was selected from the Loblolly Pine Lower Gulf Elite Population (LPLGEP) which consists of selections from a ll three southern pine tree improvement cooperatives: CFGRP, NCSUITIP, and WGFTIP. Twenty 1st generation and ten 2nd generation selections representing the Atlantic Coastal Plain, Florida, and Lower Gulf provenances were se lected from this populat ion. Two additional slow-growing parents were included to pr ovide linkage with another study. These parents were crossed in a circular dial lel mating design (Appendix A) with some additional off-diagonal crosses, resulting in a total of 70 full-sib families. On average each parent was involved in a pproximately four crosses. Seeds from the 70 families were sown in March 2000 into Ray Leach SuperCells (Stuewe and Sons, Corvallis, OR). The seedlings were grown in a greenhouse at International Paper Companys (IPC) facility in Jay, FL, and after th ree months of growth the seedlings were pruned back to a height of about 10-12 cm. Approximately 32 seedling hedges (ortets) per fu ll-sib family were transplanted into 3-gallon containers and given unique clonal identifica tions in September 2000. The hedges were repeatedly sheared in order to minimize the effects of maturation and increase the number of shoots available for the rooting trials. The ortets were randomized in a containerized hedge orchard in order to reduce spuri ous C effects at the family level. However, C effects at the clonal level could not be accounted for, b ecause all cuttings originated from a single seedling ortet.

PAGE 20

9 Table 2-1. Experimental design for 5 rooting trials of loblolly pine stem cuttings. All trials were established in randomized complete block designs with 4 to 6 blocks and 4 to 9 ramets per clone in a row plot within each block. Trial Date set No. of families No. of clones No. of plots per clone No. of cuttings per plot Total cuttings Spring01 May 711, 2001 70 2194 4 4 34,707 Summer01 July 2-6, 2001 70 2157 5 4 43,048 Winter02 Jan 1418, 2002 61 1648 6 5 49,315 Spring02 Apr 29May 3, 2002 61 1254 6 9 67,059 Summer02 June 2428, 2002 61 947 6 9 45,108 Experimental Design Two rooting trials were conducted in 2001 and three were conducted in 2002. Stem cuttings between 3 and 8 cm in length were harvested from the s eedling ortets in May 2001, July 2001, January 2002, April/May 2002, and June 2002, for trials Spring01, Summer01, Winter02, Spring02, and Summer 02, respectively. Cutting size was relatively consistent within any trial, and th e cuttings set in Winter02 were the smallest. The experimental design differe d among the trials due to numb er of families, clones and available cuttings from each ortet (Table 21). The reduction in the number of families and clones between the first and last rooting tr ials was a result of a number of factors. First, random hedge mortality was a major fact or: disease, repeated severe pruning, and uneven watering (inadequate) all contributed to the random loss of hedges. Since cuttings of a clone originated from a singl e ortet, by default mortality resulted in a truncated population for future rooting tria ls. Second, not all hedges produced an adequate number of cuttings at every harvest. The primary objective of the three settings

PAGE 21

10 in 2002 was to produce propagules for field tr ials. Therefore, cl ones that were not producing enough shoots to ultimately be pl anted across six fiel d sites were culled regardless of rooting frequency. This resulted in the reduction in the number of families in the last three rooting trials because of t oo few surviving clones in some of the families to meet the goal for field designs. We were striving for a balanced field design with 15 clones from each of 61 full-sib families. Cuttings were randomly set in 4-, 5-, or 9cutting clonal row plots (Table 2-1) into pre-formed plugs consisting of peat moss, pe rlite, and a binding resin. Each plug was approximately 13 mL in volume and was held by the V-13 HIKO tray (135 cells; Stuewe and Sons, Corvallis, OR). Cuttings were eith er treated prior to setting with a 1.0% indole-3-butyric acid and 0.5% napthalene-1-a cetic acid (NAA) basal dip or after setting using a foliar NAA application according to IP C protocols. Trays contained 15 to 30 clones depending on the trial and were randomly placed in an environmentally controlled greenhouse. There were 4 to 6 complete re plications depending on trial (Table 2-1). Rooting assessments were made 9-weeks after setting for both the Spring01 and Summer01 trials. Cuttings were measured fo r presence (1) or absence (0) of roots. Cuttings with a root 1 mm were considered rooted (Goldfarb et al. 1998, Foster et al. 2000). For trials Winter02, Spring02, a nd Summer02 assessments were at 11-weeks following setting. Cuttings in these trials we re also scored for presence or absence of roots. However, only cuttings th at had at least one visible r oot on the exterior of the plug regardless of length were considered rooted. Statistical Analyses For binomial traits, such as rooting, the unit of analysis can be the individual observations (Huber et al. 1994; Dieters et al. 1996) or plot means combined with a

PAGE 22

11 transformation such as arcsin or lo gistic (Sohn and Goddard 1979; De Souza et al. 1991). We chose to analyze the observed 0,1 data for several reasons. First, REML estimation of variance components has been shown to be robust to violations of the underlying normality assumptions (Banks et al. 1985; Westfall 1987) suggesti ng that analyses using individual observations of bi nary data yields satisfactory results. Second, simulation studies have shown that the us e of individual observations is superior to the use of plot means in REML, and that these variance component estimates perform well across mating designs and imbalanced data (Huber et al. 1994). Huber et al. (1994) showed that a lower variance among estimates was obtai ned using individual observations as compared to plot means and that this adva ntage increased with increasing imbalance. Third, when heritability is low and the incide nce close to 50%, there is little difference between heritability estimates on the binary and transformed scale (Dempster and Lerner 1950). In fact these two estimates are equiva lent when the inciden ce is exactly 50% for low heritability traits. Lopes et al. (2000) demonstrated that the Dempster and Lerner (1950) threshold model closely estimates the true underlying heritability at incidences between 25% and 75% for traits with low heritability (3 02 h ). Finally, Lopes et al. (2000) also demonstrated (for tr aits with low herita bility) that heritab ility estimates from the observed binary data without transformation of data result in predicted gain close to the realized gain, while tran sformations can suffer from issues of back transformation when one wishes to predict gains on the original scale. All variance components for rooting ability of loblolly pine stem cuttings were estimated using the individual binary obser vations using REML estimation for each of the 5-rooting trials using GAREML (Huber 1993). However, upwardly biased estimates

PAGE 23

12 of genetic variances result when variance components are estimated from single-site (trial) analyses since the estimated genetic variance also contains the genotype x environment interaction (Comstock and Moll 1963). Therefore, across-trial analyses were performed to separate the genotype x e nvironment interactions in order to remove this bias. [2-1] ijklmno lm i j lm in ilm im il lm n lm m l i j k i j i ijklmnoerror fam r fam c t sca t gca t gca t fam c sca gca gca tray R T y ) ( ) ( ) ( )) ( ( ) (* ) ( * * ) ( where yijklmno is the rooting respons e (0 or 1) of the oth ramet of the nth clone within the lmth full-sib family in the kth tray within the jth rep of the ith trial is the population mean Ti is the fixed effect of trial Rj(i) is the fixed effect of rep trayk(j(i)) is the random variable tray (incomplete block) ~ IID(0,2TRAY) m and lgca is the random variable female (l) and male (m) general combining ability (gca) ~ IID(0,2GCA) scalm is the random variable sp ecific combining ability (sca) ~ IID(0,2SCA) c(fam)n(lm) is the random variable clone within family ~ IID(0,2CLONE) t*gcail and im is the random variable test by female gca and test by male gca interaction ~ IID(0,2TESTxGCA) t*scailm is the random variable test by full-sib family interaction ~ IID(0,2TESTxFAM) t*c(fam)in(lm) is the random variable test by clone interaction ~ IID(0,2TESTxCLONE) r*famj(i)lm is the random variable rep by family interaction ~ IID(0,2REPxFAM)

PAGE 24

13 errorijklmno is the random error which include s among plot and within plot ~ IID(0,2ERROR). The single trial model is identical except all model factors with subscript i are removed (sources involving test). Genetic parameters were estimated and st andard errors were calculated according to Foster and Shaw (1988) us ing the appropriate variance co mponents from the individual or across trial model. Estimates of additive and dominance genetic variance are upwardly biased because they are confounded with fr actions of epistasi s (Cockerham 1954). Epistastic genetic variance is also only a pproximated because it co ntains only a fraction of the total epistasis plus any C effects, if they exist. [2-2] ... 4 16 1 4 1 2 AAA AA A GCA AV V V V = estimate of additive genetic variance [2-3] ... 4 4 1 2 1 2 1 2 DD AD AA D SCA DV V V V V = estimate of dominance genetic variance [2-4] ... 3 2 4 3 2 1 4 1 2 2 2 DD AD AA SCA GCA CLONE IV V V V = estimate of epistatic genetic variance [2-5] 2 2 2 2 CLONE SCA GCA GV = estimate of total genetic variance [2-6] 2 2 2 2 2 2 2 2 2 2 ERROR REPxFAM TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA PV = phenotypic variance for across tr ial model (the phenotypic va riance from the individual trial model is the same but drop2 2TESTxGCA, 2TESTxFAM, and 2TESTxCLONE). Biased and unbiased heritab ility estimates for rooting based on observed 0,1 data were derived using the estimated variance components from the single and across trial

PAGE 25

14 models, respectively. In addi tion the proportion of dominance (2 d ) and epistasis (2 i) were estimated. Standard errors of these es timates were calculated using a Taylor series expansion (Kendall and Stua rt 1963; Namkoong 1979; Huber et al. 1992; Dieters 1994). [2-7] 2 2 2 2 2 2 2 2 2 2 1 0 2 2 4 ERROR REPxFAM TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA GCAh = across-trial narrow-sense hertibility based on observed binary data [2-8] 2 2 2 2 2 2 2 2 2 2 2 2 1 0 2 2 2 ERROR REPxFAM TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA CLONE SCA GCAH = across-trial broad sense heritability based on observed binary data [2-9] 2 2 2 2 2 2 2 2 2 2 1 0 2 2 4 ERROR REPxFAM TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA SCAd = across-trial dominance proportion [2-10] 2 2 2 2 2 2 2 2 2 2 2 2 1 0 2 2 3 2 ERROR REPxFAM TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA SCA GCA CLONEi = across-trial epistatic proportion. The main problem with calculating heritabi lity on the observed 0,1 data is that the relationship between 2h on the observed scale and 2h on the underlying normal scale depends on the mean incidence ( e.g. % survival, % infected individuals, rooting percentage), and therefore th e conversion results in a bias ed estimate (Van Vleck 1972). However, this is not a problem for low heritability traits with intermediate incidences

PAGE 26

15 (Lopes et al. 2000). In order to make valid comparis ons to heritability estimates in other rooting trials that have diffe rent mean rooting percentages 2 1 0 h needs to be transformed to 2h on the underlying normal scale. Therefore, narrowand broad-sense heritability estimates on the observed 0,1-scale were tr ansformed using a threshold model to an underlying normal scale (Dem pster and Lerner 1950). [2-11] 2 2 1 0 2) 1 )( ( z p p h hN where 2Nh is the heritability on the underlying normal scale p is the rooting percent zis the ordinate of the normal density f unction which corresponds to probability p. Full-sib family mean heritability and cl onal mean heritability for rooting were estimated for both the singleand across-tria l analyses. Standard errors for these estimates were calculated using Dicker sons Method which assumes the phenotypic variance (denominator) is a know n constant (Dickerson 1969). [2-12] ctrn tr tc t t c HERROR REPxFAM TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA SCA GCA FS 2 2 2 2 2 2 2 2 2 2 2 2 2 2 = across-trial family mean heritability, wh ere, c = harmonic mean number of clones per family, t = number of trials, r = harmonic mean number of reps per test, and n = harmonic mean number of ramets per clone per plot.

PAGE 27

16 [2-13] trn tr t t t HERROR REPxFAM TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA CLONE SCA GCA CL 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 = across-trial clonal mean heritability. Type B genetic correlations for rooting acr oss all 5 trials were estimated for both additive and nonadditive components (Yamada 1962; Burdon 1977). Standard errors of Type B correlations were calculated usi ng the Taylor Series Expansion method. [2-14] 2 2 2 ge g g Br where Br is the estimate of the Type B genetic correlation, 2g is the genetic variance component (either additive or dominance), and 2ge is the G x E interaction (additive and dominance). Results and Discussion Average Rooting A total of over 239,000 stem cuttings from nearly 2,200 clones of loblolly pine were set in five rooting trials. Overall rooting across the five trials was 43% and is comparable to other rooting studies involving loblolly pine. Goldfarb et al. (1998) reported 44% rooting of loblolly pine stem cuttings from 400 seedling hedges from one open-pollinated family. Over four rooting trials, Anderson et al. (1999) reported 33% rooting from 90 clones of loblolly pine from 9 full-sib families. Foster (1990) reported overall rooting from three settings of 42% fo r 546 clones of loblolly pine derived from 54 full-sib families. However, studies with fewer families of loblolly pine have yielded

PAGE 28

17 substantially higher rooti ng percentages (Cooney and Goldfarb 1999; Murthy and Goldfarb 2001; LeBude et al. 2004). Rooting of loblolly pine cuttings was vari able over the five trials (Table 2-2). Spring cuttings rooted at greater than 50% while summer cuttings in the two summer trials averaged 38% and 24% respectively. Winter cuttings were intermediate at 45% rooting. Broad inference linear contrasts we re constructed using the estimates of the fixed effects in order to test seasonal rooti ng responses. Cuttings set in the two spring trials rooted at a significantly greater frequency then cuttings set in the summer trials ( p < 0.0001). This implies that we would always expect a greater rooti ng percent in spring settings than in summer setti ngs under this propagation system. Table 2-2. Summary of rooti ng from 5 loblolly pine trials set over two years and three seasons. Trial Rooting % Half-sib family range Full-sib family range Clone range Spring01 54% 36-70% 28-77% 0-100% Summer01 38% 24-54% 18-69% 0-100% Winter02 45% 17-60% 17-67% 0-100% Spring02 51% 36-67% 28-75% 0-98% Summer02 24% 9-43% 8-53% 0-89% Seasonal rooting responses for loblolly pine stem cuttings have been observed in other studies. Early rooting trials of loblolly pine cuttings report ed best rooting from cuttings set from September through Janua ry (Cech 1958; Reines and Bamping 1960; Grigsby 1962; Marino 1982). These early experiments concluded that increased temperatures in the greenhouse in spring and su mmer trials decreased rooting. In fact,

PAGE 29

18 Cech (1958) reported a 3-fold increase in roo ting under cool conditions rather than under warm conditions. However, Foster et al. (2000) observed an overall rooting of 50% for a rooting trial established in March, while on ly 20% rooting for a trial established in September. They hypothesized that the re duction in September rooting was due to a decrease in metabolic activity due to the decrease in photoperiod. Rowe et al. (2002a; 2002b) reported trends in rooting similar to those of the current study. They observed 59% rooting for spring cuttings versus 40% rooting for winter cuttings and 35% rooting for summer cuttings. Cooney and Goldfarb (1999) also reported hi gh rooting percentages for spring cuttings (62% and 83% in two su ccessive years). In contrast, Murthy and Goldfarb (2001) reported higher rooting percen tages for winter cuttings (85%) than for spring cuttings (60%). Winter cuttings often take longer to root but overall rooting may not be different than spring cuttings. Perh aps the slight reduction in rooting frequency for the Winter02 setting versus the two spring settings was a f unction of rate of rooting. The reduction in rooting seen here for su mmer settings may be a result of increased temperatures during the collec tion and propagation phases of th e experiment. The higher temperatures and humidity experienced during summer months may have resulted in an increased abundance or activity of pathogens and hence a higher rate of decay was observed in the two summer trials. The tim e delay between collec tion and setting of cuttings may also have contributed to this reduction in rooting. Murthy and Goldfarb (2001) reported a decline in rooting percentage with increasing drying time of cuttings. Observed Variance Components Variance components were estimated for all singleand acros s-trial analyses (Appendix B). Even with the reduction in th e number of clones and families throughout the study, there was no apparent reduction in the variance component estimates over time

PAGE 30

19 as evidenced by parameter estimates, e.g. additive genetic variance estimates were relatively constant over time (Figure 2-1) The variance associated with general combining ability was 2% of the total phenot ypic variation associ ated with rooting. Halfsib family rooting percentages ranged from a high of 36-70% for Spring01 to a low of 943% for Summer02 (Table 2-2). The proportion of the total variation in rooting that was accounted for by specific combining ability was only 0.3-1.1%. Full-sib family means for rooting for the two spring settings ra nged from 28-76%. Although overall rooting was greater for these two spring trials, the ne t difference in the range of family means was approximately the same (~45-51%). Similar ranges in family mean rooting percentages have been re ported (Foster 1990; Anderson et al. 1999). 0 0.05 0.1 0.15 0.2 0.25 Spring01Summer01Winter02Spring02Summer02AllTrial 2 1 0 h 2 1 0 d 2 1 0 i 2 1 0 H Figure 2-1. The proportion of the additive (2 1 0 h ), dominance (2 1 0 d ), and epistasis (2 1 0 i ) genetic variances on the observed binary scale for rooting of loblolly pine stem cuttings from each of the five separate trials (biased heritabilities) and from the combined analysis of all five trials. Standard error bars for broadsense heritability estimates are included.

PAGE 31

20 Clones within families accounted for appr oximately 10-17% of the total variation in rooting in the five trials (Appendix B). Foster (1990) reported that the clone within family source of variation in rooting of 546 clones of loblolly pine was 3.7 % of the total variation. However, in another study involvi ng cuttings of loblolly pine, the among clone source of variation accounted fo r nearly 22% of the total va riation (Foster 1978). In the current study, the variance associated with clone within family was 4.5-8.4 times greater than the gca variance and 13-56 times greater than the sca variance based on single-trial analyses. Just as there was a large range in root ing among families, there was also a large range in rooting among clones within families (Table 2-2). Rooting for clones within families ranged from 0-100% for the first th ree trials, and ranged from 0-98% and 0-89% in the last two trials, respectively. Anderson et al. (1999) observed ranges in rooting frequency for clones within family similar to those observed in this study. Foster (1990) reported significant variation for clones with in family with rooting percentages ranging from 6.7-85.0%. A large variation in the rooting environm ent has been reported in many rooting experiments of loblolly pine and other spec ies, with an error variance ranging from 4671% (Foster 1978; Foster 1990; Sorens en and Campbell 1980; Cunningham 1986). In the current study, the majority of the observed variance in rooting wa s attributable to the error variance. The error variance account ed for 76.3-83.7% of the total variance observed in the five trials. Each rooting tr ial was spread over an entire greenhouse due to the large size of the experiments. Appare ntly, the rooting environment was not uniform throughout the greenhouse. There are many fact ors that can contribute to a variable

PAGE 32

21 rooting environment. Differential temper ature gradients, unequal airflow, unequal misting of cuttings, edge effects, and diseas e incidence can all cont ribute to a variable rooting environment. Causal Variance Components The observed variance components were used to estimate additive, dominance and epistatic variances using equations 2-2, 2-3, and 2-4, respectively. Additive genetic variance was approximately 2 to 6 times larg er than the dominance genetic variation for rooting in the five trials (Figure 2-1). In Foster (1990), the additive genetic variance was also about 6 times larger than the domina nce genetic variance. However, Anderson et al. (1999) found three times and Foster (1978) found 2.2 times greater dominance genetic variance in rooting compared to the additive va riance in loblolly pine stem cuttings. The epistatic genetic variance was also estimated in the current study and was approximately 0.44 to 1.49 times as much as the additive genetic variance. The lowest amount of epistasis was observed for the winter sett ing and was the only setting that had a nonadditive to additive ratio less than one. C effects can lead to upwardly biased estimates of total genetic and nonadditive genetic components of variance when analyz ing clonal data (Libby and Jund 1962). If C effects are present, then total genetic va riation associated with clones will be overestimated (Libby and Jund 1962). Significant C effects are likely to occur in traits that are measured soon after propagation (such as rooting traits, early shoot growth, etc.), but apparently lessen for traits measured at later times (Libby and Jund 1962). Foster et al. (1984) used a secondary cloning approach to separate C effects from the genetic variance in rooting of western hemlock cu ttings. They found significant C effects associated with rooting and that these non-gene tic effects biased the genotypic values of

PAGE 33

22 clones. However, low or non-significant C e ffects for rooting of tamarack and balsam poplar cuttings have been reported (Farmer et al. 1989; Farmer et al. 1992). In the current study, estimates of epistatic gene tic variance components are confounded with C effects, because ramets of a clone came fr om the original seedling ortet. However, estimates of additive and dominance geneti c effects are not confounded with C effects, because the clones within families were randomized in the hedge orchard. Heritability Estimates Rooting of loblolly pine stem cuttings wa s weakly controlled by additive effects. Individual tree narrow-sense heritability using the observed variance components (2 1 0 h ) ranged from 0.075 to 0.089 in the five separate trials (Figure 2-1). These estimates are in agreement with Foster (1978) who reported 2 h as 0.07 for rooting percentage of loblolly pine. A slightly higher 2 h was reported by Foster (1990). However, Anderson et al. (1999) reported 2 h of 0.26 for rooting percentage in loblolly pine stem cuttings. However, all of these previous studies of l oblolly pine rooted cuttings analyzed data based on plot rooting percentage as opposed to using 0,1 data as in this study, and this leads to an increased estimate of heritability due to the reduction of the impact of the within plot portion of the error variance in a ny plot means or plot percentage analysis for rooting. When the data in the current st udy were analyzed based on rooting percentage, 2 h was 0.18 (from Spring01, data not shown), which falls in the middle of the range previously reported. The unbiased estimates of individual tree narrow-sense heritability (2 1 0 h ) using the 0,1 data from the pairwise test analyses ra nged from 0.045 to 0.074. When all of the data

PAGE 34

23 from the 5 trials were analyzed together, 2 1 0 h was 0.051 (Table 2-3; Figure 2-1). The proportion of dominance (2 1 0 d ) was estimated for each of the trials. The upwardly-biased estimates for this parameter ranged from 0.01 4 to 0.044 (Figure 2-1). When all of the data were analyzed together from the 5 trials, 2 d was 0.018 (Table 2-3; Figure 2-1). The epistatic proportion (2 1 0i) was also estimated and ranged from 0.032 to 0.095 (Table 2-3; Figure 2-1). Table 2-3. Genetic parameter estimates (standard error) for rooting of loblolly pine stem cuttings across 5 trials. Parameter Estimate Narrow-sense heritability on observed binary scale (2 1 0 h ) 0.051 (0.017) Broad-sense heritability on observed binary scale (2 1 0 H ) 0.101 (0.008) Narrow-sense heritability on the underlying normal scale (2Nh) 0.08 (0.027) Broad-sense heritability on the underlying normal scale (2NH) 0.16 (0.013) Narrow-sense family mean heritability ( 2FSH ) 0.833 (0.24) Broad-sense clonal mean heritability ( 2CLH ) 0.815 (0.074) Additive genetic variance (AV ) 0.0117 (0.004) Dominance genetic variance (DV ) 0.0042 (0.002) Epistatic genetic variance (IV ) 0.0074 (0.002) Phenotypic variance (PV ) 0.2297 (0.002) Type B additive genetic variance correlation 0.68 (0.23) Type B dominance genetic va riance correlation 0.61 (0.27) Type B total genetic variance correlation 0.53 (0.048) Broad-sense heritability (2 1 0 H ) ranged from approximately 0.15 to 0.22 (Figure 21). Broad-sense heritability was 0.101 when all of the data from the five trials were combined (Table 2-3; Figure 2-1). Foster (1 990) observed very little nonadditive genetic

PAGE 35

24 variance and reported 2 H of 0.13 for rooting percentage. In another study 2 H was reported as 0.23 for rooting percenta ge (Foster 1978). However, Anderson et al. (1999) reported a much higher2 H (0.63) for rooting percentage of loblolly pine. When the data in this study were analyzed based on rooting percentage, then 2 H was 0.47. Narrowand broad-sense heritability estim ates on the observed 0,1-scale were transformed to an underlying normal scale assuming a threshold model (Equation 2-11). Narrow-sense heritability based on the underlying normal scale (2Nh) ranged from 0.12 to 0.16 (Figure 2-2). Broad-sense heritabili ty based on the underlying normal scale (2NH) ranged from 0.24 to 0.36 (Figure 2-2). When all of the data from the five settings were analyzed together, then 2Nh was 0.08 and 2NH was 0.16 (Table 23; Figure 2-2). The transformation of 2 1 0 h and 2 1 0 H to the underlying normal scale allows direct comparisons of heritability estimates among rooting studie s when the rooting percentages are different. Family mean heritability ranged from 0.84 to 0.9, while clonal mean heritability ranged from 0.82 to 0.92 (Figure 2-3). When all of the data from the five trials were combined then 2FSH was 0.83 and 2CLH was 0.82 (Table 2-3; Figu re 2-3). Foster (1990) reported lower estimates of both family-mean heritability and clonal -mean heritability for a rooting study consisting of 540 clones of loblolly pine from 54 full-sib families, 0.46 and 0.40, respectively. In another loblolly pine rooting study with 27 full-sib families consisting of 10 clones each, family-mean heritability was reported as 0.31, while the clonal-mean heritability was 0.87 (Anderson et al. 1999).

PAGE 36

25 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4Spring01Summer01Winter02Spring02Summer02AllTrialNormal Scale 2Nh 2NH Figure 2-2. Narrow-sense (2Nh) and broad-sense (2NH) heritability estimates for rooting of loblolly pine stem cuttings transf ormed to the underlying normal scale using the threshold model of Equation 2-11 Standard error bars are included. 0 0.2 0.4 0.6 0.8 1 1.2Spring01Summer01Winter02Spring02Summer02AllTrialEstimate 2FSH 2CLH Figure 2-3. Full-sib family mean ( 2FSH ) and clonal mean ( 2CLH ) heritability estimates for rooting success of 2,200 clones from 70 full-sib families of loblolly pine. Standard error bars are included.

PAGE 37

26 Type B Genetic Correlations The type B correlations for additive effect s were moderately high and ranged from 0.53 to 0.91 (Table 2-4) indicating that parental rankings of the 32 parents were moderately to strongly correlated among pairs of trials. When all of the data were analyzed together from the five trials, the type B additive correlation was 0.68. The highest type B correlation was observed between the two spring settings which were one year apart, while the lowest correlations were observed between the winter setting (Winter02) and other trials. Spring and su mmer cuttings were, in general, actively growing, succulent material, while winter cuttings were generally smaller and more lignified. Perhaps, some of the genes contro lling rooting of dormant winter cuttings are different than those controlling spring and summer cuttings. Table 2-4. Type B additive and dominance variance correlations among pairs of rooting trials for loblolly pine stem cuttings (above and below diagonal, respectively). Standard errors are given in parentheses. Spring01 Summer01 Wi nter02 Spring02 Summer02 Spring01 0.701 (0.11) 0.59 (0.15) 0.911 (0.07) 0.851 (0.08) Summer01 0.823 (0.19) 0.634 (0.12) 0.673 (0.12) 0.634 (0.13) Winter02 0.192 (0.30) 0.629 (0.34) 0.656 (0.12) 0.525 (0.16) Spring02 0.493 (0.24) 0.533 (0.35) 0.724 (0.24) 0.907 (0.06) Summer02 0.777 (0.22) 0.973 (0.37) 0 0.703 (0.24) Estimated type B correlations for domina nce effects measure the correspondence of dominance across pairs of trials and were esti mated to be moderate to high (Table 2-4)

PAGE 38

27 with two exceptions. First, the type B do minance correlation between the Spring01 and Winter02 trials was 0.192. Also, no domi nance genetic variance was detected in the pairwise analysis of theWinter02 and Summer02 trials. However, when all of the data were analyzed together, the type B dominance correlation was 0.61. Selection for Rooting Before clones of loblolly pine can be deployed operationall y, two things must occur. First, the clones have to be fiel d-tested and selected for desirable traits, e.g. growth, disease resistance, and wood properties. Second, the selected clones have to be be propagated in large enough numbers for deployment. For a r ooted cutting-based program, this involves bulking up the number of hedges of a particul ar clone or group of clones through serial propagation and then producing reforestation st ock efficiently from the bulked-up clones. Only those clones that can be propaga ted easily will be economically feasible for deployment. Theref ore, selection of clones for rootability as well as field performance should be considered as part of a clonal forestry program based on rooted cutting technology (Foster et al 1985; Foster et al 2000). In the current study, selection of the top 10% of clones for r ooting (~220 clones) would result in a gain of about 24% in rooting, where trn tr t t t H i GainERROR REPxFAM TESTxCLONE TESTxSCA TESTxGCA CLONE SCA GCA CL2 2 2 2 2 2 2 2 2 2 2 (Foster 1990). Selecting the top 1% of clones for rooting w ould result in a gain of nearly 37% in rooting of loblolly pine stem cuttings in the current generation. Alternatively, genetic

PAGE 39

28 gain in rooting success in the next gene ration can be achieved through selection and breeding. trn tr t i ctrn tr tc t t c H i GainERROR REPxFAM TESTxCLONE CLONE GCA I ERROR REPxFAM TESTxCLONE TESTxSCA TESTxGCA CLONE SCA GCA FS F 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 By selecting the best rooting clone from the top 25 out of 70 families, gain in rooting success of 16.8% can be expected in the next generation by breeding these 25 selections. Conclusion Loblolly pine is the most important comm ercial tree species in the southeastern United States. Several forest industry comp anies are developing rooted cutting programs for loblolly pine in order to maximize geneti c gains through deployment of tested clones. With rooting data from 2,200 clones from 70 full-sib families, the current study gives better estimates of genetic components of variance for rooting th an several previous studies. These results show a great deal of genetic variation for rooting among families and clones of loblolly pine. Only those cl ones that can be propagated easily will be economically feasible for deployment. Therefor e, selection of clones for rooting ability, as well as field performance should be consid ered as part of a cl onal forestry program based on rooted cutting technology. Comb ined with moderate to high estimates of familyand clonal-mean heritabilities and type B correlations, these results indicate the

PAGE 40

29 potential for increasing rooting efficiency by selecting good rooting families and clones or culling poor rooters.

PAGE 41

30 CHAPTER 3 GENETIC ANALYSIS OF EARLY FIELD GR OWTH OF LOBLOLLY PINE CLONES AND SEEDLINGS FROM THE SAME FULL-SIB FAMILIES Introduction Loblolly pine ( Pinus taeda L.) is the most important comm ercial tree species in the United States with over one billio n seedlings planted annually (McKeand et al. 2003). Most commercially important tree species rema in relatively undomesticated, and loblolly pine tree improvement programs are only now beginning their 3rd generation of breeding and testing (McKeand and Bridgewater 1998). Tree improvement programs for loblolly pine have relied on recurrent selection for general combining ability for improvement of a few key traits. These programs have hist orically utilized seedling progeny trials in order to predict breeding values for these tr aits. Traditional tree improvement programs using open-pollinated seed orchard seedlings for deployment only capture additive genetic variation. However, nonadditive genetic variation may be an important component of variation for some traits, a nd additive and nonadditive genetic variation can be captured by deploying full-sib families or clones. More efficient field-testing has been impl emented in order to gain information on full-sib families. For example, the Cooperativ e Forest Genetics Research Program at the University of Florida is using a partial dialle l mating design to cross slash pine selections to generate full-sib seedlings for progeny trials (Gezan et al. 2004). Tests established with full-sib seedlings allow the genetic va riance to be partitioned into additive and dominance components (Falconer and Mackay 19 96). These trials not only provide ranks

PAGE 42

31 of parents or individuals for selection, but also of full-sib families in order to provide information for making deployment decisions. For a number of reasons, clonally replicated progeny trials have been suggested as part of a tree improvement strategy for ra diata pine (Jayawickrama and Carson 2000) and for loblolly pine (Foster and Shaw 1987; Isik et al. 2004; Byram et al. 2004). First, field trials established with clonally replicated progeny allow for further partitioning of the genetic variation into the additive, dominan ce, and epistatic genetic components (Foster and Shaw 1988). Second, clonally propagat ed seedlings can provide genetic information more efficiently and with greater precisi on than zygotic seedling progeny (Burdon and Shelbourne 1974; Isik et al. 2004). Finally, clonal testi ng and selection strategies can provide greater gain for op erational deployment than s eedling options (Shaw and Hood 1985; Mullin and Park 1994; Isik et al. 2004). Based on current technologies, several fore st industries in the southeastern United States are pursuing clonal forestry programs w ith loblolly pine (Weber and Stelzer 2002). In the initial stages of these clonal forestry programs, forest managers needed assurance that the clonal propagules growth corresponde d to that of seedlings. Therefore, earlier studies were designed to test whether cuttings grew sim ilarly to seedlings. Based on those results, it is generally accepted that cutt ings rooted from juvenile stock plants grow and perform comparably to seedlings. For example, Foster et al. (1987) reported that loblolly pine rooted cuttings should perform comparably to seedlings when the cuttings come from vigorous juvenile stock plants. In addition, McRae et al. (1993) concluded that for loblolly pine there were no signif icant differences between seedlings and rooted cutting propagules from common checklots thro ugh five years of growth. Similar results

PAGE 43

32 were obtained by Frampton et al. (2000) where they reporte d no significant differences between the means of rooted cuttings and seed lings for height, diamet er at breast height, and volume through six years in the field. Trials established with clones and seedlings from the same families provide an opportunity for comparing both half-sib and full-sib family performances across propagule types. Genetic correlations between propagule types can provide further assurance that selections made through traditional tree improvement activities for recurrent selection for general combining ab ility in seedling tests can also be used successfully in breeding families to test in a clonal forestry program. Although a number of studies have compared rooted cutting a nd seedlings, very few have been designed to estimate the genetic correlation betw een propagule types for a trait. While clonal tests derived from full-sib families provide an opportunity to estimate additive and nonadditive components of varian ce, tests should be designed with sufficient genetic structure in order to precisely quan tify the genetic variation. Several clonal studies have reported deficiencies in ma ting designs, number of parents and families (Frampton and Huber 1995; Paul et al. 1997; Isik et al. 2003). For example, Frampton and Huber (1995) reported that they had lo w power in partitioning the genetic variation because of the lack of a mating design amon g the parents of the full-sib crosses in a loblolly pine clonal study. In addition, Frampton and Foster (1993) warned that interpretation of the results may be difficult for studies that only include seedlings and cuttings from a common checklot to be comp ared to the clonal propagules from select parents and families. In this case, any diffe rences in the field performance because of

PAGE 44

33 propagule type may be confounded with th e differences in genetic improvement (Frampton and Foster 1993). The current study employs a complex geneti c structure to increase the power in quantifying the genetic variation associated with several growth traits in loblolly pine. More than 1,200 clones together with zygotic seedlings from the same 61 full-sib families were tested on multiple sites across the south eastern United States. Because of the test and mating designs, genetic correlations can be directly calculated between propagule types and within propagule types across sites. The specific objectives of this study are to 1) determine heritability estimates for various growth traits for loblolly pine clones and seedlings, 2) compare the performance betw een parents and full-sib families when grown as rooted cuttings and seedlings, and 3) de termine the extent of genotype x environment interaction by looking at the genetic correlations for parents, families, and clones across multiple sites. Materials and Methods Population The parental population consisted of tw enty first-generation and ten secondgeneration selections, subset from the Loblolly Pine Lower Gulf Elite Population. Two additional first-generation, slow-growing pare nts were included. The parental selections represent the Atlantic Coastal Plain, Florid a, and Lower Gulf provenances of loblolly pine (see FBRC 2000 for details). Briefly, th ese thirty-two loblolly pine parents were mated in a partial diallel design and created 70 full-sib families from which more than 2,000 seedling hedges were generated and give n unique clonal identifications (Appendix A). On average, each parent was involved in approximately four crosses.

PAGE 45

34 Propagation The propagation of the rooted cuttings fo r the field trials has been previously described (Chapter 2; Baltunis et al. 2005). But briefly, the seedling hedges were repeatedly sheared to slow down the effects of maturation and incr ease the number of shoots available for collection. Cuttings were collected from seedling hedges from 61 full-sib families, placed randomly in clonal-ro w plots, and replicated six times in a greenhouse in January 2002, April 2002, and Ju ne 2002. At the time of collection, the hedges were 22, 25, and 27 months old from seed, respectively. Cuttings were assessed for rooting at 11-weeks after setting (Baltunis et al. 2005). Rooted cuttings were then transplanted into Ray Leach Supercells (Steuwe and Sons, Corvalis, Oregon) and randomized into their designated field pl anting order and grown to size. The clonal propagules for a field trial came from a single rooting trial (Table 3-1). Table 3-1. Location of six fiel d trials, establishment date and total number of test trees for each test. Test Sticking Date Location Latitude Longitude Date Planted Total No. Test Trees A January 2002 Worth County, Georgia 0 2 4 4 31 N 0 5 5 5 83 W October 2002 9,216 B April 2002 Morgan County, Georgia 5 5 4 2 33 N 5 4 9 2 83 W November 2002 8,960 C April 2002 Putnam County, Florida 8 3 29 N 6 4 81W November 2002 8,960 D April 2002 Nassau County, Florida 3 2 5 4 30 N 7 2 4 5 81 W February 2003 8,960 E April 2002 Randolph County, Georgia 3 0 8 4 31 N 0 3 1 4 84 W December 2002 4,400 F June 2002 Santa Rosa County, Florida 5 0 0 5 30 N 7 5 1 1 87 W April 2003 6,912

PAGE 46

35 A single crop of seedlings, on the other hand, was used to produce all the seedlings for the field trials. Loblolly pine seed from the same full-sib families that the clones were derived from were stratified for about 30 days and then sown in May 2002. The seedlings were grown in family blocks in a different greenhouse than the cuttings. The seedlings were then moved outdoors under shade cloth and kept in their family blocks in a separate area from where the rooted cuttin gs were growing. The seedlings were not randomized into their designated fiel d order until just prior to planting. Field Design In total 47,408 measurement trees were established at six field sites across the southeastern United States (Table 3-1). Three trials each were established in Florida and Georgia (Table 3-1). An additional field trial was established in Virginia with a subset of the clones but was not included in any of these analyses. There were four replications in each of two cultural treatments (high and low intensity) in each test, except for Test E where there was only one cultural treatment a nd four replications. The goal for the high intensity treatment was to push the trees to their utmost potential by reducing competition and providing a non-limiting suppl y of nutrients, while the lo w intensive culture provides insights into family and clonal performa nce under a less optimal cultural regime (FBRC 2000). Both cultural intensities were treated similarly during the first year with cultural differences implemented at the beginning of the second growing season. Each trial contained 756-974 clones with approximately 15 clones from each of 61 full-sib families (Table 3-2). In total, more than 1,200 clones were planted in field trials (Table 3-2). The trials were designed to ha ve four zygotic seedlings from each of the same 61 full-sib families within each replicat ion. However, because of poor germination for some of the families or mortality in the nu rsery, each full-sib family is represented, on

PAGE 47

36 average, by approximately 27 zygotic seedlings per test (Table 3-2). Both the rooted cuttings and seedlings were planted in singl e-tree plots utilizi ng a resolvable alpha incomplete block design (Williams et al. 2002) in which incomp lete block size ranged from 10-14 trees. The variables measured were 1st year height, 2nd year height, height increment, and crown width. Table 3-2. Total number of clones, full-sib families, half-sib families, average number of clones per full-sib and half-sib family, and average number of seedlings per full-sib and half-sib family established at the six field trials. Test A Test B Test C Te st D Test E Test F Total Total # clones 974 941 942 956 868 756 1,212 Total # FS families 61 61 61 61 61 61a 61 Total # HS families 32 32 32 32 32 32b 32 Ave. # clones/FS family 16 15.4 15.4 15.7 14.2 12.4 19.9 Ave. # clones/HS family 60.9 58.8 58.9 59.5 54.2 47.2 75.7 Ave. # seedlings/ FS family 27.9 27 27 27 15.2 35.1 151.2 Ave. # seedlings/ HS family 106.5 103 103 103 58 106.3 576.5 a Only 47 full-sib families and b 31 half-sib families are represented in the seedling population of Test F. Statistical Analyses All growth variables, 1st year height, 2nd year height, height increment, and crown width, were analyzed in ASREML (Gilmour et al. 2002) using a parental model in order to estimate the genetic variance components a ssociated with those traits. Analyses were conducted for each trial separate ly and across all six trials for each propagule type. The

PAGE 48

37 across-trial analyses assumed a different error variance for each trial. For the clonal population, the following across-trial model was used. [3-1] ijklmnop mn ij k n ij k m ij k mn ijo ijmn ijn ijm mn io imn in im mn o mn n m ijk l ij k ij i ijklmnope fam r gca r gca r clone c t fam c t gca c t gca c t clone t fam t gca t gca t clone sca gca gca incbk R C T T y ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) (* * * * * * * * where, ijklmnopy is the measured growth trait of the p th ramet of the o th clone within the mn th full-sib family in the l th incomplete block within the k th replication of the j th cultural treatment in the i th test is the clonal population mean iT is the fixed effect of trial, 6 , 1 i ijC T is the fixed effect of the interaction between trial and culture, 2 1 j ) ( ij kR is the fixed effect of replication, 4 3 2 1 k ) ( ijk lincbk is the random variable incomplete block associated with each test ~ NIID(0, 2iINC) mgca and ngca are the random variables female ( m ) and male ( n ) general combining ability, respectively ~ NIID(0, 2GCA) mnsca is the random variable specific combining ability ~ NIID(0, 2SCA) ) (mn oclone is the random variable clone w ithin full-sib family ~ NIID(0, 2CLONE)

PAGE 49

38 imgca t and ingca t are the random variables test by female general combining ability and test by male general combining ability interactions, respectively ~ NIID(0, 2TESTxGCA) imnfam t is the random variable test by full -sib family interaction ~ NIID(0, 2TESTxFAM) ) (*mn ioclone t is the random variable test by clone within full-sib family interaction ~ NIID(0, 2TESTxCLONE) ijmgca c t * and ijngca c t * are the random variables test by culture by female general combining ability and test by culture by male general combining ability, respectively ~ NIID(0, 2TxCxGCA) ijmnfam c t * is the random variable test by cu lture by full-sib family ~ NIID(0, 2TxCxFAM) ) (* *mn ijoclone c t is the random variable test by culture by clone within full-sib family interaction ~ NIID(0, 2TxCxCLONE) m ij kgca r) (* and n ij kgca r) (* are the random variables replication by female general combining ability and replication by male general combining ability, respectively ~ NIID(0,2REPxGCA) mn ij kfam r) (* is the random variable replication by full-sib family ~ NIID(0, 2REPxFAM) ijklmnope is the random error associated with each test ~ NIID(0, 2ERROR). The single-site model for the clonal populati on is identical, except that all model factors with subscript i are removed (sources involving test). All traits were also analyzed for th e seedling population assuming a randomized complete block design (incomplete block dr opped from model). Both single-trial

PAGE 50

39 analyses and an across-trial analysis we re performed again assuming heterogeneous errors across sites. [3-2] ijklmn lm ij k m ij k l ij k ijlm ijm ijl ilm im il lm m l ij k ij i ijklmne fam r gca r gca r fam c t gca c t gca c t fam t gca t gca t sca gca gca R C T T z ) ( ) ( ) ( ) (* * * * * * * where ijklmnz is the measured growth trait of the n th seedling from the lm th full-sib family in the k th replication of the j th cultural treatment in the i th test is the seedling population m ean and the other variables are defined as above (just adjusting the appropriate subscripts). Genetic parameters were estimated and st andard errors were calculated according to Foster and Shaw (1988) us ing the variance com ponents from the appropriate model. [3-3] AAA AA A GCA AV V V V 16 1 4 1 4 2 is the estimate of additive genetic variance. [3-4] DD AD AA D SCA DV V V V V 4 1 2 1 2 1 4 2 is the estimate of dominance genetic variance. [3-5] DD AD AA SCA GCA CLONE IV V V V 4 3 2 1 4 1 3 2 2 2 2 is the estimate of epistatic gene tic variance for the clonal population. [3-6] 2 2 2 2 CLONE SCA GCA GCV is the estimate of total genetic variance for the clonal population.

PAGE 51

40 [3-7] 2 2 4 4 SCA GCA GSV is the estimate of total genetic vari ance for the seedling population (assuming no epistasis). [3-8] 6 2 2 2 2 6 1 2 2 2 2 2 2 2 2 2 2 2 2 i ERROR REPxFAM REPxGCA TxCxCLONE TxCxFAM TxCxGCA TESTxCLONE TESTxFAM TESTxGCA CLONE SCA GCA Pi CV is the estimated phenotypic variance for the across-trial model for the clonal population. [3-9] 6 2 2 2 2 6 1 2 2 2 2 2 2 2 2 2 i ERROR REPxFAM REPxGCA TxCxFAM TxCxGCA TESTxFAM TESTxGCA SCA GCA Pi SV is the estimated phenotypic variance for the across-trial model for the seedling population. Individual tree narrow-sense heritability (2 h) and broad-sense heritability (2 H ) were derived using the estimated variance com ponents for all the growth traits and each propagule type at each site and across sites. Standard errors were calculated using Taylor series expansion (Kendall and Stuart 1963; Namkoong 1979; Huber et al. 1992; Dieters 1994). The following heritability formul ae were used for the clonal data. [3-10] C CP GCA P AV V V h 4 2 2 is the across-trial estimate of individual tree narrow-sense heritability for the clonal population, where CPV is from [3-8]. [3-11] C CP CLONE SCA GCA P GV V V H 2 2 2 2 2

PAGE 52

41 is the across-trial estimate of individual tree broad-sense heritability for the clonal population, where CPV is from [3-8]. Heritability estimates were also obtaine d for the seedling data for each trial and across-trials. [3-12] S SP GCA P AV V V h 4 2 2 is the across-trial estimate of individual tree narrow-sense heritability for the seedling population, where SPV is from [3-9]. [3-13] S SP SCA GCA P GV V V H 4 4 2 2 2 is the across-trial estimate of individual tree broad-sense heritability for the seedling population, where SPV is from [3-9]. The various growth variables at each site we re also analyzed with a bivariate mixed model with the growth of the clones and seedlings as two de pendent variables. Type B genetic correlations for general combining ability (propGCABr ) and full-sib family value (propFSBr ) between cuttings and seedlings were esti mated in order to compare parental and full-sib family performance between propagul e types. The geneti c correlation between propagule types for additive effects, for ex ample, gives us an indication of whether parental ranks are dependent upon whether their progeny are grown as cuttings or seedlings, while a type B genetic correlation at the full-sib family level measures the performance of full-sib families across propagule types.

PAGE 53

42 [3-14] SEED CUT SEED CUT BGCA Var GCA Var GCA GCA Cov rpropGCA is the type B genetic correlation of additive effects between propagule types using genetic variance component estimates fr om the bivariate analysis. [3-15] SEED SEED CUT CUT SEED CUT SEED CUT BSCA Var GCA Var SCA Var GCA Var SCA SCA Cov GCA GCA Cov rpropFS 2 2 , 2 is the type B genetic correlation of full-sib family values between propagule types using genetic variance component estimates from the bivariate analysis. The extent of genotype x environment in teraction was investigated by analyzing data across trials for each propagule type using the variance components from the appropriate model. Type B gene tic correlations were calculated for additive effects, fullsib family, and the total genetic or clonal value across the trials (Yamada 1962; Burdon 1977). Standard errors of type B correlati ons were calculated using the Taylor series expansion method. [3-16] 2 2 2 TESTxGCA GCA GCA BgxeGCAr is the type B genetic correlation for additive effects across trials. [3-17] 2 2 2 2 2 2 2 2 2 TESTxFAM SCA TESTxGCA GCA SCA GCA BgxeFSr is the type B genetic correlation for full-sib families across trials. [3-18] 2 2 2 2 2 2 2 2 2 2 2 2 TESTxCLONE CLONE TESTxFAM SCA TESTxGCA GCA CLONE SCA GCA BgxeCLONEr is the type B genetic correlation for total ge netic or clonal value across trials for the clonal population.

PAGE 54

43 Ranging from 0 to 1, a value of gxeGCABr near one indicates little genotype x environment interaction and that the parents ranked the same across the trials, while a low gxeGCABr (near zero) indicates that parental ra nks were not stable across the sites and hence, genotype x environment interaction exists. A high gxeFSBr indicates that full-sib families performed similarly across the sites, while a highgxeCLONEBr indicates that the total genetic values of the clones were stable across the trials. Results and Discussion Overall Growth of Clones and Seedlings Survival of both clonal a nd zygotic propagules was hi gh across all of the field trials. Survival of the rooted cuttings range d from 91.7-98.3% at the six field trials, while seedling survival ranged from 86.5-97.4%. At the time of planting, seedlings were generally taller than rooted cuttings, and th is trend has continued through year two (Table 3-3). For example, after th e first and second growing seas ons seedlings were on average 11 cm and 10 cm taller than rooted cuttings, respectively. It has been suggested that propagule size differences at the time of plan ting may create difficulties in the analysis and interpretation of subsequent growth (F rampton and Foster 1993). However, these differences in initial propagule size may not be a problem in the current study because height increments were very similar betw een both propagule types, with site means ranging between 1.0-2.0 m and 1.0-1.9 m for r ooted cuttings and seedlings, respectively (Table 3-3). The initial and first year treatments we re the same within a site during test establishment, and there were no cultural differences between propagule types for 1st year height. There were some differences in m ean height due to the effects of cultural

PAGE 55

44 intensity during the seco nd growing season at some of the sites. However, these effects were more a function of scale. Although the overall means of the growth variables differed by cultural treatment, the ranks of pare nts, families, or clones were not affected. Type B genetic correlations exceeded 0.85 i ndicating little cultural treatment x genetic effect interaction, and therefor e, cultural effects ar e ignored for the purposes of this study. This implies that the rankings of parents, families, and clones were robust across multiple management regimes through age two. Table 3-3. Mean 1st year height, 2nd year height, height incr ement, and crown width by propagule type for each of the six fiel d trials. Although means are expressed in meters, analyses were conducted us ing measured traits in centimeters. 1st Year Height (m) 2nd Year Height (m) Height Increment (m) Crown Width (m) Clones 1.0 2.1 1.2 0.9 Test A Seedlings 1.1 2.3 1.2 1.0 Clones 0.8 1.8 1.0 0.9 Test B Seedlings 0.9 1.8 1.0 0.9 Clones 1.0 2.1 1.1 1.2 Test C Seedlings 1.1 2.3 1.2 1.3 Clones 0.7 2.0 1.3 1.0 Test D Seedlings 0.8 2.1 1.3 1.1 Clones 1.2 3.3 2.0 2.0 Test E Seedlings 1.3 3.2 1.9 2.0 Clones 0.5 1.7 1.3 1.1 Test F Seedlings 0.6 1.8 1.3 1.1 Genetic Components of Variance All of the early growth traits demonstrated genetic variat ion (Figure 3-1; Table 3-4; Appendix C). However, the genetic variat ion partitioned differently for the two propagule types. In all cases, the estimate of additive genetic vari ation was greater for

PAGE 56

45 the clones than the seedlings. Within a si ngle trial, for instance, the estimate of the additive genetic variation for 2nd year height based on seedlings was 0 to 0.58 that of additive genetic variation based on clonal data Similar trends were observed from the across-trial analyses. The estimate of add itive genetic variation for the clonal material was 4.7, 3.3, 2.9, and 2 times greater than th e additive genetic vari ation for seedlings for 1st year height, 2nd year height, height increment, and crown width, respectively (Table 34). Based on single-trial analyses, the majority of the genetic varia tion associated with all of the growth variables in the clonal population was a dditive, while in the seedling population the trend was towards nonadditive genetic variation (Figure 3-1). For example, at Test D all of the ge netic variation associated with 2nd year height of clones was additive, while for the seedlings it was dominance genetic variation (Figure 3-1). When all of the data were analyzed togeth er, then the estimates of dominance genetic variation were approximately equivalent for both propagules (Table 3-4) suggesting a large test by dominance interaction for the seedlings. Epistasis was negative for all growth variables. As a result, estimates of additive and dominance genetic variance for these traits might not be upwardly biased as indicated by the expected portions of epistatic interactions defined in Equations 3-3 and 3-4 (Foster and Shaw 1988). Isik et al. (2003) reported similar trends in the partitioning of the ge netic variation for clones and seedlings including negative epistasis estima tes for height, diameter, and volume through age six. They also reported that the add itive genetic variation for growth traits was always greater for clones than the estim ate for seedlings, while dominance genetic variation was greater in the seedling population (Isik et al. 2003).

PAGE 57

46 Table 3-4. Genetic parameter estimates for 1st year height, 2nd year height, height increment, and crown width by propagule t ype across all six tr ials. Standard errors are given in parentheses. 1st Year Height 2nd Year HeightHeight Increment Crown Width Clone Seed Clone Seed Clone Seed Clone Seed AV 69.8 (21.3) 14.7 (7.8) 401.5 (122) 119.9 (50.8) 152.6 (47) 51.8 (24.2) 110.3 (34) 55 (20) DV 12 (6.2) 17.5 (9.3) 86.4 (36.5) 83.3 (41.6) 30.3 (13.6) 41.9 (21.5) 19.7 (10.1) 20 (11.7) IV -11.5 (11.4) ---107.6 (64.3) ---39.7 (24.8) ---13.5 (18.1) --GV 70.3 (10.8) 32.1 (9.5) 380.3 (61.1) 203.1 (53.8) 143.2 (23.6) 93.65 (26.1) 116.5 (17.2) 75 (20.7) PV 442.8 (11.3) 405.2 (7.6) 1806 (62.3) 1703 (36.2) 902.6 (24.5) 972 (19.4) 640.9 (17.8) 649.5 (14.3) 2 h 0.16 (0.04) 0.04 (0.02) 0.22 (0.06) 0.07 (0.03) 0.17 (0.05) 0.05 (0.02) 0.17 (0.05) 0.08 (0.03) 2 H 0.16 (0.02) 0.08 (0.02) 0.21 (0.03) 0.12 (0.03) 0.16 (0.02) 0.1 (0.02) 0.18 (0.02) 0.11 (0.03) gxeGCABr 0.81 (0.06) 0.71 (0.2) 0.88 (0.04) 0.78 (0.12) 0.83 (0.05) 0.6 (0.15) 0.82 (0.06) 0.78 (0.12) gxeFSBr 0.8 (0.05) 0.5 (0.1) 0.88 (0.04) 0.66 (0.09) 0.83 (0.05) 0.58 (0.1) 0.82 (0.05) 0.67 (0.09) gxeCLONEBr 0.69 (0.04) --0.77 (0.03) --0.76 (0.04) --0.76 (0.03) --Genetic causes are not the only source for similarity among relatives. Non-genetic factors, such as C effects, can lead to upw ardly biased estimates of total genetic and nonadditive genetic components of variance when analyzing clonal data (Libby and Jund 1962). C effects are often assumed negligible when estimating epistatic variance (Foster and Shaw 1988). If C effects are present then the total genetic variat ion associated with

PAGE 58

47 clones will be overestimated (Libby and Jund 1962) Significant C eff ects are likely to occur in traits that are measured soon after propagation, but apparent ly lessen for traits measured at later times (Libby and Jund 1962). In the current experiment, if C effects exist, then estimates of ep istatic genetic variation will be confounded with C effects because ramets of a clone came from a singl e, non-replicated hedge. However, the estimates for epistasis were negative for all of the growth traits (Table 3-4), which suggest that interloci interactions are not an important source of genetic variation for early growth traits, and that C effects may not be a major contributing factor. Further, estimates of additive and dominance shoul d be free of confounding C effects in the current experiment since randomization of cl ones occurred at all stages including hedge establishment, propagation, and growth pr ior to and after test establishment. As Falconer and Mackay (1996) point out, re latives of all sorts may resemble one another because of sharing a common environment. The variance attributed to a common environment occurs more frequently and contri butes greater to the c ovariance of full-sibs than to the covariance of any other sort of relatives (Falconer and Mackay 1996). When a common environment effect exists, then the differences between means of families become greater than when these non-genetic fact ors are not present. In the case of fullsib families, then this will lead to biased or inflated estimates of dominance genetic variation. The results seen in the current experiment for dominance genetic variance estimates in the seedling population appear to be a result of a partitioni ng problem in that estimates of the variance due to dominance are inflated at the expense of addi tive effects (Figure 31). Seeds are often sown and seedlings grown in full-sib family blocks in the greenhouse

PAGE 59

48 Figure 3-1. The proportion of additive (a dd) and nonadditive (na) genetic variance components for clones and seedlings across the six field trials, where 2h = add and 2 H = add + na: a. 1st year height, b. 2nd year height, c. Height increment, and d. Crown width. Standard error bars for broad-sens e heritability estimat es are included. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7clone seed clone seed clone seed clone seed clone seed clone seed Test ATest BTest CTest DTest ETest F Broad-Sense Heritabilit y na add a. 1st year height 0 0.1 0.2 0.3 0.4 0.5 0.6clone seed clone seed clone seed clone seed clone seed clone seed Test ATest BTest CTest DTest ETest F Broad-Sense Heritabilit y na add b. 2nd year height 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45clone seed clone seed clone seed clone seed clone seed clone seed Test ATest BTest CTest DTest ETest F Broad-Sense Heritabilit y na add c. Height increment 0 0.1 0.2 0.3 0.4 0.5 0.6clone seed clone seed clone seed clone seed clone seed clone seed Test ATest BTest CTest DTest ETest F Broad-Sense Heritabilit y na add d. Crown width

PAGE 60

49 or nursery for progeny testing or other field testing. In order to eliminate the common environment effect at the family level, then seedlings within a full-sib family should be randomized over the environment in which th ey are being grown and tested. Although rooted cuttings were randomized at all stages of propagation, the seedlings were in fact grown in full-sib blocks, and apparently a co mmon environment effect has carried over to the field through age two measurements. Heritability Estimates First year height, 2nd year height, height increment, and crown width were all influenced by additive genetic variation. Individual tree narrow-sense heritability was always greater for the clones than for seedlings for all of the traits (Figure 3-1, Table 34). Generally, 2 h estimates of total height increased from age one to age two for both clones and seedlings (Figure 3-1, Table 3-4). Narrow-sense heritability for total height increased from 0.16 at age one to 0.22 at ag e two for the clones, while in the seedling population, 2 h increased from 0.04 to 0.07 (Table 3-4). In addition, 2 h estimates for all of the growth traits from the seedling data had larger standard errors associated with them than the estimates from clones (Table 3-4). Ye t, the differences in heritability estimates between propagule types may be related to C effects. If C effects are present, then 2 H estimates from the clonal population may be inflated. However, estimates of 2 h from the clonal population may not be upwardly biased si nce clones within full -sib families were randomized in the hedge orchard and thr oughout the study, thus reducing spurious C effects at the parental and full-sib family levels. Heritability estimates have been reported for several loblolly pine seedling and clonal populations. In a study reported by Isik et al. (2003), individual tree narrow-sense

PAGE 61

50 heritability for volume at age six was 0.3 for clones, while for seedlings 2 h was 0.06. In another loblolly pine clonal te st, heritability estimates for total height growth through age five for two factorials were simila r to the values reported here (Paul et al. 1997). They reported 2 h for 1st and 2nd year height as 0.08 and 0.17, respectively (mean of two factorials). Unfortunately, they did not include seedlings in their study. Analogous results have been reported in Eucalyptus globulus Labill. for heritability estimates for diameter for both clones and seedlings in that 2 h was always greater for the clonal population than the seedling population (Costa e Silva et al. 2004). Broad-sense heritability estimates for the va rious growth traits were always larger for the clonal material than the se edlings except for the estimates of 2 H for 1st year height at Test E and for all four variab les at Test D (Figure 3-1). However, 2 H estimates from the across-trial models were always great er for the clones than seedlings (Table 34). Because of the negative estimates of epistasis in the clonal population, 2 H estimates were equivalent to the estimates for individual tree narrow-sense heritability for all traits. As was the case with2 h, seedling estimates of 2 H had higher standard errors associated with them (Figure 3-1). Type B Genetic Correlations Between Propagule Types In order to further compare the clonal rooted cuttings and zygotic seedlings, type B genetic correlations for general combining ab ility and full-sib family value between propagule types were estimated. For all gr owth traits measured, the type B genetic correlation between propagule types for additiv e effects was high with values exceeding 0.72 (Table 3-5). For example, propGCABr for 1st year height ranged from 0.93-0.99, while

PAGE 62

51 for height increment propGCABr ranged from 0.72-0.99 (Table 3-5). These high genetic correlations imply that parental rankings for ear ly growth traits are stable regardless of whether their progeny are being tested as zygot ic seedlings or rooted cuttings (Figure 32a). Table 3-5. Genetic correlati ons between propagule types for 1st year height, 2nd year height, height increment, and crown width at the parental (propGCABr ) and full-sib family (propFSBr ) levels. 1st Year Height 2nd Year HeightHeight Increment Crown Width Test propGCABr propFSBr propGCABr propFSBr propGCABr propFSBr propGCABr propFSBr A 0.99 0.62 0.99 0.65 0.99 0.38 0.86 0.86 B 0.99 0.74 0.99 0.86 0.99 0.99 0.99 0.81 C 0.93 0.74 0.99 0.75 0.99 0.99 0.99 0.97 D 0.93 0.44 0.99 0.64 0.99 0.92 0.93 0.92 E 0.99 0.55 0.89 0.71 0.72 0.77 0.99 0.69 F 0.95 0.49 0.96 0.89 0.95 0.83 0.99 0.83 The type B genetic correlations for a dditive effects betw een propagule types observed in the current study were consistent with the expectations reported by Borralho and Kanowski (1995). In a simulation study comparing the perfor mance of clones and seedlings from the same half-sib family, Bo rralho and Kanowski ( 1995) reported that the expected correlations between propagule t ypes exceeded 0.8 when greater than 100 seedlings or propagules were tested. Additionally, in a field study comparing rooted cuttings and seedlings from four half-sib families, Foster et al. (1987) reported that family rank correlations between propagule types were positive and significant for 1st year height (0.52), 3rd year height (0.66) and 6th year height (0.70).

PAGE 63

52 Full-sib families also ranked relatively similar regardless of propagule type at most of the sites (Table 3-5; Figure 3-2b). For ex ample, type B genetic correlations at the fullsib family level ranged from 0.64-0.89 for 2nd year height. However, propFSBr was more variable from site to site for 1st year height and height incr ement. The results observed here are in accordance with those reported by Frampton et al. (2000) in that full-sib families or parental trees selected based on seedling genetic trials should also perform well as rooted cuttings. Genotype x Environment Interaction The stability of parents and full-sib families across sites was compared for the clonal and seedling populations (Table 3-4). The type B genetic correlation for additive effects across sites was always greater for the clones, although these estimates were moderately high for both populations. For example, gxeGCABr for 2nd year height was 0.88 and 0.78 for the clonal and seedling populati ons, respectively (Table 3-4). Full-sib families also ranked comparably across sites for the clonal population with gxeFSBr values exceeding 0.8 for all of the early growth trai ts (Table 3-4). These across site genetic correlations were estimated with more precisi on using clonal replicates as evidenced by lower standard errors for the estimates from the clonal population. An additional genotype x environment inte raction between test and total genetic value can be estimated using clonal tests. The total genetic values of the clones were fairly stable when all of the data was cons idered from all sites, indicating that a good clone at one site is good at al l the sites (Table 3-4). Howe ver, based on analyses from pairs of trials (data not show n), there appears to be a propa gation effect relating to the season the cuttings were rooted. For exam ple, the worst genetic correlations were

PAGE 64

53 observed between the trial esta blished with rooted cuttings from the winter setting and any of the other trials, while the best genetic correlations were obtained from the field trials that contained rooted cuttings originat ing from the spring setting. Similar results Figure 3-2. Rank-rank plots showing type B genetic corre lations between clones and seedlings from Test B based on: a. Pare ntal BLUP values, b. Full-sib family BLUP values, where full-sib BLUP values are equal to the sum of the predicted general combining ability fo r each of the two parents plus the predicted specific combining ability of the cross. 0 5 10 15 20 25 30 35 05101520253035 Test B Parental Ranks ClonesTest B Parenatl Ranks Seedlings a. Parental ranks for 2nd year height 0 10 20 30 40 50 60 70 010203040506070 Test B Full-Sib Family Ranks ClonesTest B Full-Sib Family Ranks Seedlings b. Full-sib family ranks for 2nd year height 99 0 propGCABr 86 0 FS propBr 86 0 propFSBr

PAGE 65

54 were also observed with rooting with this same popul ation in that poor genetic correlations were observed between rooting ab ility in winter and spring (Chapter 2; Baltunis et al. 2005). Conclusion Several forest industries in the southeaste rn United States ar e deploying full-sib families of loblolly pine operationally. In addition, many of these companies are pursuing clonal forestry programs with loblolly pine. Genetic field trials established with clones and seedlings from the same full-sib families provide an opportunity for comparing both half-sib and full-sib family performances for both propagules. Based on the current study, several conclusions can be dr awn. First, clonally replicated seedling trials of loblolly pine provide genetic in formation with greater precision than zygotic seedlings. Second, genetic correlations between estimates of the genetic effects associated with these growth traits between these propagule types were highly favorable. These high genetic correlations between propagul e types reassure that parental and fullsib family rankings are stable regardless of propagule type. This implies that parental and full-sib family rankings based on existi ng seedling progeny trials could be used to select parents and families that perform well when they are deployed as rooted cuttings. Third, little genotype x envir onment interaction was observed across sites at the parental, family, and clonal level for all traits. Howeve r, there appears to be a carry-over effect relating to the season in which the cuttings we re rooted for the clonal material. Finally, randomization is essential at all stages in testing when estimating genetic parameters. The lack of randomization for the seedling popu lation apparently resulted in a problem with partitioning of the genetic variance, causing full-sib families to appear more different and inflating estimates of dominance genetic variation.

PAGE 66

55 CHAPTER 4 GENETIC GAIN FROM SELECTION FO R ROOTING ABILITY AND EARLY GROWTH IN VEGETATIVELY PROPAGAT ED CLONES OF LOBLOLLY PINE Introduction Loblolly pine ( Pinus taeda L.) is the most important co mmercial tree species in the southeastern United States (McKeand et al. 2003). Genetic improveme nt of loblolly pine has been occurring since the 1950s in seve ral tree improvement programs. These programs have aimed to increase the populati on mean breeding value of a few key traits, such as stem volume, disease resistance, and wood properties, through breeding and selection of superior genot ypes. Tree improvement pr ograms are based on recurrent selection for general combining ability and ca pture only a portion of the additive genetic variation with open-pollinated seedlings fo r deployment. However, the nonadditive portion of genetic variation, dominance and ep istasis, may be important components of variation for traits. For example, Stonec ypher and McCullough (1986) reported that the estimates of nonadditive variance were approx imately equal to those of additive variance for growth traits in Douglas-fir ( Pseudotsuga menziesii (Mirb.) Franco) through age six. Additionally, Paul et al. (1997) reported that both a dditive and dominance genetic variance increased from age one to five for he ight growth for loblolly pine clones. The only manner in which to capture the tota l genetic variation is through operational deployment of clonal propagules.

PAGE 67

56 Based on current technologies, several forest industries in the southeastern United States are pursuing clonal fore stry programs for loblolly pine using either somatic embryogenesis or rooted cuttings (Weber and Stelzer 2002). Two main criteria need to be met prior to operational deployment of loblolly pine clones. First, loblolly pine clones must perform well, e.g. meet the selection criteria for the desired traits. This involves the accumulation of reliable data for the cl ones from greenhouse screening, field trials, etc. Second, the selected clones have to be propagate d in large enough numbers for deployment. For a rooted-cutting-based clonal program, this involves bulking up the number of hedges (ramets) of a particular cl one or group of selected clones through serial propagation and then producing reforestation stock efficiently from the bulked-up clones. Only those tested clones that can be propa gated easily in sufficient numbers will be economically feasible for deployment. Loblolly pine is considered a difficult to root species (Wise an d Caldwell 1994). In populations that have not experi enced any selection pressure fo r rooting ability, loblolly pine has been reported to root near 50% (Foster 1990; Baltunis et al. 2005). Previous rooting studies with loblolly pine have demonstrated substantial family and clonal variation for rooting ability (Foster 1990; Baltunis et al. 2005), indicati ng the potential for increasing rooting efficiency for both clonal deployment and through recurrent selection and breeding. Sel ection for both rooting ability and field growth will be necessary for a successful loblolly pine cl onal forestry program ba sed on rooted cutting technology (Foster et al. 1985; Foster et al. 2000; Baltunis et al. 2005). The objectives of this study were to (i ) determine the genetic correlation between rooting ability and 2nd year height, (ii) predict the gene tic gain associated with selection

PAGE 68

57 for rooting ability, (iii) predict the genetic gain associated with selection for 2nd year height, and (iv) predict the ge netic gain from combined sele ction for rooting ability and 2nd year height using a Mont e Carlo selection index. Materials and Methods Population The parental population consisted of tw enty first-generation and ten secondgeneration selections from the larger Loblolly Pine Lower Gulf Elite Population. Two additional first-generation, slow-growing parent s were included. The parental selections represent the Atlantic Coastal Plain, Florid a, and Lower Gulf provenances of loblolly pine. These thirty-two loblolly pine parent s were mated in a partial diallel design and created 70 full-sib families from which approximately 2,200 seedling hedges were generated and given unique clonal identificati ons (Appendix A). On average, each parent was involved in about four crosses. Rooting and Field Trials The propagation of the rooted cuttings for this study has previously been described (Baltunis et al. 2005; Chapter 2). But briefly, th e seedling hedges were repeatedly sheared to slow down the effects of matu ration and increase the number of shoots available for collection. Cuttings were set in five rooting trials over two years in May 2001, July 2001, January 2002, April 2002, and June 2002 in trials Spring01, Summer01, Winter02, Spring02, and Summer02, respectively Four to nine ramet clonal row plots were set in four to six replications depe nding on trial (Chapter 2) At the time of collection, the hedges were 13, 15, 22, 25, and 27 months old, respectively. Cuttings were assessed for rooting nine to elev en weeks after they were set (Baltunis et al. 2005; Chapter 2).

PAGE 69

58 Six field trials (A, B, C, D, E, F) were established with rooted cuttings from the three latter rooting trials. Th ree field trials each were esta blished in Georgia and Florida between October 2002 to April 2003 (see Table 3-1). The clonal propagules for each field trial, however, came from a single rooti ng trial. Single-tree-p lots of a clone were established in four replications in each of two cult ural treatments (high and low intensity) in each test, except for Test E where there was only one cultural treatment and four replications. The goal for the high intensit y treatment was to push the trees to their utmost potential by reducing competition and providing a non-limiting supply of nutrients, while the low intensive culture provides insights into family and clonal performance under a less optimal cultural re gime (FBRC 2000). Previous analyses of growth for this population had shown that type B genetic correlations exceeded 0.85 indicating little cultural treatme nt x genetic effect interact ion, and therefore, cultural treatment effects were ignored for th e purposes of this study (Chapter 3). Statistical Analyses A bivariate parental linear mixed-effects model was used to obtain estimates of variance and covariance component s for rooting ability and 2nd year height using ASREML (Gilmour et al. 2002): [4-1] i i q i p i o i c i f i u i n i i ii i i i i i ie q Z p Z o Z c Z f Z u Z n Z b X y where iy is the vector of ob servations indexed ( i ) by rooting ability and 2nd year height, ib is the vector of fixed effects (i.e. mean, trials and replications within trials) and iX is the known incidence matrix re lating the observations in iy to the fixed effects in ib where height root height root i ib b X 0 0 X b X,

PAGE 70

59 in is the vector of random incomplete blocks nested within replication and test effects ~MVN 2 2 ,height rootINC height INC root I 0 0 I 0 iu is the vector of random parent (female and male) general combining ability effects ~MVNA G 0 where 2 2 height rootheight rootheight rootGCA GCA GCA GCA G and A= numerator relationship matrix, if is the vector of random specifi c combining ability effects ~MVNsI S 0, where 2 2 height rootheight rootheight rootSCA SCA SCA SCA S and sI is an identity matrix of size equal to the number of full-sib families, ic is the vector of random clones wi thin full-sib family effects ~MVNcI C 0, where 2 2 height rootheight rootheight rootCLONE CLONE CLONE CLONE C and cI is an identity matrix of size equal to the number of clones, io is the vector of random test by full-sib family effects ~MVN 2 2 ,height rootTESTxFAM height TESTxFAM root I 0 0 I 0, ip is the vector of random test by cl one within full-sib family effects ~MVN 2 2 ,height rootTESTxCLONE height TESTxCLONE root I 0 0 I 0,

PAGE 71

60 iq is the vector of random replication w ithin test by full-sib family effects ~MVN 2 2 ,height rootREPxFAM height REPxFAM root I 0 0 I 0, ie is the random vector of residual terms ~MVN 2 2 ,height rootERROR height ERROR root I 0 0 I 0, inZ, iuZ, ifZ, icZ, ioZ, ipZ, and iqZ are the known incidence matrices relating the observations in iy to effects in in, iu, if, ic, io, ip, and iq, respectively, and iI is the identity matrix of dimension equal to the num ber of observations for rooting ability or 2nd year height. Causal Components of Variance Genetic parameters were estimated and sta ndard errors were calculated for rooting ability and 2nd year height from the bivariate an alysis according to Foster and Shaw (1988). [4-2] AAA AA A GCA AV V V V 16 1 4 1 4 2 is the estimate of additive genetic variance. [4-3] DD AD AA D SCA DV V V V V 4 1 2 1 2 1 4 2 is the estimate of dominance genetic variance. [4-4] DD AD AA SCA GCA CLONE IV V V V 4 3 2 1 4 1 3 2 2 2 2 is the estimate of epistatic genetic variance. [4-5] 2 2 2 2 CLONE SCA GCA GV is the estimate of total genetic variance.

PAGE 72

61 [4-6] 2 2 2 2 2 2 2 ERROR REPxFAM TESTxCLONE TESTxFAM CLONE SCA GCA PV is the estimate of the phenotypic variance. [4-7] trfcn tr tfc tf fc f VERROR REPxFAM TESTxCLONE TESTxFAM CLONE SCA GCA PHS2 2 2 2 2 2 2 is the phenotypic variance of half-sib family means, f = harmonic mean number of fullsib families per half-sib family, c = harmonic mean number of clones per full-sib family, t = number of trials, r = harmonic mean number of replications per test, and n = harmonic mean number of ramets per cl one per replica tion per test ( n = 1 for field trials). [4-8] ctrn tr tc t c VERROR REPxFAM TESTxCLONE TESTxFAM CLONE SCA GCA PFS2 2 2 2 2 2 2 2 is the phenotypic variance of full-sib family means. [4-9] trn tr t t VERROR REPxFAM TESTxCLONE TESTxFAM CLONE SCA GCA PCL2 2 2 2 2 2 2 2 is the phenotypic variance of clonal means. Heritability Estimates Heritability estimates have previously been reported for this population for rooting ability (Baltunis et al. 2005; Chapter 2) and 2nd year height (Chapter 3). However, since rooting ability and 2nd year height were analyzed usi ng a bivariate model in the current study, heritabilities were estimated based on the genetic parameter estimates from the bivariate model and differed slightly from t hose reported previously. Standard errors were calculated using the Taylor series expansion method (Kendall and Stuart 1963; Namkoong 1979; Huber et al. 1992; Dieters 1994).

PAGE 73

62 [4-10] 2 2 2 2 2 2 2 2 2 2 4 ERROR REPxFAM TESTxCLONE TESTxFAM CLONE SCA GCA GCA P AV V h is the individual tree narrow-sense heritability. [4-11] 2 2 2 2 2 2 2 2 2 2 2 2 2 ERROR REPxFAM TESTxCLONE TESTxFAM CLONE SCA GCA CLONE SCA GCA P GV V H is the individual tree broad-sense heritability. [4-12] HSP GCA HSV H 2 2 is the half-sib family mean heritability. [4-13] FSP SCA GCA FSV H 2 2 2 2 is the full-sib family mean heritability. [4-14] CLP CLONE SCA GCA CLV H 2 2 2 2 2 is the clonal mean heritability. Type B Genetic Correlations Although the same genotypes, or clones, were tested in the rooting and field trials, these measurements were not necessarily taken on the same ramet. Therefore, type B genetic correlations between rooting ability and 2nd year height for general combining ability, full-sib family value, and the total ge netic value were calcula ted. Standard errors of these estimates were calculated using the Taylor series expansi on method (Kendall and Stuart 1963; Namkoong 1979; Huber et al. 1992; Dieters 1994). [4-15] 2 2 height root rootheight GCAGCA GCA GCA Br is the type B genetic correl ation between rooting ability and 2nd year height for additive effects, and rootheightGCA is the covariance between general combining ability effects for rooting ability and 2nd year height..

PAGE 74

63 [4-16] 2 2 2 2 2 2 2 height height root root rootheight rootheight FSSCA GCA SCA GCA SCA GCA Br is the type B genetic correlation between rooting ability and 2nd year height for full-sib families, and rootheight rootheightSCA GCA 2 is the covariance between the full -sib family effects for rooting ability and 2nd year height. [4-17] 2 2 2 2 2 2 2 2 height height height root root root rootheight rootheight rootheight TGCLONE SCA GCA CLONE SCA GCA CLONE SCA GCA Br is the type B genetic correlation between rooting ability and 2nd year height for the total genetic value of clones, and rootheight rootheight rootheightCLONE SCA GCA 2 is the covariance between the total clonal value effects for rooting ability and 2nd year height. Genetic Gain Genetic gain was estimated for a number of deployment options based on various selection scenarios using the BLUP values from the bivariate analysis. All deployed populations were assumed to be propagated as rooted cuttings. An additional assumption for gain calculations was that the seventy fu ll-sib families and all clones within families were available for deployment. The deploymen t strategies considered were 1) half-sib family deployment, 2) full-sib family deploymen t, and 3) clonal deployment. For a halfsib family deployment option, the predicted value for each parent (general combining ability) was used to select the best half-sib family for deployment as rooted cuttings. Full-sib family values were calculated by summing the predicted values for the female general combining ability, male general combining ability, and the full-sib family specific combining ability. The genetic gain over the trait mean was determined for deployment of the best full-sib family.

PAGE 75

64 The total genetic value of a clone was de termined by summing the predicted values for the female general combining ability, ma le general combining ability, the full-sib family specific combining ability, and clone within full-sib family. Several clonal deployment strategies were compared. Firs t, the genetic gain was determined for deployment of the single best clone for each tr ait. A second clonal deployment strategy was considered by selecting the top ten full-sib families and then deploying the single best clone from each of these ten families. Two additional clonal deployment strategies were compared by selecting the top 10% and 1% of clones (out of 2200 possible clones) using an unrestricted selection index for 2nd year height and roo ting ability. The total genetic values of the clones were weighted with the following weights: 1:0, 0.9:0.1, 0.8:0.2, ..., 0.2:0.8, 0.1:0.9, 0:1 for 2nd year height and rooting ab ility, respectively, in the Monte Carlo index (Cotterill and Dean 1990) All gains were expressed as the percentage gain over the mean of the trait: [4-18] % 100 % i i iy y x Gain, where ix is the average predicted value for trait i of the selected population, and iy is the population mean for eith er rooting ability or height. In addition genetic gain was calculated us ing theoretical gain formulae assuming an unrelated population (Falcone r and Mackay 1996), and these estimates were compared with the predicted genetic gain based on BLUP values for the half-sib family, full-sib family, and best clone deployment opti ons. The following formulae were used:

PAGE 76

65 [4-19] HSP HS hsV H i Gain 2 is the genetic gain associated with selection of the best halfsib family, and hsi is the selection intensity corresponding to selecting one out of thirtytwo half-sib families (i = 2.07) and assuming all half-sib families are unrelated, [4-20] FSP FS fsV H i Gain 2 is the genetic gain associated with selection of the best fullsib family, and fsi is the selection intensity correspo nding to selecting one out of seventy full-sib families (i = 2.38) and assuming all full-sib families are unrelated, [4-21] CLP CL cV H i Gain 2 is the genetic gain associated with clonal selection, and ci is the selection intensity co rresponding to selecting one out of 2,206 clones (i = 3.58) and assuming all individu als are unrelated. Results and Discussion Causal Components of Variance The genetic parameter estimates from the bi variate analysis of rooting ability and 2nd year height in loblolly pine clones we re consistent with the estimates from their respective univariate analyses and both traits showed gene tic variation (Table 4-1; Appendix D). Additive genetic variation accounted for the majority of the genetic variation associated with rooting ability and 2nd year height (Table 4-1). Dominance genetic variation contributed a minor portion to the total genetic varia tion for both traits. Epistasis appears to be more important for rooting ability than 2nd year height as evidenced by a negative estimate of epistasis for 2nd year height (Table 4-1). However, estimates of epistasis may be confounded with C effects since rooted cuttings of a clone originated from a single ortet.

PAGE 77

66 Type B Genetic Correlations The bivariate analysis of five rooting tria ls and six field trials also allowed for estimation of the genetic covariance between rooting ability and 2nd year height for parental effects, full-sib family effects, a nd the total genetic value of clones within fullsib family. There was a positive genetic re lationship between rooting ability and 2nd year height at all three genetic leve ls (Table 4-1). The genetic co rrelation at the parental level between rooting ability and 2nd year height (GCABr ) was 0.32. At the full-sib family level, the genetic correlation between traits (FSBr ) was 0.39. The correlation of total genetic values of clones for rooting ability and 2nd year height (TGBr ) was 0.29. Previous studies have also reported positiv e correlations between rooting traits and growth. For example, Paul et al. (1993) reported that 1st year and 5th year height growth had strong genetic correlations with rooti ng (0.61 and 0.69, respectively) in western hemlock clones (Tsuga heterophylla (Raf.) Sarg.). In another study, Foster et al. (1985) reported that the genetic correlation between rooting ability and growth of western hemlock clones in a growth chamber was 0.37. A weak, but positive, relationship between the number of roots on loblolly pine cuttings and subsequent field growth has been reported (Foster et al. 2000). On the other hand, Goldfarb et al. (1998) reported that the phenotypic correlation between the number of roots and first year field growth in loblolly pine rooted cu ttings was negligible. C effects relating to the season in which th e cuttings were set has previously been discussed in this population (Cha pter 3). Further evidence of the presence of C effects can be seen from the genetic correlations between the total genetic value of clones for rooting ability and 2nd year height based on analyses i nvolving a single rooting and field

PAGE 78

67 trial (Table 4-2). Rooted cu ttings planted in each field trial originated from a single sticking date, and the high est genetic correlations (TGBr ) occurred between traits from these trials. For example, rooted cuttings from the Winter02 rooting trial were planted only in the Field A trial. The total genetic correlation between rooting ability from Winter02 and 2nd year height from Field A was 0.24, while 2nd year height from Field A had much lower correlations with rooting ab ility from any of the other sticking dates Table 4-1. Means, variance component estimat es, heritabilities, and genetic correlations from the bivariate analysis of rooting ability and 2nd year height. Standard errors are given in parentheses. Rooting Ability 2nd Year Height x 42 % 210 cm AV 0.0135 (0.004) 399.3 (119.6) DV 0.0014 (0.002) 77.3 (39.6) IV 0.0085 (0.002) -94.9 (63.7) GV 0.0235 (0.002) 381.7 (59.9) PV 0.2299 (0.002) 1836 (61.0) HSPV 0.0038 (0.001) 108.1 (29.6) FSPV 0.0085 (0.002) 231.1 (59.1) CLPV 0.0286 (0.002) 434.1 (59.9) 2h 0.059 (0.02) 0.22 (0.06) 2 H 0.102 (0.01) 0.21 (0.03) 2HSH 0.88 (0.04) 0.92 (0.03) 2FSH 0.84 (0.04) 0.95 (0.01) 2CLH 0.82 (0.01) 0.88 (0.02) GCABr 0.32 (0.19) FSBr 0.39 (0.17) TGBr 0.29 (0.08)

PAGE 79

68 (Table 4-2). This indicates that the vigor of the hedge at the time of rooting may have influenced rooting and subsequent field growth. The biological significance of positive correlations between rooting and 2nd year height in the current stud y is unclear. Goldfarb et al. (1998) reported that root morphological traits were not meaningful for subsequent field growth after one year of growth for loblolly pine r ooted cuttings. However, it has generally been acknowledged that stock plant vigor is related to rooting. Perhaps the vigor of the stock plant is an indicator of the metabolic activity of cutt ings during root initiation. The more metabolically active genotypes may have rooted quicker and formed be tter root systems. In the current study, rooting was assessed at nine to eleven weeks, and selection may have indirectly been for rate of rooting as opposed to strictly rooti ng ability. Clones that rooted quickly during the roo ting period may have had increased metabolic activity. A positive genetic correlation between rooting a nd growth implies that some of the same genes are responsible for the expression of each trait. There may be genes in common for rooting and height relating to metabolic ac tivity. Genotypes that have a tendency for increased rates of metabolic activity may root at higher frequencies and grow larger than genotypes that do not ha ve these alleles. Genetic Gain The importance of selecting for rooting ab ility in a clonal fore stry program based on rooted cuttings has long been recognized (Foster et al. 1984; Foster 1990; Baltunis et al. 2005). Genetic gains in field traits will not be realized if the clones can not be propagated efficiently for deployment. Therefor e, increases in rooting ability could have broad economic impacts on a clona l forestry program (Foster et al. 1984). Positively correlated traits imply that se lection for one trait should also lead to improvement in the

PAGE 80

69 second trait. In the case of rooting ability and growth, positive genetic correlations can lead to substantial gains for both traits in a clonal forestry program based on rooted cuttings. The genetic gains in rooting ability and 2nd year height based on BLUP values were compared for a number of deployment strategies Selecting the top half-sib family for rooting ability would result in a gain of 36% in rooting abili ty (Figure 4-1). Deployment of the top half-sib family selected for roo ting ability would result in a genetic gain of 5.4% in 2nd year total height (Figure 4-2). Select ing the highest ranking half-sib family for 2nd year height would result in a gain of 14.8% and 8.1% in rooting ability and 2nd year height, respectively (Figur e 4-1 and Figure 4-2). The ge netic gain predictions using theoretical calculations were slightly lower than the gain predictions based on BLUP values. The gain in rooting ability associated with selecting the best half-sib family for rooting was 26.7% (using Equation 4-19). The gain in height by selecting the top growing half-sib family was 9.4% (using Equation 4-19). Slightly higher gains can be achieved by selecting the top full-sib family. Propagation of the top full-sib family selected for rooting ability results in genetic gains in rooting ability of about 43% over the population mean (Fig ure 4-1). Gains of nearly 9% in 2nd year total height could be expected by deploying the top rooting full-sib family (Figure 4-2). If the best gr owing full-sib family was propagated and deployed, then the genetic gains in rooting ab ility (Figure 4-1) and 2nd year height (Figure 4-2) would be 8.6% and 10.1%, respectively. The theore tical gain in rooting ability for full-sib selection was equivalent to the estimate base d on BLUP values. However, the theoretical

PAGE 81

70 gain for full-sib selection for 2nd year height was 16.3% which was higher than the estimate based on BLUPs. Table 4-2. The total genetic correlation (TGBr ) between rooting ability and 2nd year height from analyses of a single rooting trial a nd field trial. Shad ed values indicate the rooting trial in which cuttings originated from for their respective field trials. Standard errors are given in parentheses. Field A Field B Field C Field D Field E Field F Spring01 0.002 (0.07) 0.17 (0.07) 0.17 (0.07) 0.21 (0.07) 0.16 (0.07) 0.25 (0.07) Summer01 0.08 (0.06) 0.07 (0.07) 0.17 (0.07) 0.21 (0.07) 0.14 (0.07) 0.12 (0.07) Winter02 0.24 (0.07) 0.16 (0.07) 0.23 (0.07) 0.27 (0.08) 0.24 (0.08) 0.18 (0.08) Spring02 0.06 (0.07) 0.25 (0.07) 0.26 (0.08) 0.34 (0.07) 0.16 (0.07) 0.24 (0.08) Summer02 0.10 (0.07) 0.21 (0.07) 0.16 (0.08) 0.21 (0.08) 0.13 (0.08) 0.28 (0.08) Clearly, the most gain for either trait woul d be achieved by selecting the single best clone (Figure 4-1 and Figure 4-2). For inst ance, by selecting the top clone for rooting ability, the genetic gain based on clonal predicted values in rooting was 96% (Figure 4-1) indicating that approxim ately 82% of the ramets would r oot from this clone (42% mean rooting plus 0.96 x 42% = 82%). However, selecting the top roo ting clone from this population would result in a decreas e (genetic loss) in overall 2nd year height (Figure 42). On the other hand, se lecting the top clone for 2nd year height would result in a gain of nearly 27% in 2nd year height (Figure 4-2) and 43% in rooting ability (Figure 4-1) over the population mean for these traits. Genetic gain in rooting ability based on theoretical

PAGE 82

71 calculations was greater than those based on BLUP values for selection of the single best clone and was 118%, while the gain in 2nd year height for the top growing clone was 31.3% (using Equation 4-21). 0 10 20 30 40 50 60 70 80 90 100 Best Half-Sib Family Best Full-Sib Family Best Clone from Best 10 Full-Sib Families Best Clone% Gain in Rooting Ability % Gain in Rooting Ability when Selecting for 2nd Year Height % Gain in Rooting Ability when Selecting for Rooting Ability Figure 4-1. The genetic ga in in rooting ability (%) over the population mean for deployment of the best half-sib family, full-sib family, best clone from the best ten full-sib families, and the single best clone when selecting for rooting ability or 2nd year height. In order to address genetic diversity issues and the risk associated with deploying a single clone, a second clonal deployment opti on was considered by selecting the best clone from each of the ten highest ranking fullsib families. This strategy resulted in genetic gains nearly double th at of the full-sib family deployment strategy. When the trait selected is rooting ability, the best clone in the top ten full-sib families would result in gains of 77% in rooting abil ity (Figure 4-1) and 5.6% in 2nd year height (Figure 4-2).

PAGE 83

72 When the trait selected is for 2nd year height, then the genetic gains would be 37.6% in rooting ability (Figure 4-1) and 18.3% for 2nd year height (Figure 4-2). -15 -10 -5 0 5 10 15 20 25 30 Best Half-Sib Family Best Full-Sib Family Best Clone from Best 10 Full-Sib Families Best Clone% Gain in 2nd Year Height % Gain in 2nd Year Height when Selecting for Rooting Ability % Gain in 2nd Year Height when Selecting for 2nd Year Height Figure 4-2. The genetic gain in 2nd year height (%) over the population mean for deployment of the best half-sib family, full-sib family, best clone from the best ten full-sib families, and the single best clone when selecting for rooting ability or 2nd year height. Deployment of well-tested clones can result in genetic gains in both rooting ability and 2nd year total height. Both tr aits should be considered fo r a successful loblolly pine clonal forestry program based on rooted cut ting technology. The re sponses to selection when considering both rooting ability and 2nd year height were compared for a number of selection indices (Figure 4-3). The gain in rooting ability associated with selecting the top 10% of clones ranged from 23.8% when only 2nd year height was considered to 57.6% when only rooting ability was consid ered (Figure 4-3). Higher gains can be

PAGE 84

73 achieved by increasing the selection intensity When only the top 1% of clones was selected, the gain in rooting abi lity ranged from 33.4% when only 2nd year height was considered to 80.6% when only rooting ability was considered (Figure 4-3). The genetic gain in 2nd year height associated with deployment of the best 10% of clones ranged from 4.8% when only rooting ability was considered to 12.6% gain when only 2nd year height was considered (Figure 4-3). Similarly, genetic gain in 2nd year height ranged from 3.8% to 18.5% when the best 1% of cl ones were selected (Figure 4-3). Clonal forestry programs that consider multiple traits need to optimize their selection strategies for deployment populations. Arbitrarily setting the optimum selection weights on 2nd year height and rooting ability to 90% of the maximum genetic gain obtainable for a single trait, then the optim um selection weights can be compared. For example, if the top 10% of clones are select ed and 90% of the maximum gain in rooting ability is considered, then the optimum weights on 2nd year height and rooting ability correspond to weights of 0.7 and 0.3, respectiv ely. These weights result in genetic gains of 53.6% in rooting ability and 8.7% in 2nd year height (Figure 4-3) If 1% of the clones are selected, then the optimu m weights are 0.6 and 0.4 on 2nd year height and rooting ability, respectively, and correspond to gains of 75.9% in rooting ability and 11% in 2nd year height (Figure 4-3). If 90% of the maximum gain obtainable for 2nd year height is the criterion, then the optimum weights for both deployment options are 0.9 and 0.1 on 2nd year height and rooting ability, respectively. The genetic gains associated with selecting the top 10% of clones utilizing these weights results in gains of 40.4% in rooting ability and 11.9% in 2nd year height (Figure 4-3). If only the top 1% of clones are deployed using these selection

PAGE 85

74 weights, then genetic gains of 54.2% in rooting ability and 17.4% in 2nd year height are obtainable (Figure 4-3). The ma ximum change in genetic gain of rooting ability results from increasing the weight on rooting ability from 0 to 0.1. In fact, an additional 17-21% gain in rooting ability can be obtained in the deployment population by increasing the weight on rooting ability from 0 to 0.1. When 10% of the clones were selected ( 220), then the number of full-sib families represented in the deployment population ra nged from 38 to 45 depending on selection index (Table 4-3). When only the top 1% of clones were selected ( 22), the number of full sib families represented in the deployment popul ation varied from 9 to 15 (Table 4-3). Maximum genetic gains can be achieved by ignoring relatedness among selections. However, if no constraints are placed on the relatedness of selections, then there is a tendency to make many selections from the better families. Although the approximate average number of clones selected per full-si b family was five when 10% of the clones were selected, nearly 45% of these selections came from only five families when only rooting ability was selected. Similar trends ca n be seen when selecting 1% of the clones. For example, when all of the selection wei ght is on rooting ability, then there was an average of 2.44 clones per full-sib family sel ected (Table 4-3). However, 15 of the 22 selections came from three families. On the other hand, when considering the number of half-sib families represented in the sele cted population, 26 to 30 out of 32 possible parents are represented when 10% of the clone s are selected. Thirteen to 17 parents are represented in the deployment population when 1% of the clones are selected (Table 4-3). Therefore, at least in this population, ther e may be sufficient gene tic diversity in the deployment population even with higher selection intensities. Isik et al. (2005) reported

PAGE 86

75 genetic gains in growth n ear 30% for loblolly clones from a different population irregardless of any restrictions on the relatedness of selections. Conclusion Rooting ability and 2nd year height are he ritable traits in lo blolly pine, and both traits showed substantial clonal variation. There was a positive genetic correlation between rooting and height at the parental, full-sib family and clonal levels. Genetic gains in rooting ability and 2nd year height are possible as demonstrated by a number of deployment strategies. For hard-to-root spec ies, like loblolly pi ne, a successful clonal 0% 10% 20% 30% 40% 50% 60% 70% 80% 90%1|0 0.9|0.1 0.8|0.2 0.7|0.3 0.6|0.4 0 .5|0.5 0 .4|0.6 0.3 |0.7 0.2 |0.8 0.1| 0.9 0|1Weight on 2nd Year Height | Weight on Rooting Ability% Gain % Gain in 2nd Year Height (10%) % Gain in 2nd Year Height (1%) % Gain in Rooting Ability (10%) % Gain in Rooting Ability (1%) Figure 4-3. Responses to sele ction in rooting ability and 2nd year height with various selection indices for two clonal deployme nt options: 10% of clones selected and 1% of clones selected.

PAGE 87

76 forestry program based on rooted cuttings must consider both rooti ng ability and growth when making selections. Because here these tr aits are positively correlated, selection for one trait should lead to positive gains in the other. Or, early culling of clones based on poor rooting will not negatively effect select ion of clones for growth when field data becomes available. Clonal forestry is not a breeding method to develop better genotypes. However, clonal forestry is a method to mass-produce we ll-tested genotypes. Short-term genetic gains may be maximized through deployment of well-tested clones, but long-term gains need to involve both clonal selection and recu rrent selection for additive genetic variation through repeated selection and breeding. Restrictions on th e relatedness of selections will be necessary when making selections fo r a breeding population or when deployment will be traditional zygotic seedlings from seed orchards in order to reduce detrimental effects of inbreeding depression.

PAGE 88

77 Table 4-3. The number of full-sib families (half-sib families) and average number of clones per full-sib family (half-sib family ) selected from selecting 10% or 1% of the top clones using the combined selection index. 10% Clones Selected 1% Clones Selected Weight on 2nd Year Height: Weight on Rooting Ability # FS Families (# HS families) Ave. # Clones per FS Family (HS Family) # FS Families (# HS families) Ave. # Clones per FS Family (HS Family) 1:0 38 (26) 5.8 (16.9) 14 (17) 1.6 (2.6) 0.9:0.1 41 (28) 5.3 (15.7) 15 (17) 1.5 (2.6) 0.8:0.2 43 (28) 5.1 (15.7) 12 (15) 1.8 (2.9) 0.7:0.3 43 (30) 5.1 (14.7) 13 (16) 1.7 (2.8) 0.6:0.4 45 (30) 4.9 (14.7) 10 (12) 2.2 (3.7) 0.5:0.5 45 (29) 4.9 (15.2) 12 (15) 1.8 (2.9) 0.4:0.6 45 (29) 4.9 (15.2) 11 (15) 2.0 (2.9) 0.3:0.7 45 (29) 4.9 (15.2) 11 (15) 2.0 (2.9) 0.2:0.8 45 (29) 4.9 (15.2) 10 (14) 2.2 (3.1) 0.1:0.9 45 (29) 4.9 (15.2) 9 (13) 2.4 (3.4) 0:1 45 (29) 4.9 (15.2) 9 (13) 2.4 (3.4)

PAGE 89

78 CHAPTER 5 CONCLUSION Loblolly pine is the most important commercial tree species in the southeastern United States. Several forest industries in the southeastern United States are deploying full-sib families of loblolly pine operationally. In addition, many of these companies are pursuing clonal forestry progr ams with loblolly pine. Clones need to be well-tested before they can be deployed. This invol ves the accumulation of reliable data from propagation, growth traits, dis ease resistance, etc. Those well-tested clones that meet the selection criteria can lead to substantial genetic gain. Several key results can be concluded from this research. With rooting data from 2,200 clones from 70 full-sib families, the curr ent study gives better es timates of genetic components of variance for rooting than severa l previous studies. These results show a great deal of genetic variation for rooting among families and clones of loblolly pine. Combined with moderate to high estimates of familyand clonal-mean heritabilities and type B correlations for rooting ability in different rooting trials, these results indicate the potential for increasing rooti ng efficiency by selecting good rooting families and clones or culling poor rooters. Field testing of clones is a necessary com ponent to clonal forestry programs based on rooted cuttings. In addition, genetic field trials establis hed with clones and seedlings from the same full-sib families provide an opportunity for comparing both half-sib and full-sib family performances for both propagul es. Based on the results in Chapter 3, several conclusions can be drawn. First, cl onally replicated seedli ng trials of loblolly

PAGE 90

79 pine provide genetic information with greater precision than zygotic seedlings. Second, genetic correlations between propagule types for the growth traits were highly favorable. These high genetic correlations between propagul e types reassure that parental and fullsib family rankings are stable regardless of propagule type. This implies that parental and full-sib family rankings based on existi ng seedling progeny trials could be used to select parents and families that perform well when they are deployed as rooted cuttings. Third, little genotype x envir onment interaction was observed across sites at the parental, family, and clonal level for all traits. Howeve r, there appears to be a carry-over effect relating to the season in which the cuttings were rooted for the clonal material. Finally, randomization is essential at all stages in testing when estimating genetic parameters. The lack of randomization for the seedling popu lation apparently resulted in a problem with partitioning of the genetic variance, causing full-sib families to appear more different and inflating estimates of dominance genetic variation. Rooting ability and 2nd year height are he ritable traits in lo blolly pine, and both traits showed substantial clonal variation. There was a positive genetic correlation between rooting and height at the parental, full-sib family and clonal levels. Genetic gains in rooting ability and 2nd year height are possible as demonstrated by a number of deployment strategies. For difficult-to-root species, like loblolly pine, a successful clonal forestry program based on rooted cuttings must consider both rooti ng ability and growth when making selections. Because here these tr aits are positively correlated, selection for one trait should lead to positive gains in the other. Or, early culling of clones based on poor rooting will not negatively affect select ion of clones for growth when field data become available.

PAGE 91

80 Clonal forestry is not a breeding method to develop better genotypes. However, clonal forestry is a method to mass-produce we ll-tested genotypes. Short-term genetic gains may be maximized through deployment of well-tested clones, but long-term gains need to involve both clonal selection and recu rrent selection for additive genetic variation through repeated selection and breeding. Restrictions on th e relatedness of selections will be necessary when making selections fo r a breeding population or when deployment will be traditional zygotic seedlings from seed orchards in order to reduce detrimental effects of inbreeding depression.

PAGE 92

APPENDIX A LOBLOLLY PINE PARTIAL DIALLEL MATING DESIGN. THIRTY-TWO PARENTS WERE CROSSED TO GE NERATE 70 FULL-SIB FAMILIES.

PAGE 93

82 1234567891011121314151617181920212223242526272829303132 1XXXX 2XXX XX 3XX X 4XX X 5XXXX 6XXX 7XX 8XXX 9XX 10X 11XX 12XXX 13 XX 14 XXXXX 15 XX 16 XX 17 X 18 X 19 20 XXX 21 XX 22 XXX 23 XX 24 X 25 XX 26 XX 27 X 28 XX 29 X 30XX 31 X 32 Parent 2Parent 1

PAGE 94

83 APPENDIX B VARIANCE COMPONENT ESTIMATES FOR ROOTING ABILITY Table B-1. Observed variance component es timates for rooting of loblolly pine stem cuttings from single-trial analyses. Spring01 Summer01 Wint er02 Spring02 Summer02 2TRAY 0.003977 0.005357 0.002891 0.010084 0.004796 2GCA 0.005018 0.004629 0.005383 0.004451 0.003817 2SCA 0.002683 0.000771 0.001363 0.001868 0.00187 2CLONE 0.041427 0.039236 0.024317 0.026364 0.023738 2REPxFAM 0.000836 0.000559 0.000637 0.000661 0.001043 2ERROR 0.190089 0.177991 0.204829 0.200698 0.142

PAGE 95

84Table B-2. Observed variance component estimates for rooting of lobl olly pine stem cuttings fr om pair-wise-trial analyses. Spring01 Summer01 Spring01 Winter02 Spring01 Spring02 Spring01 Summer02 Summer01 Winter02 Summer01 Spring02 Summer01 Summer02 Winter02 Spring02 Winter02 Summer02 Spring02 Summer02 2TRAY 0.004751 0.003637 0.007712 0.004474 0.004115 0.007929 0.005059 0.006905 0.003774 0.007824 2GCA 0.003434 0.002923 0.00433 0.003929 0.003191 0.002844 0.002557 0.003065 0.002347 0.003915 2SCA 0.001409 0.000417 0.001262 0.001876 0.00067 0.000654 0.001027 0.001107 0 0.001147 2CLONE 0.021605 0.011304 0.017015 0.014235 0.012645 0.014791 0.014151 0.011209 0.010850 0.013135 2TESTxGCA 0.001464 0.002034 0.000422 0.000688 0.001846 0.001380 0.001472 0.001609 0.002123 0.0004 2TESTxFAM 0.000302 0.001758 0.001299 0.000539 0.000396 0.000572 0.000029 0.000423 0.001480 0.000485 2TESTxCLONE 0.018862 0.021637 0.017991 0.022116 0.019564 0.018869 0.020504 0.014919 0.014083 0.012163 2REPxFAM 0.000674 0.000764 0.000703 0.000956 0.000598 0.000613 0.000797 0.000646 0.000825 0.000767 2ERROR 0.183321 0.198958 0.197306 0.162528 0.192458 0.192031 0.159335 0.202419 0.174684 0.177164

PAGE 96

85 Table B-3. Observed variance components for rooting ability from the across-trial analysis using all five rooting trials. Variance Component Estimate 2TRAY 0.005785 2GCA 0.002921 2SCA 0.001062 2CLONE 0.01638 2TESTxGCA 0.001364 2TESTxFAM 0.000691 2TESTxCLONE 0.017349 2REPxFAM 0.000713 2ERROR 0.184907

PAGE 97

86 APPENDIX C VARIANCE COMPONENT ESTIMATES FO R EARLY GROWTH TRAITS OF LOBLOLLY PINE CLONES AND SEED LINGS FROM THE SAME FULL-SIB FAMILIES Table C-1. Observed variance components fo r loblolly pine clones from the across-trial analyses of 1st year height, 2nd year height, height incr ement, and crown width. A separate error variance was modeled for each trial. 1st Year Height 2nd Year Height Height Increment Crown Width 2INC 53.4293 300.188 123.143 105.75 2GCA 17.4525 100.363 38.141 27.5703 2SCA 2.9992 21.601 7.5808 4.93 2CLONE 32.3599 157.931 59.3416 56.4437 2TESTxGCA 4.1225 13.3866 7.9053 5.9029 2TESTxFAM 1.2003 3.7766 1.2451 1.3288 2TESTxCLONE 22.4253 82.7545 29.0673 23.4036 2TxCxGCA 0.4069 4.5692 2.942 2.2437 2TxCxFAM 0.0000008 0.3873 0.0000005 1.1495 2TxCxCLONE 0.0000008 17.1395 14.179 10.7818 2REPxGCA 0.3275 1.0281 1.5638 0.6583 2REPxFAM 1.9242 5.511 1.5806 0.8574 2AERROR 512.142 1700.04 692.262 610.815 2BERROR 279.24 853.321 427.247 306.51 2CERROR 527.513 2061.14 928.903 843.627 2DERROR 264.114 1110.76 864.549 329.026 2EERROR 313.481 1124.11 682.649 423.413 2FERROR 123.318 804.457 527.064 296.22

PAGE 98

87 Table C-2. Observed variance components fo r loblolly pine seedlings from the acrosstrial analyses of 1st year height, 2nd year height, height increment, and crown width. A separate error variance was modeled for each trial. 1st Year Height 2nd Year Height Height Increment Crown Width 2GCA 3.6682 29.9688 12.9402 13.7566 2SCA 4.3654 20.8173 10.4732 5.0053 2TESTxGCA 1.5235 8.4371 8.6503 3.8956 2TESTxFAM 8.5693 25.1585 8.5047 7.9082 2TxCxGCA 0.0000007 0.0000004 0.0000002 0.5247 2TxCxFAM 0.6036 0.0000001 1.375 1.0372 2REPxGCA 0.0000001 0.0000001 0.0000001 1.2412 2REPxFAM 6.3223 55.867 26.1737 15.4417 2AERROR 599.229 1621.95 713.814 590.37 2BERROR 361.489 1104.24 567.689 325.987 2CERROR 508.95 2524.33 1246.59 1110.51 2DERROR 294.534 1356.8 1059.94 501.614 2EERROR 379.556 1766.73 1109.52 610.975 2FERROR 101.715 755.187 585.714 341.353

PAGE 99

88 APPENDIX D VARIANCE COMPONENT ESTIMATES FROM THE BIVARIATE ANALYSES OF ROOTING AND 2ND YEAR HEIGHT Table D-1. Observed variance component estimates from the bivariate analyses of rooting ability from Spring01 and 2nd year height from each of the field trials. Spring01 Field A Field B Field C Field D Field E Field F 2rootINC 0.0039 0.0039 0.004 0.0039 0.004 0.004 2heightINC 219.85 232.03 801.24 390.46 105.53 89.22 2rootGCA 0.0051 0.0052 0.0052 0.0051 0.0052 0.0051 rootheightGCA 0.0778 0.219 0.2515 0.3664 0.3268 0.1727 2heightGCA 93.85 77.71 134.98 134.81 144.59 98.15 2rootSCA 0.0027 0.0026 0.0026 0.0027 0.0026 0.0027 rootheightSCA 0.0654 -0.1556 -0.0148 0.0203 0.0815 0.1294 2heightSCA 28.34 3.7 8.13 28.18 43.77 30.08 2rootCLONE 0.0415 0.0415 0.0414 0.0415 0.0414 0.0414 rootheightCLONE -0.21 0.3412 0.4084 0.3997 0.2865 0.7351 2heightCLONE 386.08 191.46 250.52 230.24 426.68 207.67 2rootREPxFAM 0.0008 0.0008 0.0008 0.0008 0.0008 0.0008 2heightREPxFAM 0.0996 5.31 23.68 15.47 24.33 15.44 2rootERROR 0.1901 0.1901 0.1901 0.1901 0.1901 0.1901 2heightERROR 1641.17 879.86 2003.72 1127.51 1072.05 834.69

PAGE 100

89 Table D-2. Observed variance component estimates from the bivariate analyses of rooting ability fr om Summer01 and 2nd year height from each of the field trials. Summer01 Field A Field B Fiel d C Field D Field E Field F 2rootINC 0.0053 0.0053 0.0053 0.0053 0.0053 0.0053 2heightINC 220.06 232.4 801.8 389.37 105.55 89.37 2rootGCA 0.005 0.005 0.005 0.005 0.005 0.0049 rootheightGCA 0.243 0.1731 0.2933 0.3457 0.2579 0.0796 2heightGCA 98.15 79.12 135.89 138.56 147.16 98.3 2rootSCA 0.0007 0.0007 0.0007 0.0007 0.0007 0.0008 rootheightSCA 0.0397 0.0141 -0.0125 0.0078 0.0616 0.1556 2heightSCA 26.74 3.33 6.33 24.69 40.85 26.5 2rootCLONE 0.0393 0.0392 0.0393 0.0393 0.0392 0.0393 rootheightCLONE -0.0944 -0.0803 0.3137 0.4131 0.2729 0.2437 2heightCLONE 385.69 190.78 250.56 230.58 426.65 206.64 2rootREPxFAM 0.0006 0.0006 0.0006 0.0006 0.0006 0.0006 2heightREPxFAM 0.0757 5.3 23.92 15.69 24.4 15.71 2rootERROR 0.1779 0.1779 0.1779 0.1779 0.1779 0.1779 2heightERROR 1641.1 879.8 2003.35 1127.44 1072.03 834.0

PAGE 101

90 Table D-3. Observed variance component estimates from the bivariate analyses of rooting ability fr om Winter02 and 2nd year height from each of the field trials. Winter02 Field A Field B Fiel d C Field D Field E Field F 2rootINC 0.003 0.0029 0.0029 0.0029 0.0029 0.0029 2heightINC 220.76 232.19 803.25 391.48 106.31 89.66 2rootGCA 0.0052 0.0053 0.0053 0.0052 0.0052 0.0052 rootheightGCA 0.2043 0.1833 0.271 0.3 0.2017 0.0927 2heightGCA 102.17 78.17 134.82 136.55 149.43 101.99 2rootSCA 0.0014 0.0014 0.0014 0.0014 0.0014 0.0014 rootheightSCA 0.1209 0.0611 0.1015 0.1848 0.1924 0.1857 2heightSCA 28.81 4.03 7.75 27.65 44.14 26.48 2rootCLONE 0.0243 0.0243 0.0243 0.0243 0.0243 0.0243 rootheightCLONE 0.6345 0.1392 0.3584 0.3934 0.6888 0.3461 2heightCLONE 392.91 190.82 251.6 232.29 430.86 206.31 2rootREPxFAM 0.0006 0.0006 0.0006 0.0006 0.0006 0.0006 2heightREPxFAM 0.3872 5.3 23.62 15.68 24.36 15.49 2rootERROR 0.2048 0.2048 0.2048 0.2048 0.2048 0.2048 2heightERROR 1639.72 879.85 2003.09 1126.58 1071.65 834.4

PAGE 102

91 Table D-4. Observed variance component estimates from the bivariate analyses of rooting ability from Spring02 and 2nd year height from each of the field trials. Spring02 Field A Field B Field C Field D Field E Field F 2rootINC 0.0101 0.0997 0.01 0.01 0.01 0.01 2heightINC 219.97 232.63 803.37 391.27 106.4 88.85 2rootGCA 0.004 0.0041 0.004 0.004 0.004 0.0039 rootheightGCA 0.1176 0.1904 0.1485 0.2154 0.1198 0.0669 2heightGCA 97.42 78.53 136.8 140.76 148.65 97.39 2rootSCA 0.0018 0.0018 0.0018 0.0018 0.0019 0.0019 rootheightSCA 0.1254 0.1025 0.1448 0.242 0.3037 0.278 2heightSCA 28.64 3.41 6.99 26.41 41.76 32.59 2rootCLONE 0.0264 0.0265 0.0264 0.0265 0.0264 0.0264 rootheightCLONE -0.0583 0.4215 0.7019 0.8602 0.2998 0.5378 2heightCLONE 385.33 192.98 256.21 23.29 427.8 205.39 2rootREPxFAM 0.0007 0.0007 0.0007 0.0007 0.0007 0.0007 2heightREPxFAM 0 5.33 23.42 15.61 24.23 15.59 2rootERROR 0.2007 0.2007 0.2007 0.2007 0.2007 0.2007 2heightERROR 1641.16 879.54 2002.81 1127.07 1071.69 834.82

PAGE 103

92 Table D-5. Observed variance component estimates from the bivariate analyses of rooting ability fr om Summer02 and 2nd year height from each of the field trials. Summer02 Field A Field B Fiel d C Field D Field E Field F 2rootINC 0.0048 0.0048 0.0048 0.0048 0.0048 0.0048 2heightINC 219.68 232.5 802.51 390.45 105.24 88.7 2rootGCA 0.0037 0.0037 0.0037 0.0037 0.0037 0.0037 rootheightGCA 0.1006 0.2194 0.1656 0.2104 0.0913 0.1656 2heightGCA 96.96 78.41 135.51 138.45 145.59 99.21 2rootSCA 0.0017 0.0017 0.0017 0.0018 0.0017 0.0017 rootheightSCA 0.0439 0.0693 0.0831 0.101 0.2114 0.1497 2heightSCA 27.33 3.68 7.906 26.24 41.0 29.59 2rootCLONE 0.0238 0.0238 0.0238 0.0237 0.0238 0.0237 rootheightCLONE 0.2082 0.1956 0.2738 0.3435 0.2526 0.5838 2heightCLONE 386.1 190.66 249.78 229.44 426.75 206.54 2rootREPxFAM 0.001 0.001 0.001 0.001 0.001 2heightREPxFAM 0.1938 5.25 23.88 15.48 24.3 2rootERROR 0.142 0.142 0.142 0.142 0.142 2heightERROR 1641.0 879.82 2003.21 1127.52 1072.29

PAGE 104

93 Table D-6. Observed variance component estimates from the bivariate analysis of rooting ability using all fi ve rooting trials and 2nd year height using all six field trials. Variance Component Estimate 2rootINC 0.0058 2heightINC 331.45 2rootGCA 0.0034 rootheightGCA 0.1858 2heightGCA 99.83 2rootSCA 0.0004 rootheightSCA 0.1155 2heightSCA 19.33 2rootCLONE 0.0164 rootheightCLONE 0.3988 2heightCLONE 162.72 2rootTESTxFAM 0.0035 2heightTESTxFAM 30.74 2rootTESTxCLONE 0.0173 2heightTESTxCLONE 103.55 2rootREPxFAM 0.0007 2heightREPxFAM 12.1 2rootERROR 0.1849 2heightERROR 1308.39

PAGE 105

94 LIST OF REFERENCES Anderson, A.B., Frampton, L.J., and Weir, R.J. 1999. Shoot production and rooting ability of cuttings from juvenile greenhouse loblolly pine hedges. Transactions of the Illinois State Academy of Science. Volume 92(1 and 2): 1-14. Baltunis, B.S., Huber, D.A., White, T.L., Gold farb, B., and Stelzer, H.E. 2005. Genetic effects of rooting loblolly pine stem cutt ings from a partial diallel mating design. Can. J. For. Res. 35: 1098-1108. Banks, B.D., Mao, I.L., and Walter, J.P. 1985. Robustness of the restricted maximum likelihood estimator derived under normality as applied to data with skewed distributions. J. Dairy Sci. 68: 1785-1792. Borralho, N.M.G. and Kanowski, P.J. 1995. Correspondence of performance between genetically related clones and seedlings. Can. J. For. Res. 25: 500-506. Burdon, R.D. 1977. Genetic correla tion as a concept for studying genotypeenvironmentinteraction in forest tree breeding. Silvae Genet. 26: 168-175. Burdon, R.D. and Shelbourne, C.J.A. 1974. The use of vegetative propagules for obtaining genetic information. N.Z.J. For. Sci. 4: 418-425. Byram, T.D., Miller, L.G., and Raley, E.M. 2004. 52nd Progress Report of the Cooperative Forest Tree Improvement Progr am. Texas Forest Service, College Station, TX. 24 p. Cech, F.C. 1958. The vegetative propagation of Pinus taeda L. (loblolly pine). Ph.D. Dissertation. Texas A. & M., College Station, TX. Cockerham, C.C. 1954. An extension of the concept of partitioning hereditary variance for analysis of covariance among relatives when epistasis is present. Genetics 39: 859-882. Comstock, R.E. and Moll, R.H. 1963. Genotype-environment interactions. In: Statistical Genetics and Plant Breeding. Editors R.E. Hanson and H.F. Robinson. NAS-NRC Pub. 982, Washi ngton, DC. pp 53-93. Cooney, G. and Goldfarb, B. 1999. Effects of shearing height, pr uning intensity, shoot origin and family on shoot morphology and th eir effects on rooting of loblolly pine stem cuttings. In: Proceedings of the 25th Biennial Southern Forest Tree Improvement Conference. July 11-14 1999, New Orleans, LA. pp 52-53.

PAGE 106

95 Costa e Silva, J., Borralho, N.M.G., and Potts, B.M. 2004. Additive and non-additive genetic parameters from clonally re plicated and seedling progenies of Eucalyptus globules. Theor. Appl. Genet. 108: 1113-1119. Cotterill, P.P. and Dean, C.A. 1990. Succe ssful Tree Breeding with Index Selection. Center for Scientific and Industrial Re search Organization (CSIRO), Collingwood, Victoria, Australia. p. 80. Cunningham, M.W. 1986. Genetic variation in rooting ability of American sycamore cuttings. In: Proceedings of Research and Development Conference, Sept. 1986, Atlanta. Technical Associat ion of the Pulp and Paper Industry, Norcross, GA. pp. 1-6. Dempster, E.R. and Lerner, I.M. 1950. Her itability of threshold characters. Genetics 35: 212-236. De Souza, S.M., White, T.L., Hodge, G.R., a nd Schmidt, R.A. 1991. Genetic parameter estimates for greenhouse traits of slash pine artificially inoculated with fusiform rust fungus. For. Sci. 37: 836-848. Dickerson, G.E. 1962. Implications of genetic-environmental interaction in animal breeding. Animal Production 4: 47-64. Dickerson, G.E. 1969. Techniques for re search in quantitati ve animal genetics In: Techniques and Procedures in Animal Science Research. Amer. Soc. Anim. Sci., Albany, NY, pp. 36-79. Dieters, M.J. 1994. Inheritance of volume and rust resistance in slash pine. Ph.D. Dissertation,Univ. Florida, Gainesville, FL, 105 p. Dieters, M.J., Hodge, G.R., and White, T.L. 1996. Genetic parameter estimates for resistance to rust (Cronaritum quercuum) infection from full-sib tests of slash pine (Pinus elliottii), modelled as functions of rust incidence. Silvae Genet. 45: 235242. Falconer, D.S. and Mackay, T.F.C. 1996. Introduction to Quantitative Genetics. 4th Edition. Longman Group Ltd., Essex, England. 464 p. Farmer, R.E., Durst, J.T., Deng Shaotang, a nd Yang Jun-Tao. 1992. Effects of clones, primary ramets, and age of stock plants on tamarack rooting. Silvae Genet. 41: 2224. Farmer, R.E., Freitag, M., and Garlick, K. 1989. Genetic variance and C effects in balsam poplar rooting. Silvae Genet. 38: 62-65. Forest Biology Research Cooperative (F BRC) 2000. Study B: clonal biology and performance of elite genotypes of loblo lly and slash pine. FBRC Report # 13. Forest Biology Research Cooperative, Un iv. Florida, Gainesville, FL, 31 p.

PAGE 107

96 Forest Biology Research Cooperative (FBRC) 2003. Forest Biology Research Cooperative 7th Annual Report. FBRC Report # 23. Forest Biology Research Cooperative, Univ. Florida, Gainesville, FL, 100 p. Foster, G.S. 1978. Genetic variation in rooti ng stem cuttings from f our year old loblolly pine. Weyerhaeuser Co., Hot Springs AR. Tech. Rep. No. 042-3204/78/97. Foster, G.S. 1990. Genetic control of rooting ab ility of stem cuttings from loblolly pine. Can. J. For. Res. 20: 1361-1368. Foster, G.S. and Shaw, D.V. 1987. A tree improvement program to develop clones of loblolly pine for reforestation. In: Proceedings of 19th Southern Forest Tree Improvement Conference, College Station, TX. pp 17-21. Foster, G.S. and Shaw, D.V. 1988. Using cl onal replicates to expl ore genetic variation in a perennial plant species. Theor. Appl. Genet. 76: 788-794. Foster, G.S., Campbell, R.K., and Adams, W.T. 1984. Heritability, gain, and C effects in rooting of western hemlock cuttings. Can. J. For. Res. 14: 628-638. Foster, G.S., Campbell, R.K., and Adams, W. T. 1985. Clonal selection prospects in western hemlock combining rooting traits with juvenile height growth. Can. J. For. Res. 15: 488-493. Foster, G.S., Lambeth, C.C., and Greenwood, M.S. 1987. Growth of loblolly pine rooted cuttings compared with seedlings. Can. J. For. Res. 17: 157-164. Foster, G.S., Stelzer, H.E., and McRae, J.B. 2000. Loblolly pine cutting morphological traits: effects on rooting and field performance. New Forests 19: 291-306. Frampton, L.J., Jr. and Foster, G.S. 1993. Field testing vegetative propagules. In: Clonal Forestry I. Genetic and Biotechnology. Editors M.H. Ahuja and W.J. Libby. Springer-Verlag, Berlin, Germany. pp. 110-134. Frampton, L.J., Jr. and Huber, D.A. 1995. Clonal Variation in f our-year-old loblolly pine in coastal North Carolina. In: Proceedings of 23rd Southern Forest Tree Improvement Conference, Asheville, NC. June 20-22, 1995. pp. 254-264. Frampton, L.J., Jr., Goldfarb, B., and Surles, S.E. 1999. Nursery ro oting and growth of loblolly pine cuttings: effects of rooti ng solution and full-sib family. South. J. Appl. For. 23(2): 108-116. Frampton, L.J., Jr., Li, B., and Goldfarb, B. 2000. Early field growth of loblolly pine rooted cuttings and seedlings. S outh. J. Appl. For. 23: 108-115. Gezan, S., Huber, D., Medina, A., Parisi, L., and Powell, G. 2004. 46th Annual Progress Report. Cooperative Forest Genetics Res earch Program. Gainesville, FL. 30 p.

PAGE 108

97 Gilmour, A.R., Gogel, B.J., Cullis, B.R., Welham, S.J., and Thompson, R. 2002. ASReml User Guide Release 1.0. VSN International Ltd., Hemel Hempstead, HP1 1ES, UK. 267 p. Goldfarb, B., Surles, S.E., Thetford, M., and Blazich, F.A. 1998. Effects of root morphology on nursery and firstyear field growth of rooted cuttings of loblolly pine. South. J. Appl. For. 22(4): 231-234. Greenwood, M.S. and Weir, R.J. 1995. Genetic variation in rooting ability of loblolly pine cuttings: effects of auxin and fa mily on rooting by hypocotyl cuttings. Tree Physiology 15: 41-45. Grigsby, H.C. 1962. Propagation of loblolly pine by cuttings. Comb. Proc. Int. Plant Propag. Soc. 11: 33-35. Huber, D.A. 1993. Optimal mating designs and optimal techniques for analysis of quantitative traits in forest genetics. Ph .D. Dissertation, For. Res. Conserv., Univ. Florida, Gainesville, FL, 151 p. Huber, D.A., White, T.L., and Hodge, G.R. 1992. The efficiency of half-sib, half-diallel and circular mating designs in the estimati on of genetic parameters in forestry: a simulation. For. Sci. 38(4): 757-776. Huber, D.A., White, T.L., and Hodge, G.R. 1994. Variance component estimation techniques compared for two mating desi gns with forest genetic architecture through computer simulation. Theor. Appl. Genet. 88: 236-242. Isik, F., Li, B., and Frampton, L.J. 2003. Additive, dominance, and epistatic variance estimates from a replicated clonal test of loblolly pine. For. Sci. 49: 77-88. Isik, F., Li, B., Frampton, J., and Goldfarb, B. 2004. Efficiency of seedlings and rooted cuttings for testing and selection in Pinus taeda. For. Sci. 50: 44-53. Isik, F., Goldfarb, B., LeBude, A., Li, B., and McKeand, S. 2005. Predicted genetic gains and testing efficiency from two loblolly pine clonal trials. Can. J. For. Res. 35: 1754-1766. Jayawickrama, K.J.S. and Carson, M.J. 2000. A breeding strategy for the New Zealand radiata pine breeding cooperati ve. Silvae Genet. 49: 82-90. Kendall, M.G. and Stuart, A. 1963. The Advan ced Theory of Statistics. Vol. 1. Hafner, New York, 433 p. LeBude, A.V., Goldfarb, B., Blazich, F.A., Wi se, F.C., and Frampton, Jr., L.J. 2004. Mist, medium water potential, and cutting wa ter potential influen ce rooting of stem cuttings of loblolly pine. Tree Physiology 24: 823-831.

PAGE 109

98 Libby, W.J., and Jund, E. 1962. Variance a ssociated with cloning. Heredity 17: 533540. Lopes, U.V., Huber, D.A., and White, T.L. 2000. Comparison of methods for prediction of genetic gain from mass selection on bina ry threshold traits. Silvae Genet. 49: 50-56. Lush, J.L., Lamoreux, W.F., and Hazel, L.N. 1948. The heritability of resistance to death in the fowl. Poultr y Science 27(4): 375-388. Marino, T.M. 1982. Propagation of southern pines by cuttings. Comb. Proc. Int. Plant Propag. Soc. 31: 518-524. McKeand, S.E. and Bridgewater, F.W. 1998. A strategy for the th ird breeding cycle of loblolly pine in the south-eastern U.S. Silvae Genet. 47: 223-234. McKeand, S., Mullin, T., Byram, T., and Wh ite, T. 2003. Deployment of genetically improved loblolly and slash pines in the s outh. Journal of Forestry 101: 32-37. McRae, J.B., Stelzer, H.E., Foster, G.S., and Caldwell, T. 1993. Genetic results from a tree improvement program to develop clones of loblolly pine for reforestation. In: Proceedings of 22nd Southern Forest Tree Improveme nt Conference, Atlanta, GA. June 14-17, 1993. pp. 424-433. Mullin, T.J. and Park, Y.S. 1994. Estimating genetic gains from alternative breeding strategies for clonal forestry. Can. J. For. Res. 22: 14-23. Murthy, R. and Goldfarb, B. 2001. Effect of handling and water stress on water status and rooting of loblolly pine stem cuttings. New Forests 21: 217-230. Namkoong, G. 1979. Introduction to Quantitati ve Genetics in Forestry. USDA For. Serv. Tech. Bull. No. 1588. 342 p. Paul, A.D., Foster, G.S., and Le ster, D.T. 1993. Field perfor mance, C effects, and their relationship to initial rooting ability for western hemlock clones. Can. J. For. Res. 23: 1947-1952. Paul, A.D., Foster, G.S., Caldwell, T., and McRae, J. 1997. Trends in genetic and environmental parameters for height, diameter, and volume in a multilocation clonal study with loblolly pi ne. For. Sci. 43: 87-98. Reines, M. and Bamping, J.H. 1960. Seasona l rooting responses of slash and loblolly pine cuttings. Journal of Forestry 58: 646-647. Rockwood, D.L. and Goddard, R.E. 1973. Pred icted gains for fusiform rust resistance in slash pine. In: Proc. 12th Southern Forest Tree Improvement Conference, June 1213. Baton Rouge, LA. pp. 31-37.

PAGE 110

99 Roff, D.A. 1997. Evolutionary Quantitative Genetics. Chapman and Hall. New York, NY. 493 p. Rowe, D.B., Blazich, F.A., and Raper, C.D. 2002a. Nitrogen nutrition of hedged stock plants of loblolly pine. I. Tissue ni trogen concentrations and carbohydrate status. New Forests 24: 39-51. Rowe, D.B., Blazich, F.A., Goldfarb, B., a nd Wise, F.C. 2002b. Nitrogen nutrition of hedged stock plants of loblolly pine. II. Influence of carboh ydrate and nitrogen status on adventitious ro oting of stem cuttings. New Forests 24: 53-65. Shaw, D.V. and Hood, J.V. 1985. Maximizing gain per effort by us ing clonal replicates in genetic tests. Theor. Appl. Genet. 71: 392-399. Sohn, S.I. and Goddard, R.E. 1979. Influe nce of infection per cent on improvement of fusiform rust resistance in slash pine. Silvae Genet. 28: 173-180. Sorensen, F.C. and Campbell, R.K. 1980. Ge netic variation in r ootability of cuttings from one-year-old western hemlock seedli ngs. USDA For. Serv. Res. Note PNW352. Stelzer, H.E., Foster, G.S., Shaw, D.V., a nd McRae, J.B. 1998. Ten-year growth comparison between rooted cuttings and seed lings of loblolly pine. Can. J. For. Res. 28: 69-73. Stonecypher, R.W. and McCullough, R.B. 1986. Estimates of Additive and Nonadditive genetic variances from a clonal diallel of Douglas-fir Pseudotsuga mensiesii (Mirb.) Franco. In: Proceedings of the IUFRO Joint Meeting of the Working Parties for Breeding Theory, Pr ogeny Testing, and Seed Orchards, 13-17 Oct. 1986, Williamsburg, VA. North Carolina State University Industry Cooperative Tree Improvement Progr am, Raleigh, NC. pp. 211-227. Van Vleck, L.D. 1972. Estimation of heritabili ty of threshold charac ters. J. Dairy Sci. 55: 218-225. Weber, J. and Stelzer, H. 2002. Operati onal rooted cuttings in southern pines. In: National Proceedings: Forest and Conserva tion Nursery Associations 1999, 2000, and 2001. Technical coordinators R.K. Dumroese, L.E. Riley, and T.D. Landis. Proceedings RMRS-P-24. USDA Forest Service, Rocky Mountain Research Station, Ogden, Utah. pp. 91-92. Westfall, P.H. 1987. A comparison of va riance component estimates for arbitrary underlying distributions. J. Amer Stat. Assoc. 82(399): 866-874. Williams, E.R., Matheson, A.C., and Harwood, C.E. 2002. Experimental Design and Analysis for Tree Improvement. 2nd Edition. CSIRO Publishing, Collingwood, Victoria, Australia. 214 p.

PAGE 111

100 Wise, F.C. and Caldwell, T.D. 1994. Macr opropagation of conifers by stem cuttings. In: Applications of Vegetative Propagation in Forestry. Editors G.S. Foster and A.M. Diner. USDA Forest Service, S outhern Forest Experiment Station, New Orleans. Gen. Tech. Rep. SO-108. pp. 51-73. Yamada, Y. 1962. Genotype by environment interaction and genetic correlation of the same trait under different environmen ts. Jap. J. Genet. 37: 498-509.

PAGE 112

101 BIOGRAPHICAL SKETCH Brian Baltunis graduated from Southern Il linois University at Carbondale with a Bachelor of Science degree in forest resour ces management in 1995. He began graduate work at the University of Maine in August 1995 where he concentrated on forest genetics and tree improvement. He received a Master of Science degree in forestry from the University of Maine in December of 1997. For the next two and a half years, Brian worked at Boise Cascade Corporation in Loui siana where he served as the Breeding and Testing Coordinator in Boises tree improveme nt program for loblolly and slash pines. Brian began his doctoral resear ch at the University of Florida in August 2000.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20110330_AAAAMN INGEST_TIME 2011-03-30T20:47:47Z PACKAGE UFE0012944_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 25887 DFID F20110330_AACBDR ORIGIN DEPOSITOR PATH baltunis_b_Page_003.pro GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
465d672ad407627ab8143e5ae7117aff
SHA-1
95a047b60a69e9d83277f21bec88dc96fe70938f
12428 F20110330_AACAYL baltunis_b_Page_104.QC.jpg
c4bea231f1eb8070c66a98cf32f5ada2
4703e2a8b6c74646bd837e4a936813d76630108e
4271 F20110330_AACAXX baltunis_b_Page_100thm.jpg
eff2caf23bee17c23b461d2a4464bdd9
3292d45df6aa79d30bddf8606221f8341a504d5e
59529 F20110330_AACBEG baltunis_b_Page_003.jp2
97857f38bf10e80b8209a782c159e35a
d3f811ed1d6670255cad368a765bb7688cfc34b2
1849 F20110330_AACAZA baltunis_b_Page_033.txt
f087050af071b1dcf7efb01916a49549
01ead23eeb972435e9aaebf2f788c4b0cd6cd65c
26800 F20110330_AACBDS baltunis_b_Page_021.QC.jpg
730ffb8037db9082e54eff794aeb8f77
edfb7aafb73aca908291e020209bd0a659018f2e
1962 F20110330_AACAYM baltunis_b_Page_060.txt
bae5ac9c88660bf3ce5497089d92d84c
6e4800bc6adcbe6ce226e4acfa5006c3fc3fa6b2
5708 F20110330_AACAXY baltunis_b_Page_083thm.jpg
8e45db8a0127e07a8237525a5242f2a3
763ec7db33df10870156a7d4f425af627cd5dedc
86548 F20110330_AACBEH baltunis_b_Page_024.jp2
f512f3a10f3b6c56055d04a83f97c798
ece20787369f53c1d2c1a0059736e4c22194facf
47803 F20110330_AACAZB baltunis_b_Page_055.pro
d8282208b5a05f57ad9a1e773ce7a1e0
8d7a5d4bf8b82b2c2646a8fd9e4bcd70e7498ea8
68815 F20110330_AACBDT baltunis_b_Page_012.jpg
7875120516511d81164f5d53781617f3
e9f689aea983ea4775e4501acf6df03d5f780b47
39575 F20110330_AACAYN baltunis_b_Page_097.pro
cc902308891de405bdfde6c1c1142abb
e1fb8b626b64f8b3ea3abf8e40f633ba0f72f323
98584 F20110330_AACAXZ baltunis_b_Page_106.jpg
6c97f5eb530da1550da522a3e31a67e2
5ad877d6512a7d1cf546d044ebbbdadf5b465933
1053954 F20110330_AACBEI baltunis_b_Page_065.tif
529053f7cb039778f6178ffda1953957
6d46d3fca3417ef86842480183e416cf5ff0083d
113138 F20110330_AACAZC baltunis_b_Page_046.jp2
c0851343eadab8a61ea8eeb60c8df22f
4499f282a5e0e5189324ba603fdfb80f563d6fa3
25141 F20110330_AACBDU baltunis_b_Page_061.QC.jpg
6c8daf5b60c2389100027eae9b47c3f7
52cf0ed6835e78c66c7488f66e6c307075933152
6099 F20110330_AACAYO baltunis_b_Page_031thm.jpg
c14a89ee77b0161192b88176a235bb40
954c30831125dadf50ec10f2e4d8e03c026838a1
4180 F20110330_AACBEJ baltunis_b_Page_009.QC.jpg
04413a240fe830d364447bcbea437573
6e0423751e553e66a6f94e656f39008a7b9a7e9f
76721 F20110330_AACAZD baltunis_b_Page_035.jpg
eef39d40c380c36fe78fbc6a7087ea1c
8511e88e0bd032aba152829153d7f639a4021b0b
59800 F20110330_AACBDV baltunis_b_Page_078.jpg
027b2b8c4e55817962726fff140cf1bf
50376096e8fe297835c682456feb327cdb630f21
24562 F20110330_AACAYP baltunis_b_Page_043.QC.jpg
dffcaa4d73fa180674a478ea604eda53
b2baf96f7b78197ae4fc22d6d8e692709b854523
78321 F20110330_AACAZE baltunis_b_Page_073.jp2
7eac55079d418378b4a704cc529516c6
522a3a11870adf427ceef14b90a953ced80bb6f9
44164 F20110330_AACBDW baltunis_b_Page_070.jpg
8e60b0135d9bbfd0049cb523c1213ecd
d6e714fd18522e1cb0b11eedac7bd21879ab89f1
74626 F20110330_AACAYQ baltunis_b_Page_097.jpg
ac85d91f03a0bf95aee7a6169cc6ac5a
eaa70ecb39b23aca9c6c196d4abfd358de66a842
6729 F20110330_AACBEK baltunis_b_Page_106thm.jpg
51b9f65e806f2039f8d9cf80c2298fcc
86c22220a83b408c596d41ac728929b3b262e843
25907 F20110330_AACAZF baltunis_b_Page_063.QC.jpg
1241d9f3656b853ec68e4b8a32349853
fe18d20f3108a27deceeb8bc94b6c13e7325fdfb
24976 F20110330_AACBDX baltunis_b_Page_080.QC.jpg
4c492bb74046d326548dc5a3f104f91f
c0b1ff2fee8dd910d85c0312642880335765e82c
F20110330_AACAYR baltunis_b_Page_099.tif
0a6182174e433713a0e84ae05160e97a
cd3b958872aed3f2d70636633662b33fecdd6084
39472 F20110330_AACBFA baltunis_b_Page_053.pro
eba0e296ccde21134bf21e365183489e
5c9009eb7c4826e14d3e4d2aa096b5824e4a18f2
17330 F20110330_AACBEL baltunis_b_Page_052.QC.jpg
2729c9c6d9c17fd5d85382e06d9ab188
02e550688e33574ada85b027db83a9c7dd7a91b6
100050 F20110330_AACAZG baltunis_b_Page_020.jp2
5081c7622d4045972fb7c524d1db07ad
6f3b87f8b40ae5c58194de979b81459105fd2a38
18245 F20110330_AACBDY baltunis_b_Page_112.pro
90e15de5569c384e4660272368339ad5
e722dc5020510b6fc6dfbfc4a3c118eb60060ef5
1051919 F20110330_AACAYS baltunis_b_Page_031.jp2
901e4fb651d1ff2c925f71360da76b3d
931c69f5876c49b99402ebee12f7130bafb74e64
109361 F20110330_AACBFB baltunis_b_Page_077.jp2
996b861078a1f5f5d8f3ee4abd9b05c8
7bf47ea0198cb5f33054f52ef8077c43376343f6
61749 F20110330_AACBEM baltunis_b_Page_106.pro
9c0ebed2a357eb01f29842ab790627d9
70f391ab31b77670a2e2398d523d2966145e8e96
104490 F20110330_AACAZH baltunis_b_Page_005.jpg
875f4bc59fc961c18c7728126de54715
d02b95e8208dd45ff4c4e65f5c84c478336fe0d8
32349 F20110330_AACBDZ baltunis_b_Page_094.jpg
d91a9911e61f822fd023919619fce873
2a638db9c99da89aa212c6e2b9a8b626c5ce03c3
1488 F20110330_AACAYT baltunis_b_Page_064.txt
d988917e9f431a67fc48f9515a802a6d
61d87bf952b2d852bd29dbd13e365bf056463f4a
182745 F20110330_AACBFC UFE0012944_00001.xml FULL
c997eb54d587e1e9119575626f24eead
d5935e6effe64cde44d29490f67a1678dbd7ccc3
6213 F20110330_AACBEN baltunis_b_Page_014thm.jpg
7105b213af28e830cb36c78b053b3ef4
7a2b9e177fffa5f12f80ed8fb96806eebb2de84b
61762 F20110330_AACAZI baltunis_b_Page_102.jpg
f7f8c0dc54fd4d5b2cde39af5dd2bed8
bc9d61a0f3d7648bc429b5640432b5a7b8ae7216
51129 F20110330_AACAYU baltunis_b_Page_084.pro
7222a73edaa94004112c87797f226a6c
70b3ff11a2cd03e3597fa343ab4fc99ac1a0248b
49715 F20110330_AACBEO baltunis_b_Page_036.jpg
75628e6fb54ecd3d88bbac87e401b044
8d1a75b430abc6c0782c8b8aba5e167f49fdc4d1
46656 F20110330_AACAZJ baltunis_b_Page_019.pro
e1d8775f174bba44ff05a26e7f35c053
381e90bea9309226379e7094f143604f8c00648b
2030 F20110330_AACAYV baltunis_b_Page_042.txt
2b97663e7ea77e7ffbb4986d8f54f81f
f130fb63e7d2fea3bf381b02797f7839bd9808c3
1552 F20110330_AACBEP baltunis_b_Page_049.txt
bbff91e25c7876a79c8cbaee4cfb095a
ada71f29f6f35dfa0fd88ecf8618a2faae8d5657
F20110330_AACAYW baltunis_b_Page_096.tif
8519d38862dfd249fec377e4aa6f4f83
6c9ba467d13d011f2315bbdbd8cb96b97c91dec9
F20110330_AACBFF baltunis_b_Page_001.tif
ea4ecb4aca8c9168e208cfb91628a7a2
58e99b840769ec5f03a4ad2474ff85cb644711bb
26808 F20110330_AACBEQ baltunis_b_Page_087.pro
7bc81da38aaebbf3b42c73e30fea0371
de6e09cf341940e0ba02835367de56625ba26025
60089 F20110330_AACAZK baltunis_b_Page_109.pro
03a071f226a0f93cd68e50806080a8e8
0a99be1aa6bcc28b2e8a8289371673bdb7f33027
5291 F20110330_AACAYX baltunis_b_Page_030thm.jpg
c1345154a7a7885a8015f2e9c58ff91d
c63fab68eb20f1b8bc5795af83a77e72d382d6bd
F20110330_AACBFG baltunis_b_Page_003.tif
2a0fb8b8520d327c05ba0ad4dc820cec
2e6c26174f5baeeec6b8baa6c9ca2cae14046550
40991 F20110330_AACBER baltunis_b_Page_091.jp2
f148ed25c0973187188914448926f6a9
f4d568b4ee0986bfe15884054d807ede290282ac
F20110330_AACAZL baltunis_b_Page_038.tif
3b51769a0815a619ba6d6fa73eb9137e
009a03a90e1a3339804645f1c158624414277d48
78125 F20110330_AACAYY baltunis_b_Page_099.jp2
8d3d603509a8006cd10c487612f05052
0346852285edce20bc0db793e04005b68c3213da
25271604 F20110330_AACBFH baltunis_b_Page_004.tif
3010f456f337942e82a2defbbc274a6f
94798e3e80666a90873c718eef38972b3b86985a
6049 F20110330_AACBES baltunis_b_Page_019thm.jpg
6998da4b5d5ade4a5f1d2e7edd048233
ef3890f0f6194c8573affa867f267ad325b1eeb6
95735 F20110330_AACAZM baltunis_b_Page_038.jp2
2063fe06c5975702a0fd6acfd8ef7edc
19d9a2be879326f8a48a366a99ed41982562d0e9
F20110330_AACAYZ baltunis_b_Page_084.tif
ab2e1f12c4b1083fde7517dda4c3a164
8edea2a2e7b7c7c5ff24ff8ae3f9d990df50ed2a
F20110330_AACBFI baltunis_b_Page_005.tif
f8f6db757f9cd2b1ef0f4fbb86700b60
9eee314fcc8333348e3b35258e0b1a885cde5e62
F20110330_AACBET baltunis_b_Page_106.tif
124235e65f4c110e6472a1f70d60ae64
5fb83aecaeace3b945d9a13f302cd0df87223279
33848 F20110330_AACAZN baltunis_b_Page_112.jpg
7f952fb800a6d7aa2bcd34dcbf1c8313
7f73d8a90cd0d4515ff68e839423dfe4d11c1867
F20110330_AACBFJ baltunis_b_Page_007.tif
8153ee2c89605b67b96f50258031bdd1
1b6125656b99d1df8d8266c3af6a587b57368bfc
519 F20110330_AACBEU baltunis_b_Page_111.txt
7802ed43f62e19d6ed7bc71683da2e0b
024f104a2e9d6e5d67f79e28f5a8f75f4c08c898
6350 F20110330_AACAZO baltunis_b_Page_015thm.jpg
179b2d39659901710bc2610631ad246d
aae35a4351ac7580423b04acd7e96d8db3ff0d96
F20110330_AACBFK baltunis_b_Page_009.tif
f37fe85175f540462d625e40b915393e
3ebabdbbdc9218fc91e744a90cfbc47dc43d10b7
5825 F20110330_AACBEV baltunis_b_Page_028thm.jpg
d7c7907382cd9b6e86268a47b950aea3
3f2b627fbd1b22dffabdc6af0d4b6c0fca1051ee
23661 F20110330_AACAZP baltunis_b_Page_111.jpg
ee9fc0e8a4a61963dc5178b483ffbfc1
ac2583ce002ea5223cfaf379bc7ef7daa749aa07
5875 F20110330_AACBEW baltunis_b_Page_047thm.jpg
b74929762cb623d230fb72066be07c39
a634ade9ab95be7f5ee3ac1de435ab2ff2b3c99c
F20110330_AACAZQ baltunis_b_Page_068.tif
901f4e6e1fabd6b53ff61fc635e09a9e
d1a1d70c9d490d744811b1c6ae3cb691bf8e3d6c
F20110330_AACBGA baltunis_b_Page_034.tif
30901747d557de5d35531895ac0a6b1d
80539ff6bb7153701d3d3a74217e4d0ce2f039f7
F20110330_AACBFL baltunis_b_Page_012.tif
4a5b5d0c9e50500d07317bf201d829f8
bdfee86a6286b8ffd99f23852dcf5e107f64d1d5
24200 F20110330_AACBEX baltunis_b_Page_020.QC.jpg
cccb92b0d2ee5de877e93733491b23c0
d918588ac2423001cdf1b820e01a30be7cbd81d2
6060 F20110330_AACAZR baltunis_b_Page_057thm.jpg
01b7a1a2b3b4065b3576530a67c0dabf
a0e552fc40c810206545d11c890ddd0edd1203e4
F20110330_AACBGB baltunis_b_Page_035.tif
a7c9b81d199357619d499b131422ffe9
6cdd8ce46b7b2d78f3433acb47785cfdd75a0195
F20110330_AACBFM baltunis_b_Page_014.tif
7d85e7f40e76bd24c26389c9bdb0e83a
cea1ca6870d56d8802174aa628f37c88ddd97f79
F20110330_AACBEY baltunis_b_Page_008.tif
fb6b32065220018f38cba5b34547e0e7
88b419b0888b39a1197d60e2ee473e19c9bc62f5
F20110330_AACAZS baltunis_b_Page_077.tif
14b9d5fb7fa02040be23e608874a90f4
fdc49236685415e9f6bde3ae4558a762712ccd78
F20110330_AACBGC baltunis_b_Page_041.tif
00a6eb96b235bbcd22740e137d551023
ba89bdc47be27768d7a14c39db6540e5a5dad6c2
F20110330_AACBFN baltunis_b_Page_015.tif
6bcf3b66acf8d2c418fff13d967c7ec2
1bf7d6f2153c717f13b38d75dcb7875a571924d7
17887 F20110330_AACBEZ baltunis_b_Page_094.pro
5202836d28deed8b2cad0ad114d320c8
6fb3d427cff9f49e0207fe51fbc729e3804bac36
5673 F20110330_AACAZT baltunis_b_Page_096.QC.jpg
5f922d477af7661398b9571ca40271fe
12294f6a71fa62690f0ec5115521cddb947bf43a
F20110330_AACBGD baltunis_b_Page_042.tif
4067be8f7bebfe295690f45d37c3f73c
08920c4389ac11e49918a5a7bb5e1f4d2b000eb7
F20110330_AACBFO baltunis_b_Page_016.tif
612abd32a81a7df6ffdffb245a68eeeb
dd0678dde1d3d0111c70af0f7c4320b65fb8535a
76458 F20110330_AACAZU baltunis_b_Page_044.jpg
4492efaca7ce0f97c41f5f2eafc17b01
dc3ef2943a1add8ee220219a3c11ed64cdb9530f
F20110330_AACBGE baltunis_b_Page_043.tif
485b66c0ebc81bc1fc4d50b2da8f643d
e71420402b203fc1f0e49eaf7c910078acc33958
F20110330_AACBFP baltunis_b_Page_017.tif
ca7f83383fa43df4fbbda8b990c2d177
89f075f58440632cf8e70ad8137539b79a7d25a0
6475 F20110330_AACAZV baltunis_b_Page_081thm.jpg
0b4ec9629cd19e6eacb1177ac9fd4601
4e788aeab40e4839ae10e60068280fbb2900601c
F20110330_AACBGF baltunis_b_Page_044.tif
a99f9afd1d58a3339a2bdc7088501eda
2648f9d9a00072ff8c0040c46182366e060a68ed
F20110330_AACBFQ baltunis_b_Page_018.tif
dfae968fb04d6935f48a7813b263e275
2bcca59ba5dd6f0cc316e82e39f1cb5567fce4f1
59306 F20110330_AACAZW baltunis_b_Page_025.jpg
aa0749709e8ee401f0cec29caeb88df3
a1827f9535286c60508d721c7323f417511d0a2b
F20110330_AACBGG baltunis_b_Page_047.tif
49e428c099871720fd1993fa3382b836
1fbefacaa1799c00903c9d8698d9ab242545899b
F20110330_AACBFR baltunis_b_Page_020.tif
86161bdd79bbb9a75407684c6ed56802
3a67a0f76ff6a5021e1a54d96ec7412c19b18dbb
51478 F20110330_AACAZX baltunis_b_Page_015.pro
97e493d07a74f6ff9571a63c15685f1b
0a4ab5b82732ae469218fe352f64165649aee49e
F20110330_AACBGH baltunis_b_Page_048.tif
3b96b4c7f281e4834249388518b8cf4a
2f0308719c3bb568444429e9a0a1620e42915ea4
F20110330_AACBFS baltunis_b_Page_021.tif
28da527f02889927fbc882b7e471da78
cbe3f59e1cecf366e5abbff99d19554306384c94
6001 F20110330_AACAZY baltunis_b_Page_020thm.jpg
0e9360bfd84f4eb6d2cbd562e2d4590b
a206a85e1a7bf9870570f8271f7ad194af0efcea
F20110330_AACBFT baltunis_b_Page_022.tif
369e1c3603b826ae19dff955b13bdbb9
a1473a340a5467c3573d3f443cf482d18040f11c
22659 F20110330_AACAZZ baltunis_b_Page_017.QC.jpg
1d94e2a89bec1f5633f888f739ca6d28
9bd874e163515de0e8b987cd09b3c131d14ac6cc
F20110330_AACBGI baltunis_b_Page_051.tif
5261c3bcb6c8b0728000a2e3322593d9
d6bd3250d65298267d9c153b610f243fac937e8c
F20110330_AACBFU baltunis_b_Page_024.tif
8ede813b53f2770b7606661dd9bd8eee
400de942424a63c7413d1183f375c976c46c6c27
F20110330_AACBGJ baltunis_b_Page_052.tif
ad320d3a2ba66d67d5a749c841cb7d92
458502c7810409ba92d4f1caf103363140eed295
F20110330_AACBFV baltunis_b_Page_025.tif
e568da1fd18d9af16563d92e79f6583b
a2a206f24a08a7734d0fd89b888b4c6ae9d0de0c
F20110330_AACBGK baltunis_b_Page_053.tif
26ec4aadd92a710c1d4a91f3d3b68eb9
ac67da9085883b154ad144dd8194963bfd59567e
F20110330_AACBFW baltunis_b_Page_026.tif
7bc14b985f59aa35bff5d9716adb5ffc
a2af75175e9ac5c4f9fef08a1cf8dc5b5b174ed0
F20110330_AACBGL baltunis_b_Page_054.tif
4fe68eeb42a5ff5e0e4409d03a40b838
c59e6e05db75f2cfbfbd261d196a222d5c09e580
F20110330_AACBFX baltunis_b_Page_028.tif
a2e57b367dc49ffc146765d95d097d69
b284024df179f8c85e44c7ae801bbf5ce36ea72b
F20110330_AACBHA baltunis_b_Page_088.tif
56a519a0766944f297313d59f0dc776d
9a2eb8321e791a0f0e8ed80eb3840e61adfc65f3
F20110330_AACBFY baltunis_b_Page_029.tif
527506f5b5254c717465d19fed01d213
193f6706756460beacb1dc4f4137bbb1044f4581
F20110330_AACBHB baltunis_b_Page_090.tif
dae23cd796f010b2391d9892dd63fd8e
964fa0ee1a4be9fea188ed1b4d291913cfc3f9f9
F20110330_AACBGM baltunis_b_Page_055.tif
4d53e2f1789c4efc8ac15f00cf85f9d1
c27dceb975d8023ac58708664433b9b3c457c570
F20110330_AACBFZ baltunis_b_Page_030.tif
c80fac4e53423214b4f1e731e6f6c85b
91429af54aff79a6ffbf91747adc49e309f3254e
F20110330_AACBHC baltunis_b_Page_091.tif
ac1394e6de8a0cb00dda3f7b41f0d2b2
601e411ca1b3ae7f14bd0630c5f14f0cac6842ee
F20110330_AACBGN baltunis_b_Page_061.tif
ce2cf298551c9f09a9050e4d0a710afc
ccf30bf588d1444009a53c56cfc5a6e16aff1e11
F20110330_AACBHD baltunis_b_Page_092.tif
fd6878d85458e7bd0da5887b124213d0
dcc5c7d8200597ab2964d3cf7b468b6d7f3c96bb
F20110330_AACBGO baltunis_b_Page_062.tif
0ac0de26db8e27a230d12ca57046b824
d4ae249b284b899b724360c3a60f433c617f02e1
1054428 F20110330_AACBHE baltunis_b_Page_093.tif
2862dad912c2662905c7321b7a398985
4ffc661d326207fc01276f4e58f978a6628b8ca3
F20110330_AACBGP baltunis_b_Page_063.tif
e48d806877a3a4161f0825031512dfb6
37b9209dfe93d429d480a818173560ff690d1084
F20110330_AACBHF baltunis_b_Page_097.tif
cfd872caf2b10155f146e2c4773c839d
a8967bc75591b3ae95c8b0acf8af25ab921ec14c
F20110330_AACBGQ baltunis_b_Page_067.tif
54ab02af2b23d92ce0af6c6fae247f84
0fed2a44774c2e81184c6fa3b8ca98a4d5952c9c
F20110330_AACBHG baltunis_b_Page_101.tif
51f5366c85dfdfcf9234ee25ae5f3232
3ed94c4676d85d6b3289d96684051560ffbb669b
F20110330_AACBGR baltunis_b_Page_069.tif
9b766f7390110ff11ac318368f76e36d
52b984bca3beb31a226418335f0bb26548316780
F20110330_AACBHH baltunis_b_Page_103.tif
8289ce7b09ea47462ae0ec67aac5444f
74510a8476ea0a89ec79b67f3fa51d16dcaa7d90
F20110330_AACBGS baltunis_b_Page_071.tif
5456bcf1d6222b4eb6cf2cc58514740a
359074780d55abf095d781f44d4737a8b2b36dd1
F20110330_AACBHI baltunis_b_Page_105.tif
ee4defa1d2406e21ea91449bcbc89cb4
f0216ab805199b4665c3ef30596593920f649470
F20110330_AACBGT baltunis_b_Page_072.tif
a23b7e3ad1e6c4b3076d608e05a47f76
0229a56ea4ebbfe629c252843d4a23ec0507bb30
F20110330_AACBHJ baltunis_b_Page_108.tif
066e2152567b0bee700d789e897c6d99
b559c77239a1fcd2690824a3e96235461cf3ba98
F20110330_AACBGU baltunis_b_Page_073.tif
86419a499ec9ca2acc46c064ad69da26
f2e5c3eed932d26829089842243674bc964d0b75
F20110330_AACBHK baltunis_b_Page_111.tif
e84afc6ec827df9c02fc72751b31d400
3c667c86739f672b7ad10e8108189f1f579267b1
F20110330_AACBGV baltunis_b_Page_074.tif
4373b56e67024bfd15045b2a4d2f84a5
aa01b2eccbc49a5584dc057629ffe6c3c787aded
F20110330_AACBHL baltunis_b_Page_112.tif
f2a4b0a33a22767a07594a599ae1d55a
fe831f3fc15d82b986b01b344a28b9ae4f5291e7
F20110330_AACBGW baltunis_b_Page_079.tif
ed201eb4cbfd81f6049f8a4723330e22
0bb613738f4f048cae7c79173f2847c4af4b6607
1831 F20110330_AACBIA baltunis_b_Page_027.txt
f88ea84dae0319147bb970fb22618916
de5f17743f63d1aeb3658cfb036bed7e4bd12868
1091 F20110330_AACBHM baltunis_b_Page_003.txt
af24f1494266ee6ffacf3a8158da3a7b
c6aaa3d3b31ac41e08d59cda89a559fd52148727
F20110330_AACBGX baltunis_b_Page_080.tif
052eb09893ddc79b23c6ebf23481eadf
d2799d2cf1f5e43de690e1e3b24f9a155aae485b
1870 F20110330_AACBIB baltunis_b_Page_028.txt
5a939d7f4b9066287692494df9592774
c980406c0b526c8572c258cfcba9ccd4ad969974
F20110330_AACBGY baltunis_b_Page_083.tif
e96bb8cb52612629dbb50e95ae18b530
fa47ac89937ec439022ef4422d5e95c999021124
1808 F20110330_AACBIC baltunis_b_Page_035.txt
5d93b031a7403a1036b679c455afb999
6d4bb10105899021c26bcd09d6947b180db8cb9a
3625 F20110330_AACBHN baltunis_b_Page_005.txt
176451f76c954a2c646548ae5deadf0c
e4f6621f51d1f135c0d776c22d8d82d0f58190bd
F20110330_AACBGZ baltunis_b_Page_085.tif
bfa5afd4a5ac6f2bf3d5306cd8df307c
77293ad967002e42ce65bca9412fe00c2ed6331d
1811 F20110330_AACBID baltunis_b_Page_039.txt
e2d56eeabf4c6e36035805978d2e237e
5be46719c329a4e3ba63342f7af1fed5efb4e5a9
2699 F20110330_AACBHO baltunis_b_Page_006.txt
d23c3a17284205340ba3710c182507a6
88548140446d03de098a5fcd6b33781b79656b24
2099 F20110330_AACBIE baltunis_b_Page_045.txt
10ae7b6621e7cfb6639917390fc28b9e
e4a28cf2253cbaa206fb6622ef1dcd4367f3ea1f
2812 F20110330_AACBHP baltunis_b_Page_008.txt
051b29e207b979ea1ccc92103d61f117
f48670c4a1297709a1c27cfa5fc642c26ed3f4cd
1263 F20110330_AACBIF baltunis_b_Page_048.txt
03dd358d6e5e57b3684472baf403501f
30fcd66da23828bd1eb8b3bf407d08e33c9b9d3d
1454 F20110330_AACBHQ baltunis_b_Page_011.txt
05490f235921c46b3e0c4b7e1053c16f
1b1eedf84e49f4ae42578ff068fecfb4b063a541
1416 F20110330_AACBIG baltunis_b_Page_050.txt
b18f4d148413fcf686f43b65dcec5b48
691fa3a8c06fcb7d13b3c141fb7f85fc4ab44f4b
1671 F20110330_AACBHR baltunis_b_Page_012.txt
6f15c5402ea5dad7b16e11879a3d48fa
69602e207d47418c82fa89d682ccd58bb77f39f5
1412 F20110330_AACBIH baltunis_b_Page_052.txt
9f06212e1c8140250939e84c344b59d3
706cb84430d71042e4bf9e2301fbf1d2a906db00
1970 F20110330_AACBHS baltunis_b_Page_013.txt
0214df2087902153b75cd26c6440beda
48f269730a3a092b1b48d73cddd4d5f043a58852
2005 F20110330_AACBII baltunis_b_Page_059.txt
b0146b63820b9052faef303d17187e05
9b53f1da63c21936db9eb1ca4e0cefb5096595b0
1982 F20110330_AACBHT baltunis_b_Page_014.txt
66a76fd78e049e9af9cf5282f46d4e87
1b17a5cb12248f8cc19d02e910f3f19af401486b
1957 F20110330_AACBIJ baltunis_b_Page_062.txt
a9a8377eb62526e84fccd8540fa661e9
409f3e2b7c5db42922ffeffeeda599ea4cbbde8e
1949 F20110330_AACBHU baltunis_b_Page_018.txt
41b3f2d2a0278ab210882601f95e9562
845b0ab666bb3d40600a339e20e92065a1fe5ae8
1959 F20110330_AACBIK baltunis_b_Page_063.txt
c0c2a5c7478fab60b6b2cb926e6d474a
337887e63ec9ee127ee3b2bb45db8150a196f843
1873 F20110330_AACBHV baltunis_b_Page_019.txt
bb4948bbaed5c7594ee4503905a6d28e
dd5a3fe4c6c629230927f29dae324d252da8f650
2049 F20110330_AACBIL baltunis_b_Page_065.txt
30fd3b7b4febf138689b6a9b670e4bbc
a6f75b08d3e89986298550ac9d6dc6afa786265c
2404 F20110330_AACBHW baltunis_b_Page_020.txt
4d27e21c6640922f1be144e9822cb112
98169fa81567f5887ddf0d8c98e1660e48caf285
2041 F20110330_AACBIM baltunis_b_Page_067.txt
d5ec66f7cd5f020b7a3e5a9bc18ffcd1
a664bba775ef3195cdb79370b50fc6da211a8952
2044 F20110330_AACBHX baltunis_b_Page_022.txt
32ff19dc47f34b93e074aa2b382da7dd
33034c705454383c71e6f5946d7657dc1ec4062e
1616 F20110330_AACBJA baltunis_b_Page_098.txt
a5ad4b3d5fd83ca99bc3f1931fac9f6a
764b8c2121e648f1fff54689515a7c4235875e1b
1830 F20110330_AACBIN baltunis_b_Page_068.txt
e3bac5c8e48d55b659d7d86902a4997b
cf8c2ce2d73c13193daf0f457a18cb7509a14d46
1642 F20110330_AACBHY baltunis_b_Page_024.txt
ade3d07c6fec432e9b83bb927ffbfbc9
5429f99017ce52c9d2c35ea28b1e3f675731d792
1443 F20110330_AACBJB baltunis_b_Page_100.txt
7596089296ab4d1aca7a1e1533df04a1
446eaf03d2fbf55ddc02ea755b2f60d39d339b40
1591 F20110330_AACBHZ baltunis_b_Page_026.txt
9687a156f9040239ee17ee56b0b87ce4
d34400cdc246ee744dd17af52c00cf4a8bb42d63
1460 F20110330_AACBJC baltunis_b_Page_101.txt
993f1692fa7f265e295fca6c6ac20fbf
82b0a677e84e3e759d2a71a1e5f04d1836ab136b
1828 F20110330_AACBIO baltunis_b_Page_069.txt
bea9a4d15f4fe09b132e312bda4f8ec6
0d317c86f31a3ac05147f69fd2a60922390f28f6
2278 F20110330_AACBJD baltunis_b_Page_105.txt
af7b07b6c8b7cae2f9f05829a7aa02f3
41cd0175ab441b4bc54df3c7b79faa9adc41aea6
1232 F20110330_AACBIP baltunis_b_Page_070.txt
77b7ca2ec9715dfb560da9df5fe333ea
3a81222f840590bb951711a59df64a6dec12afb7
2524 F20110330_AACBJE baltunis_b_Page_106.txt
4eaa90f5de383ad418bdbf2dc7bdd6f7
15047d6e8688bfd57d783ef51fdd0ba66488534b
2215 F20110330_AACBIQ baltunis_b_Page_072.txt
c1b72ba7c35e6009d84fe33d704ad6ee
a7385adeec15734ca383367abb4841b5dfaf27be
2481 F20110330_AACBJF baltunis_b_Page_108.txt
fcfd349d6fe2833c2901a83cbaa07c1b
d84c0fdf318f39f5d1ca2aa20ff12fd636ceece2
1824 F20110330_AACBIR baltunis_b_Page_073.txt
20c5dc4f8a89193c55295db566ef9db5
b82a1ed89703cddb58f88aad39a29a37fe292449
771 F20110330_AACBJG baltunis_b_Page_112.txt
f2dd20415f39d2a74ca1df7491d13b3c
c859ae49cc25052bfec6205b7ff8961b42a29ac5
1730 F20110330_AACBIS baltunis_b_Page_076.txt
efb206971cf2377322201d6b014673ff
e38b92c63d8affc0f7486d8e2cc7ddce5e075b60
1359 F20110330_AACBJH baltunis_b_Page_002.pro
d409cdfce3ad65a032436801f2e85379
143bb1351ac23ee1512507830ca5138a08172335
2003 F20110330_AACBIT baltunis_b_Page_077.txt
f599c85f5fa795def19e454d5408b582
6baf394cb9eb77f5f9a6e559bbb387ef16bdc8b4
89537 F20110330_AACBJI baltunis_b_Page_005.pro
9dfa24b5ba631fb13e32b08d84505d01
44987fbcc07aec110c7d2990f90157dbc721bf3f
1709 F20110330_AACBIU baltunis_b_Page_082.txt
fe7fdb034a728c0037ca665fe18c4cfc
02edbd82530a90447285861c27576f30d013b4da
66209 F20110330_AACBJJ baltunis_b_Page_006.pro
39008e700d97cd187cf2a0c9bb4b3a60
4a565c5fc5f0e67581b9e500395099be85eaa13b
2012 F20110330_AACBIV baltunis_b_Page_084.txt
a12a6900ce3ac6acbd5ca4db90c81fd3
c2a00ff5bdf6e88d108210fc4ef54d1871d26238
7327 F20110330_AACBJK baltunis_b_Page_009.pro
8f68642a2e00eb04ddc73985f247d6bd
cebf666a9944cce51d236dc5f44ed7aa5cd1180e
1408 F20110330_AACBIW baltunis_b_Page_086.txt
bdd6f37fd95b4d145dc61f932500b7a0
dfd8fa063b80d32c8d58b44bb7f55a49b2208a5b
37209 F20110330_AACBJL baltunis_b_Page_010.pro
85e0d02005e44dddab65dc6e408d5e32
d1a42f69b4b910d7e98cf50292f77affcbd67764
1072 F20110330_AACBIX baltunis_b_Page_087.txt
3c05d3ab3b042f3ab59801d4a6e530b5
89cd1889438df4dc89ee3d09aaeff4a45a849ee3
44292 F20110330_AACBKA baltunis_b_Page_041.pro
39e4300e1a1f22138dfd2bee5cecb4b7
e1631b52b9f6dacda575fc5f9fb69d0d2c124bf1
49887 F20110330_AACBJM baltunis_b_Page_013.pro
0bba6d1f8d583d958842822b7f691155
9e217cfa1356ce0fc5a0e9fa91368e497f022f6b
1854 F20110330_AACBIY baltunis_b_Page_089.txt
0b4c0ad1d4ee0e3166a074b9436749cf
fb9c25bfacfcb10a35ebf0ccc30a6cdb3b4f64c4
48505 F20110330_AACBKB baltunis_b_Page_043.pro
bc0bf31ed7da7279352f3c6ee88a9ad4
0a150f1ee27143ffef37bf27000f8daf83cd09ba
31570 F20110330_AACBJN baltunis_b_Page_016.pro
b19b8d1d283fdcf30b03aaf138ee59b9
00bf12d240584528f9ead05963a030eb0b086b09
681 F20110330_AACBIZ baltunis_b_Page_091.txt
56575092706482d0554988816686f268
03f59c57ec948adc9958a9f676354743e7832e5a
53717 F20110330_AACBKC baltunis_b_Page_046.pro
6a838cfc36d993385a31d47b1b96b6f5
0bedc4f20a2127db0f8bc28612ef0225a9a98113
50796 F20110330_AACBJO baltunis_b_Page_021.pro
64acfcfa8629a4805baba16a727da86d
8d754bdb7b56cf3e34af109f1f0a1f8cc8dafa90
31395 F20110330_AACBKD baltunis_b_Page_048.pro
c6139bc9d91eec52f354d13378ba1fcf
f6ead99ae3053b670c58ce88e1df027fe00e85fd
39019 F20110330_AACBKE baltunis_b_Page_049.pro
dfd7768ccaff63e9943c0fe29d934213
5fafc9c6dbcf4123e6c1ad2297865f1d6b1601be
52059 F20110330_AACBJP baltunis_b_Page_022.pro
1ffe9d48904d73a397b9df5165d75a0f
183b12ed39d3196511daacb2884335d7d2cf9457
32037 F20110330_AACBKF baltunis_b_Page_050.pro
5346aed2d038417121e59a313d7a6851
839874c1be35302f1c3d6d5eb18ce6ae959be1dd
38501 F20110330_AACBJQ baltunis_b_Page_023.pro
c6c1def6eee381d56478c39032b240d2
99a23d10aa53af8035f3cc266613a23b51efa77a
46994 F20110330_AACBKG baltunis_b_Page_054.pro
e45b01a6987fe1bd2517811fa7144bee
9ab7a30ee2b63ebc1c3cdf1672252aab0f7e0867
35526 F20110330_AACBJR baltunis_b_Page_026.pro
de940d6661bf02ce4f82865852b097c3
c5b28184cd1b84d124501e4f0e57f983946524cc
50968 F20110330_AACBKH baltunis_b_Page_056.pro
be13a95c3dbed3d396ab7e784882ab82
ee9640f3825807c46647b46f9e0192e56cf408ff
43590 F20110330_AACBJS baltunis_b_Page_028.pro
a87bfba06124958573a7afa068f9ed2e
46bbf063569259103eeb69226dc211e3d0fe0c3f
51167 F20110330_AACBKI baltunis_b_Page_058.pro
f86ffe6bda4b7d29634ac82a0b1428ec
85743cd3317bc0d5f182b656c348b807533e924f
53585 F20110330_AACBJT baltunis_b_Page_029.pro
26b4c5e58bb358ff557f75fb972338b8
1568d4fe64ea4be37fe5d6ad5fc3e8d0c3918f67
48110 F20110330_AACBKJ baltunis_b_Page_061.pro
0c30ffee6291dc0aff1c10e82e183f77
986c8902b29ba751ee5c5a7db74bcce0e25f9e84
50933 F20110330_AACBJU baltunis_b_Page_032.pro
75d51f67e0b3cc7a77285652e8b2c38b
50bb1f17dcb4adc14b444cd74153513435b194ae
45102 F20110330_AACBKK baltunis_b_Page_062.pro
050d78bd886d83652c51e130f6dd94d6
6224c6a29f63f0ed5f61566f6fbb1a14346745b0
47424 F20110330_AACBJV baltunis_b_Page_034.pro
f608c9a9d40ffd3353f19b1d090969c4
8da7a618d6384e7a11a6a70a531fdc176d5a18ef
49597 F20110330_AACBKL baltunis_b_Page_063.pro
ed0c62ba96f007ff0f304e7a85c85da4
95657b03c6882d8f2bdf3828cae9f3b0bc282be5
18511 F20110330_AACBJW baltunis_b_Page_036.pro
e9b3f2fb0376608a82ef805039fe5139
65b61d624f740ff603aefc291edca28609e3011f
52065 F20110330_AACBLA baltunis_b_Page_085.pro
383ca0d23e144158961672f06fa1aaa6
2301c77bf3a529c05a74604787de3ea27475f011
28099 F20110330_AACBKM baltunis_b_Page_064.pro
6a089fe4656d2ae1d10a66051db489c1
c34fc66ecccb3f50acd9a18c29bf1ac6a3bef78b
43154 F20110330_AACBJX baltunis_b_Page_037.pro
a5c50a8c4baa0dced0d3b5e8cf038000
c21c642bdbc9a0f6485b0c7d00c9ec776000aadd
44972 F20110330_AACBLB baltunis_b_Page_089.pro
2f8907554072af14ea130ed4f01c9d09
4d3aa1daa5fd0924aa020aeb129b34f623392a72
51329 F20110330_AACBKN baltunis_b_Page_065.pro
ba2d480ae429eee964693b351d3cc33b
fc3bcb6b3d95f7cabf4a2d27aebaea8f18dbc0f2
44988 F20110330_AACBJY baltunis_b_Page_038.pro
b8d61e3bbeb53eed6ef92d77981328a9
477c8cdfaa7803e8c75d5f801f37058e74c36d86
51718 F20110330_AACBLC baltunis_b_Page_090.pro
b193012cff4e9fc8bf71e4ed950b83b2
36641ad8407e47f4d056998d2c7b4a416c0c145c
38762 F20110330_AACBKO baltunis_b_Page_066.pro
3149fb5aefaeac2aa199a4325e97f1b0
d6fe55d981cee860b9082b042cbe90d9093f979d
3354 F20110330_AACBJZ baltunis_b_Page_040.pro
882fb7c0986a57c93b3f33041ba8e73d
a7bc91657f9a099e055e1bd6878a8f1bc0455efd
16956 F20110330_AACBLD baltunis_b_Page_091.pro
b8220a165a4a0690a080ff587b3a4b42
3f5fa6097fa11f757df7d9806c941941f6aae901
45472 F20110330_AACBKP baltunis_b_Page_069.pro
811732949a0d195262c7eb9b306dfe66
8af7aa8399716f11e8251b4c4f60e37b8b653e08
3465 F20110330_AACBLE baltunis_b_Page_092.pro
51ea30a2c59c5371c99febed3dd29ba1
97f408baeee82ede1819bb6b9d0e3563d1dba660
27890 F20110330_AACBKQ baltunis_b_Page_070.pro
575d877725c1d6b3781813f0f48ba0c9
afc7ad868a128a681397d77a67dd4d84ce29f663
40320 F20110330_AACBLF baltunis_b_Page_093.pro
80958750e856f66af3118a788905909e
90539ad4e5d19d0c42739d98764369f2149116a4
30173 F20110330_AACBKR baltunis_b_Page_071.pro
fc419a4884ab6139762b83561a71854d
abd9830a20496ed507bb39219a5fe03ff6f91b48
10747 F20110330_AACBLG baltunis_b_Page_096.pro
6ad5ed97e893516b5de27194205761df
287dea96d0e162267a2fef5387f1df6ba973a609
45484 F20110330_AACBKS baltunis_b_Page_072.pro
079976f26f72394c82202e2333e4c95f
ef5e79ede126101ea2f2d616ca8e9383f98d1f1b
30418 F20110330_AACBLH baltunis_b_Page_098.pro
85e385de0d933c4b8cc03a85f7770a15
50a310ab756d063b9a89854ddee59dc14ad9fa23
37717 F20110330_AACBKT baltunis_b_Page_073.pro
4cf9954ff4e2eb9d9b898b2c104d11cb
be59040dcdb58785581691124e5d2e2c360ff77d
30496 F20110330_AACBLI baltunis_b_Page_101.pro
6c7d86983f4765fcdade832981d7fcab
ef9a457fac9d9d96c35cc30f50c5ba4f7c4b6ef8
42015 F20110330_AACBKU baltunis_b_Page_075.pro
ab43b91576d8424945088d084677f2ac
c29c26a77b85a9608041acfd2049b8d97411d36c
30025 F20110330_AACBLJ baltunis_b_Page_102.pro
5931c3ed0427b046d574d91469f52010
3566a5451d9dee86fe5a06a4974b70ec96dd45b7
42553 F20110330_AACBKV baltunis_b_Page_076.pro
18c1cc6501c0d999b7ae094e1d3f1c25
a88f291d8c008f0baa57790077024dff9968adde
29519 F20110330_AACBLK baltunis_b_Page_103.pro
db3451642e74956083f96cb474a4f9b4
dec6f7ff021f7aef67757ef488872ee011a6277d
50759 F20110330_AACBKW baltunis_b_Page_077.pro
ee91514c36e85483030eceb3cd833c0c
6d9c15b0e080415435dc312fb84694f6289e9b53
21394 F20110330_AACBLL baltunis_b_Page_104.pro
67866c1d7d46c04a436056236752322a
a45657557d0fb69d364f648aada0798b5299dd26
38388 F20110330_AACBKX baltunis_b_Page_078.pro
5b28097250d650ecca2150f85c53fee8
62922fac08c2103f973d2c6fda25a00349c257d5
15015 F20110330_AACBMA baltunis_b_Page_009.jpg
53908c8a4d2db2705b1c141bcdc82dea
069c4476b5beff166ae100cc2e652ac7c5a2a80b
56268 F20110330_AACBLM baltunis_b_Page_105.pro
d2034137f9ec932462e17fe9450f4680
3efb999da208d29cc42ab10b344fb465f910bf5d
36725 F20110330_AACBKY baltunis_b_Page_082.pro
de3251e5e7310e5482861f48fb6373ad
06092514b91aad1248d31d135ca5b0f852ca6e86
64576 F20110330_AACBMB baltunis_b_Page_010.jpg
12e24b2f905efe17948d9570d0a71a9d
b7ac13188751e7ac4e5f626c6ca6c9d14ba6a7a2
60826 F20110330_AACBLN baltunis_b_Page_108.pro
0ac4ae760f377f8cf04066197582c2d7
9e8f4d710e213dc156e92f77c6a784b64efa6e4a
36925 F20110330_AACBKZ baltunis_b_Page_083.pro
0cdc5939ca1542243f96abad5bfe0c5d
bd7bb0bc6ef8d38156a7ed53775514237a0c4ce9
19001 F20110330_AACBMC baltunis_b_Page_010.QC.jpg
5085623edf097b95abf6d64b5c834896
f1c0f3610589af71c1d04a97e415b0966bca0e75
63006 F20110330_AACBLO baltunis_b_Page_110.pro
126f438f16931ded2d1539595fad73e9
bab0cfbfd30cdbcffceb929ae48053983ae902a5
21442 F20110330_AACBMD baltunis_b_Page_012.QC.jpg
b1d4c07dd931e91b3d635fb7661f7255
cd9757d1e76ddd3802c01e559a60786a8f1d34cf
11706 F20110330_AACBLP baltunis_b_Page_111.pro
51f0af8a2a040b71a209b121241d4114
8c0cf38bba370342efd97f51f79206342908e415
25284 F20110330_AACBME baltunis_b_Page_013.QC.jpg
efcf154efcb594d838142d9ab43eabe6
495a0e83625dbf1c461fb5a024cfa41de47dc172
6913 F20110330_AACBLQ baltunis_b_Page_001.QC.jpg
e3f1374d818a3148305218f32453e988
279d9a9e33d91dc9215ccfcbb62f821a9713c83f
83126 F20110330_AACBMF baltunis_b_Page_014.jpg
00e7622d07d347c84d753a96519c844d
a864aad199b0df7cc127bf03a1c4fef478bfd9dd
53923 F20110330_AACBMG baltunis_b_Page_016.jpg
cfa3c31432136fe25a1a00ce600c2caf
5094b8b5347c979b00cd083af4c4ae138096e098
14430 F20110330_AACBLR baltunis_b_Page_003.QC.jpg
85adbdff91994bc0d1dc8140d0e013fd
f8f4f34021fe53f8f05ffcfabea1f54bc51aa079
16642 F20110330_AACBMH baltunis_b_Page_016.QC.jpg
d53e0687accd70ff45f77d9413c36315
e10d73cf8c770bd0df9a5095f0fa1374233d87c3
76125 F20110330_AACBLS baltunis_b_Page_004.jpg
460a36e89e59c024b6216808860521ef
497e5faeb6671e064c2ab13eaad2d55df0d71ec4
81459 F20110330_AACBMI baltunis_b_Page_018.jpg
871f8b9b3ee77e28dc9cc26d64f91df8
b42a622b60687bb34e2b39207be173456d98fe6f
16812 F20110330_AACBLT baltunis_b_Page_004.QC.jpg
226747c49b70b4cd4e867153f2d4643e
c7c18de2ffa64be8fd1360d520394c9b7c2812a1
79044 F20110330_AACBMJ baltunis_b_Page_019.jpg
04ee4e71923de18b8818997ff5899bdc
634df452779e4d86bbb4d3977c059d4afc612d91
24310 F20110330_AACBLU baltunis_b_Page_005.QC.jpg
4d7847d765acb6d0ef40746df78ee2b2
6b6e8d4cc227207eb9982294c29c2a1d3a411ab2
24777 F20110330_AACBMK baltunis_b_Page_019.QC.jpg
652ffa22cded0cdfd621b9961c9c45c1
9874dd5534bcd048532b59f50897c463bb63d772
101368 F20110330_AACBLV baltunis_b_Page_006.jpg
b61cfc307093dea9db10ca00144bc41e
d83ab855ab8217e25a122c1d914c668b0fdb8a6a
78022 F20110330_AACBML baltunis_b_Page_020.jpg
6e605664f6f967788a55001ea43dfac3
343b46a9bae781def782f7e491ed9e6998344de6
24347 F20110330_AACBLW baltunis_b_Page_006.QC.jpg
f796d310af3b584fdec74fbdef10afbf
c45d242b3523e92ad3da9dbb694ef0bb5b4f5e62
23210 F20110330_AACBNA baltunis_b_Page_037.QC.jpg
0a5ed3634531780e179f57f6a357d9fe
3acef08737407207e702d4ea5821d4e09c9f8abd
83968 F20110330_AACBMM baltunis_b_Page_021.jpg
8ce54300c9d73270796d2d9f7d5ea286
33d1046c68bdcf6927cfe370307a5414519979e0
109474 F20110330_AACBLX baltunis_b_Page_007.jpg
0a77cafc456eb2461f7490c5aa644aa1
180807d1b5f2abc7f6af4d978f270106e147b8c7
73129 F20110330_AACBNB baltunis_b_Page_038.jpg
81edcd4622eafac66a39c65321d43a2f
bf180db5d2ed12714a4b75163111a3caf37e5203
26706 F20110330_AACBMN baltunis_b_Page_022.QC.jpg
e8f4b5440145c1a773a740b4a9f58c53
0cc29a0dda14d1daa5eaf4763f36867e1250794d
28970 F20110330_AACBLY baltunis_b_Page_007.QC.jpg
bbbce4b8b058f4e0be61775c6a43a017
61a7c1857cdc934911494c280716ac7ab56d5d84
22571 F20110330_AACBNC baltunis_b_Page_038.QC.jpg
57fbb5eb5890a138db4d7186f60b64c4
7f4eadffeeac4f2aaca8c28298427ef3e9236362
63549 F20110330_AACBMO baltunis_b_Page_023.jpg
5c63d30dc48f035f09cd9db1b076de89
3dcc5683fd62974ed37eed98c441ad9599b3d1f1
107794 F20110330_AACBLZ baltunis_b_Page_008.jpg
da6555311288e97bab9004222e9f7a0f
f295df3046a11d54f9aa2ae0d4a0a96ccbdb4aaf
19193 F20110330_AACBND baltunis_b_Page_039.QC.jpg
960c6e027f009c669702056e8df039a5
228213c4450eacf0545102464116094dceb6bf28
18523 F20110330_AACBMP baltunis_b_Page_024.QC.jpg
5ebd71c5144cb889f130bfd12eaae952
25e7924d557f44d1a71b8642f589b46420abeacc
7749 F20110330_AACBNE baltunis_b_Page_040.jpg
81f438cd7a40f2fdb4b9d0c0995c1dcf
cd2c7d345e243de3df411efa65ec54b1089c98cc
18431 F20110330_AACBMQ baltunis_b_Page_025.QC.jpg
c93b61fad4f28ceb017647c0482a3a45
62e5a263c60e5af98b94949e10367f01a9bcb6ad
85867 F20110330_AACBNF baltunis_b_Page_042.jpg
2dbc1ac42c1733cb53dc90314c0e0655
5729b850680b61a6b767c81529729d7eda8eb53c
72480 F20110330_AACBMR baltunis_b_Page_028.jpg
7d24d897b0367a0732e7828408266c1e
abdb752709b0437bf606049ecb84d08de9cfa0ec
81248 F20110330_AACBNG baltunis_b_Page_043.jpg
5732a735dc31396dc5f37a57ede9b8a5
d09ab59b3fdb8131a2ce2d600e430a718660d72e
76276 F20110330_AACBNH baltunis_b_Page_045.jpg
183de98cb7c3840da39a9f6e058b797b
2ffb152c7c06a3178859fd72ca8c6359a6f0d60a
89010 F20110330_AACBMS baltunis_b_Page_029.jpg
617e7bdc4549cc3098a90678dd60aaea
571e1303898da0abdc385aef4c76564558c1b712
27853 F20110330_AACBNI baltunis_b_Page_046.QC.jpg
5710e6b1626939d6f7b5d6deb4c4c530
a694cd423785f9becd677e8f7c47d5c2076f2624
65769 F20110330_AACBMT baltunis_b_Page_030.jpg
450ac13722b08a2c1b271b18b02faa8d
7544047b9110818fdd9f51f975332187c4bd02a8
50802 F20110330_AACBNJ baltunis_b_Page_048.jpg
620b9581385dccc66e7241ea8482c802
8c9d18ec10f925046085154cf085bd676fa72a68
25013 F20110330_AACBMU baltunis_b_Page_031.QC.jpg
4c70b32a89c7cb246d84b90ffdfff272
337986aa008d0dd5f4ce4461620f7987474ceabc
16554 F20110330_AACBNK baltunis_b_Page_050.QC.jpg
1cce70ae95d4e3c0b0558da708e0efdd
e043403ae98eda7c8942c42f9aa806530d59598b
83360 F20110330_AACBMV baltunis_b_Page_032.jpg
e873e70768c918ac4939904a76e0eea9
df65f81c2b19de015e566e9d8039a7ea4856235c
53688 F20110330_AACBNL baltunis_b_Page_051.jpg
ec9e16e54d64883ca8531e86fb6ce1ed
83f19624ad97ed54458de268d2d8a52a75a7e9d9
26548 F20110330_AACBMW baltunis_b_Page_032.QC.jpg
438237cab3ddfbf7a2d7b0b51247cfff
63e2434be13127bb6888dd2a6be0367300b90b62
15966 F20110330_AACBNM baltunis_b_Page_051.QC.jpg
28e2506e3b4ab79b45598486f19467d6
9af7db7916d1ddfd60b102b7b4676e8bcfd3f6e9
23060 F20110330_AACBMX baltunis_b_Page_035.QC.jpg
2a972526221f0d9ce3b80a4bab74674d
ca6d539d6e8abdd94a96018af32514d7209e7bd7
21431 F20110330_AACBOA baltunis_b_Page_066.QC.jpg
2052498cc5714f4c082449699b35e869
4d9962c95d00e590bd93f67e50bae31f142d0997
58255 F20110330_AACBNN baltunis_b_Page_052.jpg
6164a8da03c1c883bc9f8b7e50728e0d
c2e2b7be800b254f7f97425ffca76fd32d4e7851
16220 F20110330_AACBMY baltunis_b_Page_036.QC.jpg
1168bfbceebfc92b1e6e6b6117d79f6b
41abd6c0da1a7c4eeef958e5a918464a2a712de6
84976 F20110330_AACBOB baltunis_b_Page_067.jpg
d8318063cb408a9c7cfb8c7dadf4d91e
36ffe5e918e8eefed2636a4e6eabad8d78fc6fdb
65294 F20110330_AACBNO baltunis_b_Page_053.jpg
f582176e8de63d3818c52800902ffaa0
0b198d6b2cd7fafbf919f4c7d89a9e4ea580352a
75104 F20110330_AACBMZ baltunis_b_Page_037.jpg
bdb419b5ed142bdd68688a5513e985a3
eee5f1e3caaf409fb9634ba8cbcddab19c7392cd
26224 F20110330_AACBOC baltunis_b_Page_067.QC.jpg
d3bf2595b8d16594c4a4890a7df26109
7e01f815ed5231dac77eb396ff5552434da42515
19828 F20110330_AACBNP baltunis_b_Page_053.QC.jpg
260b5213623283aa69bf5657dd07c07a
b4256a56923fd3813196bcc0553c83314cfd3a38
76174 F20110330_AACBOD baltunis_b_Page_068.jpg
f476053deb9fb7c05c8241ee5d5580e8
1c359ed207a4a7dd8de65529d7e7fdabe8f3f027
77903 F20110330_AACBNQ baltunis_b_Page_054.jpg
c9b8599728172f3b6b318721a41d3ac1
18b57a3a73cf41a3ed1feb5badbb249a60d5305a
23761 F20110330_AACBOE baltunis_b_Page_068.QC.jpg
9afe8d415e0f82d2d59374eeb21df768
ba3f5ae2e7db5c9e90fd270196b2c4ede3dfa6fa
25489 F20110330_AACBNR baltunis_b_Page_056.QC.jpg
fc029401742e86d847c8eeaa8de2d6fd
13c03e5923dacc990a043055d93cb134e9c5da10
76434 F20110330_AACBOF baltunis_b_Page_069.jpg
989940699f41f835a7e9c7f280f1ecea
5b0526c61845400daaaf0fb191de408009e1d61f
25509 F20110330_AACBNS baltunis_b_Page_058.QC.jpg
f7ed16e47577a54283f69138fc600f8a
facddb4be738f378ca4a8f028c889e21f7f0eb9e
23034 F20110330_AACBOG baltunis_b_Page_069.QC.jpg
f161d3b9849b7ac427b9d1c60865f1e6
fd9d98842fd6d2ecb262642ee78b5e9a7926c6f2
16033 F20110330_AACBOH baltunis_b_Page_071.QC.jpg
5815636f1f4243b6a1b850c91f71f880
78cd63dc96d614eceea1ecb524159203d56e0973
13530 F20110330_AACBNT baltunis_b_Page_059.QC.jpg
52ffdcf3234227f2847df50d56767c91
b901f4f5281aa532d1c1b47b9998754618f82950
64904 F20110330_AACBOI baltunis_b_Page_072.jpg
2eda1aff3b75ec253ea31c51918d558d
b574d825c6f3de6dca711f74ed3a8b6647a83560
82284 F20110330_AACBNU baltunis_b_Page_061.jpg
e3c8ab8a1e706213d6bef3f5f5d70098
5f4d16a635f498fc482809af6f8e19dbbfcf2604
19663 F20110330_AACBOJ baltunis_b_Page_072.QC.jpg
6b8ace8cc8f99c864ca759883186e284
970fdbe1014c59e2517a06a70e2fda9db85d9c44
75031 F20110330_AACBNV baltunis_b_Page_062.jpg
81fd1ba9ca058c94f3edabfb6c504422
cc7cf933cfd2cff1a10c76b87fc70b8e6d7dcd23
73924 F20110330_AACBOK baltunis_b_Page_074.jpg
2e8ea2ef37f1b713aebcf9394b7a410d
a5eb35a54e21829653cdfaaeb89d65e4044d142b
56803 F20110330_AACBNW baltunis_b_Page_064.jpg
9cd6ade9ad2a7035396ab00c2193e132
d59c45068a48a694cf491da364556ad010792a29
22023 F20110330_AACBOL baltunis_b_Page_075.QC.jpg
a0cfbd03a8607a22280df2b09eea23d0
d74cec88c71b98849cb2993928652595b1173f9b
17739 F20110330_AACBNX baltunis_b_Page_064.QC.jpg
07543c8ae90a966f245fd7e3ed6e53d1
bc2c62faaa9267518815201633660326fd3c1e1b
84335 F20110330_AACBPA baltunis_b_Page_090.jpg
bf75abb3b32f5728456d8d21c96b89d1
7a655496b498921545d9e412b09db5096ed1da60
21974 F20110330_AACBOM baltunis_b_Page_076.QC.jpg
889cdd7d9c4c72170ec4f951864ce588
fad575d6be5b7e4180d85fece546b9424919d2ea
84245 F20110330_AACBNY baltunis_b_Page_065.jpg
525f5ad0c2571cad790b8f6e3ea7d98d
676c858eb46f2f7aa9ca5b9d444663e7599769de
12076 F20110330_AACBPB baltunis_b_Page_093.jpg
91c8d9550e97c9e48cce3e2e939ddaf1
f620751e10a4df60441eb380c49890a6a70f5948
26258 F20110330_AACBON baltunis_b_Page_077.QC.jpg
e9f94201bfd97f82b12d30df96e28418
3960a5a05cacfac304b11172683249bf15ca2065
26527 F20110330_AACBNZ baltunis_b_Page_065.QC.jpg
388513c3c1c0a4cbd4540fdc54ed3713
902f13f3cd87ba5432c08ca7e1f527396fc603c3
18088 F20110330_AACBPC baltunis_b_Page_096.jpg
63f1466594dfc20897dfd9d611112de5
b435d23d9b886e49f184cba1a0ea38e94447aa9a
18563 F20110330_AACBOO baltunis_b_Page_078.QC.jpg
7c6226efecfc19b6eb0b8150f044252f
1a084d68da950f14f60f59362d7afb2d36d15331
56142 F20110330_AACBPD baltunis_b_Page_098.jpg
95699ca5fc003ca6e6d9ccfd6cb11a17
9e8ba9d7e52f24de46a4da2c4663981a2cd4eea7
83910 F20110330_AACBOP baltunis_b_Page_079.jpg
8b1da0b2991841584298b5c1e5f0cb28
1dd14e203c2b10d1d12ddc2c0e1dbe3da24dec28
68325 F20110330_AACBPE baltunis_b_Page_099.jpg
f5fd05216079a8c91c61cb56a9c17eaf
2a9ae7cc1a9b842809d71468b5a5fa814b9cb76d
81144 F20110330_AACBOQ baltunis_b_Page_080.jpg
f919fe1ee2ba2b0438707d379528c87c
b06d0ef578010340373b71d1e2d0be8529127276
63422 F20110330_AACBPF baltunis_b_Page_100.jpg
b491887d2b03dc6ca411e623125c5ec4
5555cc49ba7d3217ec162604dd93dc8b2173fe0e
24232 F20110330_AACBOR baltunis_b_Page_081.QC.jpg
cdb760f439ef71297cc490694905e59d
364fbff6d9e61a2255ab0729603cd6b580fde333
62815 F20110330_AACBPG baltunis_b_Page_101.jpg
dd54460f1241fb1c9f27bc2c8e1696ec
f67e613ac95b0f88d2f39abea1c124ffcb96b7f0
68831 F20110330_AACBOS baltunis_b_Page_082.jpg
a3ba0e5affe565a1456c22e64e720da3
2a4bcba1500b86c83647bb60616e4b53236e5af4
19395 F20110330_AACBPH baltunis_b_Page_101.QC.jpg
638df9046e983476504d873265f1bf7a
a041fe0080ea172a81ed5a2ca4988072d8d32535
22086 F20110330_AACBOT baltunis_b_Page_082.QC.jpg
550dc7ea2b1f65129e259444d8cdf636
fe3c0ee7d179769a81ba0a046687cde7738b50d1
62089 F20110330_AACBPI baltunis_b_Page_103.jpg
0cfb72e88809271a2a576677e08952be
2c421743c298a19f9826a36b62da6596ff09f6fb
111202 F20110330_AACBPJ baltunis_b_Page_107.jpg
d61252e11c922c46dfded0accdfbdd87
b5a489177f3878741629a181a5befc80b5a6dcc7
22334 F20110330_AACBOU baltunis_b_Page_083.QC.jpg
9b25e3b5ead642e01385717891d0dfa8
86d8f4b497d4d2b04d4c7e15fff1d90dda510c1f
96029 F20110330_AACBPK baltunis_b_Page_108.jpg
2de32aef772a0c95bf9aa14fef367788
18569ad62274e04eebd96579034c6bcc56b802ec
26731 F20110330_AACBOV baltunis_b_Page_084.QC.jpg
fa2ee6606bf0e626e01d183a755f0c34
e696a05014694090668195d7d043344a653ef278
26744 F20110330_AACBPL baltunis_b_Page_108.QC.jpg
4c38ccc21e07cbe3f4c83c7768c0c2ab
ee99ed5d0b4e2f4ea344fb071830829ae58c9c52
86083 F20110330_AACBOW baltunis_b_Page_085.jpg
a5b918e49f4545b12aa7cb3fdd3a0015
36465458822113a5ccfc1d2d3e3fe5d5981cbc19
82192 F20110330_AACBQA baltunis_b_Page_027.jp2
9f80a2670eb88ba3584509a9756318d0
23deaa9cabb19fdb31589d0801863b4291996ad6
1051983 F20110330_AACBPM baltunis_b_Page_004.jp2
db74d7ebcc4370669dc13d87079ed7eb
4c4fc38f3cde345961aaaee8d08c3f37f8fa4320
26577 F20110330_AACBOX baltunis_b_Page_085.QC.jpg
e040490e70bde4e9d3054c9a1ca1818a
c5124f804b6745aa6a7e875fed1bf03b25e527dd
116784 F20110330_AACBQB baltunis_b_Page_029.jp2
9e346c29e32fb05c0eb5035877cac43a
1cab1f2415333e7185e7cb52384ca4e3fd18efaf
1051986 F20110330_AACBPN baltunis_b_Page_007.jp2
cb75e609586877dfecb9108793bc3f47
ecda146f6f4d6ecfaacce6ef2acd0b511c229146
63342 F20110330_AACBOY baltunis_b_Page_086.jpg
51686f1bed2b485d1f9fa90de669214b
2cbae5ebf54f875b2c942e2816ae5a74967175fa
1051984 F20110330_AACBPO baltunis_b_Page_008.jp2
b6b44c5485eaabc3dc2635c0ccd4b798
7b7b7d9f32fa28678646ac88d6f4345afcb239c4
74005 F20110330_AACBOZ baltunis_b_Page_089.jpg
55bc23d4b19de8f4c8056d7a85c285c0
c785a8ea1ea606c72f9b6ca656f5933946f48059
102731 F20110330_AACBQC baltunis_b_Page_033.jp2
66b3cc112fd2194eecf573772c158ca9
161e3e63fe682df726efdf95b4884bb2da0595ef
284514 F20110330_AACBPP baltunis_b_Page_009.jp2
1804e76169db9c605a46142721901e47
6ef27d07cd32961632a9150a4a5e47de44f6cbdc
96082 F20110330_AACBQD baltunis_b_Page_034.jp2
2f26f442f666ad26c4418ae55f9f619d
4513a9b46d26236da6c9d9152d30a5a88f3955ac
82796 F20110330_AACBPQ baltunis_b_Page_010.jp2
105fe959326c46ccf50cc6a017b62c62
43658acd9424e2b61be1de870ade928e478aa1fe
96299 F20110330_AACBQE baltunis_b_Page_037.jp2
e0dc3d1d5c530984807e10db05d730ae
6446cbc1838bd5e63e8a0f255fa032ad550ce27f
89707 F20110330_AACBPR baltunis_b_Page_012.jp2
784094b5f4d1d53e198c641238b21748
3fe629ef5961e5de980eebef5733006f8c1dc4fe
10615 F20110330_AACBQF baltunis_b_Page_040.jp2
9f7c0c5f919842e0c6da07a9f1f1dbf2
184add2094b75c8f5a4b28c149486ea0ac3224e9
107715 F20110330_AACBPS baltunis_b_Page_013.jp2
2bf2e4726af5e55992abc3aef44937b3
dc6e36c1407b15e9221655b571974034ca12b408
97622 F20110330_AACBQG baltunis_b_Page_041.jp2
b6f2858625ac65fe003077889d548a52
51ad61c323fff56b88652d8f9ab92553bc8e09e1
111301 F20110330_AACBPT baltunis_b_Page_015.jp2
16dd2e0bd138452119ca8774f50c6d29
346a8f74e7cd0ad7786a63d02a4efd859cd35054
113054 F20110330_AACBQH baltunis_b_Page_042.jp2
9a9211f2d27b8deff8a50c5b13144e9a
cde58ea36a3d84f9f861ff180688d7fcee111111
70372 F20110330_AACBPU baltunis_b_Page_016.jp2
63617cac7a6d210d6a98423f15c80812
1d82a7c3d662539ab7c4851401c314e6b3bdf3e8
105284 F20110330_AACBQI baltunis_b_Page_043.jp2
6de7bcea67d0fe6b42aabfcf39c765e5
fb672da7d48b8b0fbbeae648bb6784c418ec1701
100526 F20110330_AACBQJ baltunis_b_Page_044.jp2
b69aef82d6bf69142b6177ac6a792488
a5ecd73496cde46a6437751ae8600effe703781f
91270 F20110330_AACBPV baltunis_b_Page_017.jp2
1efdc8cddf61874eb19bc4f2bac05406
4c1e854ab9e104a3a5c0c59b147d5d3c311cee0e
97808 F20110330_AACBQK baltunis_b_Page_045.jp2
ee94c706bf41d2dece932aa339e9451e
bf259e58d09bd3081be283537c215c33a018b440
102854 F20110330_AACBPW baltunis_b_Page_019.jp2
699ff3245a7bb17bf65cfda4b836ec1a
7d8142abae9976395804e401657463731557973a
89877 F20110330_AACBQL baltunis_b_Page_047.jp2
d502e8ba1ca93904fe13723aadcc7b5c
85e1922a68e42530614b6ffe037074657bf93b7f
109080 F20110330_AACBPX baltunis_b_Page_021.jp2
920bb68f516bf06ead327671ee9cbed3
4928e51f46ea841e14a29749dbca2b8a304610fa
64467 F20110330_AACBRA baltunis_b_Page_071.jp2
f35d1afd4afc6c236ffbac47e6565297
f5dcc6a7c956c3fea6649d2c95609ce9473a04fc
72273 F20110330_AACBQM baltunis_b_Page_051.jp2
6513d009cad3a985fd4338d0a13bf451
58173ec4324416de71466aee752869abc27d4611
113387 F20110330_AACBPY baltunis_b_Page_022.jp2
f5f6e2504bc8bb3614e68b02d6c4c7c8
5d695f12550c1013232f1b60f97984596451cda5
84071 F20110330_AACBRB baltunis_b_Page_072.jp2
a81d6d2c3394ad5bf6c8beeef90c72ba
219043167543fa6c2fddcb0c7c873243f9ba4a2e
77027 F20110330_AACBQN baltunis_b_Page_052.jp2
60ea6963d458dd681df467e2e19b2c6e
000debd43973c7b58719d4237d262dcb5a50df90
83393 F20110330_AACBPZ baltunis_b_Page_023.jp2
f366b39686d6ab09830279f78d87de7c
dd92e4f7bc12b6b473fee690b68fe73febe88338
97981 F20110330_AACBRC baltunis_b_Page_074.jp2
ea1ec21476ef5642f72fe332729b2bef
37f101d0499ad3770d9c1b32b16fbf8652d82144
89619 F20110330_AACBQO baltunis_b_Page_055.jp2
a24dc77227b35d1bd6a4c4b943788e4a
6da8eb828671ed9c65e0626bdd94c9938cf86fe5
93407 F20110330_AACBRD baltunis_b_Page_075.jp2
829e73453adb7dae0839ae1fe48a3fd5
8825e028e54581edf2f5845e0c01f9811bc89e0d
109535 F20110330_AACBQP baltunis_b_Page_056.jp2
e8d04218fbdd1a9ae9250285c56742bd
23bd9e528bbefe990cb739efaadcb7a7cb7ccd92
93653 F20110330_AACBRE baltunis_b_Page_076.jp2
d492832969eb20464520a9adc320e332
9b945fa24cb407620a620ead4d438209a27ecbbe
100281 F20110330_AACBQQ baltunis_b_Page_057.jp2
950bf8793a67e4c01e3528d674dea5f6
13805273e2c3a1b48e87bae22b32e11eb15d9f33
75742 F20110330_AACBRF baltunis_b_Page_078.jp2
f86d7be2ea5f50e298b619ed079cae0a
78360b8b251bc921361b5897dd6445c221033016
785921 F20110330_AACBQR baltunis_b_Page_059.jp2
89f66bff03d7d20f44fd25f219ed8b1a
58d23c524b2c23e1253ee1aa0955ffd0d431d378
112417 F20110330_AACBRG baltunis_b_Page_084.jp2
e5868612743e5ad8ce8f65a0ca42d0dd
52ddc4f159f1abbbd7fc664b8d9ce9a12f363cc7
104450 F20110330_AACBQS baltunis_b_Page_061.jp2
bc9ab8d5ee41ec4248af49694cccfccc
9ceb760b5b9f5b9daafc40ffd4aad5650531d90d
61913 F20110330_AACBRH baltunis_b_Page_087.jp2
83c7fe779929050dad6432e81173f1ff
6e81bee787e2782585033de841a6c32eda028acd
96482 F20110330_AACBQT baltunis_b_Page_062.jp2
3ec68da99b2d6af15096e42657d4e4b2
6f719efa16b86c0f0c47bd79a82b824a7bce0598
57824 F20110330_AACBRI baltunis_b_Page_088.jp2
1641ccafa3c0688e748a032b5ca21c5b
03c725f2c294e383d53076a4d0ed824b1cac43f7
649568 F20110330_AACBQU baltunis_b_Page_064.jp2
9a62ea2c08208ec99114bcd9c2a75aa2
607059a1158f756a2dc7e8a91d1ee925d024daa7
110779 F20110330_AACBRJ baltunis_b_Page_090.jp2
913edb7a95560d20252359ac6e27fae3
4b1672c394aa48944f5390ecb4ff4fef7c1abdb9
110311 F20110330_AACBQV baltunis_b_Page_065.jp2
0f6cc3d6df86e46ecfcc339172b91bb5
3d52ccd7675352b6c1bbf636f57a578d4122110a
74037 F20110330_AACBRK baltunis_b_Page_095.jp2
ea1e9aa0f0c319f7bdc61af0035ba752
089ec13bd8486eb094176112e0df0f73004bdbec
24006 F20110330_AACBRL baltunis_b_Page_096.jp2
a2605a02555edcb5a50a7993fc8f9624
d4e1e5e9411d716736a75efeab27d524a781b06f
87868 F20110330_AACBQW baltunis_b_Page_066.jp2
461294cacdd9ae8308b5cf72337fb398
55312826d270144d4c4460db057edcf2cd2fe0fd
6776 F20110330_AACBSA baltunis_b_Page_007thm.jpg
c1046c02e82ac0d789ee856f3582efdc
a879eb0dc46fd3f4acefdf01bcb4c60088579c81
81416 F20110330_AACBRM baltunis_b_Page_097.jp2
ee5f06a0e380b91aff95a84f00217e9f
ec51649f9be179f62dca28cf9cdf5b68d8e493c3
99417 F20110330_AACBQX baltunis_b_Page_068.jp2
f764cac3b0cc00ffe7ede8ae5b3e1d63
a626fa4912eb2f9b5656483bb109f96aa1b0eeb0
5922 F20110330_AACBSB baltunis_b_Page_008thm.jpg
d9803533a3343954ff3e772d5f4d1c04
0a87a4bede477f9aa7853d7089489dc32cd47522
61978 F20110330_AACBRN baltunis_b_Page_098.jp2
1493d50b5f055aea2c130fa35a091fa5
9343e32abefcdfff5846071690c37447efb690bf
97870 F20110330_AACBQY baltunis_b_Page_069.jp2
6c5fc4803ffc3c3f64317655cd37356d
46ff435156daebba8159fdae7f3c876c31a2c258
5230 F20110330_AACBSC baltunis_b_Page_012thm.jpg
eae8c9fa3d4ff08d8e36d7032cf5b13c
64910ff1f1accebecf3eb320441ed90daa8972bf
73535 F20110330_AACBRO baltunis_b_Page_101.jp2
367aa93420bc76b9e5cb0857c493896f
fc2b650bb48629cd11788b44e5571affea228630
60582 F20110330_AACBQZ baltunis_b_Page_070.jp2
9cfda76bf7ce111dabec575642d196f3
ea69649e007fac8fe82ac37d7d588268756fda34
6028 F20110330_AACBSD baltunis_b_Page_013thm.jpg
2075221a42345d42eea76259ca5a5c4a
6f87564b6e29b6cc644642c947a61ddcccbfc989
71594 F20110330_AACBRP baltunis_b_Page_102.jp2
e83344f1964a45d216f03aa850e76b54
fd073b5d1c04731144a8b02285d1904d9589c189
5372 F20110330_AACBSE baltunis_b_Page_017thm.jpg
991e3561e3ad2f71035fb13c00492657
192d8a729d299f6ba7c36be01c11649748f2296a
40234 F20110330_AACBRQ baltunis_b_Page_104.jp2
34daa79493aa1283e06d8b01c7c36e88
f9f889222df83542dd91342df40168fd90326200
6022 F20110330_AACBSF baltunis_b_Page_018thm.jpg
1842ef8dc80dcb9e0d87fe95cf43416b
d718499a23be0c17a792574a5808c9f111023c00
128113 F20110330_AACBRR baltunis_b_Page_106.jp2
5d8a90e9b513b55921591671e6b7bc89
fb2fe4ff8780f57b6fe76bc4231cb701b579fc98
6321 F20110330_AACBSG baltunis_b_Page_022thm.jpg
0227e4d6e15d0add04fc71740a3aa297
102e2611aeb64368ab1225babd972f5f585ef90d
123071 F20110330_AACBRS baltunis_b_Page_108.jp2
c57cecfa8be72344f851ad603fcfebc4
bdc4046c5e1cfef66e98ff25a183a2bce3b98670
5051 F20110330_AACBSH baltunis_b_Page_025thm.jpg
33f37de64a751d43bd8334103bc67777
918e3548303d82391b1879237c7297b974355866
122240 F20110330_AACBRT baltunis_b_Page_109.jp2
efb03a5e2f5aa74abb7cdc9f1ad55c33
18a138306ab6126a3907914577190a62103bb78b
6107 F20110330_AACBSI baltunis_b_Page_033thm.jpg
e2bcbfc9431579ed26aebedaf98f67a1
61e49a9a205358f6f45edabbbf53a4d28d1e816b
27638 F20110330_AACBRU baltunis_b_Page_111.jp2
d3c7eb97620f87b269e60454fea6f522
13c6112bbcafab43bd5b9b353019d876f43cf547
4692 F20110330_AACBSJ baltunis_b_Page_036thm.jpg
a7b253172d2d13200c9da35b1e87e898
09105a6d8476f580dadd12f5c5563a52848d7211
1855 F20110330_AACBRV baltunis_b_Page_001thm.jpg
f8e57c6892015c99283a63f946d83899
161aebb1e3138019da243af5aca1b0c3ca24c48e
5915 F20110330_AACBSK baltunis_b_Page_037thm.jpg
ae49f9080b4f2d2e99deea3393847cb5
1ea91fa8481dc7f0b17934721cc58f057d8ef12b
3756 F20110330_AACBRW baltunis_b_Page_003thm.jpg
986651230f78b91669b339a5db8b8c49
0cb2660772d0f7f0569be7f069a4e4651fadfcf3
5731 F20110330_AACBSL baltunis_b_Page_038thm.jpg
d10aacc1dd32e62a4236b1df77304aa9
6417fe62da59dabd14ae9bcf4d5e833c8649e586
5047 F20110330_AACBSM baltunis_b_Page_039thm.jpg
9d566a43b89953fee73673a990400e98
fca8c3b7218d563ad9f18ff9d605917d6366d813
4049 F20110330_AACBRX baltunis_b_Page_004thm.jpg
0fedca775beabfc6bd5f1ba85b16e434
ed12f094d18213cced1945a3ff858a0960197a31
6313 F20110330_AACBTA baltunis_b_Page_067thm.jpg
fc43d2f6011e9fd01214c878c6ae725d
1720f5f260437a2b96bf59e4f34e5761c751307e
793 F20110330_AACBSN baltunis_b_Page_040thm.jpg
370e318f6abee8d1c7cf8d811fa760fa
c432a0bf0d352a566b3c38d329d01e141e4d920c
5476 F20110330_AACBRY baltunis_b_Page_005thm.jpg
8c5cafd59d42c6e65ae9ac5131f45407
72603f2ff68eb2c973eb6efc59dd5ce8c5b02aca
5843 F20110330_AACBTB baltunis_b_Page_069thm.jpg
bb0b095dd6e7252283b8d66a7bc1d9a1
279d7ec2384949436cf1d484ef6e461659649fbd
5672 F20110330_AACBSO baltunis_b_Page_044thm.jpg
18da2c0fe9010bbf783e6dae8f9fa0d3
83ce5f8672d2f730fb9c28cd1fa95be6bd3562f7
5857 F20110330_AACBRZ baltunis_b_Page_006thm.jpg
543200971401963c7f70a27c119aa6c6
8bb0602437204d784c9d3cec7c47b2fa0e1e7d93
1285 F20110330_AACAQA baltunis_b_Page_104.txt
9add2c5a0e3e502db0366b173ef18748
b1288a435a6eda62cb1aa65b66d40f48450c1706
4323 F20110330_AACBTC baltunis_b_Page_070thm.jpg
933dbcf3629dd234fefefc7b85d0a843
eaf9873f62eda197dc74867b9c0dfb91fff260b7
6023 F20110330_AACBSP baltunis_b_Page_045thm.jpg
e8e7bed976fcc8fba8b696dccef40161
5b65f92e308dfd0d9518163ee33916ee0726b71f
886 F20110330_AACAQB baltunis_b_Page_094.txt
4394b9f76003f77d192fd14f5d448cac
b6e1fa5db9da30ab7c81f74bf03b4fad215ab04d
5071 F20110330_AACBTD baltunis_b_Page_072thm.jpg
743c468d50be4f671e82061c8dd974bb
419586c80721787a0aae787d08f43cf146165b07
4638 F20110330_AACBSQ baltunis_b_Page_048thm.jpg
26addaee40e432db2e5ab4f3f8965f0e
65a935f575afca93ef4f569317b9b139253aa91f
23355 F20110330_AACAQC baltunis_b_Page_033.QC.jpg
fd6c649c488fde8e112f373b1ae259bd
9aca7d1d827422b1f23f82442a9b1f141cbe6618
5105 F20110330_AACBTE baltunis_b_Page_073thm.jpg
c940f473a8ce6c386c9999927aff505f
ceefdeb5ec381a9b2506e93e951c740d6c357704
4586 F20110330_AACBSR baltunis_b_Page_051thm.jpg
0a3a04b377a88598127207a0a64af0df
68cbbfbc90e2a6e7b62b315fa9b8cffd5bce5d1f
18219 F20110330_AACAQD baltunis_b_Page_073.QC.jpg
b781975b3addad8567c8c7d586211ab8
ca5936f331deec137ecefd6ef2cbda8e3225f62f
5654 F20110330_AACBTF baltunis_b_Page_076thm.jpg
a55d4ed685203693891087358c5b02ac
92c329fd6a2d21427cff5f98cd90aadbcb2f3af4
4678 F20110330_AACBSS baltunis_b_Page_052thm.jpg
eb84d203564613ca9b3159b3f659f4be
69b177c552158fcffa28915cd90b2c72be12a225
10425 F20110330_AACAQE baltunis_b_Page_112.QC.jpg
df881ecd93ea85341c7eec933bdfe70b
c735be6adf904514a3eceab269e19e3032542d75
6223 F20110330_AACBTG baltunis_b_Page_077thm.jpg
cc84d6fb80c877371ae4d41e53dbcce1
1e5fe4aad7a9dd999744dff5a0ba4758eb5a088d
1051970 F20110330_AACAPQ baltunis_b_Page_005.jp2
504306e091dafdbb50de85d987e0956e
a004756f9b32546a8bc68efa31521ef33773d6cb
5082 F20110330_AACBST baltunis_b_Page_053thm.jpg
ca1d592c767f4671f16bfd29e9322c9c
952a0f46b3297c4b32d215845b430c8687274d6d
87788 F20110330_AACAQF baltunis_b_Page_105.jpg
c2bda14e00b823e21c8ea2f607589e16
24fefcd52a844e6518bf347ac031d93631dbfcc6
6298 F20110330_AACBTH baltunis_b_Page_079thm.jpg
db51a0f652a4bed458a91619ce57dd25
de621ca83f3cd1e02e4ab128de2b312d3eb49e3b
78305 F20110330_AACAPR baltunis_b_Page_081.jpg
259250a66f8eaa28ab872a5344d6459d
c74343623bd0b02954a2274ec000ef1f1d05b58c
3769 F20110330_AACBSU baltunis_b_Page_059thm.jpg
bad72a25c80cd179e43198be3fc26c8a
9f6f7ddb2f8acc6721f89aac33b79a79b15f2113
83044 F20110330_AACAQG baltunis_b_Page_058.jpg
1533b7e65abbc6b4417e50359006e12a
707a018d30e74ee0a65b73f1ff14e5ad9ecb5dd8
6229 F20110330_AACBTI baltunis_b_Page_080thm.jpg
dd41708ec5d7a02e000ba8bb72011a0c
5983b4617dfdd0b53d045ddd30527f7e4ab4995e
50079 F20110330_AACAPS baltunis_b_Page_020.pro
e75de0de2e4b66622bc87a051e53b0fa
aa592c19c06df84cfe7882c5628352bcf360f1df
6129 F20110330_AACBSV baltunis_b_Page_060thm.jpg
ebf5620505b103690d0b3b37903dac61
cac90c25f8174b6106f1c947f92bbcccfbc3f745
40256 F20110330_AACAQH baltunis_b_Page_012.pro
7b9fa2a95a49b1c51adb12c5c9a72f88
10c813fac6b4a8653ef9844c7595ac578c441739
5744 F20110330_AACBTJ baltunis_b_Page_082thm.jpg
b4f34f8e7c79aa9037964827b1078eb7
c94ec8d8a4d0f5c94695e185d7bb98e9581aa7db
80731 F20110330_AACAPT baltunis_b_Page_025.jp2
5665ff1f36234e7f3d67934dc333a219
f9d0fb2c97327c5c070d2786616f9c7d17a4577c
5991 F20110330_AACBSW baltunis_b_Page_062thm.jpg
ad55df28377e482a27cf615b192b7664
fdaa6dc349eb8ab41f0233c386410dad4b1d868f
F20110330_AACAQI baltunis_b_Page_095.tif
d969a79f22d4f949efaafe483a13a226
759daa459447ce1745618d1e24763c8c4ceab287
6509 F20110330_AACBTK baltunis_b_Page_084thm.jpg
e8095cac893256de14a5c2bba4fa588f
abd239df06216ba65ec16d089352e2af5755a095
F20110330_AACAPU baltunis_b_Page_049.tif
d152b22871989b589389a7ab38fce547
104b9310a9461ef439ab6ebad71b39d6833f1b26
6096 F20110330_AACBSX baltunis_b_Page_063thm.jpg
f43e0265f4d47a1864b51f95ad18bd74
e01509eae4bb2f38cdaa4e720703f2f391c868a6
F20110330_AACAQJ baltunis_b_Page_071.txt
b4de86a975c6ec690e19190175a0bec1
811cb28da413b6e500ef63d867c9c837492c76a6
6237 F20110330_AACBTL baltunis_b_Page_090thm.jpg
946f20e1041e673e867b5cdc19b5aca3
8f5a8b360970ed5f3315fb08f39e0138347020ec
6966 F20110330_AACAPV baltunis_b_Page_107thm.jpg
daa1ddb2710fd1330c3a118fe53c284c
eb2a538b9115d55845c21a58b40d13bc4f42b5b9
21015 F20110330_AACAQK baltunis_b_Page_093.jp2
82277e393b8edcadec3ffb461286a067
d22827fca1ce0756736a141b515b1a4974edee0a
2392 F20110330_AACBTM baltunis_b_Page_091thm.jpg
4f7206f9728222de7e3a8436e3cbed98
b406e30a84b9ec6056241673a2c20a10ea25feac
6194 F20110330_AACBSY baltunis_b_Page_065thm.jpg
8e1c2a07441a5b2697d92d97c1e7ba21
c52f0d44aefb7fa240e01d544a0d561fb3cb6a37
46205 F20110330_AACAQL baltunis_b_Page_059.jpg
1d386ce32ed81c1d271dc4baf360ba91
bd026c1c9bf6ff50e4ce3e2a9fa0b845bab328ec
953 F20110330_AACBTN baltunis_b_Page_092thm.jpg
8cffc31e4c81041d74499e4c6a93c41b
541184b810377f8d8775430f75d1d3416edeedc3
2001 F20110330_AACAPW baltunis_b_Page_056.txt
459a098b9b16b2be466357d63b5d2642
9426a33846830cd021201b580ba292185fe35d24
5141 F20110330_AACBSZ baltunis_b_Page_066thm.jpg
ce621aa5f7160ede3cec5c141b5c8a08
1d59b9755f615e0ce7fbc23ad49c08740bd1a32c
F20110330_AACARA baltunis_b_Page_039.tif
a102256dea96eb86d0958527cd316341
97f509ce39fce108248cbc62de2c0e0a94dfef0e
6574 F20110330_AACAQM baltunis_b_Page_109thm.jpg
836ac17649703f60c26b9610b8a3885f
32c2a5b861ebc30fb50e9f1122c6180df28ced4f
1392 F20110330_AACBTO baltunis_b_Page_093thm.jpg
0274c134686fcba0bc1367747d632093
9cff56efe489d77a4fdda0b66e5d2a482a194e1f
F20110330_AACAPX baltunis_b_Page_078.txt
cda4c1d17b69b53a532b8caa643d6270
6c65b2c3d4cf7cd02c711c25c626f3976a1b9ca7
1201 F20110330_AACARB baltunis_b_Page_009thm.jpg
35eead7f27bdc15bb85b9bc729e0d98f
40faf0968a373ce8150e0e991667cb3b2618d298
4407 F20110330_AACAQN baltunis_b_Page_078thm.jpg
393b31f9619df6b13b319b5563d43a91
ea14020c6b4c4c65f28947cceb56305d98beeb9d
2520 F20110330_AACBTP baltunis_b_Page_094thm.jpg
50cff7dc142367f63fb32c201126b648
95e72824b1bdd0aae9f342a3c934946d074737cf
2755 F20110330_AACAPY baltunis_b_Page_040.QC.jpg
d3279eb90df5958cf5e0bad905771d5a
10b0ec7722f64098e362fee59df585a5d5ca1f23
2924 F20110330_AACARC baltunis_b_Page_007.txt
ac4a94a2571b7d90c95d91d2a9f9de07
f35c93f09c2d934f2764acbc1aa7ffc5f115d56a
30882 F20110330_AACAQO baltunis_b_Page_091.jpg
7e3d5fd5a4087b6dc36d36b1ff9d455d
2ea59f4bc0ec43e3dc5d7e2fc020dd5b8b4b86a1
3772 F20110330_AACBTQ baltunis_b_Page_098thm.jpg
934637e25b956d90c90c00a4806b916e
4caf1e264f1251dac90e7236ca5ce13651d266b9
F20110330_AACAPZ baltunis_b_Page_032.tif
8ce34d9062c6472dad3a138f6635a797
ec9d361884d78634cdf77fd4ac7b8280e838fc81
5818 F20110330_AACARD baltunis_b_Page_086thm.jpg
498d8d34a5b347652beb205b883fdf07
3f0d9ddaf411453401fb4472dac97f7cde6a3785
6422 F20110330_AACAQP baltunis_b_Page_046thm.jpg
551573ebe9bfba780a52c40a270a62f6
f533484d2f06ea576a9f348f33695daa7266211e
4649 F20110330_AACBTR baltunis_b_Page_099thm.jpg
fc0ccfc4751ec629375d254e7da4714f
f07bf16cd270458c115784293c6e741de03de23f
5465 F20110330_AACARE baltunis_b_Page_075thm.jpg
19420245bc4389d7748f6040f2aea27e
cb6429c5e9f41f08a3d09330bad79f23c59d1ae3
F20110330_AACAQQ baltunis_b_Page_037.tif
3ec20933af6fc24b1fdb1315a361eea3
df1a81f8d357cdff9641ade1712e0491c40109ee
4043 F20110330_AACBTS baltunis_b_Page_101thm.jpg
f4a4e9541b2c00ae416664aebb9e87ee
337fc4a9c9ceccc1fdc681458f96dff08ce29013
16314 F20110330_AACARF baltunis_b_Page_048.QC.jpg
043541fbb27056df1cc84ad042225a13
9f124a2c4896529e6396cea2fb51f5c3bfb9cc26
F20110330_AACAQR baltunis_b_Page_102.tif
3858619f250b54d63b774eb2e91de360
0460aaceb0f12b26ca1747677b946495ee8abcec
4074 F20110330_AACBTT baltunis_b_Page_102thm.jpg
636c90742d5aa28cc5aaef6dd5f65c43
4827ed49947da3990d29fc65bc1967628420d632
885882 F20110330_AACARG baltunis_b_Page_030.jp2
3eb8ccc834e71578e69ea5ad441a9980
6c4ba694f00c7dbdbe9e1ae3fbb8526ef8807138
4983 F20110330_AACAQS baltunis_b_Page_027thm.jpg
afd5c844d192b166e694767a7c53bbc4
4fc5d22c48ad2f4bfe4a15e5e551ecb8776ffd96
F20110330_AACBTU baltunis_b_Page_105thm.jpg
8e947be19562e7468bf1b4d3374d2b93
16eac17ec9fddd0bce052d8c46d4a1c0be60059c
20100 F20110330_AACARH baltunis_b_Page_023.QC.jpg
208696865a2210c9b35e8d7625e91ba3
e653823a2c72945f2c893ead89f50dfcd2aec65a
107155 F20110330_AACAQT baltunis_b_Page_063.jp2
0c9a1a722a575ad451598499659811b7
7d69f588b06aad5135383d954ba326f48742dde0
6563 F20110330_AACBTV baltunis_b_Page_108thm.jpg
b6608154364b8b49f92a34b25267370c
2235db407e67b826b17560f26349df9fc644153d
6691 F20110330_AACBTW baltunis_b_Page_110thm.jpg
5885dd8bdeb10b97fa95b5b45c5a2de2
9fab42f1c2682fc26027c2b0be995d8e5949ed3c
21771 F20110330_AACARI baltunis_b_Page_086.QC.jpg
dfeb5ef53be27b883e08208b5600e3bb
3ae7a16630d059bf72f85802f412266d71da234d
49469 F20110330_AACAQU baltunis_b_Page_031.pro
ea06c3f7bbc1a10ac3b9e1648ad40646
301c17eecd0ceef7bc80e39b4cd04016e7370fcd
2718 F20110330_AACBTX baltunis_b_Page_112thm.jpg
e2c10f101a4e898d965673514f162788
5b53f47d5fc820ee9ad1fb107de4c80e2f2be369
1423 F20110330_AACARJ baltunis_b_Page_103.txt
da0962177111e7254fd596e3630e6131
648494a5988e3cce4c57eed3b7f5d16a023474da
39394 F20110330_AACAQV baltunis_b_Page_024.pro
1135bf933ff1f6460c60efa6492186a9
77ce0286052549283783b8c9588a92d63a6ef666
133039 F20110330_AACBTY UFE0012944_00001.mets
b0bc214e66d53b296c2f10f7cc4a2525
2f60f15bae99952710f280ec554385bbb64d8f4d
F20110330_AACARK baltunis_b_Page_046.tif
73de1e86b0cee0788aa65fe657272d7f
c2e73a3d2992f6b75b0e66a3b49d90d7bc880f7a
4914 F20110330_AACAQW baltunis_b_Page_064thm.jpg
e68c465b4023b29b1b03547624cc83ce
22b665ee09a134f0b72c9102a8a39f779ebb4ac1
82248 F20110330_AACARL baltunis_b_Page_060.jpg
887b48eef4a73b7adcd6ec6e238dac33
170ef76c7c9086158458e572a75a9bc2e78bf31a
2272 F20110330_AACARM baltunis_b_Page_081.txt
35396cfabf61ab180ff5bea6693187e1
8d85f64c950d40f852312f648ab1f2746a67dced
6085 F20110330_AACAQX baltunis_b_Page_043thm.jpg
784d8fe2d19f111b0e278b343194bbea
aed1beaef720ecba454a3d7eee6a4d8cebf526af
23614 F20110330_AACASA baltunis_b_Page_044.QC.jpg
e7d296b46a0f769fcdb5d3eaaa4d0663
8a335cd7dff95d3a21d266c2c6d3b251b2d12458
F20110330_AACARN baltunis_b_Page_070.tif
2ecc906fa12e46d6df570830ea0e0b60
927870a7d3340a9f1e0c59aef82b475a0025c6a4
14882 F20110330_AACAQY baltunis_b_Page_070.QC.jpg
dddf1239fb34b986e56c14788112aba0
325df8d3b1bf7d6ca11958fd8aeb51fb22204d0c
3211 F20110330_AACASB baltunis_b_Page_004.txt
fc08eb3864fd3e8895a7e6f8f292a168
e2412d10cf6976c392759808c330088ffd3a1465
29560 F20110330_AACARO baltunis_b_Page_088.pro
3979b54921b172ffc402165b478f1ee7
33f213b4c0726233bc86b8aa9c91313c8dd8b5b5
23310 F20110330_AACAQZ baltunis_b_Page_045.QC.jpg
40c5afea313b3da9e1c979bc382a0c4e
1478c5aae046f904b661b19fa1362bac88486899
F20110330_AACASC baltunis_b_Page_023.tif
03840e4cacd030e55750d484281627aa
257b842359e8604436152bd01afadf460c3fa50e
85522 F20110330_AACARP baltunis_b_Page_022.jpg
a8136331362ac22f35beddb4509030c1
26a898c2ad4c2c925e6a993b200c2209d959aa7e
110379 F20110330_AACASD baltunis_b_Page_014.jp2
a017a48b91d123e84363b8348117a5f2
4102ae80c7fb63cb3c40e44b5283eaa4d6a79115
F20110330_AACARQ baltunis_b_Page_027.tif
7b9c3eee120354d81632efd78c368c4a
0c526d32486171d7244038a0607000a27ba36716
F20110330_AACASE baltunis_b_Page_089.tif
0131e1e96d042af2f90afca9012e61b1
79a79db69bd58d33ace66721b8a988778c8f5e53
79423 F20110330_AACARR baltunis_b_Page_057.jpg
2e24de7818516c76e634e6a8e5b0698f
a1e8e9bff5da9c89e46573f9b3213ad0e421c688
87723 F20110330_AACASF baltunis_b_Page_046.jpg
8bf2cc387791d25313df2baf071c6de2
69432d3095e4476e7f3fd68b9111f5eaf4a3dc24
1427 F20110330_AACARS baltunis_b_Page_102.txt
826e63499ad44adf25d8fb49bf92e1d6
c2aa91eb4ea4df930635684e13fb7ec94abff4bb
50453 F20110330_AACASG baltunis_b_Page_050.jpg
3f167ce9c235a24202c25e1b6889f0d1
7a240a4744a4f65cd27349d0dfe84c3f61d34b88
F20110330_AACART baltunis_b_Page_076.tif
8e334ef3de70af36c95bd62401b63416
0b63d2d0a22f4da5d6e5985fe34fd7fe12d4a675
F20110330_AACASH baltunis_b_Page_078.tif
c2225bfa0f505f24a14e03711b3446b8
5002b98e326170127333fb8979e685bb001e5fbf
48164 F20110330_AACARU baltunis_b_Page_080.pro
26788e8044272ef130f43a8fe4f39496
70e0cf1a2abc55daa08d6178c3b00198c304236f
95469 F20110330_AACASI baltunis_b_Page_028.jp2
a153fc0b3b4a49e30d6835ba073ab035
7fccbeefbc2937af7769f2ae8fbe81f67d088749
106492 F20110330_AACARV baltunis_b_Page_018.jp2
da5c430af270307c70d91567fd520ad9
895cec191da678997623c65184f9d5c799cf30a1
36922 F20110330_AACASJ baltunis_b_Page_059.pro
0203b743bb4be209e7c0aa56dec4338a
50777ceaa1ee489fa8709974f0f8f40b63cadeab
1677 F20110330_AACARW baltunis_b_Page_075.txt
b62af9da388c628a889f46eedc6aba21
7e80b7d5635b4f132a60046f1a52a84058bce425
6181 F20110330_AACASK baltunis_b_Page_056thm.jpg
03d83de56fef678f3075bb60c76aaa0a
4ef9bb7063ece68ccf2f5b8cb455dc95ddf7bb7c
10156 F20110330_AACARX baltunis_b_Page_091.QC.jpg
5290717b3a9b0f541272c74cadf701fb
e469174d7757639bc78b750f368b8f4e2e05521f
19503 F20110330_AACASL baltunis_b_Page_011.QC.jpg
ebbe1f9f25783f6b02439b0e66efa310
3c2bdcace1b03e5055cd4f4f7f55c9d9c8538eef
197 F20110330_AACATA baltunis_b_Page_092.txt
07a31b56d58aa4d7140720de6fa641e7
47f18d6f26cbee308c6cc632336a186ee5b6ae18
37848 F20110330_AACASM baltunis_b_Page_051.pro
b2a130c1c3554c77f378fcc7061cd024
e4e501045013c258a2a32fc2d7f47ea278974c50
22275 F20110330_AACARY baltunis_b_Page_047.QC.jpg
22470026b7b91938dd5988a2861d424c
501ebcc8656067019f5e24734cfb6151fc34c147
37043 F20110330_AACATB baltunis_b_Page_095.pro
2689526de9050fd39eb58b942998259f
fb27ae15a78c0e135437adbb50f43541f00faea6
26446 F20110330_AACASN baltunis_b_Page_015.QC.jpg
9b56f3a33f6119bff772149f6714906a
e57a7e1fe1a5f24f5ec71f6d4470bb29dcf70ac5
4559 F20110330_AACARZ baltunis_b_Page_071thm.jpg
63d1940e31a5a71fcdf07fb32d7f1cae
5e3084248e41566ebf2f95d5844bf18cf55c245c
111123 F20110330_AACATC baltunis_b_Page_067.jp2
3cf69fda6897e6638a1b16576e52d1d4
d4866d600e049851fa23027ecdb8fb99f0a868da
45936 F20110330_AACASO baltunis_b_Page_045.pro
e39bddbf47cccf186cb617307b98f280
6995ad887dff53d7cdc8e1251a4f58a7eb9a1c72
52865 F20110330_AACATD baltunis_b_Page_088.jpg
25c811ba2dc94b716bf8894d8eb78e1e
43c2634e4f23a8e1da853a1bc9763506776e1bcd
24593 F20110330_AACASP baltunis_b_Page_057.QC.jpg
a6a33ea9b542c159f8c18d52a746bc71
c85198e697b822e74b49519d1bf4a39f9cee8c11
1964 F20110330_AACATE baltunis_b_Page_051.txt
8d9db78314c039dfcc8c38ebdef04199
8c2e2e465b38e13fe1755ab16e3b2b078fafb568
2009 F20110330_AACASQ baltunis_b_Page_021.txt
64a0ed13b8bb35cd68c95f2dd90b5437
d498f7890bd004fb4e634ff25d51f29e024c14bb
80139 F20110330_AACATF baltunis_b_Page_031.jpg
06fc7532f3963246f499e0a9ca2b7284
e9878255b73d777a25829695a6531af889498735
F20110330_AACASR baltunis_b_Page_006.tif
f5e3c98994eba9be6af80085dfe5902c
c5bf8a845fcf2b69acb422608b6200261d3b1ef4
6067 F20110330_AACATG baltunis_b_Page_111.QC.jpg
23ee43cd9f44a29f5f2371c7a4fe018c
e502e707b4cdbb075439a95ec3b8fade764d96f7
5786 F20110330_AACASS baltunis_b_Page_041thm.jpg
29c372e61f06479110bd5c4e419ce1ac
a7422b52555c286bf0881b9b724fe7286b20396b
26922 F20110330_AACATH baltunis_b_Page_109.QC.jpg
923a902f649c8d54a799e34410343e1a
dcbdedbe2b2d52c96e875e97cf59dfb87df2eeb5
299 F20110330_AACAST baltunis_b_Page_009.txt
a9536eaff2cda21bf3e822c107ffdba7
cca9c989915d61d5507282b4aad5a1d384b7704c
1951 F20110330_AACATI baltunis_b_Page_047.txt
5bf7065e1dd62e72c53515f61e45c707
37ff3a6dcb2e92252f68a5d09b6d140c70da273c
F20110330_AACASU baltunis_b_Page_045.tif
c9e02bbc4f4ec7f08c6829d59de935b1
8a9c82a9185628215dd4689e23282ec50a2a52d7
123 F20110330_AACATJ baltunis_b_Page_002.txt
ffd3492ce9bcb1cef3dddbe53671d08e
389137424be4a499ac59f45eb3c194c5d2edea8e
95618 F20110330_AACASV baltunis_b_Page_109.jpg
712043a31ac7a8caf5f27ea9b373ec4b
0961ca05dac53d9e802ff85cd1d18e9749f828c7
81482 F20110330_AACATK baltunis_b_Page_011.jp2
b545d10ebef9664f5856889338f5e4f9
eaa59f18f3762092d49ac7f6b65cbf617ec8e861
9984 F20110330_AACASW baltunis_b_Page_092.jpg
963b9b70f2085e345864d92a95980a39
f50b85f45055db79581b9495078908c878ce5b28
115904 F20110330_AACATL baltunis_b_Page_105.jp2
1c557423cca43393d646b065b13036a8
798b7a6da2f9dfc6e5c33741ecb11fce8b051991
F20110330_AACASX baltunis_b_Page_031.tif
c8eb908291e61df03cf3a284e0aef24f
992908e68be52984e01b547b5a468541a881b86c
68363 F20110330_AACAUA baltunis_b_Page_066.jpg
e4d47ec4aad229631ea379cb559b7a9d
11006372c4d18e9703c07f2442b09c031cb22786
68348 F20110330_AACATM baltunis_b_Page_083.jpg
f35732fb9ad45b55b512fc12fc928872
65f5739eabb56b94c4a8940ae434a328e3d44903
F20110330_AACASY baltunis_b_Page_087.tif
fb3052cae1c8036e5d613239aad572b6
44bbee7181b872bf40c42f1048064515ae010319
2083 F20110330_AACAUB baltunis_b_Page_034.txt
22d3af144954a51aa84e676d82ff01a9
ea24d1f3685fcbd63f7e35dd0f5f5bcb8b160b86
F20110330_AACATN baltunis_b_Page_109.tif
a8421c5e9fecc1cfff90542a72e2b03c
6a684a5ef0e739d0d94d894926b46a3c92014b0c
40627 F20110330_AACAUC baltunis_b_Page_094.jp2
26e5c94ac0fa69d2166cb73103905fd3
df93097fbeb60997d4a54e4b8350d1eee7f98020
112076 F20110330_AACATO baltunis_b_Page_085.jp2
d4438a4e822a4b46af3824f455a85afb
2363bc1e4fa5fb1453a533bbeb945b07b25c250c
179 F20110330_AACASZ baltunis_b_Page_040.txt
d63ceb945b5252c93458f3a1284e4838
c7209cae81b26d0aeaa1ff939e02886669a083b8
551 F20110330_AACAUD baltunis_b_Page_002thm.jpg
9a612c2855e6cc8fe9249ee5c5a2c2e5
55fcbf6c87d1ba0393368195a10f2a260185ebab
22656 F20110330_AACATP baltunis_b_Page_097.QC.jpg
073dbee1ba238c0acb4c2ff2d2cb9a38
19750dd6f32aa8c3de507d691a2e06d2ea215cc4
2778 F20110330_AACAUE baltunis_b_Page_095thm.jpg
a915537149245c0b721527f41061a838
09a4ac8d68a529a911531690bcb1ac9b23b65c1d
49777 F20110330_AACATQ baltunis_b_Page_060.pro
3e6843983213b9bc5b61c1bfa3d8574c
93efd2d255fdd616a48100675653a34d33070edb
1858 F20110330_AACAUF baltunis_b_Page_053.txt
b12391dc079ad2173a5d706e26f874ff
5c695e51b7f00433fca7f8a273ccea1b9109e951
63002 F20110330_AACATR baltunis_b_Page_027.jpg
72a934eced68dc751252fa80fcd33d6e
de7758a0b41c3c6433a982925270e9e17bfea669
2560 F20110330_AACBAA baltunis_b_Page_110.txt
f9510e690377b7d6f8fec34be090ff19
879455d8afaf4a9af0d54ecf4a32064f073d40dc
4717 F20110330_AACAUG baltunis_b_Page_010thm.jpg
c28a8809c8b993d28f2cf2481eed5309
0dae1a9fe784d0049356326a6ba2cb9f39601c4f
58194 F20110330_AACATS baltunis_b_Page_073.jpg
be23aff2959d0657fd18406b8de5eed6
e76952d6687172bf902d5ed345c2670032b00616
71583 F20110330_AACBAB baltunis_b_Page_075.jpg
689b0c469d88145b1f3a3a6a25a9fed6
6d3f2bce1b25f12849749ec620a5b08bd7ffac96
F20110330_AACAUH baltunis_b_Page_085.txt
3b2c4d4a8a66a0be6568b3c553bad7ef
c184837c7841f64d1c0d7241e6ac83cf752aeb96
2107 F20110330_AACATT baltunis_b_Page_029.txt
99310c8951579c8fcf34f90c9b8b8034
74155b5364e6d1a7fc2745fd8583c519d75154ea
40680 F20110330_AACBAC baltunis_b_Page_017.pro
c423fa34b8317c70fd3a42d864d5a6a4
45844ac21579336d26e1ad715c7fba5bbbfd1ae5
2025 F20110330_AACAUI baltunis_b_Page_015.txt
c3df75448d3df9642af760c6fc160f20
346b69df1aa4352e2536789d71affa0144fa8041
85268 F20110330_AACATU baltunis_b_Page_039.jp2
e5682a28ef56fd9fead0a81e98da77c0
18270f4c0f6c7b4a0c2759fbe4c7d3acc851811c
5794 F20110330_AACBAD baltunis_b_Page_074thm.jpg
cd806a005bdfa69c0342375f4a3f07e4
5fdade6628c017264efd75542b158aad78f08a86
51591 F20110330_AACAUJ baltunis_b_Page_079.pro
f26eac0f6f33a65889c003c41215e0b4
193896fd5bd60712c9ff12a35b353a9e15f556d3
1863 F20110330_AACATV baltunis_b_Page_095.txt
2d71f1a82c906465b58ca58a661676f2
5041976073b5c23c58b988e349f68eb58a85a3d4
1560 F20110330_AACBAE baltunis_b_Page_111thm.jpg
edd5401354dd7c8fdcf28e0fe4fe3fbe
461e052c503ae54231f5626107d886e14a38cfc2
F20110330_AACAUK baltunis_b_Page_033.tif
b28cbe82975cfffd281a82bb3267f7e7
7efae045ca597d63b89784d2e1f096df1ac24e7b
F20110330_AACATW baltunis_b_Page_011.tif
fbe87e702f73fdd9243719d3fd488dd8
c78d8a5598ee70130c5ac97fd51fe49df82e5b05
F20110330_AACBAF baltunis_b_Page_057.tif
089ce454f424fd2380d33c0d9fcea8fb
1a4742315781afef09a9bc4d9d07086c11bae838
49411 F20110330_AACAUL baltunis_b_Page_018.pro
f0c3c2bb56c377ea2462332ab93b73bf
92e446f192dabb377376676f5c93c60ba11940fb
1954 F20110330_AACATX baltunis_b_Page_031.txt
5926fd2c2110a4ae1d7f75b6abd328e5
03191f6b44d4180f5a64872ef54a8638284a0ec2
F20110330_AACAUM baltunis_b_Page_002.tif
d736224bad4d69426e9d9fe184cce3a3
f5bde3f49d29c9558308e1b1c92f6f3868fc3447
2791 F20110330_AACATY baltunis_b_Page_093.txt
3f19c280fe5b7f4b8a39b1585255d3f4
b88b06bb2379d4e780c105eb7857075421b32887
531 F20110330_AACBAG baltunis_b_Page_096.txt
e0bdd5268a4425ecad7811ca16a361f6
54bc2e553495c553e2b2aacccd353e085e782455
69004 F20110330_AACAVA baltunis_b_Page_050.jp2
57c9c40113fbe01565c9eda935ef6a1e
a5d179e5b6d7bdc998cc47780001af1973d8124c
4640 F20110330_AACATZ baltunis_b_Page_011thm.jpg
bc0021dd6c4ef4d71c92722a3a34d646
fd110287a2dc4ae76f5b1ef59d6cc9967936d444
F20110330_AACBAH baltunis_b_Page_040.tif
6f94d42dddbd5e21f747f9e3dd7297f6
22b3e001218be60687d0b1662d58530c5a0c07b6
133417 F20110330_AACAVB baltunis_b_Page_107.jp2
74441fcad15d10d2f1369e044ca04eb4
7b0c81e446b545681006173cf01289eb849d149e
F20110330_AACAUN baltunis_b_Page_098.tif
73542958ba604573b239b0fb6cc71642
f174b9f7ea735235d97101998f68fdb36554aff0
1823 F20110330_AACBAI baltunis_b_Page_037.txt
1b898961af2190974a215b673b1ad012
df9dfd08722ac7e73d471eb1fbc5ef9cea25fc0e
F20110330_AACAVC baltunis_b_Page_100.tif
349f40f5ee08657e49e40e50952aff28
9a251af3e7316cfed3792eb3272843285d03a854
1903 F20110330_AACAUO baltunis_b_Page_080.txt
f5c5b03c318457bcda3a748a9fc61347
540069e7966c9f563e8df14e2f87335f4cf915c7
45061 F20110330_AACBAJ baltunis_b_Page_047.pro
4107d0b07a678d8210a1bcffe634e49d
a2f81771813373428c2817bef27ec9abd4629510
1059724 F20110330_AACAVD baltunis_b.pdf
bec6d2977ae845d9864fd60b1768a7fe
4e87e3f5dcd2441431702adea321b59119146de3
75298 F20110330_AACAUP baltunis_b_Page_034.jpg
b5a444cbb8cfeda724c89b304d85fbfa
393945ca4c4e033ef26d62d0954b0cd8b463f7d3
83480 F20110330_AACBAK baltunis_b_Page_077.jpg
966bc569281ac1077e3fd029e093544f
d1950fe2503fabbe99b77a9ea940fc1c1ef05ead
19631 F20110330_AACAVE baltunis_b_Page_100.QC.jpg
b54dabc90d42e2a662c85a90821e6c52
8a7aa93b1e1f867f570e67c60aa6382f7d0fd18b
3003 F20110330_AACAUQ baltunis_b_Page_092.QC.jpg
3a76fdd5616a6c2640d4b0e182c1b262
9bbcdb0ece20c17dff7bb8e06b508d86d37acd6f
69998 F20110330_AACBBA baltunis_b_Page_047.jpg
4edf4bd3e2998e04a9f34b77e7002236
9ec5f7e014461b1e7355a4eec08b410708e7c87a
78243 F20110330_AACBAL baltunis_b_Page_033.jpg
42d6511318a406dba5efdc20a04530c1
a205ad26ac1fb39c8fceb905881c93ba54b5e39e
6465 F20110330_AACAVF baltunis_b_Page_029thm.jpg
95d938cea977132c9bb9766471a981e8
f66b028d07a1874cd31d7fa34682ee3926215562
37005 F20110330_AACAUR baltunis_b_Page_030.pro
9d5eca94e1cccc9c9519d2935b156f8b
ec0c96e466c8966af63adc52969cc876c277f54f
45522 F20110330_AACBBB baltunis_b_Page_068.pro
856fc7edeba4c15cd8a3fcb9f4305f84
fb1eaddcf43ed9ba571b0d5a9855cab9bc580ca9
26565 F20110330_AACBAM baltunis_b_Page_042.QC.jpg
016aee9a2a8aad6bbb8e65a8df26ee20
c164633156bc240aaf405873b1174f84f33d0990
82571 F20110330_AACAVG baltunis_b_Page_013.jpg
9aaa9be0d7ebc16712782c4e1f00109d
abd9bcf2865e823f4fa6194c5eecad5b4dd1078e
23495 F20110330_AACAUS baltunis_b_Page_041.QC.jpg
47a5e73dad1193d2fba2fdfd7f21c192
402e79c8bb13ad942664dd51738a2567f6db9dc6
30489 F20110330_AACBBC baltunis_b_Page_086.pro
44670afd41d07cea73ecfdf3c724ac41
906159ce1461d4fbf83c3b9f97c9b0d9e2cfc0f1
6174 F20110330_AACBAN baltunis_b_Page_054thm.jpg
633294913c73a31c76cfac63eac663f5
2d8444d501231b6c2df224a7a80137eba506b0a3
84888 F20110330_AACAVH baltunis_b_Page_053.jp2
c6f7c6fac663013bdbb7e1ffa4b27e13
2412d4af2ff44fa473348ac3c98c891d1838b4c5
68737 F20110330_AACAUT baltunis_b_Page_008.pro
04dc4368202bbd19ef69402d531cb6a5
b46d9de846c9f5ee82bded6019120e2a2f486ef0
4486 F20110330_AACBBD baltunis_b_Page_093.QC.jpg
6ad7a41606359cb037154addfad56938
186c4ee853d37dbc58f4d7a570de1cd2bedc8a07
25728 F20110330_AACBAO baltunis_b_Page_008.QC.jpg
859dd11f7f613d38d8f1c47d169a0e3b
1ba30956825be2c7c8727b35febb524e1535826d
1546 F20110330_AACAVI baltunis_b_Page_023.txt
70f01d1e396e32c03041a356c37a3697
e0a83286a698688bf14779638efeb796d6130061
8423998 F20110330_AACAUU baltunis_b_Page_081.tif
402ae2eb232d06af74c2174e4ccef692
3b526c0f0428726d3e66873bacba1d87ef3699e9
50216 F20110330_AACBBE baltunis_b_Page_057.pro
b7e81a94cf58bf5dc77a1afc3ef30be2
8f58c4a49fd3a40cc74b3e8c77d6ef395b5ce727
84779 F20110330_AACBAP baltunis_b_Page_084.jpg
046b1e83396b6bc9c126adc8facea9cc
dbfbe0e2b65bcf75a15adf017bd0eb67d9ff9ca4
105789 F20110330_AACAVJ baltunis_b_Page_110.jpg
f17af47c0017ce062a1b1d7755f15a5c
334e93285ea55c2cc245e486596ee023a879f697
24794 F20110330_AACAUV baltunis_b_Page_105.QC.jpg
6b2d0e5523ac400a7183de25c84117c1
096af89dde73eefe6ab83dc597aacc8664b4c1fb
36329 F20110330_AACBBF baltunis_b_Page_011.pro
5f1e792977a7f350fbd4b67e68f4561e
cb4a466fa036c712d3e849909a5cf7ea51d5ad6a
28414 F20110330_AACBAQ baltunis_b_Page_029.QC.jpg
40fe823ac261c09bea2db796c108f0e4
0c4a0d2ba5f03e4924a328c4c6d17cb4875e0552
F20110330_AACAVK baltunis_b_Page_107.tif
a033f0de6c9d88704d684949e3e4130d
a6ec91736b6381c3e2be88fc42b0633c1012aa96
65547 F20110330_AACAUW baltunis_b_Page_107.pro
05659ceae238ebac311901fb221d042b
b0fe87255b004127944b94a88d187ac319be3c98
858457 F20110330_AACBBG baltunis_b_Page_082.jp2
92a6cb63a928443f2772f3960f4f4104
6bd735fefdfc4d1f0ebb696e01a251437e50026f
71890 F20110330_AACBAR baltunis_b_Page_076.jpg
197ab945904e135b5b0112f23cf1b3c7
e1bad4b8af4d32634d3de33fef2a90feffc28714
5806 F20110330_AACAVL baltunis_b_Page_068thm.jpg
858e4f4e9f3b4d94f28c9d1c738a0f62
106a8c2b65518f4b4b17b0a564937dd345ecfaf5
1651 F20110330_AACAUX baltunis_b_Page_030.txt
649eed7e11dc0d24b03feabea9ea37af
a78831cf664c87d66bf78780dcbef07032884ced
46560 F20110330_AACAWA baltunis_b_Page_003.jpg
ccc0004b760abd03b7f06c31ebf311dc
cb820823f1212c009441b3daad581d2f25b07050
F20110330_AACBAS baltunis_b_Page_066.tif
63facc8397e2ce304552e34f8d5c6afe
1df7d9faee8610c05e4a4ffc530d9553f081d4fa
18080 F20110330_AACAVM baltunis_b_Page_026.QC.jpg
bbe461df5defc4b87f7625d284c868f7
4f2ff68e2416120bd12c80dba0b43533279648b4
24353 F20110330_AACAUY baltunis_b_Page_001.jp2
4363c9435de6910c726cd7b715626fef
35580194ca9f0325f5766a39029b1a30aa26de76
6539 F20110330_AACBBH baltunis_b_Page_085thm.jpg
c5e1f7f71736ba8f7338038194f7fb13
2b635a1e59fdd8679d4a3dce82bef49a6b9add11
4508 F20110330_AACAWB baltunis_b_Page_026thm.jpg
93e557805bff17432acd73b219722508
2fd919eb096918ea6df391e2f17c045a72fb730d
3384 F20110330_AACBAT baltunis_b_Page_088thm.jpg
fe43c6002a1db760cff0dfc30b92b841
4862e9160775a557b0ab909c027ebd360ae27c78
1505 F20110330_AACAVN baltunis_b_Page_002.QC.jpg
574913e74ebe1d5569ebf7f82991fb69
c42a957d3183773b5b76a5f1efc8ba80f2f0509c
F20110330_AACAUZ baltunis_b_Page_010.tif
ebc18b597533ed4c739004c32b1f8823
91e898e588ec757c06a064d43086a753bfdb0fbc
5202 F20110330_AACBBI baltunis_b_Page_097thm.jpg
fb7484cb7113c229160cac9a1ff3fdbb
b841ac049f4196f2f4587e0dc3da3ac21c4f1f39
25914 F20110330_AACAWC baltunis_b_Page_014.QC.jpg
32889521b571d440eacf8b17cb8618df
7126394c37f9d1662253868320cd7e52a373df33
2116 F20110330_AACBAU baltunis_b_Page_025.txt
59f50e8a2aa2a732a1a140981c7dbf22
337e129220fae8210d22f321112c4257fae3efe3
11852 F20110330_AACAVO baltunis_b_Page_092.jp2
1d41766f94ff2bd174523455f49a14db
5c0fea4a3eaf7ca6c6a7390b8641e71f38654d83
16972 F20110330_AACBBJ baltunis_b_Page_098.QC.jpg
f0ca1f435f897ee8ae0768cc379e56ef
5ed2534412721654c4a8c679c8cf6b40c7525f58
1461 F20110330_AACAWD baltunis_b_Page_083.txt
50415f7d59664c03f93d0a4a79a006d5
d997cd5fa9449797375a26e9a7cfb62efb89e9db
6235 F20110330_AACBAV baltunis_b_Page_032thm.jpg
d8f7304fe718dd15f23b2365bae4b41d
43c406a58784976f832932ce45f6e682599c655a
45642 F20110330_AACAVP baltunis_b_Page_044.pro
6071c2b96c181967ded72c39e15d9f05
51085a7c160afe0f8cd9571d4635beb2dff87a3f
19979 F20110330_AACAWE baltunis_b_Page_027.QC.jpg
e8686692c661eb162fbd86a206a25b7a
e6630e8a022b2b4289a3f37b92b784dfc657bb68
F20110330_AACBAW baltunis_b_Page_079.txt
70f3ba5a50743d3c53f01247ff43826b
9f015c717ea29c01c9c8178842dab648b6e81f5c
670668 F20110330_AACAVQ baltunis_b_Page_036.jp2
cc11644b9c4ed86857539482314a52a3
b9bc5d1067151f89b47eddb2c35bf40ecf26f27a
22683 F20110330_AACBBK baltunis_b_Page_034.QC.jpg
f9083681af2ef986455512e5cc3ed81a
ddcd348cee10488e81e9c59af9139b27d6080348
6086 F20110330_AACAWF baltunis_b_Page_061thm.jpg
ee8be02facabe6890ca19d70ecf575f7
2404e8f96d04b2b1a3211f7703fb5b277f0301e9
106700 F20110330_AACBAX baltunis_b_Page_060.jp2
e500941b5dde4f94f1846672e9fdb412
5e79b982ba168e9fb17a0765aa8e15d816260358
47022 F20110330_AACAVR baltunis_b_Page_087.jpg
a9d45dd42acd5da43eff2b4eeedb927f
37c3462e3a8def76005924ebad8e4e3eff161f96
2449 F20110330_AACBCA baltunis_b_Page_109.txt
792ceb4a269bed4bb1485a182d9809a7
af61b57cc376546cf927d8e71344dbca5622329b
60440 F20110330_AACBBL baltunis_b_Page_011.jpg
434250e4c50d81e84fac72a509e83180
90705b876bb9d719fa73feebb6180c82bb01c109
30213 F20110330_AACAWG baltunis_b_Page_100.pro
7e3d0fd391ef3185274c48dadb7d1ba8
64fb02cdd139cb661992f379bbc9cbca0708b9c6
27068 F20110330_AACBAY baltunis_b_Page_106.QC.jpg
f9562f17564326e8acede9d51f0c71ea
fe6cc8ac0cb38e91d62b221c499d3c6a7ab077d5
1260 F20110330_AACAVS baltunis_b_Page_016.txt
b74c156a5bce9994be4181320c2ae307
44ad23719c52b0452162a5b6a8b43d8a2b6b12d3
26348 F20110330_AACBCB baltunis_b_Page_060.QC.jpg
687ad3e4685eb6a3be1e4ae87edb344c
6924dbc698f63d8c1d97d435b42512a6158e7489
F20110330_AACBBM baltunis_b_Page_056.tif
0b8032d7f2cd01dd7ca46bef7d8df395
485fe8e1c5f298e69b46420e4b81d67065f42640
46682 F20110330_AACAWH baltunis_b_Page_033.pro
1727af0d45e67dac25b930d26e97e0db
3678ba4777745fa599bfae5b012fedc808978ad9
64710 F20110330_AACBAZ baltunis_b_Page_039.jpg
1cdd14765aa4aafcc75480ce0c404be4
c1201afd268aeff3c194ea7483e22e5b34113bb0
71485 F20110330_AACAVT baltunis_b_Page_100.jp2
9f4c7eafa5e729db58de47dba6dce54d
31240de7a7a9cb15f3bd580008bf2fba48f78550
110708 F20110330_AACBCC baltunis_b_Page_079.jp2
51f81f83ce8ebab93985db829450df36
cc46264e9acf58fc44cc4a8c00e79aa2b70cdae2
70271 F20110330_AACBBN baltunis_b_Page_055.jpg
94bad6ef6aca9e7c48c8977b4eb3a7f9
46f80bf6cb6391d599f3a73882f664493a4eba36
F20110330_AACAWI baltunis_b_Page_058.txt
f4990f18e50c86b97f87bfa606a1a347
5bc5eb6fa1628f7032df13bd386240effee291f6
F20110330_AACAVU baltunis_b_Page_096thm.jpg
ac1737ca4db7df05ebb8b00b4f2af519
f44e447111267f5e5491e0c5cba9caaa7b534f20
23101 F20110330_AACBCD baltunis_b_Page_089.QC.jpg
36fa64e49f128fe0dc0c201a02ad0b73
15b62e1ead7617975380d36c1936ce400ab412fd
F20110330_AACBBO baltunis_b_Page_058.tif
a95a897781c963f8a26b9d70baef6e37
8fa5c1302afcf8043e78ece99dbd856138b1f33b
F20110330_AACAWJ baltunis_b_Page_050.tif
8d628d02819ae842cb7b8fd439a764aa
06cf9b99b22a5a602dc06db747160a9a53a5c56f
65605 F20110330_AACAVV baltunis_b_Page_024.jpg
e15d83feb46ae9a386f2b4b2dc4c594f
016877a8732cb6c5cc7acbfd5ae5fec26d72d96b
80972 F20110330_AACBCE baltunis_b_Page_063.jpg
09c9b1c043bcd438c564f59bfefc59f5
dc8481d3221c2586cebf89deffb1e5b2a4a6d8f1
73851 F20110330_AACBBP baltunis_b_Page_026.jp2
91023eeaa6a28c0696da6b8d6059b405
1b47b491d5899e1b6960fb6c1bf494172d0c8ca4
1554 F20110330_AACAWK baltunis_b_Page_099.txt
ba2c48635ce8a709d78490eb06d4267e
709df3c64e61286de88e72f2b31cbb23a9a54741
69404 F20110330_AACAVW baltunis_b_Page_103.jp2
7e31fa7e4f2778a366fa7a18f07bd4e1
7bc18b8677d6314db0d32549a44c0db0de7906aa
22852 F20110330_AACBCF baltunis_b_Page_055.QC.jpg
33aac97bc46667c1067b929af5f3d9cb
226bb05d2c926dc2cbe82b042a164009894eeb19
F20110330_AACBBQ baltunis_b_Page_075.tif
654bcc4f6bf478045cd28114fa10cda1
bc19a931bb233e9b9de1a11661df0e788287cc55
1879 F20110330_AACAWL baltunis_b_Page_038.txt
5a2f4004efb1ddfb08d6533550eab3aa
fc3e5116219b494185a3a26fcf3d7e338c9af82f
4571 F20110330_AACAVX baltunis_b_Page_002.jpg
7979a04664544d4c28d89f11c07ad4be
d4241f702cae697b1fb14c059df02bcfddb9cb49
21156 F20110330_AACBCG baltunis_b_Page_099.QC.jpg
ac79df497af6ebb3fac1151c04eebaa2
b07312d8488ffbb439e09b79e1621d6fa8b203f0
82875 F20110330_AACBBR baltunis_b_Page_056.jpg
4a05658828c6dc6d95cba1f5676e9efe
55e6b111aa6b9eb111d0c14db779aa46b4e7ca11
15340 F20110330_AACAWM baltunis_b_Page_088.QC.jpg
d1371a198a86bd92d22f412ed1482d87
e7a5c247ba8b6f4580079abea8b8eb1f5738a58b
F20110330_AACAVY baltunis_b_Page_041.txt
9f447202fc42125b327ab9eebdfa88ec
c4ba69f0cfbfba43a9443cfb773177d730c20a93
69086 F20110330_AACBCH baltunis_b_Page_048.jp2
0ea3f7bbcedbaebe8799a59c889c25a3
74af61e325561cb93b4d740f79cb5107f9625a7f
19392 F20110330_AACAXA baltunis_b_Page_049.QC.jpg
924b0aa51bd6e027f6930af1d44d97a3
c645faa446261c11a0c931f4e985ff9b0883d939
2073 F20110330_AACBBS baltunis_b_Page_097.txt
217441176c0102a7305ab096ad66c40a
433995a7b7942028739f8b16f00f03c823a0d0b2
3547 F20110330_AACAWN baltunis_b_Page_087thm.jpg
2be2e302e5cdfa66611f6ccb240a91a8
b636b3ed2cbc6fb06c5c8cf7b471a110dafc0dab
2343 F20110330_AACAVZ baltunis_b_Page_055.txt
55d76dc8608fb4f82878e5403cef3946
706d1d7996cc3996b1296e774f120847333a9290
50370 F20110330_AACAXB baltunis_b_Page_014.pro
ba482b9cec6133f23b9973b76ce17595
28b96117c6c93fac10e38cac02a0435414bd2ef7
5538 F20110330_AACBBT baltunis_b_Page_034thm.jpg
f106ae53fe363a0792301eb875a6c973
503896bbadea16891e21759ccd6e689d50be50fe
F20110330_AACAWO baltunis_b_Page_064.tif
b2dde11b8564db3e6d2845c0e04ede57
c3e18939d45fa6cee769e2a04a16d74b1e727f2c
1835 F20110330_AACBCI baltunis_b_Page_044.txt
21022cc3cc48bddd4a6533ae84435250
eb99d0ddfc9ce08244269c40161a21eab27f40e9
F20110330_AACAXC baltunis_b_Page_086.tif
56bc093aa71a9cee4d47a4c1dc3d57dc
5e061bf2481a819a3e18e2bc455859ae929b3e8f
37652 F20110330_AACBBU baltunis_b_Page_104.jpg
af950f9940f9988ea6aed2b3b7dc3ac2
3276aa06131c95cfb2187b364ca1d4b0efb9bb43
1660 F20110330_AACAWP baltunis_b_Page_010.txt
da3be0e9185b15f43eada679fd9069e4
c0146d6537ceaacfeddcadb8d0eeb4cff94bf048
44928 F20110330_AACBCJ baltunis_b_Page_035.pro
9913d9a74778dd416aee16c6bc22a115
994781d5fac6c92c94e15996e13cfa9079226f47
72442 F20110330_AACAXD baltunis_b_Page_086.jp2
7c4c3e6b0ef3bc486dd14e538f712ee4
2e0244ab2b2ad1d07103172501182dcd57315ec1
23405 F20110330_AACBBV baltunis_b_Page_001.jpg
803174b5ca023ccbffa0604c0c5b5d8b
d22c3d790557963ea486c8ff25a7e7364ab35eb4
19349 F20110330_AACAWQ baltunis_b_Page_103.QC.jpg
1c63aff7db6bfaa54114ce826c2cd0ee
bc9231cd65bda88e3fe05dedc68af22efd3f720f
84056 F20110330_AACBCK baltunis_b_Page_015.jpg
32be3dd71b4afff8009abb804978ee5b
2c3242bd7d80047f0d03235c7889619a83c75083
104440 F20110330_AACAXE baltunis_b_Page_080.jp2
4a07434b805e7edf4f90f070e7b1119f
db682bc3c0c5f820ac81fdf92a3a53ac33f3b681
26168 F20110330_AACBBW baltunis_b_Page_090.QC.jpg
a7bcfab9bb721599aa0f455691f6d0c4
c019c6a798e0fed25cd940ed17d8bd949b32e0dc
5420 F20110330_AACAWR baltunis_b_Page_089thm.jpg
f6256c2dfe7785eb08f693b6b3e18c9c
9cf22bd9bd0b47414b3d6e04cb76948e195d4557
4266 F20110330_AACBDA baltunis_b_Page_103thm.jpg
a7005b4e3f17a55c3d02d4bcde2cfb26
9a5a6b703fc9785978752e68588bc28e47dab890
5138 F20110330_AACBCL baltunis_b_Page_049thm.jpg
bbdf495d6a666d278fad0ce1548dc176
84dacd7b2896700e85139625b1f133ca0b64a6ef
F20110330_AACAXF baltunis_b_Page_036.tif
3a642b988e269a8b44558a8f8b9ade18
a7c663ffb2a5e6c28dac5699f720665887a2efab
85684 F20110330_AACBBX baltunis_b_Page_049.jp2
a179422468855e0023ad4cbb7e8c6d21
6da4b4762a0efd9c19ace3738dbc2080a7a04930
6246 F20110330_AACAWS baltunis_b_Page_058thm.jpg
bd86924fe8fcaf334d6fa02616b11a3f
b2b548ca5fad3ebe5ac35ef55bf0d2817c3649d5
20759 F20110330_AACBDB baltunis_b_Page_030.QC.jpg
bab8ce51c87aafd3515807b8f2e076e2
49e1440616da0ac60176fa05899bd7032f970e23
76560 F20110330_AACBCM baltunis_b_Page_041.jpg
2a08d748059e665377b2104c07d9d67d
6bb90cb1654309744995dca3019b9388df3aead6
55191 F20110330_AACAXG baltunis_b_Page_026.jpg
11ebc15a0e25f6a159e480ac753fad07
ca24f4c8e1ff54ff4a0336d620f7f5b035dfb17e
1902 F20110330_AACBBY baltunis_b_Page_061.txt
102a3a6963ec2203576c66eafc568e4d
22cf3d53798b71d5d90d120749ff69ef7e986d1b
F20110330_AACAWT baltunis_b_Page_060.tif
b6882252af736ec0eb70abe4d6ee9fea
52567ed689e63b58704a9bcff9321efae3b09352
9841 F20110330_AACBDC baltunis_b_Page_094.QC.jpg
508eb97379d93ae7f322a2e20e91f64c
10b87a21d7ffbefb556af283ee87885670479eca
864673 F20110330_AACBCN baltunis_b_Page_083.jp2
cf2c35920cfb46c451ea44066d14ea67
c6f18305302829764edd1dbe589b8e8cb192e7ba
F20110330_AACAXH baltunis_b_Page_013.tif
1a8db6b73808f2befba4835f6fe0ceff
e39561cea68d495a1d8f52b61b936689185ee8b1
51719 F20110330_AACBBZ baltunis_b_Page_067.pro
8456c0baa9de92ed11a5d7e13647a282
47d7a292998b6c292d3ef19f3347070b1fef280a
996269 F20110330_AACAWU baltunis_b_Page_081.jp2
604a7562dadc0583570bc7826fe0f244
330e64c80c7ad21db74350529dd3cd6124a09452
48437 F20110330_AACBDD baltunis_b_Page_081.pro
bb1518a7704a4b78534c7c4ace1e75f6
680acc6b2e76447f9ff7579011de8c27b507b547
131188 F20110330_AACBCO baltunis_b_Page_110.jp2
d3cf7a618157b742a00ff910179ae21f
5faa6ec65c64306bf3a99e58c73fb7965f7d0d67
916 F20110330_AACAXI baltunis_b_Page_036.txt
8a90ee661baefb7427bd34ed32be5fc7
d380922ef58d689894c38c273971aa889057810d
F20110330_AACAWV baltunis_b_Page_088.txt
82385c2533222755289fdbf6cfb46f0a
0a04bb1dad61f99507e9060c9c53581a227942ee
F20110330_AACBDE baltunis_b_Page_066.txt
3c6daec052a92ac8f8d847aa07cb96e9
8ef78c677ffca336a65661886514f44c2ade752f
19191 F20110330_AACBCP baltunis_b_Page_102.QC.jpg
c2095d51bb21fe5d3b42da0310af097a
ded9c356d6fafc079dc5212618bf5a6840264c16
99205 F20110330_AACAXJ baltunis_b_Page_035.jp2
e31b392bb5afb611048731f18684a1b4
8fb5d8c78d85358c3fafdf89ff7bac49080756c1
32564 F20110330_AACAWW baltunis_b_Page_099.pro
279658a96359200948f55609011e2c7e
b9bdd347473d8ecf7f62cda7fa93cb620a0eb78e
72654 F20110330_AACBDF baltunis_b_Page_007.pro
24cbe8d4dafb148626c9dce05a1f1ade
d46797ccfce44fe08ec6cae8c64d670916061d1f
F20110330_AACBCQ baltunis_b_Page_082.tif
ec368824ca1bb66808165c9aa95eacbe
f51a1aa4e6bf64c316e372bd54cee258edd52cbd
6274 F20110330_AACAXK baltunis_b_Page_021thm.jpg
27d729fba5f37dda30b72c6a42e3b76f
9653b1e7a139fad6c5c118699e645bc7ca7e926a
43799 F20110330_AACAWX baltunis_b_Page_112.jp2
5bacd9647da8276f45fc824337dcca53
3e77a4be0cf5695faf459649feb792de749d680e
34223 F20110330_AACBDG baltunis_b_Page_052.pro
ad45953e7a8092ba594156a8681846ea
8bb73f69504ba16e061ef678096ea96837531627
2646 F20110330_AACBCR baltunis_b_Page_107.txt
5778e43b427f22d24248bad3c4a82aa9
1a8558abc9f418a7a0574a11f6b60e897d7a7606
1051977 F20110330_AACAXL baltunis_b_Page_006.jp2
1d23d4fbb2f17b33ab2ef87cd8f2f188
2170da098d33d109226f9cc8a8f9574f888b280c
2000 F20110330_AACAWY baltunis_b_Page_032.txt
5c1d6a1ee8a33c44f6e25de1b1a96154
9e91faa8ed42fe3b1dd4b44e14d760fafdb4adc2
39474 F20110330_AACBDH baltunis_b_Page_027.pro
b0fc3bba3d612f2003fcbb9eee793879
a7dfa3d4f3698215e0891a5d8b6fe7d87f81227f
34492 F20110330_AACAYA baltunis_b_Page_095.jpg
99559579250f7f5150b2a02d44fffdca
d7cb396b0445aec251821da80362800e347f4b48
6025 F20110330_AACBCS baltunis_b_Page_002.jp2
c42914ddfe7065f8dcf518f6f0114efc
78c2a4fd11d92402af73bad6885814222f217f66
62466 F20110330_AACAXM baltunis_b_Page_049.jpg
b07cd043504a52191ffcc3f9f2a9bc56
33ecb1ae16abe517e2dc469b8dbe87aa8c696fae
29844 F20110330_AACAWZ baltunis_b_Page_107.QC.jpg
fa9f8390f9a1eac2ab3b52f19d13e0dc
0038b123fb88060d425bf26b4f3049e345ef5e2f
44915 F20110330_AACBDI baltunis_b_Page_074.pro
0ec0b7a7f397663085cb9be5886aba8a
409c8ad7b1a98b61d98cac6d4d86d58dfed060c3
5645 F20110330_AACAYB baltunis_b_Page_055thm.jpg
d4b09c8251536648e53b3bb735860055
9075ecca144f10a7438042fa52a8e16772ff19b7
39776 F20110330_AACBCT baltunis_b_Page_039.pro
23c53a63a2c8d37a165444e8e390d822
48f726b7ec0c30420615db8ea368a4ac9887694e
11359 F20110330_AACAXN baltunis_b_Page_095.QC.jpg
5e3729f35cda462ea8681f91ea584d05
89903b316404d5ca04e36550e683c279d552aed6
457 F20110330_AACAYC baltunis_b_Page_001.txt
cc64d3b84c62b716db3193a5c184e71e
5abc7cfe35ab168c5d2b7b8088378c93e4e796d1
1915 F20110330_AACBCU baltunis_b_Page_043.txt
7d0320737cac08d07c07d0fd3fc1f80a
1949cd2e085b82ee8e962848d21a1df18f061c90
6338 F20110330_AACAXO baltunis_b_Page_042thm.jpg
d5fc247e90c75187441aa082ddabf6e7
89393027f66e661ca66f12ab588b8b589ff1de38
F20110330_AACBDJ baltunis_b_Page_050thm.jpg
ee4aa12b9676528b27ab9b9fbc77bb68
ed9c93c31f9ffc24c147686c81c39d20fb82fe7f
5526 F20110330_AACAYD baltunis_b_Page_024thm.jpg
df6bb0fe8993296dec7a00c1220a5de0
8f6b8e90f12b2bc28acfaeada982d6981a265fef
4108 F20110330_AACBCV baltunis_b_Page_016thm.jpg
43430e6d2d87938d21a11c98b31437a6
a564930572ff6ce854036ac9a5817eca34b46110
3090 F20110330_AACAXP baltunis_b_Page_104thm.jpg
3f431140bd1e63a1aa1a51ca466fd275
7ef18dabc304f255e9b0344ca70667a65ce2db03
F20110330_AACBDK baltunis_b_Page_104.tif
97a9bc398db6d7afd963619f5a912aca
9a58a34412e1790a40a49796c343d670d92332da
110762 F20110330_AACAYE baltunis_b_Page_032.jp2
7eeb1708d2042160bb43d83159b4a23f
da07060bd5a6de7855571ac9c27ced450dc7058a
5799 F20110330_AACBCW baltunis_b_Page_035thm.jpg
96eee3f02da6c27f8ca16e801eaae16b
1385b02cf3fb5fe644fb392496f257ded72ec47f
1895 F20110330_AACAXQ baltunis_b_Page_054.txt
2129f100532d04da3e1dce0a6f4d9f1a
7915e448a651ad38fd631bf4b43215856b163046
14763 F20110330_AACBDL baltunis_b_Page_087.QC.jpg
e84d974179cc41ad1eac98597db105d1
e19691ee5e0489ae91129b87f019d9b8474789dc
1705 F20110330_AACAYF baltunis_b_Page_017.txt
fb69362efc73783b4a75ae41ce349325
432ea09a544d89489d90fc666be676ba1cc1aa8c
F20110330_AACBCX baltunis_b_Page_110.tif
574ccbaef157bad1e5aba5616f225673
e0996dfd336e66994e73fb2a0a2c02d1b745187f
8430 F20110330_AACAXR baltunis_b_Page_001.pro
678dae7b9807bf1bba767b08fbe4e0f3
521ed7b4686bfa4bb0ee06b75e4e9b58a7f0f525
25265604 F20110330_AACBEA baltunis_b_Page_059.tif
58a8e2fbf0ffe054893a479b6fe1bbae
d11e3756a32fc6551fe0ae1d98b02a83da5a4d2c
71639 F20110330_AACBDM baltunis_b_Page_017.jpg
074e59fedb5a35d2c0726867673fa0be
c5a0c3f0fd010e790836f61367cea091f5141fc3
2390 F20110330_AACAYG baltunis_b_Page_057.txt
780feb375a5e4f5b8e24ba10525c1f85
9623cd29af4980f4016c5097e9593955dc143874
23465 F20110330_AACBCY baltunis_b_Page_062.QC.jpg
bbb7c07142531c333f20f3bb1a51da97
5d65c49c351579ce0120f41b7cb521e50a7854c3
47824 F20110330_AACAXS baltunis_b_Page_071.jpg
315417ec3dde45b4c34094ba072b6ac3
1b3501bf77036f98a1d10b38617c295488ced1a0
51607 F20110330_AACBEB baltunis_b_Page_042.pro
2307d533bf0d69213977c83112f078ab
49bb4edea16b14bcff23b2c2d49c14797f31569c
22723 F20110330_AACBDN baltunis_b_Page_028.QC.jpg
098a8d4f759dc163f2c0915c05dcc5f2
302dc5f0e9bf1e0b8daf129cbfd3ed4c10332915
5229 F20110330_AACAYH baltunis_b_Page_023thm.jpg
1ddd842af35f66a3c411b7e6762f3087
25637b820b05ec758a240e4df9238acc06cd32e9
79251 F20110330_AACBCZ baltunis_b_Page_004.pro
01d4b642a784ef25b2f360afe9f883dc
07ad7a03a1d8fbcbc941db0cc37636a7148b88b6
22994 F20110330_AACAXT baltunis_b_Page_074.QC.jpg
e00876e9b8bc60361ca4787cb39d1622
459997b91a2ef699caff8b9b5ba572c46e336b37
F20110330_AACBEC baltunis_b_Page_079.QC.jpg
1b023a61d2d8c564bd887a3357faa7e2
58e4018cf92a058f24e448dc94b6e9a0a934e60c
24900 F20110330_AACBDO baltunis_b_Page_018.QC.jpg
493472fc12ee49ee9944d6b38cfcaeb6
065e92ba09b0bfbcb46e9b99fbcdf4cc21ca4ec2
F20110330_AACAYI baltunis_b_Page_019.tif
b4be078122df83554e6f6358aa2db724
e88b5f74743abff81f646f500d1433de67f39946
2113 F20110330_AACAXU baltunis_b_Page_046.txt
01eebee7b8c1a753a42eec66d6eafe09
7809c6c401e5b7270d26e7302a85d3ab2015b208
28456 F20110330_AACBED baltunis_b_Page_110.QC.jpg
ef0f2276352bbca3657525dddc524c4c
ccc99caa32a0f341852f371fcac469642f3e58c8
101916 F20110330_AACBDP baltunis_b_Page_054.jp2
82ab59b13f6e14797835fc66d6fa1a6a
c18fdaaa582fed721804a0751d5063c0109797c5
F20110330_AACAYJ baltunis_b_Page_094.tif
4a15014d023b759db522c835dc1bf4e3
70cc73ccbb8fbc2fa3137d86b146335281909e0a
109477 F20110330_AACAXV baltunis_b_Page_058.jp2
56ae88cfa9cc797346c9ecbac7c66f54
fd70bce63fc3d10dd113951a41a77a153d929e0d
42575 F20110330_AACBEE baltunis_b_Page_025.pro
2f63a81c8ed7c19b9a553e69ac6c2861
f64aca991c0254a54ca1442d710441611513e830
96530 F20110330_AACBDQ baltunis_b_Page_089.jp2
3e35a27a30b92729e9f404ff7c3e8619
0024f888d74f79c2c4880bc689ed74a4ba006486
2033 F20110330_AACAYK baltunis_b_Page_090.txt
80beeece547e33b5b7a44999c8ff63f2
3251bff612f1aa5585f775919fed8fdb721bcd11
23572 F20110330_AACAXW baltunis_b_Page_054.QC.jpg
3225214a1dbf85866ff51a25138d3a3f
781668cbbf418ffdb305e1c053bb2b9c4d81646a
1905 F20110330_AACBEF baltunis_b_Page_074.txt
486d13136d6f6f6bd08b80f41edb9c5c
ab5b042ae4c362b770b9a8dcdfec862da80eba62


Permanent Link: http://ufdc.ufl.edu/UFE0012944/00001

Material Information

Title: Genetic Effects of Rooting Ability and Early Growth Traits in Loblolly Pine Clones
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0012944:00001

Permanent Link: http://ufdc.ufl.edu/UFE0012944/00001

Material Information

Title: Genetic Effects of Rooting Ability and Early Growth Traits in Loblolly Pine Clones
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0012944:00001


This item has the following downloads:


Full Text












GENETIC EFFECTS OF ROOTING ABILITY AND EARLY GROWTH
TRAITS IN LOBLOLLY PINE CLONES
















By

BRIAN STEPHEN BALTUNIS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Brian Stephen Baltunis















ACKNOWLEDGMENTS

I would like to thank Drs. Timothy White, Dudley Huber, Barry Goldfarb, Hank

Stelzer, John Davis, and Rongling Wu for serving on my advisory committee. Special

thanks go to Tim White and Dudley Huber for their guidance, advice, and support

throughout my entire program.

I also would like to thank the Forest Biology Research Cooperative for financial

support for this research, the members of the FBRC for establishing field trials, and

especially International Paper Company for providing the facilities for the propagation of

the rooted cuttings for this project. Many people were involved with bringing this study

together, and I would like to express my gratitude to all the faculty, staff, workers, and

fellow graduate students who helped. My appreciation especially goes to Brian Roth;

without his effort, this project could never have happened.

Finally, I would like to thank my wife, Jacqueline, for her patience and

encouragement these last five years.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ................................................................................................. iii

LIST OF TABLES .............. ............................................... ........ vi

L IST O F FIG U R E S .............. ............................ ............. ........... ... ........ viii

A B ST R A C T ................. .......................................................................................... x

CHAPTERS

1 IN T R O D U C T IO N ............................................................................. .............. ...

2 GENETIC EFFECTS OF ROOTING LOBLOLLY PINE STEM CUTTINGS
FROM A PARTIAL DIALLEL MATING DESIGN................................. ...............6.

Introduction ........................... .................................................... .6
M materials and M methods ................................................................. ....................... 8
P population ............................................................................................8
E xperim ental D esign ............................................................. ....................... 9
Statistical A n aly ses........... ................................................ .......... ..... ... .. 10
R results and D iscu ssion ......... ................................................................... 16
A average R ooting ........... ...... .............................................. ...... .......... ....... 16
Observed Variance Components ............... ........ ..................................... 18
C ausal V ariance C om ponents.......... ................. ......................... ........ ....... 21
H eritability E stim ates ............................ ................................ .... ...... ...... 22
Type B Genetic Correlations ............................. ......... ....... ........... 26
Selection for R ooting .......................... .................................. ...... ... ...... 27
C onclu sion ........... ......... ............................................. ... ............. 28

3 GENETIC ANALYSIS OF EARLY FIELD GROWTH OF LOBLOLLY PINE
CLONES AND SEEDLINGS FROM THE SAME FULL-SIB FAMILIES .............30

Intro du action ...................................... ................................................ 3 0
M materials and M methods ....................................................................... ..................33
Population ..................................... .......................... .... ......... 33
P ro p ag atio n ........................................................................................... ..... 3 4
F ield D esign ................................................. .................. 35
Statistical A n aly ses........... ................................................ .......... ..... .... .. 36









Results and Discussion ..................................... ...... .... ...... .. .......... 43
Overall Growth of Clones and Seedlings .................................... .................43
Genetic Components of Variance............................................... .................. 44
H eritability Estim ates ............................................. ... ....... .. .......... 49
Type B Genetic Correlations Between Propagule Types.............. .............. 50
Genotype x Environment Interaction ...................................... ............... 52
C o n c lu sio n ...................................... ............................... ................ 5 4

4 GENETIC GAIN FROM SELECTION FOR ROOTING ABILITY AND EARLY
GROWTH IN VEGETATIVELY PROPAGATED CLONES OF LOBLOLLY
P IN E ..................................................................................................................... 5 5

In tro d u ctio n ........................................................................................5 5
M materials and M methods ....................................................................... ..................57
Population ....................................................................57
R ooting and Field Trials.............................................. ............................ 57
Statistical A n aly ses........... ................................................ .......... ..... .... .. 58
Causal Com ponents of V ariance ................................ ............... ...................60
H eritability E stim ates ............................ ................................ .... ............6 1
Type B G enetic C orrelations ................................................................... ... ..62
G genetic G ain ................................................................... ............ 63
R results and D iscu ssion .............................. ......................... ... ........ .... ............65
Causal Com ponents of V ariance ................................ ............... ...................65
Type B G enetic C orrelations ................................................................... ... ..66
G genetic G ain ................................................................... ............ 68
C conclusion ...................................................................................................... ....... 75

5 C O N C L U SIO N ......... ...................................................................... ......... .. ..... .. 78

APPENDICES

A LOBLOLLY PINE PARTIAL DIALLEL MATING DESIGN. THIRTY-TWO
PARENTS WERE CROSSED TO GENERATE 70 FULL-SIB FAMILIES ............81

B VARIANCE COMPONENT ESTIMATES FOR ROOTING ABILITY .................. 83

C VARIANCE COMPONENT ESTIMATES FOR EARLY GROWTH TRAITS
OF LOBLOLLY PINE CLONES AND SEEDLINGS FROM THE SAME FULL-
SIB FA M IL IE S .......... ... ......... ......... ...................... ................. 86

D VARIANCE COMPONENT ESTIMATES FROM THE BIVARIATE
ANALYSES OF ROOTING AND 2ND YEAR HEIGHT ..........................................88

L IST O F R EFE R E N C E S ............................................................................. ............. 94

BIOGRAPHICAL SKETCH ............................................................. ............... 101
















LIST OF TABLES


Table p

2-1. Experimental design for 5 rooting trials of loblolly pine stem cuttings. All trials
were established in randomized complete block designs with 4 to 6 blocks and 4
to 9 ramets per clone in a row plot within each block..............................................9

2-2. Summary of rooting from 5 loblolly pine trials set over two years and three
se a so n s ................................. ........................................................... ............... 1 7

2-3. Genetic parameter estimates (standard error) for rooting of loblolly pine stem
cuttings across 5 trials. ........................... .................... ......................... 23

2-4. Type B additive and dominance variance correlations among pairs of rooting
trials for loblolly pine stem cuttings (above and below diagonal, respectively).
Standard errors are given in parentheses .................................... ................26

3-1. Location of six field trials, establishment date and total number of test trees for
each test. .......................................................... ................ 3 4

3-2. Total number of clones, full-sib families, half-sib families, average number of
clones per full-sib and half-sib family, and average number of seedlings per full-
sib and half-sib family established at the six field trials. .......................................36

3-3. Mean 1st year height, 2nd year height, height increment, and crown width by
propagule type for each of the six field trials. Although means are expressed in
meters, analyses were conducted using measured traits in centimeters...............44

3-4. Genetic parameter estimates for lst year height, 2nd year height, height increment,
and crown width by propagule type across all six trials. Standard errors are
given in parentheses. ........................ ...................... ... ...... .. .... ........... 46

3-5. Genetic correlations between propagule types for 1st year height, 2nd year height,
height increment, and crown width at the parental ( ro ) and full-sib family

( ropF levels ................................................51

4-1. Means, variance component estimates, heritabilities, and genetic correlations
from the bivariate analysis of rooting ability and 2nd year height. Standard errors
are given in parentheses. ........................................ ............................................67









4-2. The total genetic correlation (?, ) between rooting ability and 2"d year height
from analyses of a single rooting trial and field trial. Shaded values indicate the
rooting trial in which cuttings originated from for their respective field trials.
Standard errors are given in parentheses ....... ........ ................................... 70

4-3. The number of full-sib families (half-sib families) and average number of clones
per full-sib family (half-sib family) selected from selecting 10% or 1% of the
top clones using the combined selection index. ....................................... .......... 77

B-1. Observed variance component estimates for rooting of loblolly pine stem
cuttings from single-trial analyses....................................... ......................... 83

B-2. Observed variance component estimates for rooting of loblolly pine stem
cuttings from pair-wise-trial analyses. ........................................ ............... 84

B-3. Observed variance components for rooting ability from the across-trial analysis
using all five rooting trials. ............................................. ............................. 85

C-1. Observed variance components for loblolly pine clones from the across-trial
analyses of 1st year height, 2nd year height, height increment, and crown width.
A separate error variance was modeled for each trial. ...........................................86

C-2. Observed variance components for loblolly pine seedlings from the across-trial
analyses of 1st year height, 2nd year height, height increment, and crown width.
A separate error variance was modeled for each trial. ...........................................87

D-1. Observed variance component estimates from the bivariate analyses of rooting
ability from SpringO1 and 2nd year height from each of the field trials ..................88

D-2. Observed variance component estimates from the bivariate analyses of rooting
ability from Summer01 and 2nd year height from each of the field trials. ...............89

D-3. Observed variance component estimates from the bivariate analyses of rooting
ability from Winter02 and 2nd year height from each of the field trials .................90

D-4. Observed variance component estimates from the bivariate analyses of rooting
ability from Spring02 and 2nd year height from each of the field trials .................91

D-5. Observed variance component estimates from the bivariate analyses of rooting
ability from Summer02 and 2nd year height from each of the field trials. ...............92

D-6. Observed variance component estimates from the bivariate analysis of rooting
ability using all five rooting trials and 2nd year height using all six field trials. ......93















LIST OF FIGURES


Figure page

2-1. The proportion of the additive (h0 ), dominance (d, ), and epistasis (o )
genetic variances on the observed binary scale for rooting of loblolly pine stem
cuttings from each of the five separate trials (biased heritabilities) and from the
combined analysis of all five trials. Standard error bars for broad-sense
heritability estim ates are included ......................................................................... 19

2-2. Narrow-sense (h2 ) and broad-sense (H ) heritability estimates for rooting of
loblolly pine stem cuttings transformed to the underlying normal scale using the
threshold model of Equation 2-11. Standard error bars are included....................25

2-3. Full-sib family mean (H2 ) and clonal mean (H 2) heritability estimates for
rooting success of 2,200 clones from 70 full-sib families of loblolly pine.
Standard error bars are included........................................ ........................... 25

3-1. The proportion of additive (add) and nonadditive (na) genetic variance
components for clones and seedlings across the six field trials, where h2 = add
and H2 = add + na: a. 1st year height, b. 2nd year height, c. Height increment,
and d. Crown width. Standard error bars for broad-sense heritability estimates
are in clu ded ....................................................... ................. 4 8

3-2. Rank-rank plots showing type B genetic correlations between clones and
seedlings from Test B based on: a. Parental BLUP values, b. Full-sib family
BLUP values, where full-sib BLUP values are equal to the sum of the predicted
general combining ability for each of the two parents plus the predicted specific
com bining ability of the cross. ............. ........... ..... ..... ........ ............... 53

4-1. The genetic gain in rooting ability (%) over the population mean for deployment
of the best half-sib family, full-sib family, best clone from the best ten full-sib
families, and the single best clone when selecting for rooting ability or 2nd year
h e ig h t .................................................................................................................. 7 1

4-2. The genetic gain in 2nd year height (%) over the population mean for deployment
of the best half-sib family, full-sib family, best clone from the best ten full-sib
families, and the single best clone when selecting for rooting ability or 2nd year
height ...............................................................................72









4-3. Responses to selection in rooting ability and 2nd year height with various
selection indices for two clonal deployment options: 10% of clones selected and
1% of clones selected. .............................................. ............... .... ...... ...... 75















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

GENETIC EFFECTS OF ROOTING ABILITY AND EARLY GROWTH
TRAITS IN LOBLOLLY PINE CLONES

By

Brian Stephen Baltunis

December 2005

Chair: Timothy L. White
Major Department: Forest Resources and Conservation

Loblolly pine is the most important commercial tree species in the southern United

States with over 1.1 billion seedlings planted annually. With elite genotypes becoming

available, several forest industry companies in the southeastern United States are

developing rooted cutting and somatic embryogenesis programs aiming towards

deployment of tested clones or families. However, before clones can be deployed,

sufficient data need to be collected on the population in order to have reliable information

about the clones for deployment decisions.

This dissertation reports on the genetic effects of rooting ability and early growth

traits in nearly 2,200 clones of loblolly pine from 70-full-sib families. More than

239,000 stem cuttings were set in five rooting trials over two years. Overall rooting

success across the five trials was 43%, and significant seasonal effects were observed.

Heritability of rooting ability was estimated both on the observed binary scale and on the

transformed underlying normal scale.









Rooted cuttings from these trials along with seedlings from the same full-sib

families were established at several sites, and early growth traits through age two were

compared between propagule types. All growth traits demonstrated genetic variation,

and parental and full-sib family rankings were similar for both propagule types.

However, estimates of dominance genetic variance in the seedling population appear to

be inflated at the expense of additive effects due to a lack of randomization of seedlings

prior to field establishment. Little genotype x environment interaction was observed

across sites for all traits.

A successful clonal forestry program for loblolly pine based on rooted cutting

technology needs to consider selection for both rooting ability and subsequent growth.

There was a positive genetic correlation between rooting ability and 2nd year height at the

parental, full-sib family, and clonal levels indicating that selection for one trait will also

lead to improvement of the other. The genetic gains in rooting ability and 2nd year height

associated with several selection and deployment strategies are discussed. Moderate to

high family and clonal mean heritabilities, moderate to high type B correlations, and

substantial among-family and among-clone genetic variation indicate the potential for

increasing rooting efficiency and improving growth.














CHAPTER 1
INTRODUCTION

Loblolly pine (Pinus taeda L.) is the most important commercial tree species in the

United States with over one billion seedlings planted annually (McKeand et al. 2003).

Most commercially important tree species remain relatively undomesticated, and loblolly

pine is no exception. Genetic improvement of loblolly pine has been occurring since the

1950's in several tree improvement programs. There are three cooperative tree

improvement programs in the southern United States that focus on improvement of

southern pine species including loblolly pine: the Cooperative Forest Genetics Research

Program (CFGRP), North Carolina State University-Industry Cooperative Tree

Improvement Program (NCSUITIP), and the Western Gulf Forest Tree Improvement

Program (WGFTIP). Loblolly pine tree improvement programs in the South are

beginning their 3rd generation of breeding with gains in volume per unit area up to 30%

over unimproved loblolly pine (McKeand et al. 2003).

Long-term tree improvement programs aim to increase the population mean

breeding value of a few key traits such as stem volume, disease resistance, and wood

properties through breeding and selection of superior genotypes. Loblolly pine tree

improvement programs are based on recurrent selection for general combining ability,

which captures only the additive portion of the genetic variance. However, the

nonadditive portion of genetic variation, dominance and epistasis, may be important

components of variation for traits.









Both additive and nonadditive genetic variation can be captured by deploying full-

sib families or clones. However, deployment decisions should be based on reliable

information. Field trials established with full-sib seedlings allow the genetic variation to

be partitioned into additive and nonadditive components. These trials not only provide

ranks of parents or individuals for selection, but also of full-sib families in order to

provide information for making deployment decisions.

In any given generation of breeding, maximum genetic gains can be achieved in the

deployment population by capturing all of the genetic variation through operational

propagation and deployment of selected clones. However, clonal forestry is not a

breeding method to develop better genotypes. Clonal forestry is a method to mass-

produce well-tested genotypes. Short-term genetic gains may be maximized through

deployment of well-tested clones, but long-term gains need to involve both clonal

selection and recurrent selection for additive genetic variation through repeated selection

and breeding.

Clonal tests derived from full-sib families do provide an opportunity to estimate

additive and nonadditive components of variance associated with a particular trait or set

of traits. However, tests should be designed with a sufficient genetic structure in order to

precisely quantify the genetic variation. For example, Frampton and Huber (1995)

reported that they had low power in partitioning the genetic variation because of the lack

of a mating design among the parents of the full-sib crosses in a loblolly pine clonal

study. In another study comparing clones from 30 full-sib families derived from two

disconnected 4x4 factorials, Paul et al. (1997) concluded that future clonal studies should

include more parents in the mating design. Although Isik et al. (2003) estimated the









genetic variances from both clones and seedlings from the same nine full-sib families of

loblolly pine, they identified a weakness of their study in that there were a limited

number of parent trees used in the mating design. Finally, Frampton and Foster (1993)

warned that interpretation of the results may be difficult for studies that only include

seedlings and cuttings from a common checklot to be compared to the clonal propagules

from select parents and families. In this case, any differences in the field performance

because of propagule type may be confounded with the differences in genetic

improvement (Frampton and Foster 1993).

Clonally replicated progeny trials have been suggested as part of a tree

improvement strategy for radiata pine (Jayawickrama and Carson 2000) and for loblolly

pine (Foster and Shaw 1987; Isik et al. 2004; Byram et al. 2004) for a number of reasons.

First, field trials established with clonally replicated progeny allow for further

partitioning of the genetic variation into the additive, dominance, and epistatic genetic

variation (Foster and Shaw 1988). Second, clonally propagated seedlings can provide

genetic information more efficiently and with greater precision than zygotic seedling

progeny (Burdon and Shelbourne 1974; Isik et al. 2004). Finally, clonal testing and

selection strategies can provide greater gain than seedling options (Shaw and Hood 1985;

Mullin and Park 1994; Isik et al. 2004).

Based on current technologies, several forest industries in the southeastern United

States are pursuing clonal forestry programs with loblolly pine (Weber and Stelzer 2002).

In the initial stages of these clonal forestry programs, forest managers needed assurance

that the clonal propagules' growth corresponded to that of seedlings. Therefore, most of

the earlier studies were designed to test whether cuttings grew similarly to seedlings.









Based on those results, it is generally accepted that cuttings rooted from juvenile stock

plants grow and perform comparably to seedlings. For example, Foster et al. (1987)

reported that loblolly pine rooted cuttings should perform comparably to seedlings when

the cuttings come from vigorous juvenile stock plants. In addition, McRae et al. (1993)

concluded that for loblolly pine there were no significant differences between seedlings

and rooted cutting propagules from common checklots through five years of growth.

Similar results were obtained by Frampton et al. (2000) where they reported no

significant differences between the means of rooted cuttings and seedlings for height,

diameter at breast height, and volume through six years in the field.

Trials established with clones and seedlings from the same families provide an

opportunity for comparing both half-sib and full-sib family performances for both

propagules. Genetic correlations between propagule types can provide further assurance

that selections made through traditional tree improvement activities for recurrent

selection for general combining ability can also be used successfully in breeding families

to test in a clonal forestry program. Although a number of studies have been reported

comparing rooted cutting and seedlings, very few have been designed to estimate the

genetic correlation for a trait between propagule types.

Two main criteria need to be met prior to operational deployment of loblolly pine

clones. First, loblolly pine clones must perform well, e.g., meet the selection criteria for

the desired traits. This involves the accumulation of reliable data for the clones from

greenhouse screening, field trials, etc. Second, the selected clones have to be propagated

in large enough numbers for deployment. For a rooted cutting based clonal program, this

involves bulking up the number of hedges (ramets) of a particular clone or group of









selected clones through serial propagation and then producing reforestation stock

efficiently from the bulked-up clones. Only those clones that can be propagated easily

and in sufficient numbers will be economically feasible for deployment. A clone that

grows well in the field but roots poorly may not be economically feasible to include in a

clonal program based on rooted cutting technology.

This dissertation is unique in that a complex genetic structure was utilized in order

to increase the power in quantifying the genetic variation associated with several traits in

loblolly pine clones and seedlings. In Chapter 2 rooting ability was assessed for nearly

2,200 clones of loblolly pine from 70 full-sib families derived from a partial diallel

mating design in order to estimate genetic parameters associated with rooting. More than

1,200 of these clones along with zygotic seedlings from the same full-sib families were

established together on multiple sites across the southeastern United States. Genetic

parameter estimates are compared between propagule types for early growth traits in

Chapter 3. Finally, selection for both rooting ability and field growth is addressed in

Chapter 4.














CHAPTER 2
GENETIC EFFECTS OF ROOTING LOBLOLLY PINE STEM CUTTINGS FROM A
PARTIAL DIALLEL MATING DESIGN

Introduction

Loblolly pine is the most important commercial tree species in the southern United

States with over 1.1 billion seedlings planted annually (McKeand et al. 2003). There are

three cooperative tree improvement programs in the southern United States that focus on

improvement of southern pine species including loblolly pine: the Cooperative Forest

Genetics Research Program (CFGRP), North Carolina State University-Industry

Cooperative Tree Improvement Program (NCSUITIP), and the Western Gulf Forest Tree

Improvement Program (WGFTIP). Loblolly pine tree improvement programs in the

South are beginning their 3rd generation of breeding with gains in volume per unit area up

to 30% over unimproved loblolly pine (McKeand et al. 2003).

Long-term tree improvement programs aim to increase the population mean

breeding value of a few key traits through breeding and selection of superior genotypes.

These programs are based on recurrent selection for general combining ability which

captures only the additive portion of the genetic variance. In any given generation of

breeding, maximum gains can be achieved in the deployment population by capturing all

of the genetic variation (additive and nonadditive components) through operational

propagation of selected clones. With elite genotypes becoming available, several forest

industries in the southeastern United States are developing rooted cutting programs for









loblolly pine aiming towards deployment of tested clones or families (Weber and Stelzer

2002).

Clonal tests derived from full-sib families provide an opportunity to estimate

additive and nonadditive genetic components of variance associated with a particular trait

or set of traits (Isik et al. 2003; Isik et al. 2004). In clonal field trials traits of interest

may include height, volume, wood quality, and disease resistance. However, in order to

establish clonal field trials the clones must first be propagated. Therefore, clonal rooting

trials are important for estimating genetic variance components associated with rooting.

Previous rooting studies of loblolly pine have been relatively small in size ranging

from several hundred (Goldfarb et al. 1998; Foster et al. 2000) to several thousand

cuttings (Foster 1990; Anderson et al. 1999). Many studies contained a small number of

families from factorial mating designs (Goldfarb et al. 1998; Anderson et al. 1999;

Cooney and Goldfarb 1999; Frampton et al. 1999) and few have been designed to

estimate genetic parameters associated with rooting in loblolly pine (Foster 1978; Foster

1990; Anderson et al. 1999). The current study is unique in that a large number of

cuttings were set in each trial (> 34,000), and rooting was assessed on nearly 2,200 clones

of loblolly pine from 70 full-sib families derived from a partial diallel mating design in

order to estimate genetic parameters associated with rooting. The objectives of the study

were to (i) evaluate the rooting ability of stem cuttings from nearly 2,200 loblolly pine

clones, (ii) determine the causal components of variance in rooting of stem cuttings, (iii)

assess heritability estimates for rooting from five trials both on the observed binary scale

and the underlying normal scale, and (iv) determine the Type B genetic correlations for









both additive and dominance genetic effects to measure the correspondence in rooting

performance across five setting dates.

Materials and Methods

Population

The parental population for this study was selected from the Loblolly Pine Lower

Gulf Elite Population (LPLGEP) which consists of selections from all three southern pine

tree improvement cooperatives: CFGRP, NCSUITIP, and WGFTIP. Twenty 1st

generation and ten 2nd generation selections representing the Atlantic Coastal Plain,

Florida, and Lower Gulf provenances were selected from this population. Two additional

slow-growing parents were included to provide linkage with another study. These

parents were crossed in a circular diallel mating design (Appendix A) with some

additional off-diagonal crosses, resulting in a total of 70 full-sib families. On average

each parent was involved in approximately four crosses.

Seeds from the 70 families were sown in March 2000 into Ray Leach SuperCells

(Stuewe and Sons, Corvallis, OR). The seedlings were grown in a greenhouse at

International Paper Company's (IPC) facility in Jay, FL, and after three months of growth

the seedlings were pruned back to a height of about 10-12 cm. Approximately 32

seedling hedges (ortets) per full-sib family were transplanted into 3-gallon containers and

given unique clonal identifications in September 2000. The hedges were repeatedly

sheared in order to minimize the effects of maturation and increase the number of shoots

available for the rooting trials. The ortets were randomized in a containerized hedge

orchard in order to reduce spurious C effects at the family level. However, C effects at

the clonal level could not be accounted for, because all cuttings originated from a single

seedling ortet.









Table 2-1. Experimental design for 5 rooting trials of loblolly pine stem cuttings. All
trials were established in randomized complete block designs with 4 to 6
blocks and 4 to 9 ramets per clone in a row plot within each block.
No. of No. of
o No. of No. of Total
Tnal Date set plots per cuttings
Trial Date set families clones lotsper cuttings cuttings
clone per plot
SpringOl May 7- 70 2194 4 4 34,707
11, 2001
SummerO July 2-6, 70 2157 5 4 43,048
2001
Jan 14-
Winter02 Jan61 1648 6 5 49,315
18, 2002
Apr 29-
Spring02 May 3, 61 1254 6 9 67,059
2002
June 24-
Summer02 Ju 24- 61 947 6 9 45,108
28, 2002


Experimental Design

Two rooting trials were conducted in 2001 and three were conducted in 2002. Stem

cuttings between 3 and 8 cm in length were harvested from the seedling ortets in May

2001, July 2001, January 2002, April/May 2002, and June 2002, for trials SpringOl,

SummerOl, Winter02, Spring02, and Summer02, respectively. Cutting size was

relatively consistent within any trial, and the cuttings set in Winter02 were the smallest.

The experimental design differed among the trials due to number of families, clones and

available cuttings from each ortet (Table 2-1). The reduction in the number of families

and clones between the first and last rooting trials was a result of a number of factors.

First, random hedge mortality was a major factor: disease, repeated severe pruning, and

uneven watering (inadequate) all contributed to the random loss of hedges. Since

cuttings of a clone originated from a single ortet, by default mortality resulted in a

truncated population for future rooting trials. Second, not all hedges produced an

adequate number of cuttings at every harvest. The primary objective of the three settings









in 2002 was to produce propagules for field trials. Therefore, clones that were not

producing enough shoots to ultimately be planted across six field sites were culled

regardless of rooting frequency. This resulted in the reduction in the number of families

in the last three rooting trials because of too few surviving clones in some of the families

to meet the goal for field designs. We were striving for a balanced field design with 15

clones from each of 61 full-sib families.

Cuttings were randomly set in 4-, 5-, or 9-cutting clonal row plots (Table 2-1) into

pre-formed plugs consisting of peat moss, perlite, and a binding resin. Each plug was

approximately 13 mL in volume and was held by the V-13 HIKO tray (135 cells; Stuewe

and Sons, Corvallis, OR). Cuttings were either treated prior to setting with a 1.0%

indole-3-butyric acid and 0.5% napthalene-1-acetic acid (NAA) basal dip or after setting

using a foliar NAA application according to IPC protocols. Trays contained 15 to 30

clones depending on the trial and were randomly placed in an environmentally controlled

greenhouse. There were 4 to 6 complete replications depending on trial (Table 2-1).

Rooting assessments were made 9-weeks after setting for both the SpringOl and

SummerOl trials. Cuttings were measured for presence (1) or absence (0) of roots.

Cuttings with a root > 1 mm were considered rooted (Goldfarb et al. 1998, Foster et al.

2000). For trials Winter02, Spring02, and Summer02 assessments were at 11-weeks

following setting. Cuttings in these trials were also scored for presence or absence of

roots. However, only cuttings that had at least one visible root on the exterior of the plug

regardless of length were considered rooted.

Statistical Analyses

For binomial traits, such as rooting, the unit of analysis can be the individual

observations (Huber et al. 1994; Dieters et al. 1996) or plot means combined with a









transformation such as arcsin or logistic (Sohn and Goddard 1979; De Souza et al. 1991).

We chose to analyze the observed 0,1 data for several reasons. First, REML estimation

of variance components has been shown to be robust to violations of the underlying

normality assumptions (Banks et al. 1985; Westfall 1987) suggesting that analyses using

individual observations of binary data yields satisfactory results. Second, simulation

studies have shown that the use of individual observations is superior to the use of plot

means in REML, and that these variance component estimates perform well across

mating designs and imbalanced data (Huber et al. 1994). Huber et al. (1994) showed that

a lower variance among estimates was obtained using individual observations as

compared to plot means and that this advantage increased with increasing imbalance.

Third, when heritability is low and the incidence close to 50%, there is little difference

between heritability estimates on the binary and transformed scale (Dempster and Lerner

1950). In fact these two estimates are equivalent when the incidence is exactly 50% for

low heritability traits. Lopes et al. (2000) demonstrated that the Dempster and Lerner

(1950) threshold model closely estimates the true underlying heritability at incidences

between 25% and 75% for traits with low heritability (h2 < 0.3). Finally, Lopes et al.

(2000) also demonstrated (for traits with low heritability) that heritability estimates from

the observed binary data without transformation of data result in predicted gain close to

the realized gain, while transformations can suffer from issues of back transformation

when one wishes to predict gains on the original scale.

All variance components for rooting ability of loblolly pine stem cuttings were

estimated using the individual binary observations using REML estimation for each of

the 5-rooting trials using GAREML (Huber 1993). However, upwardly biased estimates









of genetic variances result when variance components are estimated from single-site

(trial) analyses since the estimated genetic variance also contains the genotype x

environment interaction (Comstock and Moll 1963). Therefore, across-trial analyses

were performed to separate the genotype x environment interactions in order to remove

this bias.

Yklmno =/ + + R) + trayk(j()) + gca1 + gcam + sCalm + c(fam)n(lm)
[2-1]
+ t gca', + t gca ,, + t scale + t c(fam),n(lm) + r famjn(,) + errorlklmno

where yzjklmno is the rooting response (0 or 1) of the 0th ramet of the nth clone within the

Imth full-sib family in the kth tray within thejth rep of the ith trial

,u is the population mean

T, is the fixed effect of trial

R,j( is the fixed effect of rep

tiv ,, ,, is the random variable tray (incomplete block) ~ IID(0, RA )

gcal and m is the random variable female (/) and male (m) general combining ability (gca)

^2
~IID(0, crGCA)

scalm is the random variable specific combining ability (sca) IID(0, ^SA )

c(fam)n(m) is the random variable clone within family ~ IID(0, ^ONE^ )

t'giI 'im is the random variable test by female gca and test by male gca interaction ~

IID(O, TESTxGCA)

t*scalm is the random variable test by full-sib family interaction ~ IID(O, ^ESTxFM )

t*c(fam),n(Im) is the random variable test by clone interaction IID(0, ESTxCLONE)

S"t,11n, is the random variable rep by family interaction ~ IID(0, ^rEPxFA









errorklmno is the random error which includes among plot and within plot ~

IID(O, OR OR).

The single trial model is identical except all model factors with subscript 'i' are removed

(sources involving test).

Genetic parameters were estimated and standard errors were calculated according

to Foster and Shaw (1988) using the appropriate variance components from the individual

or across trial model. Estimates of additive and dominance genetic variance are upwardly

biased because they are confounded with fractions of epistasis (Cockerham 1954).

Epistastic genetic variance is also only approximated because it contains only a fraction

of the total epistasis plus any C effects, if they exist.

[2-2] V, = 48 = + V +-V V +... = estimate of additive genetic variance

[2-3] VD = 4 VD +V + AD + DD +... = estimate of dominance genetic

variance

[2-4] VI =2 cLE-2 2 -3C2 = AA+ D DD + ... = estimate of epistatic

genetic variance

[2-5] V = 2 -cA + A +C L = estimate of total genetic variance


[2-6]

p 2 22 2 2 2 "2 "2 "2
S= 2G + SCA + CLONE + 2TESTxGCA +TESTxFAI TESTxCLONE -F + O REPxFAMI F O'ERROR

phenotypic variance for across trial model (the phenotypic variance from the individual

trial model is the same but drop 2ESTxGCA, TESTXFAM 2 and TESTXCLONE

Biased and unbiased heritability estimates for rooting based on observed 0,1 data

were derived using the estimated variance components from the single and across trial











models, respectively. In addition the proportion of dominance (d2) and epistasis (I2)

were estimated. Standard errors of these estimates were calculated using a Taylor series

expansion (Kendall and Stuart 1963; Namkoong 1979; Huber et al. 1992; Dieters 1994).

[2-7]


4&2
j2 4 GCA
0,1 ^2 2 2 2 ^ 2 2 ^2 ^2 ^2
2cGCA SCA CLONE + 2"TESTxGCA + (TESTxFAM + (TESTxCLONE + "REPxFAM -+ ERROR


across-trial narrow-sense hertibility based on observed binary data

[2-8]

^2 ^2 ^2
2 2 GCA + SCA +CLONE
0,1 2 2 ^2 ^2 2 ^2 + 2 + 2 + 2 ^2
2cGCA SCA + -CLONE + (7TESTxGCA (TESTxFAM (7TESTxCLONE (REPxFAM -V ERROR


across-trial broad sense heritability based on observed binary data

[2-9]

4 ^ 2
21 __ SCA
0,1 2 ^2 ^2 ^2 ^2 ^2 ^2 ^2
2cGCA SCA CLONE + TESTxGCA TESTxFAM TESTxCLONE + REPxFAM ERROR

across-trial dominance proportion

[2-10]

^2 ^2 ^2
'2 C_ CLONE 2 GCA SCA
0,1 22 2 2 2 2 "2 +2 "2
2a GCA + SCA CLONE + (7 TESTxGCA + (TESTxFAM (7TESTxCLONE REPxFAM ERROR


across-trial epistatic proportion.

The main problem with calculating heritability on the observed 0,1 data is that the


relationship between h2 on the observed scale and h2 on the underlying normal scale

depends on the mean incidence (e.g., % survival, % infected individuals, rooting

percentage), and therefore the conversion results in a biased estimate (Van Vleck 1972).

However, this is not a problem for low heritability traits with intermediate incidences










(Lopes et al. 2000). In order to make valid comparisons to heritability estimates in other

rooting trials that have different mean rooting percentages h0, needs to be transformed to


h2 on the underlying normal scale. Therefore, narrow- and broad-sense heritability

estimates on the observed 0,1-scale were transformed using a threshold model to an

underlying normal scale (Dempster and Lerner 1950).


[2-11] h2 ()( p), where
z

h2 is the heritability on the underlying normal scale

p is the rooting percent

z is the ordinate of the normal density function which corresponds to probability p.

Full-sib family mean heritability and clonal mean heritability for rooting were

estimated for both the single- and across-trial analyses. Standard errors for these

estimates were calculated using Dickerson's Method which assumes the phenotypic

variance (denominator) is a known constant (Dickerson 1969).

[2-12]

2 2&c +GCA + SCA
FS 22 2 2 2 2
22 ( 2 + 'CLONE 'TESTxGCA TESTxFAM TESTxCLONE REPxFAM ERROR
2 GCA + SCA + +
c t t tc tr ctrn
= across-trial family mean heritability, where, c = harmonic mean number of clones per

family, t = number of trials, r = harmonic mean number of reps per test, and n = harmonic

mean number of ramets per clone per plot.










[2-13]

22 + 2 +(2
2 GCA +SCA CLONE
CL 2 22 2 12 .2
2 (j 2 + 2 + S j2 | CO 'TESTxGCA + TESTxFAM UTESTxCLONE 'REPxFAM + ERROR
-GCA + USCA -CLONE t + + t
t t t tr trn

= across-trial clonal mean heritability.

Type B genetic correlations for rooting across all 5 trials were estimated for both

additive and nonadditive components (Yamada 1962; Burdon 1977). Standard errors of

Type B correlations were calculated using the Taylor Series Expansion method.

"2
[2-14] rB 2 g ,
g ge

where rB is the estimate of the Type B genetic correlation, !^2 is the genetic variance


component (either additive or dominance), and ^ 2 is the G x E interaction (additive and


dominance).

Results and Discussion

Average Rooting

A total of over 239,000 stem cuttings from nearly 2,200 clones of loblolly pine

were set in five rooting trials. Overall rooting across the five trials was 43% and is

comparable to other rooting studies involving loblolly pine. Goldfarb et al. (1998)

reported 44% rooting of loblolly pine stem cuttings from 400 seedling hedges from one

open-pollinated family. Over four rooting trials, Anderson et al. (1999) reported 33%

rooting from 90 clones of loblolly pine from 9 full-sib families. Foster (1990) reported

overall rooting from three settings of 42% for 546 clones of loblolly pine derived from 54

full-sib families. However, studies with fewer families of loblolly pine have yielded









substantially higher rooting percentages (Cooney and Goldfarb 1999; Murthy and

Goldfarb 2001; LeBude et al. 2004).

Rooting of loblolly pine cuttings was variable over the five trials (Table 2-2).

Spring cuttings rooted at greater than 50%, while summer cuttings in the two summer

trials averaged 38% and 24% respectively. Winter cuttings were intermediate at 45%

rooting. Broad inference linear contrasts were constructed using the estimates of the

fixed effects in order to test seasonal rooting responses. Cuttings set in the two spring

trials rooted at a significantly greater frequency then cuttings set in the summer trials (p <

0.0001). This implies that we would always expect a greater rooting percent in spring

settings than in summer settings under this propagation system.

Table 2-2. Summary of rooting from 5 loblolly pine trials set over two years and three
seasons.
Trial Rooting Half-sib Full-sib
Tnal r. ir. Clone range
% family range family range
SpringO1 54% 36-70% 28-77% 0-100%

SummerO1 38% 24-54% 18-69% 0-100%

Winter02 45% 17-60% 17-67% 0-100%

Spring02 51% 36-67% 28-75% 0-98%

Summer02 24% 9-43% 8-53% 0-89%



Seasonal rooting responses for loblolly pine stem cuttings have been observed in

other studies. Early rooting trials of loblolly pine cuttings reported best rooting from

cuttings set from September through January (Cech 1958; Reines and Bamping 1960;

Grigsby 1962; Marino 1982). These early experiments concluded that increased

temperatures in the greenhouse in spring and summer trials decreased rooting. In fact,









Cech (1958) reported a 3-fold increase in rooting under cool conditions rather than under

warm conditions. However, Foster et al. (2000) observed an overall rooting of 50% for a

rooting trial established in March, while only 20% rooting for a trial established in

September. They hypothesized that the reduction in September rooting was due to a

decrease in metabolic activity due to the decrease in photoperiod. Rowe et al. (2002a;

2002b) reported trends in rooting similar to those of the current study. They observed

59% rooting for spring cuttings versus 40% rooting for winter cuttings and 35% rooting

for summer cuttings. Cooney and Goldfarb (1999) also reported high rooting percentages

for spring cuttings (62% and 83% in two successive years). In contrast, Murthy and

Goldfarb (2001) reported higher rooting percentages for winter cuttings (85%) than for

spring cuttings (60%). Winter cuttings often take longer to root but overall rooting may

not be different than spring cuttings. Perhaps the slight reduction in rooting frequency

for the Winter02 setting versus the two spring settings was a function of rate of rooting.

The reduction in rooting seen here for summer settings may be a result of increased

temperatures during the collection and propagation phases of the experiment. The higher

temperatures and humidity experienced during summer months may have resulted in an

increased abundance or activity of pathogens, and hence a higher rate of decay was

observed in the two summer trials. The time delay between collection and setting of

cuttings may also have contributed to this reduction in rooting. Murthy and Goldfarb

(2001) reported a decline in rooting percentage with increasing drying time of cuttings.

Observed Variance Components

Variance components were estimated for all single- and across-trial analyses

(Appendix B). Even with the reduction in the number of clones and families throughout

the study, there was no apparent reduction in the variance component estimates over time









as evidenced by parameter estimates, e.g., additive genetic variance estimates were

relatively constant over time (Figure 2-1). The variance associated with general

combining ability was 2% of the total phenotypic variation associated with rooting. Half-

sib family rooting percentages ranged from a high of 36-70% for SpringO1 to a low of 9-

43% for Summer02 (Table 2-2). The proportion of the total variation in rooting that was

accounted for by specific combining ability was only 0.3-1.1%. Full-sib family means

for rooting for the two spring settings ranged from 28-76%. Although overall rooting

was greater for these two spring trials, the net difference in the range of family means

was approximately the same (-45-51%). Similar ranges in family mean rooting

percentages have been reported (Foster 1990; Anderson et al. 1999).



0.25

0.2 T I-lo2

0.15 -
H 0,1 0
0.1 -

0.05 1


SpringOl SummerOl Winter02 Spring02 Summer02 All
Trial


Figure 2-1. The proportion of the additive (0 ), dominance (d, ), and epistasis (1 )
genetic variances on the observed binary scale for rooting of loblolly pine
stem cuttings from each of the five separate trials (biased heritabilities) and
from the combined analysis of all five trials. Standard error bars for broad-
sense heritability estimates are included.









Clones within families accounted for approximately 10-17% of the total variation

in rooting in the five trials (Appendix B). Foster (1990) reported that the clone within

family source of variation in rooting of 546 clones of loblolly pine was 3.7 % of the total

variation. However, in another study involving cuttings of loblolly pine, the among clone

source of variation accounted for nearly 22% of the total variation (Foster 1978). In the

current study, the variance associated with clone within family was 4.5-8.4 times greater

than the gca variance and 13-56 times greater than the sea variance based on single-trial

analyses.

Just as there was a large range in rooting among families, there was also a large

range in rooting among clones within families (Table 2-2). Rooting for clones within

families ranged from 0-100% for the first three trials, and ranged from 0-98% and 0-89%

in the last two trials, respectively. Anderson et al. (1999) observed ranges in rooting

frequency for clones within family similar to those observed in this study. Foster (1990)

reported significant variation for clones within family with rooting percentages ranging

from 6.7-85.0%.

A large variation in the rooting environment has been reported in many rooting

experiments of loblolly pine and other species, with an error variance ranging from 46-

71% (Foster 1978; Foster 1990; Sorensen and Campbell 1980; Cunningham 1986). In

the current study, the majority of the observed variance in rooting was attributable to the

error variance. The error variance accounted for 76.3-83.7% of the total variance

observed in the five trials. Each rooting trial was spread over an entire greenhouse due to

the large size of the experiments. Apparently, the rooting environment was not uniform

throughout the greenhouse. There are many factors that can contribute to a variable









rooting environment. Differential temperature gradients, unequal airflow, unequal

misting of cuttings, edge effects, and disease incidence can all contribute to a variable

rooting environment.

Causal Variance Components

The observed variance components were used to estimate additive, dominance and

epistatic variances using equations 2-2, 2-3, and 2-4, respectively. Additive genetic

variance was approximately 2 to 6 times larger than the dominance genetic variation for

rooting in the five trials (Figure 2-1). In Foster (1990), the additive genetic variance was

also about 6 times larger than the dominance genetic variance. However, Anderson et al.

(1999) found three times and Foster (1978) found 2.2 times greater dominance genetic

variance in rooting compared to the additive variance in loblolly pine stem cuttings. The

epistatic genetic variance was also estimated in the current study and was approximately

0.44 to 1.49 times as much as the additive genetic variance. The lowest amount of

epistasis was observed for the winter setting and was the only setting that had a

nonadditive to additive ratio less than one.

C effects can lead to upwardly biased estimates of total genetic and nonadditive

genetic components of variance when analyzing clonal data (Libby and Jund 1962). If C

effects are present, then total genetic variation associated with clones will be

overestimated (Libby and Jund 1962). Significant C effects are likely to occur in traits

that are measured soon after propagation (such as rooting traits, early shoot growth, etc.),

but apparently lessen for traits measured at later times (Libby and Jund 1962). Foster et

al. (1984) used a secondary cloning approach to separate C effects from the genetic

variance in rooting of western hemlock cuttings. They found significant C effects

associated with rooting and that these non-genetic effects biased the genotypic values of









clones. However, low or non-significant C effects for rooting of tamarack and balsam

poplar cuttings have been reported (Farmer et al. 1989; Farmer et al. 1992). In the

current study, estimates of epistatic genetic variance components are confounded with C

effects, because ramets of a clone came from the original seedling ortet. However,

estimates of additive and dominance genetic effects are not confounded with C effects,

because the clones within families were randomized in the hedge orchard.

Heritability Estimates

Rooting of loblolly pine stem cuttings was weakly controlled by additive effects.

Individual tree narrow-sense heritability using the observed variance components (h1 )

ranged from 0.075 to 0.089 in the five separate trials (Figure 2-1). These estimates are in

agreement with Foster (1978) who reported h2 as 0.07 for rooting percentage of loblolly

pine. A slightly higher h2 was reported by Foster (1990). However, Anderson et al.

(1999) reported h2 of 0.26 for rooting percentage in loblolly pine stem cuttings.

However, all of these previous studies of loblolly pine rooted cuttings analyzed data

based on plot rooting percentage, as opposed to using 0,1 data as in this study, and this

leads to an increased estimate of heritability due to the reduction of the impact of the

within plot portion of the error variance in any plot means or plot percentage analysis for

rooting. When the data in the current study were analyzed based on rooting percentage,

2 was 0.18 (from SpringO1, data not shown), which falls in the middle of the range

previously reported.

The unbiased estimates of individual tree narrow-sense heritability (h ) using the

0,1 data from the pairwise test analyses ranged from 0.045 to 0.074. When all of the data









from the 5 trials were analyzed together, hk, was 0.051 (Table 2-3; Figure 2-1). The

proportion of dominance (d2 ) was estimated for each of the trials. The upwardly-biased

estimates for this parameter ranged from 0.014 to 0.044 (Figure 2-1). When all of the

data were analyzed together from the 5 trials, d2 was 0.018 (Table 2-3; Figure 2-1). The

epistatic proportion (i, ) was also estimated and ranged from 0.032 to 0.095 (Table 2-3;

Figure 2-1).

Table 2-3. Genetic parameter estimates (standard error) for rooting of loblolly pine stem
cuttings across 5 trials.
Parameter Estimate
Narrow-sense heritability on observed binary scale (h, ) 0.051 (0.017)
Broad-sense heritability on observed binary scale (H0, ) 0.101 (0.008)
Narrow-sense heritability on the underlying normal scale
1(2) 0.08 (0.027)
(h )
Broad-sense heritability on the underlying normal scale
(/2) 0.16 (0.013)

Narrow-sense family mean heritability (!H ) 0.833 (0.24)
Broad-sense clonal mean heritability (HE) 0.815 (0.074)
Additive genetic variance (VA) 0.0117 (0.004)
Dominance genetic variance (VD) 0.0042 (0.002)
Epistatic genetic variance (V) 0.0074 (0.002)
Phenotypic variance (Vp ) 0.2297 (0.002)
Type B additive genetic variance correlation 0.68 (0.23)
Type B dominance genetic variance correlation 0.61 (0.27)
Type B total genetic variance correlation 0.53 (0.048)

Broad-sense heritability (H,12 ) ranged from approximately 0.15 to 0.22 (Figure 2-

1). Broad-sense heritability was 0.101 when all of the data from the five trials were

combined (Table 2-3; Figure 2-1). Foster (1990) observed very little nonadditive genetic









variance and reported H 2 of 0.13 for rooting percentage. In another study H' 2 was

reported as 0.23 for rooting percentage (Foster 1978). However, Anderson et al. (1999)

reported a much higher 2 (0.63) for rooting percentage of loblolly pine. When the data

in this study were analyzed based on rooting percentage, then H2 was 0.47.

Narrow- and broad-sense heritability estimates on the observed 0,1-scale were

transformed to an underlying normal scale assuming a threshold model (Equation 2-11).

Narrow-sense heritability based on the underlying normal scale (h2 ) ranged from 0.12 to

0.16 (Figure 2-2). Broad-sense heritability based on the underlying normal scale (H )

ranged from 0.24 to 0.36 (Figure 2-2). When all of the data from the five settings were

analyzed together, then h2 was 0.08 and H2 was 0.16 (Table 2-3; Figure 2-2). The

transformation of h0 and Hj, to the underlying normal scale allows direct comparisons

of heritability estimates among rooting studies when the rooting percentages are different.

Family mean heritability ranged from 0.84 to 0.9, while clonal mean heritability

ranged from 0.82 to 0.92 (Figure 2-3). When all of the data from the five trials were

combined then H2 was 0.83 and H2 was 0.82 (Table 2-3; Figure 2-3). Foster (1990)
FS CL

reported lower estimates of both family-mean heritability and clonal-mean heritability for

a rooting study consisting of 540 clones of loblolly pine from 54 full-sib families, 0.46

and 0.40, respectively. In another loblolly pine rooting study with 27 full-sib families

consisting of 10 clones each, family-mean heritability was reported as 0.31, while the

clonal-mean heritability was 0.87 (Anderson et al. 1999).

















42N
AT


0.4

0.35

0.3

o 0.25
0O
-o 0.2
E
o 0.15
Z
0.1

0.05

0


SpringOl SummerOl Winter02 Spring02 Summer02 All

Trial


Figure 2-2. Narrow-sense (1) and broad-sense (/f ) heritability estimates for rooting
of loblolly pine stem cuttings transformed to the underlying normal scale
using the threshold model of Equation 2-11. Standard error bars are included.



1.2


a,

E 0.6
o,
LJ 0.4

0.2

0


SpringOl SummerOl Winter02 Spring02 Summer02
Trial


FS

CL


Figure 2-3. Full-sib family mean ( ) and clonal mean (H ) heritability estimates for
rooting success of 2,200 clones from 70 full-sib families of loblolly pine.
Standard error bars are included.









Type B Genetic Correlations

The type B correlations for additive effects were moderately high and ranged from

0.53 to 0.91 (Table 2-4) indicating that parental rankings of the 32 parents were

moderately to strongly correlated among pairs of trials. When all of the data were

analyzed together from the five trials, the type B additive correlation was 0.68. The

highest type B correlation was observed between the two spring settings which were one

year apart, while the lowest correlations were observed between the winter setting

(Winter02) and other trials. Spring and summer cuttings were, in general, actively

growing, succulent material, while winter cuttings were generally smaller and more

lignified. Perhaps, some of the genes controlling rooting of dormant winter cuttings are

different than those controlling spring and summer cuttings.

Table 2-4. Type B additive and dominance variance correlations among pairs of rooting
trials for loblolly pine stem cuttings (above and below diagonal, respectively).
Standard errors are given in parentheses.

SpringO1 SummerO 1 Winter02 Spring02 Summer02

SpringO1 0.701 (0.11) 0.59(0.15) 0.911 (0.07) 0.851 (0.08)


SummerOl 0.823 (0.19) 0.634 (0.12) 0.673 (0.12) 0.634 (0.13)


Winter02 0.192 (0.30) 0.629 (0.34) 0.656 (0.12) 0.525 (0.16)


Spring02 0.493 (0.24) 0.533 (0.35) 0.724 (0.24) 0.907 (0.06)


Summer02 0.777 (0.22) 0.973 (0.37) 0 0.703 (0.24)



Estimated type B correlations for dominance effects measure the correspondence of

dominance across pairs of trials and were estimated to be moderate to high (Table 2-4)









with two exceptions. First, the type B dominance correlation between the SpringOl and

Winter02 trials was 0.192. Also, no dominance genetic variance was detected in the

pairwise analysis of theWinter02 and Summer02 trials. However, when all of the data

were analyzed together, the type B dominance correlation was 0.61.

Selection for Rooting

Before clones of loblolly pine can be deployed operationally, two things must

occur. First, the clones have to be field-tested and selected for desirable traits, e.g.,

growth, disease resistance, and wood properties. Second, the selected clones have to be

be propagated in large enough numbers for deployment. For a rooted cutting-based

program, this involves bulking up the number of hedges of a particular clone or group of

clones through serial propagation and then producing reforestation stock efficiently from

the bulked-up clones. Only those clones that can be propagated easily will be

economically feasible for deployment. Therefore, selection of clones for rootability as

well as field performance should be considered as part of a clonal forestry program based

on rooted cutting technology (Foster et al. 1985; Foster et al. 2000). In the current

study, selection of the top 10% of clones for rooting (-220 clones) would result in a gain

of about 24% in rooting, where

Gain =
I 22 2 2 ^2 ^2 ^2 .2
i 2 2 j 2 2 2"TESTxGCA 'TESTxSCA + TESTxCLONE 'REPxFAM F ERROR
iL ^GCA SCA +CLONE
L t t t tr trn

(Foster 1990).

Selecting the top 1% of clones for rooting would result in a gain of nearly 37% in

rooting of loblolly pine stem cuttings in the current generation. Alternatively, genetic










gain in rooting success in the next generation can be achieved through selection and

breeding.

Gain =
2 2^2 ^2 ^2 ^2 ^2



(j 2
^f 2 G +2 2 + -CLONE 2s-- --xGCA TESTxSCA TESTxCL+ONE -REPxFA+M + RROR



+ il 2 ^CA&_
i, 2p 2r mO
t /2 ^2 ^2
CL2 TESTxCLONE REPxFAM RROR
V/JCLONE + +
t tr trn


By selecting the best rooting clone from the top 25 out of 70 families, gain in

rooting success of 16.8% can be expected in the next generation by breeding these 25

selections.

Conclusion

Loblolly pine is the most important commercial tree species in the southeastern

United States. Several forest industry companies are developing rooted cutting programs

for loblolly pine in order to maximize genetic gains through deployment of tested clones.

With rooting data from 2,200 clones from 70 full-sib families, the current study gives

better estimates of genetic components of variance for rooting than several previous

studies. These results show a great deal of genetic variation for rooting among families

and clones of loblolly pine. Only those clones that can be propagated easily will be

economically feasible for deployment. Therefore, selection of clones for rooting ability,

as well as field performance should be considered as part of a clonal forestry program

based on rooted cutting technology. Combined with moderate to high estimates of

family- and clonal-mean heritabilities and type B correlations, these results indicate the






29


potential for increasing rooting efficiency by selecting good rooting families and clones

or culling poor rooters.














CHAPTER 3
GENETIC ANALYSIS OF EARLY FIELD GROWTH OF LOBLOLLY PINE CLONES
AND SEEDLINGS FROM THE SAME FULL-SB FAMILIES

Introduction

Loblolly pine (Pinus taeda L.) is the most important commercial tree species in the

United States with over one billion seedlings planted annually (McKeand et al. 2003).

Most commercially important tree species remain relatively undomesticated, and loblolly

pine tree improvement programs are only now beginning their 3rd generation of breeding

and testing (McKeand and Bridgewater 1998). Tree improvement programs for loblolly

pine have relied on recurrent selection for general combining ability for improvement of

a few key traits. These programs have historically utilized seedling progeny trials in

order to predict breeding values for these traits. Traditional tree improvement programs

using open-pollinated seed orchard seedlings for deployment only capture additive

genetic variation. However, nonadditive genetic variation may be an important

component of variation for some traits, and additive and nonadditive genetic variation

can be captured by deploying full-sib families or clones.

More efficient field-testing has been implemented in order to gain information on

full-sib families. For example, the Cooperative Forest Genetics Research Program at the

University of Florida is using a partial diallel mating design to cross slash pine selections

to generate full-sib seedlings for progeny trials (Gezan et al. 2004). Tests established

with full-sib seedlings allow the genetic variance to be partitioned into additive and

dominance components (Falconer and Mackay 1996). These trials not only provide ranks









of parents or individuals for selection, but also of full-sib families in order to provide

information for making deployment decisions.

For a number of reasons, clonally replicated progeny trials have been suggested as

part of a tree improvement strategy for radiata pine (Jayawickrama and Carson 2000) and

for loblolly pine (Foster and Shaw 1987; Isik et al. 2004; Byram et al. 2004). First, field

trials established with clonally replicated progeny allow for further partitioning of the

genetic variation into the additive, dominance, and epistatic genetic components (Foster

and Shaw 1988). Second, clonally propagated seedlings can provide genetic information

more efficiently and with greater precision than zygotic seedling progeny (Burdon and

Shelbourne 1974; Isik et al. 2004). Finally, clonal testing and selection strategies can

provide greater gain for operational deployment than seedling options (Shaw and Hood

1985; Mullin and Park 1994; Isik et al. 2004).

Based on current technologies, several forest industries in the southeastern United

States are pursuing clonal forestry programs with loblolly pine (Weber and Stelzer 2002).

In the initial stages of these clonal forestry programs, forest managers needed assurance

that the clonal propagules' growth corresponded to that of seedlings. Therefore, earlier

studies were designed to test whether cuttings grew similarly to seedlings. Based on

those results, it is generally accepted that cuttings rooted from juvenile stock plants grow

and perform comparably to seedlings. For example, Foster et al. (1987) reported that

loblolly pine rooted cuttings should perform comparably to seedlings when the cuttings

come from vigorous juvenile stock plants. In addition, McRae et al. (1993) concluded

that for loblolly pine there were no significant differences between seedlings and rooted

cutting propagules from common checklots through five years of growth. Similar results









were obtained by Frampton et al. (2000) where they reported no significant differences

between the means of rooted cuttings and seedlings for height, diameter at breast height,

and volume through six years in the field.

Trials established with clones and seedlings from the same families provide an

opportunity for comparing both half-sib and full-sib family performances across

propagule types. Genetic correlations between propagule types can provide further

assurance that selections made through traditional tree improvement activities for

recurrent selection for general combining ability in seedling tests can also be used

successfully in breeding families to test in a clonal forestry program. Although a number

of studies have compared rooted cutting and seedlings, very few have been designed to

estimate the genetic correlation between propagule types for a trait.

While clonal tests derived from full-sib families provide an opportunity to estimate

additive and nonadditive components of variance, tests should be designed with sufficient

genetic structure in order to precisely quantify the genetic variation. Several clonal

studies have reported deficiencies in mating designs, number of parents and families

(Frampton and Huber 1995; Paul et al. 1997; Isik et al. 2003). For example, Frampton

and Huber (1995) reported that they had low power in partitioning the genetic variation

because of the lack of a mating design among the parents of the full-sib crosses in a

loblolly pine clonal study. In addition, Frampton and Foster (1993) warned that

interpretation of the results may be difficult for studies that only include seedlings and

cuttings from a common checklot to be compared to the clonal propagules from select

parents and families. In this case, any differences in the field performance because of









propagule type may be confounded with the differences in genetic improvement

(Frampton and Foster 1993).

The current study employs a complex genetic structure to increase the power in

quantifying the genetic variation associated with several growth traits in loblolly pine.

More than 1,200 clones together with zygotic seedlings from the same 61 full-sib families

were tested on multiple sites across the southeastern United States. Because of the test

and mating designs, genetic correlations can be directly calculated between propagule

types and within propagule types across sites. The specific objectives of this study are to

1) determine heritability estimates for various growth traits for loblolly pine clones and

seedlings, 2) compare the performance between parents and full-sib families when grown

as rooted cuttings and seedlings, and 3) determine the extent of genotype x environment

interaction by looking at the genetic correlations for parents, families, and clones across

multiple sites.

Materials and Methods

Population

The parental population consisted of twenty first-generation and ten second-

generation selections, subset from the Loblolly Pine Lower Gulf Elite Population. Two

additional first-generation, slow-growing parents were included. The parental selections

represent the Atlantic Coastal Plain, Florida, and Lower Gulf provenances of loblolly

pine (see FBRC 2000 for details). Briefly, these thirty-two loblolly pine parents were

mated in a partial diallel design and created 70 full-sib families from which more than

2,000 seedling hedges were generated and given unique clonal identifications (Appendix

A). On average, each parent was involved in approximately four crosses.









Propagation

The propagation of the rooted cuttings for the field trials has been previously

described (Chapter 2; Baltunis et al. 2005). But briefly, the seedling hedges were

repeatedly sheared to slow down the effects of maturation and increase the number of

shoots available for collection. Cuttings were collected from seedling hedges from 61

full-sib families, placed randomly in clonal-row plots, and replicated six times in a

greenhouse in January 2002, April 2002, and June 2002. At the time of collection, the

hedges were 22, 25, and 27 months old from seed, respectively. Cuttings were assessed

for rooting at 11-weeks after setting (Baltunis et al. 2005). Rooted cuttings were then

transplanted into Ray Leach Supercells (Steuwe and Sons, Corvalis, Oregon) and

randomized into their designated field planting order and grown to size. The clonal

propagules for a field trial came from a single rooting trial (Table 3-1).

Table 3-1. Location of six field trials, establishment date and total number of test trees
for each test.

Test Sticking Location Latitude Date Total No.
Test Location
Date Longitude Planted Test Trees

A January Worth County, 31'44'20"N October 9,216
2002 Georgia 83'55'50"W 2002

B April Morgan County, 3324'55"N November 8,960
2002 Georgia 83'29'45"W 2002

April Putnam County, 2938'N November
2002 Florida 8146'W 2002

D April Nassau County, 30'45'23"N February 8,960
2002 Florida 81'54'27"W 2003

E April Randolph County, 3148'03"N December 4,400
2002 Georgia 84'41'30"W 2002

F June Santa Rosa County, 30'50'05"N April
2002 Florida 87'11'57"W 2003









A single crop of seedlings, on the other hand, was used to produce all the seedlings

for the field trials. Loblolly pine seed from the same full-sib families that the clones were

derived from were stratified for about 30 days and then sown in May 2002. The

seedlings were grown in family blocks in a different greenhouse than the cuttings. The

seedlings were then moved outdoors under shade cloth and kept in their family blocks in

a separate area from where the rooted cuttings were growing. The seedlings were not

randomized into their designated field order until just prior to planting.

Field Design

In total 47,408 measurement trees were established at six field sites across the

southeastern United States (Table 3-1). Three trials each were established in Florida and

Georgia (Table 3-1). An additional field trial was established in Virginia with a subset of

the clones but was not included in any of these analyses. There were four replications in

each of two cultural treatments (high and low intensity) in each test, except for Test E

where there was only one cultural treatment and four replications. The goal for the high

intensity treatment was to push the trees to their utmost potential by reducing competition

and providing a non-limiting supply of nutrients, while the low intensive culture provides

insights into family and clonal performance under a less optimal cultural regime (FBRC

2000). Both cultural intensities were treated similarly during the first year with cultural

differences implemented at the beginning of the second growing season.

Each trial contained 756-974 clones with approximately 15 clones from each of 61

full-sib families (Table 3-2). In total, more than 1,200 clones were planted in field trials

(Table 3-2). The trials were designed to have four zygotic seedlings from each of the

same 61 full-sib families within each replication. However, because of poor germination

for some of the families or mortality in the nursery, each full-sib family is represented, on









average, by approximately 27 zygotic seedlings per test (Table 3-2). Both the rooted

cuttings and seedlings were planted in single-tree plots utilizing a resolvable alpha

incomplete block design (Williams et al. 2002) in which incomplete block size ranged

from 10-14 trees. The variables measured were 1st year height, 2nd year height, height

increment, and crown width.

Table 3-2. Total number of clones, full-sib families, half-sib families, average number of
clones per full-sib and half-sib family, and average number of seedlings per
full-sib and half-sib family established at the six field trials.

Test A Test B Test C Test D Test E Test F Total


Total # clones 974 941 942 956 868 756 1,212


Total # FS families 61 61 61 61 61 61a 61


Total # HS families 32 32 32 32 32 32b 32

Ave. # clones/FS
.# cl 16 15.4 15.4 15.7 14.2 12.4 19.9
family

Ave. # clones/HS
Ave. # conesHS 60.9 58.8 58.9 59.5 54.2 47.2 75.7
family

Ave. # seedlings/
Ave. # seedlings/ 27.9 27 27 27 15.2 35.1 151.2
FS family

Ave. # seedlings/
Ave. # seedlings/ 106.5 103 103 103 58 106.3 576.5
HS family
a Only 47 full-sib families and b 31 half-sib families are represented in the seedling
population of Test F.


Statistical Analyses

All growth variables, 1st year height, 2nd year height, height increment, and crown

width, were analyzed in ASREML (Gilmour et al. 2002) using a parental model in order

to estimate the genetic variance components associated with those traits. Analyses were

conducted for each trial separately and across all six trials for each propagule type. The









across-trial analyses assumed a different error variance for each trial. For the clonal

population, the following across-trial model was used.

[3-1]

Yjklmnop = + + T C + Rk() + incbkl(Qk) + gcam + gcan + scamn + cloneo(mn)
+ t gcat + t gca,, + t fammn + t* cloneo(mn) + t c gc + t c gca ,, + t c fammn
+t* c coIe + r g / )m +r* g )+ r famk()mn + ekmnop



where,

yjkmnop is the measured growth trait of the pth ramet of the oth clone within the mnth

full-sib family in the lth incomplete block within the kth replication of thejth cultural

treatment in the ith test

/u is the clonal population mean

T, is the fixed effect of trial, i = 1,...,6

T C, is the fixed effect of the interaction between trial and culture, j = 1,2


Rk(U) is the fixed effect of replication, k = 1,2,3,4

incbkl(,k) is the random variable incomplete block associated with each test NIID(0,

2,
INC,

gcam and gcaa are the random variables female (m) and male (n) general combining

ability, respectively NIID(0, G^CA


scamn is the random variable specific combining ability ~ NIID(0, !CA)

cloneo(mn) is the random variable clone within full-sib family ~ NIID(0, ^LONE









t ge t,, and t gca ., are the random variables test by female general combining ability

and test by male general combining ability interactions, respectively NIID(0, 2ESTxG )

t famm,. is the random variable test by full-sib family interaction NIID(0, &~ESTxFM

t chie ,, is the random variable test by clone within full-sib family interaction ~

NIID(0, cr 2)
NIID(O, TESTxCLONE

t c gca ,, and t c gca ., are the random variables test by culture by female general

combining ability and test by culture by male general combining ability, respectively ~

NIID(0, TXGC4)
NtC ,TxCxGCA)

t c famj, is the random variable test by culture by full-sib family ~ NIID(0, & CF

t c clone is the random variable test by culture by clone within full-sib family

interaction ~ NIID(0, &^ TCLNE

r gII .)m and r gII ), are the random variables replication by female general

combining ability and replication by male general combining ability, respectively ~

NIID(0, REPxGCA)

r fak(,,,,, is the random variable replication by full-sib family ~ NIID(0, 62EPxiMI)

ekm,,op is the random error associated with each test ~ NIID(0, 'ER )OR.

The single-site model for the clonal population is identical, except that all model

factors with subscript i are removed (sources involving test).

All traits were also analyzed for the seedling population assuming a randomized

complete block design (incomplete block dropped from model). Both single-trial









analyses and an across-trial analysis were performed again assuming heterogeneous

errors across sites.

[3-2]

zkimn = P + Rk() +gc +gc + T sea,, + ga +t gca c,, + t* faml,
+t*c*g I +t*c* gca,, +t*c* fam;i+m +r*gL )L +r* g ./ ) +r* famk(U)lm
+ eijklmn

where

ZUklmn is the measured growth trait of the nth seedling from the Imth full-sib family in the

kth replication of thejth cultural treatment in the ith test

/j is the seedling population mean and the other variables are defined as above (just

adjusting the appropriate subscripts).

Genetic parameters were estimated and standard errors were calculated according

to Foster and Shaw (1988) using the variance components from the appropriate model.

2= 1 1
4 16

is the estimate of additive genetic variance.
2[3-4] 4 1 1 1
[3-4] D = 4 = + AA + V-A + D-VDD
2 2 4

is the estimate of dominance genetic variance.

S= (2 (2 2 1 1 3
[3-5] V CLONE -2, GCA 0SCA = VAA + AD DD ..


is the estimate of epistatic genetic variance for the clonal population.
[3-6] = 2&c2 +s2 + o
[3-6] VG = GCA +JCSCA CLONEE

is the estimate of total genetic variance for the clonal population.










[3-7] = 4'!2 + 4A 2
GCA SCA

is the estimate of total genetic variance for the seedling population (assuming no

epistasis).

P 282 2 2 2 2 2 2
S = 2c GCA SCA + CLONE + 2 TESTxGCA TESTxFAM + TESTxCLONE + 2TxCxGCA
6
[3-8] y'-2
[3-8] ERROR,
^2 + ^2 + ^2 ^2 1=1
STxCxFAM 'TxCxCLONE + 2REPxGCA 'REPxFAM +
6

is the estimated phenotypic variance for the across-trial model for the clonal population.

P j ^2 ^ 2 ^ 2 12 ^ 2 12 ^ 2
S= 2 GC S +SCA 2 TESTxGCA + 7TESTxFAM + 2- TxCxGCA + -TxCxFAMI + 2-REPxGCA
6
[3-9] -- -2
[39] ERROR,
"2 1-i
+ -'REPxFAM + 6
6

is the estimated phenotypic variance for the across-trial model for the seedling

population.


Individual tree narrow-sense heritability (h2 ) and broad-sense heritability (H )

were derived using the estimated variance components for all the growth traits and each

propagule type at each site and across sites. Standard errors were calculated using Taylor

series expansion (Kendall and Stuart 1963; Namkoong 1979; Huber et al. 1992; Dieters

1994). The following heritability formulae were used for the clonal data.


[3-10] h2 = GCA
V V
P PC

is the across-trial estimate of individual tree narrow-sense heritability for the clonal


population, where Vp is from [3-8].


V (22 +'2 + '2
[3-111 H =V2 G GCA + CSCA + CCLONE
3- 77-









is the across-trial estimate of individual tree broad-sense heritability for the clonal

population, where V is from [3-8].

Heritability estimates were also obtained for the seedling data for each trial and

across-trials.


[3-12] h2 =A ~~C
PS PS

is the across-trial estimate of individual tree narrow-sense heritability for the seedling

population, where V is from [3-9].


[3-13] H2_ GCA + 2CA
VPS VPS

is the across-trial estimate of individual tree broad-sense heritability for the seedling

population, where Vp is from [3-9].

The various growth variables at each site were also analyzed with a bivariate mixed

model with the growth of the clones and seedlings as two dependent variables. Type B

genetic correlations for general combining ability (Br ) and full-sib family value


(BpropF ) between cuttings and seedlings were estimated in order to compare parental and

full-sib family performance between propagule types. The genetic correlation between

propagule types for additive effects, for example, gives us an indication of whether

parental ranks are dependent upon whether their progeny are grown as cuttings or

seedlings, while a type B genetic correlation at the full-sib family level measures the

performance of full-sib families across propagule types.










[3-14] Cov(GCAcrGCASED )
r-" )Var(GCAcu ) x Var (GCASEED)


is the type B genetic correlation of additive effects between propagule types using genetic

variance component estimates from the bivariate analysis.

2Cov(GCAcur GCAED) + Cov(SCAc SCARED)
[3-15] r
)(2Var(GCAcu ) + Var(SCAcu )) x (2Var(GCAED ) + Var(SCAEED))


is the type B genetic correlation of full-sib family values between propagule types using

genetic variance component estimates from the bivariate analysis.

The extent of genotype x environment interaction was investigated by analyzing

data across trials for each propagule type using the variance components from the

appropriate model. Type B genetic correlations were calculated for additive effects, full-

sib family, and the total genetic or clonal value across the trials (Yamada 1962; Burdon

1977). Standard errors of type B correlations were calculated using the Taylor series

expansion method.

^2
O~-zc
[3-16] B = 2 GCA
GCA "2 +"2
S GCA TESTxGCA

is the type B genetic correlation for additive effects across trials.

^2 ^ 2
[3-17] 'rB =~2 2 2 2
[3-17] ^ 2GCA SCA
GCA TESTxGCA SCA TESTxFAM

is the type B genetic correlation for full-sib families across trials.

j 2 '^2 ^2
[3-18] 2 GCA + SCA + CLONE
S^2 ^2 2 ^2 ^2 ^2
2-GCA + 2-TESTxGCA + -SCA + TESTxFArM + 'CLONE + UTESTxCLONE

is the type B genetic correlation for total genetic or clonal value across trials for the

clonal population.









Ranging from 0 to 1, a value of i- near one indicates little genotype x

environment interaction and that the parents ranked the same across the trials, while a

low r (near zero) indicates that parental ranks were not stable across the sites and

hence, genotype x environment interaction exists. A high B, indicates that full-sib

families performed similarly across the sites, while a high? B indicates that the total

genetic values of the clones were stable across the trials.

Results and Discussion

Overall Growth of Clones and Seedlings

Survival of both clonal and zygotic propagules was high across all of the field

trials. Survival of the rooted cuttings ranged from 91.7-98.3% at the six field trials, while

seedling survival ranged from 86.5-97.4%. At the time of planting, seedlings were

generally taller than rooted cuttings, and this trend has continued through year two (Table

3-3). For example, after the first and second growing seasons seedlings were on average

11 cm and 10 cm taller than rooted cuttings, respectively. It has been suggested that

propagule size differences at the time of planting may create difficulties in the analysis

and interpretation of subsequent growth (Frampton and Foster 1993). However, these

differences in initial propagule size may not be a problem in the current study because

height increments were very similar between both propagule types, with site means

ranging between 1.0-2.0 m and 1.0-1.9 m for rooted cuttings and seedlings, respectively

(Table 3-3).

The initial and first year treatments were the same within a site during test

establishment, and there were no cultural differences between propagule types for 1st year

height. There were some differences in mean height due to the effects of cultural









intensity during the second growing season at some of the sites. However, these effects

were more a function of scale. Although the overall means of the growth variables

differed by cultural treatment, the ranks of parents, families, or clones were not affected.

Type B genetic correlations exceeded 0.85 indicating little cultural treatment x genetic

effect interaction, and therefore, cultural effects are ignored for the purposes of this study.

This implies that the rankings of parents, families, and clones were robust across multiple

management regimes through age two.

Table 3-3. Mean 1st year height, 2nd year height, height increment, and crown width by
propagule type for each of the six field trials. Although means are expressed
in meters, analyses were conducted using measured traits in centimeters.
1st Year 2nd Year Height Crown Width
Height (m) Height (m) Increment (m) (m)
Clones 1.0 2.1 1.2 0.9
Test A
Seedlings 1.1 2.3 1.2 1.0
Clones 0.8 1.8 1.0 0.9
Test B
Seedlings 0.9 1.8 1.0 0.9
Clones 1.0 2.1 1.1 1.2
Test C
Seedlings 1.1 2.3 1.2 1.3
Clones 0.7 2.0 1.3 1.0
Test D
Seedlings 0.8 2.1 1.3 1.1
Clones 1.2 3.3 2.0 2.0
Test E
Seedlings 1.3 3.2 1.9 2.0
Clones 0.5 1.7 1.3 1.1
Test F
Seedlings 0.6 1.8 1.3 1.1


Genetic Components of Variance

All of the early growth traits demonstrated genetic variation (Figure 3-1; Table 3-4;

Appendix C). However, the genetic variation partitioned differently for the two

propagule types. In all cases, the estimate of additive genetic variation was greater for









the clones than the seedlings. Within a single trial, for instance, the estimate of the

additive genetic variation for 2nd year height based on seedlings was 0 to 0.58 that of

additive genetic variation based on clonal data. Similar trends were observed from the

across-trial analyses. The estimate of additive genetic variation for the clonal material

was 4.7, 3.3, 2.9, and 2 times greater than the additive genetic variation for seedlings for

1st year height, 2nd year height, height increment, and crown width, respectively (Table 3-

4).

Based on single-trial analyses, the majority of the genetic variation associated with

all of the growth variables in the clonal population was additive, while in the seedling

population the trend was towards nonadditive genetic variation (Figure 3-1). For

example, at Test D all of the genetic variation associated with 2nd year height of clones

was additive, while for the seedlings it was dominance genetic variation (Figure 3-1).

When all of the data were analyzed together, then the estimates of dominance genetic

variation were approximately equivalent for both propagules (Table 3-4) suggesting a

large test by dominance interaction for the seedlings. Epistasis was negative for all

growth variables. As a result, estimates of additive and dominance genetic variance for

these traits might not be upwardly biased as indicated by the expected portions of

epistatic interactions defined in Equations 3-3 and 3-4 (Foster and Shaw 1988). Isik et al.

(2003) reported similar trends in the partitioning of the genetic variation for clones and

seedlings including negative epistasis estimates for height, diameter, and volume through

age six. They also reported that the additive genetic variation for growth traits was

always greater for clones than the estimate for seedlings, while dominance genetic

variation was greater in the seedling population (Isik et al. 2003).









Table 3-4. Genetic parameter estimates for 1st year height, 2nd year height, height
increment, and crown width by propagule type across all six trials. Standard
errors are given in parentheses.
1st Year Height 2nd Year Height Height Increment Crown Width

Clone Seed Clone Seed Clone Seed Clone Seed
69.8 14.7 401.5 119.9 152.6 51.8 110.3 55
VA (21.3) (7.8) (122) (50.8) (47) (24.2) (34) (20)

12 17.5 86.4 83.3 30.3 41.9 19.7 20
D (6.2) (9.3) (36.5) (41.6) (13.6) (21.5) (10.1) (11.7)

-11.5 -107.6 -39.7 -13.5
(11.4) (64.3) (24.8) (18.1)
70.3 32.1 380.3 203.1 143.2 93.65 116.5 75
G (10.8) (9.5) (61.1) (53.8) (23.6) (26.1) (17.2) (20.7)

442.8 405.2 1806 1703 902.6 972 640.9 649.5
p (11.3) (7.6) (62.3) (36.2) (24.5) (19.4) (17.8) (14.3)

2 0.16 0.04 0.22 0.07 0.17 0.05 0.17 0.08
(0.04) (0.02) (0.06) (0.03) (0.05) (0.02) (0.05) (0.03)
S 0.16 0.08 0.21 0.12 0.16 0.1 0.18 0.11
(0.02) (0.02) (0.03) (0.03) (0.02) (0.02) (0.02) (0.03)

S0.81 0.71 0.88 0.78 0.83 0.6 0.82 0.78
erBC (0.06) (0.2) (0.04) (0.12) (0.05) (0.15) (0.06) (0.12)

B 0.8 0.5 0.88 0.66 0.83 0.58 0.82 0.67
rBsF (0.05) (0.1) (0.04) (0.09) (0.05) (0.1) (0.05) (0.09)
rB 0.69 0.77 0.76 0.76
(0.04) (0.03) (0.04) (0.03)


Genetic causes are not the only source for similarity among relatives. Non-genetic

factors, such as C effects, can lead to upwardly biased estimates of total genetic and

nonadditive genetic components of variance when analyzing clonal data (Libby and Jund

1962). C effects are often assumed negligible when estimating epistatic variance (Foster

and Shaw 1988). If C effects are present then the total genetic variation associated with









clones will be overestimated (Libby and Jund 1962). Significant C effects are likely to

occur in traits that are measured soon after propagation, but apparently lessen for traits

measured at later times (Libby and Jund 1962). In the current experiment, if C effects

exist, then estimates of epistatic genetic variation will be confounded with C effects

because ramets of a clone came from a single, non-replicated hedge. However, the

estimates for epistasis were negative for all of the growth traits (Table 3-4), which

suggest that interloci interactions are not an important source of genetic variation for

early growth traits, and that C effects may not be a major contributing factor. Further,

estimates of additive and dominance should be free of confounding C effects in the

current experiment since randomization of clones occurred at all stages including hedge

establishment, propagation, and growth prior to and after test establishment.

As Falconer and Mackay (1996) point out, relatives of all sorts may resemble one

another because of sharing a common environment. The variance attributed to a common

environment occurs more frequently and contributes greater to the covariance of full-sibs

than to the covariance of any other sort of relatives (Falconer and Mackay 1996). When

a common environment effect exists, then the differences between means of families

become greater than when these non-genetic factors are not present. In the case of full-

sib families, then this will lead to biased or inflated estimates of dominance genetic

variation.

The results seen in the current experiment for dominance genetic variance estimates

in the seedling population appear to be a result of a partitioning problem in that estimates

of the variance due to dominance are inflated at the expense of additive effects (Figure 3-

1). Seeds are often sown and seedlings grown in full-sib family blocks in the greenhouse
















a. 1styear height
0.7
0.6
0.5
- 0.4 I na


0.2
0.1








0.45
0
o (1 0 (D 0 (D 0 (D 0 (D 0 w
-5 U) O U) 0 U) 5 U) 5 U) O U)
TestA TestB TestC TestD TestE TestF


c. Height increment
0.45 -
f nlA


b. 2nd year height
0.6

S0.5

0.4
IU na
0.3
[ add
q .2

0.1



0 () 0 () 0 (1) 0 (1) 0 (1) 0 (

TestA Test B Test C Test D TestE Test F


d. Crown width
0.6
0.3 1~05-


0.3 0.4
0.25 na d ai
0 0.3
^ 0.2 Iadd ^ T add
0.15 0.2 I I
0.1
~ os 8o0.1
0 0

W ( 0 0 (W 0 (0 0 (D 0 (0 0 (W 0 (W 0 (1 0 (1 0 (1 0 (1 0 (1
-O U) -0 U) -O U) -O U) > U) > U) U) -0 U) -O U) -O U) -O U) (n U)
Test A TestB Test C Test D Test E TestF TestA TestB Test C Test D TestE TestF


Figure 3-1. The proportion of additive (add) and nonadditive (na) genetic variance components for clones and seedlings across the six
field trials, where h2 = add and H2 = add + na: a. 1st year height, b. 2nd year height, c. Height increment, and d. Crown
width. Standard error bars for broad-sense heritability estimates are included.









or nursery for progeny testing or other field testing. In order to eliminate the common

environment effect at the family level, then seedlings within a full-sib family should be

randomized over the environment in which they are being grown and tested. Although

rooted cuttings were randomized at all stages of propagation, the seedlings were in fact

grown in full-sib blocks, and apparently a common environment effect has carried over to

the field through age two measurements.

Heritability Estimates

First year height, 2nd year height, height increment, and crown width were all

influenced by additive genetic variation. Individual tree narrow-sense heritability was

always greater for the clones than for seedlings for all of the traits (Figure 3-1, Table 3-

4). Generally, h2 estimates of total height increased from age one to age two for both

clones and seedlings (Figure 3-1, Table 3-4). Narrow-sense heritability for total height

increased from 0.16 at age one to 0.22 at age two for the clones, while in the seedling

population, h increased from 0.04 to 0.07 (Table 3-4). In addition, h estimates for all

of the growth traits from the seedling data had larger standard errors associated with them

than the estimates from clones (Table 3-4). Yet, the differences in heritability estimates

between propagule types may be related to C effects. If C effects are present, then H2

estimates from the clonal population may be inflated. However, estimates of h2 from the

clonal population may not be upwardly biased since clones within full-sib families were

randomized in the hedge orchard and throughout the study, thus reducing spurious C

effects at the parental and full-sib family levels.

Heritability estimates have been reported for several loblolly pine seedling and

clonal populations. In a study reported by Isik et al. (2003), individual tree narrow-sense









heritability for volume at age six was 0.3 for clones, while for seedlings h2 was 0.06. In

another loblolly pine clonal test, heritability estimates for total height growth through age

five for two factorials were similar to the values reported here (Paul et al. 1997). They

reported h2 for 1st and 2nd year height as 0.08 and 0.17, respectively (mean of two

factorials). Unfortunately, they did not include seedlings in their study. Analogous

results have been reported in Eucalyptus globulus Labill. for heritability estimates for

diameter for both clones and seedlings in that h2 was always greater for the clonal

population than the seedling population (Costa e Silva et al. 2004).

Broad-sense heritability estimates for the various growth traits were always larger

for the clonal material than the seedlings except for the estimates of H2 for Ist year

height at Test E and for all four variables at Test D (Figure 3-1). However, H2 estimates

from the across-trial models were always greater for the clones than seedlings (Table 3-

4). Because of the negative estimates of epistasis in the clonal population, 2 estimates

were equivalent to the estimates for individual tree narrow-sense heritability for all traits.

As was the case with 2, seedling estimates of H2 had higher standard errors associated

with them (Figure 3-1).

Type B Genetic Correlations Between Propagule Types

In order to further compare the clonal rooted cuttings and zygotic seedlings, type B

genetic correlations for general combining ability and full-sib family value between

propagule types were estimated. For all growth traits measured, the type B genetic

correlation between propagule types for additive effects was high with values exceeding

0.72 (Table 3-5). For example, ro for Ist year height ranged from 0.93-0.99, while









for height increment Bprop ranged from 0.72-0.99 (Table 3-5). These high genetic

correlations imply that parental rankings for early growth traits are stable regardless of

whether their progeny are being tested as zygotic seedlings or rooted cuttings (Figure 3-

2a).

Table 3-5. Genetic correlations between propagule types for 1st year height, 2nd year
height, height increment, and crown width at the parental (rop ) and full-sib
family (ro F) levels.

1st Year Height 2nd Year Height Height Increment Crown Width

Test rBpropGCA BpropFS rBpropGCA rBpropFS rBpropGCA rBpropFS propGCA propFS

A 0.99 0.62 0.99 0.65 0.99 0.38 0.86 0.86
B 0.99 0.74 0.99 0.86 0.99 0.99 0.99 0.81

C 0.93 0.74 0.99 0.75 0.99 0.99 0.99 0.97
D 0.93 0.44 0.99 0.64 0.99 0.92 0.93 0.92

E 0.99 0.55 0.89 0.71 0.72 0.77 0.99 0.69

F 0.95 0.49 0.96 0.89 0.95 0.83 0.99 0.83



The type B genetic correlations for additive effects between propagule types

observed in the current study were consistent with the expectations reported by Borralho

and Kanowski (1995). In a simulation study comparing the performance of clones and

seedlings from the same half-sib family, Borralho and Kanowski (1995) reported that the

expected correlations between propagule types exceeded 0.8 when greater than 100

seedlings or propagules were tested. Additionally, in a field study comparing rooted

cuttings and seedlings from four half-sib families, Foster et al. (1987) reported that

family rank correlations between propagule types were positive and significant for 1st

year height (0.52), 3rd year height (0.66) and 6th year height (0.70).









Full-sib families also ranked relatively similar regardless of propagule type at most

of the sites (Table 3-5; Figure 3-2b). For example, type B genetic correlations at the full-

sib family level ranged from 0.64-0.89 for 2nd year height. However, B, was more

variable from site to site for 1st year height and height increment. The results observed

here are in accordance with those reported by Frampton et al. (2000) in that full-sib

families or parental trees selected based on seedling genetic trials should also perform

well as rooted cuttings.

Genotype x Environment Interaction

The stability of parents and full-sib families across sites was compared for the

clonal and seedling populations (Table 3-4). The type B genetic correlation for additive

effects across sites was always greater for the clones, although these estimates were

moderately high for both populations. For example, ?B, for 2nd year height was 0.88

and 0.78 for the clonal and seedling populations, respectively (Table 3-4). Full-sib

families also ranked comparably across sites for the clonal population with i- values

exceeding 0.8 for all of the early growth traits (Table 3-4). These across site genetic

correlations were estimated with more precision using clonal replicates as evidenced by

lower standard errors for the estimates from the clonal population.

An additional genotype x environment interaction between test and total genetic

value can be estimated using clonal tests. The total genetic values of the clones were

fairly stable when all of the data was considered from all sites, indicating that a good

clone at one site is good at all the sites (Table 3-4). However, based on analyses from

pairs of trials (data not shown), there appears to be a propagation effect relating to the

season the cuttings were rooted. For example, the worst genetic correlations were







53


observed between the trial established with rooted cuttings from the winter setting and

any of the other trials, while the best genetic correlations were obtained from the field

trials that contained rooted cuttings originating from the spring setting. Similar results


35
30
25
r 20
1 15
C,
10
5
0





70
60
50


a,
- 40
30
CO


a. Parental ranks for 2nd year height

*
rB =0.99










0 5 10 15 20 25 30 35
Test B Parental Ranks Clones


b. Full-sib family ranks for 2nd year height


r =0.86
rBpropFS *

0 10* *3


-*I
0 10 20 30 40 50 60 70





0 10 20 30 40 50 60 70


Test B Full-Sib Family Ranks Clones


Figure 3-2. Rank-rank plots showing type B genetic correlations between clones and
seedlings from Test B based on: a. Parental BLUP values, b. Full-sib family
BLUP values, where full-sib BLUP values are equal to the sum of the
predicted general combining ability for each of the two parents plus the
predicted specific combining ability of the cross.









were also observed with rooting with this same population in that poor genetic

correlations were observed between rooting ability in winter and spring (Chapter 2;

Baltunis et al. 2005).

Conclusion

Several forest industries in the southeastern United States are deploying full-sib

families of loblolly pine operationally. In addition, many of these companies are

pursuing clonal forestry programs with loblolly pine. Genetic field trials established with

clones and seedlings from the same full-sib families provide an opportunity for

comparing both half-sib and full-sib family performances for both propagules. Based on

the current study, several conclusions can be drawn. First, clonally replicated seedling

trials of loblolly pine provide genetic information with greater precision than zygotic

seedlings. Second, genetic correlations between estimates of the genetic effects

associated with these growth traits between these propagule types were highly favorable.

These high genetic correlations between propagule types reassure that parental and full-

sib family rankings are stable regardless of propagule type. This implies that parental

and full-sib family rankings based on existing seedling progeny trials could be used to

select parents and families that perform well when they are deployed as rooted cuttings.

Third, little genotype x environment interaction was observed across sites at the parental,

family, and clonal level for all traits. However, there appears to be a carry-over effect

relating to the season in which the cuttings were rooted for the clonal material. Finally,

randomization is essential at all stages in testing when estimating genetic parameters.

The lack of randomization for the seedling population apparently resulted in a problem

with partitioning of the genetic variance, causing full-sib families to appear more

different and inflating estimates of dominance genetic variation.
















CHAPTER 4
GENETIC GAIN FROM SELECTION FOR ROOTING ABILITY AND EARLY
GROWTH IN VEGETATIVELY PROPAGATED CLONES OF LOBLOLLY PINE

Introduction

Loblolly pine (Pinus taeda L.) is the most important commercial tree species in the

southeastern United States (McKeand et al. 2003). Genetic improvement of loblolly pine

has been occurring since the 1950's in several tree improvement programs. These

programs have aimed to increase the population mean breeding value of a few key traits,

such as stem volume, disease resistance, and wood properties, through breeding and

selection of superior genotypes. Tree improvement programs are based on recurrent

selection for general combining ability and capture only a portion of the additive genetic

variation with open-pollinated seedlings for deployment. However, the nonadditive

portion of genetic variation, dominance and epistasis, may be important components of

variation for traits. For example, Stonecypher and McCullough (1986) reported that the

estimates of nonadditive variance were approximately equal to those of additive variance

for growth traits in Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) through age six.

Additionally, Paul et al. (1997) reported that both additive and dominance genetic

variance increased from age one to five for height growth for loblolly pine clones. The

only manner in which to capture the total genetic variation is through operational

deployment of clonal propagules.









Based on current technologies, several forest industries in the southeastern United

States are pursuing clonal forestry programs for loblolly pine using either somatic

embryogenesis or rooted cuttings (Weber and Stelzer 2002). Two main criteria need to

be met prior to operational deployment of loblolly pine clones. First, loblolly pine clones

must perform well, e.g., meet the selection criteria for the desired traits. This involves

the accumulation of reliable data for the clones from greenhouse screening, field trials,

etc. Second, the selected clones have to be propagated in large enough numbers for

deployment. For a rooted-cutting-based clonal program, this involves bulking up the

number of hedges (ramets) of a particular clone or group of selected clones through serial

propagation and then producing reforestation stock efficiently from the bulked-up clones.

Only those tested clones that can be propagated easily in sufficient numbers will be

economically feasible for deployment.

Loblolly pine is considered a difficult to root species (Wise and Caldwell 1994). In

populations that have not experienced any selection pressure for rooting ability, loblolly

pine has been reported to root near 50% (Foster 1990; Baltunis et al. 2005). Previous

rooting studies with loblolly pine have demonstrated substantial family and clonal

variation for rooting ability (Foster 1990; Baltunis et al. 2005), indicating the potential

for increasing rooting efficiency for both clonal deployment and through recurrent

selection and breeding. Selection for both rooting ability and field growth will be

necessary for a successful loblolly pine clonal forestry program based on rooted cutting

technology (Foster et al. 1985; Foster et al. 2000; Baltunis et al. 2005).

The objectives of this study were to (i) determine the genetic correlation between

rooting ability and 2nd year height, (ii) predict the genetic gain associated with selection









for rooting ability, (iii) predict the genetic gain associated with selection for 2nd year

height, and (iv) predict the genetic gain from combined selection for rooting ability and

2nd year height using a Monte Carlo selection index.

Materials and Methods

Population

The parental population consisted of twenty first-generation and ten second-

generation selections from the larger Loblolly Pine Lower Gulf Elite Population. Two

additional first-generation, slow-growing parents were included. The parental selections

represent the Atlantic Coastal Plain, Florida, and Lower Gulf provenances of loblolly

pine. These thirty-two loblolly pine parents were mated in a partial diallel design and

created 70 full-sib families from which approximately 2,200 seedling hedges were

generated and given unique clonal identifications (Appendix A). On average, each parent

was involved in about four crosses.

Rooting and Field Trials

The propagation of the rooted cuttings for this study has previously been described

(Baltunis et al. 2005; Chapter 2). But briefly, the seedling hedges were repeatedly

sheared to slow down the effects of maturation and increase the number of shoots

available for collection. Cuttings were set in five rooting trials over two years in May

2001, July 2001, January 2002, April 2002, and June 2002 in trials SpringOl, SummerOl,

Winter02, Spring02, and Summer02, respectively. Four to nine ramet clonal row plots

were set in four to six replications depending on trial (Chapter 2). At the time of

collection, the hedges were 13, 15, 22, 25, and 27 months old, respectively. Cuttings

were assessed for rooting nine to eleven weeks after they were set (Baltunis et al. 2005;

Chapter 2).









Six field trials (A, B, C, D, E, F) were established with rooted cuttings from the

three latter rooting trials. Three field trials each were established in Georgia and Florida

between October 2002 to April 2003 (see Table 3-1). The clonal propagules for each

field trial, however, came from a single rooting trial. Single-tree-plots of a clone were

established in four replications in each of two cultural treatments (high and low intensity)

in each test, except for Test E where there was only one cultural treatment and four

replications. The goal for the high intensity treatment was to push the trees to their

utmost potential by reducing competition and providing a non-limiting supply of

nutrients, while the low intensive culture provides insights into family and clonal

performance under a less optimal cultural regime (FBRC 2000). Previous analyses of

growth for this population had shown that type B genetic correlations exceeded 0.85

indicating little cultural treatment x genetic effect interaction, and therefore, cultural

treatment effects were ignored for the purposes of this study (Chapter 3).

Statistical Analyses

A bivariate parental linear mixed-effects model was used to obtain estimates of

variance and covariance components for rooting ability and 2nd year height using

ASREML (Gilmour et al. 2002):

[4-1] y, =X,b, +Zn, +Z,,u, +Zff +Zcc, +Z0o, +Zp, p +Zqq, + e,,

where y, is the vector of observations indexed (i) by rooting ability and 2nd year height,

b, is the vector of fixed effects (i.e. mean, trials and replications within trials) and X, is

the known incidence matrix relating the observations in y, to the fixed effects in b,


where X,b, roo
w e 0 X height b height











n, is the vector of random incomplete blocks nested within replication and test effects


-MVN 0, rootNC
0


0
Shegh
'heightC\


u, is the vector of random parent (female and male) general combining ability effects


-MVN (0, G 0 A) where G


F "2
'GCAroot
7GCAroot~giht


GCAro t Jght and A

0*GCAhelght


numerator relationship


matrix,


f, is the vector of random specific combining ability effects -MVN (0, S 0 Is) where


SCAro

C SCAroothight


SSCA"""" and I is an identity matrix of size equal to the number of
SCAeight _


full-sib families,


c, is the vector of random clones within full-sib family effects -MVN (0, C 0 Ic) where


C= CLONEot
CLONErooteght


SCLONEroo' and Ic is an identity matrix of size equal to the number
'CLONE ght


of clones,


o, is the vector of random test by full-sib family effects



~MVN 0, i r oot a
0 I height 7 i


p, is the vector of random test by clone within full-sib family effects


-MVN 0, orroot TESTxCLONE t
I 0


0
S 2 'TESTxLONE
height TESTxCLONE~h









q, is the vector of random replication within test by full-sib family effects


~MVN 0, i'
O height T



e, is the random vector of residual terms -MVN 0, roo R he ght
0 height 'ERRORigi _

Z,Z Z, Z Zf,, ZCI, Zo, ZP, and Zq, are the known incidence matrices relating the

observations in y, to effects in n,, u,, f,, c,, o,, p,, and q,, respectively, and I, is the

identity matrix of dimension equal to the number of observations for rooting ability or 2nd

year height.

Causal Components of Variance

Genetic parameters were estimated and standard errors were calculated for rooting

ability and 2nd year height from the bivariate analysis according to Foster and Shaw

(1988).
[4-2] ~2 V 1 1
[4-2] A = VA + +- VAA ...
4 16

is the estimate of additive genetic variance.
4-3 V 2 V 1 1
[4-3] D = SCA =D + VAA + AD -VDD
2 2 4

is the estimate of dominance genetic variance.
2[4-4] V 2& C 1 1 3
[4-4] V = C^LONE 2 ^GCA -3 SC = -V +- VAD DD "
4 2 4

is the estimate of epistatic genetic variance.

[4-5] VG = 2<^cA + -SCA + UCLONE

is the estimate of total genetic variance.










j^ 2 +(2 N 2 +j2 12 1 2
[4-6] P = 2 GCA SCA C cLONE TESTxFAM TESTxCLONE REPxFAM ERROR

is the estimate of the phenotypic variance.

^2 ^2 ^2 ^2 ^2 ^2
[7 = 2 SCA CLONE TESTxFAM 'TESTxCLONE REPxFAM ERROR
[4-7] VP- dGC + + + + + +
Sf fc tf tfc tr trfcn

is the phenotypic variance of half-sib family means, f= harmonic mean number of full-

sib families per half-sib family, c = harmonic mean number of clones per full-sib family,

t = number of trials, r = harmonic mean number of replications per test, and n = harmonic

mean number of ramets per clone per replication per test (n = 1 for field trials).

^2 ^2 ^2 ^2 ^2
[A- v 2 2 CLONE TESTxFAM TESTxCLONE REPxFAM ERROR
[4-8] Vp = 2CGA + CSCA + + + + +
c t tc tr ctrn

is the phenotypic variance of full-sib family means.

^2 "2 ^2 ^2
1 2 (2 j 2 2 'TESTxFAM TESTxCLONE 'REPxFAM ERROR
[4-9] V GC = 2 A + SCA +CLONE + + + +
CL t t tr trn

is the phenotypic variance of clonal means.

Heritability Estimates

Heritability estimates have previously been reported for this population for rooting

ability (Baltunis et al. 2005; Chapter 2) and 2nd year height (Chapter 3). However, since

rooting ability and 2nd year height were analyzed using a bivariate model in the current

study, heritabilities were estimated based on the genetic parameter estimates from the

bivariate model and differed slightly from those reported previously. Standard errors

were calculated using the Taylor series expansion method (Kendall and Stuart 1963;

Namkoong 1979; Huber et al. 1992; Dieters 1994).











[4-10] h2 =A GCA
S 2 +"2 +2 +E2 2 +"2 2
S2cGCA SCA CLONE TESTxFAM TESTxCLONE REPxFAM ERROR

the individual tree narrow-sense heritability.


[4-111 2 G GCA ^SCA CLONE
^-12 2 2 2 2 12 '2 2
VP 2 GCA SCA 'CLONE + TESTxFAM + TESTxCLONE + UREPxFAM + RROR

the individual tree broad-sense heritability.

^2
[4-12] H = CA is the half-sib family mean heritability.
HS


^2 "2
[4-13] i2 2= GCA SCA i the full-sib family mean heritability.
Vp-
FS


r2 ^2 ^2
[4-14] 2 = GCA SCA CLONE is the clonal mean heritability.
CL
vpi

Type B Genetic Correlations

Although the same genotypes, or clones, were tested in the rooting and field trials,

these measurements were not necessarily taken on the same ramet. Therefore, type B

genetic correlations between rooting ability and 2nd year height for general combining

ability, full-sib family value, and the total genetic value were calculated. Standard errors

of these estimates were calculated using the Taylor series expansion method (Kendall and

Stuart 1963; Namkoong 1979; Huber et al. 1992; Dieters 1994).


OGCAotg
[4-15] ; 2 is the type B genetic correlation between rooting ability
GCAroot X GCAeIght


and 2nd year height for additive effects, and GCA is the covariance between general


combining ability effects for rooting ability and 2nd year height..










[4-16] 2r GCA4h," +SCA,,g is the type B genetic correlation
"2 2 ^2 ^ 2
S^GCA ot + SCA t 2GCAheght + 4SCA, ,

between rooting ability and 2nd year height for full-sib families, and

2 ^GCAoot&,g + SCAh is the covariance between the full-sib family effects for rooting
^ootheigh ^ footheght

ability and 2nd year height.


[2 -17] s te 2 ty2 2
I2I GCAo + SCA, + CLONE 21cGCA, + SCAg + t CLONE )

B genetic correlation between rooting ability and 2nd year height for the total genetic

value of clones, and 2,GCAooh + i -SCArooti ,g + i 'CLONErooet is the covariance between the


total clonal value effects for rooting ability and 2nd year height.

Genetic Gain

Genetic gain was estimated for a number of deployment options based on various

selection scenarios using the BLUP values from the bivariate analysis. All deployed

populations were assumed to be propagated as rooted cuttings. An additional assumption

for gain calculations was that the seventy full-sib families and all clones within families

were available for deployment. The deployment strategies considered were 1) half-sib

family deployment, 2) full-sib family deployment, and 3) clonal deployment. For a half-

sib family deployment option, the predicted value for each parent (general combining

ability) was used to select the best half-sib family for deployment as rooted cuttings.

Full-sib family values were calculated by summing the predicted values for the female

general combining ability, male general combining ability, and the full-sib family

specific combining ability. The genetic gain over the trait mean was determined for

deployment of the best full-sib family.









The total genetic value of a clone was determined by summing the predicted values

for the female general combining ability, male general combining ability, the full-sib

family specific combining ability, and clone within full-sib family. Several clonal

deployment strategies were compared. First, the genetic gain was determined for

deployment of the single best clone for each trait. A second clonal deployment strategy

was considered by selecting the top ten full-sib families and then deploying the single

best clone from each of these ten families. Two additional clonal deployment strategies

were compared by selecting the top 10% and 1% of clones (out of 2200 possible clones)

using an unrestricted selection index for 2nd year height and rooting ability. The total

genetic values of the clones were weighted with the following weights: 1:0, 0.9:0.1,

0.8:0.2, ..., 0.2:0.8, 0.1:0.9, 0:1 for 2nd year height and rooting ability, respectively, in the

Monte Carlo index (Cotterill and Dean 1990). All gains were expressed as the

percentage gain over the mean of the trait:


[4-18] %Gain = x 100%, where


x5 is the average predicted value for trait i of the selected population, and

7, is the population mean for either rooting ability or height.

In addition genetic gain was calculated using theoretical gain formulae assuming an

unrelated population (Falconer and Mackay 1996), and these estimates were compared

with the predicted genetic gain based on BLUP values for the half-sib family, full-sib

family, and best clone deployment options. The following formulae were used:









[4-19] Gain = ihs V is the genetic gain associated with selection of the best half-


sib family, and ihs is the selection intensity corresponding to selecting one out of thirty-

two half-sib families (i = 2.07) and assuming all half-sib families are unrelated,

[4-20] Gain = if,Hi V is the genetic gain associated with selection of the best full-


sib family, and if, is the selection intensity corresponding to selecting one out of seventy

full-sib families (i = 2.38) and assuming all full-sib families are unrelated,

[4-21] Gain = i -2~p is the genetic gain associated with clonal selection, and i, is

the selection intensity corresponding to selecting one out of 2,206 clones (i = 3.58) and

assuming all individuals are unrelated.

Results and Discussion

Causal Components of Variance

The genetic parameter estimates from the bivariate analysis of rooting ability and

2nd year height in loblolly pine clones were consistent with the estimates from their

respective univariate analyses, and both traits showed genetic variation (Table 4-1;

Appendix D). Additive genetic variation accounted for the majority of the genetic

variation associated with rooting ability and 2nd year height (Table 4-1). Dominance

genetic variation contributed a minor portion to the total genetic variation for both traits.

Epistasis appears to be more important for rooting ability than 2nd year height as

evidenced by a negative estimate of epistasis for 2nd year height (Table 4-1). However,

estimates of epistasis may be confounded with C effects since rooted cuttings of a clone

originated from a single ortet.









Type B Genetic Correlations

The bivariate analysis of five rooting trials and six field trials also allowed for

estimation of the genetic covariance between rooting ability and 2nd year height for

parental effects, full-sib family effects, and the total genetic value of clones within full-

sib family. There was a positive genetic relationship between rooting ability and 2nd year

height at all three genetic levels (Table 4-1). The genetic correlation at the parental level

between rooting ability and 2nd year height (~C ) was 0.32. At the full-sib family level,

the genetic correlation between traits (BFS ) was 0.39. The correlation of total genetic

values of clones for rooting ability and 2nd year height (mB ) was 0.29.

Previous studies have also reported positive correlations between rooting traits and

growth. For example, Paul et al. (1993) reported that 1st year and 5th year height growth

had strong genetic correlations with rooting (0.61 and 0.69, respectively) in western

hemlock clones (Tsuga heterophylla (Raf.) Sarg.). In another study, Foster et al. (1985)

reported that the genetic correlation between rooting ability and growth of western

hemlock clones in a growth chamber was 0.37. A weak, but positive, relationship

between the number of roots on loblolly pine cuttings and subsequent field growth has

been reported (Foster et al. 2000). On the other hand, Goldfarb et al. (1998) reported that

the phenotypic correlation between the number of roots and first year field growth in

loblolly pine rooted cuttings was negligible.

C effects relating to the season in which the cuttings were set has previously been

discussed in this population (Chapter 3). Further evidence of the presence of C effects

can be seen from the genetic correlations between the total genetic value of clones for

rooting ability and 2nd year height based on analyses involving a single rooting and field









trial (Table 4-2). Rooted cuttings planted in each field trial originated from a single

sticking date, and the highest genetic correlations (rBTG ) occurred between traits from

these trials. For example, rooted cuttings from the Winter02 rooting trial were planted

only in the Field A trial. The total genetic correlation between rooting ability from

Winter02 and 2nd year height from Field A was 0.24, while 2nd year height from Field A

had much lower correlations with rooting ability from any of the other sticking dates

Table 4-1. Means, variance component estimates, heritabilities, and genetic correlations
from the bivariate analysis of rooting ability and 2nd year height. Standard
errors are given in parentheses.
Rooting Ability 2nd Year Height
S42 % 210 cm

VA 0.0135 (0.004) 399.3 (119.6)
VD 0.0014 (0.002) 77.3 (39.6)
V, 0.0085 (0.002) -94.9 (63.7)
VG 0.0235 (0.002) 381.7 (59.9)

VP 0.2299 (0.002) 1836 (61.0)
Vp- 0.0038 (0.001) 108.1 (29.6)
Vp- 0.0085 (0.002) 231.1 (59.1)
VP~ 0.0286 (0.002) 434.1(59.9)
h2 0.059 (0.02) 0.22 (0.06)
H2 0.102 (0.01) 0.21 (0.03)

Hf2 0.88 (0.04) 0.92 (0.03)
HS
fi2 0.84 (0.04) 0.95 (0.01)
FS
fi2 0.82 (0.01) 0.88 (0.02)
CL
rB 0.32(0.19)
rB, 0.39 (0.17)
rB, 0.29 (0.08)









(Table 4-2). This indicates that the vigor of the hedge at the time of rooting may have

influenced rooting and subsequent field growth.

The biological significance of positive correlations between rooting and 2nd year

height in the current study is unclear. Goldfarb et al. (1998) reported that root

morphological traits were not meaningful for subsequent field growth after one year of

growth for loblolly pine rooted cuttings. However, it has generally been acknowledged

that stock plant vigor is related to rooting. Perhaps the vigor of the stock plant is an

indicator of the metabolic activity of cuttings during root initiation. The more

metabolically active genotypes may have rooted quicker and formed better root systems.

In the current study, rooting was assessed at nine to eleven weeks, and selection may

have indirectly been for rate of rooting as opposed to strictly rooting ability. Clones that

rooted quickly during the rooting period may have had increased metabolic activity. A

positive genetic correlation between rooting and growth implies that some of the same

genes are responsible for the expression of each trait. There may be genes in common for

rooting and height relating to metabolic activity. Genotypes that have a tendency for

increased rates of metabolic activity may root at higher frequencies and grow larger than

genotypes that do not have these alleles.

Genetic Gain

The importance of selecting for rooting ability in a clonal forestry program based

on rooted cuttings has long been recognized (Foster et al. 1984; Foster 1990; Baltunis et

al. 2005). Genetic gains in field traits will not be realized if the clones can not be

propagated efficiently for deployment. Therefore, increases in rooting ability could have

broad economic impacts on a clonal forestry program (Foster et al. 1984). Positively

correlated traits imply that selection for one trait should also lead to improvement in the









second trait. In the case of rooting ability and growth, positive genetic correlations can

lead to substantial gains for both traits in a clonal forestry program based on rooted

cuttings.

The genetic gains in rooting ability and 2nd year height based on BLUP values were

compared for a number of deployment strategies. Selecting the top half-sib family for

rooting ability would result in a gain of 36% in rooting ability (Figure 4-1). Deployment

of the top half-sib family selected for rooting ability would result in a genetic gain of

5.4% in 2nd year total height (Figure 4-2). Selecting the highest ranking half-sib family

for 2nd year height would result in a gain of 14.8% and 8.1% in rooting ability and 2nd

year height, respectively (Figure 4-1 and Figure 4-2). The genetic gain predictions using

theoretical calculations were slightly lower than the gain predictions based on BLUP

values. The gain in rooting ability associated with selecting the best half-sib family for

rooting was 26.7% (using Equation 4-19). The gain in height by selecting the top

growing half-sib family was 9.4% (using Equation 4-19).

Slightly higher gains can be achieved by selecting the top full-sib family.

Propagation of the top full-sib family selected for rooting ability results in genetic gains

in rooting ability of about 43% over the population mean (Figure 4-1). Gains of nearly

9% in 2nd year total height could be expected by deploying the top rooting full-sib family

(Figure 4-2). If the best growing full-sib family was propagated and deployed, then the

genetic gains in rooting ability (Figure 4-1) and 2nd year height (Figure 4-2) would be

8.6% and 10.1%, respectively. The theoretical gain in rooting ability for full-sib

selection was equivalent to the estimate based on BLUP values. However, the theoretical









gain for full-sib selection for 2nd year height was 16.3% which was higher than the

estimate based on BLUPs.

Table 4-2. The total genetic correlation (?B ) between rooting ability and 2nd year height
from analyses of a single rooting trial and field trial. Shaded values indicate
the rooting trial in which cuttings originated from for their respective field
trials. Standard errors are given in parentheses.
Field A Field B Field C Field D Field E Field F

l 0.002 0.17 0.17 0.21 0.16 0.25
(0.07) (0.07) (0.07) (0.07) (0.07) (0.07)

0.08 0.07 0.17 0.21 0.14 0.12
SummerO1
(0.06) (0.07) (0.07) (0.07) (0.07) (0.07)

0.24 0.16 0.23 0.27 0.24 0.18
Winter02
(0.07) (0.07) (0.07) (0.08) (0.08) (0.08)

0.06 0.25 0.26 0.34 0.16 0.24
png (0.07) (0.07) (0.08) (0.07) (0.07) (0.08)

0.10 0.21 0.16 0.21 0.13 0.28
Summer02
(0.07) (0.07) (0.08) (0.08) (0.08) (0.08)



Clearly, the most gain for either trait would be achieved by selecting the single best

clone (Figure 4-1 and Figure 4-2). For instance, by selecting the top clone for rooting

ability, the genetic gain based on clonal predicted values in rooting was 96% (Figure 4-1)

indicating that approximately 82% of the ramets would root from this clone (42% mean

rooting plus 0.96 x 42% = 82%). However, selecting the top rooting clone from this

population would result in a decrease (genetic loss) in overall 2nd year height (Figure 4-

2). On the other hand, selecting the top clone for 2nd year height would result in a gain of

nearly 27% in 2nd year height (Figure 4-2) and 43% in rooting ability (Figure 4-1) over

the population mean for these traits. Genetic gain in rooting ability based on theoretical







71


calculations was greater than those based on BLUP values for selection of the single best

clone and was 118%, while the gain in 2nd year height for the top growing clone was

31.3% (using Equation 4-21).


100

90 0% Gain in Rooting Ability when
>h Selecting for 2nd Year Height
80 -
0% Gain in Rooting Ability when
S70- Selecting for Rooting Ability


0
60

o 50
40

30
o 20
10 .. -.


Best Half-Sib Best Full-Sib Best Clone from Best Clone
Family Family Best 10 Full-Sib
Families


Figure 4-1. The genetic gain in rooting ability (%) over the population mean for
deployment of the best half-sib family, full-sib family, best clone from the
best ten full-sib families, and the single best clone when selecting for rooting
ability or 2nd year height.


In order to address genetic diversity issues and the risk associated with deploying a

single clone, a second clonal deployment option was considered by selecting the best

clone from each of the ten highest ranking full-sib families. This strategy resulted in

genetic gains nearly double that of the full-sib family deployment strategy. When the

trait selected is rooting ability, the best clone in the top ten full-sib families would result

in gains of 77% in rooting ability (Figure 4-1) and 5.6% in 2nd year height (Figure 4-2).








When the trait selected is for 2nd year height, then the genetic gains would be 37.6% in

rooting ability (Figure 4-1) and 18.3% for 2nd year height (Figure 4-2).


E % Gain in 2nd Year Height when
Selecting for Rooting Ability
0% Gain in 2nd Year Height when
Selecting for 2nd Year Height


rf]


ifi


. . . . . .


Best Half-Sib
Family


Best Full-Sib
Family


Best Clone from
Best 10 Full-Sib
Families


Pest:


'lone


Figure 4-2. The genetic gain in 2nd year height (%) over the population mean for
deployment of the best half-sib family, full-sib family, best clone from the
best ten full-sib families, and the single best clone when selecting for rooting
ability or 2nd year height.

Deployment of well-tested clones can result in genetic gains in both rooting ability

and 2nd year total height. Both traits should be considered for a successful loblolly pine

clonal forestry program based on rooted cutting technology. The responses to selection

when considering both rooting ability and 2nd year height were compared for a number of

selection indices (Figure 4-3). The gain in rooting ability associated with selecting the

top 10% of clones ranged from 23.8% when only 2nd year height was considered to

57.6% when only rooting ability was considered (Figure 4-3). Higher gains can be









achieved by increasing the selection intensity. When only the top 1% of clones was

selected, the gain in rooting ability ranged from 33.4% when only 2nd year height was

considered to 80.6% when only rooting ability was considered (Figure 4-3). The genetic

gain in 2nd year height associated with deployment of the best 10% of clones ranged from

4.8% when only rooting ability was considered to 12.6% gain when only 2nd year height

was considered (Figure 4-3). Similarly, genetic gain in 2nd year height ranged from 3.8%

to 18.5% when the best 1% of clones were selected (Figure 4-3).

Clonal forestry programs that consider multiple traits need to optimize their

selection strategies for deployment populations. Arbitrarily setting the optimum selection

weights on 2nd year height and rooting ability to 90% of the maximum genetic gain

obtainable for a single trait, then the optimum selection weights can be compared. For

example, if the top 10% of clones are selected and 90% of the maximum gain in rooting

ability is considered, then the optimum weights on 2nd year height and rooting ability

correspond to weights of 0.7 and 0.3, respectively. These weights result in genetic gains

of 53.6% in rooting ability and 8.7% in 2nd year height (Figure 4-3). If 1% of the clones

are selected, then the optimum weights are 0.6 and 0.4 on 2nd year height and rooting

ability, respectively, and correspond to gains of 75.9% in rooting ability and 11% in 2nd

year height (Figure 4-3).

If 90% of the maximum gain obtainable for 2nd year height is the criterion, then the

optimum weights for both deployment options are 0.9 and 0.1 on 2nd year height and

rooting ability, respectively. The genetic gains associated with selecting the top 10% of

clones utilizing these weights results in gains of 40.4% in rooting ability and 11.9% in 2nd

year height (Figure 4-3). If only the top 1% of clones are deployed using these selection









weights, then genetic gains of 54.2% in rooting ability and 17.4% in 2nd year height are

obtainable (Figure 4-3). The maximum change in genetic gain of rooting ability results

from increasing the weight on rooting ability from 0 to 0.1. In fact, an additional 17-21%

gain in rooting ability can be obtained in the deployment population by increasing the

weight on rooting ability from 0 to 0.1.

When 10% of the clones were selected (220), then the number of full-sib families

represented in the deployment population ranged from 38 to 45 depending on selection

index (Table 4-3). When only the top 1% of clones were selected (22), the number of full

sib families represented in the deployment population varied from 9 to 15 (Table 4-3).

Maximum genetic gains can be achieved by ignoring relatedness among selections.

However, if no constraints are placed on the relatedness of selections, then there is a

tendency to make many selections from the better families. Although the approximate

average number of clones selected per full-sib family was five when 10% of the clones

were selected, nearly 45% of these selections came from only five families when only

rooting ability was selected. Similar trends can be seen when selecting 1% of the clones.

For example, when all of the selection weight is on rooting ability, then there was an

average of 2.44 clones per full-sib family selected (Table 4-3). However, 15 of the 22

selections came from three families. On the other hand, when considering the number of

half-sib families represented in the selected population, 26 to 30 out of 32 possible

parents are represented when 10% of the clones are selected. Thirteen to 17 parents are

represented in the deployment population when 1% of the clones are selected (Table 4-3).

Therefore, at least in this population, there may be sufficient genetic diversity in the

deployment population even with higher selection intensities. Isik et al. (2005) reported










genetic gains in growth near 30% for loblolly clones from a different population

irregardless of any restrictions on the relatedness of selections.

Conclusion

Rooting ability and 2nd year height are heritable traits in loblolly pine, and both

traits showed substantial clonal variation. There was a positive genetic correlation

between rooting and height at the parental, full-sib family, and clonal levels. Genetic

gains in rooting ability and 2nd year height are possible as demonstrated by a number of

deployment strategies. For hard-to-root species, like loblolly pine, a successful clonal


90%

80% ....

70% .

60%-

7 50% -

4 40%
S/
30%

20%

10% -


^' <^ C' O' 00 b'

Weight on 2nd Year Height I Weight on Rooting Ability
% Gain in 2nd Year Height (10%) % Gain in 2nd Year Height (1%)
% Gain in Rooting Ability (10%) -% Gain in Rooting Ability (1%)



Figure 4-3. Responses to selection in rooting ability and 2nd year height with various
selection indices for two clonal deployment options: 10% of clones selected
and 1% of clones selected.









forestry program based on rooted cuttings must consider both rooting ability and growth

when making selections. Because here these traits are positively correlated, selection for

one trait should lead to positive gains in the other. Or, early culling of clones based on

poor rooting will not negatively effect selection of clones for growth when field data

becomes available.

Clonal forestry is not a breeding method to develop better genotypes. However,

clonal forestry is a method to mass-produce well-tested genotypes. Short-term genetic

gains may be maximized through deployment of well-tested clones, but long-term gains

need to involve both clonal selection and recurrent selection for additive genetic variation

through repeated selection and breeding. Restrictions on the relatedness of selections

will be necessary when making selections for a breeding population or when deployment

will be traditional zygotic seedlings from seed orchards in order to reduce detrimental

effects of inbreeding depression.









Table 4-3. The number of full-sib families (half-sib families) and average number of
clones per full-sib family (half-sib family) selected from selecting 10% or 1%
of the top clones using the combined selection index.
10% Clones Selected 1% Clones Selected
nd Ave. # Ave. #
Weight on 2nd # FS # FS#
S Clones per Clones per
Year Height: Families FS Famiy Families Fs F
FS Family FS Family
Weight on (# HSS (# HS
Rooting Ability families) Fam ) families) F
Family) Family)
1:0 38 (26) 5.8(16.9) 14 (17) 1.6 (2.6)
0.9:0.1 41 (28) 5.3 (15.7) 15 (17) 1.5 (2.6)
0.8:0.2 43 (28) 5.1 (15.7) 12 (15) 1.8 (2.9)
0.7:0.3 43 (30) 5.1 (14.7) 13 (16) 1.7 (2.8)
0.6:0.4 45 (30) 4.9(14.7) 10 (12) 2.2 (3.7)
0.5:0.5 45 (29) 4.9(15.2) 12 (15) 1.8 (2.9)
0.4:0.6 45 (29) 4.9(15.2) 11 (15) 2.0 (2.9)
0.3:0.7 45 (29) 4.9(15.2) 11 (15) 2.0 (2.9)
0.2:0.8 45 (29) 4.9(15.2) 10(14) 2.2 (3.1)
0.1:0.9 45 (29) 4.9(15.2) 9 (13) 2.4 (3.4)
0:1 45 (29) 4.9(15.2) 9 (13) 2.4 (3.4)














CHAPTER 5
CONCLUSION

Loblolly pine is the most important commercial tree species in the southeastern

United States. Several forest industries in the southeastern United States are deploying

full-sib families of loblolly pine operationally. In addition, many of these companies are

pursuing clonal forestry programs with loblolly pine. Clones need to be well-tested

before they can be deployed. This involves the accumulation of reliable data from

propagation, growth traits, disease resistance, etc. Those well-tested clones that meet the

selection criteria can lead to substantial genetic gain.

Several key results can be concluded from this research. With rooting data from

2,200 clones from 70 full-sib families, the current study gives better estimates of genetic

components of variance for rooting than several previous studies. These results show a

great deal of genetic variation for rooting among families and clones of loblolly pine.

Combined with moderate to high estimates of family- and clonal-mean heritabilities and

type B correlations for rooting ability in different rooting trials, these results indicate the

potential for increasing rooting efficiency by selecting good rooting families and clones

or culling poor rooters.

Field testing of clones is a necessary component to clonal forestry programs based

on rooted cuttings. In addition, genetic field trials established with clones and seedlings

from the same full-sib families provide an opportunity for comparing both half-sib and

full-sib family performances for both propagules. Based on the results in Chapter 3,

several conclusions can be drawn. First, clonally replicated seedling trials of loblolly









pine provide genetic information with greater precision than zygotic seedlings. Second,

genetic correlations between propagule types for the growth traits were highly favorable.

These high genetic correlations between propagule types reassure that parental and full-

sib family rankings are stable regardless of propagule type. This implies that parental

and full-sib family rankings based on existing seedling progeny trials could be used to

select parents and families that perform well when they are deployed as rooted cuttings.

Third, little genotype x environment interaction was observed across sites at the parental,

family, and clonal level for all traits. However, there appears to be a carry-over effect

relating to the season in which the cuttings were rooted for the clonal material. Finally,

randomization is essential at all stages in testing when estimating genetic parameters.

The lack of randomization for the seedling population apparently resulted in a problem

with partitioning of the genetic variance, causing full-sib families to appear more

different and inflating estimates of dominance genetic variation.

Rooting ability and 2nd year height are heritable traits in loblolly pine, and both

traits showed substantial clonal variation. There was a positive genetic correlation

between rooting and height at the parental, full-sib family, and clonal levels. Genetic

gains in rooting ability and 2nd year height are possible as demonstrated by a number of

deployment strategies. For difficult-to-root species, like loblolly pine, a successful clonal

forestry program based on rooted cuttings must consider both rooting ability and growth

when making selections. Because here these traits are positively correlated, selection for

one trait should lead to positive gains in the other. Or, early culling of clones based on

poor rooting will not negatively affect selection of clones for growth when field data

become available.









Clonal forestry is not a breeding method to develop better genotypes. However,

clonal forestry is a method to mass-produce well-tested genotypes. Short-term genetic

gains may be maximized through deployment of well-tested clones, but long-term gains

need to involve both clonal selection and recurrent selection for additive genetic variation

through repeated selection and breeding. Restrictions on the relatedness of selections

will be necessary when making selections for a breeding population or when deployment

will be traditional zygotic seedlings from seed orchards in order to reduce detrimental

effects of inbreeding depression.



























APPENDIX A
LOBLOLLY PINE PARTIAL DIALLEL MATING DESIGN. THIRTY-TWO
PARENTS WERE CROSSED TO GENERATE 70 FULL-SIB FAMILIES.





















Parent 2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
1 X X X X
2 X X X X X
3 X X X
4 X X X
5 X X X X
6 X X X
7 X X
8 XX X
9 X X
10 X
11 X X
12 X X X
13 X X
14 X X X X X
15 X X
E 16 X X
17 X
18 X
19
20 X X X t
21 X X
22 X X X
23 X X
24 X
25 X X
26 X X
27 X
28 X X
29 X
30 X X
31 X
32

















APPENDIX B
VARIANCE COMPONENT ESTIMATES FOR ROOTING ABILITY

Table B-1. Observed variance component estimates for rooting of loblolly pine stem
cuttings from single-trial analyses.

SpringO 1 SummerO 1 Winter02 Spring02 Summer02
"2
TRAY 0.003977 0.005357 0.002891 0.010084 0.004796
"2
OGCA 0.005018 0.004629 0.005383 0.004451 0.003817
"2
SCA 0.002683 0.000771 0.001363 0.001868 0.00187
"2
aCLONE 0.041427 0.039236 0.024317 0.026364 0.023738
"2
-REPxFAM 0.000836 0.000559 0.000637 0.000661 0.001043
S 0.190089 0.177991 0.204829 0.200698 0.142
ErMOR 0.190089 0.177991 0.204829 0.200698 0.142
















Table B-2. Observed variance component estimates for rooting of loblolly pine stem cuttings from pair-wise-trial analyses.

SpringO 1 SpringO 1 SpringO 1 SpringO 1 SummerO 1 SummerO 1 SummerO 1 Winter02 Winter02 Spring02
SummerO 1 Winter02 Spring02 Summer02 Winter02 Spring02 Summer02 Spring02 Summer02 Summer02

-2 0.004751 0.003637 0.007712 0.004474 0.004115 0.007929 0.005059 0.006905 0.003774 0.007824
TRAY


(GCA 0.003434 0.002923 0.00433 0.003929 0.003191 0.002844 0.002557 0.003065 0.002347 0.003915

"2
(SCA 0.001409 0.000417 0.001262 0.001876 0.00067 0.000654 0.001027 0.001107 0 0.001147

"2
'CLONE 0.021605 0.011304 0.017015 0.014235 0.012645 0.014791 0.014151 0.011209 0.010850 0.013135
J^j 2


T-ESxGCA 0.001464 0.002034 0.000422 0.000688 0.001846 0.001380 0.001472 0.001609 0.002123 0.0004


OESTxFAM 0.000302 0.001758 0.001299 0.000539 0.000396 0.000572 0.000029 0.000423 0.001480 0.000485
00
2
TESTxCLONE 0.018862 0.021637 0.017991 0.022116 0.019564 0.018869 0.020504 0.014919 0.014083 0.012163


"2
REpxFAM 0.000674 0.000764 0.000703 0.000956 0.000598 0.000613 0.000797 0.000646 0.000825 0.000767


ROR 0.183321 0.198958 0.197306 0.162528 0.192458 0.192031 0.159335 0.202419 0.174684 0.177164
0E OR 0.183321 0.198958 0.197306 0.162528 0.192458 0.192031 0.159335 0.202419 0.174684 0.177164











Table B-3. Observed variance components for rooting ability from the across-trial
analysis using all five rooting trials.

Variance Component Estimate
^2
oRAY 0.005785
^2
GCA 0.002921
^2
aSCA 0.001062
^2
rCLONE 0.01638
^2
TESTrGCA 0.001364
^2
-TESTxFAM 0.000691
^2
OTESTxCLONE 0.017349
^2
(REPxFAM 0.000713
2E
ERROR 0.184907

















APPENDIX C
VARIANCE COMPONENT ESTIMATES FOR EARLY GROWTH TRAITS OF
LOBLOLLY PINE CLONES AND SEEDLINGS FROM THE SAME FULL-SIB
FAMILIES

Table C-1. Observed variance components for loblolly pine clones from the across-trial
analyses of 1st year height, 2nd year height, height increment, and crown width.
A separate error variance was modeled for each trial.

1st Year Height 2nd Year Height Height Crown Width
Increment

INc 53.4293 300.188 123.143 105.75
"2
GCA 17.4525 100.363 38.141 27.5703
"2
(SCA 2.9992 21.601 7.5808 4.93
"2
CLONE 32.3599 157.931 59.3416 56.4437
"2
TESTxGCA 4.1225 13.3866 7.9053 5.9029
"2
TESTxFAM 1.2003 3.7766 1.2451 1.3288
"2
TESTxCLONE 22.4253 82.7545 29.0673 23.4036
"2
TxCxGCA 0.4069 4.5692 2.942 2.2437
"2
TxCxFAM 0.0000008 0.3873 0.0000005 1.1495
"2
xCxCLONE 0.0000008 17.1395 14.179 10.7818
"2
REPxGCA 0.3275 1.0281 1.5638 0.6583
"2
REPxFAM 1.9242 5.511 1.5806 0.8574
^2
jEROR 512.142 1700.04 692.262 610.815
"2
ERRORB 279.24 853.321 427.247 306.51
"2
-ERRORc 527.513 2061.14 928.903 843.627

ERRORD 264.114 1110.76 864.549 329.026
"2
ERRORE 313.481 1124.11 682.649 423.413

OR 123.318 804.457 527.064 296.22
L-RORF 123.318 804.457 527.064 296.22











Table C-2. Observed variance components for loblolly pine seedlings from the across-
trial analyses of 1st year height, 2nd year height, height increment, and crown
width. A separate error variance was modeled for each trial.

1st Year Height 2nd Year Height Height Crown Width
Increment
"2
rGCA 3.6682 29.9688 12.9402 13.7566
"2
SCA 4.3654 20.8173 10.4732 5.0053
"2
TESTxGCA 1.5235 8.4371 8.6503 3.8956
"2
TESTxFAM 8.5693 25.1585 8.5047 7.9082
"2
TxCxGCA 0.0000007 0.0000004 0.0000002 0.5247
i2
TxCxFAM 0.6036 0.0000001 1.375 1.0372
"2
REPxGCA 0.0000001 0.0000001 0.0000001 1.2412
"2
oREPxFAM 6.3223 55.867 26.1737 15.4417

RROR2 599.229 1621.95 713.814 590.37

ERRORB 361.489 1104.24 567.689 325.987
"2
ERRORc 508.95 2524.33 1246.59 1110.51
"2
ERRRD 294.534 1356.8 1059.94 501.614
"2
ERRORE 379.556 1766.73 1109.52 610.975

OR 101.715 755.187 585.714 341.353
L-RORF 101.715 755.187 585.714 341.353
















APPENDIX D
VARIANCE COMPONENT ESTIMATES FROM THE BIVARIATE ANALYSES OF
ROOTING AND 2ND YEAR HEIGHT

Table D-1. Observed variance component estimates from the bivariate analyses of
rooting ability from SpringOl and 2nd year height from each of the field trials.

SpringO1 Field A Field B Field C Field D Field E Field F

iNCr, 0.0039 0.0039 0.004 0.0039 0.004 0.004

Ui2 219.85 232.03 801.24 390.46 105.53 89.22
"2
GCArt 0.0051 0.0052 0.0052 0.0051 0.0052 0.0051

-GCAoot,, 0.0778 0.219 0.2515 0.3664 0.3268 0.1727
"2
GCA,,g 93.85 77.71 134.98 134.81 144.59 98.15

S-C, 0.0027 0.0026 0.0026 0.0027 0.0026 0.0027

SCAroo,,gh 0.0654 -0.1556 -0.0148 0.0203 0.0815 0.1294
"2
SCA,,,,, 28.34 3.7 8.13 28.18 43.77 30.08

CLONEt 0.0415 0.0415 0.0414 0.0415 0.0414 0.0414

CLONEroo,_tgi, -0.21 0.3412 0.4084 0.3997 0.2865 0.7351
"2
CLONEh,,ght 386.08 191.46 250.52 230.24 426.68 207.67

REPxFAMt 0.0008 0.0008 0.0008 0.0008 0.0008 0.0008
"2 0.0996 5.31 23.68 15.47 24.33 15.44
"2
OERRORo 0.1901 0.1901 0.1901 0.1901 0.1901 0.1901
2OR 1641.17 879.86 2003.72 1127.51 1072.05 834.69
E ORhgt, 1641.17 879.86 2003.72 1127.51 1072.05 834.69











Table D-2. Observed variance component estimates from the bivariate analyses of
rooting ability from SummerO 1 and 2nd year height from each of the field
trials.

Summer01 Field A Field B Field C Field D Field E Field F
"2
JiNC 0.0053 0.0053 0.0053 0.0053 0.0053 0.0053

CU2 220.06 232.4 801.8 389.37 105.55 89.37
"2
GCAot 0.005 0.005 0.005 0.005 0.005 0.0049

'GCAoot, 0.243 0.1731 0.2933 0.3457 0.2579 0.0796
"2
GCA,, t 98.15 79.12 135.89 138.56 147.16 98.3

2sCA 0.0007 0.0007 0.0007 0.0007 0.0007 0.0008

sCAo,,, ,, 0.0397 0.0141 -0.0125 0.0078 0.0616 0.1556
"2
sCA,,,g 26.74 3.33 6.33 24.69 40.85 26.5
^2
CLONEt 0.0393 0.0392 0.0393 0.0393 0.0392 0.0393

CLONErootgi -0.0944 -0.0803 0.3137 0.4131 0.2729 0.2437
2
'CLONE,,gh, 385.69 190.78 250.56 230.58 426.65 206.64
2
REPxFAMt 0.0006 0.0006 0.0006 0.0006 0.0006 0.0006
U2 0.0757 5.3 23.92 15.69 24.4 15.71

"2
OERRORo 0.1779 0.1779 0.1779 0.1779 0.1779 0.1779
OR 1641.1 879.8 2003.35 1127.44 1072.03 834.0
OEMRORho gh 1641.1 879.8 2003.35 1127.44 1072.03 834.0