Citation
Role of Uneven-Aged Silviculture and the Soil Seed Bank in Restoration of Longleaf Pine-Slash Pine (Pinus palustris-Pinus elliottii) Ecosystems

Material Information

Title:
Role of Uneven-Aged Silviculture and the Soil Seed Bank in Restoration of Longleaf Pine-Slash Pine (Pinus palustris-Pinus elliottii) Ecosystems
Creator:
Sharma, Ajay
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (156 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Forest Resources and Conservation
Committee Chair:
Jose, Shibu
Committee Co-Chair:
Bohn, Kimberly Kirsten
Committee Members:
Andreu, Michael Gardner
Cropper, Wendell P
Miller, Deborah L
Graduation Date:
8/11/2012

Subjects

Subjects / Keywords:
Ecosystems ( jstor )
Forests ( jstor )
Overstory ( jstor )
Plantations ( jstor )
Seeds ( jstor )
Simulations ( jstor )
Species ( jstor )
State forests ( jstor )
Understory ( jstor )
Vegetation ( jstor )
Forest Resources and Conservation -- Dissertations, Academic -- UF
conversion -- light -- restoration -- seedbank -- thinning -- understory -- uneven
Blackwater River State Forest ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Forest Resources and Conservation thesis, Ph.D.

Notes

Abstract:
Concern over the loss of longleaf pine-slash pine (Pinus palustris-Pinus elliottii) ecosystems and the associated high level of biodiversity in the southeastern United States has led to an increased interest in their restoration and development of best management practices. This research, conducted using both simulation analysis and field data from multiple sites in north-west and north-central Florida, addressed three important aspects related to their restoration and management: 1) evaluating various silvicultural regimes to restore overstory structure and meet multiple objectives, 2) quantifying dynamics of the biophysical environment, specifically light levels with various silvicultural treatments and species composition, and 3) examining soil seed banks for their potential to restore understory biodiversity in degraded stands. Simulation analysis of 49 scenarios of different silvicultural regimes identified possible management options to restore degraded mature slash pine plantations to multifunctional uneven-aged stands which could maximize the provision of multiple benefits including timber production, structural diversity, and carbon storage. Appropriate options include converting slash pine plantations using the irregular shelterwood method with basal area of 4.6 square meter per hectare on a cutting cycle of 20 years and regeneration of 741 or more seedlings/hectare. For management at higher residual basal area of 11.5 square meter per hectare, several different silvicultural regimes maximized the provision of multiple benefits but more so with high levels of regeneration following harvests. Among the various uneven-aged silvicultural methods evaluated in field trials, group selection method resulted in the creation of understory light conditions desirable for both tree regeneration and understory biodiversity. The understory light availability created by group selection method was comparable to that of shelterwood method but had higher variability at the stand level. Light availability also varied with overstory composition, with higher understory light availability in pure longleaf pine stands than those of slash pine. Examination of seed bank in degraded, partially restored, and restored stands in pine flatwoods sites found a total of 26, 39, and 64 species respectively, mainly consisting of sedges and rushes. Unlike other stands, seed density in degraded stands was higher at a depth of 5-10 cm in soil. ( en )
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.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Jose, Shibu.
Local:
Co-adviser: Bohn, Kimberly Kirsten.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-08-31
Statement of Responsibility:
by Ajay Sharma.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Sharma, Ajay. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2014
Classification:
LD1780 2012 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

1 ROLE OF UNEVEN AGED SILVICULTURE AND THE SOIL SEED BANK IN RESTORATION OF LONGLEAF PINE SLASH PINE ( P inus palustris P inus elliottii ) ECOSYSTEMS By AJAY SHARMA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSI TY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Ajay Sharma

PAGE 3

3 To my teachers friends and family

PAGE 4

4 ACKNOWLEDGMENTS The work was possible because of support, c ooperation, and efforts of many people. I thank Dr s Shibu Jose and Kimberly Bohn, the chairs of my supervisory committee for the opportunity mentorship support, encouragement, time, and concern during the study. I thank my committee members Dr s Michae l Andreu, Wendell Cropper, and Debbie Miller, for their expert guidance and inputs during the study. Dr. Andreu is also thanked for providing me extra opportunities for teaching and research. The study would not have been possible but for the financial and logistical support of the Cooperative for Conserved Forest ecosystems: Outreach and Research (CFEOR), and McIntire Stennis Funding. Kent Perkins, Barry Davis and Susan Carr at the Florida Museum of Natural History Herbarium patiently helped with timely id entification of plant specimens. Drs. Dale Brockway and Ken neth Outcalt of USDA Forest Service are acknowledged for the opportunity to utilize the trials at the Goethe State Forest and the Blackwater River State Forest and extending all cooperation. Mr. Da vid Morse, Florida Forest Service is acknowledged for his support and inputs for the Special thanks to Dr. Tim Martin for providing some of the equipment for the study. Willie Wood, Paul Proctor, and Matt Pollard are gratefully acknowledged for all their help. I had great lab mates, friends, and field crew which made my life and work very enjoyable. I especially thank Puneet Dwivedi, Melissa Kreye, Michael Morgan, John Roberts, Joseph Culen, Don Hagan, Lyn n Proenza Nilesh Timilsina and other lab members /friends for all their support and well wishes.

PAGE 5

5 My wife Pritika Sharma supported throughout the study. Our son Arul Sharma was born during the study in March 2010. He has filled our world with joy and is a s ource of constant inspiration. I feel blessed to have studied in the School of Forest Resources and Conservation at the University of Florida. Coming to the United States of America and studying here have been one of the nicest experiences of my life. Than ks to all the people here in the School of Forest Resources and Conservation and broader University of Florida I express my sincere apology if I have, inadvertently, forgotten to acknowledge anyone

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ ........ 10 L IST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 2 CONVERTING MATURE SLASH PINE PLANTATIONS TO UNEVEN AGED STANDS: A SIMULATION ANALYSIS OF SILVICULTURAL REGIMES ................ 20 Materials and Methods ................................ ................................ ............................ 24 Site and Data ................................ ................................ ................................ .... 24 Model Description ................................ ................................ ............................. 25 Silvicultural Regimes ................................ ................................ ........................ 27 Evaluation of Silvicultural Regimes ................................ ................................ .. 29 Statistical Analyses ................................ ................................ .......................... 30 Results ................................ ................................ ................................ .................... 31 Initial Stand Conditions ................................ ................................ ..................... 31 Evaluation of Silvicultural Regimes ................................ ................................ .. 31 Discussion ................................ ................................ ................................ .............. 36 Comparison of Silvicultural Regimes ................................ ................................ 36 Assumptions and Limitations of the Simulation Model ................................ ..... 39 3 UNDERSTORY LIGHT DYNAMICS IN LONGLEAF PINE SLASH PINE ECOSYSTEMS IN CONVERSION ................................ ................................ ......... 54 Material and Methods ................................ ................................ ............................. 59 Study Sites ................................ ................................ ................................ ....... 59 Stand Characteristics and Treatments ................................ ............................. 62 Data Collection ................................ ................................ ................................ 63 Data Analysis ................................ ................................ ................................ ... 65 Results ................................ ................................ ................................ .................... 68 Effect of Management Systems ................................ ................................ ........ 68 Effect of Overstory Species Composition ................................ ......................... 70 Discussion and Conclusions ................................ ................................ ................... 70 Comparison of Management Systems on Light Transmittance ........................ 70 Effect of Overstory Species Composition ................................ ......................... 75

PAGE 7

7 Conclusions ................................ ................................ ................................ ...... 76 4 SEED BANK DYNAMICS AND UNDERSTORY RESTORATION IN PINE FLATWOODS ECOSYSTEMS ................................ ................................ ............... 86 Materials and Methods ................................ ................................ ............................ 89 Study Area ................................ ................................ ................................ ........ 89 Study Stands ................................ ................................ ................................ .... 90 Vegetation and Soil Seed Bank Sampling ................................ ........................ 91 Seedling Emergence Method ................................ ................................ ........... 93 Data Analysis ................................ ................................ ................................ ... 94 Results ................................ ................................ ................................ .................... 94 Aboveground Vegetation ................................ ................................ .................. 94 Seed Bank Structure and Composition ................................ ............................ 96 Relationship between Aboveground Vegetation and Seed Bank ..................... 98 Discussion and Conclusions ................................ ................................ ................... 99 Aboveground Vegetation ................................ ................................ .................. 99 Seed Bank Structure and Composition ................................ .......................... 100 Relationship between Aboveground Vegetation and Seed Bank ................... 103 Implications for Restoration of Understory ................................ ...................... 103 5 SUMMARY AND CONCLUSIONS ................................ ................................ ........ 117 APPENDIX A DETAILS OF THE SCENARIOS USED IN SIMULATING STAND CONVERSION ................................ ................................ ................................ ...... 122 B ESTIMATES OF STRUCTURAL DIVERSITY, CARBON STOCKS AND TIMBER PRODUCTION IN DIFFERENT SI MULAT ION SCENARIOS ................. 134 C RANKING OF THE SIMULATION SCENARIOS USED IN THE STUDY .............. 139 LIST OF REFERENCES ................................ ................................ ............................. 143 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 156

PAGE 8

8 LIST OF TABLES Table page 2 1 Control variables inpu ts specified to FVS sn variant to control the simulation for the study. ................................ ................................ ................................ ....... 44 2 2 Overview of the scenarios used in simulating conversion of slash pine plantations to uneven aged stands ................................ ................................ .... 45 2 3 Main effects and interactions of the harvest types, residual basal areas, cutting cycles, and regeneration on structural diversity, carbon stocks, and timber production. ................................ ................................ ............................... 46 2 4 Structural diversity, carbon stocks, and timber production in top scenarios at 11.5 m 2 ha 1 basal area ................................ ................................ ...................... 47 2 5 Structural diversity, carbon stocks, and timber production in top scenarios at 4.6 m 2 ha 1 basal area ................................ ................................ ........................ 48 3 1 Summary of the stands treated to different reproduction cutting systems at Blackwater River St ate Forest and Goethe State Forest, FL ............................. 78 3 2 Number of measurement plots and dates of acquisition of Digital Hemispherical Phot ographs ................................ ................................ .............. 79 3 3 Estimates (MeanSE) of leaf area index (LAI), fraction of Absorbed Photosynthetically Active Radiation (fAPAR), cover fraction, and direct fAPAR: diffuse fAPAR ratio in longleaf pine slash pine stands ........................... 80 3 4 Effect of overstory species composition on understory light availability (least square means) in longleaf pine slash pine stands ................................ .............. 81 4 1 History of the three stand conditions representing a restoration gradient at ................................ ................................ .............. 105 4 2 Attributes of the three stand conditions representing a restoration gradient at ................................ ................................ ....... 105 4 3 Presence/absence list from the vegetation sampling and seed bank examination from the three stand conditions representing a restoration ................................ ........................... 106 4 4 Estimates (means) of seed density, number of species, and diversity (Shannon Index) observed in the soil seed bank samples .............................. 111 4 5 The most common species germinated from the seed bank collected at different depths in the soil profile ................................ ................................ ..... 112

PAGE 9

9 4 6 vegetation for three stand conditions representing a restoration gradient at FL. ................................ ................................ ............. 112 A 1 Details of the scenarios used in simulating conversion of slash pine plantation to an uneven aged stand ................................ ................................ 123 B 1 Estimates (Mean Standard Error) of structural diversity, carbon stocks, and timber production in the 49 scenarios simulated in the study .......................... 135 C 1 Ranks of the scenarios simulated in the study for their ability to provide multiple benefit s ................................ ................................ ............................... 140

PAGE 10

10 LIST OF FIGURES Fi gure page 2 1 Forest Vegetation Simulator (FVS) processing as used in simulating different silvicultural regimes for converting even aged plantations to uneven aged stands. ................................ ................................ ................................ ................ 49 2 2 Input stand conditions inventoried in 2009. ................................ ........................ 50 2 3 Sensitivity of structural diversity to regeneration when the stand is treated unde r different harvest types, basal areas and cutting cycles. ........................... 51 2 4 Sensitivity of carbon stocks to regeneration when the stand is treated under different harvest types, basal areas and cutting c ycles. ................................ ..... 51 2 5 Sensitivity of total merchantable timber produced to regeneration when the stand is treated under different harvest types, basal areas and cutting cycles. .. 52 2 6 Sensitivity of saw timber production to regeneration when the stand is treated under different harvest types, basal areas and cutting cycles. ........................... 52 2 7 Change in stand structural diversity (Shannon index) during the simulation period in top scenarios when maintained at higher basal area (11.5 m 2 ha 1 ). ... 53 2 8 Change in stand structural diver sity (Shannon index) during the simulation period in top scenarios when maintained at a lower basal area (4.6 m 2 ha 1 ). .... 53 3 1 Location of the three study sites at Blackwater River State Forest Goethe FL ................................ .................... 82 3 2 Acquisition of digital hemispherical photograph using Nikon 5000 FC E8 camera ................................ ...... 82 3 3 Digital hemispherical photograph (left), classified image (middle), and gap fraction image (right) of longleaf pine stands ................................ ................... 83 3 4 Understory light conditions (sky, cover fraction, direct fraction of Absorbed Photosynthetically Active Radiation (fAPAR), and diffuse fAPAR) in various areas of the group selection system plots. ................................ ......................... 84 3 5 Understory light conditions (leaf area index (LAI), and ratio of direct fraction of Absorbed Photosynthetically Active Radiation (fAPAR) and diffuse fAPAR) in in various areas of the group selection system plots. ................................ ..... 85 4 1 ............................... 113

PAGE 11

11 4 2 State Forest, FL. ................................ ................................ ............................... 114 4 3 Hell State Forest, FL. ................................ ................................ ........................ 115 4 4 Aboveground seed bank species richness relationship observed in mesic wet flatwoods sites in different stand conditions representing a restoration ................................ ........................... 116

PAGE 12

12 LIST OF ABBREVIATION S amsl Above mean sea level BDq Residual Basal Area Maximum Diameter Diminution quotient approach BRSF Blackwater River State Forest C Carbon CFEOR Co operative for Conserved Forest Ecosystems: Outreach and Research DBH Diameter at Breast Height DHP Digital Hem ispherical Photograph(y) DOF Division of Forestry, Florida fAPAR fraction of Absorbed Photosynthetically Active Radiation FL Florida GSF Goethe State Forest ha hectare LAI Leaf area index THSF USA United States of America USDA Unit ed States Department of Agriculture

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 ROLE OF UNEVEN AGED SILVICULT URE AND THE SOIL SEED BANK IN RESTORATION OF LONGLEAF PINE SLASH PINE ( Pinus palustris Pinus elliottii ) ECOSYSTEMS By Ajay Sharma August 2012 Chair: Shibu Jose Cochair: Kimberly Bohn Major: Forest Resources and C onservation Concern over the loss of lon gleaf pine slash pine ( Pinus palustris Pinus elliottii ) e cosystems and the associated high level of biodiversity in the southeastern United States has led to an increased interest in the ir restoration and development of best management practices. This research conducted using both simulation analysis and field data from multiple sites in north west and north central Florida addressed three important aspects related to their restoration and management : 1) evaluating various silvicultural regimes to res tore o verstory structure and m eet multiple objectives 2) quantifying dynamics of the biophysical environment specifically light levels with various silvicultural treatments and species composition, and 3) examining soil seed banks for their potential to restor e understory biodiversity in degraded s tands S imulation analysis of 49 scenarios of different silvicultural regimes identified possible management options to restore degraded mature slash pine plantation s to multifunctional uneven aged stand s which could maximize the provision of multiple benefits including timber production, structural diversity and carbon s torage Appropriate options include converting slash pine plantations using the irregular

PAGE 14

14 sh e lterwood method with basal area of 4.6 m 2 ha 1 on a cutting cycle of 20 year s and regeneration of 741 or more seedlings/hectare For m anagement at higher residual basal area of 11.5 m 2 ha 1 several different silvicultural regimes maximized the provision of multiple benefits but more so with high level s of regeneration following harvests Among the various uneven aged silvicultural methods evaluated in field trials group selection method resulted in the creat ion of understory light conditions desirable for both tree regeneration and understory biodiversity The understory light availability created by group selection method was comparable to that of shelterwood method but had high er variability at the stand level. Light availability also varied with overstory composition, with higher understory light availa bility in pure longleaf pine stands than those of slash pine. Examination of seed bank in degraded, partially restored, and restored s tands in pine flatwoods sites found a total of 26, 39, and 64 species respectively mainly consisting of sedges and rushes Unlike other s tands, seed density in degraded s tands was higher at a depth of 5 10 cm in soil.

PAGE 15

15 CHAPTER 1 INTRODUCTION Longleaf pine ( Pinus palustris ) ecosystems were once the most dominant forest type in the southeastern United States occupying an estim ated 37 million hectare (ha) Out of these 37 million ha, 23 million ha consisted of pure longleaf pine while 14 million ha cons isted of longleaf pine mixed with the other pines such as slash pine ( Pinus elliottii ) The se ecosystems were distributed all ov er the southern C oastal P lains ranging from Virginia to Texas through central Florida to the montane areas in northern Alabama and occup ied a variety of sites ranging from xeric sandhills to wet poorly drained flatwoods ( Frost, 2006 ). In fact on wetter s ites, especially in panhandle of Florida, the overstory in these ecosystems had substantial amount s of slash pine making it the dominant species in the overstory. However, all of the se ecosystems were associated with high level of species richness mainly contributed by understory wherein as many as 40 species/m 2 were observed, leading them to be recognized as the ecosystems with the highest species richness outside the tropics ( Peet and Allard, 1993) T hese ecosystems have now been reduced to a fraction of their original extent mainly due to their conversion to commercial pine plantations or through degradation as a result of fire exclusion (Frost, 1993; Landers et al., 1995; Frost, 2006) currently classifying them among the critically threatened ecosyst ems of North America (Means and Grow, 1985; Noss, 1989; Noss and Peters, 1995) Loss of habitat resulting from the decline in the extent of these ecosystems has led to rarity of 191 vascular plant taxa (Hardin and White 1989 ; Walker 1993) and several vert ebrate species Commercial pine plantations managed under intensive silvicultural practices constitute a substantial portion of area formerly occupied by longleaf pine ecosystems

PAGE 16

16 (Frost, 2006) Intensive silviculture and timber intensive management in thes e sites have led to degradation resulting in a serious loss of species richness. In Florida, particularly in the wetter sites, a large extent of these ecosystems was replaced by intensively managed slash pine plantations for the sole production of timber a nd pulpwood. These high density plantations with intense understory shade have poor to non existing herbaceous cover and thick growth of shrubs and hardwood midstory. Concern over the loss of longleaf pine ecosystems and the associated species richness has led to an unprecedented interest and efforts to manage and wherever possible restore the se ecosystems (Alavalapati et al., 2002; Brockway et al., 2005b; Jose et al., 2006 a ; Kirkman et al., 2007 ) Although historically, these forests have been successfull y regenerated following even aged shelterwood reproduction methods, u neven aged reproduction methods have received considerable attention in the recent past because o f the perception that s e lection management has potential to sustainably meet diverse objec tives of timber production, biodiversity enhancement, habitat conservation, recreation and carbon sequestration ( Brockway et al., 2005 a ; Jose et al., 2006 a ) Uneven aged management, actually may be more suitable for th ese ecosystems because it emulates so me of the natural disturbances that have historically sustained them ( Brockway et al., 2005 a ; Brockway et al., 2006) For example, the group selection and the single tree selection methods simulate mortality caused by lightning strikes or small insect outb reaks. The irregular shelterwood method represents circumstances where a partial stand of longleaf pine is left following a catastroph ic event, such as a hurricane (Brockway et al., 2006) An uneven aged management system that creates biophysical environme nt favorable for natural regeneration and

PAGE 17

17 understory biodiversity can possibly meet the multiple objects of management including provision of timber and environmental services The second chapter in this dissertation deals with restoring overstory structur e in a mature unthinned slash pine plantation established on a former longleaf pine flatwoods site in the panhandle of Florida. We use a simulation analysis approach to evaluat e a suite of silvicultural regime s to convert a mature slash pine plantation to an uneven aged stand and evaluate these d ifferent regimes for their effectiveness in terms of creating stand structural diversity carbon s torage and timber production. We, then, discuss merits and ecological feasibility of the potential silvicultural re gimes and suggest optimal conversion scenarios to meet the objects of management and restoration The restoration and sustainable management of these ecosystems also require a better understanding of the biophysical factors that influence regeneration dyna mics and understory succession during the conversion process and subsequent uneven aged management. For example, understory light availability is one of most important factors that have been determined to affect t he survival and growth of longleaf pine reg eneration and the maintenance of understory species richness ( Wolters, 1981; Palik et al., 1997; McGuire et al., 2001; Gagnon et al., 2003 ; Platt et al., 2006 ). Greater understory light has also been associated with increased growth in basal area and crown width in longleaf pine (Harrington and Edwards, 1999; Harrington, 2006). However, diversity in growing conditions may be needed to maintain the broad range of species in the understory in these ecosystems (Harrington, 2006). In the third chapter, we use h igh resolution Digital Hemispherical Photography (DHP) to examine understory light regimes (indicated by canopy biophysical measures such as LAI, cover fraction,

PAGE 18

18 and fraction of Absorbed Photosynthetically Absorbed Radiation) in the longleaf pine slash p ine ecosystems being treated to uneven aged reproduction cuttings, and discuss implications of using selection silviculture in managing these unique ecosystems While restoration of the overstory can be accomplished by stand conversion and planting if req uired the restoration of species rich understory poses a real challenge (Cohen et al., 2004). In these degraded off site pine plantations where most of the natural species rich understory has been lost, the seed bank may provide a valuable opportunity for restoration (Cohen et al., 2004; Simpson et al., 1989). A s eed bank contains propagules of species, desirable or undesirable which can colonize after restoration activities such as thinning, burning, or other mechanical treatments following decades of su ppression (Abella and Springer, 2008). A s eed bank containing desirable species, if present, can help prevent the costs associated with sowing and transplanting of native species. Seed banks also become more valuable when geographically suitable seed of th e desired species is not readily available. Additionally, natural seed dispersal of desired understory species may be unreliable when rare community types are isolated (Van der Valk and Pederson, 1989; Augusto et al., 2001). The seed bank dynamics in pine flatwoods sites in the restoration context is discussed at length in the fourth chapter We examine structure and composition of soil seed bank relationships between seed bank and existing vegetation and how these relationships change during the process of ecosystem restoration Finally, we discuss if the soil seed banks can be relied on for restoration of species rich understory in the degraded flatwoods sites in panhandle of Florida.

PAGE 19

19 T hus, the overall objective of t h is study was to develop a better und erstanding of three main aspects related to restoration of overstory and understory in these ecosystems: (1) How can we optimally convert off site plantations to uneven aged stands that will optimize the provision of commodity (timber production) as well a s non commodity benefits (environmental services such as structural diversity and carbon s torage ) ?; (2) How does stand conversion and uneven aged management affect the biophysical environment (understory light availability ) in the se stand s which affect re generation as well as understory succession ? and; (3) Can a n impoverished understory be restored in situ from the soil seed bank in the degraded plantation s tands? The study was conducted at three sites in north west and north central Florida viz., Blackw State Forest (GSF). These state forests are the sites for long term experiments to convert pine plantations to uneven aged stands, and evaluat e uneven aged silvicultural methods in the ir management

PAGE 20

20 CHAPTER 2 CONVERTING MATURE SLASH PINE PLANTATIONS TO UNEVEN AGED STANDS : A SIMULATION ANALYSIS OF SILVICULTURAL REGIMES Recent decades have witnessed an increasing interest in managing southern pinelands as uneven aged forests because of the perception that uneven aged silviculture has potential to optimize forest resource use for multiple objectives in a sustainable manner (Brockway et al., 2005 a ; Jose et al., 2006 a ; Division of Forestry, 2007). From a timber and fiber production stand point, even aged systems are believed to be the most efficient and profitable in the short term. However, timber and fiber production are no longer considered the only products to be derived from forestland and increasingly, they are not the primary object ive of management or even the most profitable product or service produced (Tarp et al, 2005; Pukkala et al., 201 0 ; Laiho et al., 2011). Groundcover biodiversity, w ildlife habitat recreation, and environmental services such as carbon sequestration are beco ming more important on some lands (Sedjo, 2001; McConnell, 2002; Gorte, 2009), and may be incompatible with the forest structure associated with even aged stands (McConnell, 2002) Further, h igh structural diversity with many elements, such as tree sizes, canopy layers and various types of decaying wood is highly correlated to the species diversity of a forest stand (MacArthur and MacArthur, 1961; Thomas, 1979; Laiho et al., 2011; Marion et al., 2011) Additionally public acceptance of even aged methods su ch as clearcutting has changed (Gobster 1996; Lindhagen, 1996; Gundersen and Frivold, 2008), and the stand conditions associated with uneven aged forest structure are viewed as aesthetically favorable (Silvennoinen et al., 2001). As a consequence, many age ncies now have a mission to convert the plantations into uneven aged stands (Division of Forestry, 2007; Brockway and Outcalt 2010).

PAGE 21

21 However, their experience with such conversion and uneven aged silviculture is limited. Since most of the silvicultural re search in the past century has focused on even aged silviculture, there is uncertainty as to the feasibility of uneven aged silviculture in many southern species and its implications for the provision of multiple benefits including timber and environmental services. Several strategies have been suggested to convert and subsequently manage plantations as uneven aged stands which invariably have involved partial cuttings either uniformly over the stand or in patches, heavy enough to allow successful establish ment of multi age cohorts, and eventually allowing the application of uneven aged silvicultural systems such as single tree selection or group selection system (O'Hara, 2001; Nyland, 2 003; Loewenstein, 2005). Broadly, one approach to partial cutting of eve n aged stand involves heavily thinning the even aged stand in the beginning by removing the inferior or suppressed trees to create a vigorous residual stand to which a new age class will be added following natural regeneration (Nyland, 2003; Lo e wenstein, 2 005) Irregular shelterwood, a two aged regeneration method, is similar to that approach. Another approach that ha s been suggested is to apply the selection cut based on diameter regulation which aims at creating a reverse J shaped diameter distribution di rectly from the beginning (Della Bianca and Beck, 1985) T he field trials to evaluate these strategies generally span over a period of many decades before comparisons can be made on their relative effectiveness. Simulation models, however, can be used to p redict or conduct sensitivity analyses on different silvicultural regimes in lieu of evidence that will take several decades to develop with experimental field trials (Vanclay, 1994; Weiskittel et al., 2011).

PAGE 22

22 rest Vegetation Simulator (FVS) model (Crookston and Dixon, 2005) to evaluate different s ilvicultural regimes to convert a mature slash pine ( Pinus elliottii ) plantation to an uneven aged stand. Slash pine is one of the commercially most im portant native p ine species in Coastal P lains of the southeastern United States and is also a natural part of many of the more mesic and hydric ecosystems in the region In 2007, total acreage of slash pine including mixed stands with longleaf pine ( Pinus palustris ) fores ts in the southeastern United States was 5.3 million hectares (Smith et al., 2009), with a majority of slash pine about 79% of the total acreage concentrated in Florida and Georgia (Barnett and Sheffield, 2005). Out of that, about 69% of slash pine stand s exist as intensively managed plantations with rotation age s of 30 years or less ( Barnett and Sheffield, 2005). Currently, there is an increasing interest in man ag ing these plantations u nder alternat e reproduction systems using uneven aged silviculture H owever, our experiences with uneven aged management are limited. Owing to the strong shade intolerant nature of slash pine (Baker, 1949; Lohrey and Kossuth, 1990), transition to an uneven aged stand could possibly require a low residual basal area that wo uld allow an adequate amount of understory light for regeneration and seedling growth A low basal area approaching about 4 to 5 m 2 ha 1 as would be used with the shelterwood method has been suggested to manage slash pine (Langdon and Bennett, 1976) Suc h low level s of residual basal area will allow abundant light to fall on the forest floor, which is likely to result in adequate natural regeneration a nd help maintain a species rich understory in these ecosystems (Dickens et al., 2004; Jose et al., 2006 a ) However a residual basal area of up to 11.5 m 2 ha 1

PAGE 23

23 has also been reported to have no adverse effect on the early seedling establishment or growth of slash pine (McMinn, 1981) and may thus be one of the possible levels of appropriate residual basal area s to maintain in uneven aged stands. T he success of uneven aged silvicultural methods also depends on its ability to obtain regeneration of the desired species and to have that regeneration develop into merchantable size classes However information about the minimum level of regeneration required for successful application of selection silviculture is the biggest gap in our understanding (Guldin, 2006). S ince slash pine in the southeastern United States has been mostly cultivated as intensive ly managed hi gh density short rotation plantations for the purpose of timber and fiber scientific literature is critically deficient in studies on its natural regeneration, particularly in uneven aged stands. Because regeneration in southern pine forests is known to b e influenced by a number of factors related to management, including residual basal area, site preparation, shrub and hardwood control measures, and prescribed fire (Dickens et al., 2004; Jose et al., 2006), an appropriate level of regeneration, when known can be targeted by manipulating management practices and/or supplementing natural regeneration with planting. The overall objective of the study was to evaluate different silvicultural regimes, over a range of levels of regeneration for their feasibilit y to successfully convert plantations and lead to sustainable uneven aged stands The FVS model was used for its versatility to model a variety of forest types and stand structures ranging from even aged to uneven aged and to specify any combination of man agement activities during any cycle of the simulation period. We evaluated two harvest approaches for stand

PAGE 24

24 conversion at two harvest intensities and cutting cycles. The regimes were assessed in terms of their long term ecological and economic benefits. We specifically evaluated stand structural diversity carbon (C) stocks and merchantable timber production over a 100 year simulation Materials and Methods Site and Data State about 820 km 2 of poorly drained lowland mesic hydric flatwoods between the Apalachicola and Ochlockonee Rivers in the panhandle of Florida. At an elevation ranging from 0 to 10 m above mean sea level, it has nearly level topography. The climate is humid subtropical with annual precipitation totaling about 147 cm, of which about 49% is received during June to September (Gilpin and Vowell, 2006) Although more than 40 unique so il types occur within the forest, four groups account for the majority of the soils, namely, (a) Scranton Rutlege, (b) Plummer Surrency Pelham, (c) Meadowbrook Tooles Harbeson, and (d) Pamlico Pickney Maurepas (Gilpin and Vowell, 2006) All are poorly dr ained hydric soils. The site was once a swampy mosaic of wet prairies, cypress ( Taxodium spp.) sloughs, Atlantic White Cedar ( Chamaecyparis thyoides ) forests and other wetland and pine flatwoods communities, but large scale silvicultural operations and hyd rological manipulations during 1960s through 1980s converted extensive areas of native habitats to slash pine plantation. Stands were established following intensive mechanical site preparation, bedding and planting at high densities. Fertilizers, particul arly nitrogen and phosphorus, were applied at mid rotation. These large scale disturbances and manipulations to the habitats altered the

PAGE 25

25 ecological communities and hydrology of the area (Gilpin and Vowell, 2006). Currently, one of the main land management maintain these plantations as low density uneven aged stand structures that will maintain biological diversity, while integrating public use (Division of Forestry, 2007). The input data were collected fr om a mature unthinned slash pine plantation scheduled for active conversion to an uneven aged stand. We established 5 sample plots of 25 m x 25 m in the stand. Within each plot, we recorded diameter at breast height (dbh) and species for all the trees grea ter than 10 cm dbh. Within the 25m x 25 m plot, we established two 5 m x 5 m plots at diagonally opposite corners in which we measure d trees smaller than 10 cm dbh. Model Description We used the FVS model Version 4280 Southern U.S. PROGNOSIS RV: 12/20/1 1, developed by USDA Forest Service. FVS is a distance independent, individual tree forest growth model which uses individual stands as the basic units of projection. The applications of FVS spans a wide variety of forest types and stand structures ranging from even aged to uneven aged and single to mixed species It can model single stand or landscape and large regional assessments involving thousands of stands. The key variables of each tree are its count, species, diameter, height, crown ratio, diameter growth, and height growth while the key variables for each sample point, or plot, include slope, aspect, elevation, density, and a measure of site potential. The model has a self calibration feature that automatically adjusts the internal growth models to match the measured growth rates of a particular location consistent with the source of the input data. Thus, it is able to modify predictions for local conditions. The

PAGE 26

26 growth cycles in the simulator are set to 5 or 10 year time steps with up to 40 cycles a llowed per simulation (Dixon, 2002; Crookston and Dixon, 2005). The model has the ability to simulate almost any kind of harvest and other silvicultural activities. The u ser has the options to specify how many trees or the basal area from which size classe s should be retained at the end of any simulation cycle along with other activities such as shrub removal or artificial planting. Typical growth and mortality rates are predicted as functions of post harvest conditions (residual densities/basal area etc.). Broadly, the FVS model has four primary components: diameter growth, height growth, mortality, and crown change. Diameter growth, height growth, and crown change models are further divided into submodels that pertain to tree growth relationships are driven by tree height from which diameter growth is then estimated. On the other hand, the large tree growth relationships are driven by diameter first and then height growth is estimated from diameter growth and other tree and stand variables. The FVS sn model is a popular model and has extensive documentation. More details about its contents, description, structure, and applications can be found in McGaughey (1997), Van Dyck and Smith (2000), Donnelly et al. (2001), Dixon (2002), Crookston and Dixon (2005), and Keyser (2008). FVS sn does not have equations for predicting natural regeneration of slash pine. The user must schedule its natural regeneration by specifying input va lues for density and size of expected new trees based on other published literature and/or its understanding of the species ecology.

PAGE 27

27 The inventory data collected from the slash pine plantation were used to initialize the simulations in FVS model using spec ific control variables (Table 2 1, Figure 2 1 ).The outputs generated from FVS were used to calculate structural diversity, carbon s tocks and merchantable timber production as described below Silvicultural R egimes We followed two approaches in developing silvicultural regimes for stand conversion. One approach was the direct application of the uneven aged BDq method beginning with the first cut The BDq method is a common approach used in developing marking tallies for single tree selection system S pecifications of B D and q constitute a unique reverse J shaped distribution of diameter, which, theoretically, characterizes an ideal balanced uneven aged stand T he diameter distribution of the stand to be cut is compared to that of the ideal, balanced uneven aged stand, and a marking tally is generated that brings the existing stand closer to the balanced uneven aged structure (Farrar and Gersonde, 2004; Guldin, 2006; Keyser, 2008). The other conversion approach we used involved a traditional even aged or two aged approach by first cutting of trees of lower crown position in initial cuts to create a re sidual stand (Nyland, 2003). In these scenarios, all the trees of small size were cut first followed by larger sized trees till a desired residual basal area was achieved. This approach is essentially a or ning al., 1997) but approaches a shelterwood regeneration method at very low residual basa l areas We simulated a total of 49 scenarios, each replicated thrice, which formed various silvicultural regimes encompass ing a range of combinations of harvest t ypes residual

PAGE 28

28 basal areas cutting cycles, and regeneration densities (Table 2 2 and Appendix A ). with the he first cut area to either 4.6 or 11.5 m 2 ha 1 (h ) on a 10 or 20 year cutting cycles during the sim ulation period. Beginning with the third cut, however, all scenarios were treated to single tree selection system which inherently implements the BDq approach. Regeneration scenarios varied from abundant regeneration of as many as 2224 seedlings/ha in increments of about 495 seedlings/ha The simulation was run for a total of 100 years after the first cut in 2012 to 2111 at simula tion cycle length of 5 years. The naming scheme for the scenarios consisted of a sequence of four variables in the form basal area harvest type cutting cycle regeneration thin 10 al area of 11.5 m 2 ha 1 treated with low thinning in the beginning a nd cutting cycle of 10 years which results in 247 seedlings/ha of regeneration following every harvest. The following outputs from the model were generated: tree list, C arbon (C) report, stand summary, and cut list (Dixon, 2002). The individual trees in the tree list were then classified into diameter classes and height classes. We used a total of 18 diameter classes with class width of 5.08 cm, ranging from size 0 (less than 2.54 cm) to 8 6.4 + cm (equal to or greater than 83.8 cm). For height, we used a total of 16 height classes with class width of 3.048 m, ranging from 0 (less than 1.524 m) to 45.72 + (equal to or

PAGE 29

29 greater than 44.196 m). In both the cases, the classes represented the mid points of the class widths. Evaluation of Silvicultural Regimes The scenarios were evaluated based on their ability to create stand structural diversity, store C stocks and produce merchantable timber. Stand structural diversity: At every cycle of the sim ulation period, we calculated Shannon diversity Index for both the diameter class and height class distributions with respect to the basal area constituted by them. Shannon index was calculated as equal P i ln P i 2 is the proportion of basal a rea constituted by a diameter class (Magurran, 1988). Stand structural diversity was then obtained as the average of Shannon Indices for both diameter and height classes. Every simulation had 2 1 values of Shannon indices for diameter and height distrib utions each, one for every cycle plus one at the beginning of the simulation. Values were plotted over time and the average of Shannon indices at all cycles during a simulation was calculated as average stand structural diversity under a given scenario. C s tocks : The outputs of the C reports obtained from the model were grouped into aboveground stored C (consisting of aboveground live tree, standing dead trees, down dead wood, forest floor, and understory), belowground stored C (consisting of belowground li ve tree, and belowground dead tree), and total C removed in harvest at each year during the simulation period (Hoover and Rebain, 2011). The following variables were calculated: (a) Total stand C in the beginning of simulation (year 0) = Total aboveground s tored C in year 0 + Total belowground stored C in year 0 (b) Total stand C at the end of simulation (year 100 ) = Total aboveground stored C in year 100 + Total belowground stored C in year 100, and (c) Total additional C stored during

PAGE 30

30 simulation period = (Total stand C at the end of simulation (year 100 ) + sum of C harvest ed during different cycles in simulation period) Total stand C in the beginning of simulation (year 0 ). Average annual C stock for each scenario was then obtained by dividing the total additional C s tored by 100. Timber production: From the FVS summary output generated in simulations, we calculated values of the following variables: (a) Total merchantable timber produced during simulation period = Total merchantable timber removed during simulation + Total merchantable timber left standing after year 100 Total merchantable timber in year 0 and (b) Total sawtimber produced during simulation period = Total sawtimber removed during simulation + Total sawtimber left standing after year 100 Total sawtimber in year 0 Average annual productions of these variables were then obtained by dividing the totals by 100. Statistical A nalyse s We ran 3 simulations for each scenario, manually reseeding the random number generator, and thus, producing v ariation in the projection results. Given low and stable variance s across the simulations, three replications were considered adeq uate for the statistical analyse s (Hamilton, 1991). Mean and standard error were calculated for average annual values of struc tural diversity, C s tocks merchantable timber, and sawtimber for each scenario. Analyses of variance (ANOVA) were carried out for all the 4 9 scenarios residual basal area, cutting cyc le, regeneration and their interaction s ANOVA were done separately also for the scenarios with different levels of basal area i.e. within the groups of scenarios with 11.5 or 4.6 m 2 ha 1 basal area

PAGE 31

31 Significant Difference) test was in all of the analyses. For each level of basal area, w e then selected the top four scenarios based on the means that maximized structural diversity, C s tocks and merchantable timber each. Addition ally, the t op four scenarios leading to maximization of multiple benefits (i.e., collective provision of structural diversity, carbon storage and merchantable timber production) were identified for each level of residual basal area For th is purpose, we fi rst scaled the values of all three criteria variables as percent of their maximum among all scenarios and summed them for each scenario ; t he higher the sum, the greater the ability to provide multiple benefits All these top scenarios were then discussed f or their relative ecological feasibility and merit. Results Initial Stand C onditions The overstory was dense (1136 trees/ha) with an average basal area and mean quadratic diameter of a pproximately 28.7 m 2 ha 1 and 18 cm respectively. The diameter s at breas t height of initial slash pine plantation stand exhibited typical bell shape d distribution associated with even aged stands (Figure 2 2 A, B). Occasionally sweetbay ( Magnolia virginiana ) and pond cypress ( Taxodium ascendens ) were found in the overstory. Sl ash pine made about 63% of tree density and 96% of stand basal area. The lower diameter classes were dominated by sweetbay which had regenerated under the dense canopy of slash pine. Natural regenera tion of slash pine was lacking. Evaluation of Silvicultur al Regimes None of the scenarios maximized all criteria variables simultaneously. There were tradeoffs involved between structural diversity and C s tocks as well as merchantable

PAGE 32

32 timber production in different scenarios. The scenarios that led to higher st ructural diversity generally resulted in lower C s tocks and merchantable timber production, and vice versa ( Appendix B, Table 2 4 and 2 5 ) Not surprisingly, either harvesting approach followed by complete failure of regeneration (no regeneration scenarios ) was among the worst in terms of structural diversity, C stocks wherein no cutting and no regeneration occurred during the period of simulation, also led to low values of these variable s. Values for stand structural diversity, C s tocks and merchantable timber production overall similar or slightly superior to the harvest scenarios with no regeneration Stand structural diversity: In general, scenarios with high residual basal area had higher average structural diversity than low residual basal area (Shannon index as high as 2.058) ( F ( 1, 117 ) = 984, p<0.001) and scenarios with long cutting cycle led to lower average structural diversity ( Shannon Index as low as 0.749) ( F ( 1, 117 ) = 960, p<0.001) (Table 2 3) Average structural diversity increased significantly with increasing level of regeneration ( increase in Shannon index as much as 0.75 to 1.77 ) (p<0.001) when scenarios shifted from no regeneration t o 247 seedlings/ha regeneration scenarios for all harvest types and cutting cycles (Figure 2 3). Any f urther increases in the amount of regeneration did not significantly change structural diversity except causing a slight but significant decrease when reg eneration level reached 1730 seedlings/ha or more (P<0.001) higher average structural and regeneration ( Table 2 3, Figure 2 3).

PAGE 33

33 At basal area 11.5 m 2 ha 1 year cutting cycle and regeneration varying between 247 t hr o ugh 1730 seedlings/ha resulted in the highest average structural diversity ( Shannon Index = 2.06) (Table 2 4 ). At basal area 4.6 m 2 ha 1 with 10 year cutting cycle and regeneration of either 247 or 1236 seedlings/ha led to the highest average structural diversity (Shannon Index = 1.99) (Table 2 5) over th e simulation period, they had achieved comparable level by the end of simulation and sometimes by year 50 (Figure 2 7, 2 8). The low overall average structural diversity sim ulation period. C s tocks : L eaving a high er residual basal area s ignificantly increased C s tocks ( F ( 1 117 ) = 2552, p<0.001) For the most part, t annual average C s tocks (0 to 0.8 metric t on ha 1 year 1 higher than BDq ) for a given residual basal area, cutting cycle, and amount of regeneration except for scenarios with no regeneration (Figure 2 4). Annual average C stocks increased significantly and substantially ( as much as three times) when shifting from no regeneration to 247 seedlings/ha (p<0.001). Thereafter, a significant increase was observed only when regeneration became 1230 seedlings/ha or more (p<0.001). less sensitive to length o f cutting cycle; were sensitive to an increase in cutting cycle particularly at a low residual basal area The sensitivity of C s tocks to an increase in regeneration was

PAGE 34

34 only evident when basal area was low and cutting cycle wa Evaluation of the top scenarios a t basal area 11.5 m 2 ha 1 indicated that annual average C stocks was highest (2.07 metric ton ha 1 year 1 ) a 20 year cutting cycle and regeneration within a range of 741 to 1236, and 2224 seedlings/ha (Table 2 4). At basal area 4.6 m 2 ha 1 h ighest annual average C s tocks (1.88 metric ton ha 1 year 1 ) was led by the year c utting cycle and high regeneration of 1730 to 2224 seedlings/ha (Table 2 5) Merchantable timber production: A nnual a verage merchantable timber production during the simulation period followed more or less the same pattern as the annual average C s tocks T he scenarios with high residual basal area had significantly higher merchantable timber production than all the scenarios with low basal area except at no regeneration (p<0.001). timber production for a given basal area, cutting cycle and level of regeneration (p<0.001). The g reatest significant increase (as high as 4.2 m 3 ha 1 year 1 ) in annual average merchantable timber production was observed when the scenarios shifted from no regeneration to 247 seedlings/ha. Thereafter, at higher basal area, significant but s light increase s w ere observed in a few scenarios only when regeneration was 123 6 seedlings/ha or higher while in some other scenarios increases in regeneration over 1236 seedlings/ha led to decreases in total merchantable timber production A distinct greater sensitivity to increase s in regeneration was observed in the low thin low basal scenario with long cutting cycle (Figure 2 5). Interestingly, sawtimber production revealed very clear distinction s in scenarios with different residual basal area s ; high

PAGE 35

35 basal area scenarios clearly producing s ignificantly and s ubstantially higher amount of annual average sawtimber than the low basal area scenarios However, harvest type* basal area interactions were not significant (Table 2 3) After a regeneration of 247 seedlings/ha, sawtimber production was relative ly less sensitive to increases i n regeneration for most part except for higher basal area at higher levels of regeneration which led to decrease in annual average sawtimber production (Figure 2 6). At basal area 11.5 m 2 ha 1 a short cutting cycle (10 year) and high regeneration of 1730 to 2224 seedlings/ha, and with a long cutting cycle (20 year) but low regeneration of 741to 1236 seedlings/ha resulted in maximum average annual merchantable timber production of 6.26 to 6.39 m 3 ha 1 year 1 (Table 2 4). At basal a rea 4.6 m 2 ha 1 h ighest average annual merchantable timber production (4.86 to 5.11 m 3 ha 1 year 1 ) was achieved in the year cutting cycle with regeneration bet ween 1236 to 2224 seedlings/ha (Table 2 5) Topmost scenarios for mu ltiple benefits : At basal area 11.5 m 2 ha 1 s cenarios that maximized multiple objective s a 1 0 year cutting cycle and level of regeneration varying from 741 to 1730 a 10 year cutting cycle and high level of regeneration (22 24 seedlings/ha) ( Appendix C, Table 2 4). It should be noted that higher ranking of BDq scenarios was due to relatively greater contribution of structural diversity In other words were mor e effective at creating structural diversity as compared to timber production or C s torage For example, average structural diversity ranged

PAGE 36

36 At basal area 4.6 m 2 ha 1 m longer cutting cycle with a range of regeneration between 741 to 2224 seedlings/ha ( Appendix C, Table 2 5). ranked among the top 4 in providing multiple benefit s at lower basal area. Discussion Compariso n of S i lvicultural R egimes The best scenarios for creating high average structural diversity at both low and high residual basal area s invariably involved BDq cut from the beginning. This is because BDq approach specifically aims at creating a residual stand closer to the balanced uneven aged structure with a typical reverse J shaped diameter distribution ( Smith et al., 1998; Guldin, 2006) In doing so to a mature even aged stand in the initi ation of conversion, h owever, a BDq cut also tend s to cut trees which are among the best grown trees in the stand leaving a residual stand of inferior suppressed trees (Kelty et al., 2003 ; Lo e wenstein, 2005) resulted in low average s tructural diversity, they approached to similar structural selectively removed during that cu tting cycle leading to concentration of large sized trees which reduced overall structural diversity However this reduction was restricted to the first cutting cycle only and thereafter structural diversity increasing rapidly to become comparable to BDq s cenarios.

PAGE 37

37 Though a higher structural diversity is desirable for wildlife habitat enhancement in pine forests of Florida ( McConnell, 2002; Marion et al., 2011), our understanding of an optimum range of structural diversity is inadequate. It is highly probab le that the optimum structural diversity required to maintain habitat as well as aesthetics is not necessarily the maximum possible that could be achieved in the uneven aged stands. Our goals of achieving a level of structural diversity in this sense, are fuzzy. Also, our structural diversity considered in our study is essentially tree size diversity. Structural diversity, as it is understood, will also include snags, litter accumulation, coarse woody debris, tree spacing which all together create a comple x structure ( Mc Elhinny et al., 2005). These additional attributes were not examined in our study. The scenarios which led to high C stocks in general, also resulted in high merchantable timber and saw timber production While other factors such as stocki ng density and forest productivity determine the amount of C stored in the forest biomass, the length of the rotation determines how long the C remains in the forest system (Kaipainen et al., 2004). This is consistent with our model simulations as low thin ning with a long cutting cycle maximized C s tocks by retaining large r trees which added more C than the respective BDq scenarios with similar cutting cycles From a private land sale of timber and earning C credits from the efforts to sequester C these scenarios are highly desirable. Given that all methods and computations of C accounting by FVS model are consistent with Intergovernmental Panel on Climate Change (IPCC) Good Prac tice Guidance and U.S. voluntary C accounting rules and guidelines (Hoover and Rebain, 2011); the simulation outputs have a legitimate appeal.

PAGE 38

38 Most of the public agencies interested in managing forests for larger societal, environmental, and econom ic bene fits, as well as their ability to comply with global obligations to mitigate climate change, require strategies that optimize the provision of these benefits where none of the benefits is of greater or lesser importance. Our approach to scale the values of all three criteria variables as percent of their maximum among all scenarios, and then ranking the scenarios for maximizing the provision of these benefits, takes this into consideration However, the use of scaling for this purpose should be interpreted with a caution as scaling assumes that values of a variable are distributed uniformly within its range. The multiple benefits were best provided a t high basal area of 11.5 m 2 ha 1 by high regeneration with short cutting cycle However, t he realization of these outcomes in a practical sense may be uncertain. For instance, t even a moderate level of regeneration in initial harvests when t he residual stand consists of a large proportion of suppressed inferior trees (Kelty et al., 2003) level of regeneration in a stand with high basal are a may al s o be difficult to achieve as the high light requirements of the species may not be met in dense stand conditions For these scenarios to succeed, it will probably be required to supplement the regeneration with planting in the initial few cutting cycles. Another reasonab le option regeneration (Rank 5, see Table 2 4). This has higher probability of success as there is greater chance that some moderate level of regen eration will be obtained or

PAGE 39

39 compensated when cutting cycle is long g iven our understanding of theses ecosystems (Dickens et al., 2004). Despite our stated expectations, our understanding of natural regeneration in uneven aged stands in slash pine is still very poor. At low basal area of 4.6 m 2 ha 1 any low thin scenario with a longer cutting cycle a nd regeneration of more than 741 seedlings/ha acre will optimize the provision of multiple benefits; better regeneration leading to greater benefits. This is a highly likely to succeed scenario as this represents stand condition with an open canopy very much in alignment with the ecological requirements of intolerant slash pine. A regeneration of a minimum of 741 seedlings/ha at a cutting cycle of 20 years with a residual basal area of 4.6 m 2 ha 1 is a very reasonable expectation Assumptions and Limitations of the Simulation M odel Though the simulation analysis provided a good broad understanding of the conversion process, and the tradeoffs involved in the diffe rent scenarios, it had several limitations particularly with respect to regeneration The scenarios simulated in the study were intentionally kept simple. Because FVS model is not spatially explicit, it could not appropriately be used to simulate group se lection harvest cuts when gap dynamics have been known to influence regeneration and growth in pine stands ( Brockway and Outcalt 1998; Jose et al., 2006 a ) The group selection or the patch harvests of varying size s are also considered to be one of the bes t means to convert stand structure (Kelty et al., 2003; Nyland et al., 2003). A model that can effectively simulate distance dependent and gap phase dynamics is highly desirable for evaluating group selection harvests regimes for converting stand structure Our model lacked that capability.

PAGE 40

40 FVS carbon reporting in the study did not include C fluxes such as management related emissions (e.g., C emitted from equipment use when harvesting and transporting timber, residual stand debris, etc.). A complete C foot print analysis would include life cycle analysis of all aspects of forest management, such as the emissions associated with the production, transportation, and application of fertilizers, if any (Hoover and Rebain, 2011). Also, the impact of possible clima te change during the simulation period was not accounted for in simulation projections. Economic analyses in terms of costs and benefits of activities and products in different scenarios were also not done which could be a significant factor in finally dec iding about appropriate silvicultural regime Fire which is an important ecological disturbance in these ecosystems was not considered, nor were other shrub and hardwood control measures. However, we did assess a range of regeneration scenarios from compl ete failure of regeneration to abundant regeneration that would span the likely range of regeneration densities with and without these additional control measures The knowledge of the effect of these factors on natural regeneration can help mangers in m an ipulating management efforts to achieve desirable amount of regeneration. For example, seed b ed preparation and removal of competition by removing shrub and saw palmetto ( S erenoa repens ) significantly improves slash pine seedling establishment (Langdon and Bennett, 1976). Occasional fire in slash pine stands helps to create a mineral seed bank and reduce competition from hardwoods to enhance germination (Landers, 1991; Lohrey and Kossuth, 1990). Though y oung slash pine is susceptible mature trees are quite fire resistant (Brown and Davis, 1973). A fire interval of at least 5 6 years allows the young

PAGE 41

41 slash pine trees to develop fire resistance. The seedlings grow fast and in 10 12 years slash pine is resistant to fire that does not crown (Wright and Bailey, 1982). Cattle grazing, which is extensive in pine flatwoods in southeastern United States, is not known to cause any serious harm except in first few years of seedling growth (Pearson et al., 1971). Thus, m anagers can influence regeneration by burning a sl ash pine stand before an expected good seed year or before a cut to prepare a seed bed to enhance germination Young slash pine regeneration cohort should not be burned for first 5 years of age or when they are at least 4 5 m high (Langdon and Bennett, 197 6; McCulley 1950). A cutting cycle length of 10 and 20 years will provide enough growth periods to new cohorts to achieve these sizes. Additionally, c attle grazing, chemical or mechanical treatments can be used to reduce the buildups of the fi n e fuel and h ardwood competition until the new regeneration is resistant to light fire. Though the manipulation of factors affecting regeneration can help obtaining desired regeneration level, t he BDq scenarios may not provide a vigorous source of seed trees. BDq cut a pplied to a mature stand in the beginning essentially lead to high grading c reat ing a residual stand consisting of poorly grown trees (Smith et al., 1997; Kelty et al., 2003; Nyland et al., 2003). Such a stand with poorly developed crown structure is likel y to have reduced growth rate and may not result in production of adequate seed crop following the cut. Owing to this reason, BDq approach has been considered inappropriate for converting stand structures (Kelty et al., 2003; Lo e wenstein, 2005). Artificial regeneration following first two cuts will possibly be needed for these scenarios to succeed However, once an uneven aged structure has been achieved, BDq approach can be successfully applied for maintaining uneven aged

PAGE 42

42 structure (Kelty et al., 2003 ; Lo e wenstein 2005; Guldin, 2006; Brockway and Outcalt, 2010 ). On the other hand, the top scenarios involving low thinning in the beginning will retain the best grown trees (dominants and co dominants) in the residual stand (Smith et al., 1997; Nyland, 2003) Such a low thinned stand illuminates the understory by removing low shade (Nyland, 2001). The vigorous residual stand with bright understory seems more likely to result in successful natural regeneration. In that sense, scenarios have greater ec To conclude, o verall our simulation analysis reveals that no single stand conversion scenario will lead to maximization of all products and services simultaneously. There are always tradeoffs involved. The sc enarios that result in higher structural diversity generally le a d to lower C stocks and merchantable timber production. Given the need to achieve multiple objectives, the scenarios under the assumptions of our simulation that would best achieve that at res idual basal area of 11.5 m 2 ha 1 year cutting cycle requiring moderate level of regeneration (1236 seedlings/ha). At low basal area of 4.6 m 2 ha 1 year cutting cycle which results in min imum regeneration of 741 seedlings/ha is the best scenario. However, these results should also be taken into which ranked among the top at high basal area were not co nsidered feasible due to their high regeneration requirements and the uncertainty of obtaining those levels of regeneration. An important observation to note however, is that the greatest gains for all of the variables in all of the scenarios were made wh en as low as 247 seedlings/ha were regenerated. Further gains in most of the scenarios were insignificant or were very

PAGE 43

43 small when significant. Owing to this, if as low as 247 seedlings/ha can be regenerated following a harvest, reasonable multiple benefits still can be achieved under most of the scenarios. In worst case scenario of complete failure of regeneration, planting of 247 seedlings/ha is apparently not economically prohibitive, though our study did not evaluate economic aspects of the simulation sc enarios.

PAGE 44

44 Table 2 1 Control variables inputs specified to FVS sn variant to control the simulation for the study Parameter or attribute Input setting Location Code 80501 (Apalachicola National Forest in Florida, Latitude : 30.44 N, Long itude : 84. 28W) Ecological Unit codes 232Dd (Gulf coastal lowlands) Slope (%) 0 Aspect 0 Elevation(m) 9.15 Site Species Codes (Alpha code) SA, PC, MG,BY Site Index(m) for slash pine 27.43 Maximum density (trees/ha) 1075 Number of projection cycles 21 Project ion cycle length 5 Volume equations National Volume Estimator Library Volume Specifications: Minimum DBH/ Top Diameter Inside Bark (cm) Stump Height (cm) Pulpwood 10.2/10.2 30.5 Sawtimber 25.4/17.8 30.5 Sampling Design Large trees (fixed area p lot) 6.5 Small trees (fixed area plot) 81 Breakpoint DBH(cm) 10.2

PAGE 45

45 Table 2 2. Overview of the scenarios used in simulating conversion of slash pine plantations to uneven aged stands. Each of the harvest type was simulated over each combination of resid ual basal area, cutting cycle, and level of regeneration described below. All the scenarios had Maximum Diameter (D) > 86.4 cm, and Diminution quotient (q) = 1.5 for simulating BDq cuts in final stages Harvest type Description of silvicultural regime Res idual b asal reas (m 2 ha 1 ) Cutting cycles (y ea rs) Regeneration r anges ( seedlings / hectare ) BDq cut from first cutting cycle onwards 4.6, 11.5 10, 20 0, 247, 741, 1236, 1730, and 2224 Low t 1. Thinning from below during first cutting cycle 2. Thinning across d iameter classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.6, 11.5 10,20 0, 247, 741, 1236, 1730, and 2224 No action No cut of any kind NA NA No harvest and no regeneration were simulated in thi s scenario. The regeneration is defined as the seedlings of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle The name of a scenario consisted of four parts representing respectively harvest type cutti ng cycle regen eration

PAGE 46

46 Table 2 3 Main effects and interactions of the harvest types, residual basal area s cutting cycles, and regeneration on s tructural diversity, carbon s tocks and timber productio n Structural diversity C stocks Total merchantable timber S awtimber P ulpwood Df F value Pr(>F) F value Pr(>F) F value Pr(>F) F value Pr(>F) F value Pr(>F) Harvest type 1 7793 < 0.001 95 < 0.001 48 < 0.001 17 < 0.001 239 < 0.001 Residual basal area 1 984 < 0.001 2552 < 0.001 3573 < 0.001 6707 < 0.001 6 0.017 Cutting cycle 1 960 < 0.001 81 < 0.001 7 0.007 21 < 0.001 1 0.399 Regeneration 5 3132 < 0.001 1033 < 0.001 1021 < 0.001 1006 < 0 .001 281 < 0.001 Harvest type: Basal area 1 8 0.006 29 < 0.001 27 < 0.001 3 0.088 39 < 0.001 Harvest type: Cutting cycle 1 12 < 0.001 170 < 0.001 57 < 0.001 21 < 0.001 44 < 0.001 Harvest type :Regeneration 5 721 < 0.001 95 < 0.001 82 < 0.001 80 < 0.001 31 < 0.001 Cutting cycle :Regeneration 5 67 < 0.001 6 < 0.001 2 0.193 1 0.224 2 0.169 Basal area: Cutting cycle 1 0 0.918 53 < 0.001 109 < 0.001 32 < 0.001 101 < 0.001 Basal area :Regeneration 5 22 < 0.001 25 < 0.001 28 < 0.001 4 5 < 0.001 5 < 0.001 Residuals 117

PAGE 47

47 Table 2 4. Structural diversity, carbon s tocks and timber production in top scenarios at 11.5 m 2 ha 1 basal area. Four top scenarios for each of the variable that maximized their value and four scenarios that maximized multiple benefits were selected. Six scenarios overlapped resulting in a total of 10 scenarios Scenario Average stand structural diversity A nnual a verage C s tocks (metric ton/ha/year) Annual average merchantable wood production (m 3 /ha/year) Annual a verage sawtimber production (m 3 /ha/year) Annual a verage pulpwoo d production (m 3 /ha/year) Rank (provision of multiple benefits) 11.5 BDq 10 0247 (D) 2.0540.011 a** 1.7660.013 g,h 5.771 0.047 d,e 5.021 0.053 a,b,c 0.750 0.006 10 11.5 BDq 10 0741 (D,M) 2.0520.003 a 1.8510.014 e,f,g 6.002 0.040 b,c,d 5.067 0.094 a,b,c 0.934 0.055 4 11.5 BDq 10 1236 (D,M) 2.0490.006 a 1.8920.034 d,e 6.105 0.123 a,b,c 5.040 0.114 a,b,c 1.065 0.016 1 11.5 BDq 10 1730 (D,M) 2.0580.001 a 1.8710.018 d,e,f 5.999 0.062 b,c,d 4.884 0.046 b,c,d 1.115 0.077 2 11.5 Thin10 1730 (T) 1.8340.001 c 1.9100.00 6 c,d,e 6.258 0.011 a,b 4.932 0.063 b,c 1.327 0.052 7 11.5 Thin10 2224 (T,M) 1.8210.004 c 1.9680.026 b,c,d 6.387 0.073 a 4.879 0.068 b,c,d 1.507 0.014 3 11.5 Thin20 0741 (C,T) 1.7190.007 d 2.0090.004 a,b,c 6.259 0.031 a,b 5.298 0.004 a 0.960 0.028 9 11.5 Thin2 0 1236 (C,T) 1.7180.006 d 2.0680.008 a 6.280 0.027 a,b 5.111 0.026 a,b 1.168 0.035 5 11.5 Thin20 1730 (C) 1.7160.003 d 2.0530.009 b 5.936 0.022 c,d 4.813 0.034 b,c,d 1.123 0.020 11 11.5 Thin20 2224 (C) 1.7040.004 d 2.0780.006 a 5.716 0.025 d,e 4.575 0.061 d 1. 141 0.037 12 *D, C, and T represent one of the top scenarios in structural diversity, carbon s tocks and merchantable timber production. M represents the top scenarios with multiple benefits ** The values with the same letter are not significantly differe nt

PAGE 48

48 Table 2 5 Structural diversity, carbon s tocks and timber production in top scenarios at 4.6 m 2 ha 1 basal area Four top scenarios for each of the variable that maximized their value and four scenarios that maximized multiple benefits were selected Eight scenarios overlapped resulting in a total of 8 s cenarios. Scenario Annual a verage stand structural diversity Annual a verage C sequestration (metric ton/ha/year) Annual a verage merchantable wood production (m 3 /ha/year) Annual a verage sawtimber produ ction (m 3 /ha/year) Annual a verage pulpwood production (m 3 /ha/year) Rank (provision of multiple benefits) 4.6 BDq 10 0247 (D) 1.9920.008 a 1.1310.008 e 3.234 0.022 c 2.788 0.019 b,c 0.446 0.004 14 4.6 BDq 10 0741 (D) 1.9620.007 b 1.1610.028 e 3.292 0.10 c 2. 682 0.098 c 0.610 0.005 13 4.6 BDq 10 1236 (D) 1.9740.008 a,b 1.2130.001 d,e 3.448 0.016 c 2.749 0.015 b,c 0.699 0.030 9 4.6 BDq 10 2224 (D) 1.9290.006 c 1.2890.033 d 3.594 0.106 c 2.794 0.060 b,c 0.800 0.069 5 4.6 Thin20 0741 (C,T,M) 1.6340.001 d 1.5230.02 7 c 4.439 0.077 b 3.235 0.087 a 1.204 0.017 4 4.6 Thin20 1236 (C,T,M) 1.6000.003 e 1.7020.026 b 4.862 0.089 a 3.260 0.053 a 1.602 0.043 3 4.6 Thin20 1730 (C,T,M) 1.5550.007 f 1.7780.022 a,b, 4.906 0.088 a 3.040 0.093 a,b 1.866 0.081 2 4.6 Thin20 2224 (C,T,M) 1 .5170.002 g 1.8820.023 a 5.107 0.067 a 3.144 0.078 a 1.962 0.046 1 *D, C, and T represent one of the top scenarios in structural diversity, carbon s tocks and merchantable timber production. M represents the top scenarios with multiple benefits ** The value s with the same letter are not significantly different

PAGE 49

49 Figure 2 1 Forest Vegetation Simulator (FVS) processing as used in simulating different silvicultural regimes for converting even aged plantations to uneven aged stands.

PAGE 50

50 A B Figure 2 2 Input stand conditions inventoried in 2009. A). Pinus elliottii exhibit ed bell shaped diameter distribution typical of even aged stands. However, due to dense stand conditions and absence of any hardwood control measures, Magnolia virginiana ha d grown de nsely as an understory and was fast becoming a part of midstory. B). More than 95 percent of basal area wa s constituted by Pinus elliotti i The stand basal area wa s 28.7 m 2 ha 1

PAGE 51

51 Figure 2 3 Sensitivity of structural diversity to regeneration when the stand is treated under different harvest types, basal area s and cutting cycles Figure 2 4 Sensitivity of carbon s tocks to regeneration when the stand is treated under different harvest types, basal areas and cutting cycles

PAGE 52

52 Figure 2 5 Sensitivity of total merchantable timber produced to regeneration when the stand is treated under different harvest types, basal areas and cutting cycles Figure 2 6 Sensitivity of saw timber production to regeneration when the stand is treated under different ha rvest types, basal areas and cutting cycles

PAGE 53

53 Figure 2 7 Change in stand structural diversity (Shannon index) during the simulation period in top scenarios when maintained at higher basal area (11.5 m 2 ha 1 ) Figure 2 8 Change in stand structural di versity (Shannon index) during the simulation period in top scenarios when maintained at a lower basal area (4.6 m 2 ha 1 )

PAGE 54

54 CHAPTER 3 UNDERSTORY LIGHT DYNAMICS IN LONGLEAF PINE SLASH PINE ECOSYSTEMS IN CONVERSION Restoration and management of longleaf pin e ( Pinus palustris ) ecosystems of the southeastern United States is currently of high ecological and economic concern (Alavalapati et al., 2002; Brockway et al., 2005b; Jose et al., 2006 a ; Kirkman et al., 2007). Longleaf pine ecosystem types range from xer ic uplands to hydric flatwoods, and vary in overstory composition from pure longleaf to mixed species stands. In fact, in hydric flatwoods slash pine ( Pinus elliottii ) may comprise a significant component of the overstory. However, all of these ecosystems are characterized by some of the highest species richness among ecosystems of North America, mainly attributable to an understory wherein over 40 species per square meter have been reported (Peet and Allard, 1993). These ecosystems once occupied an estima ted 37 million hectare (ha) in the southeastern United States, with 23 million of pure longleaf pine and 14 million ha of mixed pine species, but have been reduced to a fraction of their original extent mainly due to conversion to commercial pine plantatio ns or through degradation as a result of fire exclusion (Frost, 1993; Landers et al., 1995; Frost, 2006). Thus, currently, these ecosystems have been designated as critically threatened (Means and Grow, 1985; Noss, 1989; Noss and Peters, 1995) and strategi es to restore and manage them are being developed. Successful restoration and management of these diverse ecosystems, however, is dependent upon an understanding of the biophysical processes that affect species performance. The opening of the canopy follow ing cutting treatments generally lead s to an increase in the availability of light (Palik et al., 1997; McGuire et al., 2001; Gagnon et al., 2003), soil moisture (Brockway and Outcalt, 1998), and nitrogen (Palik et al., 1997).

PAGE 55

55 Cutting method and intensity also influences the amount and spatial distribution of fuels within the stand (Brockway and Outcalt, 1998; Brockway et al., 2006; Mitchell et al., 2009), which is important in these fire driven ecosystems because of the effect on fire dynamics and subseque nt reductions in competing vegetation. All these environmental factors and their interactions may affect the survival and growth of longleaf pine regeneration and the maintenance of understory species richness. While soil moisture may be critical to longle af pine regeneration in xeric sandhills sites (Brockway and Outcalt, 1998), light is certainly an important factor in mesic sites with richer soils where soil moisture and nutrient availability are not limiting factors (Palik et al., 1997; McGuire et al., 2001; Gagnon et al., 2003). In general, light and its distribution across a stand have been determined to be a good indicator of regeneration, growth, and maintenance of understory species richness in these ecosystems (Palik et al., 1997; McGuire et al., 2 001; Gagnon et al., 2003; Palik et al., 2003; Gagnon et al., 2004). Light is one of the primary factors that limit growth of longleaf pine seedlings both in artificial (Gagnon et al., 2003) as well as naturally created gaps (Gagnon et al., 2004). Greater u nderstory light has also been associated with a greater abundance of understory vegetation, particularly herbaceous plants (Wolters, 1973; Wolters, 1981; Platt et al., 2006), and increased growth in basal area and crown width in longleaf pine (Harrington a nd Edwards, 1999; Harrington, 2006). However, diversity in growing conditions may be needed to maintain the broad range of species in the understory in these ecosystems (Harrington, 2006). Various silvicultural management systems, including uneven aged rep roduction methods such as single tree or group selection methods, as well as traditional even or

PAGE 56

56 two aged shelterwood methods, are being proposed as appropriate techniques to restore and sustainably manage these ecosystems for the provision of multiple be nefits including timber production and biodiversity because they are suggested to mimic natural disturbance regimes that historically maintained these ecosystems (Masters et al., 2003; Brockway et al., 2005a,b; Van Lear et al., 2005; Jose et al., 2006 a ). H owever, these proposed management systems vary in the amount and distribution of residual basal area across a stand (Smith et al., 1997), which leads to varying level s of horizontal as well as vertical heterogeneity in the distribution of canopy gaps. This results in alteration of light availability not only at the stand but also at the tree level (Beaudet and Messier, 2002; Davi et al., 2008). Understanding these changes in light regimes created by different management systems is critical for guiding appro priate management actions. Use of light transmittance, typically measured by leaf area index (LAI), has recently gained popularity as a management tool to predict regeneration and tree growth following silvicultural activities (Lieffers et al., 1999; Hale, 2003; Yirdaw and date have examined light availability as a function of thinning in even aged stands. In most cases, effects of thinning on light regime were studied by simply observing relationships between transmittance and overstory basal area for both conifer and broadleaf forests (Kuusipalo, 1985; Cutini, 1996; Mitchell and Popovich, 1997; Hale, 2003; Davi et al., 2008) and occasionally as a function of both stand de nsity and basal area (Hale et al., 2009). More sophisticated, spatially explicit models to predict canopy transmittance and understory light have also been developed for some conifer species,

PAGE 57

57 such as MAESTRO for Sitka spruce (Wang and Jarvis, 1990) and tRA Yci for Douglas fir (Brunner, 1998). Still, the effect of thinning in even aged stands may not be the same as using uneven aged methods that can alter both the horizontal and vertical distribution of the canopy and residual basal area. Overstory understor y interactions may also operate differently in these stands (Oliver and Larson, 1996). The use of only basal area as a management recommendation ignores the spatial distribution of trees and gaps, and age or size related variations and thus may be unsuita ble for estimating light regimes in stands managed under two aged or uneven aged systems. Light regimes have been studied in longleaf pine ecosystems but have sometimes given conflicting results possibly as a result of using different sampling techniques o r due to differences in site conditions. Brockway and Outcalt (1998) reported that light availability in a naturally regenerated uneven aged longleaf pine forest in sandhill sites was uniformly distributed across the forest floor of entire canopy gaps beca use the light stands. However, other studies carried out in mature second growth longleaf pine stands on more productive sites have found significantly higher light availa bility in the gap centers than at the gap edges as a result of tree removal (Palik et al., 1997; McGuire et al., 2001; Battaglia et al., 2002; Battaglia et al., 2003; Palik et al., 2003). The most comprehensive study among these was conducted by Battaglia et al. (2002, 2003) who estimated light availability in longleaf pine stands treated to three different retention harvest treatments (single tree, small group, and large group selection) and found that spatial structure of the longleaf pine stand strongly regulated understory light and its distribution across the stand. In that study, treatments differed in spatial distribution of

PAGE 58

58 residual trees but not residual basal area (11.9 to 12.3 m 2 ha 1 ) and the large group selection cut led to the highest light ava ilability. Since proposed management systems for longleaf pine encompass a broader range of residual basal areas from as low as 5 m 2 ha 1 (for shelterwood systems) to approximately 13 m 2 ha 1 (for selection systems) at the stand level (Franklin, 1997; Broc kway et al., 2005a; Brockway et al., 2006; Johnson and Gjerstad, 2006), there is still a gap in knowledge regarding the entire range of proposed management systems and canopy biophysical variables such as LAI and canopy cover fraction related to light regi mes in forest stands. Additionally, restoration of plantations to natural longleaf pine ecosystems often results in different mixes of species composition during the transition phase. Longleaf pine ecosystems also often occur naturally as mixed stands with slash pine in its distribution range (Peet and Allard, 1993). Because canopy transmittance may vary among species (Canham et al., 1994; Yirdaw and Luukkanen, 2004), light availability in mixed stands could be different from pure stands for a given basal a rea. Kirkman et al. (2007) reported that the canopy transmittance characteristics are significantly different in longleaf pine and slash pine stands. Thus the understanding of understory light availability in longleaf pine ecosystems as influenced by basal area, spatial distribution of trees, and overstory composition is required to fine tune silvicultural recommendations so that favorable light conditions could be created in the stands. The first objective of this study was to quantify the effects of silvi cultural systems on understory light regimes in longleaf pine ecosystems. We characterized light regimes in longleaf pine stands managed under single tree selection, group selection, and shelterwood system s along with an uncut control. The second objectiv e of the study

PAGE 59

59 was to evaluate whether light regimes in these ecosystems were affected by overstory species composition (proportional stocking of longleaf pine and slash pine). The study was carried out in a variety of stand conditions in north west and no rth central Florida, and employed high resolution Digital Hemispherical Photography (DHP) for greater accuracy (Frazer et al., 2001). Material and Methods S tudy Sites The study was carried out at the Blackwater River State Forest (BRSF), Goethe State Fores 3 1). Blackwater River State Forest (30.94N, 86.81W) is a well drained middle coastal plain upland site covering about 848.1 km 2 in the western Florida panhandle. It has an elevation of a pproximately 60 m above mean sea level (amsl) with a nearly level to gently inclined topography (0 5 percent). Soils are deep well drained and sandy, low in organic matter and nutrients and low to moderate in water holding capacity. Site index for longleaf pine was estimated to be 25 m at 50 years (Brockway, personal communication). The study site was occupied by second growth longleaf pine forest that naturally regenerated following clearcutting of the original forest during 1920s and was subjected to impr ovement cuts in 1981 and 1991prior to installation of the study treatments as well as some salvage harvesting in 2005 following hurricane Ivan. The overstory was dominated by longleaf pine with lesser components of slash pine and sporadic hardwoods (primar ily oaks ( Quercus spp.)). Tree seedlings common in the understory were mostly oaks and longleaf pine. Understory vegetation consisted of variety of shrubs, primarily dangleberry ( Gaylussacia frondosa ), blueberries ( Vaccinium spp.), blackberries ( Rubus spp. ), wax myrtle ( M orella cerifera ), and gallberry ( Il ex

PAGE 60

60 glabra ). The herbaceous layer was well developed and consisted of a variety of forbs, such as silverthread goldenaster ( Pityopsis graminifolia ), morning glory ( Ipomoea spp.), noseburn ( Tragia urens ), an Hypericum spp.) and grasses, such as wiregrass ( Aristida beyrichiana ), broomsedge bluestem ( Andropogon virginicus ), and witchgrass ( Dichanthelium spp.). The site had been managed under prescribed burning since 1970. The prescribed fire r egime had been winter season burning on a two to three year cycle since 1970, and was changed to late summer burning in 1995 and finally to spring burning in 2003 (Brockway and Outcalt, 2010; Brockway, personal communication). Goethe State Forest (29.19N, 82.57W) covers about 217 km 2 of poorly drained lower coastal plain flatwood s site in north central Florida. The stands used in the study had an elevation of approximately 15 m amsl with nearly level (0 2 percent) topography. The soils belong ed to Smyrna so il series, which were deep and characterized by poor drainage, low organic matter and nutrients and low water holding capacity. Site index of these areas ranged from 21 to 25 m at 50 years for longleaf pine. On stands clearcut in the late 19th and early 20 th century, sporadic planting was done during the period of 1940 to 1980 with subsequent thinnings between 1997 and 2004. The overstory was dominated by longleaf pine with some slash pine and sporadic hardwoods (mostly oaks). Tree seedlings were infrequent in the understory and included mostly longleaf pine, slash pine, water oak ( Quercus nigra ) and sweetgum ( Liquidambar styraciflua ). Shrubs, primarily saw palmetto ( Serenoa repens ) and gallberry, dominated the understory vegetation, and the herbaceous layer consisted of some grasses, mainly wiregrass, witchgrass, broomsedge bluestem, nodding fescue ( Festuca obtusa ) and

PAGE 61

61 panic grass ( Panicum spp.). Forbs were infrequently present. The prescribed fire regime is currently winter season burning on a three year cy cle (Brockway and Outcalt, 2010; Brockway, personal communication). The site experienced an aggressive lightning initiated fire in April June 2011 which burned nearly 2000 ha of forest area, affecting some of the study plots and severely burning one of the control plots. 2 of poorly drained lowland hydric flatwood s site between the Apalachicola and Ochlockonee rivers in panhandle Florida. It has elevation ranging from 0 to 10 m amsl with nea rly level topography. Four poorly drained hydric soil types account for the majority of the soils at this site namely, (a) Scranton Rutlege, (b) Plummer Surrency Pelham, (c) Meadowbrook Tooles Harbeson, and (d) Pamlico Pickney Maurepas (Gilpin and Vowel l, 2006). The site was once a swampy mosaic of wetlands and pine flatwoods communities, but large scale silvicultural operations and hydrological manipulations during 1960s through 1980s converted extensive areas to slash pine plantation using intensive me chanical site preparation, bedding and planting at high densities. Current restoration efforts include thinning planted pines and shrubs and conducting prescribed burns to restore and maintain a fire frequency of every one to four years since 1994. The stu dy stands consisted of pure slash pine plantations subjected to prescribed burns 1 to 3 times in the past 10 years. The understory mainly consisted of titi ( Cliftonia monophylla ), swamp titi ( Cyrilla racemiflora ), gallberry, and fetterbush ( Lyonia lucida ), saw poorly developed in denser stands and well developed in open stands and consisted mainly of bracken fern ( Pteridium aquilinum ), flatsedges ( Cyperus spp.), beak rushes

PAGE 62

62 ( R hynchospora spp.), wiregrass, bluestem grass ( Andropogon spp.), and yellow eyed grass ( Xyris spp.). Stand Characteristics and Treatments To examine the effect of management system on light availability, we utilized the long term experimental treatments est ablished by Brockway and Outcalt (2010) at Blackwater River and Goethe State Forests. Single tree selection (STS), group selection (GS), and shelterwood (SW) cutting treatments were each replicated on three 9 ha treatment plots, with basal area ranging fro m 10.3 to 12.9, 9.1 to 12.6, and 4.2 to 7.4 m 2 ha 1 respectively across the study sites(Table 3 1). Goethe State Forest had an additional three replications of shelterwood treatments compared to Blackwater, which were intended to further evaluate tradition the traditional shelterwood treatment had not taken place when our measurements were taken, we considered these treatments equivalent and grouped the treatment plots in our analyses (Table 3 2). Longleaf pine represented 30% to 100% of the overstory in different stands with mostly slash pine and hardwood trees as the remainder. At Blackwater River and Goethe State Forests, tre e marking for selection systems was based on the Proportional B method (Brockway and Outcalt, 2010; Brockway et al., 2011) where the proportion of residual basal area among three diameter classes (<15 cm, 15 30 cm, and >30 cm) is retained in the ratio of 1 :2:3. In all of the cutting treatments, slash pine was selectively marked and harvested in the stands to increase the proportion of longleaf pine in the stands. The single tree selection cuttings removed individual trees to create openings for regeneration with gap sizes less than 0.1 ha. The group selection cuttings created gaps in the stands ranging from 0.1 to 0.8 ha in size

PAGE 63

63 and of variable shape. Shelterwood treatments were applied using a seed cut that reduced the basal area uniformly across the stand. No cutting treatments were done in the experimental control. To evaluate the effect of species composition at various levels of residual basal described above. Stands her e were established in the 1970s and consisted of mature pure slash pine plantations which were either uniformly thinned to varying levels of residual basal area or were unthinned. Basal area ranged from 5.5 to 40 m 2 ha 1 Tree marking in thinned plots foll owed the principles of low thinning (Smith et al., 1997). Data Collection Digital Hemispherical Photography (DHP), recognized as one of the most accurate and robust techniques for studying canopy transmittance and understory light availability (Battaglia e t al., 2003; Valladares and Guzmn, 2006; Garrigues et al., 2008; Khabba et al., 2009), was employed to quantify understory light availability. We used a Nikon 5000 camera equipped with a Nikon fisheye FC E8 0.21X lens and calibrated it to compute the opti cal center of the camera fisheye lens system (Baret, 2004). The camera had an adaptable rotating LCD panel which was pulled out from the camera body, swung out horizontally 180 degree, and rotated up and over so that it faced the same direction as the lens towards the canopy. Orienting the camera lens assembly magnetic north, we placed a two way spirit level over the LCD panel to take perfectly horizontal photographs ( Figure 3 2 ). This method eliminated the need of using a tripod to level the camera, which can be cumbersome in forest with dense understory. The photographs were taken with F1 setting of the camera at the widest zoom setting (Nikon User Manual). All the photographs were acquired by a single person, ensuring that all

PAGE 64

64 photographs were acquired at a single height of 1.6 m from forest floor. The pictures were recorded and saved in TIF format at the highest possible resolution (2560 x1920 pixels, 14.1 megabytes). The DHPs were acquired in Dec 2009, June 2010, and later in October 2011 (Table 3 2), ne arly 3 5 years after the last reproduction cuttings. To collect light data as related to management systems, we utilized only the experimental treatment plots in Blackwater River State Forest and Goethe State Forest. For each management system or uncut con trol, we randomly assigned 25m x 25m measurement plots across any of the treatment plot replications. Each measurement plot was inventoried for its composition and basal area, and a DHP was acquired at the center. The number of measurement plots varied by cutting treatments and site, and ranged from 8 to 26 per treatment at Blackwater River and 9 to 48 at Goethe State Forests (Table 3 2). We established substantially higher number s of measurement plots on some treatments than other s and particularly in she lterwood treatment plots at Goethe State Forest because they provided opportunities for data collection on light regimes in various proportions of mixed slash pine longleaf pine to be used for the second part of our study. Similarly, control plots at Black water River Forest and single tree selection at Goethe State Forests were largely of pure longleaf pine. Group selection plots at Goethe State Forest also had a relatively greater number of measurement plots to account for the heterogeneous spatial distrib ution of residual trees in these treatments. For additional analysis to evaluate light availability with regard to different locations in the group selection plots at Goethe State Forest, we classified these measurement plots into three classes, namely, 1) gap, 2) matrix, and 3) new regeneration cohort. We defined matrix portion as the portion of group selection stands

PAGE 65

65 at least 10 m distant from a gap edge. This led to allocation of 8 gap, 17 matrix, and 3 regeneration measurements plots out of a total of 3 3 plots in group selection, while 5 measurement plots did not meet criteria to fall in any of these classes because they were not located at least 10 m away from the gap edge. In October 2011, we also revisited a heavily burned control plot in Goethe State Forest to assess the effect of an April 2011 wildfire on understory light. To further assess the effects of overstory species composition on understory light we also randomly established twenty three 25m x 25m measurement plots in five pure slash pine sta Data Analysis We used CAN_EYE V5.0 software (Avignon, INRA, France) to analyze the DHPs. CAN_EYE allowed interactive classification of the images exploiting the three colors viz., red, blue and green. After the pixels were classified either as sky or vegetation (leaves and branches), gap fraction was computed for 10 of azimuth angle and 5 of zenith angle sectors, which was a substantial improvement over previous studies where an azimuth and zenith resolution of 20 was use d (Battaglia et al., 2003). Additionally, only the portions of DHPs lower than 60 zenith angle were considered to prevent problems associated with a large fraction of mixed pixels containing both gaps and foliage (Davi et al., 2008) (Figure 3 3 ). To minim ize the observation errors, all the photographs were processed by one person using the same classification and analysis criteria. The following estimates were obtained from CAN_EYE: (1) proportion of sky visible through the canopy (called sky), (2) daily i ntegrated black sky fAPAR, also called direct fAPAR, (3) white sky fAPAR, also called diffuse fAPAR, (4) true leaf area index (LAI), and (5) cover fraction. These variables, calculated from gap fraction

PAGE 66

66 measurements (Weiss et al., 2004; Weiss and Baret, 20 10), are often used to quantify light regimes in forest ecosystems (Canham, 1988; Canham et al., 1990; Roxburgh and Kelly, 1995; Khabba et al., 2009). Sky simply refers to the proportion of pixels representing sky in an upward looking canopy hemispherical photograph and thus is a measure of canopy openness. fAPAR is the fraction of Absorbed Photosynthetically Active Radiation by the vegetation and is actually constituted of direct and diffuse components, also called black sky fAPAR and white sky fAPAR respe ctively (Weiss et al., 2004). The ratio of these components may influence the ratio of red and far red radiation (Pecot et al., 2005). LAI is defined as one half the total leaf area per unit ground surface area (see Jonckheere et al., 2004). LAI is related to Photosynthetically Active Radiation (PAR) (400 700nm) transmission as given by the Beer Lambert law (Vose et al., 1995). Cover fraction, on the other hand, is the fraction of sky covered by vegetation viewed in the nadir direction integrated over a ra nge of zenith angles (0 100). Overall, greater sky implies greater light availability at the forest floor while greater LAI, fAPAR, and cover fraction generally mean lesser canopy transmittance to the forest floor. Among the various measures of LAI compute d by CAN_EYE, we chose true LAI to present in our results because it takes into account the clumping index of the overstory species. Also, it may be pointed out that the term LAI has been used as proxy for PAI (Plant Area Index) because all parts of the tr ees including stems, leaves and branches are accounted for by CAN_EYE in the analysis (Weiss and Baret, 2010). Effect of management systems on light availability : Mean and coefficient of variation for each of the five response variables (sky, cover fractio n, LAI, direct fAPAR,

PAGE 67

67 diffuse fAPAR) were calculated for each of the management systems and control plots at Blackwater River and Goethe State Forests, and analysis of variance (ANOVA) was carried out to evaluate the effect of management on each response v ariable for each site separately, with each measurement plot forming one replication of the treatment. The number of replications varied among the treatments (Table 3 differences. Similarly, ANOVA were carried out to evaluate the effect of position (gap, matrix, and regeneration cohort) within group selection plots, and the effect of wildfire on each Effect of overstory species composition on light a vailability : For this portion of the analysis, we utilized the measurement plots across each of the three study sites, which differed in proportional overstory species composition as well as basal area. Regardless of site or cutting treatment, we selected and classified the measurement plots into the following five categories based on proportional overstory species composition: (1) pure longleaf pine (>90 % basal area constituted by longleaf pine pine), (2) 70% longleaf pine (65 75 % basal area constituted by longleaf pine), (3) 50% longleaf pine (45 55 % basal area constituted by longleaf pine), (4) 30% longleaf pine (25 35 % basal area constituted by longleaf pine), and (5) pure slash pine (>90% basal area constituted by slash pine). A total of 164 measure ments consisting of 94 for pure longleaf pine, 22 for 70% longleaf pine, 14 for 50% longleaf pine, 11 for 30% longleaf pine, and 23 for pure slash pine were considered. We then used analysis of covariance

PAGE 68

68 (ANCOVA) with the five proportional species composi tions as a categorical variable, each of the response variables as a dependent variable, and basal area as the as the covariate (Crawley, 2005). All the data were analyzed using the statistical program R 2.14.0 or XLstat 2012 (http://www.xlstat.com, Addins oft). Results Effect of Management S ystems As would be expected, the cuttings associated with the management systems at both Blackwater River State Forest and Goethe State Forest led to significant decreases in LAI, direct fAPAR and indirect fAPAR, and inc rease in sky compared to the uncut control plots (p<0.001). At Blackwater River State Forest, cover fraction was lower in only group selection as compared to the uncut control (Table 3 3). Although the mean values of some of the response variables differed between the two sites, overall trends between the management systems and patterns of the coefficient of variation were largely consistent between the two sites. LAI at both sites averaged 1.7 1.8 for the control plots and ranged from 0.3 to 0.9 for the va rious management systems. At Blackwater River State Forest, the group selection and shelterwood treatment resulted in significantly higher values of sky than single tree selection and control, and correspondingly lower values for LAI, and direct as well as diffuse fAPAR (Table 3 3). The group selection showed the highest variability (coefficient of variation) in the values of cover fraction, LAI, and direct as well as diffuse fAPAR, while sky was more variable in single tree selection Though the ratio of d irect: diffuse fAPAR was not significantly different among the various management systems, the greatest variability was again observed in group selection system at this site.

PAGE 69

69 At the Goethe State Forest, shelterwood treatment had significantly higher value of sky and lower value of direct fAPAR than both of the other treatments and the control. As at Blackwater, LAI and diffuse fAPAR were similar in shelterwood and group selection but were significantly lower than single tree selection and control. Cover fra ction did not differ significantly among the treatments, but was lower than the control. The coefficient of variation was highest in group selection only for LAI, direct fAPAR and diffuse fAPAR. At this site, shelterwood treatment had the highest coefficie nt of variation in cover fraction and ratio between direct to diffuse fAPAR, while single tree selection observed highest coefficient of variation for sky. Within the group selection plots at Goethe State Forest, the position with regard to gap, matrix, an d regeneration cohort significantly altered light conditions (p<0.001). The new cohorts of regeneration in gap openings, after nearly five years since the last reproduction cuttings, led to a significant decrease in sky and increase in cover fraction, dire ct fAPAR, and diffuse fAPAR, within the regeneration clusters compared to open gap area and matrix portion in the group selection plots. LAI in regeneration clusters was similar to the matrix portion but, as expected, higher than at gap portion. Sky and co ver fraction in regeneration clusters were also significantly higher than in the uncut control s whereas LAI and diffuse fAPAR were significantly lower. Direct fAPAR were similar in regeneration clusters and uncut control (Figure 3 4 and 3 5 ). The incident of wildfire in the control plot of Goethe in 2011 significantly increased understory light availability (p<0.001) by reducing basal area to approximately 2.7 m 2 ha 1 This resulted in significant changes in all of the response variables (values of LAI = 0. 16; cover fraction= 0.02; direct fAPAR= 0.10; diffuse fAPAR= 0.10) as compared to

PAGE 70

70 the unburned control. The mean and variability of light conditions created by the wildfire were closest to the shelterwood systems in Goethe State Forest. Effect of Overstory Species C omposition Overstory species composition and basal area had a significant effect on understory light availability (p<0.001) (Table 3 4). For a given basal area, pure longleaf pine stands had the highest amount of light availability while pure sla sh pine stands had the least understory light availability. Sky, LAI, and diffuse fAPAR differed significantly in pure longleaf pine stands and pure slash pine stands, while cover fraction was similar among different composition categories. Direct fAPAR sh owed inconsistent patterns possibly due to difference in dates of DHP acquisition. However, total fAPAR (obtained by summing direct fAPAR and diffuse fAPAR for each observation) was also found to differ significantly between pure slash pine and pure longle af pine. Stands containing longleaf pine and slash pine in varying proportions of basal area generally resulted in significant reductions in light availability relative to pure longleaf pine only when the proportion of slash pine in the overstory was 70 pe rcent or higher. Discussion and Conclusions Comparison of Management Systems on L ight T ransmittance Understory light availability in forest ecosystems can be critical for successful tree regeneration and growth as well as maintenance of understory species richness and abundance (Canham and Marks, 1985; Parrotta, 1995; Yirdaw and Luukkanen, 2004; Platt et al., 2006). This is of particular concern in ecosystems with high biodiversity and where the principal overstory species are light demanding, such as are slash and longleaf pine (Wahlenberg, 1946; McGuire et al., 2001; Platt et al., 2006). The management challenges in longleaf pine ecosystems are to regulate the canopy

PAGE 71

71 structure and light transmittance in a manner that will create favorable light conditions for successful regeneration and maintenance of understory biodiversity while also retaining sufficient overstory stocking to provide fine fuels to maintain periodic fires or to provide some acceptable level of timber production where desired. Historically longleaf pine has been managed as even aged stands and natural regeneration has been successfully obtained using shelterwood system. More recently, however, there has been significant interest in using other regeneration systems that mimic smaller scale natural disturbance in longleaf pine ( Brockway et al., 2005b) Our comparisons between management systems showed a predictable response in both the mean and variability of the light responses as can be attributed to the level of residual basal area and the distribution of residual trees across the stands for each cutting method applied. Not surprisingly, shelterwood systems resulted in the highest understory light availability and generally with the least amount of variability as these treatments had the lo west residual basal area and residual trees were purposefully left distributed evenly across the stands leading to homogenous canopy openings. Interestingly, the means of several light responses, e.g. fAPAR (direct and diffuse) or LAI, in the shelterwood s ystem were not significantly different from those of group selection at Blackwater River State Forest or Goethe State Forest, where group selection treatments had greater stand basal area. That result likely occurred because the mean values for group selec tion included data from both open gaps as well as matrix forest such that the high light conditions from open gap conditions skewed the mean values downward and high variability made mean separation difficult Differences in light availability between gap and matrix forest were quite large in some cases

PAGE 72

72 (Figure 3 4 and 3 5 ) and thus led to the highest variability in light responses as a result of the aggregated distribution of trees following group selection cuts. Single tree selection, which maintained the same amount of residual basal area at the stand level as the group selection, in general had smaller variability in light measurements on par with shelterwood method as a result of the more even dispersal of residual trees. Though the measures of light av ailability used in our study did not allow us a direct comparison to other studies (using different measures of light) in longleaf pine and other forest types, the overall patterns appear to be in alignment with these studies. For example, understory light availability was found to vary inversely with increasing overstory basal area in longleaf pine (Kirkman et al., 2007), and Sitka spruce and Scots pine (Hale et al., 2009). Also, for the same level of residual basal area, mean stand light availability (mea sured as gap light index) in longleaf pine increased significantly from 56 percent in the single tree cut to 63 percent in large group harvest (Battaglia et. al, 2003; Palik et al., 2003). Similarly, in another study conducted in France on a variety of for est types, LAI decreased from 60 to 33 percent with a similar decrease in basal area, with higher LAI variability resulting from group and seed cuts than the uniform cuts (Davi et al., 2008). While uniform higher light levels can have positive benefits for many species in longleaf ecosystems, heterogeneity of light across a stand has also been suggested to be more conducive to creating, improving, or maintaining understory species diversity, stand structural diversity, microenvironment and habitat (Holmes a nd Smith, 1977) by availability at the stand level than other management systems (Palik et al., 2003).

PAGE 73

73 Spatial aggregation in group selection systems, which creates l arger gaps, also exposes the stand to the maximum length of time over which understory and related fauna can respond to increased light because larger openings are likely to fill in slower than the smaller openings created by other types of systems (Spruge l et al., 2009). After 5 years post cutting in the group selection plots, dense clusters of new cohorts of longleaf pine regeneration originated in the centers of the gaps. While the large gaps were favorable for longleaf regeneration, the reduction in lig ht availability we observed could, at least temporarily, have a negative effect on localized understory biodiversity enhancement and maintenance. However, this negative effect would be localized to the gap centers and might not have a large impact on biodi versity across the entire stand. Our field observations and preliminary data three years following the reproduction cuttings suggest that group selection cuttings produce desirable conditions similar to those of the traditionally used shelterwood cuttings by promoting longleaf pine regeneration and enhancing understory species richness (Brockway and Outcalt 2010; Brockway, personal communication). On the other hand, long term management of longleaf pine following group selection may have more concerns than short term regeneration and understory dynamics. For example, as Mitchell et al. (2009) has noted, group selection might cause disruption s in fuel continuity and create hard edges which in turn might affect fire behavior leading to unpredictable consequen ces. Only long term studies will establish the suitability of group selection systems in managing longleaf pine ecosystems. Further, management recommendations typically use stand level measures of stocking, such as basal area or more currently LAI as surr ogates for regulating resource

PAGE 74

74 availability; however, variability in light conditions such as would be created by group selection may also have strong implications for understory dynamics. For example, regeneration and growth responses under a mature fores t canopy are not linearly related to LAI and thus use of mean values may not be appropriate where variability of LAI across a stand is high (Davi et al., 2006). Also, the proportion of direct and indirect radiation can affect the ratio of red (660 nm) and far red (730 nm) light (R: FR) in the stand as diffuse radiation is low in far red light (Holmes and Smith, 1977; Pecot et al., 2005). This ratio has been reported to regulate important developmental processes including seed germination, specific leaf area and stem elongation (Pecot et al., 2005). Although the reproduction cutting techniques did not differ significantly in mean ratio of direct to diffuse fAPAR in Blackwater River State Forest, the group selection cuttings did show the highest variability. The ratio has been reported to vary in longleaf pine ecosystems with respect to sky conditions, overstory stocking, and solar angles (Pecot et al., 2005). Though, our study did not examine these aspects, the inconsistency in the pattern of direct and diffu se fAPAR observed at the Goethe State Forest in contrast to Blackwater River State Forest was possibly because the data from the Goethe State Forest was collected in different months of year (June and October) when sun angles were different. The shelterwoo d and uneven aged management systems evaluated in our study were designed to mimic natural disturbances that have historically governed the dynamics of the natural longleaf pine ecosystems. For example, the shelterwood method represent s circumstances where a partial stand of longleaf pine is left following a catastrophic event such as wildfire or damage from tropical storms, while group

PAGE 75

75 selection and single tree selection represent s small er scale localized disturbances such as those resulting from insect an d pathogen attacks or lightning strikes (Brockway and Outcalt 2010; Brockway, personal communication). These disturbances create canopy gaps that allow higher resource availability at the forest floor resulting in longleaf pine regeneration (Platt et al., 19 8 8 b ). Although the change in resource availability as a result of disturbance will vary with its intensity and frequency, the incident of wildfire at Goethe State Forest led to increased availability of light, more than any of our treatments. The monito ring of understory succession and longleaf pine regeneration in the wildfire affected stand at Goethe State Forest offers an opportunity for comparison with the treatments implemented at Goethe State Forest. Effect of Overstory Species C omposition The prop ortion of longleaf pine and slash pine in the overstory altered light conditions especially at higher proportions of slash pine. The capacity of a species to intercept light is affected by its crown structure including characteristics and spatial distribut ion of needles, shoots and branches (Stenberg et al., 1994). Clearly, the crown morphology, needle length and foliage clustering are quite distinct between the two species. While longleaf pine has a sparse porous crown consisting of long needles, slash pi ne has dense crown of short er needles. In total, longleaf pine typically has less leaf area than slash pine for trees of the same sapwood area (Gonzalez Benecke et al., 2011). These characteristics allow longleaf pine stands to transmit more light than tho se of slash pine. Light at a given point in a measurement plot (photopoint) was possibly affected by overstory composition outside the plot. Though we tried to control this effect by selecting measurement plots in more uniform stands, some variability in t he observations is bound to occur. This probably has resulted in high variability in our

PAGE 76

76 results leading to significant changes in light availability observed only when slash pine occurred to the extent of 70 percent or more in the overstory. More studies are required to tease out the specific difference in transmittance in these ecosystems. Given that much of the area currently under slash pine plantations in the southeastern United States was formerly occupied by longleaf pine, restoration of these sites will involve a series of mixed stand stages of slash pine and longleaf pine in varying proportions during the process (Kirkman et al., 2007). Currently, the pure and mixed stands are managed under a basal area regulation approach without regard to overstor y composition. As suggested in both our study and in Kirkman et al. (2007), the residual basal area required to maintain favorable understory light regimes in mixed stands will be different than in the pure stand s of longleaf pine or slash pine because of the difference in light transmittance of these species. Thus a basal area regulation approach must account for overstory composition to create desirable light regimes while restoring and managing these ecosystems. However, our results seem to suggest that appropriate recommendations based on species proportions may only be most critical during early restoration efforts when slash pine is the most dominant component of the overstory. Conclusions Our study examined the canopy structure and light transmittance 3 5 years following three different management systems. In general, management systems that create larger gaps in the canopy, i.e. group selection and shelterwood, result in greater amounts of light available to the understory. Differences in both the qua ntity and variability of understory light availability suggest that there may be trade offs in determining which management systems may be most appropriate for meeting multiple

PAGE 77

77 objectives including tree regeneration and growth as well as understory diversi ty. For instance, shelterwood methods which led to high, uniform light levels may be conducive for promoting significant tree regeneration and establishment, while group selection system which created both overall greater amounts as well as variability in light conditions likely to result in a more heterogeneous and diverse understory species response. The group selection system also retains greater stocking than the shelterwood system. We suggest that understory and longleaf pine regeneration response shou ld be monitored over a long period in these treatments to establish the suitability of group selection in managing longleaf pine ecosystems. Additionally, since the relative proportion of longleaf pine and slash pine affected the light conditions especiall y at higher proportions of slash pine, the residual basal area should be adjusted appropriately in mixed stands to create favorable light conditions.

PAGE 78

78 Table 3 1. Summary of the stands treated to different reproduction cutting systems at Blackwater River St ate Forest and Goethe State Forest, FL, USA (compiled from Brockway and Outcalt, 2010; Brockway, personal communication). The values before and after the comma (,) represent the Blackwater River State Forest and Goethe State Forest respectively Reproducti on cutting method/Management system Control Single tree selection Group selection Shelterwood Origin(year) 1920, 1935 48 1920s, 1935 48 1920s, 1935 48 1920s, 1935 61 Time of reproduction cuttings Nov Dec 2006 Nov Dec 2006 Nov Dec 2006 Nov Dec 2006 A pproximate slope (%) 0 2, 0 5 0 2, 0 5 0 2, 0 5 0 2, 0 5 Mean BA prior to cut (m 2 ha 1 ) 11.0, 15.6 14.0, 16.5 16.7, 16.7 11.7, 14.8 15.3 Mean BA in 2009 3 years after cut (m 2 ha 1 ) 10.3 13.3, 16.0 18.3 10.3 12.6,10.8 12.9 11.5 12.6, 9.1 11.5 5.8 7.4, 4.2 5.0 Approximate species composition (% basal area constituted by LLP)* 85 95, 30 70 70 80, 80 100 70 75, 40 90 70 90, 65 90 Site index (m) at 50 years 24.4, 21.30 24.0 24.4, 21.3 24.4, 21.3 22.2 24.4, 21.3 24.3 The values represent the approxim ate percent basal area constituted by longleaf pine (LLP) at stand level. The remaining basal area was mostly constituted by hardwoods and some slash pine at Blackwater River State Forest, and slash pine at Goethe State Forest. The measurement plots had va riable composition ranging from 25% to 100% longleaf pine by basal area.

PAGE 79

79 Table 3 2. Number of measurement plots and dates of acquisition of Digital Hemispherical Photographs (DHPs) at Blackwater River State Forest (BRSF) and Goethe State Forest (GSF) use d in the analysis of the effect of management system on light availability. Reproduction cutting method/ Management system* Number of measurement plots/treatment Date of acquisition of DHPs BRSF GSF BRSF GSF Control 26 9 Dec 2009 June 2010 STS 9 2 4 Dec 2009 June 2010, Oct 2011 GS 8 33 Dec 2009 Oct 2011 SW 10 48** Dec 2009 June 2010, Oct 2011 *STS= S in gle tree selection system, GS= G roup selection system SW= S helterwood system **Shelterwood treatment at Goethe State Forest consisted of uniform s helterwood and irregular shelterwood system. At the time of our study, these treatments were similar and thus grouped into a single shelterwood treatment. This also resulted in high number of measurement plots relative to the other treatments.

PAGE 80

80 Table 3 3 Estimates (Mean SE ) of leaf area index (LAI), fraction of Absorbed Photosynthetically Active Radiation (fAPAR), cover fraction, and direct fAPAR: diffuse fAPAR ratio in longleaf pine slash pine stands treated to different reproduction cutting systems at Bl ackwater River State Forest (BRSF) and Goethe State Forest (GSF), FL, USA The coefficient of variation is given in the parentheses. The values of a response variable with the same letter did not differ significantly. Site Management system* Basal area (m 2 ha 1 ) Sky Cover fraction LAI Direct fAPAR Diffuse fAPAR Direct fAPAR: Diffuse fAPAR BRSF Control 17.661.17 (33.65) 0.470.01 a (8.20) 0.300.03 b (46.69) 1.810.06 c (17.10) 0.790.01 c (5.97) 0.280.00 c (4.68) 2.850.01 a (2.39) STS 11.220.46 (12.29) 0.6 80.02 b (10.91) 0.190.04 a b (67.64) 0.920.11 b (36.85) 0.550.04 b (19.73) 0.200.01 b (18.68) 2.740.01 a (1.59) GS 9.823.38 (68.74) 0.760.02 c (9.22) 0.050.04 a (222.53) 0.350.13 a (104.30) 0.300.10 a (89.31) 0.110.02 a (55.79) 2.411.25 a (147.34) SW 7.580.91 (37.93) 0.760.01 c (3.14) 0.200.07 a,b (104.14) 0.500.06 a (39.14) 0.390.02 a (13.91) 0.140.01 a (19.91) 2.870.11 a (11.70) GSF Control 22.072.61 (35.55) 0.490.01 a (6.24) 0.190.04 b (57.41) 1.730.12 c (20.09) 0.460.02 d (10.48) 0.550.01 c (7.4 0) 0.830.01 a (4.97) STS 11.290.52 (22.37) 0.730.01 b (7.09) 0.070.02 a (116.01) 0.620.05 b (39.84) 0.290.01 c (15.70) 0.290.02 b (26.85) 1.060.05 a,b (21.35) GS 9.830.90 (35.70) 0.790.01 c (5.82) 0.070.02 a (143.09) 0.370.04 a (67.01) 0.220.01 b (31 .41) 0.190.01 a (35.02) 1.210.02 b (11.87) SW 4.420.48 (52.16) 0.840.00 d (3.26) 0.050.01 a (149.68) 0.280.02 a (52.99) 0.170.01 a (26.69) 0.150.01 a (30.05) 1.140.05 b (27.41) *STS= Single tree selection system, GS= Group selection system, SW= Shelter wood system

PAGE 81

81 Table 3 4. Effect of overstory species composition on understory light availability (least square means) in longleaf pine slash pine stands pooled from data collected State Forest, FL, USA. The values of a response variable w ith the same letter did not differ significantly Proportional overstory composition* Sky Cover fraction LAI Direct fAPAR Diffuse fAPAR Total fAPAR* Pure LLP 0.68 b 0.16 a 0.92 a 0.43 b,c 0.23 a 0 .66 a 70 % LLP 0.71 b 0.10 a 0.84 a 0.32 a 0.29 a 0.60 a 50% LLP 0.69 b 0.15 a 0.77 a 0.35 a,b 0.28 a 0.62 a 30 % LLP 0.59 a 0.15 a 1.26 a 0.53 c 0.29 a 0.82 b Pure SLP 0.55 a 0.16 a 2.38 b 0.40 a,b,c 0.46 b 0.86 b *Total fAPAR was obtained by summing direct fAPAR and diffuse fAPAR for each observation

PAGE 82

82 Figure 3 1. Location of the three study sites at Blackwater River State Forest, Goethe Figure 3 2. A cquisition of digital hemispheric al photograph using Nikon 5000 FC E8 camera Photo courtesy of Joseph Culen.

PAGE 83

83 Figure 3 3. Digital hemispherical photograph (left), classified image (middle), and gap fraction image (right) of longleaf pine stands managed under different systems viz ., A) uncut control, B) shelterwood C) group selection, and D) single tree selection

PAGE 84

84 Figure 3 4. Understory light conditions (sky, cover fraction, direct fraction of Absorbed Photosynthetically Active R adiation (fAPAR), and diffuse fAPAR) in various areas of the group selection system plots as compared to the uncut control plots approximately 5 years following reproduction cuttings at Goethe State Forest, FL, USA. The values with the same letter did not differ significantly.

PAGE 85

85 Figure 3 5. Understory light conditions (leaf area index (LAI), and ratio of direct fraction of Absorbed Photosynthetically Active Radiation (fAPAR) and diffuse fAPAR) in in various areas of the group selection system plots as co mpared to the uncut control plots approximately 5 years following reproduction cuttings at Goethe State Forest, FL, USA. The values with the same letter did not differ significantly.

PAGE 86

86 CHAPTER 4 SEED BANK DYNAMICS AND UNDERSTORY RESTORATION IN PINE FLATWOOD S E CO SYSTEMS Pine flatwoods ecosystem s are among the most species rich containing the highest concentrations of threatened and endangered species in the southeastern United States (Peet and Allard, 1993; Cohen et al., 2004; Jose et al., 2006 b ). Due to the reduction in their historical extent caused by intensive pine plantations, land conversion, and fire exclusion, the species rich understor ies in these ecosystems have been replaced by shrubby hardwoods. While restoration of the overstory can be accomplishe d by planting and / or stand conversion, the restoration of a species rich understory poses a real challenge (Cohen et al., 2004). In degraded forest stands where a natural species rich understory has been lost, the seed bank may provide a valuable opportuni ty for restoration ( Simpson et al., 1989 ; Cohen et al., 2004 ). A s eed bank contains propagules of species, which may be desirable or undesirable which can germinate and colonize after restoration activities such as thinning, burning, or other mechanical t reatments following decades of suppression (Abella and Springer, 2008). A s eed bank containing desirable species, if present, can prevent or diminish the costs associated with sowing and transplanting of native species. Seed banks become more valuable when geographically suitable seed of the desired species is not readily available. Additionally, natural seed dispersal of desired understory species may be unreliable when rare community types are isolated (Van der Valk and Pederson, 1989; Augusto et al., 200 1). T he presence of a seed bank and its composition and structure could be an important clue to understanding the demography of past and potential future ecosystem s For example, seed banks can inform about past aboveground vegetation

PAGE 87

87 and histories of inva sion and disturbance, and overlapping generations (Price et al., 2010). In addition, knowledge of seed bank size and composition can facilitate proactive management of restoration and ecosystem management by providing early warnings of potential exotic spe cies which could become management concern (Glass, 1989). Despite the high importance of seed banks in restoration efforts very few studies have fully characterize d them, especially in pine dominated communities of the southeastern United States (Carringt on, 1997; Maliakal et al., 2000; Cohen et al., 2004; Coffey and Kirkman, 2006; Ruth et al., 2008; Andreu et al., 2009). Of those studies done, variable results across sites suggest that a seed bank with desirable species and density can n ot always be relied upon for in situ recruitment of understory in degraded sites For example, Cohen et al. (2004), observed a persistent seed bank (seed viability more than a year) in longleaf pine ( Pinus palustris ) ecosystems of the Carolina Coastal Plains h owever, seed b ank s lacked native desirable species in longleaf pine stands in Georgia (Coffey and Kirkman, 2006; Andreu et al., 2009) and Florida (Ruth et al., 2008 ) Several variables may explain discrepancies between results as the longevity of seeds in soil varies wi th site conditions, cultural operations, moisture level and a variety of other factors related to management (Warr et al., 1993 ; Bossuyt and Honnay, 2008). T hese studies also differed in their sampling designs and procedures including sampling at differen t depths and during different seasons. Some s tudies have shown that seed bank composition can vary dramatically with depth in the soil (Warr et al. 1993) and this may be especially true in degraded sites For example, it is possible that a greater concent ration of seed of desirable species may only be present in deeper layers in long degraded sites A s upper soil and litter layers are added later during the

PAGE 88

88 degraded phase these layers may lack propagules of historical vegetation as those species die back The studies assessing seed bank only in top thin layer of soil profile in excessively degraded site might not have accounted for potential seed bank in the deeper layers. Several studies have attempted to understand relationships between the seed bank and aboveground vegetation (Cohen et al., 2004; Ruth et al., 2008; Andreu et al., 2009) or other parameters such as basal area, canopy cover, litter depth, and light levels ( Warr et al 1993 ; Abella and Springer, 2008). In general, little correlation has bee n found between the seed bank and existing aboveground vegetation in degraded sites. However, i n consistently disturbed sites influenced by fire and harvesting, there is generally a greater correlation between aboveground vegetation and the seed bank (Warr et al. 1993). This suggests that seed bank dynamics and its relationship with aboveground vegetation may differ during different stages of restoration as different disturbance factors interact to influence succession. An understanding of seed bank ecolog y relevant to restoration of pine flatwoods ecosystems is critically lacking. It i s very difficult to theorize seed bank dynamics in pine flatwoods ecosystems from a few isolated studies made in the southern pine ecosystems. In order to ensure the success of restoration projects which are often very expensive and time consuming it is imperative that we have sound quantified information about the status of seed banks and understanding of the factors that are potentially related to seed banks such as stan d parameters, and successional stage. In this study, we examined the soil seed bank in wet to mesic pine flatwoods sites of Florida panhandle Three stand conditions representing a restoration gradient

PAGE 89

89 degraded, partially restored, and restored were exa mined for their seed bank and aboveground vegetation Materials and Methods Study Area The study was conducted at the in Northwest Florida (Figure 4 1). It consists of about 820 km 2 of poorly drained lowla nd mesic hydric flatwood s sites between the Apalachicola and Ochlockonee Rivers in Florida panhandle The elevation ranges from 0 to 10 m amsl with nearly level topography. The climate is humid subtropical with annual precipitation totaling about 147 cm, o f which about 49% is received during June to September (Gilpin and Vowell, 2006) Although more than 40 unique soil types occur within the forest, four groups account for the majority of the soils, namely, (a) Scranton Rutlege, (b) Plummer Surrency Pelha m, (c) Meadowbrook Tooles Harbeson, and (d) Pamlico Pickney Maurepas (Gilpin and Vowell, 2006) All are poorly drained hydric soils. The site was once a swampy mosaic of wet prairies, cypress ( Taxodium spp.) sloughs, Atlantic White Cedar ( Chamaecyparis thy oides ) forests and other wetland and pine flatwoods communities. Large scale silvicultural operations and hydrological manipulations during 1960s through 1980s converted extensive areas of native habitats to slash pine plantation. As many as 1300 km of unp aved logging roads were constructed during this time. Roadside ditches were excavated to drain excessive water from the adjacent wetland habitats and the pine stands were established following intensive mechanical site preparation, bedding and planting at high densities. Fertilizers, particularly Nitrogen and Phosphorus, were applied at mid rotation. These large scale disturbances and manipulations to the habitats altered the ecological communities and hydrology of the

PAGE 90

90 area (Gilpin and Vowell, 2006). Curren Hell State Forest is to restore native pine ecosystems on these plantations and maintain biological diversity, while integrating public use. The Florida Forestry Service has begun restoring some of these land s by thinning planted pines and shrubs and conducting prescribed burns to restore and maintain a fire frequency of every one to four years since past few decades. Depending on the intensity of restoration activities (prescribed burning and thinning), a var iety of stand conditions at different stages of restoration currently exist Study S tands We identified stands along a restoration gradient from highly degraded conditions to stands with desirable conditions, and arbitrarily classified them into the follo wing three stand conditions (Table 4 1) Degraded: H ighly degraded condition s included three unthinned stands of mature slash pine plantation established in 1970s (Figure 4 2a). These stands ha d never been burned since establishment, except one stand that w as accidentally burnt once about 5 years prior to th is study. The understory consisted of thick vegetation of hardwood species mainly black titi ( Cliftonia monophylla ), white titi ( Cyrilla racemiflora ), gallberries ( Ilex spp.), fetterbush ( Lyonia spp.) an Hypericum spp.). Occasionally, pond cypress ( Taxodium ascendens ), bald Cypress ( Taxodium distichum ), black gum ( Nyssa sylvatica ) and sweet bay ( Magnolia virginiana ) appeared in the overstory mainly localized in patchy depressions, and a lso formed a major component of the midstory. The herbaceous g round layer was very poorly developed to non existent. Partially restored: Th ese conditions included three stands of mature slash pine plantations established in 1970s which had been subjected t o low to moderate

PAGE 91

91 restoration activities (Figure 4 2b) T he restoration activities started in early 2000 when these stands were thinned and later prescribed burned in 2006, 2008, and 2011. T he understory consisted of moderate cover of shrubs like gallberri es, fetterbush, inkberries ( Vaccinium consisted of species like meadows beauty ( Rhexia spp.), bracken fern ( Pteridium aquilinum ), flatsedges ( Cyperus spp.), saw palmetto ( Serenoa rep ens ), bluestems ( Andropogon spp.), wiregrass ( Aristida stricta ) etc. Restored: R estored conditions included three slash pine plantation stands established during 1960 70s which had been subjected to intense restoration activities in past two decades (Figur e 4 2c) The stands had been subjected to heavy thinnings forming open savanna and had been maintained on a three year prescribed burn regime. The ground layer was species rich and consisted of a variety of grasses, sedges, and a large diversity of forbs c onsidered representative of wet flatwoods communities (Florida Natural Areas Inventory, 2010) The primary ground layer species were wiregrass, gallberry, witchgrasses ( Dichanthelium spp.) pineland rayless goldenrod (Bigelowia nudata), saw palmetto, Brack en fern, meadows beauty, flatsedges, beakrushes, bluestem grasses etc. Vegetation a nd Soil S eed B ank S ampling We used a modified version of the sampling strategy used by Andreu et al. (2009). A total of 9 random overstory plots of size 50 m x 2 0 m were est ablished, one in each of the nine stands representing 3 conditions along a restoration gradient (Figure 4 3 ). Within each of these 50 m x 2 0 m overstory plots, four replicate sets of three circular nested plots (100 m 2 for midstory, 10 m 2 for understory an d 1 m 2 for herbaceous cover) were established to quantify midstory, understory, and herbaceous plant communities

PAGE 92

92 respectively. Diameter at breast height (DBH) for all woody species cm DBH was measured in overstory plots. In midstory plots (circular, 100 m 2 ), diameter of all woody species < 10 cm DBH and > 1. 37 m tall was recorded. In understory plots (circular, 10 m 2 ), all woody species <1. 37 m tall were recorded and percent foliar cover per species was visually estimated and placed into one of the 11 cover scale classes: <1%, 1 5%, 6 10%, 11 15%, 16 20%, 21 25%, 26 35%, 36 50%, 50 75%, 75 95%, and >95%. Similarly, percent foliar cover of all herbaceous species was recorded in 1m 2 p lots. Heights of the shrub layer and the midstory were measured as their modal values. Additionally, litter depth was measur ed at 12 random points in each 50 m x 2 0 m plot. The vegetation sampling was done in June 2011, and was repeated in October 2011 to positively identify a few species while in flowering stage in October. Soil samples were collected in June 2011 at the same time when the vegetation sampling was done first. A total of 25 soil cores including duff and litter were taken in a nested plot, fi ve randomly within one meter of each of five systematic positions at the center and four cardinal quadrants in each nested plot (Figure 4 3 ) using a 2 cm diameter soil core to a depth of 15 cm These 25 soil cores per nested plot were carefully fractionate d (to avoid contamination) and pooled into 3 depth classes: 0 5 cm, 5 10 cm, and 10 15 cm resulting in 3 soil samples for each nested plot, one for each depth class. Each 50 m x 20 m plot, thus, had a total of 12 samples Since there were three s tand condi tions with three replications of the larger 50m x 20m plots there were a total of 108 soil samples for the study. Soil samples were then transported to the laboratory in separate sealed plastic bags and cold stratified at 3.3 o C for 40 days.

PAGE 93

93 Seedling Emerg ence Method The s eedling emergence method, one of the most common and reliable method s for seed bank assessments (Thomson and Grime, 1979; Price et al., 2010), was used for this study After cold stratification, each sample was thoroughly mixed and a subsa mple of 300 g (240 260 cm 3 ) was obtained and germinated in germination trays (25 cm x 25 cm x 10 cm approx.) using Fafard 3B mix (45% peat, 25% pine bark, 10% vermiculate, 20% perlite) in an institutional greenhouse at the University of Florida. For germ ination, we evenly spread the soil seed sample over the mix which formed an approximately 0.4 cm thick layer. Additionally we placed 5 trays containing only the Fafard 3B germination mix as experimental control. The trays were kept in the greenhouse in a single tier on iron benches at a height of 1m and arranged randomly to minimize biases caused by small variations in light and moisture intensities (Cohen et al. 2004). They were watered regularly. An exclusive and dedicated greenhouse chamber was used in the study to prevent any chance of contamination from the extraneous sources. The greenhouse was subjected to natural photoperiod a nd temperature ranging between 5 o C (lowest in winter) to 36 o C (highest in summer) over the observation period. The temperatu re, light, and photoperiod closely matched with those of natural habitat conditions (personal observation). As the seedlings emerged, they were identified, counted and removed on a biweekly basis. Species difficult to identify at emergence stage were trans planted and grown until they could be identified. Professional services for species identification were solicited from the University of Florida Herbarium, Florida Museum of Natural History, and nomenclature largely followed Wunderlin and Hansen (2003). Th e trial continued for 10 months before the greenhouse was damaged by a tree fall due to tropical storm

PAGE 94

94 by this time. Some of the species are still being grown in pots to f lowering stage for positive identification. Data Analysis W e characterized the aboveground sampled vegetation for its basal area, density, species richness, and percent cover We also calculated seed density, species richness, and total germinants from the emerged seedlings in the seed bank for each depth class and s tand condition (Magurran, 1988). The 50 m x 2 0 m plot was considered a replication, and thus every s tand condition had three replications. Analysi s of variance was carried out to test the main effects of stand conditions and soil depth and their interactions on seed density 0.05. 2a/ (b + c) ) was used to compare the species composition of seed bank and the ex isting aboveground the number of species common to both the seed bank and the aboveground vegetation, species detected in the seed bank and the corresponding aboveground vegetation, respectively (Cox, 1985; Arroyo et al., 1999). Results Above ground Vegetation The basal area in the degraded s tands was approximately 40% higher than the partially degraded s t ands and about 4 times higher than the restored s tands (Table 4 2). T he dominant species in the overstory in all of the s tands was slash pine which accounted for about 90 95% in degraded and partially degraded s tands and 100% of the basal area in restored s tands In degraded s tands which w ere mature unthinned

PAGE 95

95 plantation s few hardwood trees had regenerated to become part of overstory and midstory. The p artially restored s tands had rare occurrence s of hardwood in the midstory while the restored s tands did n ot have any hardwood species in the overstory or midstory. The hardwood species primarily consisted of black gum, pond cypress, sweet bay, and titi which had established in the low depressions in in the degraded s tands and constituted about 5 10% (2.33 m 2 ha 1 ) of basal area. Partially restored and restored s tands had about 2% (0.33 m 2 ha 1 ) and 7% (0.51 m 2 ha 1 ) basal area constituted by midstory. The midstory in restored s tands was slash pine only. Many of the species in the understory were common among the three s tand conditions but they varied in percent cover. The dominance of shrubs was evident in degraded s tands though most of the shrubby species had grown above 1.37 m height and so were classified as midstory (Table 4 2) Nevertheless, there was a s ubstantial shrubby layer in degraded s tands primarily consisting of gallberry, fetterbush, giant gallberry ( Ilex coriacea ), titi, and sweet pepperbush ( Clethra alnifoila ). The p artially restored s tands due to a recent burn, had less shrub cover in June bu t had developed dense shrubby cover by end of the growing season in October primarily consisting of gallberry, fetterbush, St. John s worts, dangleberries ( Gaylussacia spp.) and sweet bay. The restored s tands w ere the most diverse with the highest overall foliar cover mainly contributed by the herbaceous layer with high similarity to the characteristic wet flatwoods communities (Table 4 3) (Florida Natural Areas Inventory, 2010) Though the overall shrub cover was less in restored s tands it had few additi onal shrub species compared to the other s tand conditions consisting of dangleberries, blueberries, saw palmetto etc.

PAGE 96

96 A clear pattern of transition in dominance from shrubs to herbaceous layer (grasses, herbs, forbs) was evident from degraded to restored s tand conditions While d egraded and p artially r estored s tands had high shrub cover (20 80%) the r estored s tands had as much as 150% herbaceous cover formed by overlapping herbaceous plants and a maximum of 25% shrub cover (Table 4 2 ). The r estored s tands w ere dominated by native species typical of natural pine wet mesic flatwoods sites consisting of wiregrass, saw palmetto, blueberries, gayfeathers ( Liatris spp.), yellow eyed grasses, and a variety of other sedges, rushes and grasses (Peet and Allard, 1993 ; Jose et al., 2006b ; Florida Natural Areas Inventory, 2010 ). Several rare species were also observed, most notably bear grass ( Nolina atopocarpa ), and saltmarsh umbrella sedge ( Fuirena breviseta ). Overall, 2 4 59, and 124 species were identified in degrad ed, partially restored, and restored s tand conditions A complete species list is summarized in Table 4 3. Seed Bank Structure and Composition A total of 3 131 seedlings representing 83 species were recorded over the 10 month period of observation (Table 4 4) Two germinants of American sweet gum ( Liquidambar styraciflua ) were found in the control germination trays containing only the germination mix, so this species was excluded from analysis. Some notable species that germinated were small butterwort ( Ping uicula pumila ), a highly rare species, observed in restored site. Other notable species were Drosera spp., and salt marsh umbrella sedge No tree species was observed. The main effect of s tand conditions on seed density as well as seed species diversity (S hannon index) was not significant. A ll the three s tand conditions had almost similar total seed densit ies (Table 4 4) but the number of species that germinated

PAGE 97

97 increased from degraded (26) to partially restored (39) to the restored s tand condition which ha d the highest number of species (64). Notably, although species richness in the seed bank was higher in partially restored and restored s tands than the degraded s tands the degraded s tands had a substantially higher above ground to seed bank species richnes s ratio (Figure 4 4) Only a f ew species constituted the majority of the germinants, primarily grassleaf rush ( Juncus marginatus ) clustered mille grains ( Oldenlandia uniflora ) St. Johns wort, primrose willow ( Ludwigia spp .), and ovateleaf f latsedge ( Cype rus ovatus ) (Table 4 5) Approximately two third of the total germinants were of only these species. On the other hand, white violet ( Viola primulifolia ), saltmarsh umbrella sedge dense blazing star ( Liatris spicata ) fringed meadowbeauty ( Rhexia petiolat a ) fairy beaksedge ( Rhynchospora pusilla ) fewflower beaksedge ( Rhynchospora rariflora ) and American burnweed ( Erechtites hieracifolia ) formed only a few seedlings. One of the most striking observations in the study was the number of germinants and speci es distributed across the depth in the soil profile. While partially restored and restored s tands had the highest concentration of seeds and species in the topmost 0 5 cm layer (144 168 seeds/kg soil) degraded s tands had the highest density (142 seeds/kg soil) in the middle 5 10 cm layer (Table 4 4). In all of the cases, however, the lowest number of germinants was observed at a depth of 10 15 cm. The main effect of soil depth was significant for both seed density (F (2, 18) = 4.926, p=.019) and species div ersity (F (2, 18) = 4.789, p=.0215). However, seed density (p=0.020) and seed species diversity (p=0.017) were significantly different only between soil depths of 0 5 cm and 10 15 cm. The s tand condition *soil depth interaction effect was not significant for both

PAGE 98

98 seed density and seed species diversity. The higher seed density in 5 10 cm layer in the degraded s tands was mostly contributed by grassleaf rush (27%) St. Johns wort (23%) and primrose willow (19%) (Table 4 5) all of which did not exist above groun d on the degraded s tands at the time of sampling. Although the seed density in degraded s tands was higher in 5 10 cm deep layer, diversity, as evidenced by Shannon index, was highest in 0 5 cm layer and decreased with increasing depth across all stand cond itions (Table 4 4). Relationship between Aboveground Vegetation and Seed Bank Out of the total 83 species germinated from the seed bank, only 23 were common to the aboveground vegetation across all the s tand conditions The greatest correspondence between aboveground vegetation and seed bank was in the restored and partially restored s tands and extremely low in degraded s tands (Table 4 6) Most notably, wiregrass was absent in seed bank from the restored s tands where it formed the most dominant ground cover Slash pine did not germinate from the seed bank in all the s tand conditions though it was the most dominant species forming the overstory across all the s tands Similarly, rare species like small butterwort was observed in the seed bank when it was abse nt in the aboveground vegetation. Bear grass, a rare species observed in aboveground vegetation was not observed in seed bank. In all of the stand conditions except the degraded, species richness of the aboveground vegetation was higher than the seed bank (Table 4 3 Figure 4 4 ). The Sorensen similarity index indicated that the similarity between the aboveground vegetation and soil seed banks was the low est in degraded s tands (0. 13 ) P artially restored (0.29) and restored (0.28) s tands also had low correspo ndence between seed

PAGE 99

99 bank and aboveground vegetation, but were higher than the degraded s tands (Table 4 6) Discussion and Conclusions Aboveground Vegetation The typical pine flatwood ecosystem is described as an open savanna of pines over a carpet of grass es, sedges, and a large divers ity of forbs (Peet and Allard, 1993 ; Florida Natural Areas Inventory, 2010 ). The restored site in our study fits well in this description. These desirable conditions in our restored sites have been maintained by low residual b asal area and frequent low moderate intensity prescribed fires. A good mix of w iregrass, saw palmetto, sedges, rushes and a variety of other forbs are typical of desirable understory and ground cover conditions in wet mesic flatwoods communities ( Peet and Allard, 1993 ; Jose et al., 2006 b ; Florida Natural Areas Inventory, 2010 ). In the degraded s tands the formation of a dense hardwood midstory and the loss of fire adapted shade intolerant understory species are promoted by the high basal area in these unt hinned pine plantation stands as well as the exclusion of fire for a long time (Frost, 2006; Jose et al., 2006 b ). An exclusion of fire for as little as 7 10 years from a closed canopy pine community could result in loss of the ground cover (Frost, 1990). B lack gum, titi, gallberries, and fetterbush are typical hardwoods that colonize these degraded conditions in wet mesic flatwood s ecosystems (Peet and Allard, 1993). The partially restored s tands had characteristics in between those of degraded and restored s tands The se stands had moderate level s of basal area, shrub cover, species richness, and burn history. Despite the three prescribed burns in past 5 years on this s tands the relatively higher density of shrubs is likely due to the fact that the

PAGE 100

100 burns we re mostly dormant season which tend to promote the persistence of shrubs more so than growing season burns (Lewis and Harshbarger, 1976 ; Waldrop et al., 1992; Drewa et al., 2002 ). Seed Bank Structure and Composition Th ough we calculated seed density per m ass basis of soil th e seed densities and the species richness observed in this study are comparable to the range reported in other studies made on southern pines F or example, 0 204 /m 2 seed constituting 10 taxa in 5 cm deep soil samples were detected in sand pine scrub ecosystems (Carrington, 1997), and in another study 0 170 /m 2 seed s were detected in sand pine scrubs and longleaf pine flatwoods containing 28 taxa in top 5 cm layer (Ruth et al., 2008). A dditional studies have observed 56 (Andreu et al. 2009) and 43 species (Cohen et al. 2004) in the seed bank collected from loblolly longleaf pine forests. However, these comparisons must be made with caution as these studies involved different sampling intensities, sizes, soil depth, and season, which ca n have substantial effect on the results (Warr et al., 1993 ; Bossuyt and Honnay, 2008 ). We also observed a high amount of similarity in the species observed in seed banks in our study and some other studies in similar sites (Cohen et al., 2004; Ruth et al. 2008) such as grassleaf rush, St. Johns wort, sundews, rosette grasses, small butterwort, polygala or milkwort, fringed meadowbeauty, beakrushes, juniper leaf ( Polypremum procumbens ) and yellow eyed grasses. A review of a number of studies in Europe has shown that the dominating genera in the seed banks are rather similar across different communities (Bossuyt and Honnay, 2008). Additionally, Juncus spp. has been found to be the most dominant species in seed bank of longleaf pine flatwoods in North America (Ruth et al., 2008) and a variety of ecosystems in Europe including marshes, forests, heathlands and

PAGE 101

101 grasslands (Bossuyt and Honnay, 2008). Another notable observation is the absence of wiregrass in seed bank germination studies made in southern pine ecos ystems, even when wiregrass was observed in sampled vegetation (Maliakal et al., 2000; Cohen et al., 2004; Ruth et al., 2008; Andreu et al., 2009). The absence of wiregrass and some other dominant aboveground species such as narrowleaf silkgrass ( Pityopsis gramin i folia ) could be due to the effects of season of burn or due to lack of some special germination requirements such as extended cold stratification or high temperature (Platt et al., 1988a; Clewell, 1989). Like other studies, herbaceous species domin ated the soil seed banks in our study. The high density of seeds of herbaceous species in the seed bank may be related to species traits as seeds of herbaceous plants have been reported to stay viable for longer periods of time (Ghersa and Martinez Ghers a, 2000) and thus have greater probability of being observed in the seed bank. Additionally in dense stands such as degraded site conditions in our study, the increasingly shaded conditions may inhibit germination of light demanding herbaceous species resu lting in their greater seed density in soil. P revious studies in other forest types have reported lower seed densities with increase in depth in the soil (Warr et a l., 1993; Thompson et al., 1997; Price et al., 2010). This was true for partially restored a nd restored s tands in our study too. However, higher seed density in the lower 5 10 cm layer than 0 5 cm in the degraded s tands in our study may be due to of the buildup of a high amount of litter. Seed density in the litter/humus layer above the soil is s tated to be extremely variable (Warr et al., 1993). In a study in plantation stands in China, although greater seed densities were

PAGE 102

102 recorded in upper 0 5 cm layer, higher density of herbaceous species were reported at 5 10 cm depth (Wang et al., 2009). In a nother study made in mixed spruce forests in central Estonia, negligible difference was observed in seed densities between 0 5 cm and 5 10 cm deep soil layers (Zobel et al., 2007). Although the composite soil sampling methodology tends to represent the ove rall population better (Warr et al., 1993 ; Bossuyt and Honnay, 2008 ) and was followed in the study, the reported seed bank density for flatwoods community in our study has some limitations The seedling emergence method tends to underestimate the densities of the viable seeds in the soil, as conditions for germination may not be met for some species (Simpson et al. 1989; Warr et al., 1993; Carrington, 1997; Price et al. 2010). One way of accounting for ungerminated seeds could be to subject the soil used in seed ling emergence trails after germination to seed extraction/floatation for detecting species that failed to germinate (Price et al. 2010). This was planned in the overall scheme of this study but could not be done because of the damage done to the s tudy greenhouse by a tree Further, after a lull in germination during the months of December to March, we did begin to observe germination of a few new additional species in April onwards. Some of those sp ecies were shortleaf rose gentian ( Sabatia brevifolia ), false b uttonweed ( Spermacoce sp. ) hyssopleaf sandmat ( Chamaesyce hyssopifolia ), and polygala ( Polygala sp. ). These were likely species that typically germinate in the spring under field conditions. A lthough we had already observed germination of over 95% of the species, it is possible that a very few additional species were missed because of the early termination of the study.

PAGE 103

103 Relationship between Aboveground Vegetation and Seed Bank M ost studies have indicated a generally poor correspondence between the existing aboveground vegetation and the soil seed bank across a range of the ecosystem types (Thompson and Grime 1979; Warr et al., 1993; Augusto et al., 2001; Bossuyt and Honnay, 2008; Ruth et al., 2 008; Andreu et al., 2009 ). Th e greater similarity observed in restored and partially restored s tands than the degraded s tands in our study is likely because of the frequent disturbances of prescribed fire and open light conditions in restored s tands Frequ ently disturbed sites generally have greater correspondence between seed bank and extant vegetation (Jensen 1969 ; Warr et al., 1993; Bossuyt and Honnay, 2008 ). The lack of similarity in degraded s tands may be because the light demanding, early successiona l species such as Juncus spp., Cyperus spp and Ludwigia spp. were lost from the ground flora due to high density intensive plantations but survived in the seed bank. Shade tolerant shrubby species like gallberry, fetterbush, and sweetbay generally reprod uce through vegetative means and produce generally require long cold treatment which may not have met in our study. Implications for Restoration of Understory The dete ction of 26 species in the seed bank of our degraded s tands is an encouraging finding suggesting seed bank can potentially be used to restore some of the divers ity of the understory. Although some of the species observed were ruderal, no exotic invasives w ere detected. The genera Dichanthelium Rhynchospora and Hypericum observed in the seed bank, are recognized as typical of wet flatwoods communities (Florida Natural Areas Inventory, 2010). The partial ly restored s tands in our study had a similar history as the degrade d s tands except for the recent restoration

PAGE 104

104 activities (thinning and introduction of prescribed burning) in past decade s, and these s tands had greater species richness in both the seed bank as well as aboveground vegetation. Though there is n o information about the initial seed bank of this partial ly restored s tands it does suggest that seed bank composition (or at least its response in seedling emergence) does change with progression of restoration. If it is any sign, we expect the degraded s tands to successfully restore the ground cover following the initiation of restoration activities. There could be some inputs from seed rain that could also influence the understory restoration. However, such assessments are very difficult to make or pred ict. Since greater seed density in the degraded s tands was observed at 5 10 cm depth in soil, the restoration activities that can intermix the soil to expose this layer can also aid in fast er restoration of the groundcover. For example, prescribed burning can burn 10 cm layer.

PAGE 105

105 Table 4 1. History of the three s tand conditions representing a restoration gradient at Site Origin Thinning hi story Burn years Time since last burn Burn characteristics Degraded 1976 77 Never thinned 2006 5 Only one stand burned,; Moderate 0.5 3m flame length,; Aerial head, flanking, backing Partially Restored 1976 3 rd row thinned Jan 2006, Feb 2008, April 20 11 2 months Intense 1 2m flame length,; Aerial and hand burn head, flanking, backing Restored 1964 Heavily thinned 2000, 2002, ,July 2006 Growing, Feb 2008, June 2008, June 2010 I year Moderate 1 2m flame length,; Hand burn head, flank, backing Tabl e 4 2. Attributes of the three s tand conditions representing a restoration gradient at the examination Site Degraded Partially R estored Restored Overstory basal area (m 2 ha 1 ) (1000 m 2 plot ) 30 22 7 Herbaceous cover (%) (1 m 2 plot) 0 20 30 90 80 150 Species richness ( species / m 2 ) 1 4 5 10 8 20 Understory shrub cover (%) (10 m 2 plot) 20 60 30 80 10 25 Midstory basal area (m 2 ha 1 ) (100 m 2 plot) 2.33 0.33 0.51 Mean litter depth (cm) 13.9 3.8 0.42 Average shrub height (m) 1.35 0.55 0.8

PAGE 106

106 Table 4 3. Presence/absence list from the vegetation sampling and seed bank examination from the three s tand conditions representing a restoration gradient at tate Forest, FL. (0 and 1 repre sent absent and present respectively) Species Degraded site Partially Restored Restored Vegetation Seed bank Vegetation Seed bank Vegetation Seed bank Acalypha sp. 0 0 0 1 0 1 Agalinus sp. 0 0 0 0 1 1 Agrostis sp. 0 0 1 0 1 0 Andropogon ar ctatus 0 0 1 0 0 0 Andropogon sp. 0 0 1 0 1 1 Andropogon sp. 0 0 1 0 1 0 Andropogon sp. 0 0 1 0 0 0 Andropogon virginicus 0 0 1 0 1 0 Aristida stricta 0 0 1 0 1 0 Bidens mitis 0 0 0 0 1 0 Bigelowia nudata 0 0 0 0 1 0 Bulbostylis ca ppilarris 0 0 0 0 0 1 Bulbostylis sp2. 0 1 0 0 0 1 Carex sp. 1 0 0 0 0 0 Cardamine flexuosa 0 1 0 1 0 1 Carphephorus odorattisimma 0 0 0 0 1 0 Carpheph o rus sp. 0 0 0 0 0 1 Centella asiatica 0 0 0 0 1 0 Chamaesyce hyssopifolia 0 1 0 1 0 1 Clethra alnifolia 1 0 1 0 1 0 Cliftonia monophylla 1 0 1 0 1 0 Ctenium aromaticum 0 0 0 0 1 0 Cyperus haspan 0 1 0 0 0 0 Cyperus ovatus 1 1 0 1 0 0 Cyperus sp 0 0 1 1 0 1 Cyperus sp2. 0 1 0 0 0 0 Cyrilla racemiflora 1 0 1 0 0 0 Dichanthelium 0 1 0 0 0 0 Dichanthelium ensifolium 0 0 0 0 1 0 Dichanthelium leucothrix 0 0 1 1 1 1 Dichanthelium sp2. 0 0 1 1 1 1 Dichanthelium sp3. 0 0 1 1 1 1 Dichanthelium sp4. 0 0 1 1 1 1 Dichanthelium sp5. 0 0 1 1 1 1 Dichanthe lium sp6. 0 0 1 0 1 1 Dichanthelium sp7. 0 0 0 0 1 1

PAGE 107

107 Table 4 3. Continued Species Degraded site Partially Restored Restored Vegetation Seed bank Vegetation Seed bank Vegetation Seed bank Dichanthelium strigosum 0 0 1 1 1 1 Digitaria bicorni s 0 1 0 1 0 1 Diodia sp. 0 0 1 0 0 0 Drosera spp. 0 0 0 0 0 1 Eleocharis sp. 0 0 0 0 0 1 Eleocharis sp2. 0 0 0 0 0 1 Erechtites hieracifolia 0 1 0 0 0 0 Erigeron vernus 0 0 0 0 1 0 Eriocaulon compressus 0 0 0 0 1 0 Eriocaulon decang ulare 0 0 1 0 0 0 Eupatorium compositifolium 0 1 0 1 1 1 Eurybia eryngifolia 0 0 0 0 1 0 Fuirena breviseta 0 0 0 1 1 1 Fuirena sp. 0 0 0 0 1 0 Galium sp1. 0 1 0 1 0 1 Galium sp2. 0 0 0 0 0 1 Gaylussacia dumosa 0 0 0 0 1 0 Gaylussaci a frondosa 0 0 1 0 1 0 Gaylussacia mosieri 0 0 0 0 1 0 Gaylussacia sp. 0 0 1 0 0 0 Gelsimium sempervirens 0 0 0 0 1 0 Helenium sp. 0 0 0 0 1 0 Helianthus angustifolius 0 0 0 0 1 0 Helianthus radula 0 0 0 0 1 0 Hibiscus aculeatus 0 0 0 0 1 0 Hieracium gronovii 0 0 0 0 1 0 Hypericum ci stifolium 0 0 0 0 1 0 Hypericum hyper i coides 0 1 0 1 1 1 Hypericum sp. 0 0 0 1 1 0 Hypericum sp2. 0 0 0 0 1 1 Hypericum sp3. 0 1 0 0 0 0 Ilex cassine 1 0 0 0 0 0 Ilex coriacea 1 0 1 0 1 0 Ilex glabra 1 0 1 0 1 0 Ilex myrtifolia 1 0 0 0 0 0 Ilex opaca 0 0 1 0 0 0 Juncus dichotomus 0 0 0 1 0 0 Juncus elliottii 0 0 0 0 1 1 Juncus marginatus 0 1 0 1 1 1

PAGE 108

108 Table 4 3. Continued Species Degraded site Partially Restored Re stored Vegetation Seed bank Vegetation Seed bank Vegetation Seed bank Juncus sp 1. 0 0 0 0 0 1 Juncus sp 2 0 1 0 1 0 1 Lachnanthes caroli ana 1 0 1 0 1 0 Lachnocaulon anceps 0 0 1 0 1 0 Lespedeza sp. 0 0 0 0 1 0 Liatris sp. 0 0 0 0 1 0 Liatris sp2. 0 0 0 0 1 0 Liatris sp3. 0 0 0 0 1 0 Liatris spicata 0 0 0 0 1 1 Lobelia paludosa 0 0 0 0 1 0 Ludwigia linifolia 0 0 0 1 0 1 Ludwigia maritima 0 0 0 0 1 0 Ludwigia sp1 0 1 0 1 1 1 Ludwigia sp2. 0 1 0 1 0 1 Ludwigia sp3 0 0 0 0 0 1 Ludwigia sp4. 0 0 0 1 0 1 Ludwigia virgata 0 0 0 0 1 0 Ludwigi a hirtella 0 0 0 0 1 0 Lyonia lucida 1 0 1 0 1 0 Magnolia virginiana 1 0 1 0 0 0 Morella cerifera 0 0 1 0 1 0 Morella inodora 1 0 0 0 0 0 Myrica caroliniens is 0 0 1 0 1 0 Nolina a to po carpa 0 0 0 0 1 0 Nyssa sylvatica 1 0 1 0 0 0 Oldenlandia sp2. 0 1 0 1 0 1 Oldenlandia uniflora 0 0 1 1 1 1 Osmanthus americanus 1 0 0 0 0 0 Osmundastrum cinnamomeum 0 0 0 0 1 0 Oxalis cornulata 0 1 0 1 0 1 Oxypolis filiformis 0 0 0 0 1 0 Panicum sp. 0 0 1 1 0 0 Panicum verrucosum 0 0 0 1 0 1 Persea palustris 1 0 0 0 1 0 P inguicula pumila 0 0 0 0 0 1 Pinus elliottii 1 0 1 0 1 0 Pityopsis graminifolia 0 0 1 0 1 0 Pluchea baccharis 0 1 0 0 1 1

PAGE 109

109 Table 4 3. Continued Species Degraded site Partially Restored Restored Vegetation Seed bank Vegetation Seed bank Vegetation Seed bank Pluchea sp. 0 0 1 0 1 0 Polygala cruciata 0 0 0 0 1 0 Polygala cymosa 0 0 0 0 1 1 Polygala lept ocaulis 0 0 0 0 1 0 Polygala lutea 0 0 0 0 1 0 Polygala ramosa 0 0 0 0 1 1 Polygala sp. 0 0 0 0 1 0 Polypremum procumbens 0 0 0 1 0 1 Pteridium aquilinum 0 0 0 0 1 0 Pterocaulon pycnostachyum 0 0 0 0 1 0 Quercus minima 0 0 0 0 1 0 Q uercus myrtifolia 0 0 0 0 1 0 Quercus sp. 1 0 1 0 0 0 Rhexia alifanus 0 0 1 0 1 0 Rhexia lutea 0 0 0 0 1 0 Rhexia mariana 0 0 0 0 1 0 Rhexia petiolata 0 0 1 1 1 0 Rhexia sp2. 0 0 1 0 1 0 Rhexia sp3. 0 0 1 0 1 0 Rhynchospora fasicula ris 0 1 1 0 1 1 Rhynchospora pusilla 0 0 0 0 1 1 Rhynchospora rariflora 0 0 0 0 0 1 Rhynchospora sp1. 0 1 0 1 1 1 Rhynchospora plumosa 1 0 0 0 1 0 Rubus erecta 0 0 1 0 0 0 Sabatia brevifolia 0 0 0 0 1 1 Saccharum sp. 0 0 0 0 1 0 Sch izacharium scoparium 0 0 0 0 1 0 Serenoa repens 0 0 1 0 1 0 Seriocarpus tortifolius 0 0 0 0 1 0 Smilax auriculata 0 0 0 0 1 0 Smilax bonanox 0 0 0 0 1 0 Smilax laurifolia 1 0 1 0 1 0 Smilax lucida 1 0 1 0 1 0 Smilax sp 1 0 1 0 0 0 Solidago odora 0 0 0 0 1 0 Solidago semp e r virens 0 0 0 0 1 0 Solidago sp1. 0 0 0 1 0 1

PAGE 110

110 Table 4 3. Continued Species Degraded site Partially Restored Restored Vegetation Seed bank Vegetation Seed bank Vegetation Seed bank Sphagnum spp. 1 1 1 1 1 1 Spermacoce sp1. 0 0 0 0 0 1 Stillingia aquatica 0 0 0 0 1 0 Symphyotrichum chapmanii 0 0 0 0 1 0 Symphyotrichum dumosum 0 0 0 0 0 1 Syng o nanthus sp. 0 0 0 0 1 0 Syngonanthus flavidulu 0 0 0 0 1 0 Taxodium ascendens 1 0 1 0 1 0 Taxodium distichum 0 0 0 0 1 0 Tephrosia hispidula 0 0 0 0 1 0 Unidentified 1 (monocot) 0 1 0 1 0 1 Unidentified 10 (dicot) 0 1 0 0 0 0 Unidentified 2 (Asteraceae) 0 0 0 0 0 1 Unidentified 3 (monocot) 0 0 0 0 0 1 Unidentified 4 (monoco t) 0 0 0 1 0 1 Unidentified 4 (monocot) 0 0 0 0 0 1 Unidentified 5 (monocot) 0 0 0 0 0 1 Unidentified 6 (monocot) 0 0 0 0 0 1 Unidentified 7 (monocot) 0 0 0 0 0 1 Unidentified 8 (monocot) 0 0 0 1 0 0 Unidentified 9 (monocot) 0 0 0 1 0 0 Vaccinium corymbosus 1 0 0 0 0 0 Vaccinium darrowii 0 0 1 0 1 0 Vaccinium myrsinites 0 0 1 0 1 0 Vaccinium sp 0 0 1 0 1 0 Vaccinium sp. 0 0 0 0 1 0 Viola primulifoila 0 0 0 0 0 1 Vitis rotundifolia 1 0 1 0 1 0 Wo o dwardia virginica 0 0 1 0 0 0 Xyris smalliana 0 0 1 0 1 0 Xyris sp. 0 0 1 1 1 1 Xyris sp1. 0 0 1 0 1 0 Xyris sp2. 0 0 0 0 1 0 Xyris sp3. 0 0 0 0 1 0 Xyris sp4. 0 1 0 0 1 0 Xyris sp5. 0 0 0 0 1 0 Total number of species 2 4 26 59 39 124 64

PAGE 111

111 Table 4 4. Estimates (means) o f s eed density, number of species, and diversity (Shannon Index) observed in the soil seed bank samples collected at different depths in the soil profile (0 5 cm, 5 10 cm, and 10 15 cm) from the three st and conditions represe nting a restoration gradient at Site Degraded Partially restored Restored Depth (cm) 0 5 5 10 10 15 0 5 5 10 10 15 0 5 5 10 10 15 Seed density (Seeds/kg) 90 142 47 168 84 37 144 112 46 Average number of specie s 6.5 5.9 4.58 8.08 5.83 3.58 10.3 8.25 4.08 Shannon Index 1.51 1.34 1.25 1.47 1.33 0.96 1.94 1.64 0.98 Total no. of seedlings 324 512 168 604 303 133 517 404 166 *The overall average seed density for Degraded, Partially Restored, and Restored S tands w ere 93, 96 and 101 respectively *The overall average Shannon Index for Degraded, Partially Restored, and Restored Sites were 1.36, 1.25 and 1.52 respectively.

PAGE 112

112 Table 4 5 The most common species germinated from the seed bank collected at different depths in the soil profile (0 5 cm, 5 10 cm, and 10 15 cm) from the three s tand conditions representing a restoration gradient Forest, FL Site Degraded Partially restored Restored Depth (cm) 0 5 Juncus marginatus (19%) Cyperus ovatus (1 2%) Ludwigia sp1. (11.5%) Juncus marginatus (43%) Oldenlandia uniflora (13%) Oldenlandia sp2. (10%) Xyris spp. (10%) Juncus marginatus (26%) Oldenlandia uniflora (18%) Oldenlandia sp2. (13%) 5 10 Juncus marginatus (27%) Hypericum spp (23%) Ludwigia sp1. (19%) Juncus marginatus (52%) Oldenlandia uniflora (12%) Oldenlandia sp2. (6%) Juncus marginatus (33%) Oldenlandia uniflora (20%) Oldenlandia sp2. (14%) 10 15 Hypericum spp. (20%) Juncus marginatus (19%) Ludwigia sp1 (15%) Cyperus ovatus (14%) Juncus mar ginatus (48%) Oldenlandia uniflora (8%) Cardamine flexuosa (8%) Juncus marginatus (44%) Oldenlandia uniflora (23%) Oldenlandia sp2. (7%) Table 4 6 vegetation for three s tand conditi ons representing a restoration gradient at Site similarity coefficient Degraded 0. 13 Partially Restored 0.29 Restored 0.28

PAGE 113

113 Figure 4 1. Location of the study area

PAGE 114

114 Figure 4 2. The three stand conditions representing a restoration gradient State Forest, FL A) Degraded stand condition B) Partially r estored s tand condition and C) Restored s tand condition Photos courtesy of author.

PAGE 115

115 Figure 4 3. Modified Andreu et al. (2009) sampling design used in the study site at out 1000 m 2 (50 m x20 m) plot. Four 100 m 2 circular nested plots were laid out systematically in the 1000 m 2 plot for midstory sampling. Within each 100 m 2 plot were laid one circular 10 m 2 nested plot (grid) for understory sampling, and one square 1 m 2 plots were laid within each of the 10 m 2 plots for sampling herbaceous plants. The soil seed bank samples wer e collected at locations marked with a dot; each dot represents a composite sample of five soil cores collected within 1m distance. For a nested plot, all the subsamples were fractionated into 0 5 cm, 5 10 cm, and 10 15 cm depths and then combined to form three samples, one for each 0 5 cm, 5 10 cm, and 10 15 cm depths. Thus, finally, each nested plot formed 3 samples and there were 12 samples for each plot.

PAGE 116

116 Figure 4 4. Aboveground seed bank species richness relationship observed in mesic wet flatwood s sites in different s tand conditions representing a restoration gradient species richness ratio increased from degraded s tands to restored s tands as restoration progressed.

PAGE 117

117 CHAPTER 5 SUMMARY A ND CONCLUSIONS Restoration of longleaf pine slash pine ecosystems ( Pinus palustris Pinus elliottii ) of the southeastern United States is of significant economic and ecological concern. These ecosystems which once dominated the southeastern United States covering about 37 million hectare s, have now reduced to a fraction of their original extent mainly due to their conversion to other pine monocultures and exclusion of fire in their management. Interest in the management and restoration of these ecosystems has renewed in the past decade s and best management practices for their management are being developed. There is high advocacy for managing these ecosystems using uneven aged silvicultural methods. This research has shown that longleaf pine slash pine ecos ystems can be restored and managed using uneven aged silviculture, that the biophysical environment including understory light and leaf area index varies among the uneven aged methods with most heterogeneous environment created by group selection method, a nd soil seed banks can be a potential source for restoring at least some of the native species in the understory For restoring slash pine plantations to stands with uneven aged structures, the study used dat a from a mature even aged slash pine stand at Ta FL, and used Forest Vegetation Simulator (FVS) model to evaluate a total of 49 scenarios that encompassed a range of combinations of harvest types, intensity (residual basal area of 4.6 or 11.5 m 2 ha 1 ), frequency (cutting cycle of 10 or 20 years), and regeneration (0 to 2224 established seedlings per hectare) for their effectiveness to create structural diversity, s tore carbon, and produce merchantable timber over a period

PAGE 118

118 of 100 years. The results showed that the scenarios which re sulted in higher structural diversity generally led to lower carbon s tocks as well as merchantable timber production, and vice versa. The scenarios creating a traditional uneven aged diameter distribution with the first cut resulted in higher average struc tural diversity, but low thinning during first cut s led to slightly greater annual average merchantable timber product ion over the simulation period Longer cutting cycle s resulted in decrease d structural diversity in all scenarios. For the most part, carb on s tocks and merchantable timber production by the end of simulation increased with increase in length of cutting cycle. When maintained at a low residual basal area of 4.6 m 2 ha 1 using low thinning at the first cut with a 20 year cutting cycle and assu ming regeneration of 741 or more seedlings/hectare optimized the provision of multiple benefits i.e. collective benefits of s tructural diversity, carbon s torage and merchantable timber At higher residual basal area of 11.5 m 2 ha 1 low thinning at first cut with 10 year cutting cycle and regeneration of 22 24 seedlings per hectare was one of the feasible opti ons to opti mize provision of multiple benefits, though the BD q method on 10 y ea r cutting cycles ranked highest because of the higher structural diver sity values. Given our knowledge of the species, and our understanding of cultural operations such as prescribed burning, shrub control by mechanical means or the application of herbicides, that influence regeneration in these ecosystems, obtaining these l evels of regeneration for successful application of uneven aged silviculture appears possible. Even in highly unlikely case of complete failure of regeneration, planting of 247 to 741 seedlings per hectare is, apparently, not an economically prohibiting op tion. However, t he results of the study should be interpreted considering the limitations of the study. For example, t he scenarios

PAGE 119

119 simulated in the study were intentionally kept simple. Additionally, t he impact of possible climate change during the simulat ion period could affect the simulation projections. Economic analyses in terms of costs and benefits of activities and products in different scena rios will have huge impact o n decision making. We recommend such evaluations are also addressed in future stud ies. Among the various biophysical factors known to affect longleaf pine slash pine regeneration and understory species richness u nderstory light regime is considered one of the most important particularly in mesic to wet sites where moisture is not limi ting factor. Our study, using Digital Hemispherical Photography (DHP) approach, carried out at multiple sites in north west and north central Florida examined how different management regimes affect ed light levels in longleaf slash pine forests treated wit h shelterwood and uneven aged systems relative to uncut control plots. Basal area in these stands ranged from approximately 5.0 m 2 ha 1 to 40 m 2 ha 1 and species composition ranged from pure longleaf pine to pure slash pine in the overstory As expected, t hese management systems led to significant decreases in leaf area index (LAI), cover fraction, direct fraction of Absorbed Photosynthetically Active Radiation (fAPAR) and diffuse fAPAR, and increase s in visible sky as compared to uncut controls These chan ges indicated significant increase s light availability in shelterwood and uneven aged stands compared to uncut control stands. Mean LAI ranged between 1.7 1.8 for control plots and from 0.3 to 0.9 for the various management systems. Shelterwood systems gen erally had the highest amount of understory light availability, while the greatest variability was observed in group selection system. Greater variability in light conditions may suggest advantages of group selection management over the

PAGE 120

120 other management sy stems when understory restoration is the primary objective The overstory species composition also affected the understory light availability. For a given basal area, longleaf pine stands had greater understory light availability than those of slash pine. Light availability in the mixed species stands differed significantly from pure longleaf pine stands when the proportion of slash pine basal area was 70% or higher. Thus we recommend that basal area regulation approach must account for overstory compositio n to create desirable light regimes while restoring and managing these ecosystems. This however, may be most critical during early restoration efforts when slash pine is the most dominant component of the overstory Another main objective of restoring the se ecosystems is to restore species rich understory, as these ecosystems are among the most species rich ecosystems in North America. In situ restoration of understory following restoration activities requires that adequate seed bank of the desirable speci es exists in the soil. Our assessment of seed bank in the degraded s tands has confirmed the presence of seed bank of at least some of the native desirable species. In fact, species richness in seed bank in degraded s tands was found to be higher than that o f existing vegetation. Additionally seed dynamics in partially restored and restored s tands in these ecosystems have provided critical understanding of the seed ecology and vegetation relationships in these ecosystems. The detection of 26 species in the s eed bank of our degraded s tands is an encouraging finding suggesting seed bank can potentially be used to restore at least some of the diverse understory. The partial ly restored s tands in our study had similar history as the degraded s tands except the rest oration activities (thinning and introduction of prescribed burning) in the past decade but had greater species richness

PAGE 121

121 in seed bank as well as aboveground vegetation. Though we have no information about past seed bank in the partial ly restored sites it does suggest that seed bank composition (or at least its response in seedling emergence) does change with succession of restor ed s tands If it is any sign, we expect the degraded sites to successfully restore the ground cover following the initiation of re storation activities. Since greater seed density in the degraded s tands was observed at 5 10 cm depth in soil, care should be taken to protect this layer during restoration activities. Additionally, the restoration activities that can intermix the soil to expose this layer can also aid in speed recovery of groundcover. For example, prescribed burning can burn the excess 10 cm layer.

PAGE 122

122 APPENDIX A DETAILS OF THE SCENA RIOS USED IN SIMULAT ING STAND CONVERSION

PAGE 123

123 Table A 1. Detai ls of the s cenarios used in simulating conversion of slash pine plantation to an uneven aged stand. Each of the harvest type was simulated over each combination of residual basal area, cutting cycle, and level of regeneration described below. All the scena rios had Maximum Diameter (D) > 86.4 cm, and Diminution quotient (q) = 1.5 for simulating BDq cuts in final stages The name of a scenario consisted of four parts representing harvest type cutting cycle Scenario Desc ription Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Regeneration input 4.6 BDq 10 0000 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 10 No natural regeneration occurred 4.6 BDq 10 02 47 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 10 247 seedlings/hectare of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 BDq 10 0741 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 10 74 1 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 BDq 10 1236 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 10 1236 seedlings/ha of slash pine with average height of about 60 cm su rviving 3 years after each cutting cycle 4.6 BDq 10 1730 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 10 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 BDq 10 2224 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 10 2224 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle

PAGE 124

124 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diame ter (cm) (D) Diminution quotient (q) Cutting cycle Regeneration input 4.6 BDq 20 0000 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 20 No natural regeneration occurred 4.6 BDq 20 0247 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 20 2 47 seedlings/hectare of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 BDq 20 0741 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 20 741 seedlings/ha of slash pine with average height of about 60 c m surviving 3 years after each cutting cycle 4.6 BDq 20 1236 BDq cut from first cutting cycle onwards 4.59 >86.4 1.5 20 1236 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 BDq 20 1730 BDq cut from first cutting cycle onwards 4.59 >86.44 1.5 20 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 BDq 20 2224 BDq cut from first cutting cycle onwards 4.59 >86.44 1.5 20 2224 seedli ngs/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 Thin10 0000 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 10 No natural regeneration occurred

PAGE 125

125 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Regeneration input 4.6 Thin10 0247 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 10 247 seedlings/hectare of slash pine with average height of about 60 cm surviving 3 years after e ach cutting cycle 4.6 Thin10 0741 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 10 741 seedlings/ha of slash pine with average heig ht of about 60 cm surviving 3 years after each cutting cycle 4.6 Thin10 1236 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 10 1236 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle

PAGE 126

126 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Reg eneration input 4.6 Thin10 1730 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 10 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 Thin10 2224 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 10 2224 see dlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 Thin20 0000 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutt ing cycle onwards 4.59 >86.4 1.5 20 No natural regeneration occurred

PAGE 127

127 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Regeneration input 4.6 Thin20 0247 1. Thi nning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 20 247 seedlings/hectare of slash pine with average height of about 60 cm surviving 3 years a fter each cutting cycle 4.6 Thin20 0741 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 20 741 seedlings/ha of slash pine with averag e height of about 60 cm surviving 3 years after each cutting cycle 4.6 Thin20 1236 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 20 1236 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle

PAGE 128

128 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cyc le Regeneration input 4.6 Thin20 1730 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 20 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 4.6 Thin20 2224 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 4.59 >86.4 1.5 20 2 224 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 10 0000 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 10 No natural regeneration occurred 11.5 BDq 10 0247 BDq cut from fir st cutting cycle onwards 11.48 >86.4 1.5 10 247 seedlings/hectare of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 10 0741 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 10 741 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle

PAGE 129

129 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Regeneration input 11.5 BDq 10 1236 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 10 1236 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 10 1730 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 10 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 10 2224 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 10 2224 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 20 0000 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 20 No natural regeneration occurred 11.5 BDq 20 0247 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 20 247 see dlings/hectare of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 20 0741 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 20 741 seedlings/ha of slash pine with average height of about 60 cm su rviving 3 years after each cutting cycle 11.5 BDq 20 1236 BDq cut from first cutting cycle onwards 11.48 >86.4 1.5 20 1236 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 20 1730 BDq cut from first cutting cycle onwards 11.48 >86.44 1.5 20 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 BDq 20 2224 BDq cut from first cutting cycle onwards 11.48 >86.44 1.5 20 2224 se edlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle

PAGE 130

130 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Regenerat ion input 11.5 Thin10 0000 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 10 No natural regeneration occurred 11.5 Thin10 0247 1. Th inning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 10 247 seedlings/hectare of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 Thin10 0741 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 10 741 seedlings/ha of slash pine with av erage height of about 60 cm surviving 3 years after each cutting cycle

PAGE 131

131 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Regeneration input 11.5 Thin10 1236 1 Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 10 1236 seedlings/ha of slash pine with average height of about 60 cm surviving 3 year s after each cutting cycle 11.5 Thin10 1730 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 10 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 Thin10 2224 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 10 2224 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle

PAGE 132

132 T able A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutti ng cycle Regeneration input 11.5 Thin20 0000 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 20 No natural regeneration occurred 11.5 Thin20 0247 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 20 247 seedlings/hectare of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 Thin20 0741 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 20 741 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle

PAGE 133

133 Table A 1. Continued Scenario Description Residual Basal rea (m 2 ha 1 ) (B) Maximum Diameter (cm) (D) Diminution quotient (q) Cutting cycle Regeneration input 11.5 Thin20 1236 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 20 1236 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 Thin20 1730 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 20 1730 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle 11.5 Thin20 2224 1. Thinning from below during first cutting cycle 2. Thinning across dbh classes during second cutting cycle 3. BDq cut from third cutting cycle onwards 11.48 >86.4 1.5 20 2224 seedlings/ha of slash pine with average height of about 60 cm surviving 3 years after each cutting cycle No action No cut of any kind NA NA NA NA NA

PAGE 134

134 APPENDIX B ESTIMATES OF STRUCTU RAL DIVERSITY, CARBO N STOCKS AND TIMB ER PRODUCTION IN DIFFER ENT SI MULATION SCENARIOS

PAGE 135

135 Table B 1. Estimates ( Mean Standard Error ) of s tructural diversity, carbon s tocks and timber production in the 49 scenarios simulated in the study. The values with the same letter d o not differ significant ly. Scenario Average Shannon Index( DBH diversity) Average Shannon Index (Height diversity) Average stand structural diversity Average C s tocks (metric ton/ha/year) Average merchantable wood production (m 3 /ha/year) Average sawtimber production (m 3 /ha/year) 4.6 BDq 10 0000 2.0060.005 1.0560.021 1.5310.012 r 0.6600.005 w,x 1.517 0.018 t 1.496 0.013 t 4.6 BDq 10 0247 2.2390.016 1.7450.006 1.9920.008 b,c 1.1310.008 s,t,u 3.234 0.022 p,q,r,s 2.788 0.019 p,q,r,s 4.6 BDq 10 0741 2.1650.009 1.7600.020 1.9620. 007 c,d,e 1.1610.028 r,s,t,u 3.292 0.103 o,p,q,r,s 2.682 0.098 o,p,q,r,s 4.6 BDq 10 1236 2.1590.002 1.7890.017 1.9740.008 c,d 1.2130.001 q,r,s,t 3.448 0.016 n,o,p,q,r 2.749 0.015 n,o,p,q,r 4.6 BDq 10 1730 2.0900.018 1.7610.010 1.9260.014 d,e,f 1.2640.007 o,p,q 3.560 0.027 l,m,n,o,p 2.859 0.027 l,m,n,o,p 4.6 BDq 10 2224 2.0850.016 1.7730.007 1.9290.006 d,e,f 1.2890.033 o,p,q 3.594 0.106 k,l,m,n,o 2.794 0.060 k,l,m,n,o 4.6 BDq 20 0000 2.0330.014 1.1610.017 1.5970.014 p,q 0.7150.017 w 1.724 0.067 t 1.662 0.0 76 t 4.6 BDq 20 0247 2.1500.011 1.5000.018 1.8250.008 h,I,j 1.0860.007 u,v 3.041 0.032 s 2.767 0.039 s 4.6 BDq 20 0741 2.1140.015 1.5340.014 1.8240.001 h,I,j 1.1540.010 r,s,t,u 3.309 0.032 o,p,q,r,s 2.790 0.044 o,p,q,r,s 4.6 BDq 20 1236 2.0840.017 1.54 30.026 1.8140.019 h,I,j,k 1.2190.023 q,r,s 3.514 0.070 m,n,op,q 2.948 0.076 m,n,o,p,q 4.6 BDq 20 1730 2.0490.013 1.5310.021 1.7900.012 i,j,k,l 1.2690.022 o,p,q 3.665 0.073 k,l,m,n 2.895 0.153 k,l,m,n

PAGE 136

136 Table B 1. Continued Scenario Average Shannon In dex( DBH diversity) Average Shannon Index (Height diversity) Average stand structural diversity Average C s tocks (metric ton/ha/year) Average merchantable wood production (m3/ha/year) Average sawtimber production (m3/ha/year) 4.6 BDq 20 222 4 2.0220.008 1.5360.007 1.7790.008 j,k,l 1.3250.008 o,p 3.866 0.041 k,l 2.927 0.021 k,l 4.6 Thin10 0000 1.0710.011 0.4270.023 0.7490.015 t 0.3990.001 y 0.619 0.004 u 0.626 0.003 u 4.6 Thin10 0247 2.0300.020 1.4940.006 1.7620.008 k,l,m 1.1150.005 t,u,v 3.236 0.014 p,q,r,s 2.736 0.007 p,q,r,s 4.6 Thin10 0741 1.9400.027 1.5370.011 1.7380.019 l,m,n 1.2380.021 p,q,r 3.609 0.062 k,l,m,n,o 2.807 0.051 k,l,m,n,o 4.6 Thin10 1236 1.9540.022 1.5240.006 1.7390.011 l,m,n 1.2740.028 o,p,q 3.713 0.103 k,l,m,n 2.728 0 .086 k,l,m,n 4.6 Thin10 1730 1.9060.016 1.5150.002 1.7110.009 m,n 1.3210.021 o,p 3.829 0.078 k,l,m 2.725 0.092 k,l,m 4.6 Thin10 2224 1.8670.010 1.5080.006 1.6880.004 n,o 1.3560.014 o 3.903 0.061 k 2.597 0.074 k 4.6 Thin20 0000 1.1490.016 0.4280.015 0.7 890.013 t 0.4160.001 y 0.679 0.004 u 0.688 0.004 u 4.6 Thin20 0247 1.9550.016 1.4240.011 1.6890.005 n 1.2520.015 p,q,r 3.608 0.054 k,l,m,n,o 3.088 0.066 k,l,m,n,o 4.6 Thin20 0741 1.8610.005 1.4070.007 1.6340.001 o,p 1.5230.027 n 4.439 0.077 j 3.235 0.087 j 4.6 Thin20 1236 1.8100.002 1.3900.008 1.6000.003 p,q 1.7020.026 l,m 4.862 0.089 i 3.260 0.053 i 4.6 Thin20 1730 1.7530.008 1.3570.009 1.5550.007 q,r 1.7780.022 i,j,k,l 4.906 0.088 i 3.040 0.093 i 4.6 Thin20 2224 1.7040.001 1.3290.004 1.5170.002 r 1.8 820.023 d,e,f,g,h 5.107 0.067 i 3.144 0.078 i 11.5 BDq 10 0000 2.1750.010 1.2050.012 1.6900.011 n 1.0220.004 v 3.057 0.013 s 2.871 0.040 s 11.5 BDq 10 0247 2.2600.008 1.8480.016 2.0540.011 a 1.7660.013 j,k,l 5.771 0.047 d,e,f,g,h 5.021 0.053 d,e,f,g,h

PAGE 137

137 Table B 1. Continued Scenario Average Shannon Index( DBH diversity) Average Shannon Index (Height diversity) Average stand structural diversity Average C s tocks (metric ton/ha/year) Average merchantable wood production (m3/ha/year) Average sawtimbe r production (m3/ha/year) 11.5 BDq 10 0741 2.2320.008 1.8710.001 2.0520.003 a 1.8510.014 e,f,g,h,I, j 6.002 0.040 b,c,d,e,f 5.067 0.094 b,c,d,e,f 11.5 BDq 10 1236 2.2490.004 1.8490.011 2.0490.006 a 1.8920.034 d,e,f 6.105 0.123 a,b,c,d 5.040 0.114 a,b,c,d 11.5 BDq 10 1730 2.2300.008 1.8850.006 2.0580.001 a 1.8710.018 d,e,f,g,h ,i 5.999 0.062 b,c,d,e,f 4.884 0.046 b,c,d,e,f 11.5 BDq 10 2224 2.2010.002 1.8640.017 2.0330.008 a,b 1.8670.017 d,e,f,g,h ,I,j 5.883 0.064 d,e,f,g 4.782 0.020 d,e,f,g 11.5 BD q 20 0000 2.1930.007 1.1930.021 1.6930.014 n 1.0600.004 u,v 3.146 0.009 r,s 2.985 0.020 r,s 11.5 BDq 20 0247 2.2190.004 1.5870.007 1.9030.005 f 1.6190.020 m,n 5.146 0.066 i 4.619 0.074 i 11.5 BDq 20 0741 2.1950.002 1.5840.012 1.8900.005 f,g 1.7330.007 k,l 5.492 0.024 h 4.884 0.021 h 11.5 BDq 20 1236 2.2190.008 1.6070.009 1.9130.008 e,f 1.7670.030 j,k,l 5.581 0.085 g,h 4.945 0.149 g,h 11.5 BDq 20 1730 2.1950.024 1.5860.025 1.8910.024 f,g 1.7890.029 g,h,I,j,k, l 5.583 0.095 g,h 4.857 0.108 g,h 11.5 BDq 20 2224 2.1960.011 1.6080.008 1.9020.009 f 1.7840.033 h,I,j,k,l 5.502 0.083 h 4.744 0.076 h 11.5 Thin10 0000 1.5070.004 0.3710.011 0.9390.007 s 0.5790.000 x 1.541 0.001 t 1.566 0.002 t 11.5 Thin10 0247 2.1130.013 1.5000.005 1.8070.006 h,I,j,k 1.7220.014 k,l 5.764 0.042 e,f,g,h 5.040 0.051 e,f,g,h 11.5 Thin10 0741 2.1040.017 1.5250.018 1.8140.011 h,I,j,k 1.8370.012 e,f,g,h,I, j 6.100 0.052 a,b,c,d,e 5.028 0.032 a,b,c,de 11.5 Thin10 1236 2.1160.009 1.5740.007 1.8450.008 g,h 1.8870.010 d,e,f,g 6.246 0.047 a, b,c 4.911 0.078 a,b,c

PAGE 138

138 Table B 1. Continued Scenario Average Shannon Index( DBH diversity) Average Shannon Index (Height diversity) Average stand structural diversity Average C s tocks (metric ton/ha/year) Average merchantable wood production (m3 /ha/year) Average sawtimber production (m3/ha/year) 11.5 Thin10 1730 2.0830.002 1.5850.002 1.8340.001 h,i 1.9100.006 c,d,e 6.258 0.011 a,b,c 4.932 0.063 a,b,c 11.5 Thin10 2224 2.0530.015 1.5880.013 1.8210.004 h,I,j 1.9680.026 b,c,d 6.387 0.073 a 4.879 0.068 a 11.5 Thin20 0000 1.5490.007 0.3770.004 0.9630.005 s 0.6140.002 w,x 1.655 0.006 t 1.683 0.006 t 11.5 Thin20 0247 2.0510.008 1.3600.010 1.7060.002 n 1.8050.015 f,g,h,I,j, k 5.796 0.045 d,e,f,g,h 5.216 0.058 d,e,f,g,h 11.5 Thin20 0741 2.0460. 009 1.3910.005 1.7190.007 m,n 2.0090.004 a,b,c 6.259 0.031 a,b,c 5.298 0.004 a,b,c 11.5 Thin20 1236 2.0240.005 1.4130.007 1.7180.006 m,n 2.0680.008 a,b 6.280 0.027 a,b 5.111 0.026 a,b 11.5 Thin20 1730 2.0040.005 1.4280.003 1.7160.003 m,n 2.0530.009 a,b 5.936 0.022 c,d,e,f 4.813 0.034 c,d,e,f 11.5 Thin20 2224 1.9890.002 1.4190.009 1.7040.004 n 2.0780.006 a 5.716 0.025 f,g,h 4.575 0.061 f,g,h No action 2.2060.021 0.9910.003 1.5990.011 p,q 1.1960.007 q,r,s,t 3.218 0.013 q,r,s 2.478 0.016 q,r,s

PAGE 139

139 APPENDI X C R ANKING O F THE SIMULATION SCENARIO S USED IN THE STUDY

PAGE 140

140 Table C 1. R anks of the scenarios simulated in the study for their ability to provide multiple benefits. Ranking was done using sum of the scaled values (percent of the maximum value of the variable ) of stand structural diversity, carbon stocks, and total merchantable production Overall ranks as well as rank within a residual basal area group (11.5 m 2 ha 1 and 4.6 m 2 ha 1 ) are provided Scenario Structural diversity (Shannon Index) (1) C stocks (2) Total merchantable production (3) Sum ( 1 +2 + 3 ) Rank (within residual basal area group) Overall Rank 50 BDq10 500 0.996 0.911 0.956 2.862 1 1 50 BDq10 700 1.000 0.900 0.939 2.839 2 2 50 Thin10 900 0.885 0.947 1.000 2.832 3 3 50 BDq10 300 0.997 0.891 0.940 2.828 4 4 50 Thin20 500 0.835 0.995 0.983 2.814 5 5 50 BDq10 900 0.988 0.898 0.921 2.807 6 6 50 Thin10 700 0.891 0.919 0.980 2.790 7 7 50 Thin10 500 0.897 0.908 0.978 2.783 8 8 50 Thin20 300 0.835 0.966 0.980 2.782 9 9 50 BDq10 100 0.998 0.850 0.904 2.751 10 10 50 Thin20 700 0.834 0.988 0.929 2.751 11 11 50 Thin20 900 0.828 1.000 0.895 2.723 12 12 50 Thin10 300 0.882 0.884 0.955 2.721 13 13 50 BDq20 500 0.930 0.850 0.874 2.654 14 14 50 BDq20 700 0.919 0.861 0.874 2.654 15 15 50 BDq20 900 0.924 0.858 0.861 2.644 16 16 50 BDq20 300 0.918 0.834 0.860 2.612 17 17

PAGE 141

141 Table C 1. Continued Scenario Structural diversity (Shannon Index) (1) C stocks (2) Total merchantable production (3) Sum ( 1 +2 + 3 ) Rank (within residual basal area group) Ove rall Rank 50 Thin10 100 0.878 0.829 0.902 2.609 18 18 50 Thin20 100 0.829 0.868 0.907 2.605 19 19 50 BDq20 100 0.925 0.779 0.806 2.510 20 20 20 Thin20 900 0.737 0.906 0.800 2.442 1 21 20 Thin20 700 0.756 0.856 0.768 2.380 2 22 20 Thin20 500 0.778 0.8 19 0.761 2.358 3 23 20 Thin20 300 0.794 0.733 0.695 2.222 4 24 20 BDq10 900 0.937 0.620 0.563 2.120 5 25 20 BDq20 900 0.865 0.637 0.605 2.107 6 26 20 BDq10 700 0.936 0.608 0.557 2.102 7 27 20 Thin10 900 0.820 0.652 0.611 2.084 8 28 20 BDq10 500 0.959 0.584 0.540 2.083 9 29 20 Thin10 700 0.831 0.636 0.600 2.066 10 30 20 BDq20 700 0.870 0.610 0.574 2.054 11 31 20 Thin10 500 0.845 0.613 0.581 2.039 12 32 20 BDq10 300 0.954 0.559 0.515 2.028 13 33 20 BDq10 100 0.968 0.544 0.506 2.019 14 34 20 BDq20 500 0.881 0.586 0.550 2.018 15 35 20 Thin10 300 0.845 0.596 0.565 2.006 16 36 20 Thin20 100 0.821 0.602 0.565 1.988 17 37 20 BDq20 300 0.886 0.555 0.518 1.960 18 38 20 Thin10 100 0.856 0.536 0.507 1.899 19 39 20 BDq20 100 0.887 0.523 0.476 1.886 20 40

PAGE 142

142 Table C 1. Continued Scenario Structural diversity (Shannon Index) (1) C stocks (2) Total merchantable production (3) Sum ( 1 +2 + 3 ) Rank (within residual basal area group) Overall Rank No action 0.777 0.575 0.504 1.856 21 41 50 BDq20 000 0.823 0. 510 0.493 1.825 22 42 50 BDq 10 000 0.821 0.492 0.479 1.792 23 43 20 BDq20 000 0.776 0.344 0.270 1.390 21 44 20 BDq 10 000 0.744 0.318 0.238 1.299 22 45 50 Thin20 000 0.468 0.295 0.259 1.023 24 46 50 Thin10 000 0.457 0.279 0.241 0.976 25 47 20 Thin20 000 0.383 0.200 0.106 0.690 23 48 20 Thin10 000 0.364 0.192 0.097 0.653 24 49

PAGE 143

143 LIST OF REFERENCES Abella, S. R., Springer, J.D., 2008. Estimating soil seed bank characteristics in ponderosa pine forests using vegetation and forest floor data. Res. Note. RMRS RN 35, United States Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fort Collins, CO, p. 7. Alavalapati, J.R.R., Stainback, G.A., Carter, D.R., 2002. Restoration of the longleaf pine ecosystem on private lands in the US So uth: An ecological economic analysis. Ecol. Econ. 40, 411 419. Andreu, M.G., Hedman, C.W., Friedman, M. H., Andreu, A. G., 2009. Can managers bank on seed banks when restoring Pinus taeda L. plantations in southwest Georgia? Rest. Ecol. 17, 586 596. Arroyo, M., Cavieres, L., Castor, C., Humana, A., 1999. Persistent soil seed bank and standing vegetation at a high alpine site in the central Chilean Andes. Plant Ecology 119, 126 132. Augusto, L., Dupouey, J.L., Picard, J.F., Ranger, J., 2001. Potential contribu tion of the seed bank in coniferous plantations to the restoration of native deciduous forest vegetation. Acta Oecologica 22, 87 98. Baker, F.S., 1949. A revised tolerance table. J. Forest. 47, 179 181. Baret, F., 2004. A simple method to calibrate hemisph erical photographs. http://w3.avignon.inra.fr/valeri/fic_htm/methodology/main.php ( Accessed December 26, 2011). Barnett, J.P., Sheffield, R.M., 2005. Slash pine: characteristics, history, status, and trends. In: Dickens, E.D., Barnett, J.P., Hubbard, W.G., Jokela, E.L., (Eds), Slash Pine: still growing and growing! Proceedings of the slash pine symposium. Gen. Tech. Rep. SRS 76. U.S. Department of Agriculture, Forest Service, Southern Research Station, Asheville, NC pp. 1 6. Battaglia, M.A., Mitchell, R.J. Mou, P.P., Pecot, S.D., 2003. Light transmittance estimates in a longleaf pine woodland. For. Sci. 49, 752 762. Battaglia, M.A., Mou, P., Palik, B., Mitchell, R.J., 2002. The effect of spatially variable overstory on the understory light environment of a n open canopied longleaf pine forest. Can. J. For. Res. 32, 1984 1991. Beaudet, M., Messier, C., 2002. Variation in canopy openness and light transmission following selection cutting in northern hardwood stands: An assessment based on hemispherical photogr aphs. Agric. For. Meteorol. 110, 217 228.

PAGE 144

144 Bossuyt, B., Honnay, O., 2008. Can the seed bank be used for ecological restoration? An overview of seed bank characteristics in European communities. J Veg Sci 19, 875 884. Boyer, W.D., 1999. Longleaf pine: Nat ural regeneration and management. Alabama's Treasured Forests 18, 7 9. Brockway, D.G., and Outcalt, K.W., 1998. Gap phase regeneration in longleaf pine wiregrass ecosystems. For. Ecol. Manage. 106, 125 139. Brockway, D.G., Loewenstein, E.F., Outcalt, K.W., 2011. Pro B selection method for uneven aged management of longleaf pine forests. Annual meeting of the Natural Areas Association, Tallahassee, FL, November 1 3, 2011. (Abstract). Brockway, D.G., Outcalt, K.W., 2010. Comparative analysis of reproduction t echniques for sustainable management of longleaf pine forest ecosystems. Report to the Florida Division of Forestry concerning the CART Research Project conducted at the Goethe State Forest and Blackwater River State Forest. Southern Research Station, USDA Forest Service, Auburn, AL. May 11, Final Report, p. 55. Brockway, D.G., Outcalt, K.W., Boyer, W.D., 2006. Longleaf pine regeneration ecology and methods. In: Jose, S., Jokela, E.J., Miller, D.L. (Eds.), The longleaf pine ecosystem: Ecology, silviculture and restoration. Springer, New York, pp. 95 133. Brockway, D.G., Outcalt, K.W., Guldin, J.M., Boyer, W.D., Walker, J.L., Rudolph, D.C., Rummer, R.B., Barnett, J.P., Jose, S., Nowak, J., 2005a. Uneven aged management of longleaf pine forests: A scientist an d manager dialogue. Gen. Tech. Rep. SRS 78. U.S. Department of Agriculture, Forest Service, Southern Research Station, Asheville, NC, p. 38. Brockway, D.G., Outcalt, K.W., Tomczak, D.J., Johnson, E.E., 2005b. Restoration of longleaf pine ecosystems. USDA F or. Serv. Gen. Tech. Rep. SRS 83. U.S. Department of Agriculture, Forest Service, Southern Research Station, Asheville, NC, p. 34. Brown, A. A. Davis, K.P., 1973. Forest fire control and use. 2nd ed. New York: McGraw Hill, p. 686. Brunner, A., 1998. A ligh t model for spatially explicit forest stand models. For. Ecol. Manage. 107, 19 46. Canham, C.D., Finzi, A.C., Pacala, S.W., Burbank, D.H. 1994 Causes and consequences of resource heterogeneity in forests: Interspecific variation in light transmission by canopy trees. Can. J. For. Res. 24,337 349. Canham, C.D., 1988. An index for understory light levels in and around canopy gaps. Ecology 69, 1634 1638.

PAGE 145

145 Canham, C.D., Denslow, J.S., Platt, W.J., Runkle, J.R., Spies, T.A., White, P.S., 1990. Light regimes ben eath closed canopies and tree fall gaps in temperate and tropical forests. Can. J. For. Res. 20, 620 631. Canham, C.D., Marks, P.L., 1985. The response of woody plants to disturbance: Patterns of establishment and growth. In: Pickett, S. T. A, and White, P .S., (Eds.), The ecology of natural disturbance and patch dynamics. Academic Press, Orlando, Fla., pp. 197 216. Carrington, M.E., 1997. Soil seed bank structure and composition in Florida sand pine scrub. Amer. Midl. Natural. 137, 39 47. Clewell, A.F., 198 9. Natural history of wiregrass (Aristida stricta Michx., Graminae). Nat. Areas J. 9, 223 233. Coffey, K.L., Kirkman, L.K., 2006. Seed germination strategies of species with restoration potential in a fire maintained pine savanna. Nat. Areas J. 26, 289 299 Cohen, S., Braham, R., Sanchez, F., 2004. Seed bank viability in disturbed longleaf pine sites. Rest. Ecol. 12, 503 515. Cox, G.W., 1985. Laboratory Manual of General Ecology, 7th ed. McGraw Hill, New York, p. 248. Crawley, M.J., 2005. Statistics: An Int roduction using R. John Wiley, New York, p. 333. Crookston, N.L., Dixon, G.E., 2005. The Forest Vegetation Simulator: A review of its structure, content, and applications. Comput. Electron. Agric. 49, 60 80. Cutini A., 1996. The influence of drought and t hinning on leaf area index estimates from canopy transmittance method. Ann. Sci. For. 53, 595 603. Davi, H., Baret, F., Huc, R., Dufr e ne, E., 2008. Effect of thinning on LAI variance in heterogeneous forests. For. Ecol. Manage. 256, 890 899. Davi, H., Bour iaud, O., Dufr e ne, E., Soudani, K., Pontailler, J.Y., Le Maire, G., Francois, C., Breda, N., Granier, A., Le Dantec, V., 2006. Effect of aggregating spatial parameters on modelling forest carbon and water fluxes. Agric. For. Meteorol. 139, 269 287. Della B ianca, L., Beck, D.E., 1985. Selection management in southern Appalachian hardwoods. South. J. App. For. 9, 191 196. Dickens, E.D., Barnett, J.P., Hubbard, W.G., Jokela, E.J., 2004. Slash pine: still growing and growing Proceedings of the slash pine sympos ium. Gen. Tech. Rep. SRS 76. U.S. Department of Agriculture, Forest Service, Southern Research Station, Asheville, NC, p. 145.

PAGE 146

146 Division of Forestry, 2007. Ten State Forest, Franklin and Liberty counties. Fl orida Division of Agriculture and Consumer Services, Division of Forestry, Carrabelle, FL, p. 73. Internal Report. U.S. Department of Agriculture, Forest Service, Forest M anagement Service Center, Fort Collins, CO, p. 244. (Revised: December 16, 2011) Donnelly, D., Lilly, B., Smith, E., 2001. The southern variant of the Forest Vegetation Simulator. U.S. Department of Agriculture, Forest Service, Forest Management Service Ce nter, Fort Collins, CO, p. 61. Drewa, P B. Platt, W J. Moser, E.B ., 2002. Fire effects on resprouting of shrubs in headwaters of southeastern longleaf pine savannas. Ecology 83 755 767 Farrar, R.M., 1996. Fundamentals of uneven aged management in south ern pine. Misc. Publ. No. 9,Tall Timbers Research Station, Tallahassee, FL, p. 68. Florida Natural Areas Inventory, 2010. Guide to the natural communities of Florida. Florida Natural Areas Inventory, Tallahassee, FL, p. 223. Franklin, R.M., 1997. Stewardsh ip of longleaf pine forests: A guide for landowners. Longleaf Alliance Report no. 2, Andalusia, AL, p. 41. Frazer, G.W., Fournier, R.A., Trofymow, J.A., Hall, R.J., 2001. A comparison of digital and film fisheye photography for analysis of forest canopy st ructure and gap light transmission. Agric. For. Meteorol. 109, 249 263. Frost, C.C., 1990. Natural diversity and status of longleaf pine communities. In: Forestry Annual Business Meeting of the Appalachian Society of American Foresters, 24 26 January 1990, Pinehurst, NC, pp. 26 56 Frost, C.C, 1993. Four centuries of changing landscape patterns in the longleaf pine ecosystem. In: Hermann, S.M. (Ed.), Proc. Tall Timbers Fire Ecol. Conf. Tall Timbers Research Station, Tallahassee, FL, pp. 17 43. Frost, C.C., 2006. History and future of the longleaf pine ecosystem. In: Jose, S., Jokela, E.J., Miller, D.L. (Eds.), The longleaf pine ecosystem: Ecology, silviculture and restoration. Spri nger, New York, pp. 9 48. Gagnon, J.L., Jokela, E.J., Moser, W.K., Huber, D.A., 2003. Dynamics of artificial regeneration in gaps within a longleaf pine flatwoods ecosystem. For. Ecol. Manage. 172, 133 144.

PAGE 147

147 Gagnon, J.L., Jokela, E.J., Moser, W.K., Huber, D .A., 2004. Characteristics of gaps and natural regeneration in mature longleaf pine flatwoods ecosystems. For. Ecol. Manage. 187, 373 380. Garrigues, S., Shabanov, N.V., Swanson, K., Morisette, J.T., Baret, F., Myneni, R.B., 2008. Intercomparison and sensi tivity analysis of Leaf Area Index retrievals from LAI 2000, AccuPAR, and digital hemispherical photography over croplands. Agric. For. Meteorol. 148, 1193 1209. Ghersa, C.M., Martinez Ghersa, M.A., 2000. Ecological correlates of weed seed size and persist ence in the soil under different tilling systems: Implications for weed management. Field Crops Res 67, 141 148. Gilpin, T.T., Vowell, J.L., 2006. Restoring hydroperiod on previously intensively managed forestlands. In: Proceedings of the International Co nference, Hydrology and Management of Forested Wetlands, ASAE Pub #701P0406, 8 12 April 2006, New Bern, NC, pp. 94 98. Glass, S., 1989. The role of soil seed banks in restoration and management. Rest. Manage. Notes 7, 24 29. Gobster, P.H., 1996. Forest aes thetics, biodiversity, and the perceived appropriateness of ecosystem management practices. In: M. Brunson, L. Kruger, C. Tyler, and S. Schroeder (Eds.), Defining social acceptability in ecosystem management: a workshop proceedings. Gen Tech Rep PNW 369 U.S. Forest Service, Portland, OR, pp. 77 97. Gonzalez Benecke, C.A., Martin, T.A., Cropper, W.P., 2011. Whole tree water relations of co occurring mature Pinus palustris and P. elliottii. Can. J. of For. Res. 41, 509 523. Gorte, R.W., 2009. Carbon seque stration in forests: U.S. Library of Congress, Congressional Research Service Report, USA. Guldin, J.M., 2006. Uneven aged silviculture of longleaf pine. In: Jose, S., Jokela, E.J., Miller, D.L. (Eds.), The longleaf pine ecosystem: Ecology, silviculture an d restoration. Springer, New York, pp. 217 249. Gundersen, V.S., Frivold, L.H., 2008. Public preferences for forest structures: A review of quantitative surveys from Finland, Norway and Sw eden. Urban For. Urban Green. 7, 241 58. Hale, S.E., 2003. The effec t of thinning intensity on the below canopy light environment in a Sitka spruce plantation. For. Ecol. Manage. 179, 341 349. Hale, S.E., Edwards, C., Mason, W.L., Price, M., Peace, A., 2009. Relationships between canopy transmittance and stand parameters i n Sitka spruce and Scots pine stands in Britain. Forestry 82, 503 513.

PAGE 148

148 H amilton, D.A., 1991 Implications of random variation in the Stand Prognosis Model. Res. Note INT 394, U. S. Department of Agriculture, Forest Service, Intermountain Research Station, Ogden, UT, p. 11. Hardin, E.D., White, D.L., 1989. Rare vascular plant taxa associated with wiregrass (Aristida stricta) in the southeastern United States. Nat Areas J 9, 234 245. Harrington, T.B., 2006. Plant competition, facilitation, and other oversto ry understory interactions in longleaf pine ecosystems. In: Jose, S., Jokela, E.J., Miller, D.L. (Eds.), The longleaf pine ecosystem: Ecology, silviculture and restoration. Springer, New York, pp. 135 156. Harrington, T.B., Edwards, M.B., 1999. Understory vegetation, resource availability, and litterfall responses to pine thinning and woody vegetation control in longleaf pine plantations. Can. J. For. Res. 29, 1055 1064. Holmes, M.G., Smith, H., 1977. The function of phytochrome in the natural environment I I. The influence of vegetation canopies on the spectral energy distribution of natural daylight. Photochem. Photobiol. 25, 539 545. Hoover, C.M., Rebain, S.A., 2011. Forest carbon estimation using the Forest Vegetation Simulator: Seven things you need to k now. Gen Tech Rep NRS 77 USDA Forest Service, Northern Research Station, Newtown Square, PA, p. 16. Jelaska S.D., Antonic, O., Bozic, M., Krizan, J., Kusan, V, 2006. Responses of forest herbs to available understory light measured with hemispherical p hotographs in silver fir beech forest in Croatia. Ecol Model. 194, 209 218. Jensen, H.A., 1969. Content of buried seed in arable soil in Denmark and its relation to the weed population. Danske Botaniske Arkiv 27, 1 55. Johnson, A.S., Landers, J.L., 1982. Habitat relationships of summer resident birds in slash pine flatwoods. J. Wild. Manage. 46, 416 428. Johnson, R., Gjerstad, D., 2006. Restoring the overstory of longleaf pine ecosystems. In: Jose, S., Jokela, E.J., Miller, D.L. (Eds.), The longleaf pine e cosystem: Ecology, silviculture and restoration. Springer, New York, pp. 271 295. Jonckheere, I., Fleck, S., Nackaerts, K., Muys, B., Coppin, P., Weiss, M., Baret, F., 2004. Review of methods for in situ leaf area index determination Part I. Theories, sens ors and hemispherical photography. Agri. For. Meteorol. 121, 19 35. Jose, S., Jokela, E.J., Miller, D.L. (Eds.), 2006 a Longleaf pine ecosystems: Ecology, silviculture, and restoration. Springer Verlag, New York, p. 438. Jose, S., Jokela, E. J. Miller, D. L 2006 b The longleaf pine ecosystem. In: Jose, S., Jokela, E.J., Miller, D.L. (Eds.), The longleaf pine ecosystem: Ecology, silviculture and restoration. Springer, New York, pp. 3 8.

PAGE 149

149 Kaipainen, T., Liski, J., Pussinen, A., Karjalainen, T., 2004. Managing carbon sinks by changing rotation length in European forests. Environ. Sci. Policy 7, 205 2 19. Kelty, M.J., Kittredge, D.B., Kyker Snowman, T, Leighton, A.D., 2003. The Conversion of even aged stands to uneven aged structure in southern New England. North J. Appl. For. 20 109 116. Keyser, C.E., 2008. Southern (SN) Variant overview Forest Vegetation Simulator. Internal Rep. U. S. Department of Agriculture, Forest Service, Forest Management Service Center, Fort Collins, CO, p. 70. (Revised: October 4, 20 11) Khabba, S., Duchemin, B., Hadria, R., Ezzahar, J., Chehbouni, A., Lahrouni, A., Hanich, L., 2009. Evaluation of digital hemispherical photography and plant canopy analyser for measuring vegetation area index of orange orchards. J. Agron. 8, 67 72. Kirk man, L.K., Mitchell, R.J., Kaeser, M.J., Pecot, S.D., Coffey, K.L., 2007. The perpetual forest: using undesirable species to bridge restoration. J. Appl. Ecol. 44, 604 614. Kuusipalo, J., 1985. On the use of tree stand parameters in estimating light condit ions below the canopy. Sil. Fenn. 19, 185 196. Laiho, O., Lahde, E., Pukkala, T., 2011. Uneven vs. even aged management in Finnish boreal forests. Forestry 84, 547 556. Landers, J.L., 1991. Disturbance influences on pine traits in the southeastern United States. In: Proceedings, 17th Tall Timbers fire ecology conference 1 989 May 18 21 T all Timbers Research Station Tallahassee, FL., pp. 61 95 Landers, J.L., Van Lear, D.H., Boyer, W.D., 1995.The longleaf forests of the Southeast: Requiem or renaissance? J For. 93, 1139 44. Langdon, O.G., Bennett, F., 1976. Management of natural stands of slash pine. Res. Pap. SE 147. U.S. Department of Agriculture, Forest Service, Southeastern Forest E xperiment Station, Ashville, NC, p 12. Lewis, C.E., Harshbarger, T.J. 1976. Shrub and herbaceous vegetation after 20 years of prescribed burning in the South Carolina coastal plain. J. Range Manage. 29, 13 18. Lieffers, V.J., Messier, C., Stadt, K.J., Gendron, F., Comeau, P.G., 1999. Predicting and managing light in the un derstory of boreal forests. Can. J. For. Res. 29, 796 811. Lindhagen, A., 1996. An approach to clarifying public preferences about silvicultural systems: A case study concerning group selection and clear cutting. Scand. J. For. Res. 11, 375 87.

PAGE 150

150 Loewenstein E.F., 2005. Conversion of uniform broadleaved stands to an uneven age d structure. For. Ecol. Manage. 215, 103 112. Lohrey, R.E., Kossuth, S.V., 1990. Pinus elliottii Engelm. slash pine. In: Burns, R.M., Honkala, B.H., (Tech. coords.), Silvics of North Am erica Volume 1. Conifers. Agricultural Handb. 654. U.S. Department of Agriculture, Forest Service, Washington, D.C., pp. 338 347. MacArthur, R.H., MacArthur, J.W., 1961. On bird species diversity. Ecology 42, 594 598. Magurran, A.E., 1988. Ecological diver sity and its measurement. Princeton University Press, Princeton, New Jersey, p. 192. Maliakal, S.K., Menges, E.S., Denslow, J.S., 2000. Community composition and regeneration of lake wales ridge wiregrass flatwoods in relation to time since fire. J. Torr. Bot. Soc. 127 125 138. Marion, W.R., Tanner, G.W., Ober, H.K., Werner, M, 2011 Management of pine forests for selected wildlife in Florida Circular 706 (Revised) Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, Univer sity of Florida. Available online at http://edis.ifas.ufl.edu/uw098 (Accessed June 2012) Masters, R.E., Robertson, K., Palmer, B., Cox, J., McGorty, K., Green, L., Ambrose, C., 2003. Red Hills forest stewardsh ip guide. Miscellaneous Publication no. 12, Tall Timbers Research Station, Tallahassee, FL, p. 78. McConnell, W.V., 2002. Initiating uneven aged management in longleaf pine stands: Impacts on red cockaded woodpecker habitat. Wildl. Soc. Bull. 30, 1276 1280 McCulley, R.D., 1950. Management of natural slash pine stands in the Flatwoods of south Florida and north Florida. Circular No. 845, U.S. Department of Agriculture, Washington, DC p. 57. Mc Elhinny, C., Gibbons, P., Brack, C., Bauhus, J., 2005. Forest a nd woodland stand structural complexity: its definition and measurement. For. Ecol. Manage. 218, 1 24. McGaughey, R.J. 1997. Visualizing forest stand dynamics using the stand visualization system. In: Proceedings of 1 997 ASPRS ACSM RTI Spring Convention 1 997 April 7 10 Seattle, WA. American Congress on Surveying and Mapping Bethesda, MD 4 248 257. McGuire, J.P., Mitchell, R.J., Moser, E.B., Pecot, S.D., Gjerstad, D.H., Hedman, C.W., 2001. Gaps in a gappy forest: plant resources, longleaf pine regenerat ion, and understory response to tree removal in longleaf pine savannas. Can. J. For. Res. 31, 765 778.

PAGE 151

151 McMinn, J.W., 1981. Site preparation for natural regeneration of slash pine. South. J. Appl. For. 5, 10 12. Means, D.B., Grow, G., 1985. The endangered l ongleaf pine community. ENFO Newsletter 85, 1 12. Mitchell, J.E., Popovich, S.J., 1997. Effectiveness of basal area for estimating canopy cover of ponderosa pine. For. Ecol. Manage. 95, 45 51. Mitchell, R.J., Hiers, J.K., O'Brien, J., Starr, G., 2009. Ecol ogical forestry in the Southeast: Understanding the ecology of fuels. J. For. 107, 391 397. Noss, R.F., Peters, R.L., 1995. Endangered ecosystems of the United States: A status report and plan for action. Defenders of Wildlife, Washington, D.C., p. 151. No ss, R.F., 1989. Longleaf pine and wiregrass: Keystone components of an endangered ecosystem. Nat. Areas J. 9, 211 213. Nyland, R.D., 2003. Even to uneven aged: the challenges of conversion. For. Ecol. Manage. 172, 291 300. 04. Stocking control concepts in uneven aged silviculture. Forestry 77, 131 143. O'Hara, K.L., 2001. The silviculture of transformation -a commentary. For. Ecol. Manage. 151, 81 86. O'Hara, K.L., Stancioiu, P.T., Spencer, M.A., 2007. Understory stump sprou t development under variable canopy density and leaf area in coast redwood. For. Ecol. Manage. 244, 76 85. Oliver, C.D., Larson, B.C., 1996. Forest stand dynamics, Wiley, N ew Y ork p. 537. Palik, B. J Mitchell, R.J., Pecot, S., Battaglia, M., Pu, M., 2003. Spatial distribution of overstory retention influences resources and growth of longleaf pine seedlings. Ecol. Appl. 13, 674 686. Palik, B.J., Mitchell, R.J., Houseal, G., Pederson, N., 1997. Effects of canopy structure on resource availability and seedlin g responses in a longleaf pine ecosystem. Can. J. For. Res. 27, 1458 1464. Parrotta, J.A., 1995. Influence of overstory composition on understory colonization by native species in plantations on a degraded tropical site. J. Veg. Sci. 6, 627 636. Pearson, H .A., Whitaker, L.B., Duvall, V.L., 1971. Slash pine regeneration under regulated grazing. J. For. 69, 744 746.

PAGE 152

152 Pecot, S.D., Horsley, S.B., Battaglia, M.A., Mitchell, R.J., 2005. The influence of canopy, sky condition, and solar angle on light quality in l ongleaf pine woodland. Can. J. For. Res. 35, 1356 1366. Peet, R.K., Allard, D.J., 1993. Longleaf pine dominated vegetation of the southern Atlantic and eastern Gulf Coast region, USA. In: Hermann, S.M., (Ed.), Proceedings of the Tall Timbers Fire Ecology C onference, Tall Timbers Research Station, Tallahassee, FL, 18, 45 81. Platt, W.J., Evans, G.W., Davis, M.M., 1988a. Effects of fire season on flowering of forbs and shrubs in longleaf pine forests. Oecologia 76, 353 363. Platt, W.J., Evans, G.W., Rathbun, S.L., 1988 b The population dynamics of a long lived conifer (Pinus palustris). Am. Naturalist 131, 491 525. Platt, W.J., Carr, S.M., Reilly, M., Fahr, J., 2006. Pine savanna overstorey influences on ground cover biodiversity. App. Veg. Sci. 9, 37 50. Pric e, J.N., Wright, B.R., Gross, C.L., Whalley, R.D.B., 2010. Comparison of seedling emergence and seed extraction techniques for estimating the composition of soil seed banks. Meth. Ecol. Evol. 1, 151 157. Pukkala, T., Lhde, E., Laiho, O. 2010. Optimizing the structure and management of uneven aged stands in Finland. Forestry 83 129 142. Roxburgh, J.R., Kelly, D., 1995. Short communication uses and limitations of hemispherical photography for estimating forest light environments. N. Z. J. Ecol. 19, 213 217 Ruth, A.D., Jose, S., Miller, D.L., 2008. Seed bank dynamics of sand pine scrub and longleaf pine flatwoods of the gulf coastal plain (Florida). Ecol. Rest. 26, 19 21. Sedjo, R.A., 2001. Forest carbon sequestration: some issues for forest investments. Di scussion Paper # 01 34 Washington, DC: Resources for the Future. Silvennoinen, H., Alho, J., Kolehmainen, O., Pukkala, T., 2001. Prediction models of landscape preferences at the forest stand level. Landsc. Urban Plan. 56, 11 20. Simpson, R.L., Leck, M.A., Parker, V.T., 1989. Seed banks: General concepts and methodological issues. In: Leck, M.A., Parker, V.T, Simpson, R.L. (Eds.), Ecology of soil seed banks. Academic Press, San Diego, CA, pp. 3 8. Smith, D.M., Larson, B.C., Kelty, M.J., Ashton, P.M.S., 1997 The Practice of Silviculture: Applied Forest Ecology. John Wiley and Sons, New York, p. 537.

PAGE 153

153 Smith, W., Miles, P., Perry, C., Pugh, S., 2009. Forest resources of the United States, 2007: A technical document supporting the Forest Service 2010 RPA assessm ent. Gen. Tech. Rep. WO 78. U S Department of Agriculture, Forest Service, Washington Office, Washington, DC. 2009. Spatially explicit modeling of overstory manipulations in young forests: Effects on stand structure and light. Ecol. Model. 220, 3565 3575. Stenberg, P., Kuuluvainen, T., Kellomki, S., Grace, J.C., Jokela, E.J., Gholz, H.L., 1994. Crown structure, light interception and productivity of pine trees and stands. Ecol. Bull. 43, 20 34. Tarp, P., Buongiorno, J., Helles, F., Larsen, J.B., Meilby, H., Strange, N, 2005. Econom ics of converting an even aged Fagus sylvatica stand to an uneven aged stand using target diameter harvesting. Scand. J. For. Res. 20, 63 74. Thomas, J.W. 1979. Wildlife Habitats in Managed Forests the Blue Mountains of Oregon and Washington. Agriculture Handbook No. 553. U.S. Department of Agriculture, Forest Service, p. 512. Thompson, K., Grime, J.P., 1979. Seasonal variation in the seed banks of herbaceous species in ten contrasting habitats. J. Ecol. 67, 893 921. Thompson, K., Bakker, J.P., Bekker, R.M ., 1997. The soil seed banks of North West Europe: Methodology, density, and longevity. Cambridge University Press, Cambridge, UK, p. 276. Valladares, F., Guzmn, B., 2006. Canopy structure and spatial heterogeneity of understory light in an abandoned Holm oak woodland. Ann. For. Sci. 63, 749 761. Van der Valk, A.G., Pederson, R.L., 1989. Seed banks and the management and restoration of natural vegetation. In: Leck, M. A., Parker, V. T, Simpson, R. L. (Eds.), Ecology of soil seed banks. Academic Press, San Diego, CA, pp. 329 346. Van Dyck, M.G., Smith, E.E., 2000. Keyword reference guide for the forest vegetation simulator. Int Rep. U. S. Department of Agriculture, Forest Service, Forest Management Service Center Fort Collins, CO p. 120. (Revised: March 2 012) Van Lear, D.H., Carroll, W., Kapeluck, P., Johnson, R., 2005. History and restoration of the longleaf pine grassland ecosystem: Implications for species at risk. For. Ecol. Manage. 211, 150 165. Vanclay, J.K. 1994. Modelling Forest Growth and Yield: A pplications to Mixed Tropical Forests CAB International, Wallingford, UK, p. 312.

PAGE 154

154 Vose, J.M., Sullivan, N.H., Clinton, B.D., Bolstad, P.V., 1995. Vertical leaf area distribution, light transmittance, and application of the Beer Lambert Law in four mature h ardwood stands in the southern Appalachians. Can. J. For. Res. 25, 1036 1043. Wahlenberg, W.G., 1946. Longleaf pine: its use, ecology, regeneration, growth and management. Charles Lathrop Park Forestry Foundation and U.S. Forest Service, Washington, D.C., p. 429. Waldrop, T.A. White, D.L. Jones, S.M. 1992. Fire regimes for pine grassland communities in the southeastern United States. For Ecol Manage 47 195 210 Walker, J.L, 1993. Rare vascular plant taxa associated with the longleaf pine ecosystem. I n: Hermann, S. M. (Ed.), Proceedings of the 18th Tall Timbers Fire Ecology Conference. Tall Timbers Research Station, Tallahassee, FL, pp. 105 125. Wang, J., Ren, H., Yang, L., Li, D., Guo, Q., 2009. Soil seed banks in four 22 year old plantations in South China: Implications for restoration. For Ecol. Manage. 258, 2000 2006. Wang, Y.P., Jarvis, P.G., 1990. Description and validation of an array model MAESTRO. Agric. For. Meteorol. 51, 257 280. Warr, S.J., Thompson, K., Kent, M., 1993. Seed banks as a neglected area of biogeographic research: A review of literature and sampling techniques. Prog. Phys. Geogr. 17, 329 347. Weiskittel, A.R., Hann, D.W., Kershaw, J., Vanclay, J.K., 2011. Forest Growt h and Yield Modeling. Wiley Blackwell, Chicester, UK, p. 430. Weiss, M., Baret, F., 2010. CAN EYE V6.1 User Manual. EMMAH laboratory (Mediterranean environment and agro hydro system modelisation). French National Institute of Agricultural Research (INRA), p. 47. Weiss, M., Baret, F., Smith, G.J., Jonckheere, I., Coppin, P., 2004. Review of methods for in situ leaf area index determination. Part II. Estimation of LAI, errors and sampling. Agri. For. Meteorol. 121, 37 53. Wolters, G.L., 1973. Southern pine ov erstories influence herbage quality. J. Range Manage. 26, 423 426. Wolters, G.L., 1981. Timber thinning and prescribed burning as methods to increase herbage on grazed and protected longleaf pine ranges. J. Range Manage. 34, 494 497. Wright, H.A., Bailey, A.W., 1982. Fire ecology: United States and southern Canada. John Wiley & Sons New York, p. 501

PAGE 155

155 Wunderlin, R.P., Hansen, B.F. 2003. Guide to the vascular plants of Florida, 2nd ed. U. Press of Florida, Gainesville, FL, p 787. Yirdaw, E., Luukkanen, O., 2004. Photosynthetically active radiation transmittance of forest plantation canopies in the Ethiopian highlands. For. Ecol. Manage. 188, 17 24. Zobel, M., Kalamees, R., P u ssa, K., Roosaluste, E. Moora, M. 2007. Soil seed bank and vegetation in mixed co niferous forest stands with different disturbance regimes. For Ecol Manage 250 71 76.

PAGE 156

156 BIOGRAPHICAL SKETCH Ajay Sharma was born in Kathua, a small town in the state of Jammu and Kashmir, India. He earned his B.S and M.S. degrees in f orestry from India. In fall 2008, he joined doctoral degree program in the School of Forest Resources and Conservation at the University of Florida. Before joining this program he w orked as a professional forester and a research fellow at the Indian Institute of Remote Sens ing, Dehradun, and the Indian Institute of Integrative Medicine, Jammu, in India. His research interests include restoration ecology and ecological restoration of degraded ecosystems, uneven aged silviculture, and forest dynamics modeling. Ajay Sharma is m arried to Pritika Sharma and has a 2 year old son. He likes traveling and reading autobiographies.