<%BANNER%>

Genotype X Environment Interactions in Selected Loblolly (Pinus taeda L.) and Slash Pine (p. elliottii Engelm. Var. elli...

Permanent Link: http://ufdc.ufl.edu/UFE0024254/00001

Material Information

Title: Genotype X Environment Interactions in Selected Loblolly (Pinus taeda L.) and Slash Pine (p. elliottii Engelm. Var. elliotttii) Plantations in the Southeastern United States
Physical Description: 1 online resource (158 p.)
Language: english
Creator: Roth, Brian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: biomass, genotype, gxe, loblolly, slash
Forest Resources and Conservation -- Dissertations, Academic -- UF
Genre: Forest Resources and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Few studies have quantified the combined effects of silvicultural treatment, planting density, and genetic improvement in full-sib family blocks of loblolly (Pinus taeda L.) and slash pine (P. elliottii Engelm. var. elliottii). This information is critical for the proper deployment of improved genotypes of southern pine and ultimately how they respond to specific silvicultural treatments in a changing climate across a range of soils. This study employed a series of replicated factorial experiments and family block plantings in Florida and Georgia which manipulated gradients in planting density (1334 versus 2990 trees ha-1), understory competition and soil nutrient availability. Age-two loblolly pine accumulation and distribution of biomass and nitrogen were more responsive to treatment than slash pine. Total slash pine biomass accumulation varied from 4.7 to 11.6 Mg ha-1 with responses limited to the main effects of silviculture, density, and family. Loblolly pine total biomass accumulation varied from 8.6 to 15.5 Mg ha-1, with several interactions: silviculture x density, silviculture x location, and family x density. Loblolly pine distribution of biomass was influenced by: family x density, family x silviculture and silviculture x density x location. Total nitrogen content ranged from 33.0 to 83.3 kg ha-1 and varied with silviculture, family and density x location for slash pine. Loblolly pine total N accumulation ranged from 23.3 to 101.5 kg ha-1 with two interactions: family x density and silviculture x density x location. At age five, families varied in needlefall, leaf area index, fraction of light intercepted, light extinction coefficient, intercepted photosynthetically active radiation (IPAR) and radiation use efficiency (RUE). There were small differences among families in RUE (1.08 to 1.16 g MJ-1 PAR), although they were stable across locations. Age five basal area and standing volume for both species demonstrated significant interactions: family x location, family x silviculture and silviculture x density. Family x silviculture interactions were positive and the best overall families responded the greatest to intensive silviculture. This research demonstrates that genotype x environment interactions exist in certain southern pine families. These results reinforce the need for understanding the biotic and abiotic mechanisms that drive these interactions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Brian Roth.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Jokela, Eric J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0024254:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024254/00001

Material Information

Title: Genotype X Environment Interactions in Selected Loblolly (Pinus taeda L.) and Slash Pine (p. elliottii Engelm. Var. elliotttii) Plantations in the Southeastern United States
Physical Description: 1 online resource (158 p.)
Language: english
Creator: Roth, Brian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: biomass, genotype, gxe, loblolly, slash
Forest Resources and Conservation -- Dissertations, Academic -- UF
Genre: Forest Resources and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Few studies have quantified the combined effects of silvicultural treatment, planting density, and genetic improvement in full-sib family blocks of loblolly (Pinus taeda L.) and slash pine (P. elliottii Engelm. var. elliottii). This information is critical for the proper deployment of improved genotypes of southern pine and ultimately how they respond to specific silvicultural treatments in a changing climate across a range of soils. This study employed a series of replicated factorial experiments and family block plantings in Florida and Georgia which manipulated gradients in planting density (1334 versus 2990 trees ha-1), understory competition and soil nutrient availability. Age-two loblolly pine accumulation and distribution of biomass and nitrogen were more responsive to treatment than slash pine. Total slash pine biomass accumulation varied from 4.7 to 11.6 Mg ha-1 with responses limited to the main effects of silviculture, density, and family. Loblolly pine total biomass accumulation varied from 8.6 to 15.5 Mg ha-1, with several interactions: silviculture x density, silviculture x location, and family x density. Loblolly pine distribution of biomass was influenced by: family x density, family x silviculture and silviculture x density x location. Total nitrogen content ranged from 33.0 to 83.3 kg ha-1 and varied with silviculture, family and density x location for slash pine. Loblolly pine total N accumulation ranged from 23.3 to 101.5 kg ha-1 with two interactions: family x density and silviculture x density x location. At age five, families varied in needlefall, leaf area index, fraction of light intercepted, light extinction coefficient, intercepted photosynthetically active radiation (IPAR) and radiation use efficiency (RUE). There were small differences among families in RUE (1.08 to 1.16 g MJ-1 PAR), although they were stable across locations. Age five basal area and standing volume for both species demonstrated significant interactions: family x location, family x silviculture and silviculture x density. Family x silviculture interactions were positive and the best overall families responded the greatest to intensive silviculture. This research demonstrates that genotype x environment interactions exist in certain southern pine families. These results reinforce the need for understanding the biotic and abiotic mechanisms that drive these interactions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Brian Roth.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Jokela, Eric J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0024254:00001


This item has the following downloads:


Full Text

PAGE 1

1 GENOTYPE X ENVIRONMENT INTERACTIONS IN SELECTED LOBLOLLY ( Pinus taeda L.) AND SLASH PINE ( P elliottii Engelm. var. elliotttii) PLANTATIONS IN THE SOUTHEASTERN UNITED STATES. By BRIAN EDWARD ROTH 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 2010

PAGE 2

2 2010 Brian Edward Roth

PAGE 3

3 Man always kills the thing he loves, ... and so we the pioneers have killed our wilderness. Some say we had to. Be that as it may, I am glad I shall never be young without wild country to be young in. Aldo Leopold, A Sand County Almanac 1949 I' m truly sorry man's dominion, H as broken nature's social union, Robert Burns To A Mouse, On Turning Her Up In Her Nest With The Plough. 1785 What do we know? How do we know it? And how do we apply it? Congressman Greg Walden (R OR), Subcommittee on Forests and Forest Health Chairman. 24th February 2006

PAGE 4

4 ACKNOWLEDGMENTS Many, many people and organizations have contributed in some way to my PhD program and while this space is too small to properly recognize everyone, I will make an attempt. My apologies to anyone who was missed; I could not have done this alone. I wish to thank my major advisor, Dr. Eric Jokela, who gave me the freedom to be curious and explore while at the same time keeping me on track whenever I became distracted. Special thanks to Drs.: Tim Martin for helping me dis tinguish sense from nonsense, Ken Portier for his excitement and encouragement, Wendell Cropper for expanding the scope of my work, Shibu Jose for his professionalism, Clint Slatton for demonstrating how to be an over achiever in the face of adversity Ti mothy White for brin g ing me to Gainesville, Dudley Huber for always having time to discuss the structure of a statistical model, Gary Peter f or exploring some new territory together PK Nair for his interest and support, McNair Bostick and Carmen Valero Ar acama for their strength, courage and integrity. Last, but not least, the entire Faculty and Staff at the School of Forest Resources and Conservation for their collegiality and friendship. My efforts here were made possible by members of the Forest Biolog y Research Cooperative (FBRC) and the Institute of Food and Agricultural Sciences (IFAS) at the University of Florida. Many people assisted with trial design, installation, data collection and sampling in the field including: Marshall Jacobson, Robbie Cannon, Steve Kennerly, David Adams Mike Cunningham Alan Wilson Ben Cazell Edgar Barrs, Greg Powell, Duncan Wilson, Angela Kegley, Andreas Coveny, Debbie Prokop, Bobby Hutchinson, Chris Cabrera, Emilio Ancaya, Vanessa Tischler, Vance Overton, Paul Proctor and Kathy Slifer. Finally I wish to thank my parents for their support Dorota for all her sacrifices, and my children for their never ending love

PAGE 5

5 TABLE OF CONTENTS ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .......................................................................................................................11 ABSTRAC T ...................................................................................................................................13 CHAPTER 1 INTRODUCTION ..................................................................................................................15 Background .........................................................................................................................15 Problem ...............................................................................................................................17 Approach and Objectives ....................................................................................................19 2 INFLUENCE OF GENOTYPE AND ENVIRONMENT ON BIOMASS ACCUMULATION, DISTRIBUTION AND NITROGEN CONTENT IN SELECTED LOBLOLLY ( Pinus taeda L.) AND SLASH PINE ( P. elliotti E NGELM VAR elliotttii) PLANTATIONS IN THE SOUTHEASTERN UNITED STATES ........................21 Introduction .........................................................................................................................21 Treatmen t Descriptions ...................................................................................................23 Sampling Procedures .......................................................................................................26 Results .................................................................................................................................29 Biomass Acc umulation ....................................................................................................29 Biomass Distribution to Foliage, Bole, and Belowground ..............................................31 Nitrogen Content .............................................................................................................32 Discussion ...........................................................................................................................34 Conclusions .........................................................................................................................41 3 FAMILY DIFFERENCES IN LIGHT INTERCEPTION AND RADIATION USE EFFICIENC Y IN SELECTED LOBLOLLY ( Pinus taeda L.) PLANTATIONS IN THE SOUTHEASTERN UNITED STATES .................................................................................57 Introduction .........................................................................................................................57 Materials and Methods ........................................................................................................62 Experimental Site and Design .........................................................................................62 Inventory and Aboveground Biomass Estimates ............................................................64 Needlefall, Leaf Area Index, and Aboveground Net Primary Production ......................64 Fraction of PAR Intercepted, Light Extinction Coefficient ............................................65 Amount of PAR Intercepted and Radiation Use Efficiency ............................................67 Statistical Analyses ..........................................................................................................67 Results .................................................................................................................................69 Needlefall, Leaf Area Index, and Aboveground Net Primary Production ......................69

PAGE 6

6 Fraction of PAR Intercepted, Light Extinction Coefficient ............................................70 Amount of PAR intercepted and Radiation Use Efficiency ............................................71 Discussion ...........................................................................................................................71 Conclusions .........................................................................................................................78 4 GENOTYPE X ENVIRONMENT INTERACTIONS IN SELECTED LOBLOLLY ( Pinus taeda L.) AND SLASH PINE ( P. elliottii ENGLM. VAR. elliottii) PLANTATIONS IN THE SOUTHEASTERN UNITED STATES. ......................................86 Introduction .........................................................................................................................86 Methods ...............................................................................................................................89 Study Description ............................................................................................................89 Experimental Design .......................................................................................................90 Treatment Descriptions ...................................................................................................90 Inventory, Yield and Biom ass Estimates .........................................................................93 Analysis ...........................................................................................................................95 Results .................................................................................................................................97 Genotype x Site Interactions ...........................................................................................97 Genotype x Silviculture Interactions ...............................................................................98 Silviculture x Density Interactions ..................................................................................99 Location x Density Interactions .......................................................................................99 Species and Deployment Interactions ...........................................................................100 Effects of Disease and Hurricanes .................................................................................100 Discussion .........................................................................................................................101 Genotype x Silviculture .................................................................................................101 Genotype x Location .....................................................................................................102 Silviculture x Density ....................................................................................................104 Species and Deployment Interactions ...........................................................................105 Conclusions .......................................................................................................................105 5 SUMMARY AND FUTURE RESEARCH ..........................................................................122 Summary ...........................................................................................................................122 Biomass Production, Distribution, and Nitrogen Content at Age Two .........................122 Interception and Efficiency of PAR at Ages Four and Five ..........................................124 Basal Area, Stem Volume and Biomass Production at Age Five ..................................125 Focus of Future Research ..................................................................................................127 APPENDIX A GLOSS ARY OF TERMS, UNITS AND DESCRIPTIONS ................................................130 B HISTORICAL ANNUAL RAINFALL, AVERAGE TEMPERATURE, MINIMUM TEMPERATURE, MAXIMUM TEMPERATURE DATA BY LOCATION ....................131 C GENERAL SOIL PROPERTIES OF THE FIVE PPINES EXPERIMENTAL LOCATIONS IN SOUTHEAST GEORGIA AND NORTHEAST FLORIDA ..................133

PAGE 7

7 D RAW DATA UTILIZED TO DEVELOP ALLOMETRIC EQUATIONS .........................134 LIST OF REFERENCES .............................................................................................................145 BIOGRAPHICAL SKETCH .......................................................................................................158

PAGE 8

8 LIST OF TABLES 21 Characteristics of the five PPINES experimental locations in southeast Georgia and nor theast Florida ................................................................................................................42 22 Elemental application rates of fertilizers supplied to the PPINES locations through the end of the second growing season. ...............................................................................42 23 Parameter estimates and standard errors of the estimates for foliage, bole and belowground biomass equations for twoyear old loblolly and slash pine. .......................43 24 Summary of statistical significance to test loblolly and slash pine foliage, branches, bark, stemwood, belowground, and total biomass accumulation at age two years. ..........44 25 Summary of statistical significance to test loblolly and slash pine distribution of biomass to foliage, stemwood, and belowground components at age two y ears. .............45 26 Least squared means for slash pine nitrogen accumulation among various biomass components as influenced by full sib family at age two years. .........................................46 27 Slash pine nitrogen accumulation in foliage, aboveground and total biomass components by planting density and location at twoyears. ..............................................46 28 Loblolly pine nitro gen accumulation in foliage, aboveground, belowground, and total biomass components by family and planting density at twoyears. ...................................47 29 Loblolly pine nitrogen accumulation in foliage, aboveground, belowground, and total biomass components by silvicultural treatment, planting density, and location at age two years. ...........................................................................................................................48 31 Cumulative elemental nutrient application rates for the PPINE S intensive silvicultural treatments through five growing seasons. ......................................................79 32 Summation of incoming monthly shortwave radiation per month in MJ m2, for two locations and two years ......................................................................................................79 33 Least squares means for needlefall, light extinction coefficient, amount of PAR intercepted, aboveground net primary productivity, and radiation use efficiency for six fullsib loblolly pine families averaged across two locations and two years. ..............79 34 Least squares means for needlefall, leaf area index, light extinction coefficient, fraction of PAR intercepted, amount of PAR intercepted, aboveground net primary productivity for the interaction between location and year ..............................................80 35 Least squares means for Leaf Area Index (LAI) for six full sib loblolly pine families averaged across two locations ............................................................................................80 41 Characteristics of the PPINES experimental locations. ...................................................107

PAGE 9

9 42 Cumulative elemental nutrient application ra tes for the PPINES intensive silvicultural treatments through five growing seasons .....................................................107 43 Description of biomass harvest data used to develop the allometric e quations ...............108 44 Parameter estimates and standard errors of the estimate aboveground bio mass equati ons for loblolly and slash pine ...............................................................................109 45 Summary of statistica l significance and associated degrees of freedom from ANOVA to test loblolly pine basal area, stem volume and aboveground biomass at age two threeand five years. .......................................................................................................110 46 Summary of statistical significance and associated degrees of freedom from ANOVA to test slash pine basal area, stem volume and aboveground biomass at age two, threeand five years. .......................................................................................................112 47 Variance component s and associated statistical significance for individual mode l results in tables 4 5 and 46 .............................................................................................114 48 Summary of mensurational characteristics by species, silvicultural treatment and planting density at age two three and five years. .........................................................115 49 Age five contrasts between slash pine families grown in mixtures versus grown in pure plots. .........................................................................................................................116 A 1 Glossary of terms, units and descriptions. .......................................................................130 B 1 Historical annual rainfall, average temperature, minimum temperature, maximum temperature data by location. ...........................................................................................131 B 2 Palmer Drought Severity Index annual values for three regions in FL and GA. .............132 C 1 General soil properties of the five PPINES ex perimental locations in southeast Georgia and northeast Florida. .........................................................................................133 D 1 Raw data utilized to develop allometric equations from the Bunnell, FL location at age two ............................................................................................................................134 D 2 Raw data utilized to develop allometric equations from the Perry, FL location at age two ....................................................................................................................................135 D 3 Raw data utilized to develop allometric equations from the Waldo, FL location at age two ....................................................................................................................................136 D 4 Raw data utilized to develop allometric equations from the Sanderson, FL location at age two ............................................................................................................................137 D 5 Raw data utilized to develop allometric equations from the Waverly, GA location at age two ............................................................................................................................138

PAGE 10

10 D 6 Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, operational silviculture and family L4. .................................................................139 D 7 Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, operational silviculture and family L7. .................................................................140 D 8 Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, intensive silviculture and family L4. .....................................................................141 D 9 Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, intensive silviculture and family L7. .....................................................................142 D 10 Raw data utilized to dev elop allometric equations from the Waverly, GA location at age 5, intensive silviculture and family L4. .....................................................................143 D 11 Raw data utilized to develop allometric equations from the Waverly, GA location at age 5, intensive silviculture and family L7. .....................................................................144

PAGE 11

11 LIST OF FIGURES 21 Typical two year old slash pine that was harvested from the intensive silvicu lture treatment at the Waldo, FL location ..................................................................................49 22 Least squares means for the main effect of full sib family on slash pine biomass accumulation in foliage bark and branches, stemwood, and belowground components at age two years .............................................................................................50 23 Least squares means for the main effect of silvicultural intensity on loblolly pine total biomass accumulation as influenced by planting density and location.. ...................51 24 Least squares means for loblolly pine total biomass accumulation demonstrating an interaction between family and planting density ...............................................................52 25 Least squares means for the main effect of full sib family on slash pine biomass distribution to foliage, stemwood, and belowground components ....................................53 26 Least squares means for the main effect of full sib family on loblolly pine biomass distribution to foliage stemwood, and belowground components ....................................54 27 Loblolly pine biomass distribution to foliage stemwo od and belowground components demonstrating a three way interaction ..........................................................55 28 Belowground nitrogen content at age twoyears demonstrating a significant three way interaction between silvicultura l intensity, planting density, and location ................56 32 Relationship between annual aboveground net primary production and intercepted radiation for two loblolly pine full sib families .................................................................82 33 Relationship between RUE and tree density at age five years for lob lolly pine ...............83 34 Relationship between fraction of light intercept ed and light extinction coefficient by month of year for two locations. ........................................................................................84 35 Relationship between the fraction of light intercepted by month of year for six loblolly pine families .........................................................................................................85 41 Standing crop biomass at age five demonstrating a genotype x location interaction for loblolly pine and slash pine .......................................................................................117 42 Standing volume at age five demonstrating a genotype x silviculture interaction for loblolly pine and slash pine ..............................................................................................118 43 Aboveground biomass accretion by silvicultural treatment for loblolly pine at Sanderson, FL, Waverly, GA, and slash pine across both slash pine locations...............119

PAGE 12

12 44 Standing crop biomass at age five demonstrating a species x silviculture interaction for loblolly and slas h pine ................................................................................................120 45 I ncidence of fusiform rust and wind damage per plot at age five for slash pine demonstrating a significant genotype x location interaction for slash pine .....................121

PAGE 13

13 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 GENOTYPE X ENVIRONMENT INTERACTIONS IN SELECTED LOBLOL LY ( Pinus taeda L.) AND SLASH PINE ( Pinus elliottii Englm. var. elliottii) PLANTAT IONS IN THE SOUTHEASTERN UNITED STATES. By Brian Edward Roth December 2010 Chair: Eric J. Jokela Major: Forest Resources and Conservation Few studies have quantified the c ombined effects of silvicultural treatment, planting density, and genetic improvement in full sib family blocks of loblolly ( Pinus taeda L ) and slash pine ( P elliottii Engelm. var. elliottii) This information is critical for the proper deployment of improved genotypes of southern pine and ultimately how they respond to specific silvicultural treatments in a changing climate across a range of soils This study employ ed a series of replicated factorial experiments and family block plantings in Florida and Georgia which manipulat ed gradients in planting density (1334 versus 2990 trees ha1Age two loblolly pine accumulation and distribution of biomass and nitrogen were more responsive to treatment than slash pine. Total slash pine biomass accumulation varied from 4.7 to 11.6 Mg ha ) understory competition and soil nutrient availability 1 with responses limited to the main effects of silviculture, density, and family. Loblolly pine total biomass accumulation varied from 8.6 to 15.5 Mg ha1, with several in teractions: silviculture x density, silviculture x location, and family x density. Loblolly pine distribution of biomass was influenced by: family x de nsity, family x silviculture and silviculture x density x location. Total nitrogen content ranged from 33.0 to 83.3 kg ha1 and varied with

PAGE 14

14 silviculture, family and density x location for slash pine. Loblolly pine total N accumulation ranged from 23.3 to 101.5 kg ha1At age fiv e, families varied in needlefall, leaf area index, fraction of light intercepted, light extinction coefficient, intercepted photosynthetically active radiation (IPAR ) and radiation use efficiency (RUE). There were small differences among families in RUE ( 1.08 to 1.16 g MJ with two interactions: family x density and silviculture x density x location. 1Age five b asal area and standing volume for both species demonstrated significant interactions: family x location, family x silvicultur e and silvicultur e x density. Family x silvicultur e interactions were positive and the best overall families respond ed the greatest to intensive silviculture This research demonstrates that genotype x environment interactions exist in certain southern pine families. These results reinforce the need for understanding the biotic and abiotic mechanisms that drive these interactions. PAR ), al though they were stable across locations

PAGE 15

15 CHAPTER 1 INTRODUCTION Background The southeastern United States contains some of the most intensively managed pine plantations in the world (Fox et al. 2007). C onsiderable gains in the productivity of loblolly ( Pinus taeda L.) and slash pine ( Pinus elliottii Englm. var. elliottii) plantations have been achieved over the past 30 years and these plantations now produce a wide range of ecosystem services, such as fiber, carbon sequestration, and biofuels Growth responses to intensive silvicultural practices (Fox et al. 2007) typically range from 2to 3.5 fold for loblolly pine in the southeastern USA (Jokela et al. 2004) While the absolute amount of biomass produced is i mportant, so too is the distribution of biomass and nutrients to various components such as the foliage, branches, bark, bole, and belowground, particularly when new sources of bioenergy are concerned (Scott and Dean, 2006). Demonstrated increases in unit area production have been realized through combinations of silvicultural treatments such as site preparation, fertilization, competition control, and density management (Colbert et al. 1990; Albaugh et al. 1998; Martin and Jokela, 2004b; Fox et al. 2007) Preparation of the planting site prior to plantation establishment has demonstrated benefits on seedling survival in the short term and stand growth and development in the long term (Nilsson and Allen, 2003). On flatwood sites in the southeastern Unit ed States, site preparation treatments generally involve combinations of bedding and chemical control of competing vegetation. Collectively, these treatments have had a substantial effect on longterm growth responses (Jokela et al. 2000; Zhao et al. 2008). From a mechanistic perspective, post planting management of associated vegetation has led to large gains in plantation productivity as site resources such as soil moisture and nutrient supply is made available to the crop trees (Roth

PAGE 16

16 and Newton, 1996; Miller et al. 2003). Competing vegetation can be controlled either mechanically or chemically and long term gains in wood volume in southern pine plantations have ranged from 14 to 5840% (Wagner et al. 2006). Since the early 1970s, forest fertilization has been a common management practice throughout much of the southern United States. During the period from 1969 to 2004, over 6.5 million ha of southern pine stands were fertilized in the southeastern United States mainly with nitrogen and phosphor us (Albaugh et al. 2007). Fertilization treatments are aimed at meeting the nutrient demands of rapidly growing plantations on sites where inherent nutrient supplies are low (Comerford et al. 2006). Nutrient availability is often adequate during the fi rst few years following plantation establishment because of enhanced decomposition and mineralization of organic matter H owever, by the time of crown closure, nutrients become limiting due to decreases in soil supplies and immobilization of nutrients in microbial biomass and woody vegetation in addition to the increased demands of these rapidly growing trees Loblolly pine tends to be more responsive to fertilization treatments than slash pine plantations (Jokela et al. 2000). Historically, southern pi ne plantations have been established at densities ranging from about 1500 to 2240 trees ha1 and subsequently thinned to between 625 and 1000 tre e s ha1 at around age 10 to 14 years. Recently, there has been a trend in the southeastern United States towar ds the establishment of plantations at lower densities with the goal of producing chipnsaw or sawtimber without thinning (Huang and Kronrad, 2004). However, associated with these shifts in initial planting density, come implications for nutrient management (Barron Gafford et al., 2003), the efficiency of resource use (Will et al. 2001), and the deployment of elite genotypes (Fang et al. 1999).

PAGE 17

17 Advances in tree breeding over the past 50 years have been focused on improving volume growth, disease resistance, tree form and wood quality (White et al. 2007). Plantations established with firstgeneration seed developed by tree breeding programs ha ve demonstrated 8 to 12% gains in unit area volume production at harvest age (Squillace, 1989). Gains in volume production due to the establishment of plantations with secondgeneration seed sources are estimated to be in the range of 10 to 30% over unimproved sources (Li and McKeand, 1989; McKeand et al. 2003a) When a combination of elite genetic materials are combined with site specific silvicultural treatments, mean annual increments of up to 20 m3ha1yr1Problem have been documented (Allen et al. 2005c) Most plantations in the Southeast are currently established using open pollinated, half sib families; however full sib family blocks are becoming more common and deployment of clones via rooted cuttings or somatic plantlets are moving to become an operational reality (Sutton, 2002). Genotypes may respond differentially to silvicultural treatments, locations and climatic conditions. This genotype x environment interaction is more likely to occur given increasing levels of tree improvement (i.e. secondgeneration breeding cycles), the deployment of more uniform genotypes (i.e. full sib families and clona l materials) (McKeand et al. 2006), and especially when deployed in combinations of intensive silvicultural management (Roth et al. 2007a). These genotype x environment interactions may be manifest as rank changes among genotypes when deployed across na tural or man made environmental gradients in resource availability, or as scale effects in which the absolute differences among genotypes change with environment (Knight, 1970). Tree breeding programs have recognized that genotype x environment interacti ons are a possibility, which could dictate a need for site specific breeding efforts (McKeand et al. 1997b).

PAGE 18

18 The general approach has been the deployment of a large number of genotypes across a wide range of sites, under fairly uniform silvicultural trea tments aimed at minimizing within site variation. In this situation it is important to test how genotypes respond to environmental variation due to localized edaphic, climatic and disease conditions, since there is a desire to move genotypes long distance s from their origins in order to increase yields (Lambeth et al. 2005). Often these tests are evaluating the performance of progeny in single tree plots and rarely has the field performance of these elite genotypes been verified using pure family block plantation trials (Vergara et al. 2004; Dean, 2007). R esearch studies aimed at quantifying the response of southern pines with the combined effects of silvicultural treatments and genetic improvement in loblolly and slash pine in pure family block planting s are limited and documented genotype x silviculture interactions on stand level attributes have not always been evident. For example, in a 12year old loblolly pine family x vegetation control study in Georgia, no genotype x environment interaction was f ound for total standing volume (Martin and Shiver, 2002). Also, loblolly pine whole tree biomass at age five in a genotype x fertilization experiment in North Carolina did not demonstrate a significant interaction (Retzlaff et al. 2001). However, in a 5 year old loblolly pine experiment, families interacted with silviculture, planting density and location for basal area, unit area stem volume and aboveground biomass (Roth et al. 2007a) and in the same trial there was a family x density interaction for w ood quality traits (Roth et al. 2007b). The current research effort is driven by the need to supply information to resource managers who are faced with difficult decisions about where to deploy elite genetic materials, given the understanding that these g enotypes may respond differently to silvicultural treatments, locations, and changing climatic conditions. Given this information, site specific and

PAGE 19

19 silviculture specific treatment prescriptions for individual genotypes could be developed. In addition, t his information may influence breeding strategies in order to capture the full advantage of unique genotypes that respond differentially with environmental conditions. The work described in this investigation is aimed at partially filling this need by exa mining how genotype x environment interactions influence biomass accretion, distribution, nitrogen content, stem volume, basal area, and radiation use efficiency in selected full sib families of loblolly and slash pine in the southeastern United States. Ap proach and Objectives Th e approach taken in this investigation was to use various subsets of a large series of replicated factorial experiments which were established on multiple locations in Florida and Georgia to provide insight into understanding genoty pe x environment interactions in southern pines. The experimental design was unique in that it used large full sib family block plots, where stand level attributes could be investigated across combinations of silvicultural management intensity, planting density, and locations. The larger trial series consisted of: contrasting silvicultural treatment intensities (operational versus intensive), planting densities (1334 versus 2990 trees ha1The overall objectives of this study were to investigate and quantify the magnitude and nature of genotype x environment interactions for selected elite full sib families of loblolly and slash pine across a range of contrasting locations on stand level traits Specific objectives were to document significant interactions between th e factors of: genotype x silviculture, genotype x density, genotype x location, silviculture x density, silviculture x location, location x density, and any higher level interactions among these factors for biomass accretion, distribution, nitrogen ) and elite fullsib families (seven for loblolly and six for slash pine). These treatments were designed to provide large contrasts in planting density, competition from associated vegetation, soil nutrient availability and genetic selection

PAGE 20

20 content stem volume, basal area, and radiation use efficiency. Results are pre sented in three main chapters: C hapter two biomass production, distribution and nitrogen content at age two, C hapter three interception and efficiency of PAR at ages four and five and C hapter four basal area, stem volume and biomass production at age five. This research provides a valuable contribution towards the management of intensively managed loblolly and slash pine plantations in the southeastern United States, in that it is the only one of its kind where the combined effects of genotype, silviculture, and planting density can be examined singly or in combination across a range of site conditions in a family block planting design. This information will aid r esource manager s through an understanding of how elite genotypes respond on a unit area basis to intensive silvicultural treatments when deployed across man made and naturally occurring environmental gradients

PAGE 21

21 CHAPTER 2 INFLUENCE OF GENOTYP E AND ENVIRONMENT ON BIOMASS ACCUMULATION, DISTRIBUTION AND NIT ROGEN CONTENT IN SEL ECTED LOBLOLLY ( Pinus taeda L.) AND SLASH PINE ( P. elliotti ENGELM. VAR. elliotttii) PLANTATIONS IN THE SOUTHEASTERN UNITED STATES Introduction The forested portion of the southeastern United States contains almost one half of the world's industrial forest plantations and produces more industrial timber than any other region of the world (Allen et al. 2005c). Combinations of species selection, silvicultural management intensity, planting density, and deployment of improved genotypes have greatly improved the productivity of these plantations over the past several decades (Jokela et al. 2004). These managed plantations now produce a wide range of ecosystem services, such as fiber, carbon sequestration and biofuels Much of the response to these treatments occurs early in the development of these stands when rapid shifts in the accumulation and distribution of biomass among tree components occurs (Adegbidi et al. 2005). During this early stage of st and development, biomass is preferentially accumulated in the foliage (FOL) and belowground (BELOW) components, at the expense of stemwood (BOLE) (Adegbidi et al. 2002). As stand development progresses this trend reverses, with the bole becoming a more s ignificant component over time (Martin and Jokela, 2004b). Quantification of accumulation and distribution of biomass and nutrients in various tree components such as the FOL, branches (BRANCH), bark (BARK), BOLE, and BELOW is important for predicting and managing alternative sources of bioenergy (Scott and Dean, 2006) and in the development of process based growth models (Landsberg and Waring, 1997; Zhang and Borders, 2004). Nitrogen is the most common nutrient limiting in these systems and was examined i n this study. M uch of the early research on the distribution of biomass in southern

PAGE 22

22 pine species has been focused on individual seedlings (Ledig et al. 1970; Johnson, 1990; Li et al., 1991b; Jose et al. 2003), or trees (Neary et al. 1990), and where unit area production has been examined (Jokela et al. 1989), it often does not include the BELOW component (Jokela and Martin, 2000), with a few exceptions (Gholz and Fisher, 1982; Adegbidi et al ., 2005). Shifts in biomass distribution over time have been a ttributed to several factors, including inherent site productivity (Keyes and Grier, 1981; Stape et al. 2004), soil properties (Burkes et al., 2003), nutrient additions (Keyes and Grier, 1981; Axelsson and Axelsson, 1986; Gower et al., 1992; Haynes and Go wer, 1995; Albaugh et al. 1998; du Toit, 2008), irrigation (Axelsson and Axelsson, 1986; Gower et al. 1992; Samuelson et al. 2004), stand density (Litton et al. 2003; Burkes et al. 2003), weed control (Britt et al. 1990), stand structure (Vanlear and Kapeluck, 1995; Dean, 2001), species (Johnson, 1990; Griffin et al. 1995), and genotype within species (Bongarten and Teskey, 1987; Crawford et al. 1991; St Clair, 1994). Previous work has documented how biomass accumulation and distribution can be affe cted by single factors such as fertilization and genotype; however, studies investigating the effects of these factors in combination are limited, especially for southern pines (Crawford et al. 1991; Li et al. 1991b; Gower et al. 1993; Retzlaff et al. 2001). It has been demonstrated that much of the response to these treatments occurs early in the development of these intensively managed stands, when rapid shifts in the accumulation and distribution of biomass among tree components occurs. The strengt h of the current investigation was that stand level attributes were examined on a unit area basis in pure family blocks across species, contrasting soil types, and silvicultural treatments. In this investigation biomass accumulation, distribution and N content as influenced by the combined effects of silvicultural management intensity, planting density, and family in two year old family block plantings of loblolly ( Pinus taeda L .) and slash pine ( P elliottii

PAGE 23

23 Engelm. var. elliotttii) across a range of locations was examined. The main objective of this study was to examine the nature and extent of any genotype x environment interactions in loblolly and slash pine such as: genotype x silviculture, genotype x density, genotype x location, silviculture x densi ty, silviculture x location, location x density, and any higher level interactions among these factors. Within this larger framework, specific objectives were to: 1) quantify the accumulation of biomass and N content within the components of FOL, BARK and BRANCH, BOLE, and BELOW components, 2) quantify the distribution of biomass and N content in the FOL, BOLE and BELOW components, and 3) determine if family performance and biomass distribution was stable across combinations of locations, silvicultural man agement intensities, and planting densities. Treatment Descriptions Five trials, three with loblolly and two with slash pine, were installed at locations previously supporting southern pine plantations (Table 21). The topography is nearly flat, with le ss than a 1% slope. Soil series for the five sites were: Sanderson, FL Leon (sandy, siliceous, thermic Aeric Alaquods); Waverly, GA Bladen (mixed, semiactive, thermic Typic Albaquults); Bunnell, FL Myakka ( sandy, siliceous, hyperthermic Aeric Alaquo ds ); Perry, FL Leon (sandy, siliceous, thermic Aeric Alaquods); Waldo, FL Newnan (sandy, siliceous, hyperthermic Ultic Haplohumods). Trials were installed on sites that held recently harvested southern pine plantations. Associated woody vegetation common to all sites included sawtooth palmetto [ Serenoa repens (B.) Small.], wax myrtle ( Myrica ceriferea L.), runner oak ( Quercus pumila Walt.), blueberries ( Vaccinium spp.), gallberry [ Ilex glabra (L.) Gray], and St. John's wort [ Hypericum fasciculatum (La m.)]. Herbaceous plants in the understory commonly included bluestem grasses ( Andropogon spp.), panic grasses ( Panicum spp.), sedges ( Carex spp. and Cyperus spp.), and dogfennel [ Eupatorium capillifolium (Lam.) Small.]. All study locations

PAGE 24

24 shared a subtropical and humid climate with long hot wet summers and mild dry winters. Long term (1931 2000) precipitation has averaged 1384 mm yr1 Each installation was double bedded using separate passes following 2.75 meter spacing between plantin g rows. A ll installations were treated i n the late summer/early fall of 1999 with pre plant herbicides consisting of Chopper (NOAA, 2002) (imazapyr) at 1.02 ha1 and Garlon (triclopyr) at 7.02 ha1, with the goal of removing all woody competition and reducing ini tial levels of herbaceous vegetation. The operational silvicultur al treatment was represent ative of a management regime commonly utilized by forest industry throughout the s outheastern U nited States These plots also received a single banded, or broadcas t, application of 280 kgha1In contrast the intensive treatment received annual fertiliz er additions following early vegetation control. For two years following planting competing vegetation was controlled using directed applications of Arsenal diammonium phosphate at the time of planting. (imazapyr) at 0.28 ha1 (limited to loblolly pine installations) and Oust (sulfometuron methyl) at 0.14 ha1 on all installations. A t the time of planting t he intens ive plots were fertilized with 5 60 kgha1Planting density was t he second treatment factor applied at the whole plot level: 1334 treesha of 101010 plus micronutrients, and was followed by site specific annual applications of macroand micronutrient fertilizers using prescriptions based on foliar analyses. The total amounts of nutrients applied to each installation through age two are pr esented in Table 22. 1 planted at a spacing of 2.75 m x 2.75 m, and 2990 treesha1 planted at a spacing of 1.22 m x 2.75 m. Within the 2990 treesha1 subpl ots trees were arranged in eight beds of 16 planting positions for a total of 128 trees per gross treatment plot. A two tree border existed around the perimeter result ing in a 48 tree interior measurement plot that was 0.016 ha in size

PAGE 25

25 The 1334 trees ha1At the subplot level, genetic entries consisted of first generation elite fullsib families. On loblol ly pine sites, there were seven entries of full sib families and on slash pine sites there were six entries, each including a poor grower In order to reduce confounding effects of disease, a ll genetic entries in the study were selected a priori based on knowledge from breeding programs and progeny tests from sources exhibiting average to excellent resistance to fusiform rust [ Cronartium quercum (Berk.) Miyabe ex Shirai f. sp. f usiforme ]. Seedlings were grown in 66 ml cell subplots of each genetic entry were arranged in eight beds of 10 planting positions each, for a total of 80 trees per gross plot. This arrangement, with a single tree buffer around the perimeter result ed in a 48 tree interior measurement plot of 0.036 ha. Tree survival was over 95% in all treatments at the end of the first growing season, despite an ongoing drought at the time of establishment. 1 Ray Leach Cone tainerTMInsecticides were uniformly applied across all treatments on loblolly pine installations in an effort to control damage from N antucket pine tip moth ( Rhyacionia frustrana [ Comstoc k ]). Treatments were applied on a monthly basis over the first two growing seasons, beginning in March and ending in September. Alternating applications of the following chemicals and application rates were applied aerially or by hand: Pounce cell s (Stuewe and Sons, Inc Corvallis, OR), and were planted in January of 2000. 3.2EC (62 ml product 1 water), Warrior T (39 ml product 1 water), Dimilin 25W (62 ml product 1 water), and Mimic 2LV (125 ml product 1A fungicide was uniformly applied across all treatments on slash pine installations in order to control for the confounding effects of potential fusiform rust infection. Treatments were applied by ground application on a monthly basis over a four month period (beginning in March) during water).

PAGE 26

26 t he first two growing seasons. The fungicide product and rate consisted of 12 g of 50% co ncentrated DF BayletonSampling Procedures with 65ml of Agri Dex in 18.9 of water. Biomass estimation Biomass prediction equations were developed using data collected from a destructive harvest completed at the end of the second growing season in November 2001. The sample trees were located in border row s which surrounded the sub plots Three loblolly and two slash pine experimental trials were sampled. At the loblolly pine locations, three replicates from each of two predetermined families (families L2 and L4), on each of the four silviculture x densit y whole plots, were sampled. A total of 70 trees were harvested across three locations. The slash pine experimental trials were sampled (families S1 and S6) in a similar manner, with the exception that four replicates were chosen for a total of 60 trees across two locations. Sample trees were free of damage and disease and were selected utilizing random sampling stratified by diameter class within each treatment and location. Prior to harvest, an inventory was completed on each sample tree consisting of total height, diameter at breast height, diameter at ground line, and crown width at the widest point parallel to and perpendicular to the bed (Figure 21). Sample trees were felled at ground line using a hand saw, placed on a tarp and separated into aboveground components: FOL, BRANCH, BARK and BOLE; there were no dead branches on these trees due to their young age. The total fresh weight of each component was measured in the field. To determine the fraction of BARK to BOLE, a 6 cm disk of wood was cut from the base of each of three equally spaced stem segments along the full length of the stem. The BARK component was separated from each disk and the fresh weight of each was determined in the field. There was no subsampling; all tissue samples were transported from the field and dried to a constant weight at 700C.

PAGE 27

27 Due to the considerable effort involved in belowground sampling, loblolly pine harvests for these components were limited to two out of three replicates, while slash pine was sampled on all three replicates, both from the same trees that were sampled for the aboveground components (n=32 slash, n=21 loblolly). Entire taproots were extracted and included coarse roots within one square meter that were greater than two millimeters in diameter to a depth of 40 cm. Entire samples were dried and weighed; there was no sub sampling. Biomass components for taproot and coarse roots were pooled and the combined component was identified as BELOW. Nitrogen concentrations were determined using a subsampl e of ground tissue from each biomass component. Tissue samples were chipped, following drying, using a small gas powered chipper. A homogenized sample of chips was then processed through a Wiley mill using a 2 mm sieve and then analyzed using a NCS 2500 Elemental Analyzer (Fisons Instrument, Milan, Italy). There was one exception, at the Bunnell, FL location, where the foliar tissue samples were lost. As a substitute, a sample of foliage was collected during the same month as the biomass sampling from e ight random living trees across each family and plot that was sampled. L ogarithmically transformed linear allometric equations were developed for each biomass component according to the base model (Equation 21) (Crow, 1988): ln(Y 01ln X1w here ln is the natural logarithm, Y is the dry weight of each biomass component (FOL, BRANCH. BARK, BOLE, BELOW) expressed in kg tree1, 0) is the mean of the tree within each species, X1 is the product of the c ombined variables of DBH squared times HT for each tree expressed in dm3 aboveground biomass for each tree. ( 21)

PAGE 28

28 The need for separate groups of equations by species, locations, silvicul tural management intensity families, and planting density was examined utilizing PROC MIXED (Littel et al. 1996) in SAS. These were evaluated by beginning with a pooled dataset and systematically decomposing the general model by entering treatment varia bles and their interactions. At each step, slopes and intercepts of the resulting equations were evaluated through covariate analysis. Probability plots of the residuals indicated that the normality assumption w ere satisfied and plots of residuals versus predicted values showed no obvious pattern, suggesting that the assumptions of independence and equal variance were met Corrections for bias in the transformation of logarithmic units to arithmetic units, were completed (Baskerville, 1972). Species specific equations for each biomass component (Table 23) were applied to the age two individual tree inventory data and summed for each plot ( expressed in Mg ha1 Analysis of dry matter ) Percentage FOL, BOLE, and BELOW were calculated using the sum of all individua l biomass components (TOTAL) in the denominator. Nitrogen contents were determined as the product of nitrogen concentration and the dry weight for each component. Components were pooled and summed to obtain N content for the FOL, aboveground (ABOVE), BEL OW and TOTAL components. All analys es were performed using PROC MIXED (Littel et al. 1996) in SAS. To test for differences in stand level attributes among treatments, separate a nalys e s of variance ( ANOVA) w ere performed for loblolly and slash pi ne using a mixed linear model for data pooled across sites within each species (Equation 22) : Yijklmn Si + b(s)ij + Ck + Dl + CDkl + Fm + CFkm + DFlm + CDFklm + SCik + SDil + CDikl + SFim + SCFik m + SDFilm + SCDFik lm + b(s)Cijk + b(s)Dijl + b(s)CDijkl + b(s)Fijm + b(s)CFijkm + b(s)DFijlm + b(s)CDFijklm + b(s)Sij + b(s)SCijk + b(s)SDijl + b(s)CDijkl + b(s)SFijm + b(s)SCFijkm + b(s)SDFijlm + wijklmn (2 2)

PAGE 29

29 where Yijklmn is the response variable ( FOL BRANCH, BARK, BOLE or TOTAL ) of the nth plot of the m th family of the l th planting density of the k th silvicultural intensity of the j th block of the i th location ( i = 1,2 ,; j = 1,2, 4; k = 1,2; l = 1,2; m = 1,2,, 6 for slash and 7 for Si is the fixed effect of the i th location; b(s)ij is the random interaction effect of the j th block within the i th location; Ck is the fixed effect of the k th silvicultural intensity; Dl is the fixed effect of the l th planting density; Fm is the fixed effect of the m th family and wijklmn is the random error. Blocks were nested within locations, while the factors of silvicultur al management intensity (C), planting density (D), and genotype (F) were crossed. All terms containing b(s)ij were considered to be random effects in the model and were pooled as appropriate for each variable tested using the procedure described by Bancroft and Han ( 1983) The only exception was b(s)CDijkl, w hich was never pooled as it wa s used as the error term to test the main effects of Si, Ck and Dl. Individual variance components were pooled when the probability of a greater F statistic was 0.25 or larger As noted by Bancroft and Han ( 1983) the significance level for the F test is much higher than conventional levels of 0.01 or 0.05 and is a conservative measure of the relative efficiency of pooling the sources of variation. Results Biomass Accumulation Sl ash pine Slash pine TOTAL biomass accumulation varied significantly at age two years by the main effects of silvicultural intensity (p=0.0021), planting density (p<0.0001) and family (p<0.0001), with no interactions among these factors or locations (Table 24). Total above and belowground biomass (TOTAL) accumulation under intensive silviculture, when averaged across both densities and all families, was 14% greater than the operational treatment (7.6 vs. 8.7 Mg ha1). As expected, a near doubling of the number of stems planted per ha from 1334 to 2990 trees ha1

PAGE 30

30 (44%) resulted in a 1.46fold increase in TOTAL biomass accumulation (4.7 vs. 11.6 Mg ha1). There was considerable variation among slash pine families in TOTAL biomass accumulation. The absolut e differences in accumulation between the two bottom and the single top performing families were around 18% (7.6 versus 8.9 Mg ha1Similar to the TOTAL component, slash pine biomass accumulation in the components of FOL, BRANCH, BARK, BOLE and BELOW varied by the main effects of silvicultural intensity (p<0.0024), planting density (p<0.0001), and family (p<0.0001), with no interactions among these factors or between locations (Table 2 4). Silvicultural intensity increased the accumulation of biomass to the BELOW, FOL, and BOLE components by 11, 13 and 17%, respectively. The effect of increasing planting density, from 1334 to 2990 trees ha families S3 and S4, versus S2, respectively) (Figure 2 2). 1, increased the FOL component the least (1.44fold) and t he BOLE component the most (1.48fold). There were large differences among the two bottom and the top performing families in the accumulation of biomass for the various tree components; BOLE was the most responsive at 21% (1.2 versus 1.4 Mg ha1, families S3 and S4, versus S2) and BELOW the least responsive at 16% (1.8 versus 2.1 Mg ha1Loblolly pine ; families S3 and S4, versus S2) (Figure 22). There were three strong and significant (p<0.001) twoway interactions for loblolly pine biomass accumulation a t age 2 for all components: silviculture x density silviculture x location, and family x density (Table 24). The effect of si lvicultur al intensity on TOTAL biomass accumulation was not stable a s planting densit y increased from 1334 to 2990 trees ha1 (i .e. silviculture x density). The response to the intensive s ilvicultural treatment was most pronounced on the high density plots (13.9 versus 17.2 Mg ha1) ( Figure 2 3a ). The form of this interaction was a non proportional response between planting densi ties to increasing silvicultural

PAGE 31

31 intensity. Under the operational silviculture treatment, a near doubling of the planting density resulted in a 2.2fold increase in TOTAL biomass, compared to a 2.4 fold increase under the intensive silvicultur al treatment Total biomass accumulation was influenced by a combination of silvicultural intensity and location (silviculture x location). The most productive combination occurred under the intensive silviculture treatment at the S anderson, FL location (15.5 Mg ha1), with the least productive under the operational treatment at B unnell, FL (8. 6 Mg ha1) (F igure 23b). Within each of the B unnell, FL and W averly, GA locations, there was no significant difference in TOTAL biomass accumulation between the s ilvicultural treatments; however, at the SA location there was a statistically significant 36 % difference in TOTAL biomass accumulation (11.3 versus 15.5 Mg ha1Family performance for TOTAL biomass accumulation was not stable across densities (i.e., a significant interaction of family x density). In general, families with high growth rates at low densities performed better than expected at higher densities (Figure 2 4). The form of this interaction was a non proportional response among families with increasing planting densities, with the exception of families L4 and L8. These two families were top performers for biomass accumulation and demonstrated a rank change. Family L4 outperformed L8 at 1334 trees ha operational and intensive treatments, respectively) 1, av eraged across locations (7.7 vs. 7.3 Mg ha1); however, this trend was reversed at 2990 trees ha1 (17.2 vs. 17.6 Mg ha1Biomass Distribution to Foliage Bole, and Belowground ). Slash pine The main effects of silvicultural intensity (p<0.0017) and family (p<0.0001) were statistically significant in slash pine for the distribution of biomass to all components (Table 2 5). Intensive silvicultural treatments minutely increased the distribution of biomass to BOLE

PAGE 32

32 (15.2 to 15.7% operational and intensive trea tments, respectively) By contrast, biomass distribution in the FOL (38.5 and 38.1% ) and BELOW components (24.3 and 24.0% ) decreased in response to the intensive silvicultural treatments Families varied significantly in their distribution of biomass to FOL (38.0 to 38.4% families S2 and S5, respectively), BOLE (15.5 to 15.8% families S2 and S6, respectively ) and the BELOW (23.9 to 24.2% families S2 and S6, respectively) (Figure 2 5). The effects of silviculture and family did not interact and were st able across planting densities and locations. Loblolly pine The d istribution of loblolly pine biomass among FOL, BOLE, and BELOW components demonstrated a main effect of family (p<0.0001) and a three way interaction of silviculture x density x location (p< 0.0001) (Table 2 5) Families varied in their distribution of biomass to FOL (19.6 to 20.8% families L6 and L4, respectively), BOLE (30.0 to 30.9 % families L4 and L6 respectively ) and BELOW (25.7 to 26.6% families L4 and L6, respectively) (Figure 2 6). The three way interaction was complex and involve d a differential response in biomass distribution between silvicultur al intensity and planting density across locations. In general, this interaction was driven by an anomaly which occurred at the B unne ll, FL location within the 1334 trees ha1Nitrogen Content planting density At this planting density, the intensive silvicultural treatment had less biomass distributed to the BELOW component (18.7 versus 19.7% intensive and operational respectively), and greater distr ibution to FOL (31.7 versus 30.9% ) and BOLE (27.3 versus 26.5% ) (Figure 2 7c). Slash pine Two years after planting, there was a significant and strong main effect of silvicultural treatment intensity on N content for all slash pine biomass components (p<0.0001). On average,

PAGE 33

33 total N accumulation in the above and belowground tissue components increased by about 40% under the intensive silvicultural treatment (from 46.8 to 65.5 kg ha1 under operational and intensive treatments, respectively ). Foliar N accumulation s averaged 30.9 and 40.7 kg ha1 (operational and intensive) and N accumulation in the ABOVE component were 39.4 kg ha1 for the operational and 54.0 kg ha1There were significant differ ences among slash pine families for N contents of all biomass components: FOL (p=0.0481), ABOVE (p=0.0479), BELOW (p=0.0462), and TOTAL (p=0.0490) (Table 26). Family variation in total N content ranged from a low of 52.0 kg ha for the intensive treatments, respectively. 1 for family S3 to a high of 61.3 kg ha1 Significant interactions between planting density and location for N content of the FOL, ABOVE and TOTAL compone nts (p<0.0001) were detected. At 1334 trees ha for family S2 (Table 2 6). Families were stable and did not vary in combination with any other treatment factors for N content. 1, total N content averaged 33.0 kg ha1 for each of the Perry, FL and Waldo, FL locations independently. However, when the planting density was increased at the same conditions to 2990 trees ha1, the Perry FL location was more responsive than the Waldo, FL location (88.3 versus 70.5 kg ha1 of N, respectively) (Table 2 7). There was a threeway interaction for N content in the BELOW component between silvicultural intensity, planting density, and location (Figure 2 8). The N content for the BELOW component at the Waldo, FL location, under the 2990 trees ha1 treatment, was considerably higher than that at the Perry, FL location ( 11.4 versus 9.6 kg ha1Loblolly pine respectively both under the intensive silvicultural treatment ). Nitrogen content in all biomass components demonstrated two types of interactions: genotype x density, and silviculture x density x location. Families were not stable across the two planting densities tested in this experiment f or N content for FOL (p=0.0235), ABOVE (p=0.0364),

PAGE 34

34 BELOW (p=0.0057), and TOTAL (p=0.0250) (i.e. a family x density interaction). In general, this interaction was due to scale effects among families, with the notable exception of two families (L4 and L8) which changed rank in N content between planting densities. For example, at the 2990 trees ha1 density, N contents for families L4 and L8 by component were as follows: FOL (59.4 and 60.3 kg ha1), ABOVE (79.3 and 80.8 kg ha1), BELOW (15.1 and 15.3 kg h a1), and TOTAL (94.3 and 96.1 kg ha1). Nitrogen contents for families L4 and L8 by component at the 1334 trees ha1 planting density were as follows: FOL (27.4 and 26.0 kg ha1), ABOVE (37.5 and 35.7 kg ha1), BELOW (6.4 and 6.0 kg ha1), and TOTAL (43.9 and 41.7 kg ha1The three way interaction (silviculture x density x location) was caused by an unequal response to silvicultural management intensity between planting densities at the three locations examined (FOL, p<0.0001; ABOVE, p<0.0001; BELOW, p=0.0001; TOTAL, p<0.0001). At the Bunnell, FL location, there were much lower than expected N contents in the FOL, ABOVE, and TOTAL components in the intensive silvicultural treatment combination at the 1334 trees ha ) (Table 2 8). 1 planting density (Table 2 9). For each component, with the exception of BELOW, the intensive silviculture treatment had lower values of N than the operational silvicultural treatment (FOL 14.5 and 19.7 kg ha1, ABOVE 18.1 and 24.1 kg ha1, BELOW 5.2 and 5.0 kg ha1, and TOT AL 23.3 and 29.1 kg ha1Discussion respectively). This was not observed at either the Sanderson, FL or Waverly, GA locations. Understanding genotype x environment interactions is becoming increasingly important as the level of genetic selection in sou thern pines increases in combination with the intensity of silvicultural management. Questions remain as to how best to deploy this elite genetic material, given uncertainty in predicting its response to a wide variety of anthropogenic and abiotic

PAGE 35

35 environmental factors (Fox, 2000). The experimental design in the current study provided a powerful opportunity to examin e unit area biomass accumulation, N content, and distribution in selected elite families of loblolly and slash pine as influenced by a combination of silvicultural management intensities planting densit ies and locations After two growing seasons significant interactions were evident for biomass accumulation, N content and distribution to various stand components. Generally, loblolly pine de monstrated a greater number of and higher level interactions than did slash pine. This was not unexpected, since loblolly pine has been shown to be considerably more responsive than slash pine to silvicultural treatments, such as weed control and fertiliz ation than slash pine (Xiao et al. 2003; Cobb et al. 2008). Significant interactions in this study included: genotype x location, genotype x density, and silviculture x density x location. The ability to detect these interactions was a function of the high statistical power of the experimental design and the precision associated with the intensity of sample measurements. Biomass accumulation in s lash pine at age two was responsive t o the main effects of silvicultural intensity, planting density, and fam ily with no interactions Age two total biomass accumulation was in the range of 6 Mg ha1and was consistent with that published by others when extrapolated from age four data (Colbert et al. 1990; Jokela and Martin, 2000). Since the response to these t reatments did not interact in slash pine, the accumulation of biomass should be predictable at an early age; however, this stability may be influenced over time as stands develop and are subjected to a variety of biotic and abiotic factors such as disease and catastrophic weather events (Roth et al. 2007a). As reported previously by others, slash pine response to intensive silviculture was mostly evidenced by increases in BOLE biomass despite modest

PAGE 36

36 increases in FOL biomass (Colbert et al. 1990) T he g reatest variation among the families in this experiment occurred in the BOLE component Total biomass accumulation values for loblolly pine at age two years in this region were in general agreement with previously published estimates of around 10 Mg ha1Loblolly pine response in biomass accumulation to silvicultural management intensity was not stable across planting densities; and the largest response occurred under the narrowest spacing. While the mechanism behind this effect could not be tested in this investigation, there may be a synergistic effect of increased resource availability with increasing stand density. Increases in individual tree biomass with increasing stand density have been documented (Scott et al. 1998) ; however, this was not observed in the current study on the individual tree level. S ever al mechanisms have been proposed to explain this including micro site availability of water and nutrients (Woodruff et al. 2002), and early signaling of intra tree competition (Ritchie, 1997; Ballar, 1999). ( Adegbidi et al. 2002). There were three types of interactions evident for loblolly pine biomass accumulation at agetwo: silviculture x density, silviculture x location, and genotype x density. The silviculture x location interaction was in teresting in that it was driven by a single location: Sanderson, FL. The two locations were contrasting in water availability and soil characteristics (see Appendix B and C). The Sanderson location has lower water availability and t he soil type at this location was a Spodosol, which tends to be infertile, and is in contrast to the Ultisol at the Waverly, GA location which tends to have available nutrients concentrated near the soil surface. It has been well documented that loblolly pine productivity is highly dependant on nutrient availability (Jokela et al. 2000) and on the nutrient poor Sanderson, FL location, the

PAGE 37

37 intensive silvicultural treatment created a large contrast in biomass accumulation due to nutrient addition, when compared to the relatively nutrient rich Waverly, GA location. The genotype x density interaction for TOTAL biomass accumulation observed in this investigation was surprising. Previous research has shown that stand density has no effect on the growth performance of individual lobl olly pine families (McCrady and Jokela, 1996); however, there are limited exceptions and these were generally of the scale type effects for a few genotypes grown at very high densities (Land et al. 2004). While much of the interaction observed in the cur rent investigation was of the scale type, there was a rank change between the two top producing families (L4 and L8) as planting density increased. The fact that a limited number of highly reactive families were responsible for this effect is consistent w ith most genotype x environment interactions described in the literature (Zas et al. 2004; Roth et al. 2007a). Of interest, is that these two families have been identified as putative crop and competition ideotypes (L8 and L4, respectively) at age 6 in this same study series (Staudhammer et al. 2009). Family L4 tends to have a large and deep crown structure (Chmura and Tjoelker, 2008), which may have provided a competitive advantage in the widely spaced 1334 trees ha1Shifts in biomass distribution amo ng tree components occur as trees and stands develop (Ledig et al. 1970; Jokela and Martin, 2000; King et al. 2007). For example, soon after plantation establishment, biomass accumulation was primarily distributed in the FOL and BELOW components, but as trees grow in size the BOLE component becomes an increasingly larger and more significant component (Adegbidi et al. 2005). Treatments such as silvicultural treatment. If this contrast betw een putative ideotypes continues to hold over time, then the results documented in this investigation would suggest that competition dynamics between these genotypes began at an extremely early stage of stand development.

PAGE 38

38 management intensity, plantation density, and family also influence the distribution of biomass a mong tissue components, but the separation of these treatment effects from developmental effects is difficult since these treatments simultaneously increase growth and advance stand development (Coyle and Coleman, 2005). This investigation does not attempt to separate these effects, since the primary goal was to examine genotype x environment interactions among combinations of treatments, rather than make inferences about the causation of differences in biomass distribution between treatments. Slash pine biomass distribution was influenced by the main effects of family and silvicultural treatment. Family differences in biomass distribution within various components were relatively small, less than one percent; however, a t this early stage of stand developm ent small differences can lead to a competitive growth advantage in the long term. Increasing the intensity of silvicultural treatments had the effect of greater biomass distribution to BOLE in relation to FOL, which was supported by the work of others (C olbert et al. 1990). However, in this study, distribution of biomass to BOLE was not influenced by increasing planting density as was found in a four year old slash pine experiment in GA (Burkes et al. 2003). Loblolly pine family distribution of biomass to various components was stable across all treatments and locations. This finding is supported by that documented in a five year old loblolly pine family block experiment where biomass allocation of fastand slow growing families were similar across fe rtilizer treatments (Retzlaff et al. 2001). However, others have found that selected loblolly pine families (Li et al. 1991b) and clones (Tyree et al. 2009) distribute biomass differentially between roots and shoots under contrasting fertilization regi mes. Among the limited number of families examined in this investigation, those with the highest

PAGE 39

39 amount of TOTAL biomass accumulation had the least amount of biomass allocated to FOL and the greatest to BOLE, which may reflect ontogenetic effects. There was a significant threeway interaction influencing the distribution of loblolly pine biomass, which involved a combination of silvicultural intensity, planting density, and location. The Bunnell, FL location was the primary driver behind this interaction. The topography at this location is nearly flat and soils are deep, sandy Spodosols. On September 14thNitrogen accumulation in slash pine components demonstrated two main effects (silvicultural treatment intensity and family) and one significant three way interaction: silviculture x density x location for N content to the BELOW component. For slash pine, the addition of around 240 kg ha 2001, after the first growing season, tropical storm Gabrielle traversed the study area with large amounts of precipitation and sustained high winds. From observational data following the storm, it was evident that the effects of this storm influenced biomass distribution in the year following, especially in the treatment combination of wide spacing and intensive silviculture. Trees in this treatment c ombination appeared to be most affected (primarily root damage) by the effects of wind, since fertilization produced larger crowns and the heavy precipitation contributed to erosion of the beds and instability of the root systems. The patterns of biomass distribution in these treatments demonstrated a higher distribution to FOL and BOLE, which ran contrary to the patterns observed for the same treatment combinations at the other two locations. Presumably, these factors collectively influenced the expressi on of this three way interaction in loblolly pine. 1 of N above that added in the operational silvicultural intensity treatment was associated with an increase in total N content by about 20 kg ha1 over the two years following planting. As has been reported previously in the literature, the families with the greatest biomass production also tended to have the greatest N contents (Crawford et al. 1991). There was a twoway interaction

PAGE 40

40 between location and planting density The cause was at the Perry, FL location, which had larger total N content at the 2990 trees ha1 density than at Waldo, FL. While nutritional inputs were similar between locations, with the exception of sulfur (Table 2 2), response s to this difference should have been evidenced via a silviculture x location interaction, which was not detected. To complicate this, there was a threeway interaction for N content in the BELOW component between silvicultural treatment intensity, planting density and location, which may be a statistical anomaly. For some yet unexplained reason, N accumulation in the BELOW component was greater at Waldo, FL than at Perry, FL. This occurred despite similar biomass accumulations, under the operational silvicultural treatment i ntensity at the 2990 trees ha1Nitrogen contents of loblolly pine biomass components were within 10% of those reported for stands of similar age (Adegbidi et al. 2005) and varied in combination with the factors of families, locations, p lanting densities and silvicultural treatments. This was especially evident at the inherently nutrient poor Bunnell, FL location, where the influence of the intensive silvicultural treatment had a negative impact on N accumulation at the 1334 trees ha planting density. 1The genotype x density interaction for N content offered insight into why particula r genotypes responded differentially to intra specific competition. The genotype x density interaction was similar as observed in the accumulation and distribution of biomass; the two families involved in the rank change for N content were the same ( famil ies L4 and L8). While subtle, it appears that the two families responsible for the differences in N content may have differential uptake rates between planting densities which were not explained by belowground biomass, excluding fine planting density. This effect was driven more by a lower accumulation of biomass than N concentration, and possibly reflective of hurricane damage from the previous season.

PAGE 41

41 roots. While this c urrent investigation did not examine fine roots this may be worthy of future research efforts in an attempt to better understand N uptake. Conclusions After two growing seasons, biomass accumulation, N content and distribution were influenced by combinati ons of silvicultural intensity, planting density, family and locations in plantations of loblolly and slash pine. Significant genotype x density, silviculture x density, and silviculture x location interactions existed for loblolly pine biomass accumulation, yet none existed for slash pine. When distribution of this accumulated biomass to various components was examined, the only interaction that was significant was silviculture x density x location for loblolly pine. Nitrogen accumulation and distributi on was influenced by significant genotype x density and silviculture x density x location interactions for loblolly pine and density x location interactions for slash pine. Evidence of these complex interactions, as early as age two, serves to underscore t he importance of understanding how to best deploy elite genotypes of loblolly and slash pine given the uncertainty in predicting its response to a wide variety of man made and abiotic environmental factors For example, the only three way interaction occ urring for slash pine may have been due to the effects of a hurricane after the first growing season. While distribution of accumulated biomass did not appear to be involved in driving these interactions, variation in crown structure (McCrady and Jokela, 1996) or resource use efficiency may play a role (McKeand et al. 1997a) and deserves further investigation.

PAGE 42

42 Table 21. Characteristics of the five PPINES experimental locations in southeast Georgia and northeast Florida Site location Species Latitude ( 0 ) Longitude ( 0 ) Soil order Elevation (m) Sanderson, FL Loblolly 3 0.28 82.33 Spodosol 45 Waverly, GA Loblolly 31.13 81.75 Ultisol 10 Bunnell, FL Loblolly 29.28 81.31 Spodosol 8 Perry, FL Slash 30.17 83.73 Spodosol 15 Waldo, FL Slash 29.80 82.21 Spodosol 50 PPINES : Pine Productivity Interactions on Experimental Sites. Table 22. Elemental application rates (kgha1 Site location ) of fertilizers supplied to the PPINES locations through the end of the second growing season ( intensive silvicultural manag ement treatments only). N P K Mg Ca S B Zn Mn Fe Cu Sanderson, FL 290 103 121 45 45 28 0.9 2.7 2.2 5.4 1.5 Waverly, GA 290 103 121 45 45 28 0.9 2.7 2.2 5.4 1.5 Bunnell, FL 236 100 125 34 34 40 0.9 2.2 2.2 5 .3 0.9 Perry, FL 298 90 116 56 45 132 1.1 3.0 3.0 7.3 1.5 Waldo, FL 292 103 124 63 56 27 1.7 2.5 2.7 3.4 1.0 Operational silviculture treatments all received 45 kgha1 N and 50 kgha1 of P in the form of diammonium p hosphate at the time of planting only.

PAGE 43

43 Table 2 3. Parameter estimates and standard errors of the estimates for foliage, bole and belowground biomass (kgtree1 Component/ ) equations for two year old loblolly and slash pine. Equations were generated using destructive harvest data. 0 1 Overall Model Estimates Species Estimate SE p value Estimate SE p value R 2 RMSE n FOL Loblolly 0.06034 0.04752 0.2065 0.44997 0.03387 <0.0001 0.582 0.31682 130 Slash 0.01654 0.04622 0.7209 0.44997 0.03387 <0.0001 0. 582 0.31682 130 BRANCH Loblolly 0.69577 0.05997 <0.0001 0.55626 0.04275 <0.0001 0.658 0.39985 130 Slash 1.18661 0.05833 <0.0001 0.55626 0.04275 <0.0001 0.658 0.39985 130 BARK Loblolly 1.88773 0.03 359 <0.0001 0.48311 0.02378 <0.0001 0.806 0.22095 129 Slash 1.36949 0.03227 <0.0001 0.48311 0.02378 <0.0001 0.806 0.22095 129 BOLE Loblolly 0.54397 0.03351 <0.0001 0.59587 0.02337 <0.0001 0.866 0.22042 130 Slash 0.93778 0.03220 <0.0001 0.59587 0.02337 <0.0001 0.866 0.22042 130 BELOW Loblolly 0.19065 0.06869 0.0077 0.44718 0.04403 <0.0001 0.694 0.27061 53 Slash 0.42350 0.05038 <0.0001 0.44718 0.04403 <0.0001 0.694 0.27061 53 Regression 1 regression coefficients (intercept and slope, respectively), X = DBH2 x HT in dm3 Abbreviations for biomass components are as follows: FOL = foliage, BRANCH = branches, BARK = bark, BOLE = stemwood, and BELOW = belowground. The belowground component consists of taproot plus coarse roots greater than 2 mm in diameter. Fine roots were excluded.

PAGE 44

44 Table 2 4. Summary of statistical significanc e (prob. >F) and associated degrees of freedom from ANOVA to test loblolly and slash pine foliage, branches bark stemwood, belowground, and total biomass accumulation at age two years Source of Variation Num. df Den. df Foliage p value Branches p value Bark p value stemwood p value Belowground p value Total p value Loblolly Silviculture (C) 1 9 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Density (D) 1 9 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 C x D 1 9 0.0003 0.0003 0.0003 0.0003 0.0003 0.0003 Family (F) 6 54 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 C x F 6 154 0.9598 0.9252 0.9517 0.9052 0.9604 0.9445 D x F 6 154 0.0212 0.0175 0.0199 0.0164 0.0213 0.0192 C x D x F 6 154 0.7393 0 .7188 0.7336 0.7096 0.7398 0.7293 Location (S) 2 9 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 S x C 2 9 0.0007 0.0004 0.0006 0.0003 0.0007 0.0005 S x D 2 9 0.0519 0.0553 0.0527 0.0571 0.0519 0.0534 S x C x D 2 9 0.0524 0.0658 0.0568 0.0701 0.0520 0.0596 S x F 12 54 0.1136 0.1035 0.1119 0.0970 0.1137 0.1094 S x C x F 12 154 0.7983 0.7881 0.7970 0.7799 0.7983 0.7933 S x D x F 11 154 0.2319 0.2991 0.2524 0.3250 0.2302 0.2663 S x C x D x F 11 154 0.8269 0.791 5 0.8159 0.7784 0.8278 0.8086 Slash Silviculture (C) 1 18 0.0024 0.0017 0.0021 0.0016 0.0024 0.0021 Density (D) 1 18 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 C x D 1 18 0.1448 0.1269 0.1385 0.1215 0. 1454 0.1370 Family (F) 5 119 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 C x F 5 119 0.6414 0.6066 0.6305 0.5931 0.6423 0.6277 D x F 5 119 0.4192 0.3432 0.3965 0.3135 0.4210 0.3883 C x D x F 5 119 0.8823 0.8705 0.8789 0.8652 0. 8825 0.8778 Location (S) 1 6 0.7625 0.7919 0.7716 0.8026 0.7617 0.7746 S x C 1 18 0.8989 0.8950 0.8974 0.8940 0.8991 0.8973 S x D 1 18 0.4731 0.4363 0.4607 0.4244 0.4741 0.4574 S x C x D 1 18 0.8746 0.8876 0.8794 0.8908 0.8742 0.87 98 S x F 5 119 0.1622 0.1658 0.1631 0.1673 0.1621 0.1637 S x C x F 5 119 0.8304 0.8590 0.8400 0.8682 0.8296 0.8425 S x D x F 5 119 0.6339 0.6525 0.6396 0.6595 0.6334 0.6415 S x C x D x F 5 119 0.9980 0.9982 0.9981 0.9982 0.9980 0.99 81 P values significant at the 95% level of confidence are shown in bold type

PAGE 45

45 Table 25. Summary of statistical significance (prob. >F) and associated degrees of freedom from ANOVA to test loblolly and slash pine distribution of biomass to foliage, stem wood, and belowground components at age two years. Source of Variation Num. df Den. df Percent foliage p value Percent stemwood p value Percent belowground p value Loblolly Silviculture (C) 1 121 <0.0001 <0.0001 <0.0001 Density (D) 1 114 0.15 85 0.1763 0.1539 C x D 1 121 0.0066 0.0072 0.0065 Family (F) 6 114 <0.0001 <0.0001 <0.0001 C x F 6 121 0.9964 0.9963 0.9963 D x F 6 114 0.4562 0.4743 0.4510 C x D x F 6 121 0.4958 0.5062 0.4926 Location (S) 2 9 0.0002 0.0002 0.0002 S x C 2 121 <0.0001 <0.0001 <0.0001 S x D 2 114 <0.0001 <0.0001 <0.0001 S x C x D 2 121 <0.0001 <0.0001 <0.0001 S x F 12 114 0.2868 0.2765 0.2898 S x C x F 12 121 0.8920 0.8938 0.8913 S x D x F 11 114 0.0550 0.0550 0.0550 S x C x D x F 11 1 21 0.5210 0.5159 0.5223 Slash Silviculture (C) 1 12 0.0016 0.0014 0.0016 Density (D) 1 6 0.1587 0.1620 0.1549 C x D 1 12 0.9003 0.8960 0.9017 Family (F) 5 30 <0.0001 <0.0001 <0.0001 C x F 5 89 0.1854 0.1832 0.1860 D x F 5 89 0.6592 0.6697 0.6554 C x D x F 5 89 0.3632 0.3586 0.3648 Location (S) 1 6 0.8900 0.8954 0.8880 S x C 1 12 0.9454 0.9455 0.9455 S x D 1 6 0.3675 0.3623 0.3691 S x C x D 1 12 0.9030 0.9049 0.9023 S x F 5 30 0.1510 0.1465 0.1525 S x C x F 5 89 0.7210 0.7814 0.7674 S x D x F 5 89 0.8835 0.8828 0.8837 S x C x D x F 5 89 0.9895 0.9897 0.9894 P values significant at the 95% level of confidence are shown in bold type.

PAGE 46

46 Table 26. Least squared means for s lash pine nitrogen accumula tion ( kg ha1 ) among various biomass components as influenced by full sib family at age two years Family S1 S2 S3 S4 S5 S6 FOL 37.0 ab 38.8 a 33.2 b 33.7 b 34.7 ab 37.3 ab ABOVE 48.3 abc 50.9 a 43.3 c 43.9 bc 45.1 abc 48.6 ab BELOW 9.8 ab 10.3 a 8.8 b 8.9 b 9.3 ab 9.9 ab TOTAL 58.1 ab 61.3 a 52.0 b 52.9 b 54.4 ab 58.5 ab Values within biomass components (rows) having the same letter are not significantly different at the 95% level of confidence using Bonferroni's least s igni ficant d i fference (LSD). Abbreviations and pvalues for the percent distribution by component are as follows: FOL = foliage (p=0.0481) ABOVE = aboveground (p=0.0479) BELOW = belowground (p=0.0462) and TOTAL = total (p=0.0490) The belowground component consist s of taproot plus coarse roots greater than 2 mm in diameter. Fine roots were excluded. Table 2 7. Slash pine nitrogen accumulation in foliage, aboveground and total biomass components by planting density and location at two years (kg ha1 Component/ Location ). There was a significant two way interaction (p<0.0001) between planting d ensity and location for each component. 2990 trees ha 1 1334 trees ha 1 FOL Perry, FL 57.5 a 21.2 c Waldo, FL 44.2 b 20.5 c ABOVE Perry, FL 74 .4 a 28.0 c Waldo, FL 57.5 b 26.9 c TOTAL Perry, FL 88.3 a 33.0 c Waldo, FL 70.5 b 33.0 c Values within each biomass component (i.e. two rows following each component) ha vi ng the same letter are not significantly different at the 9 5 % level of confidence using Bonferroni's least significant d ifference (LSD). Abbreviations by component are as follows: FOL = foliage, ABOVE = aboveground, and TOTAL = total.

PAGE 47

47 Table 28. Loblolly pine nitrogen accumulation in foliage, aboveground, belowground, and total biomass components by family and planting density at two years (kg ha1 Component/ ). There was a significant two way interaction between family and density (p=0.0235, p= 0.0364, p=0.0057, and p=0.0250 for each component, respectively) Famil y Planting density L1 L2 L4 L5 L6 L7 L8 FOL 1334 trees ha 1 22.3 f 24.6 def 27.4 d 22.8 ef 21.0 f 25.8 de 26.0 de 2990 trees ha 1 50.5 bc 55.6 ab 59.4 ab 52.2 abc 44.4 c 55.9 ab 60.3 a ABOVE 1334 tree s ha 1 30.2 ef 33.3 de 37.5 d 30.8 ef 28.2 f 35.0 de 35.7 d 2990 trees ha 1 66.9 bc 73.7 abc 79.3 a 69.6 abc 59.1 c 74.5 ab 80.8 a BELOW 1334 trees ha 1 5.3 fg 5.8 defg 6.4 d 5.4 efg 5.0 g 6.2 de 6.0 def 2990 trees ha 1 12. 8 bc 14.1 ab 15.1 a 13.2 abc 11.3 c 14.2 ab 15.3 a TOTAL 1334 trees ha 1 35.5 fg 39.2 def 43.9 d 36.2 efg 33.2 g 41.2 de 41.7 de 2990 trees ha 1 79.7 bc 87.8 ab 94.3 a 82.8 abc 70.4 c 88.7 ab 96.1 a Values within each biomass component (i.e. two rows following each component) ha vi ng the same letter are not significantly different at the 9 5 % level of confidence using Bonferroni's least significant d ifference (LSD). Abbreviations for each component are as follows: FOL = foliage, ABOVE = aboveground, BELOW = belowground, and TOTAL = total.

PAGE 48

48 Table 2 9. Loblolly pine nitrogen accumulation in foliage, aboveground, belowground, and total biomass components by silvicultural treatment, planting density and location at age two years (kg ha1 Component/ ). 2990 trees ha 1 1334 trees ha 1 Location Intensive Operational Intensive Operational FOL Bunnell, FL 53.7 bc 43.8 d 14.5 g 19.7 fg Sanderson, FL 62.7 a 43.9 cd 33.0 e 21.2 f Waverly, GA 62. 2 ab 57.7 ab 32.4 e 24.8 f ABOVE Bunnell, FL 66.4 bc 53.5 de 18.1 g 24.1 g Sanderson, FL 86.5 a 60.1 cd 47.4 e 29.5 fg Waverly, GA 85.8 a 79.5 ab 44.7 e 33.9 f BELOW Bunnell, FL 13.5 abc 11.8 abc 5.2 cd 5.0 d Sanderson, FL 14.2 ab 6.5 c 11.1 abc 4.4 d Waverly, GA 15.7 ab 7.7 bc 16.0 a 5.7 cd TOTAL Bunnell, FL 79.9 a 65.3 ab 23.3 d 29.1 d Sanderson, FL 100.7 a 71.2 a 53.8 b 33.8 cd Waverly, GA 101.5 a 95.5 a 52.4 b 39.6 c There was a significant threeway interaction between silvicultural management intensity, planting density and location. Abbreviations and pvalues for N content by component are as follows: FOL = foliage (p<0.0001) ABOVE = aboveground ( p<0.0001 ) BELOW = belowground (p= 0.0001 ) and TOTAL = total (p<0.0001) Values within each biomass component (i.e. two rows following each component) ha vi ng the same letter are not significantly different at the 9 5% level of confidence using Bonferro ni's least significant d ifference (LSD).

PAGE 49

49 Figure 2 1. Typical twoyear old slash pine that was harvested from the intensive silviculture treatment at the Waldo, FL location, demonstrating the rapid development at this early age.

PAGE 50

50 1.3 1.4 1.2 1.2 1.2 1.3 2.0 2.0 1.7 1.7 1.8 1.9 3.2 3.4 2.9 2.9 3.0 3.2 2.0 2.1 1.8 1.8 1.9 2.0 -3.0 0.0 3.0 6.0 9.0 S1 S2 S3 S4 S5 S6 Family Biomass (Mg ha -1) ab ab ab a b b Figure 2 2. Leas t squares means for the main effect of full sib family (p<0.0001) on slash pine biomass accumulation (Mg ha 1 ) in foliage (FOL), bark and branches (BARK & BRANCH), stemwood (BOLE), and belowground (BELOW) components at age two years when averaged across tw o locations and planting densities. For each family with the same letter, each biomass components was not significantly different at the 95% level of confidence using Bonferronis least significant difference (LSD).

PAGE 51

51 A 5 7 9 11 13 15 17 19 Operational Intensive Total biomass (Mg ha-1) 1334 tph 2990 tph A B C D B 7 9 11 13 15 17 Operational Intensive Total biomass (Mg ha -1) Sanderson, FL Waverly, GA Bunnell, FL A BC C C B B Figure 2 3. Least squares mea ns for the effect of silvicultural intensity on loblolly pine total biomass accumulation as influenced by A) planting density (p=0.0003) and B) location (p=0.0005). Data points within graphs followed by the same letter are not significantly different at the 95% level of confidence using Bonferronis least significant difference (LSD). Trees ha 1 is abbreviated as tph.

PAGE 52

52 5 8 11 14 17 1334 tph 2990 tph Total biomass (Mg ha -1) L1 L2 L4 L5 L6 L7 L8 a c a ab b a ab abc c d Figure 2 4. Least squares means for loblolly pine total biomass accumulation demonstrating an interaction between family and planting density (p=0.0192). Each data point is averaged across two silvicultural treatment intensities and three locations. While not statistically significant, notice the rank change between families L4 and L8 when planting density increased from 1334 to 2990 trees ha1. Trees ha1 is abbreviated as tph. Data points within planting densities followed by the same letter are not significantly different at the 95% level of confidence using Bonferronis least significant difference (LSD).

PAGE 53

53 15.4 15.8 15.4 15.3 15.3 15.5 38.3 38 38.4 38.4 38.4 38.3 24.1 23.9 24.2 24.2 24.2 24.2 -30 -15 0 15 30 45 60 S1 S2 S3 S4 S5 S6 Family Percent of total biomass b a a a a a a a a b b b b b b b b b Figure 2 5. Lea st squares means for the main effect of full sib family (p<0.0001) on slash pine biomass distribution to foliage (FOL), stemwood (BOLE), and belowground (BELOW) components expressed as a percentage of total biomass. Branch and bark data followed similar p atterns and are not shown for simplicity. Data are at age two years and was averaged across two locations and planting densities. For each family having the same letter, biomass components were not significantly different at the 95% level of confidence using Bonferronis least significant difference (LSD).

PAGE 54

54 20.1 20.4 20.8 20.3 19.6 20.7 20.8 30.6 30.3 30.0 30.4 30.9 30.1 30.1 25.8 25.8 26.6 26.1 25.7 26.0 26.2 -30.0 -15.0 0.0 15.0 30.0 45.0 60.0 L1 L2 L4 L5 L6 L7 L8 Family Percent of total biomass abc cd a bcd a b a b d Figure 2 6. Least squares means for the main effect of full sib family (p<0.0001) on loblolly pine biomass distribution to foliage (FOL), stemwood (BOLE), and belowground (BELOW) components expres sed as a percentage of total biomass. Branch and bark data followed similar patterns and are not shown for simplicity. Data are at age two years and was averaged across two locations and planting densities. For each family having the same letter, biomas s components were not significantly different at the 95% level of confidence using Bonferronis least significant difference (LSD).

PAGE 55

55 A 16 18 20 22 24 26 28 30 32 34 1334 tph 2990 tph Percent of total biomass FOL BOLE BELOW Intensive Operational B 16 18 20 22 24 26 28 30 32 34 1334 tph 2990 tph Percent of total biomass FOL BOLE BELOW Intensive Operational C 16 18 20 22 24 26 28 30 32 34 1334 tph 2990 tph Percent of total biomass FOL BOLE BELOW Intensive Operational Figure 2 7. Loblolly pine biomass distribution to foliage (FOL), stemwood (BOLE) and belowground (BELOW) components demonstrating a three way interaction (p<0.0001) between silvicultural management intensity, planting density, and location: A) Sanderson, FL B) Waverly, GA and C) Bunnell, FL. Planting density was expressed in trees ha1 (tph)

PAGE 56

56 0 4 8 12 16 20 2990 tph 1334 tph Nitrogen Content (kg ha-1) Intensive Operational A a cd e f 0 4 8 12 16 20 2990 tph 1334 tph Nitrogen Content kg ha-1) Intensive Operational B b c de ef Figure 2 8. Belowground nitrogen content at age two years demonstrating a significant three way interaction (p=0.0012) between silvicultural intensity, planting density, and location: A) Perry, Fl, and B) Waldo, FL. Data were averaged across six full sib families. Bars with the same letter across both locations were not significantly different at the 95% level of confidence using Bonferronis least significant difference (LSD).

PAGE 57

57 CHAPTER 3 FAMILY DIFFERENCES I N LIGHT INTERCEPTION AND RADIATION USE EFFICIENCY IN SEL ECTED LOBLOLLY ( Pinus taeda L.) PLANTATIONS IN THE SOUTHEASTERN UNITED STATES Introduction Over the past three decades plantation forest productivity in the southeastern United States has been enhanced significantly through the active management of nutrition, weed control, genetic improvement and density management (Colbert et al. 1990; Borders and Bailey, 2001; Jose et al. 2003; Jokela et al. 2004; Martin and Jokela, 2004b; Allen et al. 2005c; Fox et al. 2007) The aim has been to increase aboveground net primary productivity (ANPP) through rapid development of site occupancy and canopy leaf area index (LAI). Aboveground net primary productivity has been shown to be positively and linearly related to the amount of photosynthetically active radiation i ntercepted (IPAR) by the crowns (Cannell et al., 1987; Dalla Tea and Jokela, 1991; McCrady and Jokela, 1998; Will et al., 2005). Forest managers can manipulate LAI through a combination of genetic selection and the application of silvicultural treatments that ameliorate site resource deficiencies. There is evidence that specific crown attributes in individual trees may make them more efficient at resource use and these traits could be exploited in tree breeding and production purposes (Martin et al. 2001; Emhart et al. 2007) While this approach has been successfully implemented in the case of annual cereal crops (Green, 1989; Siddique et al., 1989), it has yet to be demonstrated in forested plantations (Cannell, 1989). The amount of photosyntheticall y active radiation ( PAR ) intercepted by a canopy is largely determined by the amount of foliage, as well as its distribution and orientation in the canopy (Sinclair and Knoerr, 1982; Colbert et al. 1990; McCrady and Jokela, 1996) The slope of the relati onship between ANPP and IPAR describes the efficiency of IPAR use in biomass

PAGE 58

58 production (Waring, 1983; Jarvis and Leverenz, 1983; Cannell et al. 1987; Cannell et al. 1988; Landsberg and Wright, 1989; Green et al. 2001; Binkley et al. 2004) (Equation 31) and is often referred to as Radiation Use Efficiency (RUE) (expressed in g MJ1ANPP = IPAR x RUE (3 1) of PAR intercepted) (Monteith, 1977; Sinclair and Muchow, 1999) As LAI and IPAR in forest plantations approaches a maximum, through the combination of deployment of elite genotypes and intensive silvicultural management, information about the efficiently of IPAR will become increasingly important. Recently, there has been evidence that RUE is influenced by canopy structural properties in fores t plantations (Chmura et al., 2007; Duursma and Makela, 2007, Duursma et al., 2010). The deep crowns of trees mean s that the relationship between individual leaf photosynthesis and absorbed light is strongly nonlinear and any increase in incident radiati on should theoretically lead to a decrease in RUE as a greater proportion of individual leaves in the canopy become light saturated (Monteith, 1977; Hollinger, 1989). However, in practice, this does not occur since not all the leaves in the canopy are lig ht saturated due to shading from leaves higher in the canopy (Wang and Jarvis, 1990; Hilker et al., 2008). Therefore, in order to influence RUE the entire canopy light environment should be optimized via a combination of leaf acclimation to the light envi ronment (Dewar et al., 1998) and a modification of canopy structure (Jarvis and Leverenz, 1983; Leverenz and Hinckley, 1990) This 'optimization' theory states that the greatest number of leaves should receive sufficient nonsaturating energy to function near their photosynthetic potential (Hollinger, 1989). This has been demonstrated in tropical trees in South America, where leaf physiology, morphology and orientation were optimized such that leaf level RUE was maximized (Posada, 2003).

PAGE 59

59 A quantifiable me asure of leaf orientation and canopy structure is the light extinction coefficient (k), which describes the rate of decreased irradiance vertically through a canopy (Kira et al. 1969; Falster and Westoby, 2003) This variable is influenced by leaf orient ation, foliage clumpiness and canopy structure (Okerblom and Kellomaki, 1983; Gower et al. 1999; de Castro and Fetcher, 1999) An erect display of foliage homogenizes the penetration of light through the canopy, which is especially effective at high levels of leaf area (Terashima and Hikosaka, 1995) In theory, RUE should increase with declining values of k (i.e. trending towards an erectophile orientation of leaves). Previous research in agricultural systems has demonstrated this effect in cultivars of wheat ( Triticum aestivum ) in Western Australia (Siddique et al., 1989) and rice ( Oryza sativa ) in Japan (Hayashi and Ito, 1962). It follows that the light extinction coefficient should play an important role in determining how much PAR is intercepted by a forested canopy (Gholz et al., 1991) and therefore should influence RUE in forest plantations (Dalla Tea and Jokela, 1991; Sands, 1996; McCrady and Jokela, 1998; Medlyn, 1998). This theory was supported in a study comparing native and hybrid poplar clones grown at high densities in southern Wisconsin (Green et al., 2001). However, canopy structure and light interception are often difficult to measure in detail, especially across wide areas and short time periods. Emerging remote sensing technologies such as Lidar (light detection and ranging) have advanced to the stage where detailed measurements of crown structure (Roth et al. 2007c) and light interception are now possible (Ahl et al., 2004; Lee et al., 2009) and should allow for more detailed investigations of variation in RUE in forested ecosystems. Radiation use efficiency in forests is known to be influenced by site conditions such as water availability, fertility, and climatic conditions (Monteith, 1977; Runyon et al. 1994; Prince and Goward, 1995; Balster and Marshall, 2000), which vary in time, elevation, soil type, and

PAGE 60

60 geographic location (Ares and Fownes, 2001; Stape, 2002; Stape et al. 2004; Pangle et al. 2009) Process based growth models modify RUE according to the stresses imposed by s ite resource limitations and climatic extremes in order to generate accurate estimates of production (Esprey et al. 2004) For example, drought stress has been shown to reduce RUE in sweetgum ( Liquidambar styaciflura ) and sycamore ( Platanus occidentalis ) in Georgia (Allen et al., 2005a; Allen et al., 2005b), tropical ash ( Fraxinus uhdei [Wenzing] Lingelsh) and Eucalyptus ( Eucalyptus camaldulensis ) in Hawaii (Harrington and Fownes, 1995; Ares and Fownes, 2001) and E. grandis x urophylla in Brazil (Stape et al., 2004). Likewise, nutrition has been shown to have a positive influence on RUE in forests globally. For example, significant gains in RUE have been attributed to increased soil nutrient availability in loblolly pine ( Pinus taeda L.) in the southeaste rn United States (Martin and Jokela, 2004a), Douglas fir ( Pseudotsuga menziesii var. glauca (Beissn.) Franco) in British Columbia (Balster and Marshall, 2000), Eucalyptus grandis in South Africa (du Toit and Dovey, 2005; du Toit, 2008), and Sitka spruce ( P icea sitchensis [ Bong.] Carr. ) in Central Scotland (Wang et al. 1991) However, in contrast, some investigations concluded that RUE was not influenced by nitrogen fertilization. For example, while nitrogen inputs increased aboveground productivity in st ands of sweetgum and s ycamore in Georgia, it did not influence RUE (Allen et al., 2005b). The same effect was reported in an experiment with loblolly and slash pine ( P elliottii Engelm. var. elliotttii) in north central Florida (Dalla Tea and Jokela, 1991). In these studies, the authors concluded that changes in growth following fertilization were due to increased LAI and light capture, rather than changes in RUE. Genotype x environment studies of RUE are uncommon in forested ecosystems and results have been inconclusive. Radiation use efficiency did not vary between a mixed species

PAGE 61

61 deciduous and a coniferous plantation forest type along a 260 meter elevation gradient in North Carolina (Pangle et al., 2009). While there were significant differences amon g forest types along a transect in Oregon, the experimental design did not allow for the examination of the stability of RUE among these genotypes across environmental gradients (Runyon et al., 1994). Radiation use efficiency was not stable among four tro pical species in Hawaii when grown under contrasting plantation establishment regimes (i.e. planted versus coppiced) (Harrington and Fownes, 1995) Similarly, RUE among four species (loblolly pine, slash pine, sweetgum, and sycamore) was dependant upon t he treatments of fertilization and irrigation (i.e. species x silviculture) in a trial in Georgia (Allen et al., 2005a). In that study, fertilization and irrigation increased RUE in sycamore, yet sweetgum was only influenced by fertilization, and the pine s were not affected. In a factorial species x irrigation x fertilization study in north central Florida, there were differences in RUE between loblolly and slash pine, but these did not interact with irrigation nor fertilization at age 6 years (Martin and Jokela, 2004a). The current investigation of light interception, crown attributes and RUE builds on that of McCrady and Jokela (1998), who found differences in RUE among loblolly pine families, but did not test for the stability of th ese traits across an environmental gradient The current investigation is unique, in that it is the first to examine the stability of loblolly pine performance grown under intensive silviculture in large full sib family block plots across locations of contrasting site propert ies The main objective of this study was to detect and describe the nature and extent of genotype x location interactions in loblolly pine I PAR and RUE at age four and five years Within this larger framework, specific objectives were to: 1) quantify NF LAI and k among families and between locations 2) quantify ANPP, the fraction of light intercepted ( f ), and I PAR among families and between locations, 3) quantify RUE among families and between

PAGE 62

62 locations, and 4) determine if family performance in these parameters wa s stable across locations. Materials and Methods Experimental Site and Design This investigation was carried out at two study locations in Sanderson, FL (29.28 0 N 82.33 0 W) and Waverly, GA (31.28 0 N 81.75 0 W). The topography at both loca tions was nearly flat with slopes less than 1%. The soils at the S anderson, FL and W averly, GA locations were mapped as the Leon (sandy, siliceous, thermic Aeric Alaquods; formed in thick beds of acid sandy marine sediments) and Bladen series (mixed, semi active, thermic Typic Albaquults; formed from deposits of clayey fluvial or marine sediments), respectively. The sandy soils at the Sanderson, FL location were generally poor in fertility and trees responded well to nutrient additions. The clay loam soils at the Waverly, GA location were acutely phosphorus deficient, but supported high and sustained levels of productivity when fertilizer additions were made (Pritchett and Comerford, 1982; Jokela et al. 1989) The climate varied little between locations a nd was characterized as humid and subtropical, with average annual temperatures ranging from 19 to 21 0The field experiments used in this investigation were part of a larger set of genotype x silviculture x planting density studies (Roth et al. 2007a) In this analysis, four complete blocks per location were sampled and consisted of six full sib loblolly pine families grown under intensive silvicultura l management and at an initial planting density of 2990 trees ha C. Long term annual precipitation, from 1931 to 2000, averaged 1384 mm across locations (NOAA, 2002). 1. Trees within a family block were planted at a 1.22 m x 2.75 m spacing and were arranged in eight beds of 16 planting positions each, for a total of 128 trees per gross treatment plot. A two tree border around the perimeter resulted in a 48 tree interior measurement plot of 0.016 ha. Prior to

PAGE 63

63 planting, the study sites were double bedded on separate passes following a 2.75 meter spacing. In the late summer/early fall of 1999, each locatio n was treated with preplant herbicides consisting of Chopper (imazapyr) at 1.02 1 and Garlon (triclopyr) at 7.02 1, with the goal of removing all woody competition and reducing initial levels of herbaceous vegetation. The objective was to min imize within location environmental variation caused by competition for site resources from associated vegetation For two years following planting, competing vegetation was controlled using directed applications of Arsenal (imazapyr) at 0.28 1 and Oust (sulfometuron methyl) at 0.14 1, with the objective of maintaining ground cover below a 30% threshold. At the time of planting, the plots were fertilized with 560 kg ha1All genetic entries in the study were selected from s ources exhibiting moderate to excellent resistance to fusiform rust [ Cronartium quercum (Berk.) Miyabe ex Shirai f. sp. fusiforme] based upon a priori knowledge from tree breeding programs and progeny tests This was done in order to reduce the confoundin g effects of disease incidence. Seedlings were grown in Ray Leach 'Cone tainer'TM cells (Stuewe and Sons, Inc Corvallis, OR), consisting of 66 ml cell of 101010 plus micronutrients, which was followed by annual applications of macro and micronutrient fertilizers based on prescriptions developed from foliar analyses. The total amounts of nutrients applied on each installation through age five are presented in Table 31. 1Insecticides were uniformly applied across all treatments on loblolly pine installations in an effort to control damage from the Nantucket pine tip moth ( Rhyacionia frustrana [Comstock]). Treatments were applied on a monthly basis over the first two growing seasons, beginning in March and ending in September. Alternating applications of the following and were hand planted over a two day period in January 2000. Overall survival following planting was greater than 95 % with low mortality rates in the subsequent year s

PAGE 64

64 chemicals and application rates were applied aerially or by hand: Pounce 3.2EC (62 ml pr oduct 1 water), Warrior T (39 ml product 1 water), Dimilin 25W (62 ml product 1 water), and Mimic 2LV (125 ml product 1Inventory and Aboveground Biomass Estimates water). Diameter (DBH) was measured annually at ages three (2003), four (2004) and five (2005) years on all trees in the 48 tree measurement plots. Total height (HT) was measured on all trees at age 3 and a random 20 % subsample at ages 4 and 5. Individual tree HT at ages 4 and 5 were estimated using site specific HT vs. DBH regressi on models developed from this subsample. Aboveground biomass estimates (AGB), expressed in units of Mg ha1Needlefall, Leaf Area Index, and Ab oveground Net Primary Production of dry matter, were developed using linear allometric equations developed from these same studies (Roth et al. 2007a) Additional variables such as surviving tree density, height to live crown, crown ratio, and projected crown area at age five years were measured in order to help explain any associations with treatment related differences via covariate analysis. Needlefall (NF) was collected an average of six times per year over a three year period (June 2003 to February 2007) to estimate leaf area index (LAI). Within each family measurement plot, six circular littertraps (1.0 m2 each) were deployed, with half randomly positioned along the bed and inner bed positions. Litter was almost exclusively pine foliage (> 98%) and was separated from other pine material (branch, bark, twigs and cones), oven dried at 700Foliar biomass accretion was modeled using logistic equations fitted to measured NF data (Kinerson et al. 1974; Dougherty et al. 1995). This approach assumed that: (1) within a given year all needles that forme d had died and senesced from the tree by February 28, two calendar years following their formation (i.e. maximum needle lifespan was no greater than two years), C and weighed to the nearest 0.1 g.

PAGE 65

65 2 1 1 2 t t t tNF AGB AGB ANPP and (2) that the accretion curve was not affected by family or location, (3) and the phenological year for foliage accreti on began on March 1 (Martin and Jokela, 2004a). Therefore, total NF in a given phenological year (March 1 February 28) represented total needle production in the previous year. In this study, we were able to calculate foliage biomass accretion over a t wo year period (ages 4 to 5 yrs). Needlefall data were corrected for losses due to senescence, which were estimated to be 14% for loblolly pine (Dalla Tea and Jokela, 1991). Projected LAI was calculated as the product of NF mass (based on 1.0 m2 litter traps) and family level specific leaf areas (SLA) measured within each location. Specific leaf area (m2 g1 (3 2) ) was determined using the volume displacement method (Johnson, 1984) on foliar samples that were collected from individual families in the fall of 2 003. Aboveground net primary productivity (ANPP) was calculated as the difference between unit area standing biomass (AGB) on an annual basis, which included annual NF (Equation 32): While mortality was accounted for, herbivory w as as sumed to be insignificant in these young trees during this period, and was not included in this calculation of ANPP (Clark et al. 2001) Fraction of PAR Intercepted, Light Extinction Coefficient The fraction of PAR intercepted ( f ) was calculated from meas urements of above and below canopy PAR (Equation 3 3): f = 1 (PAR transmitted through canopy/ PAR incident at top of canopy). (3 3) The amount of PAR transmitted through the canopy was measured using a handheld Sunfleck Ceptometer (Decagon Devices 1987), which measured below canopy PAR of each family block plot (at 1.37 meter elevation). Incident PAR above the canopy was measured in an adjacent open area using a quantum light sensor (LI COR 190SA) linked to a data recorder which continuously

PAGE 66

66 recorded P AR on a 5 second interval, and stored averages per minute. At regular intervals during the sampling period, measurements of PAR with the two sensors were conducted side by side to ensure concurrence. Six permanent transects running perpendicular to the beds were sampled and recorded for each plot, which included gaps in the canopy. A mean of eight subsamples were averaged along each transect and the six transects averaged to obtain a per plot value. Data were collected on a regular interval during the growing seasons (March through October) of 2004 and 2005 under cloud free conditions between 1100 and 1400 h solar time (Table 3 2). The light extinction coefficient ( k ) was empirically determined using an equation derived from the Beer Lambert law (Equat ion 34) and measured values of projected LAI (all sided/pi;(Grace, 1987) and f ): k = (log( 1 f ))/ LAI (3 4) Since k varies by the angle of the sun (Stenberg et al. 1994), s eason specific sun angle values of the extinction coefficient k were cosine corrected ( k /cos 500) for the solar zenith noon at each sampling date. The Beer Lambert law assumes a homogenous uniform distribution of leaves along the path of light (Kira et al. 1969). This assumption is easily violated in pine forests, especially i n mature stands where the canopy is highly aggregated (Gholz et al. 1991). However, given that the current investigation was carried out in dense young stands planted to a regular spa cing with a uniform canopy, it wa s assumed that the Beer Lambert assump tions were met and that the estimates of k were valid for comparing relative family and silvicultural treatment effects (Gower et al. 1999; de Castro and Fetcher, 1999). In a related study the methodology for estimating IPAR and k using the Beer Lambert law w as validated in a young loblolly pine experiment in north central Florida (Martin and Jokela, 2004a).

PAGE 67

67 Amount of PAR Intercepted and Radiation Use Efficiency Estimates of intercepted PAR over each growing season (MJ m2 y1) were calculated by summing monthly estimates of incoming shortwave radiation (total direct and diffuse solar radiation received on a horizontal surface ) and multiplying by the plot level fraction of PAR intercepted ( f ), centered around each sampling period (monthly values interpola ted) (Table 3 2). The s atellite derived solar estimates were obtained for each study site using 10 km modeled grid data available from the National Solar Radiation Data Base ( NSRDB ), 1991 2005 update ( ftp://ftp.ncdc.noaa.gov/pub/data/nsrdbsolar Last accessed March 2009) We assumed that incoming PAR was 50% of the total incoming shortwave radiation (Landsberg and Waring, 1997; Waring and Running, 1998). Growing season (March to October) PAR interception was used since ceptometer measurements were problematic when the sun angle was low during the winter months. The consequence of this was that values of RUE would be slightly higher than published estimates of RUE using year long IPAR data. Aboveground radiation use efficiency of each family was determined annually by dividing ANPP by the sum of intercepted PAR over the growing season (March October) and was expressed as g MJ PAR1Statistical Analyses Analysis of variance (ANOVA) was used to test for stand level differences in the variables of interest between locations and among families, including their interactions at ages 4 and 5 years using PROC MIXED (Littel et al. 1996) in SAS. Two forms of the mixed linear model were used, since some variables were measured on a monthly basis and others on an annual basis. The response variables of LAI, k and f were examined utilizing Equation 3 5: Yijklm n i + b(l)ij + Fk + Yl + LFik + LYil + b(l)Fijk + b(l)Yijl + FYkl + Mm + FMkm + LMim + wijklmn (3 5)

PAGE 68

68 where Yijklmn is the response variable (LAI, k and f ) of the nth plot of the m th month of the l th year of the k th family of the j th block of the i th location ( i = 1,2; j = 1,2, 4; k = 1,2, 6; l = 1,2; m = 1,2, 4; and n = 1); is the overall mean; Li is the fixed effect of the i t h location; b(l)ij is the random interaction effect of the j th block within the i th location; Fk is the fixed effect of the k th family; Yl is the fixed effect of the l th year; Mm is the fixed effect of the m th month; and wijklmnThe response variables of NF, ANPP, and RUE were examined utilizing Equation 36: is the random error. Yijklm i + b(l)ij + Fk + Yl + LFik + LYil + b(l)Fijk + b(l)Yijl + LFYikl + COVijklm + w (3 6) ijklm where Yijklm is t he response variable ( NF, ANPP, and RUE ) of the m th plot of the l th year of the k th family of the j th block of the i th location ( i = 1,2; j = 1,2, 4; k = 1,2,, 6; l=1,2; and m = i is the fixed effect of the i th location; b(l )ij is the random interaction effect of the j th block within the i th location; Fk is the fixed effect of the k th family; Yl is the fixed effect of the l th year; COV is the covariate of interest where applicable; and wijklmFor both models, b locks were nested within locations. Block within locations b(l) is the random error. ij was considered a random effect, as are all terms containing b(l)ij. Where a covariate was significant at p=0.05 it was included in the model, otherwise it was excluded (the only significant covariate was age five stand density on RUE). Where assumptions of normality and constant variance were not met the variables were log transform ed prior to analysis Significant effects were examined comparing least squares means using Bonferroni s adjusted significance level.

PAGE 69

69 Results Needlefall, Leaf Area Index, and Aboveground Net Primary Production Variation in annual NF was dependant on the combined effects of location and year (p <0.0001, location x year interaction), and to a lesser degree f or the main effect of family (p=0.1054). In 2004, the greatest NF occurred at the Sanderson, FL location (5.12 versus 3.88 Mg ha1 yr1 for Sanderson, FL and Waverly, GA respectively); however, in 2005 this trend had reversed, with the Waverly, GA locati on producing the greatest NF (3.72 versus 5.05 Mg ha1 yr1 for Sanderson, FL and Waverly, GA respectively). Differences in family level NF were stable across years (p=0.1332, genotype x year) and locations (p=0.7533, genotype x location). Needlefall am ounts among families, averaged across both locations and years, varied relatively little and ranged from 4.05 to 4.63 Mg ha1 yr1There were significant location x year and location x month interactions for mean projected LAI (p=0.0002 and p<0.0001, respectively). The location x year interaction was caused by differences in LAI at the Sanderson, FL location between years. While LAI in 2004 was similar between locations, the amount of LAI in 2005 had decli ned by approximately 20% at the Sanderson, FL location (Table 34). Plotting the trend over time (Figure 31a) demonstrated that this decline at the Sanderson, FL location began in September 2004. By the following year (2005), the decline in LAI at Sande rson, FL had stabilized and then tracked parallel to the trends observed at the Waverly, GA location. The significant location x month interaction was caused by a steeper decline in LAI at the Sanderson location in the fall months (Figure 31b). for families L1 and L7, respectively (Table 3 3). Additiona lly, mean projected LAI between years was not stable among families (p=0.0218, genotype x environment interaction). While LAI decreased for all families, LAI significantly deceased in 2005 for families L5, L7 and L2 (19, 17, and 14%, respectively) (Table 3 5). In contrast, family L1 was the most stable with an LAI of 1.9 m2 m2 in 2004 vs. 1.7 m2 m2 in 2005.

PAGE 70

70 Unlike NF and LAI response, there were no interactions between factors for ANPP; only the main effects of year (p=0.0079) w as statistically signific ant. There was a 10% decline in overall ANPP from year to year (20.9 to 18.8 Mg ha1 yr1Fraction of PAR Intercepted, Light Extinction Coefficient for 2004 and 2005, respectively). Family performance for ANPP remained stable across locations and years with subtle differences (10%) between the top producing and poorest growing family (T able 33). There w ere location x year (p=0.0001) location x month (p=0.0007), and family x month (p=0.0053) interaction s for the fraction of PAR intercepted. As with L AI, the variable response in f across time occurred at a single location (Sanderson, FL). When averaged across families, t he fraction of PAR intercepte d at the Sanderson, FL location had declined from 0.739 in 2004 to 0.667 in 2005 (Table 34). Averaged across years, the Sanderson, FL location had a steeper decline in f in the month following peak LAI (August) than did the Waverly, GA location (Figure 3 4a). Families were not stable across months averaged between years with the overall greatest instabi lity occurring in March and the least in September. Overall, family L4 was the most stable with families L1 and L2 the least and families L5, L7, and L8 intermediate (Figure 35). Lo cations were un stable across years and months for the sun angle cosin e corrected light extinction coefficient ( k ) (p<0.0001) The location x year interaction was again influenced by the Sanderson, FL location, where significantly higher k values existed in 2005 than at the contrasting Waverly, GA location (0.83 versus 0.44) (Table 3 4). Values of k between locations in 2004 were similar averaging between 0.42 a nd 0.44. The location x month interaction was also due to instability at the Sanderson, FL location. Averaged across both years, k was significantly higher in the month of May at the Sanderson, FL location ( 0.88) than at the Waverly, GA location (0.60) (Figure3 4b). There were large differences in growing season k

PAGE 71

71 among families which ranged from 0.43 for family L4 and 0.67 for family L1 (Table 33). Low values o f k (tending toward an erectophile display of foliage ) allow for a deeper penetration of light through the canopy than higher values (tending toward planophile). Amount of PAR intercepted and Radiation Use Efficiency The total amount of PAR intercepted by the canopy during the growing season varied by the main effect of family (p=0.0291) and demonstrated a location x year interaction (p=0.0002). Family variation in IPAR ranged from 1719 to 1857 MJ m2 yr1 for families L1 and L8, respectively (Table 3 3) Family L7 performance was similar to L8 and it intercepted 1856 MJ m2 yr1. The Sanderson, FL location was the most reactive across years for I PAR (Table 3 4) At this location, a reduction in the amount of PAR intercepted occurred in 2005 (from 1639 to 1819 MJ m2 yr1Aboveground net primary productivity was positively related to IPAR (Figure 32) and the slope of this relationship (RUE) was different among families (p=0.0161), with efficiency values ranging from 1. 08 to 1.16 g MJ from 2004 to 2005, respectively) 1 (families L8 and L4 respectively) (Table 3 3). For a given amount of I PAR, family L4 produced more aboveground biomass than did family L8 when averaged across locations and years. There were no location x year interactions for RUE and families did not vary in RUE across locations or years. However, covariate analysis indicated a strong positive influence of surviving tree density in 2005 on RUE (p=0.0005) (Figure 33). This was the only covariate that was significant for any variable examined. Radiation use efficiency increased by about 0.2 g MJ1 across the range of observed tree density (1334 to 2990 trees ha1Discussion ) in this study at age five. This investigation provided the unique opportunity to examine the stability of PAR i nterception and RUE in selected elite full sib families of loblolly pine when deployed in

PAGE 72

72 uniform blocks and grown under an intensive silvicultural treatment regime over time and contrasting locations (see Appendix B and C) This study provided information about the extent that genotype x environment interactions influenced RUE, which can be particularly useful in the development and utilization of process based growth models (Landsberg and Hingston, 1996; Landsberg and Waring, 1997; Esprey et al. 2004). While there was variation among families, there was no evidence for genotype x environment interactions (i.e. genotype x location, genotype x year or genotype x month) in RUE The annual NF observed in this investigation ranged from 3.7 to 5.1 Mg ha1, w hich corresponds with published literature for loblolly pine: 4.6 Mg ha1 at age six (DallaTea and Jokela, 1991; Jokela and Martin, 2000) and 4.7 to 5.7 Mg ha1 among selected families at age five (McCrady and Jokela, 1998). Annual NF at age four was about 5.0 Mg ha1Projected LAI varied among years and locations and ranged from a low of 1.7 to a maximum of 3.3 m for family L4 in a separate experiment (Adegbidi et al. 2005). There was a large decline in NF at the Sanderson, FL location in 2005, and climatic conditions have been shown to alter levels of annual NF (Dougherty et al. 1995). This effe ct was most likely due to early leaf senescence related to the effects of hurricanes Frances and Jeanne in 2004. Since the Waverly, GA location was further away from the storm paths, it was largely unaffected by these severe windstorm events (Roth et al., 2007a). The effect of declining NF following these hurricanes was also observed and documented in another trial in Florida (Li et al. 2007). 2 m2. These v alues were lower than those reported for loblolly pine of around 4.0 m2 m2 at a similar stage of stand development (Allen et al. 2005a) and yet were within the upper limit of 3.2 m2 m2 observed in a series of intensively managed loblolly pine plantation s at age four years (Adegbidi et al. 2002). Following the normal pattern of leaf

PAGE 73

73 accretion over a growing season, values of LAI increased to a maximum in August and declined to a minimum in March. However, as with the observed pattern of NF, the decline in LAI in the fall at the Sanderson, FL location was much more dramatic than at Waverly, GA. Since the Sanderson, FL location is inherently less fertile, this may be due in part to the curtailment of fertilization in 2003 and the hurricanes in 2004 (fert ilization was resumed in 2005). In a nearby trial, loblolly pine LAI declined significantly within a year following curtailment of fertilization and declined by as much as 12.6% after four years (Martin and Jokela, 2004b). It follows that the amounts of nutrients added in 2003 may have also been insufficient to meet the demands of these rapidly developing crowns and stands, especially on the inherently nutrient poor site at Sanderson, FL. Examination of LAI on the family level between years, demonstrate d that certain families were able to hold higher levels of LAI despite the overall decline in 2005 across both locations. As predicted by McCrady and Jokela ( 1998), it follows that these were the most productive families and could be characterized as hav ing the crown structural properties of crop ideotypes as described by Martin and others (2001). Crown structural properties influence the amount of light intercepted by individual trees within a forest canopy and play important roles in determining the competitiveness of individual trees (Cannell, 1978). Traits such as the distribution of leaves within the canopy are under some level of genetic control (Emhart et al. 2007; Chmura and Tjoelker, 2008). For example, in a companion study, which compared elite families of loblolly and slash pine in the Western Gulf region of the United States, family L4 exhibited the largest number of flushes and had a different crown shape with longer branches in the midcrown position (Chmura et al. 2007). The authors suggested that this most likely led to better lightcapture and greater carbon assimilation and aboveground biomass accumulation.

PAGE 74

74 The current investigation examined differences in the light extinction coefficient k and how it was influenced by the combined effects of family, location and year. Cosine corrected values of k averaged about 0.44 across locations and years, with the exception of a significantly high value which occurred at the Sanderson, FL location in 2005 ( k =0.83). With the exception of this outlying value, these estimates were in general agreement with the published literature. A mean of 0.33 was documented among five loblolly pine families in a four year old experiment (McCrady and Jokela, 1998). In a 14year old loblolly pine experiment k values ranged from 0.36 to 0.64 (Sampson and Allen, 1998), while a k of 0.46 was documented in a 15year old loblolly pine plantation in North Carolina (Sinclair and Knoerr, 1982). Across several lodgepole pine stands in Wyoming, k varied from 0.3 to 0.7 for a wide range of stand and site conditions, with lower k values associated with higher values of LAI (Sampson and Smith, 1993). This was also documented among five elite families of loblolly pine, where families with the largest LAI also had the low est values of k (McCrady and Jokela, 1998). In theory, low values of k will allow for a deeper penetration of light into the canopy and therefore greater RUE given similar amounts of IPAR. In support of this theory, variation in k has been shown to be re lated to RUE among five cultivars of wheat (Green, 1989) and five hybrid poplar clones (Green et al. 2001). In these experiments, RUE was greatest in cultivars and clones with lower values of k despite similar amounts of IPAR. There was some evidence i n the current investigation to support this trend as family L4 exhibited the greatest values of RUE combined with the lowest k Aboveground net primary productivity was similar among the families examined in this investigation. This likely resulted becaus e the families were all chosen for superior performance in growth. Similarly, the silvicultural treatments applied in this experiment, which combined high planting density with intensive silvicultural management, likely evened the inherent

PAGE 75

75 differences in site quality between the locations. Presumably, this may account for the lack of observed differences in ANPP between locations. The lack of a genotype x location interaction in ANPP was supported in a recent study which compared deciduous and pine fores t types along an elevational gradient in North Carolina (Pangle et al. 2009). The authors noted no differences between forest types for ANPP; however, there were large differences in RUE. Individual families varied in their RUE, but were stable between l ocation s and year s (i.e. no genotype x environment interaction). Family level growing season RUE values in this study ranged from 1.08 to 1.16 g MJ1; however, due to differences in methodology in the calculation of IPAR (most studies report RUE on an annual basis), direct comparisons of RUE among other studies require s interpolation. With this in mind, the RUE values reported in this investigation were generally consistent with others reported for loblolly pine. They were bounded by annual estimates rep orted by Martin and Jokela (2004a) and Dalla Tea and Jokela (1991) of 0.40 g MJ1 and 0.81 g MJ1, respectively in north central Florida. The values of RUE in the current investigation were also greater than those reported for loblolly pine in Georgia of 0.795 g MJ1 (Allen et al. 2005a). In contrast, McCrady and Jokela (1998) documented RUE ranging from 1.33 to 1.48 g MJ1 among five loblolly pine families in South Carolina, and Allen et al. (2005b) reported values ranging from 1.39 to 1.74 g MJ1 for s ycamore and sweetgum in Georgia, respectively T he total amount of variation in RUE among the families tested in this experiment was marginal. A similar effect was found among poplar clones; however, those were from random sources which had not been sele cted for improved growth (Landsberg and Wright, 1989). The lack of large differences in RUE among the families in the current experiment may be due to the highly productive nature of the elite loblolly pine families that were selected for inclusion in the experiment which did not vary in ANPP At a species level, RUE and growth

PAGE 76

76 efficiency among genotypes tends to vary the greatest where contrasting species (Allen et al. 2005a; Cobb et al. 2008) and families (McCrady and Jokela, 1998) were tested. Evide nce from other studies suggests the greatest responses in RUE have occurred where the most limiting site resources have been manipulated (Stape, 2002; du Toit, 2008). For example, fertilization increased RUE of stemwood production of Douglas fir across a series of locations with wide range of site quality in the in terior region of the Pacific Northwest USA (Balster and Marshall, 2000) and in an eight year old loblolly pine experiment on a nutrient poor, deep sandy soil in North Carolina (Albaugh et al. 2004). However, there was no effect of fertilization and weed control on the RUE of loblolly and slash pine on three locations in the Western Gulf of the United States (Chmura and Tjoelker, 2008) possibly since all three were productive sites In contrast this hypothesis was not supported for two experiments growing on nutrient poor locations where fertilization did not influence RUE (Dalla Tea and Jokela, 1991; Allen et al. 2005a). Similar to anthropogenic manipulations of site resources, natural variat ion along resource gradients can influence RUE in forests (Stape et al. 2004). While RUE was stable along a 260 m elevational gradient in two deciduous and pine forest types (Pangle et al. 2009), this was not the case along a much steeper elevational gr adient in Hawaii (Ares and Fownes, 2001) In this latter investigation conducted with tropical ash, RUE was reduced at the higher elevations where site resources were most limited. In keeping with the theory that the greatest interactions for RUE occur under conditions where resources are most limiting, RUE varied the most on nutrient poor Andisols, as compared to the relatively nutrient rich Histosols (Ares and Fownes, 2001). The stability of RUE across locations documented in the current investigation, despite inherent differences in productivity between the soil types, may be due to the leveling effects of the

PAGE 77

77 intensive silviculture treatments which ameliorated limiting site resources on the least productive site (see chapter 2) The current investiga tion noted a trend of increasing RUE with increasing tree density after five years from planting (Figure 3 3). This effect has been documented previously where closely spaced stands tended to be more efficient at capturing and utilizing PAR than widely sp aced stands (Burkes et al. 2003), perhaps due to less variation in horizontal needle distribution (Smith and Long, 1989; Smith et al. 1991; Smith and Long, 2001). Part of this phenomenon may be that stands growing at higher densities tend to have a more even distribution of foliage within their canopies, which leads to increased efficiency in light interception (Will et al. 2001; Will et al., 2005). However, it has also been demonstrated that growth efficiency varies with changes in biomass allocation due to increasing planting densities (Burkes et al. 2003). Yet, in older stands, RUE in loblolly pine was found to decrease with stand basal areas exceeding 18 m2 ha1 (Martin and Jokela, 2004a). The authors demonstrated that this association was driven by decreases in woody biomass increment due to stand aging at full s tocking. This was also demonstrated among poplar clones in China, where growth efficiency declined at the highest planting densities (Fang et al. 1999). Results from the current study suggest that there is room to exploit differences in RUE among selected families of loblolly pine and that RUE may be stable across years and locations when grown under intensive silvicultural management. It follows that, a s LAI reaches a biological maxiu m due to the combined effects of the deployment of elite genotypes grown under intensive silvicultural treatments, manipulation of RUE may become more important for forest production, and therefore the mechanisms behind this phenomenon warrant further investigation.

PAGE 78

78 Conclusion s This investigation examined the influence of location and year on NF, LAI, k f IPAR, ANPP, and RUE among six elite full sib loblolly pine families grown in fullsib family blocks grown under close spacing and intensive silvicultur al management tested at two unique locations and soil types G enotype by environment interactions were evident : i.e., there w as a genotype x year interaction for LAI and also a genotype x month interaction for the fraction of light intercepted. Trends over time were not stable between locations, with location x year interactions evident for all variables, with the exception of ANPP and RUE. Additionally, f and k were not stable across months between locations (i.e. location x month interaction). Radiati on use efficiency among families was stable across time and locations, which is in general agreement with results from other studies where RUE was stable across a range of environmental conditions (Pangle et al. 2009). The current investigation makes an important contribution to the field by documenting family level loblolly pine stability of RUE through time and across locations of contrasting soil types. Future investigations at these locations should examine the interacting influence of silvicultural intensity, planting density, family and locations on canopy structure and RUE. Investigations which link specific canopy structural traits involved in light capture and photosynthesis with variation in natural and anthropogenic factors that influence RUE would be informative decisions regarding the deployment of elite southern pine genotypes

PAGE 79

79 Table 31. Cumulative elemental nutrient application rates for the PPINES intensive silvicultural treatments through five growing seasons (kg ha1 L ocation ). N P K Mg Ca S B Zn Mn Fe Cu Sanderson, FL 369 128 121 45 45 35 0.9 2.7 2.2 14.7 3.9 Waverly, GA 369 128 121 45 45 35 0.9 2.7 2.2 14.7 3.9 Table 32. Summation of i ncoming monthly shortwave radiation per month in MJ m2 L ocation for two locations ( Sanderson, FL and Waverly, GA) and two years (2004 and 2005). Each location is part of a larger family block study investigating the effects of silvicultural treatment intensity and planting density on the performance of full sib loblolly pine families. Year March April May June July August September October S anderson, FL 2004 592 662 756 650 655 537 413 447 2005 490 654 691 573 680 589 554 414 W averly, GA 2004 607 669 777 649 684 551 419 462 2005 489 665 710 639 709 618 5 55 408 Data were obtained for each study site using the 10 km modeled grid data available from the NSRDB (The National Solar Radiation Data Base) (1991 2005). ftp://ftp.ncdc.noaa.gov/pub/data /nsrdbsolar Last accessed March 2009 Table 33. Least squares means for needlefall, light extinction coefficient, amount of PAR intercepted aboveground net primary productivity, and radiation use efficiency for six fullsib loblolly pine families averaged across two locations (Sanderson, FL and Waverly, GA) and two years (2004 2005). Families were stable across locations and years (i.e. no genotype x environment interactions). Family NF k IPAR ANPP RUE L4 4.58 a 0.43 a 1808 ab 21.0 a 1.16 a L1 4.05 a 0.67 c 1719 b 19.3 a 1.12 ab L2 4.33 a 0.60 bc 1798 ab 19.6 a 1.09 ab L5 4.46 a 0.52 abc 1791 ab 19.5 a 1.09 ab L8 4.60 a 0.47 ab 1857 a 20.2 a 1.08 ab L7 4.63 a 0.51 ab 1856 a 19.5 a 1.08 b NF is mean annual needlefall in Mg ha1, k is cosine corrected light extinction coefficient, IPAR is the sum of PAR intercepted by the canopy (MJ m2), and ANPP is expressed in Mg ha1. Radiation use efficiency (RUE) is expressed in g MJ1PAR and is adjusted for the covari ate of tree density which was fixed at 2840 trees ha1. Values with the same letter within columns are not significantly different at the 95% level of confidence using Bonferronis LSD.

PAGE 80

80 Table 34. Least squares means for needlefall, leaf area index, lig ht extinction coefficient, fraction of PAR intercepted, amount of PAR intercepted, aboveground net primary productivity for the interaction between location (S anderson, FL and W averly, GA ) and year (2004 and 2005). Location Year NF LAI k f IPAR Sanderson FL 2004 5.12 a 2.08 a 0.44 a 0.739 a 1819 a 2005 3.72 b 1.66 b 0.83 b 0.667 b 1639 b Waverly GA 2004 3.88 b 2.22 a 0.42 a 0.735 a 1888 a 2005 5.05 a 3.26 ab 0.44 a 0.740 a 1872 a Needlefall (NF) is mean annual needlefall in Mg ha1, LAI is mean growing season projected, k is cosine corrected light extinction coefficient, f is the mean fraction of light intercepted during the growing season, and IPAR is the sum of PAR intercepted by the canopy expressed in MJ m2 Values with the same letter within columns are not significantly different at the 95% level of confidence using Bonferronis LSD. Table 3 5. Least squares means for Leaf Area Index (LAI) for six full sib loblolly pine families in 2004 and 2005, averaged across two locations (Sanderson, FL and Waverly, GA). Families were stable across locations (i.e. no genotype x location interaction). Family 2004 2005 L4 2.1 ab 1.8 bc L8 2.2 ab 1.9 abc L5 2.1 ab 1.7 c L7 2.3 a 1.9 bc L2 2.1 ab 1.8 c L1 1.9 abc 1.7 c Leaf area index is on a projected basis Values with the same letter within variables are not significantly different at the 95% level of confidence using Bonferronis LSD.

PAGE 81

81 1.0 1.5 2.0 2.5 3.0 Jan-04 Apr-04 Aug-04 Nov-04 Feb-05 May-05 Sep-05 Dec-05 Mar-06 LAI (m2 m-2) Sanderson, FL Waverly, GA A 1.2 1.8 2.4 3.0 2 3 4 5 6 7 8 9 10 11 Month LAI (m2 m-2) Sanderson, FL Waverly, GA A B BC DE C D E E B Figure 3 1. A) Projected leaf area index (LAI) over a two year period for the Sanderson, FL and Waverly, GA locations. Within each location, the data were averaged across six full sib loblolly pine families. The location x year interaction for mean LAI was significant at p <0.0001. Note the effects of hurricanes Frances and Jeanne were evident for the periods immediately following the September 2004 data point at the Sanderson, FL location. B) Relationship between projected leaf area and month demonstrating the location by month inter action (P<0.0001). Each data point represents a single location averaged across six families and two years (2004 and 2005).

PAGE 82

82 12 14 16 18 20 22 24 26 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 IPAR (MJ m-2 yr -1) ANPP (MG ha -1 yr -1) L1 L2 L4 L5 L7 L8 Figure 3 2. Relationship between annual aboveground net primary production and intercepted radiation for loblolly pine ful l sib families L4 (solid line) and L7 (dotted line) Each data point with the same symbol represents a single fullsib family plot (n=6) at one of two years (2004 and 2005) and two locations (S anderson, FL and W averly, GA ). There was a statistically sign ificant difference in the slopes (RUE) between families L4 and L7 (see Table 3 3 ).

PAGE 83

83 0.7 0.9 1.1 1.3 1.5 2200 2400 2600 2800 3000 Density (trees ha-1) RUE (g MJ-1) Figure 33. Relationship between RUE and tree density at age five years for loblolly pine. Each data point represents a single family plot at one of two locations ( Sanderson, FL and Waverly, GA) and the solid line represents the least squared regression. RUE = 0.0003 (density) + 0.3606 (r2 = 0.1507)

PAGE 84

84 0.60 0.65 0.70 0.75 0.80 0.85 2 3 4 5 6 7 8 9 10 11 Month Fraction of light intercepted Sanderson, FL Waverly, GA A DE AB BC CD F EF AB A 0.0 0.2 0.4 0.6 0.8 1.0 2 3 4 5 6 7 8 9 10 11 Month Extinction Coefficient Sanderson, FL Waverly, GA A B BC F CD D E E B Figure 3 4. Relationship between : A) fraction of light intercepted ( f ), and B ) light extinction coeffici ent ( k ), by month of year for two locations (p<0.0007 and p=0.0001 respectively). Each data point represents a single location averaged across six families and two years (2004 and 2005).

PAGE 85

85 0.6 0.7 0.7 0.8 0.8 0.9 2 3 4 5 6 7 8 9 10 11 Month Fraction of light intercepted L1 L2 L4 L5 L7 L8 AB BC A CD D D A A A A A A A B C Figure 3 5. Relationship between the fraction of light inter cepted ( f ) by month of year for six loblolly pine families (p=0.0053). Each data point represents a single family averaged across two locations (Sanderson, FL and Waverly, GA) and two years (2004 and 2005). RUE and tree density at age five years for lobl olly pine.

PAGE 86

86 CHAPTER 4 GENOTYPE X ENVIRONMENT INTERAC TIONS IN SELECTED LOBLOLLY ( Pinus taeda L.) AND SLASH PINE ( P. elliottii ENGLM. VAR. elliottii ) PLANTATIONS IN THE SOUTHEASTERN UNITED STATES1. Introduction Considerable gains in the productivity of loblolly (Pinus taeda L.) and slash pine ( Pinus elliottii Englm. var. elliottii ) plantations in the southeastern United States have been achieved over the past 30 years. Demonstrated increases in unit area production have been realized through silvicultural inputs of fertilization, competition control, and density management. These treatments were designed to relieve site res ource limitations in order to focus growth on crop trees (Colbert et al. 1990; Jokela et al. 2000; Borders and Bailey, 2001; Martin and Jokela, 2004b). Growth responses to intensive silvicultural practices range from 2to 3.5-fold at age 15 for loblolly pine in the southeastern USA (Jokela et al. 2004). Additionally, tree breeding programs have increased volume production by 10 to 30% over unimproved sources (Li and McKeand, 1989; McKeand et al. 2003a). When a combination of elite genetic materials are combined with site-specific silvicultural trea tments, mean annual increments of up to 20 m3ha1yr-1 have been documented (Allen et al. 2005c). However, as resource managers begin to deploy selected full-sib families or clones (Bridgwater et al. 2005), there is a greater likelihood that genotype x environmental (GxE) interactio ns will occur, especially under conditions of increased silvicultural intensity (McKeand et al. 2006). These interactions may be manifest as rank changes among genotypes when grown under different environments/treatment conditions, 1 Reprinted with permission from Roth,B.E., Jokela,E.J., Martin,T.A., Huber,D.A., White,T.L., 2007. Genotype x environment interactions in selected loblolly and slash pi ne plantations in the southeastern United States. Forest Ecology and Management 238, 175-188.

PAGE 87

87 or as scale effects in which the absolute differences among genotypes cha nge with environment. Research studies aimed at quantifying the combined effects of silvicultural treatments and genetic improvement on unit area production in loblolly and slash pine are rare. Earlier studies indicate that GxE would not be of major conse quence for the majority of genotypes being deployed under traditional silvicultural systems (McKeand et al. 2006) For example, no GxE was found for total standing volume at age 12 in a loblolly pine genotype x vegetation control study in Georgia, USA (M artin and Shiver, 2002) and none was found at age four for five openpollinated loblolly pine families grown under two spacings in South Carolina, USA (McCrady and Jokela, 1996) Likewise, an analysis of whole tree biomass of five year old loblolly pine from two seed sources did not demonstrate GxE using a factorial genotype x fertilization experiment in North Carolina, USA (Retzlaff et al. 2001) Tree improvement programs have historically assessed GxE interactions for determining the need for site specific breeding efforts (McKeand et al. 1997b) Generally, in these investigations, a large number of genotypes are tested across a range of sites. Environmental variance in these breeding programs is due to localized climatic, edaphic and disease conditi ons, rather than to specific silvicultural treatments that manipulate site resources. Few studies have documented GxE interactions among silvicultural treatments, but available evidence suggests that when GxE did occur in these situations, it was caused b y a limited number of genotypes in the population that were highly sensitive to environmental variation (Zas et al. 2004) For example, (Duzan et al. 1988) found modest rank changes in family performance across a variety of sites in the southeastern USA while (Yeiser et al. 2001) showed instability in volume production at ages five and ten among loblolly pine families from some, but not all, seed zones

PAGE 88

88 in the Western Gulf region of the U nited States Similarly, a large GxE was documented for growth tr aits in loblolly pine families from Florida sources when moved northward one USDA (United States Department of Agriculture) Plant Hardiness Zone (Atwood et al. 2002; Sierra Lucero et al. 2002) ; yet, none was observed for other provenances (McKeand et al. 1990; Sierra Lucero et al. 2002) It appears that GxE may become significant only under extremes in seed source movement and/or site productivity and that relatively few genotypes from the population contribute to this response. The intensity of geneti c selection and silvicultural treatments is expected to increase in the future as resource managers move seed sources long distances in an effort to increase yields (Lambeth et al. 2005) Likewise, the probability of GxE becoming significant in the futur e is real and site/genotype specific silvicultural prescriptions may be needed to maximize volume and value production (Allen et al. 2005c) It follows that resource managers will benefit from an understanding of how elite genotypes respond across natura lly occurring and manmade environmental gradients ( i.e. fertilization, density management and associated vegetation control), as well as how soil physical, chemical and biological processes affect productivity (Fox, 2000). The overall objectives of this study were to investigate and quantify the magnitude and nature of GxE in full sib families of loblolly and slash pine. This was accomplished by using a series of replicated factorial experiments and family block plantings established in Florida and Georg ia that manipulated gradients in planting density, understory competition and soil nutrient availability.

PAGE 89

89 Methods Study D escription In January of 2000, the Forest Biology Research Cooperative ( http://www.sfrc.ufl.edu/fbrc/index.html Last accessed February 2009), located at the University of Florida, established a series of field research installations that were designed to examine the interactions of full sib loblolly and slash pine families with sev eral environmental factors, such as: location, silvicultural treatment intensity, and planting density (Roth et al. 2002) This trial series, referred to as PPINES (Pine Productivity INteractions on Experimental Sites), is the only one of its kind where the combined effects of species, genotype, silviculture and planting density can be examined singly or in combination across a range of site conditions in the southeastern U nited States Large family block plots, combined with contrasting treatments provi de a unique opportunity to examine GxE using stand level variables ( i.e. basal area, stem volu me, and aboveground biomass). Four study locations were included in this trial series representing two distinctly contrasting soil types ( Table 4 1). The topogr aphy is nearly flat, with less than a 1% slope. Soil series for the four sites were: Sanderson, FL Leon (sandy, siliceous, thermic Aeric Alaquods); Waverly, GA Bladen (mixed, semiactive, thermic Typic Albaquults); Perry, FL Leon (sandy, siliceous, t hermic Aeric Alaquods); Waldo, FL Newnan (sandy, siliceous, hyperthermic Ultic Haplohumods). Trials were installed on sites that held recently harvested southern pine plantations. Associated woody vegetation common to all sites included sawtooth palme tto [ Serenoa repens (B.) Small.], wax myrtle ( Myrica ceriferea L.), runner oak ( Quercus pumila Walt.), blueberries ( Vaccinium spp.), gallberry [ Ilex glabra (L.) Gray], and St. John's wort [ Hypericum fasciculatum (Lam.)]. Herbaceous plants in the understory commonly included bluestem grasses ( Andropogon spp.), panic grasses ( Panicum spp.), sedges ( Carex spp. and

PAGE 90

90 Cyperus spp.), and dogfennel [ Eupatorium capillifolium (Lam.) Small.]. All study locations shared a subtropical and humid climate with long hot wet summers and mild dry winters. Long term (1931 2000) precipitation has averaged 1384 mm yr1Experimental D esign (NOAA, 2002) The PPINES series was composed of two installations each of loblolly and slash pine. Within each installation, the experimental design wa s a 2 x 2 x 8 (silviculture x planting density x genetic entry) factorial which is planted in a randomized complete block, split plot design. Each site had four complete blocks which consist ed of four silviculture density whole plots. At the whole plot level, the two contrasting silvicultural treatments were operational versus intensive, while the two planting density treatments were 1334 treesha1 versus 2990 treesha1Treatment D escriptions Within each of these whole plot treatment combinations, there were eig ht subplots representing the genetic entries. Throughout this chapter genetic entries are alpha numerically coded using the prefix letter L for loblolly and S for slash pine. Each installation ha d 13 312 trees, on 128 plots, which were distributed on approximately 10 ha of experimental area. Prior to planting, each installation was double bedded on separate passes following a 2.75 m spacing pattern. In the late summer/early fall of 1999, all installations were treated with pre pla nt herbicides consisting of Chopper (imazapyr) at 1.02 ha1 and Garlon (triclopyr) at 7.02 ha1 with the goal of removing all woody competition and reducing initial levels of herbaceous vegetation. The objective was to provide a site with resources suitable for optimum growth while minimizing the variation within individual study sites. The operational silviculture treatment represented a typical regime utilized by forest industry throughout the southeastern USA at the time. After receiving a common site preparation treatment, the operationally treated

PAGE 91

91 plots received a single banded, or broadcast, application of 280 kgha1The contrasting intensive treatment was mainly driven by early vegetation control and annual fertilization. On these plots, competing ve getation was controlled for two years following planting using directed applications of Arsenal diammonium phosphate at the time of planting. (imazapyr) at 0.28 ha1 (limited to loblolly pine installations) and Oust (sulfometuron methyl) at 0.14 ha1 on all installations. For the follow up trea tments ground cover was kept below a 30% threshold through age three. By age five, the tree crowns had closed canopy and the herbaceous component was limited due to light availability. The intens ive plots were fertilized with 5 60 kgha1The second treatment factor at the wholeplot level wa s contrasting planting density: 1334 treesha of 101010 plus micronutrients at the time of planting, which was followed by annual applications of macroand micronutrient fertilizers using prescriptions based on foliar analyses. The total amounts of nutrients applied on each installation through age five are presented in Table 4 2. 1 planted at a spacing of 2.75 m x 2.75 m, and 2990 treesha1 planted at a spacing of 1.22 m x 2.75 m. The 2990 treesha1 subplots of each genetic entry were arranged in eight beds of 16 planting positions each, for a total of 128 trees per gross treatment plot. A two tree border around the perimeter result ed in a 48 tree interior measurement plot of 0.016 ha. The 1334 treesha1At the subplot level, genetic entries consisted of first generation elite fullsib families. On loblolly sites there were seven entries of fu llsib families, which included a previously identified subplots of each genetic entry were arranged in eight beds of 10 planting positions each, for a total of 80 trees per gross plot. A single tree buffer around the perimeter result ed in a 48 tree interior measurement plot of 0.036 ha. Despite an ongoing droug ht at the time of establishment, survival was over 95% in all treatments at the end of the first growing season.

PAGE 92

92 poor grower, and an intimate mixture of the top six full sib families. The entries were similar for the slash pine sites, with the exception of one less full sib family to make room for the loblolly mixture in addition to a slash mixture. This allow ed for a direct comparison of species performance across spacing and silvicultural treatments on these two slash locations. All genetic entries in the study were selected from sources which exhibit ed moder ate to excellent resistance to fusiform rust [ Cronartium quercum (Berk.) Miyabe ex Shirai f. sp. f usiforme ] based upon a priori knowledge from breeding programs. This was done in order to reduce the confounding effects of disease incidence. Seedlings wer e grown in 66 mlcell1 Ray Leach 'Cone tainer'TMInsecticides were uniformly applied across all treatments on loblolly pine installations in an effort to control damage from N antucket pine tip moth ( Rhyacionia frustrana [ Comstock ]). Treatments were applied on a monthly basis over the first two growing seasons, beginning in March and ending in September. Alternating applications of the following chemicals and application rates were applied aerially or by hand: Pounce cells (Stuewe and Sons, Inc Corvallis, OR). Each site was hand planted over a two day period in January 2000. 3.2EC (62 ml product 1 water), Warrior T (39 ml product 1 water), Dimilin 25W (62 ml product 1 water), and Mimic 2LV (125 ml product 1A fungicide was uniformly applied across all treatments on slash pine installations in order to control for the confounding effects of potential fusiform rust infection. Treatments were applied by ground application on a monthly basis over a four month period (beginning in March) during the first two growing seasons. Treatments a rates consisted of 12 g of 50% concentrated D F Bayleton water). with 65ml of Agri Dex in 18.9 of water.

PAGE 93

93 Inventory, Yield and Biomass E stimates Annual measurements of DBH were made at ages two three, and five years on all trees in the measurement plots. Total height (HT) was measured on every tree at ages two and three, but was limited to a re presentative 20% subsample at age five. Individual tree HT at age five was determined from site and treatment specific HT v ersus DBH relationships developed from this subsample. Abiotic and biotic tree damage was assessed at the time of measurement. B asal area (BA) was calculated on a per family plot basis (m2ha1), which accounts for variation due to mortality. Since the trees were relatively small at these ages, a simple index of individual tree stem volume was utilized : the sum of a cylinder from the tree base to breast height (BH = 1.37 m) and of a cone from BH to the top if the tree. Individual surviving trees per plot were summed to yield total standing stem volume per plot (VOL) and wa s expressed in m3 ha1Aboveground biomass equations were developed using a treatment specific dataset from this experiment along with supplemental data from previous regional studies of similar ag e and treatment history (Table 4 3). Biomass harvests on PPINES were conducted at age two and five and covered the f ull range of locations and silvicultural treatments. Due to resource limitations, allometric equations were developed that were common to the fullsib families represented in the study. This was done in order to increase the power to detect differences b etween species, locations, densities and silvicultural intensities. Two families were harvested within each species at age two. At age five, due to further resource limitations, the harvest was restricted to two locations of loblolly pine and further limited to a subset of silvicultural treatments and families. Trees selected for each harvest originated from border rows. These border rows were buffering the effects of genotype only since there were additional buffers separating the density and silvicultural intensity treatments. The age two harvest consisted of 47 loblolly pine (families L2 and L4) and 60 slash pine (S1 and S6) trees, which were harvested across each of the four

PAGE 94

94 culture/density whole plots on two sites from each species. The age five harvest consisted of 54 loblolly pine trees from the 2290 treesha1Within each harvest year, sample trees free of damage and disease, were selected at random from across the diameter distribution representative of each treatment and site. Prior to harvest, an inventory was completed on each sample tree which consist ed of HT, DBH, diameter at ground line, and crown width at the widest point parallel to, and perpendicular to the planting bed. Sample trees were felled at ground line using a hand saw, placed on a tarp and separated into four aboveground components: foliage, branches, stem and dead branches. The total fresh we ight of each component was measured separately in the field. The fraction of bark to stem components were estimated from six cm disks of wood, cut from the base of each of three equally spaced stem segments along the full length of the stem. Bark was se parated from each disk and the fresh weight of each was determined in the field. Tissue samples were transported from the field and dried to a constant weight at 70 planting density over both contrasting silvicultural intensities at Sanderson, FL and the intensive cultural treatment at Waverly, GA. 0Logarithmically transformed linear allometric equations were developed using a combina tion of the biomass harvests and the regional data set according to the base model ( Equation 41) (Crow, 1988) C. ln(Yi0)i 0 1ln Xi 1 iW here ln is the natural logarithm, Y ) ( 41) i is the dry weight of the unit area aboveground biomass (AGB) of the i th sample tree expressed in kgtree10)i 0 is the mean of the i th sample tree within each species, Xi 1 is the product of the combined variables of DBH squared times HT for the i th sample tree expressed in dm3i is the random error associated with estimating the weight of the aboveground biomass of the i th sample tree. The need for separate groups of

PAGE 95

95 equations by species, location, silviculture, and density was examined utilizing PROC MIXED (Littel et al. 1996) in SAS. These were evaluated by beginning with a pooled dataset and systematically decomposing the general model by entering treatment va riables and their interactions. At each step slopes and intercepts of the resulting equations were evaluated through covariance analysis. The large sample size in the pooled regional dataset (n = 432 harvest trees) yielded tests with many degrees of freedom, thereby increasing the power to detect differences in parameters between treatments. Variables were included in the development of the model if they treat ments in the PPINES trial resulted in individual trees of much larger size than those from the regional dataset. Therefore, results could not be extrapolated for those individual treatment combinations. As a result, only the variable of species was included in the allometric relationships (Table 44 ). Probability plots of the residuals indicated that the normality assumption was satisfied and plots of residuals versus predicted values showed no obvious pattern, suggesting that the assumptions of independence and equal variance were met. Corrections for bias in the transformation of logarithmic units to arithmetic units were applied (Equation 42) (Baskerville, 1972) : (4 2) where is the estimated aboveground biomass in arithmetic units of th e skewed Y distribution at X (Equation 41) AGB was calculated for each plot and is expressed in Mgha1 Analysis of dry matter. All analyses were performed using PROC MIXED (Littel et al. 1996) in SAS. To test for differences in stand level attributes among treatments, separate analyses of variance (ANOVA) were performed for loblolly and slash pine using a mixed linear model for data pooled across two sites within each species ( Equation 43): ) 2 / (2 e Y

PAGE 96

96 Yijklmn i + b(s)ij + Ck + Dl + CDkl + Fm + CFkm + DFlm + CDFklm + SCik + SDil + CDikl + SFim + SCFikm + SDFilm + SCDFiklm + b(s)Cijk + b(s)Dijl + b(s)CDijkl + b(s)Fijm + b(s)CFijkm + b(s)DFijlm + b(s)CDFijklm + b(s)Sij + b(s)SCijk + b(s)SDijl + b(s)CDijkl + b(s)SFijm + b(s)SCFijkm + b(s)SDFijlm + wijklmn (4 3) where Yijklmn is the response variable (BA, VOL, or AGB) of the nth plot of the m th family of the l th planting density of the k th silvicultural intensity of the j th block of the i th site ( i = 1,2; j = 1,2, mean; Si is the fixed effect of the i th location; b(s)ij is the random interaction effect of the j th block within the i th location; Ck is the fixed effect of the k th silvicultural intensity; Dl is the fixed effect of the l th planting density; Fm is the fixed effect of the m th family and wijklmn is the random error. Blocks were nested within locations, while the factors of silviculture (C), planting densi ty (D), and genotype (F) were crossed. All terms containing b(s)ij were considered to be random effects in the model and were pooled as appropriate for each variable tested using the procedure described by (Bancroft and Han, 1983) The only exception was b(s)CDijkl, which was never pooled as it is used as the error term to test the main effects of Si, Ck and DlSince the analysis of each variable has a differing model constr uct, the variance components for each model are presented in a separate table subsequent to the traditional ANOVA tables in the results section. Assumptions of equal variance between the two planting density treatments were violated for all variables exam ined. This was due to heterogeneity in the covariance structure associated with planting density; there was greater variation within the 2290 treesha Individual variance components were pooled when the probability of a greater F statistic was 0.25 or larger. As noted by (Bancroft and Han, 1983) the significance level for the F test is much higher than conventional levels of 0.01 or 0.05 and is a conservative measure of the relative efficiency of pooling the sources of variation. 1 treatment. To account for this heterogeneous variance, the residual was grouped by the

PAGE 97

97 fixed effect o f density (Bozivich et al. 1956) Where significant effects were found, least squares means were generated between levels of the factors of interest. Where multiple non planned comparisons were made, a Bonferroni's adjusted significance level was used. Single degree of freedom contrasts were performed to test for differences between species (mixed loblolly vs. mixed slash pine plots) and also method of deployment (mixed vs. pure plots) Results Strong and significant GxE in BA, VOL, and AGB was apparent in this experiment for both species. The strength of the experimental design enabled the detection of three types of unit area production interactions: genotype x site, genotype x silviculture, and silviculture x density (Tables 45, 4 6 and 47 ). There were no significant three way interactions involving genotype, site and silviculture. Some combinations of treatments interacted as early as age two and all increased in significance with time. Despite the high statistical power to detect interactions, there was no evidence for genotype x density interactions of any kind, despite the extremes in planting density combined with the contrasting silvicultural treatments and locations. Genotype x Site I nteractions At age two there were strong and significant interactions between sites and loblolly pine families for BA, VOL and AGB (p=0.0474, p=0.0390, p=0.0440, respectively); by age five the significance of these interactions had increased (p=0.0271, p=0.0224, 0.0388) ( Table 4 5). For slash pine, GxE between sites was not evident at age two but became significant for BA, VOL and AGB by age three (p=0.0039, p=0.0046, p=0.0158, respectively) and gained in significance over those for loblolly by age five (p=0.0127, p=0.0157, p=0.0158) ( Table 4 6). The varying pe rformance of families across sites was largely due to scale effects, with certain families performing better or worse than their peers when grown together on contrasting sites. For example, at age five, the difference between sites in AGB for loblolly pin e family L5 was 13%

PAGE 98

98 (versus a 19% average for all other families) (Figure 41a). In terms of absolute production, family L4 was the top performer across both sites. Similar effects were observed for slash pine families between sites. Generally, these can be split into three groups based on their pe rformance. The first group (S4 and S6) was the most sensitive across locations, and despite varying yields, both families had similar slopes representing the degree of performance across locations (Figure 4 1b). The second group had intermediate sensitivities despite a wide range of yields in AGB (S1, S2, and S3). Family S5 had similar levels of AGB at both locations which resulted in a rank change (Figure 4 1b). Genotype x Silviculture I nteractions G enotype x silviculture interactions were not as strong as the interactions of genotype x location Significant GxE became apparent by age three among the loblolly pine families for VOL (p=0.0421) ( Table 4 5) and grew stronger with time (p=0.0019 at age five). The significance of the interaction for loblolly pine in BA (p=0.0541) and AGB (p=0.0502) at age five was not as strong as was for VOL In contrast, elite families of slash pine were not as responsive to silviculture as was loblolly pine. Similarly, the pe rformance among slash pine families was more stable when grown under contrasting silvicultural regimes. In slash pine, GxE (as driven by silvicultur al treatment inten sity ) was not significant until age five and then only for VOL (p=0.0126); BA was weakly significant at p=0.0589 ( Table 3 6). As with genotype x location interactions, the instability of family performance across contrasting silvicultural treatments was mainly the result of scale effects, where certain families either outperformed or underperf ormed their peers with increasing intensity of silvicultural treatment. Examination of least squares means for VOL at age five showed that loblolly family L4 was most responsive to increasing silvicultural intensity (75% increase), while family L5 was one of the least responsive families (55% increase) (Figure 4 2a). Family L5 was also the family

PAGE 99

99 that exhibited the least difference in volume growth across contrasting locations (13% difference). All other families were intermediate in their response. For slash pine, families S2 and S6 were the most responsive in VOL at age five to increasing intensity of silvicultural treatment (63% increase), with all other families exhibiting a lower response (combined 55% increase) (Figure 4 2b). Silviculture x Density I nteractions Interactive effects of silvicultural treatment intensity and planting density for loblolly pine were highly significant (p<0.0001) for all growth metrics at age two, and continued through age five. Similar effects were noted for slash pine, but they did not become significant until age three (p<0.005). In all cases, the interactions were due to larger responses to increasing silvicultural treatment intensities under conditions of increasing planting density. For example, on the slash pine s ites, the intensive silvicultural treatment increased AGB by 5.7 Mgha1 at 1334 treesha1, versus 12 Mgha1 at 2990 treesha1 (Figure 4 3c ). However, there was one case where this two way interaction at age five for loblolly AGB was dependent on locat ion (three way interaction, p = 0.0007). In this case (Sanderson, FL), the 2990 treesha1Location x Density I nteractions operational treatment produced a much lower than expected response in AGB than that at the Waverly GA location (Figures 43a and 43b). All other combinations of silvicultur al treatment intensity and planting density between locations had similar responses for AGB at age five. There were significant location x density interactions for all variables at ages three and five for slash pi ne (p< 0.05) ( Table 4 6) but not for loblolly pine ( Table 4 5). In general, mortality was greatest in slash pine, with the majority occurring between ages three and five. Despite similar survival between the two slash pine locations ( Table 4 8), the Perr y, FL location had

PAGE 100

100 greater VOL at age five than the Waldo, FL location at the 2990 treesha1Species and Deployment I nteractions planting density (p<0.0197). There were strong and significant species x location interactions (p<0.0001) for all variables. Loblolly pine was more responsive in aboveground biomass than slash pine on the two locations where a direct comparison was possible (Figure 44). Top performing full sib loblolly pine families expressed similar yields at age five whether grown in intima te mixtures or pure blocks. However, slash pine tended to have greater BA, VOL, and AGB on a unit area basis when grown in mixtures, as compared to pure plots of the same full sib families ( Table 4 9). Effe cts of Disease and H urricanes Plot level incidenc e of fusiform rust and wind damage at age five was examined in an attempt to partially explain genotype x location interactions. Despite the fact that all families in the study were selected to have some level of fusiform rust resistance, based on a prior i knowledge, there were significant rank changes among slash pine families in fusiform rust occurrence between locations at age five (p=0.0189). Similar results have been previously documented in slash pine (Schmidt and Allen, 1998) Of the six slash pine families in the experiment, three (S4, S5, and S6) demonstrated GxE in fusiform rust incidence, with the Waldo, FL location having the highest incidence levels (Figure 45a ). The other three families had a similar, but low overall incidence of fusiform rust between locations. Loblolly pine families generally had low incidence of fusiform rust and no significant interactions were found. In the summer of 2004 two hurricanes, Frances and Jeanne, passed in close proximity to the Waldo, FL location. While damage was not extensive, there were a substantial proportion of trees toppled or leaning at varying degrees throughout the study area. Damage from these storms

PAGE 101

101 was minimal at the Perry, FL location and barely evident at either of the two loblolly pine loc ations. There was significant GxE for wind damage in slash pine between locations (p<0.0001) (Figure 4 5b). Trees on the slash pine locations may have toppled due to indirect effects of weak root systems in combination with relatively large canopies. Di seased trees may have broken due to fusiform rust galls located on tree stems. Discussion This experiment provided the opportunity to quantify the combined effects of silvicultural treatments and genetic improvement on unit area production in selected full sib loblolly and slash pine families. The GxE observed in this study occurred at the twoway level: genotype x location and genotype x silviculture. While the genotype x density interactions were not significant, as reported for another loblolly pine ex periment (McCrady and Jokela, 1996), there were significant silviculture x density interactions for unit area production, which occurred independent of mortality. The variety of interactions evident in this study was not surprising given the range of cont rasting elite genotypes, silvicultural treatments and study locations established. When combined with the high statistical power associated with the complex experimental design, we had the ability to detect significant differences in the responses of thes e elite genotypes to various environmental conditions in plantations of loblolly and slash pine in the southeastern U nited S tates Genotype x S ilviculture G enotype x environment issues in southern forestry will not be of importance unless silviculture or p ropagule type changes significantly from those currently in use (McKeand et al. 2006) Therefore, it was somewhat surprising that the genotype x location interactions were more significant and consistent than the genotype x silviculture interactions. This is even more striking given the extremes in the silvicultural treatment intensities employed in this study.

PAGE 102

102 However, the magnitude increase in productivity with increasing silviculture likely overpowered the statistical significance of this interaction as certain families tended to show a greater response than others. One example was loblolly family L4 which is widely deployed operationally across the southeastern U nited S tates Its plasticity with regard to intensive management demonstrates responsi veness considerably greater than its peers. While not of the same magnitude, the same is true for select families of slash pine in this experiment (S2 and S6). This effect of similar relative differences in yield, yet larger absolute differences with inc reasing silvicultural intensity has been previously demonstrated in loblolly pine (McKeand et al. 1997a) It follows that this variation in GxE across locations and silvicultural treatments could potentially be exploited if the relatively few responding genotypes were to be identified and deployed on the proper sites in combination with appropriate site specific silvicultural treatments. Genotype x L ocation The strongly significant genotype x location interaction, even after accounting for the extremes in silvicultural treatments, is an indication that variation in soils, climate, edaphic variables, and pests (even across relatively short distances) are important regardless of the level of silvicultural intensity. As other researchers have suggested, soil conditions that regulate the ability to supply moisture and nutrients (Fox, 2000) may be partly responsible for the GxE observed in this experiment (see Appendix B and C) Growth response to nutrition has been shown to vary by family, especially for l oblolly pine (Li et al. 1991b; Samuelson, 2000) There is also evidence that carbon allocation to aboveground and belowground tissues is sensitive to soil fertility and varies with provenance and family (Crawford et al. 1991; Wu et al. 2004) For exam ple, in an loblolly pine fertilization experiment, families varied in the production of fine root s under low nitrogen (N) treatments, but not under high N levels (Samuelson, 2000)

PAGE 103

103 Examination of foliar nutrition at age five on the current experiment did not explain the GxE observed in production at age five (unpublished data). Genetic variability within a population allows for the potential to buffer against the effects of disease and weather, and is an important aspect of family stability. This becomes critical in areas where there are extremes in localized climatic conditions and/or pathogen populations. In the current study, through examination of damage codes made at the time of inventory, we were able to partially explain the GxE across locations fo r slash pine, but not for loblolly pine. In slash pine, the occurrence of fusiform rust and hurricane damage influenced the genotype x location interaction. Two of the three families responsible for the age five GxE in fusiform rust occurrence (S4 and S6) (Figure 45a ) corresponded to the GxE between locations in AGB (Figure 4 1b). It was somewhat surprising that fusiform rust incidence did not explain the genotype x location interactions in loblolly pine given that the performance of resistant families of this species are the most unpredictable across sites (McKeand et al. 2003b) Since all test locations were located within USDA Plant Hardiness Zone 8b, adaptation problems across sites should not be expected in this experiment (Schmidtling, 2001; Lambe th et al., 2005) One anomaly is the single family (S5), which had a greater incidence of fusiform rust occurrence at Waldo, FL (Figure 4 5a ), yet similar biomass production when compared across locations (Figure 4 1b). The explanation for this anomaly m ay lie with its relative stability to the severe winds of 2004 (Figure 45b). In contrast, family S6 had the highest incidence of weather damage at the Waldo, FL location (42.4%), in combination with a fairly high occurrence of fusiform rust (30.1%). Whi le there were large scale effects from wind damage, there were no changes in rank among the slash pine families (Figure 45b). Occurrence of pitch canker, insect damage, and forking was examined, but did not explain t he GxE observed in this study.

PAGE 104

104 The sig nificant genotype x location interactions as demonstrated in this study, with limited genotypes and locations, serves to emphasize the importance of carefully considering deployment and management of elite genotypes in the future. In some cases, existing expert local knowledge of site conditions, including those not foreseen such as catastrophic insect, disease or climatological variation, may provide critically important information needed to make successful deployment decisions. For locations with extre me site conditions or unknown climate variability, it may be desirable to emphasize pest resistance over growth when selecting genotypes to deploy, which could minimize the risk of unexpected growth performance. Silviculture x D ensity Interactions betwee n silvicultural treatments and stand density are well known to occur and have been described using several conceptual models that link silviculture with ecology (Long et al., 2004) We noted significant silviculture x density interactions, with the greate st response in production occurring under conditions of intensive silviculture and high initial planting density (Figure 4 3). This interactive effect is due to better and earlier site resource capture at higher planting densities. Treatments planted at 2 990 treesha1 closed canopy a minimum of two years earlier than the lower densities. The low density plots were not able to take full advantage of the extra resources made available through the intensive silvicultural treatment, which was applied to both planting densities. The location x silviculture x density interaction noted for loblolly AGB is likely a function of differences in the inherent productivity of the two contrasting locations examined in combination with the relatively high nutrient deman ds of loblolly pine (Jokela et al. 2000) Inherent productivity differences between locations are demonstrated using a surrogate of average tree height at age five (averaged across families and densities) on the operational treatments (6.64 m at Sanderson, FL versus 7.65 m at Waverly, GA). The nutrient poor, sandy soil at Sanderson, FL is clearly unable to supply the nutrients demanded for

PAGE 105

105 maximum growth in the absence of nutrient additions and has been documented in other studies (Adegbidi et al. 2005) Nutrient limitations are exacerbated when tree density, and resulting unit area AGB, is dramatically increased to levels approaching 2990 treesha1Species and Deployment I nteractions (Burkes et al. 2003) Resource managers will need to be aware that plantations in the southeastern U nited S tates growing on nutrient poor sites at higher densities will be in critical need of nutrient amendments much earlier in their rotations than previously thought. As seen in other experiments, where limiting site resources were ameliorated through co mbinations of competing vegetation control and nutrient applications, loblolly pine productivity was close to its predicted biological maximum, regardless of the inherent site quality (Jokela et al. 2004; Sayer et al. 2004). It is curious that loblolly performance was similar regardless of deployment in mixtures or pure plots, while slash pine performed better when grown in mixtures. It has been documented in other ecosystems that contrasting species can exploit diffe rent resource strata and therefore have greater yields when grown together on the same site. Perhaps the families chosen for the slash pine installations are truly an example of this. A more likely explanation is that there may be differential pest or en vironmental stress between the mixed and pure plots. Conclusions The significant genotype x location interactions that were found in this study, despite limited genotypes and locations, serves to emphasize the importance of carefully considering deploymen t strategies of improved genotypes of loblolly and slash pine in the southeastern United States. As resource managers make decisions about where to deploy this elite genetic material, they also will need to know how these genotypes will respond to intensi ve silvicultural treatments in association with localized pest and climatic conditions. For example, as

PAGE 106

106 silvicultural treatments become more effective at ameliorating limiting site resources, the efficiency of nutrient uptake and utilization among genotyp es will likely play a larger role in their differentiation of performance (Li et al. 1991a) Variation in crown structure could also lead to significant GxE (McCrady and Jokela, 1996) This issue is certain to increase in importance as advances in clonal forestry occur (McKeand et al. 2003a; Bouvet et al. 2005) In certain cases where intensive silviculture and advanced breeding strategies are combined, it may become necessary to develop site specific silvicultural treatments for particular genotypes o r to modify breeding strategies in order to capture the full advantage of the GxE interaction.

PAGE 107

107 Table 41. Characteristics of the PPINES experimental locations. PPINES Pine Productivity INteractions on Experimental Sites. All locations w ere planted in January of 2000. Table 42. Cumulative elemental nutrient application rates for the PPINES intensive silvicultural treatments through five growing seasons (kg ha1 Site location ). N P K Mg Ca S B Zn Mn Fe Cu Sanderson, FL 369 128 121 45 45 35 0.9 2.7 2.2 14.7 3.9 Waverly, GA 369 128 121 45 45 35 0.9 2.7 2.2 14.7 3.9 Perry, FL 373 112 115 56 45 139 1.1 3.0 3.0 15.5 5.2 Waldo, FL 370 124 124 63 56 33 1.7 2.5 6.1 6.1 4.4 Operational silviculture treatments all received 45 kg ha1 N and 50 kg ha1 Site location of P in the form of diammonium p hosphate at the time of planting only. Species Latitude Longitude Soil order Elevation (m) Sanderson, FL Lobl olly 30 .28 0 82.33 0 Spodosol 45 Waverly, GA Loblolly 31.13 0 81.75 0 Ultisol 10 Perry, FL Slash 30.17 0 83.73 0 Spodosol 15 Waldo, FL Slash 29.80 0 82.21 0 Spodosol 50

PAGE 108

108 Table 43. Description of biomass harvest data used to develop the allometric equations displayed in Table 44. Silvicultural intensity Planting density Age Year Sample ( n ) Reference Loblolly Intensive 495 2 2003 24 FBRC 2004 Intensive 540 2 2001 17 This chapter Intensive 608 4 1999 8 Adegbidi et al 2002 Intensive 625 4 1986 27 Colbert et al. 1990 Intensive 1200 2 2001 18 This cha pter Intensive 1200 5 2004 72 This chapter Operational 495 2 2003 24 FBRC 2004 Operational 540 2 2001 16 This chapter Operational 608 3 2000 24 Adegbidi et al 2002 Operational 608 4 1999 24 Adegbidi et al 2002 Operational 625 4 1986 6 Colbert et al. 1990 Operational 1200 2 2001 18 This chapter Operational 1200 5 2004 36 This chapter Slash Intensive 495 2 2003 12 FBRC 2004 Intensive 540 2 2001 14 This chapter Intensive 625 4 1986 27 Colbert et al. 1990 Intensive 1200 2 2001 14 This c hapter Operational 495 2 2003 12 FBRC 2004 Operational 540 2 2001 16 This chapter Operational 625 4 1986 7 Colbert et al. 1990 Operational 1200 2 2001 16 This chapter Silvicultural intensity is a generalized grouping of cultural treatments found in the individual studies that closely approximates that found in the current PPINES investigation: The operational treatment represents silviculture best practices in the southeast at the end of the 20th century, receiving a common site preparation treatment and single banded application of 280 kgha1 Diammonium Phosphate at the time of planting. The contrasting intensive treatment is driven mainly by early complete vegetation control and annual fertilization. Planting density is expressed as the number of treesha1. Age is the age in years of the trees at time of the biomass harvest. Year is year of the biomass harvest.

PAGE 109

109 Table 44. Parameter estimates and standard errors of the estimate aboveground biomass (kgtree1 ) equations for loblolly and s lash pine. Data used to develop equations were generated from regional trials in FL and GA, two to five years in age. 0 1 Species Estimate SE p value Estimate SE p value Loblolly 0.63065 0.02679 <0.0001 0.53480 0.00774 <0.0001 Slash 0.56723 0.02678 <0.0001 0.53480 0.00774 <0.0001 01ln X where ln= natural logarithm, Y = abov eground total dry weight (kgtree10 1 = regression coefficients (intercept and slope, respectively), X = DBH2HT in dm3. Overall model R2 = 0.929, RMSE = 0.26574, n = 432.

PAGE 110

110 Table 45. Summary of statistical significance (prob. >F) and associated degrees of freedom from ANOVA to test loblolly pine basal area, stem volume and aboveground biomass at ag e two threeand five years Source of Basal area Stem volume Aboveground biomass variation Num. df Den. df p value Num. df Den. df p value Num. df Den. df p value Age 2 Silviculture (C) 1 76 <0.0001 1 76 <0.0001 1 76 <0.0001 Density (D) 1 6 <0.0001 1 6 <0.0001 1 6 <0.0001 C x D 1 72 <0.0001 1 72 <0.0001 1 72 <0.0001 Family (F) 6 76 <0.0001 6 76 <0.0001 6 76 <0.0001 C x F 6 76 0.3689 6 76 0.2596 6 76 0.4582 D x F 6 72 0.1361 6 72 0.1015 6 72 0.1785 C x D x F 6 72 0.4031 6 72 0.3705 6 72 0.3433 Location (S) 1 6 0.0263 1 6 0.0219 1 6 0.0276 S x C 1 76 <0.0001 1 76 <0.0001 1 76 0.0002 S x D 1 6 0.1887 1 6 0.1672 1 6 0.1537 S x C x D 1 72 0.0097 1 72 0.0094 1 72 0.0055 S x F 6 76 0.0474 6 76 0.0390 6 76 0.0440 S x C x F 6 76 0.8238 6 76 0.7674 6 76 0.8742 S x D x F 5 72 0.7599 5 72 0.7067 5 72 0.5411 S x C x D x F 5 72 0.2597 5 72 0.2553 5 72 0.2679 Age 3 Silviculture (C) 1 6 <0.0001 1 6 <0.0001 1 6 <0.0001 Density (D) 1 6 <0.0001 1 6 <0.0001 1 6 <0.0001 C x D 1 72 <0.0001 1 6 0.0024 1 72 <0.0001 Family (F) 6 70 <0.0001 6 70 <0.0001 6 70 <0.0001 C x F 6 70 0.0701 6 70 0.0421 6 70 0.1081 D x F 6 72 0.5611 6 66 0.6152 6 72 0.8293 C x D x F 6 72 0.4365 6 66 0.3392 6 72 0.3136 Location (S) 1 6 0.0023 1 6 0.0042 1 6 0.0061 S x C 1 6 0.1132 1 6 0.2190 1 6 0.0290 S x D 1 6 0.1526 1 6 0.1837 1 6 0.1664 S x C x D 1 72 0.0715 1 6 0.2083 1 72 0.0178 S x F 6 70 0.0550 6 70 0.1195 6 70 0.1899 S x C x F 6 70 0.7820 6 70 0.7901 6 70 0.9084 S x D x F 5 72 0.8857 5 66 0.9280 5 72 0.6851 S x C x D x F 5 72 0.4151 5 66 0.3647 5 72 0.4449 Age 5 Silviculture (C) 1 6 <0.0001 1 6 <0.0001 1 6 <0.0001 Density (D) 1 6 <0.0001 1 6 <0.0001 1 6 <0.0001 C x D 1 6 0.0014 1 6 0.0011 1 142 <0.0001 Family (F) 6 136 <0.0001 6 136 <0.0001 6 142 <0.0001 C x F 6 136 0.0541 6 136 0.0019 6 142 0.0502 D x F 6 136 0.1022 6 136 0.1149 6 142 0.4576 C x D x F 6 136 0.8249 6 136 0.6683 6 142 0.5154 Location (S) 1 6 0.0021 1 6 0.0028 1 6 0.0032 S x C 1 6 0.0056 1 6 0.0038 1 6 0.0005 S x D 1 6 0.1092 1 6 0.1314 1 6 0.07 08 S x C x D 1 6 0.4445 1 6 0.2368 1 142 0.0007 S x F 6 136 0.0271 6 136 0.0224 6 142 0.0388 S x C x F 6 136 0.3847 6 136 0.2075 6 142 0.5364 S x D x F 5 136 0.4779 5 136 0.5922 5 142 0.4878 S x C x D x F 5 136 0.6594 5 136 0.5897 5 142 0.43 61

PAGE 111

111 Table 4 5 Continued. Different models were constructed for each variable within each age with varying random affects in the variance terms; hence the need for different numerator and denominator degree s of freedom in the mixed model Basal area is expressed in m2ha1. Stem volume is expressed in m3ha1 and is calculated as the sum of per tree measurements of the volume of a cylinder to 1.37 m and the volume of a cone from 1.37 m to the top if the tree. Aboveground biomass is expressed in met ric tons per hectare and was calculated using indiv idual tree allometric equations P values significant at the 95% level of confidence are shown in bold type.

PAGE 112

112 Table 46. Summary of statistical significance (prob. >F) and associated degrees of freedom from ANOVA to test slash pine basal area, stem volume and aboveground biomass at age two threeand five years. Source of Basal area Stem volume Aboveground biomass variation Num. df Den. df p value Num. df Den. df p value Num. df Den. df p val ue Age 2 Silviculture (C) 1 18 0.0002 1 18 0.0004 1 18 0.0018 Density (D) 1 18 <0.0001 1 18 <0.0001 1 18 <0.0001 C x D 1 18 0.0659 1 18 0.0760 1 18 0.1300 Family (F) 5 119 <0.0001 5 119 <0.0001 5 119 <0.0001 C x F 5 119 0.4326 5 1 19 0.4252 5 119 0.6137 D x F 5 119 0.1267 5 119 0.1027 5 119 0.3591 C x D x F 5 119 0.8362 5 119 0.8065 5 119 0.8731 Location (S) 1 6 0.9236 1 6 0.9826 1 6 0.7859 S x C 1 18 0.7668 1 18 0.8279 1 18 0.8956 S x D 1 18 0.2432 1 18 0.2885 1 18 0.4431 S x C x D 1 18 0.7714 1 18 0.8336 1 18 0.8855 S x F 5 119 0.1953 5 119 0.1890 5 119 0.1649 S x C x F 5 119 0.9212 5 119 0.9251 5 119 0.8537 S x D x F 5 119 0.7424 5 119 0.7318 5 119 0.6487 S x C x D x F 5 119 0.9946 5 119 0.9951 5 119 0.9982 Age 3 Silviculture (C) 1 18 <0.0001 1 18 <0.0001 1 18 <0.0001 Density (D) 1 18 <0.0001 1 18 <0.0001 1 18 <0.0001 C x D 1 18 0.0003 1 18 0.0007 1 18 0.0037 Family (F) 5 119 <0.0001 5 119 <0.0001 5 119 <0.0001 C x F 5 119 0.1182 5 119 0.0797 5 119 0.4432 D x F 5 119 0.0641 5 119 0.0559 5 119 0.1259 C x D x F 5 119 0.6940 5 119 0.6627 5 119 0.7740 Location (S) 1 6 0.0257 1 6 0.0410 1 6 0.0937 S x C 1 18 0.0121 1 18 0.0240 1 18 0.1037 S x D 1 18 0.011 8 1 18 0.0159 1 18 0.0369 S x C x D 1 18 0.1990 1 18 0.2184 1 18 0.3651 S x F 5 119 0.0039 5 119 0.0046 5 119 0.0158 S x C x F 5 119 0.0549 5 119 0.0608 5 119 0.2363 S x D x F 5 119 0.4222 5 119 0.4083 5 119 0.4500 S x C x D x F 5 119 0.9283 5 119 0.9148 5 119 0.9953 Age 5 Silviculture (C) 1 6 <0.0001 1 6 <0.0001 1 6 <0.0001 Density (D) 1 12 <0.0001 1 12 <0.0001 1 12 <0.0001 C x D 1 12 0.0007 1 12 0.0002 1 12 0.0037 Family (F) 5 116 <0.0001 5 116 <0.0001 5 116 <0.0001 C x F 5 116 0.0589 5 116 0.0126 5 116 0.4432 D x F 5 116 0.2837 5 116 0.1763 5 116 0.1259 C x D x F 5 116 0.4665 5 116 0.5684 5 116 0.7740 Location (S) 1 6 0.0024 1 6 0.0037 1 6 0.0937 S x C 1 6 0.1441 1 6 0.1880 1 6 0.1037 S x D 1 12 0.0439 1 12 0.0197 1 12 0.0369 S x C x D 1 12 0.2945 1 12 0.2869 1 12 0.3651 S x F 5 116 0.0127 5 116 0.0157 5 116 0.0158 S x C x F 5 116 0.0510 5 116 0.0790 5 116 0.2363 S x D x F 5 116 0.7333 5 116 0.5427 5 116 0.4500 S x C x D x F 5 116 0.9229 5 116 0.8777 5 116 0.9953

PAGE 113

113 Table 4 6. Continued. Different models were constructed for each variable within each age with varying random affects in the variance terms; hence the need for different numerator and denominator degrees of freedom in the mixed model Basal area is expressed in m2ha1. Stem volume is expressed in m3ha1 and is calculated as the sum of per tree measurements of the volume of a cylinder to 1.37 m and the volume of a cone from 1.37 m to the top if the tree. Aboveground biomass is expressed in metric tons per hectare and was calculated using individual tree all ometric equations P values significant at the 95% level of confidence are shown in bold type.

PAGE 114

114 Table 47. Variance components and associated stati stical significance (prob. > |Z|) for individual model results in tables 45 and 46. Random variance components and their interactions were pooled when the P value was greater than 0.25 and are either left blank or not shown in this table. Source of Sla sh Pine Loblolly Pine Variation Basal area Stem volume Aboveground biomass Basal area Stem volume Aboveground biomass Age2 Location (Block) 0.0928 0.0965 0.0967 0.1129 0.1205 0.1442 S x D(Block) 0.1177 0.1164 0.1192 S x C x D(Block) 0.0268 0.0245 0.0311 S x C x F(Block) 0.0647 0.0609 0.0364 Residual at 3440 tph <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Residual at 1240 tph <0.0001 <0.0001 <0.0001 0.0090 0.0103 0.0415 Age3 Location (Blo ck) 0.0915 0.0908 0.1037 0.2663 0.2733 0.2539 S x C(Block) 0.1119 0.1691 0.1380 S x D(Block) 0.1024 0.1654 0.0978 S x C x D(Block) 0.0153 0.0131 0.0197 0.2337 S x C x F(Block) 0.2404 0.1763 0.2309 Residual at 3440 tph <0.0001 <0.0001 < 0.0001 <0.0001 <0.0001 <0.0001 Residual at 1240 tph <0.0001 <0.0001 <0.0001 0.0026 0.0038 0.0143 Age5 Location (Block) 0.3690 0.3663 0.3769 0.1579 0.1594 0.1661 S x C(Block) 0.1667 0.1468 0.2149 0.1374 0.1356 0.1026 S x D(Block) 0.1707 0.1746 0.0992 S x C x D(Block) 0.0736 0.0688 0.1142 0.2415 0.2037 Residual at 3440 tph <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Residual at 1240 tph <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Sources of variation are Location ( S), planting density (D), silvicultural intensity (C), and family (F). Due to heterogeneity in the covariance structure with respect to the fixed effect of planting density and greater variation within the 2290 treesha1 treatment, the residual was group ed according to the planting density treatment. Basal area is expressed in m2ha1. Stem volume is expressed in m3ha1 and is calculated as the sum of per tree measurements of the volume of a cylinder to 1.37 m and the volume of a cone from 1.37 m t o the top if the tree. Aboveground biomass is expressed in metric tons per hectare and was calculated using individual tree allometric equations

PAGE 115

115 Table 48. Summary of mensurational characteristics by species, silvicultural treatment and planting dens ity at age two three and five years. (n = 56 plots for loblolly and n = 48 plots for slash pine when averaged across sites and families). Values in parentheses are one standard error of the mean. Silvicultural Treatments Operational Intensive Age and planting density DBH (cm) Height (m) treesha 1 DBH (cm) Height (m) treesha 1 Loblolly pine Age 2 1334 TPH 2.8(0.06) 2.68(0.04) 1328(1) 3.5(0.07) 2.88(0.04) 1329(0) 2990 TPH 2.7(0.06) 2.76(0.04) 2990 (1) 3.3(0.07) 2.97(0.04) 2990(0) Age 3 1334 TPH 6.5(0.12) 4.24(0.06) 1324(3) 8.5(0.10) 4.66(0.05) 1328(0) 2990 TPH 5.7(0.12) 4.27(0.07) 2984(2) 7.3(0.07) 4.67(0.05) 2984(3) Age 5 1334 TPH 11.0(0.21) 7. 10(0.08) 1232(9) 13.9(0.10) 7.73(0.05) 1227(8) 2990 TPH 8.6(0.18) 6.78(0.10) 2769(21) 11.1(0.10) 7.77(0.05) 2742(20) Slash pine Age 2 1334 TPH 2.4(0.07) 2.00(0.03) 1325(2) 2.7(0.07) 2.04(0.03) 1324(2) 2990 TPH 2.6(0.06) 2.10(0.02) 2985(3) 2.9(0.05) 2.10(0.02) 2982(4) Age 3 1334 TPH 5.5(0.12) 3.14(0.05) 1309(6) 6.2(0.14) 3.24(0.05) 1317(4) 2990 TPH 5.1(0.09) 3.32(0.04) 2973(8) 6.1(0.12) 3.43(0.04) 2973(6) Age 5 1334 TPH 10.5(0.17) 6.22(0.06) 1072(45) 13.1(0.10) 6.50(0.03) 1107(23) 2990 TPH 8.7(0.11) 6.31(0.04) 2690(35) 10.9(0.14) 6.81(0.04) 2608(36) *The operational treatment represents silviculture best practices in the southeast at the end of the 20th century, receiving a common site preparation treatment and single banded application of 280 kgha1 diammonium p hosphate at the time of planting. The contrasting intensive treatment is driven mainly by early complete vegetation con trol and annual fertilization. Planting density and subsequent density are expressed as the number of treesha1.

PAGE 116

116 Table 49. Age five contrasts between slash pine families grown in mixtures versus grown in pure plots. Deployment Variable Mixed P ure p value Basal Area 17.2 16.7 0.1016 Stem Volume 54.1 52.5 0.1089 Aboveground Biomass 38.4 37.4 0.0754 Intimate mixtures of all top performing slash pine fullsib families (S1, S2, S4, S5 and S6) were contrasted with the average of the same full sib families grown in pure blocks across two locations (Waldo and Perry), two planting densities, and two silvicultural intensities. Basal area is expressed in m2ha1. Stem volume is expressed in m3ha1 and is calculated as the sum of per tree measurements of the volume of a cylinder to 1.37 m and the volume of a cone from 1.37 m to the top if the tree. Aboveground biomass is expressed in metric tons per hectare and was calculated using individual tree allometric e quations.

PAGE 117

117 A 38 40 42 44 46 48 50 52 54 Sanderson Waverly Aboveground biomass (Mg ha -1) L1 L2 L4 L5 L6 L7 L8 A A AB AB B AB B B 30 32 34 36 38 40 42 44 Waldo Perry Aboveground biomass (Mg ha -1) S1 S2 S3 S4 S5 S6 A AB B C A A A B Figure 4 1. Standing crop biomass (metric tons per hectare) at age five demonstrating a genotype x location interaction for: A) loblolly pine (p=0.0388) and B) slash pine (p=0.0158). Data points within sites with the same letter are not significantly di fferent at the 90% level of confidence using Bonferroni's Least Significant Difference (LSD).

PAGE 118

118 A 40 45 50 55 60 65 70 75 80 85 90 95 Operational Intensive Stem volume (m3 ha -1) L1 L2 L4 L5 L6 L7 L8 A A AB AB B A AB B B B B 35 40 45 50 55 60 65 70 75 Operational Intensive Stem volume (m3 ha -1) S1 S2 S3 S4 S5 S6 A A AB C A A B C C C Figure 4 2. Standing volume (m 3 ha 1 ) at age five demonstrating a genotype x silviculture interaction for:A) loblolly pine (p=0.0019) and B) slash pine (p=0.0126). Data points within species and cultures having the same letter are not significantly different at the 90% level of confidence using Bonferroni's Least Significant Difference (LSD).

PAGE 119

119 A 0 10 20 30 40 50 60 70 2 3 4 5 Aboveground biomass (Mg ha -1) Intensive 2990 tph Intensive 1334 tph Operational 2990 tph Operational 1334 tph B A C B B A C D A B B B 0 10 20 30 40 50 60 70 2 3 4 5 Aboveground biomass (Mg ha -1) C B A D C B A D A B A B C 0 10 20 30 40 50 60 70 2 3 4 5 Age (years) Aboveground Biomass (Mg ha -1) C B A C B A D C A B C C Figure 4 3. Aboveground biomass accretion by silv icultural treatment for loblolly pine at A) Sanderson, FL, B) Waverly, GA, and C) slash pine across both locations with slash pine. Loblolly pine is expressed by location for ease of presentation due to a three way, location x silviculture x density inter action. There was no three way interaction for slash pine. Data points within ages on each graph having the same letter are not significantly different at the 90% level of confidence using Bonferroni's Least Significant Difference (LSD).

PAGE 120

120 30 35 40 45 50 55 60 65 Operational Intensive Loblolly Slash Aboveground biomass (Mg ha -1) Figure 4 4. Standing crop biomass (metric tons per hectare) at age five demonstrating a species x silviculture interaction for loblolly and slash pine (p<0.0001). Mixed family plots across two locations and two levels of silvicultural treatment intensity were compa red. Y error bars represent plus and minus one standard error of the mean.

PAGE 121

121 A 0 10 20 30 40 50 Waldo Perry Rust incidence (%) S1 S2 S3 S4 S5 S6 AB A ABC BCD CD D A A A B B B 0 10 20 30 40 50 Waldo Perry Wind damage (%) S1 S2 S3 S4 S5 S6 A A B B C C C Figure 4 5. A) Percent incidence of fusiform rust per plot at age five demonstrating a genotype x location interaction for slash pine (p=0.0189). Trees were considered i nfected if galls were noted on the branches or the main stem. B) Percent incidence of wind damage per plot at age five, also demonstrating a significant genotype x location interaction for slash pine (p<0.0001). Trees were considered to be impacted by wi nd if they were leaning by more than 22 degrees from vertical or had a broken top. Data points within sites with the same letter are not significantly different at the 90% level of confidence using Bonferroni's Least Significant Difference (LSD).

PAGE 122

122 CHAPTE R 5 SUMMARY AND FUTURE RESEARCH Summary R esearch studies aimed at quantifying the response of southern pines with the combined effects of silvicultural treatments and genetic improvement in loblolly and slash pine in pure family block plantings are very limited and documented genotype x silviculture interactions have not always been evident. The main goal of this investigation was to examine the stability of loblolly and slash pine full sib family performance across a contrasting range of locations, planti ng densities and silvicultural management intensities. Performance variables measured included: above ground and belowground biomass accumulation and distribution, nitrogen content, unit area stem volume, leaf area index light extinction coefficient, int erception of photosynthetically active radiation and radiation use efficiency. This study makes use of a series of replicated factorial experiments and family block plantings established in Florida and Georgia that manipulated gradients in planting densit y, understory competition and soil nutrient availability. Results for genotype x environment interactions (Table 51) were pre sented in three main chapters: C hapter two biomass production, distribution and nitrogen accumulation at age two, C hapter three interception and efficiency of PAR at ages four and five, and C hapter four basal area, stem volume and biomass production at age five. Biomass Production, Distribution, and Nitrogen Content at Age Two Chapter two makes usage of the full complement of five experimental locations in southeastern Georgia and northeastern Florida (t hree loblolly and t wo slash pine), two levels of silvicultural management intensity (operational and intensive), seven loblolly and six slash pine families (L1, L2, L4, L5, L6, L7, L8 and S1, S2, S3, S4, S5, S6), two levels of planting density (1334 and 2990 trees ha1). Biomass accumulation of the foliage, branch & bark, bole and

PAGE 123

123 belowground components were examined. In addition, the nitrogen accumulation of the foliage, above ground and belowground, and total components were quantified and tested across treatments The distribution of this biomass among the components was also examined. The objective was to determine for each species independently, if there were interactions between the factors of location, family, silvicultural management intensity, and planting density. After two growing seasons, biomass accumulation, N content and distribution were influenced by the combinations of silvicultural intensity, planting density, family and locations in plantations of loblolly and slash pine. Significant genotype x density, silviculture x density, and silviculture x location interactions existed for loblolly pine biomass accumulation, yet none were evident for slash pine. Total biomass accumulation in slash pine ranged from 4.7 to 11.6 Mg ha1 across planting densities. Loblolly pine TOTAL biomass effects of silviculture were not stable across planting density or locations. The response to the intensive silvicultural treatment was most pronounced on the high density plots (13.9 versus 17.2 Mg ha1, operational and intensive, respectively) and at the Sanderson, FL location (15.5 Mg ha1When distribution of this accumulated biomass to various components was examined, the only interaction that was significant was a three way silviculture x density x location for loblolly pine. This interaction was driven by a differential response to silviculture and planting density at the Bunnell, FL location. At this location, the pattern of biomass distribution within the low density treatment was the inverse of those expected for the intensive silvicultural treatment which had less biomass distributed to the BOLE component Loblolly pine f amilies varied in their distribution of biomass to FOL (19.6 to 20.8%, families L6 and L4 respectively), BOLE (30.0 to 30.9%, families L4 and L6) and BELOW (25.7 to 26.6%, families L4 and L6) ).

PAGE 124

124 Nitrogen accumulation in all biomass components demonstrated two types of interactions: genotype x density, and si lviculture x density x location. Families were not stable across the two planting densities tested in this experiment for N content for FOL (p=0.0235), ABOVE (p=0.0364), BELOW (p=0.0057), and TOTAL (p=0.0250) (i.e. a family x density interaction). In general, this interaction was due to scale effects among families, with the notable exception of two families (L4 and L8) which changed ranks in N content between planting densities. The threeway interaction (silviculture x density x location) was caused by an unequal response to silvicultural management intensity between planting densities at the three locations examined (FOL, p<0.0001; ABOVE, p<0.0001; BELOW, p=0.0001; TOTAL, p<0.0001). At the Bunnell, FL location, there were much lower than expected N co ntents in the FOL, ABOVE, and TOTAL components in the intensive silvicultural treatment combination at the 1334 trees ha Taken together, evidence of these complex interactions as early as agetwo, serves to underscore the importance of understanding how to best deploy elite genotypes of loblolly and slash pine given the uncertainty in predicting its response to a wide variety of man made and abiotic environmental factors For example the only three way interaction occurring for slash pine may have been due to the effects of a hurricane after the first growing season. While distribution of accumulated biomass did not appear to be involved in driving these interactions, variation in crown structure (McCrady and Jokela, 1996) or resource use efficiency may play a role (McKeand et al. 1997a) and was investigated in C hapter three. 1 Interception and Efficiency of PAR at Ages Four and Five Chapter three examines the stability of I PAR and RUE in selected elite full sib families of loblolly pine deployed in uniform blocks grown under intensive silviculture over time and locations. This chapter utilizes a subset of the experimental material from C hapter two since logistical constraints limited the experimental design to two loblolly pine locations (Sanderson,

PAGE 125

125 FL a nd Waverly, GA), six full sib families (L1, L2, L4, L5, L6, L7, and L8), at a combination of 2990 trees ha1The annual NF observed in this investigation ranged from 3.7 to 5.1 Mg ha and the intensive silvicultur al treatment While there was variation among families, there was no evidence for genotype x environment interactions (i.e. genotype x location or genotype x year). However, there were significant interactions for location x year. 1 across years and locations. There was a large decline in NF at t he Sanderson, FL location in 2005 which was thought to be due to early leaf senescence related to the effects of hurricanes Frances and Jeanne in 2004. Projected LAI varied among years and locations from a low of 1.7 to a maximum of 3.3 m m2. Cosine cor rected values of k averaged about 0.44 across locations and years with the exception of a significantly high value which occurred at the Sanderson, FL location in 2005 ( k =0.83). Aboveground net primary productivity was similar among families and likely d ue to the fact that they were all chosen for superior performance in growth. Similarly, the silvicultural treatment combination applied in this experiment, which combined high planting density with intensive silvicultural management, likely evened the inh erent differences in site quality between the locations ; hence, the lack of significant differences in ANPP between locations. There was a statistically significant effect of family on growing season RUE and values ranged from 1.08 to 1.16 g MJ1Basal Area, Stem Volume and Biomass Production at Age Five ( families L8 and L4, respectively). The more d ensely spaced plots tended to be more efficient at capturing and utilizing PAR Results from this experiment suggest that RUE in fullsib families of loblolly pine was stable across contrasting soil types early in the stage of stand development. The fourth chapter examined genotype x environmental responses in biomass production and stem volume through age five years. This investigation utilized two locations from within each species, co mbining silvicultural treatment intensity (operational v ersus intensive), planting

PAGE 126

126 density (1334 v ersus 2990 treesha1Age five unit area aboveground biomass ( AGB ) and stem volume ( VOL ) varied significantly in response to several treatment combinations: genotype x location genotype x silvicultur e, and silvicultur e x density for BA and VOL. Family interactions with s ilvicultur al intensity were positive and the best overall performing families respond ed the most under the intensive silvicultural treatment. For slash pine families S2 and S6 were the most responsive in VOL (63% increase due to silvicultural intensity). Loblolly pine family L4 was the most respo nsive to silvicultural treatment (75% increase in VOL), while L5 was the least responsive (55% increase). Accumulation of bi omass at age five interacted with silvicultural management intensity and planting density. Response to silvicultural treatment was the greatest under the narrowest planting density. This was evidenced for slash pine where the intensive silvicultural treatment increased aboveground standing biomass by 5.7 Mg ha ) and full sib families (seven loblolly and six slash pine families). 1 at 1334 trees ha1 versus 12 Mg ha1 at 2990 trees ha1. There was a t hree way interaction for loblolly pine aboveground biomass. At the Sanderson, FL location, the 2990 trees ha1However, t here were changes in slash pine family rankings between locations, which were partly explained by reductions in growth associated with a combination of fusiform rust infection [ Cronartium quercum (Berk.) Miyabe ex Shirai f. sp fusiforme ] and wind damage from t he 2004 density under the operational silviculture treatment produced a much lower response than the same treatment combination at the Waverly, GA locat ion. This was attributed to the Sanderson, FL location having a lower inherent level of productivity than the Waverly, GA location. No three way interactions, which included family, were evident and all genetic sources were stable across the contrasting planting densities.

PAGE 127

127 hurricane season. Loblolly pine generally had a low incidence of rust infection and no significant interactions were found. At age five, lobloll y pine outperformed slash pine in aboveground biomass accumulation, especially under the intensive silvicultural management intensity. While loblolly performance was similar whether deployed in mixtures or pure family blocks, slash pine tended to be more productive in intimate mixtures than when grown in pure family blo cks. Focus of Future Research Future research should build on the strengths and limitations of this investigation and provide insight into the biological mechanisms that drive genotype x environment interactions in southern pines. The strengths of this st udy were the large contrasts within the treatments of silvicultur al treatment intensity and planting density across locations with contrasting soil types. Also, elite fullsib families were deployed in large family block planting s allowing for unit area estimates over time. Moreover, t he experimental design had a high statistical power for the detection of interactions among these multiple treatment combinations. However, the study could have been more informative if a larger number of families were tes ted especially those which had contrasting traits for biomass accumulation, distribution, and resource use efficiency. A wider distribution of locations across regions may have also provided more information as to how these elite families may vary with local climate and disease conditions. Informative research investigations of genotype x environment interactions could focus on particular biological mechanisms of forest plantation productivity, such as nutrient uptake or light capture, in large block planting experiments. These experiments should incorporate a wide range of genotypes which have been selected for contrasting traits which have been identified as having an influence on these biological mechanisms For example, genotypes with a large contras t in the distribution and structure of fine roots could provide insight into genotype x

PAGE 128

128 environment interactions for nutrient uptake. Similarly, genotypes with contrasting crown architecture could yield valuable information regarding the mechanisms drivin g light capture and use. Additionally, increasing the uniformity within genotypes (i.e. full sib families to clones within families), while increasing the uniformity of experimental material, is thought to lead to increased genotype x environmental interactions and should be utilized in future studies where appropriate. While this investigation tested several genotypes across contrasting soil types (Spodosols versus Ultisols), these were were in close geographic proximity to each other with similar environmental conditions Despite this, there were large genotype x environment interactions which were driven by the effects of localized disease incidence and hurricane proximity. Future research should test selected genotypes (clones) across a wider range of contrast ing physical and environmental conditions Environmental and physical attributes should contrast: soil types ( i.e. rooting volume, physical properties and nutrient availability), climate ( i.e. temperature, relative humidity, light regime, and th e distribution of annual precipitation) and pest incidence ( i.e. fusiform rust and Nantucket tip moth incidence). This wide gradient of environmental conditions could be achieved by establishing trials across contrasting regions where southern pine has been deployed, such as Hawaii, South America and the southeastern United States ).

PAGE 129

129 Table 5 1. Significant interactions for loblolly and lash pine by variable and year as described in this investigation. Trait Age s Interaction Loblolly pine Tot al b iomass accumulation 2 silviculture x density, silviculture x location, genotype x density Biomass distribution 2 silviculture x density x location Nitrogen accumulation 2 genotype x density, silviculture x density x location Needlefall 4 &5 locat ion x year Leaf area index 4&5 location x year Fraction of PAR intercepted( f ) 4&5 location x year Light extinction coefficient ( k ) 4&5 location x year Intercepted PAR 4&5 location x year Basal area 2,3&5 genotype x location silviculture x densi ty x family Stem volume accumulation 3&4 silviculture x density Stem volume accumulation 2& 5 genotype x location Stem volume accumulation 2 silviculture x density x location Aboveground biomass accumulation 3 silviculture x density x location Abov eground biomass accumulation 2,3,& 5 genotype x location Slash pine Aboveground n itrogen accumulation 2 density x location Belowground nitrogen accumulation 2 silviculture x density x location Basal area 3&5 silviculture x density location x density Basal area 3 silviculture x location Basal area 3 & 5 genotype x location Stem volume accumulation 3 &5 silviculture x density, location x density Stem volume accumulation 3 silviculture x location Stem volume accumulation 3 & 5 genotype x location Stem volume accumulation 5 genotype x silviculture Aboveground biomass accumulation 3&5 genotype x location, silviculture x density, location x density

PAGE 130

130 APPENDIX A GLOSSARY OF TERMS, UNITS AND DESCRIPTIONS Table A 1. Glossary of t erms unit s and descriptions. Term Unit Description Diameter at breast height cm The diameter of individual trees measured (DBH). Height m Total height of individual trees (HT). Stem volume m 3 ha 1 Unit area volume of the bole component (VOL). B asal area m 2 ha 1 The cross sectional area of all trees per hectare measured at 1.37 meters above the ground level (BA). Biomass Mg ha 1 Unit area biomass of various tree components (foliage, branches & bark, bole, aboveground, belowground and total). Nitrogen content kg ha 1 Nitrogen content of various tree components (foliage, branches & bark, bole, aboveground, belowground and total). Aboveground net primary productivity Mg ha 1 yr 1 Aboveground unit area biomass production including needl efall (ANPP). Needlefall Mg ha 1 yr 1 Amount of needles senesced per unit area over time (NF). Leaf area index m 2 m 2 Measure of the projected surface area of leaves per unit area of ground (LAI). Fraction of photosynthetically active ra diation fraction Fraction of photosynthetically active radiation below the canopy to that above the canopy ( f ). Intercepted photosynthetically active radiation MJ m 2 yr 1 Sum of photosynthetically active radiation (400 to 700 nm wavelength) intercep ted per unit area over time (IPAR). Light extinction coefficient unit less A quantitative description of the rate of decreased irradiance vertically through a canopy ( k ). Radiation Use Efficiency g MJ 1 The weight of biomass produced per unit i nput of intercepted photosynthetically active radiation (RUE).

PAGE 131

131 APPENDIX B HISTORICAL ANNUAL RA INFALL, AVERAGE TEMP ERATURE, MINIMUM TEM PERATURE, MAXIMUM TEMPERATURE DATA BY LOCATION Table B 1. Historical annual rainfall, average temperature, minimum temperature, maximum temperature data by location Variable/Year Sanderson, FL Waverly, GA Bunnell, FL Perry, FL Waldo, FL Annual Precipitation (mm yr 1 ) 2000 802 1118 1076 895 872 2001 1075 1073 1948 1051 1070 2002 1355 1094 1594 1578 1 405 2003 1610 933 1327 2076 1184 2004 1914 1119 1906 1521 1483 mean 1961 to 1990 1409 1260 1424 1477 1316 A verage t emperature ( 0 C) 2000 20.2 19.5 21.2 19.9 20.1 2001 20.6 20.2 21.4 20.2 20.6 2002 20.7 20.5 21.7 20.4 20.7 2003 2 0.2 19.6 21.8 21.3 20.4 2004 20.3 19.9 21.3 20.3 20.7 mean 1961 to 1990 19.9 19.8 21.1 20.4 20.3 Minimum t emperature ( 0 C) 2000 6.7 5.0 5.6 9.4 7.2 2001 5.0 5.0 4.4 7.8 5.6 2002 5.0 3.3 3.3 7.8 5.0 2003 7.8 7.2 7. 8 7.2 6.7 2004 5.0 2.8 3.9 6.7 5.0 Maximum t emperature ( 0 C) 2000 38.3 39.4 36.7 37.8 42.2 2001 36.7 36.1 35.6 36.7 35.6 2002 37.2 36.7 36.1 37.2 37.2 2003 35.0 35.0 35.6 35.0 34.4 2004 35.6 35.6 38.3 36.1 36.7 Noaa: Sander son = L ake City 2B, Waverly = Brunswick McKinnon AP, Bunnell = Deland 1 SSE, Perry = Perry, Waldo = Gainesville Regional AP. (NOAA National climatic data center http://www.ncdc.noaa.gov/oa/ncdc.html Last accessed April 22nd 2009)

PAGE 132

132 Table B 2. Palmer Drought Severity Index (PDSI) annual values for three regions in FL and GA. Y ear Waverly, GA Sanderson, Perry, and Waldo, FL Bunnell FL 2000 2.64 2.96 3.63 2001 2.37 2.38 1.79 2002 2.1 5 0.59 0.83 2003 0.82 1.20 2.19 2004 0.24 0.60 0.89 Note: This index quantifies the severity of a wet or dry spell. This is based on the principles of a balance between moisture supply and demand. Man made changes were not considered in this calcul ation. The index generally ranges from 6 to +6, with negative values denoting dry spells and positive values indicating wet spells. There are a few values in the magnitude of +7 or 7. PDSI values 0 to .5 = normal; 0.5 to 1.0 = incipient drought; 1 .0 to 2.0 = mild drought; 2.0 to 3.0 = moderate drought; 3.0 to 4.0 = severe drought; and greater than 4.0 = extreme drought. Similar adjectives are attached to positive values of wet spells. This is a meteorological drought index used to assess th e severity of dry or wet spells of weather.

PAGE 133

133 APPENDIX C GENERAL SOIL PROPERT IES OF THE FIVE PPINES EXPERIMENTAL LOCATIONS IN SOUTHEAST GEORGIA AND NORTHEAST FLORIDA Table C 1. General soil properties of the five PPINES experimental locations in southea st Georgia and northeast Florida Sanderson, FL Waverly, GA Bunnell, FL Perry, FL Waldo, FL Soil o rder Spodosol Ultisol Spodosol Spodosol Spodosol Soil s eries Leon sand (sandy, siliceous, thermic Aeric Alaquod s ) Bladen (mixed, semiactive, therm ic Typic Albaquults ) Myakka sand ( s andy, siliceous, hyperthermic Aeric Alaquods ) Leon (sandy, siliceous, thermic Aeric Alaquods) Newnan (sandy, siliceous, hyperthermic Ultic Haplohumods) Depth class Very deep Very deep Deep, to very deep Very deep Very d eep Drainage class Poorly, to very poorly drained Poorly drained Poorly, to very poorly drained Poorly, to very poorly drained Somewhat poorly drained Reactivity Extremely acidic, to slightly acidic Extremely acidic, to strongly acidic Acidic, to strong ly acidic Extremely acidic, to slightly acidic Extremely, to moderately acidic USDANRCS Soil Survey Division, Official soil series descriptions. PPINES : Pine Productivity Interactions on Experimental Sites.

PAGE 134

134 APPENDIX D RAW DATA UTILIZED TO DEVELOP ALLOMETRIC EQUATIONS Table D 1. Raw data utilized to develop allometric equations from the Bunnell, FL location at age 2. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 1334 Intensive L2 1 70 2. 42 3.8 0.2 0 10.2 2.08 2.15 1.9993 1.4494 0.375104 1.517796 1334 Intensive L2 3 77 2.20 2.7 0.2 0 7.9 1.64 1.52 1.3489 0.7946 0.23855 0 0.83815 0 1334 Intensive L2 4 7 1.94 1.5 0.01 6.9 1.65 1.51 1.1313 0.624 0 0.15906 0 0.47294 0 0.416 0 0.691 0 1334 Operat ional L2 1 73 2.25 1.9 0.3 0 6.9 1.28 1.29 0.6434 0.4233 1334 Operational L2 3 79 2.80 3.2 0.19 8.3 1.82 1.62 1.2531 0.7825 0.18036 0 1.16124 0 1334 Operational L2 4 1 2.31 1.8 0.21 7.5 1.61 1.76 1.2195 0.6061 0.199396 0.783704 0.2853 0.5816 2990 Int ensive L2 1 12 3.32 4.3 0.37 10 .0 1.7 0 1.55 1.2563 0.9118 0.460987 1.750213 0.6858 0.9406 2990 Intensive L2 3 4 2.58 3.7 0.23 9.3 1.73 1.67 1.4409 1.0869 0.311846 1.471054 2990 Intensive L2 4 6 2.32 1.8 0.21 7.2 1.43 1.62 0.958 0 0.4584 0.199242 0.64235 8 2990 Operational L2 1 6 3.21 3.2 0.23 7.6 1.45 1.5 0 1.0635 0.574 0 0.311367 1.159533 2990 Operational L2 3 127 2.04 1.5 0.46 4.8 0.88 0.86 0.2103 0.1331 0.100299 0.293601 2990 Operational L2 4 125 2.49 2.6 0.21 8.5 1.86 1.98 1.4054 0.9211 0.24309 0 0.90941 0 0.7699 0.9218 1334 Intensive L4 1 20 2.90 4.2 0.05 10.8 2.36 2.08 2.6396 1.689 0 0.353197 1.752203 1334 Intensive L4 3 5 2.20 1.9 0.32 8.2 1.7 0 1.38 1.4466 0.872 0 0.232867 1.039833 1334 Operational L4 1 9 2.05 1.7 0.01 6.6 1.35 1.64 1.2691 0.8082 0.153526 0.486474 1334 Operational L4 3 41 2.95 3.3 0.41 8.5 1.75 1.59 1.0709 0.8363 0.285786 1.095514 1334 Operational L4 4 7 2.23 1.8 0.27 7.3 1.2 0 1.32 0.8104 0.3953 0.177883 0.573817 0.1556 0.3609 2990 Intensive L4 1 81 3.44 4.4 0.25 9.8 1.9 0 2.35 1.8884 1.6298 0.372667 2.060933 0.3749 1.0445 2990 Intensive L4 3 3 2.27 2 .0 0.25 7.8 1.6 0 2.02 1.3426 0.9867 0.159796 0.686904 2990 Intensive L4 4 115 2.32 2.1 0.28 6.7 1.36 1.62 0.8189 0.4482 0.124543 0.650657 2990 Operational L4 1 122 3.48 3.6 0.21 7.7 1.8 0 1.72 1.0685 0.8756 0.290712 1.477088 2990 Operational L4 3 115 2.54 2.9 0.39 6.9 1.36 1.79 0.508 0 0.4693 0.258857 0.735943 2990 Operational L4 4 11 2.03 1.5 0.32 6.6 1.35 1.42 0.6676 0.3894 0.156072 0.461528 0.2889 0.3794 Note : For all tables in Appendix D; DENSITY is expressed in trees per hectare, FAM refers to Family (L is loblolly and S is slas h pine), TREE is tree number, HEIGHT is total height in meters, DBH is in centimeters, HTLC is the height to the base of the live crown in meters, GLD is groundline diameter in centimeters, CWAL is crown width along the beds in meters and CWAC is crown width across the beds in meters, FOLIAGE is the total dry weig ht of green foliage in kilograms, BRANCH is the total dry weight of bra nches in kilograms, BARK is the total dry weight of bark in kilograms, BOLE is the total dry weight of the bole wood in kilograms, CROOT is the total dry weight of roots greater than 2 millimeters in diameter centered on a 1 square meter section to 40 cent imeters depth in kilograms, and TAP is the total dry weight of the taproot in kilograms.

PAGE 135

135 Table D 2. Raw data utilized to develop allometric equations from the Perry, FL location at age 2. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 1334 Intensive S6 1 96 1.99 2.5 0.3 0 8.4 1.25 1.5 0 2.1453 0.8269 0.301743 0.547257 1334 Intensive S6 2 4 2.50 4.1 0.3 0 8.9 1.45 1.95 2.2553 0.7892 0.540589 0.911011 0.4005 0.8964 1334 Intensive S6 3 72 3.2 3 4.9 0.21 9.6 1.79 2.11 3.0797 1.2585 0.694913 1.472187 0.7344 1.0651 1334 Operational S6 1 2 2.51 3.3 0.25 7.2 1.44 1.71 1.658 0 0.4443 0.307644 0.805956 1334 Operational S6 2 1 2.08 3.2 0.39 6.6 1.36 1.43 1.166 0 0.4177 0.378365 0.407935 0.1735 0.5498 1334 Operational S6 3 74 1.84 2.1 0.41 6 .0 1.12 0.99 0.6873 0.26 00 0.20584 0 0.31916 0 0.1278 0.3591 1334 Operational S6 4 2 2.61 3.7 0.51 8.1 1.64 1.71 1.7699 0.6999 0.469514 0.708886 2990 Intensive S6 1 116 1.85 2.4 0.45 6.9 1.24 1.4 0 0.9379 0.2818 0 .217668 0.460232 2990 Intensive S6 2 121 2.29 3 .0 0.21 8.3 1.9 0 2 .00 2.0256 0.7007 0.356322 0.740278 0.4097 0.7402 2990 Intensive S6 3 127 2.98 4.3 0.23 9.3 1.57 1.8 0 2.0559 0.7573 0.569735 1.196065 0.7409 0.8588 2990 Operational S6 1 12 1.58 1.3 0.3 0 5.4 1.1 0 1.08 0.4652 0.1507 0.126927 0.154173 2990 Operational S6 2 113 2.69 4.5 0.3 0 10.0 1.73 2.03 3.0237 1.444 0 0.597148 1.329352 0.5761 1.3678 2990 Operational S6 3 15 2.46 4.1 0.29 9.3 1.56 1.4 0 1.7066 0.706 0 0.488642 0.837158 0.4557 0.8823 2990 Operational S6 4 6 2.29 3.4 0.24 8.2 1.79 1.69 1.4725 0.6223 0.385816 0.626684 1334 Intensive S1 1 126 1.79 2 .0 0.35 6.6 1.14 1.22 0.827 0 0.2408 0.191367 0.345933 1334 Intensive S1 2 3 1.80 2.7 0.31 8 .0 1.44 1.39 1.5151 0.4442 0.317754 0.452846 0.38 69 0.5581 1334 Intensive S1 3 5 2.74 4.5 0.19 8.6 1.84 1.81 2.4361 0.8204 0.615309 1.093291 0.5377 0.8622 1334 Operational S1 1 32 2.30 3.1 0.24 8.6 1.64 1.74 1.4823 0.5066 0.374399 0.714601 1334 Operational S1 2 10 1.70 1.6 0.19 6.2 1.15 1.02 0.7542 0.1615 0.182176 0.236924 0.1895 0.3045 1334 Operational S1 3 31 2.32 3.7 0.18 9 .0 1.35 1.37 1.4716 0.4939 0.439875 0.804725 0.219 0 1.1023 1334 Operational S1 4 75 2.46 3.4 0.2 0 9.3 1.47 1.64 1.6562 0.7018 0.409141 0.664759 2990 Intensive S1 1 3 1.74 2 .2 0.28 6.2 1.12 1.51 0.894 0 0.2096 0.194299 0.337801 2990 Intensive S1 2 13 2.54 4.1 0.23 8.9 1.74 1.76 1.9971 0.6639 0.575313 0.959287 0.5628 0.9062 2990 Intensive S1 3 13 2.60 4.2 0.3 0 9.4 1.7 0 1.9 0 2.0986 0.686 0 0.628359 0.990141 0.3803 0.7521 299 0 Operational S1 1 124 2.50 3.6 0.23 7.9 1.18 1.42 1.1744 0.359 0 0.470901 0.679999 2990 Operational S1 2 5 1.66 1.9 0.39 5.7 1.17 1.14 0.5927 0.1621 0.149687 0.215313 0.075 0 0.2642 2990 Operational S1 3 3 2.07 3.1 0.24 8.1 1.54 1.7 0 1.4365 0.5125 0.334 085 0.641515 0.4098 0.6234 2990 Operational S1 4 126 2.55 3.5 0.24 7.8 1.51 1.48 1.2111 0.3375 0.388529 0.777771

PAGE 136

136 Table D 3. Raw data utilized to develop allometric equations from the Waldo, FL location at age 2. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 1334 Intensive S6 1 10 2.20 4 .0 0.17 9 .0 2 .00 1.6 0 3.0356 1.2928 0.520445 0.907655 0.4454 0.8535 1334 Intensive S6 2 74 2.45 3.3 0.38 7.5 1.59 1.65 2.0905 0.6763 0.561258 1.015342 1334 Intensive S6 3 71 1.77 2.1 0.22 6.2 1.6 0 1.75 1.3901 0.4907 0.190888 0.406612 0.1111 0.3983 1334 Intensive S6 4 8 2.10 3.5 0.19 9.1 1.6 0 1.9 0 2.0303 0.672 0 0.388469 0.654631 1334 Operational S6 1 74 2.20 2.6 0.05 6.9 1.9 0 1.5 0 1.4639 0.4683 0.2 47864 0.531136 0.2465 0.5603 1334 Operational S6 2 123 2.61 3.2 0.1 0 7.1 1.74 1.19 1.3896 0.524 0 0.440034 0.707166 1334 Operational S6 3 3 1.91 2.1 0.29 5.3 1.22 1.15 0.6997 0.1994 0.216668 0.302732 0.1823 0.2682 1334 Operational S6 4 9 2.18 2.6 0.1 0 6.8 1.4 0 1.3 0 0.9933 0.3173 0.356395 0.447005 2990 Intensive S6 1 126 2.02 3.2 0.04 7.7 1.9 0 1.7 0 2.0282 0.9394 0.237175 0.509125 0.4734 0.5047 2990 Intensive S6 2 30 1.83 2 .0 0.24 6.1 1.19 1.2 0 0.818 0 0.2029 0.206956 0.308944 2990 Intensive S6 3 11 7 2.69 4 .0 0.27 8.3 1.63 1.5 0 2.0611 0.7134 0.576441 1.001759 0.3921 0.8283 2990 Intensive S6 4 8 2.55 4.1 0.1 0 8.7 1.6 0 1.8 0 1.9619 0.6508 0.626494 1.077106 2990 Operational S6 1 3 2.29 3.3 0.02 8.1 1.6 0 1.9 0 1.7945 0.7292 0.411277 0.844523 0.5452 0.5 278 2990 Operational S6 2 11 2.51 2.6 0.02 7.3 1.4 0 1.15 1.25 00 0.6663 0.347706 0.628894 2990 Operational S6 3 9 1.66 1.9 0.12 5 .0 1 .00 0.98 0.4264 0.1144 0.140267 0.213333 0.0449 0.1802 2990 Operational S6 4 4 2.56 3 .0 0.25 6.8 1.64 1.59 1.1806 0.317 0 0.342947 0.606753 1334 Intensive S1 1 4 2.08 3.5 0.18 8.2 1.9 0 2 .00 2.2294 0.8992 0.444257 0.656243 0.4862 0.7035 1334 Intensive S1 2 80 2.63 4 .0 0.21 8 .0 1.79 1.68 2.1209 0.6998 0.552539 1.037061 1334 Intensive S1 3 6 2.20 3.6 0.18 8.3 1.37 1.39 1.5683 0.4347 0.450499 0.730001 0.2928 0.595 0 1334 Intensive S1 4 10 1.55 1.6 0.35 5.1 1 .00 0.8 0 0.4784 0.0802 0.142004 0.179696 1334 Operational S1 1 78 2.21 3.4 0.2 0 8.5 1.7 0 2 .00 2.0702 0.7119 0.414243 0.709157 0.3604 0.8613 1334 Operational S1 2 2 0 1.61 1.3 0.17 5.1 1.03 1.04 0.5805 0.1069 0.150655 0.175545 1334 Operational S1 3 76 2.38 3.5 0.15 8 .0 1.57 1.69 1.9282 0.7629 0.525205 0.783095 0.4418 0.8052 1334 Operational S1 4 5 2.10 2.2 0.3 0 5.9 0.9 0 1 .00 0.713 0 0.1463 0.225553 0.321547 2990 Intensive S1 1 126 1.45 1.3 0.02 6.2 1.3 0 1.3 0 0.9496 0.2259 0.188882 0.238618 0.1602 0.2887 2990 Intensive S1 2 41 2.31 2.7 0.15 7.1 1.6 0 1.6 0 1.2378 0.3772 0.411945 0.535155 2990 Intensive S1 3 123 2.60 3.9 0.03 7.8 1.16 1.8 0 1.4101 0.3549 0.420664 0.724836 0.4343 0.5717 2990 Intensive S1 4 1 2.55 3.6 0.35 7.5 1.5 0 1.4 0 1.5596 0.469 0 0.460796 0.801904 2990 Operational S1 1 3 1.75 2 .0 0.08 7 .0 1.3 0 1.9 0 1.0707 0.4046 0.266551 0.384949 0.3382 0.3761 2990 Operational S1 2 2 2.36 3.6 0.07 7.4 1.4 0 1 .35 1.2435 0.3661 0.458898 0.619702 2990 Operational S1 3 11 2.17 2.6 0.19 7 .0 1.06 1.3 0 0.8249 0.182 0 0.335544 0.524456 0.1717 0.3276 2990 Operational S1 4 15 2.80 4.1 0.22 7.4 1.46 1.21 1.3039 0.4125 0.540188 0.816012

PAGE 137

137 Table D 4. Raw data utilized to develop allometric equations from the Sanderson, FL location at age 2. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 1334 Intensive L2 1 4 2.95 3.5 0.25 8.7 1.64 1.84 2.0391 0.9833 0.1935 95 1.701805 1334 Intensive L2 2 2 3.80 4.3 0.25 9.8 2.12 2.31 3.0527 2.2758 0.570441 1.965059 0.927 0 1.6907 1334 Intensive L2 3 40 2.09 1.6 0.16 6.6 1.5 0 1.52 1.1152 0.621 0 0.084507 0.578693 1334 Operational L2 1 5 3.64 3.8 0.18 9.3 2.4 0 2.32 2.4474 1.8532 0.500576 1.478524 1334 Operational L2 2 8 3.07 3.6 0.22 9.8 2.1 0 1.78 2.0887 1.2268 0.261541 1.715159 0.2953 1.3505 1334 Operational L2 3 1 2.52 2.3 0.19 6.9 1.6 0 1.65 1.3519 0.8934 0.161003 0.745697 2990 Intensive L2 1 10 3.22 3.6 0.51 7.8 1.28 1.8 0 1.4088 0.6836 0.299616 1.792484 2990 Intensive L2 2 126 2.72 2.4 0.4 0 8.2 1.95 1.49 1.4938 0.687 0 0.248792 0.792008 0.5125 0.6553 2990 Intensive L2 3 4 3.56 4.3 0.23 9.7 1.83 2.24 2.0789 1.2958 0.489258 2.462842 2990 Operational L2 1 125 2 .61 2.4 0.19 6.6 1.24 1.64 1.135 0 0.4955 0.129914 0.673286 2990 Operational L2 2 14 3.04 2.9 0.19 7 .0 1.35 1.38 1.1446 0.5655 0.216974 0.804326 0.32 00 0.6055 2990 Operational L2 3 124 3.09 3.3 0.23 10.9 1.67 1.85 1.2378 0.8821 0.32214 0 1.45946 0 1334 Intensive L4 1 8 3.93 5.7 0.22 11.5 2.7 0 2.55 3.994 0 2.9101 0.562096 3.771904 1334 Intensive L4 2 76 3.14 2.8 0.31 8.5 1.76 1.9 0 1.9891 1.3182 0.193095 0.988005 1334 Intensive L4 3 8 2.96 3.6 0.22 8.9 2.33 2.1 0 2.6945 1.8854 0.34836 0 1.87854 0 133 4 Operational L4 1 73 2.84 3.4 0.25 8.4 1.85 1.82 2.4216 1.4661 0.250277 1.675123 1334 Operational L4 2 71 2.76 2.5 0.02 7.1 1.46 1.47 1.1628 0.6002 0.143202 0.628898 0.376 0 0.8027 1334 Operational L4 3 6 3.34 3.6 0.32 9 .0 2.45 2.42 2.3765 1.9574 0.316 95 1.38205 0 2990 Intensive L4 1 126 3.50 4.6 0.02 9.5 2.35 1.8 0 3.1121 1.85 00 0.491626 2.906674 2990 Intensive L4 2 2 3.94 5 .0 0.28 9.8 2.05 2.3 0 2.4143 1.6231 0.51932 2.04058 0 0.6239 1.1592 2990 Intensive L4 3 128 2.85 3.3 0.38 7.9 1.44 1.46 1.2372 0.7529 0.204887 1.050213 2990 Operational L4 1 120 3.58 4.3 0.14 8.6 1.68 2.15 1.9006 1.2712 0.2801 00 1.3678 00 2990 Operational L4 2 117 2.60 2.1 0.22 6.6 1.35 1.58 0.8465 0.5332 0.180528 0.542572 0.8747 0.5137 2990 Operational L4 3 12 2.03 1.8 0.1 2 5.8 1.17 1.28 0.713 0 0.3459 0.0826 00 0.4169 00

PAGE 138

138 Table D 5. Raw data utilized to develop allometric equations from the Waverly, GA location at age 2. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 1334 Intensive L2 1 8 2.11 1.8 0.23 5.9 1.12 1.23 1.1391 0.8627 0.096729 0.417871 1334 Intensive L2 2 3 2.95 3.1 0.14 8.7 1.55 1.69 2.2978 1.1845 0.361774 1.247126 0.5656 1.0484 1334 Intensive L2 3 73 2.34 2.8 0.17 7.2 1.23 1.35 1.4804 0.699 7 0.179847 0.675153 1334 Operational L2 2 9 3.14 3.3 0.15 7.9 1.38 1.49 1.924 0 0.6339 0.308182 1.173318 0.269 0 0.8558 1334 Operational L2 3 73 2.99 3.5 0.17 7 .0 1.2 0 1.13 1.2734 0.559 0 0.155815 0.709185 2990 Intensive L2 1 2 2.60 3.3 0.26 8.1 1.88 1 .45 2.1357 1.1878 0.195979 0.925121 2990 Intensive L2 2 125 2.24 2.1 0.35 5.7 1.14 1.34 0.6316 0.2473 0.148617 0.387783 0.2052 0.3724 2990 Intensive L2 3 9 2.82 3.3 0.15 7.8 1.55 1.73 1.2008 0.7586 0.23251 0 0.95329 0 2990 Operational L2 1 125 2.30 2. 3 0.21 5.2 1.04 1.19 0.5506 0.2421 0.106025 0.318075 2990 Operational L2 2 120 2.95 2.9 0.25 6.9 1.36 1.67 1.4353 0.7634 0.221482 0.884018 0.5576 0.538 0 2990 Operational L2 3 124 3.39 3.5 0.21 7.4 1.38 1.4 0 1.3602 0.5697 0.200931 1.142569 1334 Inten sive L4 1 77 3.27 4.4 0.13 9.8 1.98 2.1 0 2.4741 1.7733 0.459487 1.688713 1334 Intensive L4 2 72 2.91 2.9 0.34 8.2 1.64 1.72 2.1399 0.792 0 0.244071 1.015429 1334 Intensive L4 3 2 3.44 4.5 0.23 10 .0 1.88 1.82 2.4949 1.7038 0.407901 1.716099 1334 Ope rational L4 1 5 3.24 3.4 0.19 7 .0 1.66 1.41 1.6942 0.8265 0.209492 0.801308 1334 Operational L4 2 78 2.53 2.2 0.02 5.6 1.36 1.31 0.8475 0.3751 0.151125 0.468875 0.1405 0.5245 1334 Operational L4 3 8 2.88 3.1 0.19 6.8 1.37 1.42 0.714 0 0.4606 0.241195 0. 853005 2990 Intensive L4 1 8 3.12 3.5 0.27 7.2 1.39 1.58 1.6439 0.9477 0.238393 0.954707 2990 Intensive L4 2 116 3.55 3.9 0.36 8.3 1.5 0 1.6 0 1.8904 0.9539 0.274692 1.291208 0.4717 0.9361 2990 Intensive L4 3 4 2.41 2.5 0.35 6.8 1.22 1.36 1.0046 0.496 7 0.14701 0 0.52529 0 2990 Operational L4 1 5 2.72 3.2 0.14 7.2 1.43 1.54 1.2515 0.5328 0.172311 0.702089 2990 Operational L4 2 122 3.04 3 .0 0.21 6.5 1.4 0 1.33 1.3144 0.592 0 0.155577 0.849323 0.5117 0.5961 2990 Operational L4 3 3 2.71 2.7 0.24 7.2 1.5 2 1.63 1.8726 0.8959 0.252374 0.758826

PAGE 139

139 Table D 6. Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, operational silviculture and family L4. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRA NCH BARK BOLE CROOTS TAP 2990 Operational L4 1 117 8.90 10.9 4 .0 15.1 1.7 2.1 2.95649 0 4.089143 2.543035 13.55602 2.0684 4.03 0 2990 Operational L4 1 119 8.70 9.5 3.8 13.1 1.2 1.9 1.319228 1.712753 1.935347 9.597816 1.5594 3.12 0 2990 Op erational L4 1 124 8.60 11.1 3.6 16.0 1.9 2.2 2.397552 3.003228 2.725137 11.71901 2.0003 4.03 0 2990 Operational L4 2 2 7.30 8.7 2.7 13.4 1 .0 1.3 1.773862 2.318415 1.602432 7.561728 1.2511 3.085 2990 Operational L4 2 7 6.60 7.5 3.3 10.5 0.9 1.6 0.835146 1 .163054 1.128934 4.595382 0.9319 1.795 2990 Operational L4 2 10 7.20 9.4 3.2 14 .0 0.9 2.3 1.364494 2.465949 1.918021 7.986881 0.9411 3.2 00 2990 Operational L4 3 3 6.70 8.5 1.2 13.0 1.6 2.6 1.863535 2.598732 1.816049 5.918531 1.8901 2.455 2990 Operationa l L4 3 8 6.60 8.2 2.3 12.2 1.3 1.8 1.362917 1.757734 1.222061 5.475812 1.1872 1.935 2990 Operational L4 3 10 5.20 6.2 2.1 10 .0 1.2 1.3 0.700546 0.642764 0.711522 3.098149 0.8232 0.98 0 2990 Operational L4 1 117 8.90 10.9 4 .0 15.1 1.7 2.1 2.95649 0 4.089143 2.543035 13.55602 2.0684 4.03 0 2990 Operational L4 1 119 8.70 9.5 3.8 13.1 1.2 1.9 1.319228 1.712753 1.935347 9.597816 1.5594 3.12 0 2990 Operational L4 1 124 8.60 11.1 3.6 16.0 1.9 2.2 2.397552 3.003228 2.725137 11.71901 2.0003 4.03 0 2990 Operational L 4 2 2 7.30 8.7 2.7 13.4 1 .0 1.3 1.773862 2.318415 1.602432 7.561728 1.2511 3.085 2990 Operational L4 2 7 6.60 7.5 3.3 10.5 0.9 1.6 0.835146 1.163054 1.128934 4.595382 0.9319 1.795 2990 Operational L4 2 10 7.20 9.4 3.2 14.0 0.9 2.3 1.364494 2.465949 1.918 021 7.986881 0.9411 3.2 00 2990 Operational L4 3 3 6.70 8.5 1.2 13.0 1.6 2.6 1.863535 2.598732 1.816049 5.918531 1.8901 2.455 2990 Operational L4 3 8 6.60 8.2 2.3 12.2 1.3 1.8 1.362917 1.757734 1.222061 5.475812 1.1872 1.935 2990 Operational L4 3 10 5.20 6.2 2.1 10 .0 1.2 1.3 0.700546 0.642764 0.711522 3.098149 0.8232 0.98 0

PAGE 140

140 Table D 7. Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, operational silviculture and family L7. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 2990 Operational L7 1 1 7.30 10.2 3.3 14.2 1.6 2.1 1.463778 1.885822 2.091422 8.969746 0.9572 2.98 0 2990 Operational L7 1 5 6.50 7 .0 2.6 10.8 1.1 1.3 1.1655 00 1.29108 0 1.1048 00 4.36 6182 1.0887 1.935 2990 Operational L7 1 11 7.90 10.1 3.4 14.5 1.3 1.8 2.081492 2.928984 2.279269 10.58874 2.5761 4.55 0 2990 Operational L7 2 7 7.00 9.3 3.4 12.6 1.4 1.5 1.221102 1.7017 00 1.719716 7.117963 1.4154 2.61 0 2990 Operational L7 2 10 5.60 7 .0 2 .2 11.5 0.8 1.5 0.848667 1.335285 1.119273 3.87303 0 0.8682 1.68 0 2990 Operational L7 2 12 6.70 10 3 .0 13.2 1.4 1.6 1.69439 0 2.316637 2.080382 7.018864 1.0027 2.955 2990 Operational L7 3 116 5.40 6.8 2.1 10.2 1.2 1.7 0.878688 0.88775 0 0.958651 2.858964 0. 8464 1.525 2990 Operational L7 3 120 4.30 4.7 1.7 7.7 1.1 0.8 0.50406 0 0.438053 0.561154 1.1095 00 0.3732 0.82 0 2990 Operational L7 3 125 5.70 7 .0 2.2 11.1 1.5 1.8 1.375891 1.405923 1.221622 3.507611 0.8432 1.735 2990 Operational L7 1 1 7.30 10.2 3.3 14. 2 1.6 2.1 1.463778 1.885822 2.091422 8.969746 0.9572 2.98 0 2990 Operational L7 1 5 6.50 7 .0 2.6 10.8 1.1 1.3 1.1655 00 1.29108 0 1.1048 00 4.366182 1.0887 1.935 2990 Operational L7 1 11 7.90 10.1 3.4 14.5 1.3 1.8 2.081492 2.928984 2.279269 10.58874 2.5761 4 .55 0 2990 Operational L7 2 7 7.00 9.3 3.4 12.6 1.4 1.5 1.221102 1.7017 00 1.719716 7.117963 1.4154 2.61 0 2990 Operational L7 2 10 5.60 7 .0 2.2 11.5 0.8 1.5 0.848667 1.335285 1.119273 3.87303 0 0.8682 1.68 0 2990 Operational L7 2 12 6.70 10.0 3 .0 13.2 1.4 1 .6 1.69439 0 2.316637 2.080382 7.018864 1.0027 2.955 2990 Operational L7 3 116 5.40 6.8 2.1 10.2 1.2 1.7 0.878688 0.88775 0 0.958651 2.858964 0.8464 1.525 2990 Operational L7 3 120 4.30 4.7 1.7 7.7 1.1 0.8 0.50406 0 0.438053 0.561154 1.1095 00 0.3732 0.82 0 2990 Operational L7 3 125 5.70 7 .0 2.2 11.1 1.5 1.8 1.375891 1.405923 1.221622 3.507611 0.8432 1.735

PAGE 141

141 Table D 8. Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, intensive silviculture and family L4. DENSITY CU LTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 2990 Intensive L4 1 2 9.70 14.7 3.1 21.7 1.9 3 .0 4.365611 6.154241 4.193221 22.68302 3.1143 8.185 2990 Intensive L4 1 10 10.20 12 .0 5 .0 17.8 1.6 2.3 2.27 3807 3.483918 2.725506 17.49223 1.7567 5.695 2990 Intensive L4 1 14 9.50 10.7 4.6 16.0 1.7 1.9 2.473326 3.329021 2.314326 14.56822 2.3031 4.075 2990 Intensive L4 2 4 9.10 12.5 4.5 17.5 0 .1 1.5 2.353397 4.592274 3.343303 15.32264 2.3967 4.95 0 2990 Intens ive L4 2 9 8.50 10.1 3 .0 14.1 1.5 1.9 2.348419 2.641727 2.092892 10.37911 1.6156 4.375 2990 Intensive L4 2 16 8.70 11.6 3.3 16.3 1.6 2 .0 2.805349 4.603774 2.295119 13.76613 2.1191 4.03 0 2990 Intensive L4 3 5 9.80 13.4 3.9 19.0 2.1 3.3 3.74935 0 5.8157 00 2 .729314 20.02824 3.244 0 8.37 0 2990 Intensive L4 3 8 9.10 10.6 3.9 14.7 1.7 2.8 2.751041 4.683094 1.975358 11.60802 1.6952 4.395 2990 Intensive L4 3 12 8.20 8.6 3.9 13 .0 1.3 2.2 1.647227 2.840481 1.25858 0 7.105862 1.8196 2.52 0 2990 Intensive L4 1 2 9.70 14.7 3.1 21.7 1.9 3 .0 4.365611 6.154241 4.193221 22.68302 3.1143 8.185 2990 Intensive L4 1 10 10.20 12.0 5 .0 17.8 1.6 2.3 2.273807 3.483918 2.725506 17.49223 1.7567 5.695 2990 Intensive L4 1 14 9.50 10.7 4.6 16.0 1.7 1.9 2.473326 3.329021 2.314326 14.56822 2.3031 4.075 2990 Intensive L4 2 4 9.10 12.5 4.5 17.5 0 .1 1.5 2.353397 4.592274 3.343303 15.32264 2.3967 4.95 0 2990 Intensive L4 2 9 8.50 10.1 3 .0 14.1 1.5 1.9 2.348419 2.641727 2.092892 10.37911 1.6156 4.375 2990 Intensive L4 2 16 8.70 11.6 3.3 16.3 1.6 2 .0 2.805349 4.603774 2.295119 13.76613 2.1191 4.03 0 2990 Intensive L4 3 5 9.80 13.4 3.9 19.0 2.1 3.3 3.74935 0 5.8157 00 2.729314 20.02824 3.244 0 8.37 0 2990 Intensive L4 3 8 9.10 10.6 3.9 14.7 1.7 2.8 2.751041 4.683094 1.975358 11.60802 1.6952 4.395 2990 Intensive L4 3 12 8.20 8.6 3.9 13 .0 1.3 2.2 1.647227 2.840481 1.25858 0 7.105862 1.8196 2.52 0

PAGE 142

142 Table D 9. Raw data utilized to develop allometric equations from the Sanderson, FL location at age 5, intensive silviculture and family L7. DENSITY CULT URE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 2990 Intensive L7 1 115 8.70 12.6 3.8 16.7 1.7 2 .0 2.225849 3.273179 3.076216 16.27494 2.552 0 6.95 0 2990 Intensive L7 1 121 7.90 11.8 2.4 16.3 1.6 1.7 2.10 8314 4.298014 2.827464 11.91885 2.0844 4.78 0 2990 Intensive L7 1 127 9.80 12.5 3.9 19.5 1.7 2.2 3.181229 4.871542 3.515428 20.98008 3.1436 8.545 2990 Intensive L7 2 115 8.60 12.2 3.7 16.2 1.2 1.9 3.328359 4.565272 3.140607 16.50768 2.3114 6.285 2990 Int ensive L7 2 119 7.60 10.6 2.4 14.5 1.5 2.3 2.138471 3.443857 2.147685 10.78626 1.9542 3.585 2990 Intensive L7 2 122 8.40 12.6 3 .0 17.1 0.8 2 .0 2.820264 4.374569 2.959585 13.57723 2.3371 5.39 0 2990 Intensive L7 3 118 6.90 9.4 3 .0 13.5 0.9 2.1 1.291142 2.4 51272 1.516088 6.812977 1.8101 3.41 0 2990 Intensive L7 3 120 7.70 12.6 2.9 17.1 1.8 1.7 3.051658 4.498341 2.743148 13.54697 2.5773 5.44 0 2990 Intensive L7 3 123 8.60 13.8 3.5 18.7 1.6 1.5 3.079408 4.314661 3.339809 18.57347 3.6666 6.77 0 2990 Intensive L 7 1 115 8.70 12.6 3.8 16.7 1.7 2 .0 2.225849 3.273179 3.076216 16.27494 2.552 0 6.95 0 2990 Intensive L7 1 121 7.90 11.8 2.4 16.3 1.6 1.7 2.108314 4.298014 2.827464 11.91885 2.0844 4.78 0 2990 Intensive L7 1 127 9.80 12.5 3.9 19.5 1.7 2.2 3.181229 4.871542 3 .515428 20.98008 3.1436 8.545 2990 Intensive L7 2 115 8.60 12.2 3.7 16.2 1.2 1.9 3.328359 4.565272 3.140607 16.50768 2.3114 6.285 2990 Intensive L7 2 119 7.60 10.6 2.4 14.5 1.5 2.3 2.138471 3.443857 2.147685 10.78626 1.9542 3.585 2990 Intensive L7 2 122 8.40 12.6 3 .0 17.1 0.8 2 .0 2.820264 4.374569 2.959585 13.57723 2.3371 5.39 0 2990 Intensive L7 3 118 6.90 9.4 3 .0 13.5 0.9 2.1 1.291142 2.451272 1.516088 6.812977 1.8101 3.41 0 2990 Intensive L7 3 120 7.70 12.6 2.9 17.1 1.8 1.7 3.051658 4.498341 2.743148 13.54697 2.5773 5.44 0 2990 Intensive L7 3 123 8.60 13.8 3.5 18.7 1.6 1.5 3.079408 4.314661 3.339809 18.57347 3.6666 6.77 0

PAGE 143

143 Table D 10. Raw data utilized to develop allometric equations from the Waverly, GA location at age 5, intensive silviculture and family L4. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 2990 Intensive L4 2 115 7.40 13.4 2.9 17.2 1.6 1.7 4.7108 00 5.812981 2.393065 13.12115 3.024 4.65 0 2990 Intensive L4 2 118 7.80 11.1 3.5 16.5 0.9 1.7 2.385796 2.601123 1.74223 0 12.12738 1.132 3.885 2990 Intensive L4 2 125 8.20 14.5 3.2 19.5 1.5 1.7 4.038553 4.951684 3.290382 17.85689 2.677 4.5 00 2990 Intensive L4 3 117 7.80 11.0 3.1 15.6 0.9 1.5 2.816326 3.501507 1.63923 0 10.32345 1. 164 3.63 0 2990 Intensive L4 3 120 7.80 11.5 3.2 15.5 1.2 1.7 3.644605 3.752476 1.910576 11.58636 2.873 4.19 0 2990 Intensive L4 3 125 9.30 13.1 3.6 18.8 1.3 1.8 4.65473 0 5.814757 2.607345 18.63877 3.62 0 5.985 2990 Intensive L4 4 102 8.70 12.0 3.2 14.5 1. 2 2.2 2.203734 3.052889 2.255072 10.13622 1.917 2.785 2990 Intensive L4 4 113 9.50 13.1 3.9 17.1 1.4 3 .0 3.91903 0 5.868425 2.980239 17.35341 2.705 4.055 2990 Intensive L4 4 116 8.80 13.1 3.5 16.7 1.6 1.9 4.641446 5.521206 3.479358 18.14702 1.862 3.975 2 990 Intensive L4 2 115 7.40 13.4 2.9 17.2 1.6 1.7 4.7108 00 5.812981 2.393065 13.12115 3.024 4.65 0 2990 Intensive L4 2 118 7.80 11.1 3.5 16.5 0.9 1.7 2.385796 2.601123 1.74223 0 12.12738 1.132 3.885 2990 Intensive L4 2 125 8.20 14.5 3.2 19.5 1.5 1.7 4.0385 53 4.951684 3.290382 17.85689 2.677 4.5 00 2990 Intensive L4 3 117 7.80 11 .0 3.1 15.6 0.9 1.5 2.816326 3.501507 1.63923 0 10.32345 1.164 3.63 0 2990 Intensive L4 3 120 7.80 11.5 3.2 15.5 1.2 1.7 3.644605 3.752476 1.910576 11.58636 2.873 4.19 0 2990 Intensiv e L4 3 125 9.30 13.1 3.6 18.8 1.3 1.8 4.65473 0 5.814757 2.607345 18.63877 3.62 0 5.985 2990 Intensive L4 4 102 8.70 12.0 3.2 14.5 1.2 2.2 2.203734 3.052889 2.255072 10.13622 1.917 2.785 2990 Intensive L4 4 113 9.50 13.1 3.9 17.1 1.4 3 .0 3.91903 0 5.868425 2.980239 17.35341 2.705 4.055 2990 Intensive L4 4 116 8.80 13.1 3.5 16.7 1.6 1.9 4.641446 5.521206 3.479358 18.14702 1.862 3.975

PAGE 144

144 Table D 11. Raw data utilized to develop allometric equations from the Waverly, GA location at age 5, intensive silviculture and family L7. DENSITY CULTURE F AM BLOCK TREE HT D BH HTLC G LD CWAL CWAC FOLIAGE BRANCH BARK BOLE CROOTS TAP 2990 Intensive L7 2 7 7.20 10.5 3.3 14.5 0.6 0.8 1.895508 1.987311 1.906627 8.523747 0.76 0 2.445 2990 Intensive L7 2 10 7.90 14 .0 2.7 19 .0 1.6 1.5 4.931632 5.993063 3.86407 0 17.62493 2.933 5.605 2990 Intensive L7 2 16 7.80 10.9 3.2 15.6 0.9 1.1 1.863997 2.03719 0 1.935102 9.755023 1.602 2.37 0 2990 Intensive L7 3 3 8.30 12.3 3.2 15.5 1.2 1.4 3.811197 4.731141 2.734752 14.83388 1 .523 4.175 2990 Intensive L7 3 7 6.20 8 .0 3.7 11.5 0.6 0.6 0.924774 1.010446 1.046071 4.56671 0 0.848 2.035 2990 Intensive L7 3 13 8.10 13.1 3.2 17.9 1 .0 1.3 3.585558 5.289763 2.885409 17.25368 3.422 5.1 00 2990 Intensive L7 4 6 7.70 12.8 3 .0 16.7 1 .0 1.4 3.062652 4.871665 2.897013 14.69069 2.96 0 3.965 2990 Intensive L7 4 9 8.00 15.2 2.9 22.1 1.6 1.6 5.274806 8.861452 4.01182 0 19.44468 3.864 5.41 0 2990 Intensive L7 4 13 7.00 11 .0 2.8 15 .0 1.1 1.5 2.711424 3.156577 2.576949 9.253309 1.677 3.155 2990 Inte nsive L7 2 7 7.20 10.5 3.3 14.5 0.6 0.8 1.895508 1.987311 1.906627 8.523747 0.76 0 2.445 2990 Intensive L7 2 10 7.90 14.0 2.7 19.0 1.6 1.5 4.931632 5.993063 3.86407 0 17.62493 2.933 5.605 2990 Intensive L7 2 16 7.80 10.9 3.2 15.6 0.9 1.1 1.863997 2.03719 0 1.935102 9.755023 1.602 2.37 0 2990 Intensive L7 3 3 8.30 12.3 3.2 15.5 1.2 1.4 3.811197 4.731141 2.734752 14.83388 1.523 4.175 2990 Intensive L7 3 7 6.20 8 .0 3.7 11.5 0.6 0.6 0.924774 1.010446 1.046071 4.56671 0 0.848 2.035 2990 Intensive L7 3 13 8.10 13.1 3.2 17.9 1 .0 1.3 3.585558 5.289763 2.885409 17.25368 3.422 5.1 00 2990 Intensive L7 4 6 7.70 12.8 3 .0 16.7 1 .0 1.4 3.062652 4.871665 2.897013 14.69069 2.96 0 3.965 2990 Intensive L7 4 9 8.00 15.2 2.9 22.1 1.6 1.6 5.274806 8.861452 4.01182 0 19.44468 3.86 4 5.41 0 2990 Intensive L7 4 13 7.00 11 .0 2.8 15 .0 1.1 1.5 2.711424 3.156577 2.576949 9.253309 1.677 3.155

PAGE 145

145 LIST OF REFERENCES Adegbidi,H.G., Jokela,E.J., Comerford,N.B., 2005. Factors influencing production efficiency of intensively managed loblolly pine plantations in a 1to 4year old chronosequence. Forest Ecology and Management 218, 245258. Adegbidi,H.G., Jokela,E.J., Comerford,N.B., Barros,N.F., 2002. Biomass development for intensively managed loblolly pine plantations growing on Spodosols in the s outheastern USA. Forest Ecology and Management 167, 91102. Ahl,D.E., Gower,S.T., Mackay,D.S., Burrows,S.N., Norman,J.M., Diak,G.R., 2004. Heterogeneity of light use efficiency in a northern Wisconsin forest: implications for modeling net primary production with remote sensing. Remote Sensing of Environment 93, 168178. Albaugh,T.J., Allen,H.L., Dougherty,P.M., Johnsen,K.H., 2004. Long term growth responses of loblolly pine to optimal nutrient and water resource availability. Forest Ecology and Management 1 92, 319. Albaugh,T.J., Allen,H.L., Dougherty,P.M., Kress,L.W., King,J.S., 1998. Leaf area and above and belowground growth responses of loblolly pine to nutrient and water additions. Forest Science 44, 317328. Albaugh,T.J., Allen,H.L., Fox,T.R., 2007. H istorical patterns of forest fertilization in the southeastern United States from 1969 to 2004. Southern Journal of Applied Forestry 31, 129137. Allen,C.B., Will,R.E., Jacobson,M.A., 2005a. Production efficiency and radiation use efficiency of four tree s pecies receiving irrigation and fertilization. Forest Science 51, 556569. Allen,C.B., Will,R.E., McGarvey,R.C., Coyle,D.R., Coleman,M.D., 2005b. Radiationuse efficiency and gas exchange responses to water and nutrient availability in irrigated and fertil ized stands of sweetgum and sycamore. Tree Physiology 25, 191200. Allen,H.L., Fox,T.R., Campbell,R.G., 2005c. What is ahead for intensive pine plantation silviculture in the south? Southern Journal of Applied Forestry 29, 6269. Ares,A., Fownes,J.H., 2001. Productivity, resource use, and competitive interactions of Fraxinus uhdei in Hawaii uplands. Canadian Journal of Forest Research 31, 132142. Atwood,R.A., White,T.L., Huber,D.A., 2002. Genetic parameters and gains for growth and wood properties in Flori da source loblolly pine in the southeastern United States. Canadian Journal of Forest Research 32, 10251038. Axelsson,E., Axelsson,B., 1986. Changes in carbon allocation patterns in spruce and pine trees following irrigation and fertilization. Tree Physio logy 2, 189204.

PAGE 146

146 Ballar,C.L., 1999. Keeping up with the neighbours: phytochrome sensing and other signalling mechanisms. Trends in Plant Science 4, 201. Balster,N.J., Marshall,J.D., 2000. Eight year responses of light interception, effective leaf area index, and stemwood production in fertilized stands of interior Douglas fir (Pseudotsuga menziesii var. glauca). Canadian Journal of Forest Research 30, 733743. Bancroft,T.A., Han,C.P., 1983. A note on pooling variances. Journal of the American Statistical A ssociation 78, 981983. Barron Gafford,G.A., Will,R.E., Burkes,E.C., Shiver,B., Teskey,R.O., 2003. Nutrient concentrations and contents, and their relation to stem growth, of intensively managed Pinus taeda and Pinus elliottii stands of different planting densities. Forest Science 49, 291300. Baskerville,G.L., 1972. Use of logarithmic regression in the estimation of plant biomass. Canadian Journal of Forest Research 2, 49 53. Binkley,D., Stape,J.L., Ryan,M.G., 2004. Thinking about efficiency of resource us e in forests. Forest Ecology and Management 193, 516. Bongarten,B.C., Teskey,R.O., 1987. Dry weight partitioning and its relationship to productivity in loblolly pine seedlings from 7 sources. Forest Science 33, 255 267. Borders,B.E., Bailey,R.L., 2001. L oblolly pine Pushing the limits of growth. Southern Journal of Applied Forestry 25, 6974. Bouvet,J.M., Vigneron,P., Saya,A., 2005. Phenotypic plasticity of growth trajectory and ontogenic allometry in response to density for Eucalyptus hybrid clones and families. Annals of Botany 96, 811 821. Bozivich,H., Bancroft,T.A., Hartley,H.O., 1956. Power of analysis of variance test procedures for certain incompletely specified models.1. Annals of Mathematical Statistics 27, 10171043. Bridgwater,F., Kubisiak,T., Byram,T., McKeand,S., 2005. Risk assessment with current deployment strategies for fusiform rustresistant loblolly and slash pines. Southern Journal of Applied Forestry 29, 8087. Britt,J.R., Zutter,B.R., Mitchell,R.J., Gjerstad,D.H., Dickson,J.F., 1990. Influence of herbaceous interference on growth and biomass partitioning in planted loblolly pine ( Pinus taeda). Weed Science 38, 497503. Burkes,E.C., Will,R.E., BarronGafford,G.A., Teskey,R.O., Shiver,B., 2003. Biomass partitioning and growth efficiency of intensively managed Pinus taeda and Pinus elliottii stands of different planting densities. Forest Science 49, 224234. Cannell,M.G.R., 1978. Improving per hectare forest productivity. In: Hollis,C.A., Squillace,A.E. (Eds.), Proc. Fifth North American Forest Biology Workshop. Gainesville, FL, pp. 120148.

PAGE 147

147 Cannell,M.G.R., 1989. Physiological basis of wood production: a review. Scandinavian Journal of Forest Research 4, 459490. Cannell,M.G.R., Milne,R., Sheppard,L.J., Unsworth,M.H., 1987. Radiation inter ception and productivity of willow. Journal of Applied Ecology 24, 261278. Cannell,M.G.R., Sheppard,L.J., Milne,R., 1988. Light use efficiency and woody biomass production of poplar and willow. Forestry 61, 125136. Chmura,D.J., Rahman,M.S., Tjoelker,M.G., 2007. Crown structure and biomass allocation patterns modulate aboveground productivity in young loblolly pine and slash pine. Forest Ecology and Management 243, 219230. Chmura,D.J., Tjoelker,M.G., 2008. Leaf traits in relation to crown development, lig ht interception and growth of elite families of loblolly and slash pine. Tree Physiology 28, 729742. Clark,D.A., Brown,S., Kicklighter,D.W., Chambers,J.Q., Thomlinson,J.R., Ni,J., 2001. Measuring net primary production in forests: Concepts and field methods. Ecological Applications 11, 356370. Cobb,W.R., Will,R.E., Daniels,R.F., Jacobson,M.A., 2008. Aboveground biomass and nitrogen in four short rotation woody crop species growing with different water and nutrient availabilities. Forest Ecology and Manage ment 255, 40324039. Colbert,S.R., Jokela,E.J., Neary,D.G., 1990. Effects of annual fertilization and sustained weedcontrol on dry matter partitioning, leaf area, and growth efficiency of juvenile loblolly and slash pine. Forest Science 36, 9951014. Come rford,N.B., Cropper,W.P., Li,H., Smethurst,P.J., Van Rees,K.C.J., Jokela,E.J., Adegbidi,H., Barros,N.F., 2006. Soil supply and nutrient demand (SSAND): A general nutrient uptake model and an example of its application to forest management. Canadian Journal of Soil Science 86, 665673. Coyle,D.R., Coleman,M.D., 2005. Forest production responses to irrigation and fertilization are not explained by shifts in allocation. Forest Ecology and Management 208, 137152. Crawford,D.T., Lockaby,B.G., Somers,G.L., 1991. Genotype nutrition interactions in field planted loblolly pine. Canadian Journal of Forest Research 21, 15231532. Crow,J.R., 1988. A guide to using regression equations for estimating tree biomass. Northern Jourrnal of Applied Forestry 5, 1522. Dalla Te a,F., Jokela,E.J., 1991. Needlefall, canopy light interception, and productivity of young intensively managed slash and loblolly pine stands. Forest Science 37, 12981313. de Castro,F., Fetcher,N., 1999. The effect of leaf clustering in the interception of light in vegetal canopies: theoretical considerations. Ecological Modelling 116, 125134.

PAGE 148

148 Dean,C.A., 2007. Genotype and population performances and their interactions for growth of coastal Douglas fir in Western Washington. Forest Science 53, 463472. Dean,T.J., 2001. Potential effect of stand structure on belowground allocation. Forest Science 47, 6976. Dewar,R.C., Medlyn,B.E., McMurtrie,R.E., 1998. A mechanistic analysis of light and carbon use efficiencies. Plant Cell and Environment 21, 573588. Dough erty,P.M., Hennessey,T.C., Zarnoch,S.J., Stenberg,P.T., Holeman,R.T., Wittwer,R.F., 1995. Effects of stand development and weather on monthly leaf biomass dynamics of a loblolly pine ( Pinus taeda L.) stand. Forest Ecology and Management 72, 213227. du Toi t,B., 2008. Effects of site management on growth, biomass partitioning and light use efficiency in a young stand of Eucalyptus grandis in South Africa. Forest Ecology and Management 255, 23242336. du Toit,B., Dovey,S.B., 2005. Effect of site management on leaf area, early biomass development, and stand growth efficiency of a Eucalyptus grandis plantation in South Africa. Canadian Journal of Forest Research 35, 891900. Duursma,R.A., Makela,A., 2007. Summary models for light interception and light use efficiency of nonhomogeneous canopies. Tree Physiology 27, 859870. Duursma,R.A., Makela,A., Reid,D.E.B., Jokela,E.J., Porte,A.J., Roberts,S.D., 2010. Self shading affects allometric scaling in trees Functional Ecology 24, 723730. Duzan,H.W.JR., Williams, C.G., 1988. Matching loblolly pine families to regeneration sites Southern Journal of Applied Forestry 12, 166169. Emhart,V.I., Martin,T.A., White,T.L., Huber,D.A., 2007. Clonal variation in crown structure, absorbed photosynthetically active radiation and growth of loblolly pine and slash pine. Tree Physiology 27, 421430. Esprey,L.J., Sands,P.J., Smith,C.W., 2004. Understanding 3PG using a sensitivity analysis. Forest Ecology and Management 193, 235250. Falster,D.S., Westoby,M., 2003. Leaf size and angl e vary widely across species: what consequences for light interception? New Phytologist 158, 509525. Fang,S.Z., Xu,X.Z., Lu,S.X., Tang,L.Z., 1999. Growth dynamics and biomass production in short rotation poplar plantations: 6year results for three clones at four spacings. Biomass & Bioenergy 17, 415425. Fox,T.R., 2000. Sustained productivity in intensively managed forest plantations. Forest Ecology and Management 138, 187202.

PAGE 149

149 Fox,T.R., Jokela,E.J., Allen,H.L., 2007. The development of pine plantation si lviculture in the southern United States. Journal of Forestry 105, 337347. Gholz,H.L., Fisher,R.F., 1982. Organic matter production and distribution in slash pine ( Pinus elliottii) plantations. Ecology 63, 18271839. Gholz,H.L., Vogel,S.A., Cropper,W.P., McKelvey,K., Ewel,K.C., Teskey,R.O., Curran,P.J., 1991. Dynamics of canopy structure and light interception in Pinus elliottii stands, North Florida. Ecological Monographs 61, 3351. Gower,S.T., Haynes,B.E., Fassnacht,K.S., Running,S.W., Hunt,E.R., 1993. I nfluence of fertilization on the allometric relations for 2 pines in contrasting environments. Canadian Journal of Forest Research 23, 17041711. Gower,S.T., Kucharik,C.J., Norman,J.M., 1999. Direct and indirect estimation of leaf area index, f (APAR), and net primary production of terrestrial ecosystems. Remote Sensing of Environment 70, 2951. Gower,S.T., Vogt,K.A., Grier,C.C., 1992. Carbon dynamics of Rocky Mountain Douglas fir influence of water and nutrient availability. Ecological Monographs 62, 4365. Grace,J.C., 1987. Theoretical ratio between "one sided" and total surface area for pine needles. New Zealand Journal of Forest Science 17, 292296. Green,C.F., 1989. Genotypic differences in the growth of triticum aestivum in relation to absorbed solar radiation. Field Crops Research 19, 285295. Green,D.S., Kruger,E.L., Stanosz,G.R., Isebrands,J.G., 2001. Light use efficiency of native and hybrid poplar genotypes at high levels of intracanopy competition. Canadian Journal of Forest Research 31, 1030 1037. Griffin,K.L., Winner,W.E., Strain,B.R., 1995. Growth and dry matter partitioning in loblolly and ponderosa pine seedlings in response to carbon and nitrogen availability. New Phytologist 129, 547556. Harrington,R.A., Fownes,J.H., 1995. Radiation inter ception and growth of planted and coppice stands of 4 fast growing tropical trees. Journal of Applied Ecology 32, 1 8. Hayashi,K., Ito,H., 1962. Studies on the form of plant in rice varieties. Crop Science Society Japan 37, 329333. Haynes,B.E., Gower,S.T. 1995. Belowground carbon allocation in unfertilized and fertilized red pine plantations in northern Wisconsin. Tree Physiology 15, 317325. Hilker,T., Coops,N.C., Schwalm,C.R., Jassal,R.S., Black,T.A., Krishnan,P., 2008. Effects of mutual shading of tree crowns on prediction of photosynthetic light use efficiency in a coastal Douglas fir forest. Tree Physiology 28, 825834.

PAGE 150

150 Hollinger,D.Y., 1989. Canopy organization and foliage photosynthetic capacity in a broadleaved evergreen montane forest. Functional Ecology 3, 5362. Huang,C.H., Kronrad,G.D., 2004. Economic analysis of pruning and low density management compared to traditional management of loblolly pine plantations in east Texas. Southern Journal of Applied Forestry 28, 1220. Jarvis,P.G., Leverenz,J .W., 1983. Productivity of temperate, deciduous and evergreen forests. In: Lange,O.L., Nobel,P.S., Osmond,C.B., Ziegler,H. (Eds.), Physiological Plant Ecology IV. Encyclopedia of Plant Physiology, New Series, Vol. 12D. Springer Verlag, Berlin, pp. 234280. Johnson,J.D., 1984. A rapid technique for estimating total surface area of pine needles. Forest Science 30, 913921. Johnson,J.D., 1990. Dry matter partitioning in loblolly and slash pine effects of fertilization and irrigation. Forest Ecology and Manag ement 30, 147157. Jokela,E.J., Dougherty,P.M., Martin,T.A., 2004. Production dynamics of intensively managed loblolly pine stands in the southern United States: a synthesis of seven long term experiments. Forest Ecology and Management 192, 117130. Jokela ,E.J., Harding,R.B., Nowak,C.A., 1989. Long term effects of fertilization on stem form, growth relations, and yield estimates of slash pine. Forest Science 35, 832842. Jokela,E.J., Martin,T.A., 2000. Effects of ontogeny and soil nutrient supply on product ion, allocation, and leaf area efficiency in loblolly and slash pine stands. Canadian Journal of Forest Research 30, 1511 1524. Jokela,E.J., Wilson,D.S., Allen,J.E., 2000. Early growth responses of slash and loblolly pine following fertilization and herbac eous weed control treatments at establishment. Southern Journal of Applied Forestry 24, 2330. Jose,S., Merritt,S., Ramsey,C.L., 2003. Growth, nutrition, photosynthesis and transpiration responses of longleaf pine seedlings to light, water and nitrogen. Forest Ecology and Management 180, 335344. Keyes,M.R., Grier,C.C., 1981. Above ground and below ground net production in 40year old douglas fir stands on low and high productivity sites. Canadian Journal of Forest Research 11, 599605. Kinerson,R.S., Higgi nbotham,K.O., Chapman,R.C., 1974. Dynamics of foliage distribution within a forest canopy. Journal of Applied Ecology 11, 347353. King,J.S., Giardina,C.P., Pregitzer,K.S., Friend,A.L., 2007. Biomass partitioning in red pine ( Pinus resinosa) along a chronosequence in the Upper Peninsula of Michigan. Canadian Journal of Forest Research 37, 93102.

PAGE 151

151 Kira,T., Shinozak,K., Hozumi,K., 1969. Structure of forest canopies as related to their primary productivity. Plant and Cell Physiology 10, 129142. Knight,R., 1970. The measurement and interpretation of genotype environment interactions. Euphytica 19, 225 235. Lambeth,C., McKeand,S., Rousseau,R., Schmidtling,R., 2005. Planting nonlocal seed sources of loblolly pine Managing benefits and risks. Southern Journal of Applied Forestry 29, 96104. Land,S.B., Roberts,S.D., Duzan,H.W.JR., 2004. Genetic and spacing effects on loblolly pine plantation development through age 17. In: Connor,K.F. (Ed.)., Proceedings of the 12th biennial southern silvicultural research confer ence. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern Research Station. 594 p., pp. 413419. Landsberg,J.J., Hingston,F.J., 1996. Evaluating a simple radiation/dry matter conversion model using data from Eucalyptus globulus plantations in Western Australia. Tree Physiology 16, 801808. Landsberg,J.J., Waring,R.H., 1997. A generalised model of forest productivity using simplified concepts of radiationuse efficiency, carbon balance and partitioning. Forest Ecology and Management 95, 209228. Landsberg,J.J., Wright,L.L., 1989. Comparisons among populus clones and intensive culture conditions, using an energy conversion model. Forest Ecology and Management 27, 129147. Ledig,F.T., Bormann,F.H., Wenger,K.F., 1970. The distribution of dry matter growth between shoot and roots in loblolly pine. Botanical Gazette 131, 349359. Lee,H., Slatton,K.C., Roth,B.E., Cropper,W.P., 2009. Prediction of forest canopy light interception using three dimensional airborne LiDAR data. International Journal of Remote Sensing 30, 189207. Leverenz,J.W., Hinckley,T.M., 1990. Shoot structure, leaf area index and productivity of evergreen conifer stands. Tree Physiology 6, 135 149. Li,B., McKeand,S.E., 1989. Stability of loblolly pine families in the southeastern United States. Silvae Genetica 38, 96 101. Li,B., McKeand,S.E., Allen,H.L., 1991a. Genetic variation in nitrogen use efficiency of loblolly pine seedlings. Forest Science 37, 613626. Li,B.L., Allen,H.L., McKeand,S.E., 1991b. Nitrogen and family effects on biomass allocation of loblolly pine seedlings. Forest Science 37, 271 283. Li,J.H., Powell,T.L., Seiler,T.J., Johnson,D.P., Anderson,H.P., Bracho,R., Hungate,B.A., Hinkle,C.R., Drake,B.G., 2007. Impacts of Hurricane Frances on Florida scruboak

PAGE 152

152 ecosystem processes: defoliation, net CO2 exchange and interactions with elevated CO2. Global Change Biology 13, 11011113. Littel,R.C., Milliken,G.A., Stroup,W.W., Wolfinger,R.D., 1996. SAS system for mixed models. SAS Institute Inc., Cary, N.C. Litton,C.M., Ryan,M .G., Tinker,D.B., Knight,D.H., 2003. Belowground and aboveground biomass in young postfire lodgepole pine forests of contrasting tree density. Canadian Journal of Forest Research 33, 351363. Long,J.N., Dean,T.J., Roberts,S.D., 2004. Linkages between silvi culture and ecology: examination of several important conceptual models. Forest Ecology and Management 200, 249261. Martin,S.W., Shiver,B.D., 2002. Impacts of vegetation control, genetic improvement and their interaction on loblolly pine growth in the Southern United States age 12 results. Southern Journal of Applied Forestry 26, 3742. Martin,T.A., Johnsen,K.H., White,T.L., 2001. Ideotype development in southern pines: Rationale and strategies for overcoming scale related obstacles. Forest Science 47, 2 1 28. Martin,T.A., Jokela,E.J., 2004a. Developmental patterns and nutrition impact radiation use efficiency components in southern pine stands. Ecological Applications 14, 18391854. Martin,T.A., Jokela,E.J., 2004b. Stand development and production dynamic s of loblolly pine under a range of cultural treatments in northcentral Florida USA. Forest Ecology and Management 192, 3958. McCrady,R.L., Jokela,E.J., 1996. Growth phenology and crown structure of selected loblolly pine families planted at two spacings Forest Science 42, 4657. McCrady,R.L., Jokela,E.J., 1998. Canopy dynamics, light interception, and radiation use efficiency of selected loblolly pine families. Forest Science 44, 64 72. McKeand,S., Mullin,T., Byram,T., White,T., 2003a. Deployment of gen etically improved loblolly and slash pines in the south. Journal of Forestry 101, 3237. McKeand,S.E., Amerson,H.V., Li,B., Mullin,T.J., 2003b. Families of loblolly pine that are the most stable for resistance to fusiform rust are the least predictable. Ca nadian Journal of Forest Research 33, 1335 1339. McKeand,S.E., Crook,R.P., Allen,H.L., 1997a. Genotypic stability effects on predicted family responses to silvicultural treatments in loblolly pine. Southern Journal of Applied Forestry 21, 8489. McKeand,S. E., Eriksson,G., Roberds,J.H., 1997b. Genotype by environment interaction for index traits that combine growth and wood density in loblolly pine. Theoretical and Applied Genetics 94, 10151022.

PAGE 153

153 McKeand,S.E., Li,B., Hatcher,A.V., Weir,R.J., 1990. Stability parameter estimates for stem volume for loblolly pine families growing in different regions in the Southeastern United States. Forest Science 36, 10 17. McKeand,S.E., Jokela,E.J., Huber,D.A., Byram,T.D., Allen,H.L., Li,B., Mullin,T.J., 2006. Performance of improved genotypes of loblolly pine across different soils, climates, and silvicultural inputs. Forest Ecology and Management 227, 178184. Medlyn,B.E., 1998. Physiological basis of the light use efficiency model. Tree Physiology 18, 167176. Miller,J.H., Zutter,B.R., Zedaker,S.M., Edwards,M.B., Newbold,R.A., 2003. Growth and yield relative to competition for loblolly pine plantations to midrotation A southeastern United States regional study. Southern Journal of Applied Forestry 27, 237 252. Monteith,J.L., 1977. Climate and efficiency of crop production in Britain. Philosophical Transactions of the Royal Society of London Series B Biological Sciences 281, 277294. Neary,D.G., Rockwood,D.L., Comerford,N.B., Swindel,B.F., Cooksey,T.E., 1990. Importance of weed control, fertilization, irrigation, and genetics in slash and loblolly pine early growth on poorly drained Spodosols. Forest Ecology and Management 30, 271281. Nilsson,U., Allen,H.L., 2003. Short and longterm effects of site preparation, fertilization and vegetation control on growth and stand development of planted loblolly pine. Forest Ecology and Management 175, 367377. NOAA, 2002. State, Regional, and National Monthly Precipitation: Area Weighted Monthly Normals, 19712000 (and previous periods ) Historical Climatography Series 42. Okerblom,P., Kellomaki,S., 1983. Effect of grouping of foliage on the withinstand and withincrown light regime Comparison of random and grouping canopy models. Agricultural Meteorology 28, 143155. Pangle,L., Vos e,J.M., Teskey,R.O., 2009. Radiation use efficiency in adjacent hardwood and pine forests in the southern Appalachians. Forest Ecology and Management 257, 10341042. Posada,J.M., 2003. Optimization of light use by tropical canopy trees through acclimation of leaf physiology and orientation. PhD Dissertation 115 pp. Prince,S.D., Goward,S.N., 1995. Global primary production: A remote sensing approach. Journal of Biogeography 22, 815 835. Pritchett,W.L., Comerford,N.B., 1982. Long term response to phosphorus f ertilization on selected southeastern Coastal plain soils. Soil Science Society of America Journal 46, 640644.

PAGE 154

154 Retzlaff,W.A., Handest,J.A., O'Malley,D.M., McKeand,S.E., Topa,M.A., 2001. Whole tree biomass and carbon allocation of juvenile trees of lobloll y pine ( Pinus taeda): influence of genetics and fertilization. Canadian Journal of Forest Research 31, 960970. Ritchie,G.A., 1997. Evidence for red:far red signaling and photomorphogenic growth response in Douglas fir (Pseudotsuga menziesii ) seedlings. Tr ee Physiology 17, 161168. Roth,B.E., Jokela,E.J., Martin,T.A., Huber,D.A., White,T.L., 2007a. Genotype x environment interactions in selected loblolly and slash pine plantations in the southeastern United States. Forest Ecology and Management 238, 175188. Roth,B.E., Li,X., Huber,D.A., Peter,G.F., 2007b. Effects of management intensity, genetics and planting density on wood stiffness in a plantation of juvenile loblolly pine in the southeastern USA. Forest Ecology and Management 246, 155162. Roth,B.E., Ma rtin,T.A., Jokela,E.J., White,T.L., 2002. Finding the keys to unlock the productivity of southern forests. Florida Forests 6, 1820. Roth,B.E., Newton,M., 1996. Survival and growth of Douglas fir relating to weeding, fertilization, and seed source. Western Journal of Applied Research 11, 6269. Roth,B.E., Slatton,K.C., Cohen,M.J., 2007c. On the potential for highresolution lidar to imp rove rainfall interception estimates in forest ecosystems. Frontiers in Ecology and the Environment 5, 421428. Runyon,J., Waring,R.H., Goward,S.N., Welles,J.M., 1994. Environmental limits on net primary production and light use efficiency across the Orego n transect. Ecological Applications 4, 226237. Sampson,D.A., Allen,H.L., 1998. Light attenuation in a 14year old loblolly pine stand as influenced by fertilization and irrigation. Trees Structure and Function 13, 8087. Sampson,D.A., Smith,F.W., 1993. Influence of canopy architecture on light penetration in lodgepole pine ( Pinus contorta var latifolia) forests. Agricultural and Forest Meteorology 64, 6379. Samuelson,L.J., 2000. Effects of nitrogen on leaf physiology and growth of different families of lo blolly and slash pine. New Forests 19, 95107. Samuelson,L.J., Johnsen,K., Stokes,T., 2004. Production, allocation, and stemwood growth efficiency of Pinus taeda L. stands in response to 6 years of intensive management. Forest Ecology and Management 192, 5970. Sands,P.J., 1996. Modelling canopy production .3. Canopy light utilisation efficiency and its sensitivity to physiological and environmental variables. Australian Journal of Plant Physiology 23, 103114.

PAGE 155

155 Sayer,M.A.S., Goelz,J.C.G., Chambers,J.L., Tan g,Z., Dean,T.J., Haywood,J.D., Leduc,D.J., 2004. Long term trends in loblolly pine productivity and stand characteristics in response to thinning and fertilization in the West Gulf region. Forest Ecology and Management 192, 7196. Schmidt,R.A., Allen,J.E., 1998. Spacial stability of fusiform rust resistance in slash pine in the coastal plain of the southeastern USA. In: Finnish For. Res. Inst., Saariselka, Finland, pp. 219229. Schmidtling,R.C., 2001. Southern pine seed sources. In: USDA, Forest Service, Southern Research Station. Scott,D.A., Dean,T.J., 2006. Energy trade offs between intensive biomass utilization, site productivity loss, and ameliorative treatments in loblolly pine plantations. Biomass & Bioenergy 30, 10011010. Scott,W., Meade,R., Leon,R., Hyink,D., Miller,R., 1998. Planting density and tree size relations in coast Douglas fir. Canadian Journal of Forest Research 28, 7478. Siddique,K.H.M., Belford,R.K., Perry,M.W., Tennant,D., 1989. Growth, development and light interception of old and modern wheat cultivars in a mediterranean type environment. Australian Journal of Agricultural Research 40, 473487. Sierra Lucero,V., McKeand,S.E., Huber,D.A., Rockwood,D.L., White,T.L., 2002. Performance differences and genetic parameters for four coastal p rovenances of loblolly pine in the southeastern United States. Forest Science 48, 732 742. Sinclair,T.R., Knoerr,K.R., 1982. Distribution of photosynthetically active radiation in the canopy of a loblolly pine plantation. Journal of Applied Ecology 19, 183 191. Sinclair,T.R., Muchow,R.C., 1999. Radiation use efficiency. Advances in Agronomy, Vol 65 65, 215265. Smith,F.W., Long,J.N., 1989. The influence of canopy architecture on stemwood production and growth efficiency of Pinus contorta var. latifolia Journal of Applied Ecology 26, 681691. Smith,F.W., Long,J.N., 2001. Age related decline in forest growth: an emergent property. Forest Ecology and Management 144, 175181. Smith,F.W., Sampson,D.A., Long,J.N., 1991. Comparison of leaf area index estimates fro m tree allometrics and measured light interception. Forest Science 37, 16821688. Squillace,A.E., 1989. Tree improvement accomplishments in the South. In: Clemson Univ., Clemson, SC, June 2630, 1989. 463 p., pp. 920. St Clair,J.B., 1994. Genetic variatio n in tree structure and its relation to size in Douglas fir. 1.Biomass partitioning, foliage efficiency, stem form, and wood density. Canadian Journal of Forest Research 24, 12261235.

PAGE 156

156 Stape,J.L., 2002. Production ecology of clonal Eucalyptus plantations i n northeastern Brazil. PhD Dissertation, Colorado State University, Ft. Collins, USA. Stape,J.L., Binkley,D., Ryan,M.G., 2004. Eucalyptus production and the supply, use and efficiency of use of water, light and nitrogen across a geographic gradient in Braz il. Forest Ecology and Management 193, 1731. Staudhammer,C.L., Jokela,E.J., Martin,T.A., 2009. Competition dynamics in pure versus mixed family stands of loblolly and slash pine in the southeastern United States. Canadian Journal of Forest Research 39, 396409. Stenberg,P., Linder,S., Smolander,H., Flowerelles,J., 1994. Performance of the lai 2000 plant canopy analyzer in estimating leaf area index of some scots pine stands. Tree Physiology 14, 981995. Sutton,B., 2002. Commercial delivery of genetic impr ovement to conifer plantations using somatic embryogenesis. Annals of Forest Science 59, 657661. Terashima,I., Hikosaka,K., 1995. Comparative ecophysiology of leaf and canopy photosynthesis. Plant Cell and Environment 18, 11111128. Tyree,M.C., Seiler,J.R ., Maier,C.A., 2009. Short term impacts of nutrient manipulations on leaf gas exchange and biomass partitioning in contrasting 2year old Pinus taeda clones during seedling establishment. Forest Ecology and Management 257, 18471858. Vanlear,D.H., Kapeluck ,P.R., 1995. Above and below stump biomass and nutrient content of a mature loblolly pine plantation. Canadian Journal of Forest Research 25, 361367. Vergara,R., White,T.L., Huber,D.A., Shiver,B.D., Rockwood,D.L., 2004. Estimated realized gains for first generation slash pine ( Pinus elliottii var. elliottii) tree improvement in the southeastern United States. Canadian Journal of Forest Research 34, 25872600. Wagner,R.G., Little,K.M., Richardson,B., McNabb,K., 2006. The role of vegetation management for en hancing productivity of the world's forests. Forestry 79, 5779. Wang,Y.P., Jarvis,P.G., 1990. Influence of crown structural properties on PAR absorption, photosynthesis and transpiration in Sitka spruce: application of a model (Maestro). Tree Physiology 2 97316. Wang,Y.P., Jarvis,P.G., Taylor,C.M.A., 1991. Par absorption and its relation to aboveground dry matter production of Sitka spruce. Journal of Applied Ecology 28, 547 560. Waring,R.H., 1983. Estimating forest growth and efficiency in relation to canopy leaf area. Advances in Ecological Research 13, 327354. Waring,R.H., Running,S.W., 1998. Forest Ecosystems: Analysis at Multiple Scales. Academic Press, New York.

PAGE 157

157 White,T.L., Adams,W.T., Neale,D.B., 2007. Forest Genetics. CABI, Wallingford, Oxfordshire UK. Will,R.E., Barron,G.A., Burkes,E.C., Shiver,B., Teskey,R.O., 2001. Relationship between intercepted radiation, net photosynthesis, respiration, and rate of stem volume growth of Pinus taeda and Pinus elliottii stands of different densities. Forest Ec ology and Management 154, 155163. Will,R.E., Narahari,N.V., Shiver,B.D., Teskey,R.O., 2005. Effects of planting density on canopy dynamics and stem growth for intensively managed loblolly pine stands. Forest Ecology and Management 205, 2941. Woodruff,D.R ., Bond,B.J., Ritchie,G.A., Scott,W., 2002. Effects of stand density on the growth of young Douglas fir trees. Canadian Journal of Forest Research 32, 420427. Wu,R., Grissom,J., McKeand,S., O'Malley,D., 2004. Phenotypic plasticity of fine root growth incr eases plant productivity in pine seedlings. BMC Ecology 4, 14. Xiao,Y., Jokela,E.J., White,T.L., 2003. Species differences in crown structure and growth performance of juvenile loblolly and slash pine. Forest Ecology and Management 174, 295313. Yeiser,J.L ., Lowe,W., Van Buijtenen,J.P., 2001. Stability and seed movement for loblolly pine in the Western Gulf Region. Silvae Genetica 50, 81 88. Zas,R., Merlo,E., Fernandez Lopez,J., 2004. Genotype x environment interaction in maritime pine families in Galicia, northwest Spain. Silvae Genetica 53, 175 182. Zhang,Y.J., Borders,B.E., 2004. Using a system mixedeffects modeling method to estimate tree compartment biomass for intensively managed loblolly pines an allometric approach. Forest Ecology and Management 194, 145157. Zhao,D., Kane,M., Borders,B., Harfison,M., 2008. Pine growth response to different site preparation methods with or without post plant herbaceous weed control on North Florida's Lower Coastal Plain. Forest Ecology and Management 255, 25122523.

PAGE 158

158 BIOGRAPHICAL SKETCH Brian Edward Roth was born in Edmonton, Alberta, Canada in 1969 and learned to appreciate nature at an early age. He joined Junior Forest Wardens, a forestry youth program similar to 4 H, when he was nine, which fostered his inte rest in forestry and outdoor activities. Brian enrolled at the University of Alberta in 1987 in the newly founded Biotechnology program, but transferred to the Forestry program a year later. While an undergrad, he took every advantage of summer work oppor tunities to gain a variety of experience in his chosen field. This included a summer as a helicopter rappel forest firefighter and cruising timber for the province of Alberta. He also spent a summer in northern Sweden working for the Swedish University of Agricultural Sciences on a National Forest Inventory. Brian left Alberta for Oregon State University in Corvallis in the Spring of 1991, where he studied forest regeneration and forest herbicides. He purchased and rehabilitated 6.5 acres of degraded for estland in the Coastal Mountain Range and eventually earned a M .S. in forest science in 1994. Over the following several years, Brian worked as a field forester in an oldgrowth logging camp on Vancouver Island for MacMillan Bloedel, Inc. and in Seattle, Washington as a research forester for Weyerhaeuser Company. In 2000, Brian left Washington State for Gainesville, Florida to work at the School of Forest Resources and Conservation. While employed as the Program Manager for the Forest Biology Research Coo perative and later as an independent forestry consultant, Brian labored toward the lofty goal of a Doctorate in the area of forest production ecology.