<%BANNER%>

Floristic and Environmental Variation of Pyrogenic Pinelands in the Southeastern Coastal Plain

Permanent Link: http://ufdc.ufl.edu/UFE0021711/00001

Material Information

Title: Floristic and Environmental Variation of Pyrogenic Pinelands in the Southeastern Coastal Plain Description, Classification, and Restoration
Physical Description: 1 online resource (184 p.)
Language: english
Creator: Carr, Susan C
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: classification, florida, grasslands, longleaf, pinelands, restoration, vegetation
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Until recent times, the landscape of north and central Florida was dominated by fire-dependent pineland savanna vegetation with sparse canopies of longleaf pine (Pinus palustris). Economic development coupled with fire suppression lead to the drastic decline in the distribution and integrity of these natural communities. I present a vegetation classification of natural pineland communities in this highly fragmented landscape based on data collected over large gradients of environmental and geological variation. I collected field data that quantified species composition and abundance from 293 plots (from 103 sites) distributed throughout the northern two thirds of Florida. After omission of species that occurred in < 3% of plots, a total of 677 plant species were used in numerical analyses. I developed a vegetation classification based on floristic similarity using K-means cluster and indicator species analyses. Three ecological series were described corresponding to idealized moisture conditions. These were further divided into 16 species associations. Floristic variation was related to geographic separation between the panhandle and peninsula regions of Florida. I hypothesized that the numerous plant species that have limited distributional ranges contribute to compositional patterns. Similar geographic trends were apparent in a model of compositional variation related to environment and spatial variation. Local environmental factors, including location on a topographic/moisture gradient and soil fertility, were important correlates of local floristic variation. Regional variation was correlated with soil texture and nutrient availability. A much greater proportion of the explained variance was provided by environmental variables than by pure spatial variables. The model revealed that both regional factors (climate, edaphic, and geographic) and local factors (topographic position, soil chemistry) were correlates of with floristic variation. In addition to spatial variation, natural pineland communities undergo temporal variation in response to periodic fires and changes in timber stand structure. Central questions regarding ecological restoration of Coastal Plain pinelands are: how resilient are these communities following anthropogenic alterations? Will ecological restoration affect vegetation succession within the range of 'natural' temporal variation? I studied ground cover vegetation response to removal of woody biomass and reintroduction of natural fire regimes as it related to a program of ecological restoration in a degraded pine savanna remnant. Treatment plots were thinned for timber or else un-thinned as a control. Prescribed fire was applied at two subsequent times, and changes in species composition were monitored over an 8-year period. Species richness was enhanced by mechanical woody reduction in the first two years, compared to sites that were burned only (not logged). This response largely reflected increases in detectable graminoid species. However, species richness of treatments converged within 8 years, following two prescribed fires 3 years apart. Species composition responded similarly, converging between treatments over time. Succession was toward pre-settlement conditions, as suggested by comparisons to reference sites and historical data. Community composition appears to be robust to temporary alterations in fire regime and changes in timber stand structure within the range of conditions studied. Spatial variation in species composition of pineland communities may be relatively stable over time. Woody biomass reduction via careful mechanical logging does not appear to adversely affect pineland vegetation recovery, and may expedite overall community restoration.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Susan C Carr.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Tanner, George W.
Local: Co-adviser: Robertson, Kevin.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021711:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021711/00001

Material Information

Title: Floristic and Environmental Variation of Pyrogenic Pinelands in the Southeastern Coastal Plain Description, Classification, and Restoration
Physical Description: 1 online resource (184 p.)
Language: english
Creator: Carr, Susan C
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: classification, florida, grasslands, longleaf, pinelands, restoration, vegetation
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Until recent times, the landscape of north and central Florida was dominated by fire-dependent pineland savanna vegetation with sparse canopies of longleaf pine (Pinus palustris). Economic development coupled with fire suppression lead to the drastic decline in the distribution and integrity of these natural communities. I present a vegetation classification of natural pineland communities in this highly fragmented landscape based on data collected over large gradients of environmental and geological variation. I collected field data that quantified species composition and abundance from 293 plots (from 103 sites) distributed throughout the northern two thirds of Florida. After omission of species that occurred in < 3% of plots, a total of 677 plant species were used in numerical analyses. I developed a vegetation classification based on floristic similarity using K-means cluster and indicator species analyses. Three ecological series were described corresponding to idealized moisture conditions. These were further divided into 16 species associations. Floristic variation was related to geographic separation between the panhandle and peninsula regions of Florida. I hypothesized that the numerous plant species that have limited distributional ranges contribute to compositional patterns. Similar geographic trends were apparent in a model of compositional variation related to environment and spatial variation. Local environmental factors, including location on a topographic/moisture gradient and soil fertility, were important correlates of local floristic variation. Regional variation was correlated with soil texture and nutrient availability. A much greater proportion of the explained variance was provided by environmental variables than by pure spatial variables. The model revealed that both regional factors (climate, edaphic, and geographic) and local factors (topographic position, soil chemistry) were correlates of with floristic variation. In addition to spatial variation, natural pineland communities undergo temporal variation in response to periodic fires and changes in timber stand structure. Central questions regarding ecological restoration of Coastal Plain pinelands are: how resilient are these communities following anthropogenic alterations? Will ecological restoration affect vegetation succession within the range of 'natural' temporal variation? I studied ground cover vegetation response to removal of woody biomass and reintroduction of natural fire regimes as it related to a program of ecological restoration in a degraded pine savanna remnant. Treatment plots were thinned for timber or else un-thinned as a control. Prescribed fire was applied at two subsequent times, and changes in species composition were monitored over an 8-year period. Species richness was enhanced by mechanical woody reduction in the first two years, compared to sites that were burned only (not logged). This response largely reflected increases in detectable graminoid species. However, species richness of treatments converged within 8 years, following two prescribed fires 3 years apart. Species composition responded similarly, converging between treatments over time. Succession was toward pre-settlement conditions, as suggested by comparisons to reference sites and historical data. Community composition appears to be robust to temporary alterations in fire regime and changes in timber stand structure within the range of conditions studied. Spatial variation in species composition of pineland communities may be relatively stable over time. Woody biomass reduction via careful mechanical logging does not appear to adversely affect pineland vegetation recovery, and may expedite overall community restoration.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Susan C Carr.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Tanner, George W.
Local: Co-adviser: Robertson, Kevin.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021711:00001


This item has the following downloads:


Full Text








FLORISTIC AND ENVIRONMENTAL VARIATION OF PYROGENIC PINELANDS INT THE
SOUTHEASTERN COASTAL PLAIN: DESCRIPTION, CLASSIFICATION, AND
RESTORATION
















By

SUSAN CATHERINE CARR


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

UNIVERSITY OF FLORIDA

2007






































O 2007 Susan Carr





























To Mike, my husband and great love









ACKNOWLEDGMENTS

I thank the members of my committee, Dr. Debbie Miller, Dr. Doria Gordon, and Dr.

Wiley Kitchens for their patience and support. I especially thank my advisors, Dr. George

Tanner and Dr. Kevin Robertson. They were really there for me, and I am grateful for their

support.

Many people were involved in every stage of this proj ect, including initial information

gathering, site selection, vegetation sampling, data entry and management, and data analysis.

For assistance with the initial ecoregion delineation and site selection process, I thank Dr. Bruce

Means, Dr. Bill Platt, Wilson Baker, Andy VanLoan, Ann Johnson, Carolyn Kindell, Dr. Louis

Provencher, and Brenda Herring. Many employees of state, federal, and private land

management agencies provided invaluable assistance with logistical matters ranging from

permitting paperwork to site selection and access. These include:, Dr. Dennis Hardin, Charlie

Pederson, Dr. Ann Cox, Ace Haddick, Tom Serviss, Bobby Cahal, Vince Morris, Scott Crosby,

Dan Pearson, John McKenzie, Craig Parenteau, Ginger Morgan, Rosie Mulholland, Alice Bard,

Ken Alvarez, Terry Hingtgen, Bobby Hattaway, Donna Watkins, Mark Latch, Roy Ogles, Carla

Jean Ogles, Dr. Jean Huffman, Louise Kirn, Dr. Guy Aglin, Jim Ruhl, Lorraine Miller, Kevin

Love, Greg Seamon, Monica Folk, Sandy Woiak, Bee Pace, Lynn Askins, Kevin Hiers, Steve

Orzell, Edwin Bridges, Amanda Stevens, Jerry Pitts, Kristen Wood, Ray and Patricia Ashton,

Raymond Bass, and Gary Maxwell.

I thank those that helped with field data collection, and data entry, management and

analysis, including Dr. Joel Gramling, Dr. Ann Johnson, Kevin Hiers, Brian Mealor, Steve

Orzell, Edwin Bridges, John Brubaker, Dr. Jeff Glitsenzen, Dr. Donna Streng, Maynard Hiss, Dr.

Brian Strom, Dr. Jean Huffman, Dr. Bill Platt, Dr. Robert Peet and Deb Cupples. I extend









special gratitude to Jessica Kaplan, Christine Carlson, Brian Strom, and Dr. Joel Gramling for

their long hours toiling in the Hield and at the computer.

I am grateful to the faculty and staff of the University of Florida Herbarium, for allowing

me to use their facilities and exploit their knowledge including Kent Perkins, Dr. Norris

Williams, Trudy Lindler and Dr. Walter Judd. I especially thank Richard Abbot, Patrick

McMillian, and Brenda Herring for their help with plant identifications. I sampled vegetation on

the Ordway-Swisher Biological Station which is owned and managed by the University of

Florida. Thanks to Steve Coates for allowing access and helping with site selection. I thank the

staff of the Department of Wildlife Ecology and Conservation for their acceptance and support

during my tenure as a graduate student at UF. Elaine Culpepper, Dana Tomasevic, and Delores

Tillman helped with Einal dissertation preparation and presentation.

The Florida portion of this work was funded by the Florida Fish and Wildlife

Conservation Commission. I thank Dr. Robert Peet, the project Principal Investigator, for

providing this research opportunity. The Abita Preserve research was funded by the Louisiana

Field Offce of The Nature Conservancy. I thank the TNC employees who initiated the proj ect

and helped along the way: Nelwyn McInnis, Latimore Smith, Richard Martin, Judy Teague,

David Moore, and David Baker.

Finally, I thank my parents, Drs. Thomas and Glenna Carr. They taught me the value of

knowledge and scientific inquiry. Lastly, I thank my husband Mike Hoganson. Without his

love, support, and tolerance, I never would have finished this proj ect. He is my biggest fan.












TABLE OF CONTENTS

IM Le


ACKNOWLEDGMENT S .............. ...............4.....


LIST OF TABLES .........__.. ..... .___ ...............8....


LI ST OF FIGURE S .............. ...............9.....


AB S TRAC T ............._. .......... ..............._ 10...


CHAPTER


1 INTRODUCTION ................. ...............13.......... .....


2 VEGETATION CLASSIFICATION OF FLORIDA'S PYROGENIC PINELANDS ...........19
Introduction ........._..... ...._... ............... 19....
Methods............... ...............23

Study Area ................... ...............23..
Selection of Sample Sites............... ...............25.
Field M ethods .............. ...............27....
Numerical Analy sis............... ...............29
R e sults........._...... ...... .. ._. ...............3 1.....
Series 1: Dry Uplands ........._.._.. ...._... ...............33...
Series 2: M esic Flatwoods .............. ...............41....
Series 3: Wetlands............... ...............45
D discussion .................... .... ......... .......... .............5

Comparisons to Other Classifications............... ............5


3 GEOGRAPHIC, ENVIRONMENTAL AND REGIONAL VARIATION IN FLORISTIC
COMPO SITION OF FLORIDA PYROGENIC PINELAND S.....__ ............. ...... .........75
Introduction ............ _. .... ...............75....
Methods............... ...............78

Study Region............... .. .... ....... ........7
Vegetation and Environmental Data .....__.....___ ..........._ ............7
Numerical Data Assembly and Analysis .............. ...............81....
Re sults..................... .....__...... ...............86
Variation Partitioning Model s............... ...............87.
Environmental Explanatory Variables............... ...............8
Discussion ............ _. .... ...............90....


4 ECOLOGICAL RESTORATION OF A LONGLEAF PINE SAVANNA IN THE
SOUTHEASTERN COASTAL PLAIN ................. ...............105................
Introduction ................. ...............105................
M ethods............... ....... .... .. ..........0

Study Site and Reference Sites ................. ......... ...............109 ....












Restoration Treatments and Sampling Methods ................. .......... ................1 11
D ata Analysis ................. ..... .......... ............. ............11
Comparisons of ACP data to Reference Data ................. ..............................119
R e sults................ ............. .. ..... ...... ... ..........12
Trends in Species Richness and Woody Stems .............. ...............120....
Trends in Species Composition .............. .... ...... ..............12
ACP Treatment Responses vs. Reference Conditions ................. ............ .........124
Discussion .........._.... ... ... ._._. ........__. ..................12
ACP Restoration Compared to Reference Model ....._.__._ ........___ ................128
Management and Conservation Implications............... .............13


5 CONCLUSION ................. ...............143._._._.......


APPENDIX


A LOCATIONS OF FLORIDA VEGETATION PLOTS ........._..... ...._... ........_.._.....146


B LIST OF FREQUENT AND ABUNDANT SPECIES BY COMMUNITY
AS SOCIATION ............_ ..... ..__ ...............154...


C ABITA CREEK PRESERVE PLANT SPECIES ................. ...............163..............


LIST OF REFERENCES ................. ...............170................


BIOGRAPHICAL SKETCH ................. ...............184......... ......










LIST OF TABLES
Table page

2-1 Means and standard errors for soil and site variables by community series......................60

2-2 Means and standard errors of soil and site variables by community association. .............61

2-3 Common woody shrubs in midstory and understory strata by association........................62

2-4 Indicator species of Dry uplands and Mesic flatwoods associations ............... ... ........._...65

2-5 Indicator species of Wetland associations ...........__......_ ....___ ...........6

3-1 List of variables included in RDA and partial RDA ordinations ................. ................. .97

3-2 Monte Carlo tests of canonical axes for all RDA and partial RDA ordinations ................99

4-1 ANOVA tables for models of species richness and stem density ................. ................1 33

4-2 Results of Monte Carlo permutation tests for RDA ordinations ................. ................. 134










LIST OF FIGURES
Figure page

2-1 Physiographic landforms of Florida ................. ...............71........... ...

2-2 Florida plot locations by associations ................. ...............72...............

2-3 Non-metric multidimensional ordination of Florida species data ................. ................. 74

3-1 Venn diagrams of variation partition model ........._.._.. ...._... ............ ........0

3-2 Biplots of RDA ordination constrained by edaphic variables ................. ................ ..101

3-3 Biplot of partial RDA ordination constrained by edaphic variables ............... .... ........._..102

3-4 Biplots of RDA and pRDA constrained by climate variables ................. ................ ..103

3-5 Contour maps derived from constrained ordination axis scores displaying geographic
variation in variation partitions from the model of environmental-compositional
correlations............... .............10

4-1 Abita Creek Preserve : (a) pre-treatment in 1997, (b) immediately after logging in
1998, and (c) after logging and first prescribed fire in 2000 .............. .....................3

4-2 Least squares means and standard errors of small stems/1000-m2 Sample...........__........136

4-3 Least square means and standard errors of species richness by treatment and year........137

4-4 PCA ordination of ACP species data (1000-m2 Scale) ................. .........................13 8

4-5 Constrained RDA ordinations of pre-logged vs. first post-year species data ..................1 39

4-6 Constrained RDA ordinations of ACP species: all sample years ................. .................140

4-7 Number species per log sample area (m2). mean Species counts from ACP
treatments vs. Penfound and Lake Ramsey species richness............... .................4

4-8 Successional trends of ACP species data compared to Penfound and Lake Ramsey
reference site data .............. ...............142....









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

FLORISTIC AND ENVIRONMENTAL VARIATION OF PYROGENIC PINELANDS INT THE
SOUTHEASTERN COASTAL PLAIN: DESCRIPTION, CLASSIFICATION, AND
RESTORATION

By

Susan Catherine Carr

December 2007


Chair: George Tanner
Cochair: Kevin M. Robertson
Major: Wildlife Ecology and Conservation

Until recent times, the landscape of north and central Florida was dominated by fire-

dependent pineland savanna vegetation with sparse canopies of longleaf pine (Pinus palustris).

Economic development coupled with fire suppression lead to the drastic decline in the

distribution and integrity of these natural communities. I present a vegetation classification of

natural pineland communities in this highly fragmented landscape based on data collected over

large gradients of environmental and geological variation. I collected Hield data that quantified

species composition and abundance from 293 plots (from 103 sites) distributed throughout the

northern two thirds of Florida. After omission of species that occurred in < 3% of plots, a total

of 677 plant species were used in numerical analyses. I developed a vegetation classification

based on floristic similarity using K-means cluster and indicator species analyses. Three

ecological series were described corresponding to idealized moisture conditions. These were

further divided into 16 species associations. Floristic variation was related to geographic

separation between the panhandle and peninsula regions of Florida. I hypothesized that the

numerous plant species that have limited distributional ranges contribute to compositional










patterns. Similar geographic trends were apparent in a model of compositional variation related

to environment and spatial variation. Local environmental factors, including location on a

topographic/moistur gradient and soil fertility, were important correlates of local floristic

variation. Regional variation was correlated with soil texture and nutrient availability. A much

greater proportion of the explained variance was provided by environmental variables than by

pure spatial variables. The model revealed that both regional factors (climate, edaphic, and

geographic) and local factors (topographic position, soil chemistry) were correlates of with

floristic variation.

In addition to spatial variation, natural pineland communities undergo temporal variation

in response to periodic fires and changes in timber stand structure. Central questions regarding

ecological restoration of Coastal Plain pinelands are: how resilient are these communities

following anthropogenic alterations? Will ecological restoration affect vegetation succession

within the range of "natural" temporal variation? I studied ground cover vegetation response to

removal of woody biomass and reintroduction of natural fire regimes as it related to a program of

ecological restoration in a degraded pine savanna remnant. Treatment plots were thinned for

timber or else un-thinned as a control. Prescribed fire was applied at two subsequent times, and

changes in species composition were monitored over an eight year period. Species richness was

enhanced by mechanical woody reduction in the first two years, compared to sites that were

burned only (not logged). This response largely reflected increases in detectable graminoid

species. However, species richness of treatments converged within eight years, following two

prescribed fires 3 years apart. Species composition responded similarly, converging between

treatments over time. Succession was toward pre-settlement conditions, as suggested by

comparisons to reference sites and historical data. Community composition appears to be robust










to temporary alterations in fire regime and changes in timber stand structure within the range of

conditions studied. Spatial variation in species composition of pineland communities may be

relatively stable over time. Woody biomass reduction via careful mechanical logging does not

appear to adversely affect pineland vegetation recovery, and may expedite overall community

restoration.









CHAPTER 1
INTTRODUCTION

Pinelands of the Southeastern Coastal Plain are exceptional, both for their overall

biodiversity and degree of biotic endemism. The combination of large climatic gradients, long

growing seasons, variable geology and large species pools creates a prime environment for high

local and landscape scale floristic variation. Large compositional variation has been documented

both across geographic and local gradients (Peet and Allard 1993, Bridges and Orzell 1989).

Such high degrees of alpha, beta and gamma diversity (sensu Whittaker 1967) belie the

exceptional habitat specialization and regionalization of many species of the Coastal Plain.

Seemingly imperceptible topographic-moisture gradients coincide with almost complete changes

in plant species composition (Peet and Allard 1993). At larger scales, there is evidence of

regionalization of biota concurrent with geology, physiography, soils, and historical

biogeography. Accordingly, biogeographers have recognized and delineated distinct

"ecoregions" in Florida based on differences in environmental conditions and vegetation patterns

(Davis 1967, Brooks 1982).

In addition to biodiversity variation related to local and landscape gradients, levels of

endemism are exceptionally high in the Southeastemn Coastal Plain, and in Florida pinelands

specifically. Sorrie and Weakley (2002) report over 1600 taxa of plants endemic to the

Southeastern Coastal Plain. In addition to wide-ranging Coastal Plain endemics, many "narrow"

endemic species inhabit very restricted geographic regions and/or habitats. Florida is notable

both for the number of narrow endemic species it harbors, and the number of "centers of

endemism" (sensu Sorrie and Weakley 2002) located within the State, particularly relative to

plant species. In addition, Florida is home to over 2500 native plant species, many of which are

restricted in range of distribution, or habitat specificity (Wunderlin 2000).









Pine savannas and woodlands native to the Southeastern Coastal Plain are among the

most imperiled ecosystems in North America (Walker and Peet 1983, Croker 1987, Noss 1988,

Frost 1993, Peet and Allard 1993). Although they once dominated the landscape, native

pinelands now occupy less than three percent of their former range (Frost 1993, Outcalt and

Sheffield 1996). Of this, an even smaller area contains vegetation composition and structure

similar to that of pre-settlement conditions (Simberloff 1993). The rapid range reduction of

longleaf pinelands coincided with extensive logging, agricultural land use, and expanding rural

settlement in the 19th and 20th centuries (Croker 1987, Frost 1993).

Most contemporary native pinelands are small and fragmented, and are no longer subj ect

to the natural processes under which constituent species evolved. Most notably, this precludes

the natural occurrence of frequent, low intensity fires that historically swept across the landscape

(Frost 1993, Simberloff 1993, Platt 1999, VanLear et al. 2005). Fire suppression of longleaf pine

natural areas has contributed to large scale species replacement, as less fire tolerant pines and

hardwoods invade these pyrogenic communities (Glitzenstein et al. 1995, Platt 1999, Provencher

et al. 2000, VanLear et al. 2005). In the absence of frequent fire, thick growths of woody plants

compete with herbaceous vegetation for light and other resources, affecting succession and

community structure (Brockway and Lewis 1997, Provencher et al. 2001, VanLear et al. 2005).

Vegetation classification plays a key role in many areas of conservation, land

management and scientific research. Classification of vegetation delimits the number of relevant

natural communities to provide a conceptual framework for understanding the natural variation.

The process of delimitation is subj ective by nature. Much of this subjectivity resides in deciding

which data to use, what quantitative methods to use, and how to interpret the resulting solutions.









Some vegetation classifications of large landscapes also explicitly incorporate information about

geography (Peet and Allard 1993, Newell and Peet 1998, Wimberly and Spies 2001).

To date, classification systems developed specifically for Southeastern pineland

communities have been quantitatively rigorous but local in scope, or wide-ranging but

qualitative. Examples of the former are ecological classifications of vegetation are limited to a

specific management areas, usually on the scale of several thousand hectares (Carter et al. 1999,

Grace et al. 1999, Goebel et al. 2001, Abella and Shelburne 2004). Additionally, many

vegetation descriptions of Southeastern plant communities only describe woody species (e.g.,

Harcombe et al. 1993, others), thus missing important floristic "information" residing in the

ground flora (e.g. Bridges and Orzell 1989, Peet and Allard 1993, DeCoster et al. 1999, Schmitz

et al. 2002, Drewa et al. 2002). Classifications that do include explicit descriptions of

herbaceous vegetation are generally subj ective and anecdotal in nature. The exception is the

quantitative treatment of Feet and Allard (1993), which includes a regional classification of

pineland vegetation of the Coastal Plain emphasizing ground cover vegetation.

Traditionally, ecologists have studied the distribution of plant species according to

environmental factors (Bray and Curtis 1957, Peet 1978, Newell and Peet 1998). However,

recent studies have underscored the need for spatially explicit models of environmental-

composition variation (Legendre and Fortin 1989). Spatial trends are relevant in such models for

three reasons: 1) failure to account for spatially auto-correlated response data leads to biased

interpretations of environmental effects, 2) environmental determinants of vegetation

composition may be spatially structured, and 3) spatial autocorrelation independent of

environment suggests other control mechanisms of community composition (Legendre 2005).









Little is know about temporal variation in composition of pyrogenic pineland vegetation,

particularly compared to spatial variation. Studies of grasslands of other regions suggest that

temporal variation in species distributions is large compared to species-area relationships (Adler

et al. 2005). Longitudal studies of pineland community structure are rare and generally address

successional responses to specific treatments. Natural pinelands temporarily altered by unnatural

fire regimes and forest structure also provide opportunities to study the resiliency of pineland

vegetation to such alterations by quantifying community responses to restoration of natural

conditions (Walker and White 2006). A better understanding of temporal changes under typical

and degraded conditions will contribute to the applied models of pineland restoration, as well as

the models of "natural" variation used by conservationists.

I present a two-step process of vegetation classification and description of Florida

pyrogenic pineland flora. First, I classify pineland communities based on vegetation data alone

(primarily herbaceous ground cover vegetation). Second, I present a model of environmental-

composition correlations in a spatially explicit context. Finally, I present results of an ecological

restoration program of a degraded pineland remnant, and interpret vegetation responses in the

context of life history traits.

The classification of pyrogenic Florida pineland vegetation is based on 293 vegetation

plots (58.6 ha total) collected over a broad range of environmental conditions throughout the

range of longleaf pine in panhandle and peninsular Florida. The classification, derived from

floristic data alone, is presented as a system that can be used in land survey and management.

Sixteen community associations are described by environmental characteristics, diagnostic and

indicator species, general appearance and landscape context. I discuss floristic differences









between community association, and how range-restricted and endemic taxa influence

community variation.

A spatially explicit model of environmental and historical determinants is presented with

regard to the composition and diversity of pyrogenic pineland vegetation. Environmental factors

included edaphic, topographic, and climate variables, presumed to be operating at different

spatial and temporal scales. Variation related to pure spatial autocorrelation is hypothesized to

be indicative of biotic processes (not related to environmental determinants). Biogeographic

patterns were assessed by testing an "ecoregion" hypothesis of regionalization of community

variation. Significant environmental-composition correlations were used to generate hypotheses

regarding controls of community variation in Coastal Plain pinelands.

An experimental and longitudinal study of ecological restoration underscored the

resiliency of pyrogenic pineland groundcover plant communities. Changes mediated by

restoration treatments affected succession toward desired reference conditions. Furthermore, this

study suggested stability in succession even in atypical conditions of long fire-free intervals.

Temporal dynamics in pineland plant communities is quite variable (as are spatial trends), and it

is hypothesized that life history adaptations of typical plant species buffet the community over a

range of atypical environmental conditions.

As part of the restoration program, tree stand structure and frequent fire were restored in

a degraded pineland savanna remnant to resemble pre-settlement conditions. I measured the

effects of two restoration treatments on composition and diversity of native ground cover

vegetation. Restoration was measured as changes in composition relative to that of reference

sites which represented desired restored conditions. The larger question involved resiliency of a

specific pineland community to temporal changes in fire regime and timber stand structure.









From that, it may be concluded that pyrogenic pineland communities in general might be

relatively stable over time and over a range of conditions. Secondly, the study demonstrated

resiliency of native pineland vegetation following decades of man-induced fire suppression and

contributes to predictions of restoration success relative to starting conditions.









CHAPTER 2
A VEGETATION CLASSIFICATION OF FLORIDA' S PYROGENIC PINELANDS

Introduction

Fire-dependent pineland vegetation once dominated the landscape of the Southeastemn

Coastal Plain, ranging from southern Virginia south to the tip of Florida and westward to eastern

Texas. Frequent fires perpetuated the open aspect of pine savannas and woodlands, promoting

development of species-rich herbaceous ground cover vegetation. It is estimated that prior to

European settlement of the Gulf and lower Atlantic Coastal Plain regions, fire return intervals in

upland pinelands averaged once per 2-3 years (Martin et al. 1993, Olson and Platt 1995, Platt

1999, Glitzenstein et al. 2003). Following disruption of fire regimes, these communities are

rapidly colonized by fire-intolerant woody growth, prompting drastic alteration of community

composition and dynamics (Glitzenstein et al. 1995, Platt 1999, Glitzenstein et al. 2003).

Economic development removed native pineland vegetation from much of its former range in the

Coastal Plain, particularly from the finer-textured soils that readily support agriculture (Frost

1993, Frost 2006). Native longleaf pinelands currently occupy less than three percent of their

former range (Frost 1993, Outcalt and Sheffield 1996). Even rarer are Coastal Plain pineland

communities managed with fire regimes that mimic those of pre-settlement conditions

(Simberloff 1993, Varner et al. 2005).

Fire-dependent pineland communities Florida are exceptional both for their overall

biodiversity and the degree of biotic endemism. Over 1600 plant taxa are endemic to the

Southeastern Coastal Plain, and over 250 of these are endemic or near-endemic to Florida (Ward

1979, Kautz and Cox 2001, Sorrie and Weakley 2001, Sorrie and Weakley 2006). The Florida

peninsula has a complex geologic history of inundation and land expansion related to sea level

change and glaciation. Ancient islands isolated during sea level rise gave rise to many endemic










species of contemporary highlands and ridge provinces and other regions served as glacial

"refugia" (Webb 1990). Florida is notable for the number of "centers of endemism" (sensu

Sorrie and Weakley 2002) located in the State. More than 2500 plant species are native to

Florida, representing a mixture of temperate and tropical species that changes with latitude

(Holdridge 1967, Ward 1979, Wunderlin 1998).

The combination of large climatic gradients, long growing seasons, variable geology and

large species pools in Florida creates a prime environment for exceptional floristic variation at

local and landscape scales. Florida has the third richest flora of all States (Wunderlin and

Hansen 2000). Plant species richness of Florida pinelands are among the highest recorded at

small scales (Walker and Peet 1983, Peet 2006). In addition, subtle topographic-moisture

gradients can harbor almost complete turnover in plant species composition (Bridges and Orzell

1989, Abrahamson and Hartnett 1990, Peet and Allard 1993, Platt 1999). Such a high degree of

"beta" and "gamma" diversity (sensu Whittaker 1962, 1967) belies the exceptional habitat

specialization and regionalization of many pineland species. On a landscape scale, there is

evidence of aggregation in floristic and community similarity associated with specific regions.

Accordingly, Florida "ecoregions" have been recognized and delineated based on similarity of

edaphic, geologic, physiognomic, and vegetative features (Puri and Vernon 1964, Davis 1967,

Brooks 1982, Brown et al. 1990).

Floristic classification systems provide a conceptual framework for understanding natural

variation across environmental and geographic gradients. Such systems are widely applied in

ecological inventory, conservation, and management (1990, Grossman et al. 1998, Comer et al.

2003). To be useful in the field, a vegetation classification should provide detailed information

regarding frequent and abundant species, as well as those that are diagnostic of specific









associations (i.e., "indicator species"; Defrene and Legendre 1997). Ideally, a classification

would also describe relevant environmental attributes, including typical ranges of variation. A

comprehensive account of floristic types and variation could aide ecological restoration

programs by providing a range of reference conditions and guiding land conservation priorities

(White and Walker 1997, Walker and Silletti 2006).

Vegetation classification based on quantitative data is very much dependent on sampling

design, intensity, and breadth (Nekola and White 1999, Cooper et al. 2006). Random and area-

proportionate sampling designs are often not practically possible in large regions containing

fragmented landscapes with variable natural conditions and mixed land ownership, land use

history, and degree of public access. However, subjective bias can be minimized by application

of a stratified sampling design which promotes balanced sampling intensity and effort across

gradients of interest (Leps and Smilauer 2007). Such a design may not yield an unbiased

representation of variation of pre-settlement natural vegetation, but may facilitate a

representative sample of contemporary natural vegetation in a highly modified landscape (e.g.

most of Florida).

Classification systems differ in many respects, including geographic and environmental

scope, and type and quality of input data. Many vegetation classifications are strictly qualitative

and descriptive (FNAI 1990, Grossman et al. 1998, Comer et al. 2003), although widely used for

community classification and conservation policy guidelines in Florida. These works are based

on expert accounts offloristic variation over a large region. Conversely, quantitative

classifications typically incorporate site specific vegetation data. Depending on program

obj ectives, abiotic environmental attributes are either explicitly included in the classification or

are presented as descriptors or explanatory factors offloristically defined types. "Ecosystem









classification" and "ecological landtype phases" typify the former approach (Cleland et al. 1993,

Hix and Pearcy 1997, Goebel et al. 2001, Abella et al. 2003). Regional vegetation classifications

that include all or part of Florida are of the latter type, based on quantitative data of species

abundances collected using standardized sampling methodology (Peet and Allard 1993, Peet

2006). In the present study, abiotic variables are descriptors of community classifications,

including soil properties and geology. The quantitative delineation of floristic data approach has

several advantages: 1) it encourages objectivity in classification partitioning, 2) it allows a

posteri examination of relationships between abiotic variables and community types, which can

be useful for inventory and predictive modeling; 3) it may uncover "unexplained" gradients of

floristic variation, stimulating generation of hypotheses regarding determinants of biodiversity

(McCune and Grace 2002, Leps and Smilauer 2003, Legendre et al. 2005, Leps and Smilauer

2007).

I present a quantitative classification of fire-adapted pineland vegetation of northern and

central Florida. The study region includes the entire historic range of longleaf pine in Florida.

My focus was the classification of natural communities: i.e. frequently burned (at least 2-3 times

over the past two decades) vegetation of pinelands and associated communities relatively

unaltered by soil disturbance or severe fire suppression. My objective was to characterize plant

communities based on floristic assemblages alone, followed by descriptions of geographic

distribution, topographic context, and soil characteristics. Community descriptions include

identification of dominant and diagnostic plant species, facilitating easy field recognition of

characteristic vegetation. My sampling design, coupled with an obj ective approach to cluster

analysis, yielded a comprehensive yet manageable classification of 16 associations. I describe

edaphic and landscape features that are useful for field identification, such as soil texture









attributes and landscape context. Furthermore, I describe geographic and environmental trends

in floristic similarity among pineland associations as they relate to distribution and identification

of community types.

Methods

Study Area

The study area included the entire Florida Panhandle and most of central and northern

Peninsular Florida. This area extends south from the State border to a southern boundary

extending from roughly 260 70' latitude on the west coast to 280 80' on the east coast (Figure 2-

1). This area roughly coincides with the current range of longleaf pine in Florida (Figure 2-2(a)).

This range is thought to represent the historic longleaf pine range in Florida (Platt 1999 and

references within), although there is some evidence that historic distribution extended farther

south. The southern boundary also approximates the southern extent of the "warm temperate

moist forest" bioclimate zone, separating it from the "subtropical moist forest" zone (Holdridge

1967).

Three generalized land units of Puri and Vernon (1964) subdivide the Florida study

region according to common geologic history. These generalized land units describe geographic

regions: 1) Northem Highlands, 2) Central Highlands, and 3) Coastal Lowlands (Figure 2-1).

These are further subdivided according to physiographic landforms, which describe maj or soil

types, geology and prevailing landscape features (Puri and Vemnon 1964, Myers 1990). These

are 1) Highlands; 2) Ridges, Hills, Inclines and Slopes; and 3) Lowlands, Gaps, and Valleys.

The Northern Highlands of the upper panhandle lie north of a prominent ancient

Pleistocene shoreline known as the Cody Scarp (Myers and Ewel 1990). This region is

distinguished by broad expanses of continuous highlands. The Western and Tallahassee









Highlands, New Hope and Grand Ridges, and Marianna Lowlands landforms comprise the

Northern Highlands land unit (Puri and Vernon 1964). The first two have dissected topography

and plastic sediments of mainly Appalachian origin from the Miocene epoch (20 to 5 million

years before present; Puri and Vernon 1964, Brown et al. 1990, Myers 2000). The Marianna

Lowlands landform contains outcrops of Eocene and Oligocene carbonates in a low lying

anticline (Puri and Vernon 1964, Brown et al. 1990). Although lower than the first two

landforms, it is higher than the Coastal Lowlands, and is generally well-drained owing to sandy

soils shallowly overlying limerock perforated by sink holes (Brown et al. 1990). Ultisols are

common upland soils of Northland Highlands, although Entisols typify Citronelle Formation

uplands in the Western Highland portion as well as the sandy uplands of central panhandle

Ridges. The Central Highlands land unit contains discontinuous highlands of the central

Peninsular ridge system amid lower and flatter landforms (Figure 2-1). The former are

landforms of the Ridges, Uplands, and Slopes and Highlands types while the latter are Lowlands,

Gaps, Valleys and Plains (Puri and Vernon 1964).

The Central Highlands and the Northern Highlands approximate the emergent portion of

the Wicomico shoreline, an early Pleistocene shoreline of high sea level. This region was once

an integrated highland that has since been partitioned by erosion and solution (Puri and Vernon

1964). The Ridges and Uplands of the peninsula arose from ancient shorelines, dune systems,

barrier islands, and associated terraces (Puri and Vernon 1964). Larger ridge systems of the

Central Highlands include the Brookville, Deland, Trail, Mount Dora and Lake Wales Ridge

physiographic landforms, and maj or Uplands include Sumter, Polk, Marion, Duval and Lake

landforms. Soils are mainly coarse, excessively drained Entisols and loamy Ultisols. Soils of









Lowlands landforms are typically Spodosols underlain by limestone of the Florida peninsula

platform (Brown et al., 1990).

The Coastal Lowlands land unit includes the southern tier of the panhandle below the

Cody Scarp, in addition to the coastal regions of the peninsula (Figure 2-1). Much of this region

has been subj ected several marine inundations during the Late Miocene to the Early Pliocene

(Puri and Vernon 1964, Webb 1990). Most of the Coastal Lowlands region contains Lowlands,

Gaps, Valleys, and Plains physiographic landforms. These are broad plains with little relief,

containing poorly drained Spodosols (Brown et al. 1990).

Selection of Sample Sites

The focus of this study was fire-dependent plant communities of Florida containing

herbaceous-dominated ground cover vegetation. This included many types of pine woodlands

and savannas, variously labeled pine flatwoods, sandhills, high pine, piney woods, mesic

flatwoods, wet flatwoods, and scrubby (or xeric) flatwoods. Also included were fire-dependent

herbaceous dominated communities associated with pinelands, such as prairies, bogs, lake

margins, and seepage slopes. These communities are naturally characterized by frequent, low-

intensy fires in which herbaceous vegetation and litter provide the dominant fuel matrix (Platt

1999). I omitted scrub and maritime pinelands of Central Florida and coastal regions, which are

typically characterized by crown fires in the shrub or tree layers and have relatively longer fire-

free intervals (Myers and Ewel 1990).

Although the Florida range of longleaf pine is the large scale region of interest of this

study, descriptions of pyrogenic communities were not restricted to longleaf pine dominated

sites. The geographic and habitat scope of this study included all pineland and associated

communities within the longleaf pine range of Florida. Sites lacking pine overstory were









included in the study based on their similarity in ecosystem processes and herbaceous ground

cover structure and diversity to pine-dominated sites. Such sites often represented topographic-

moisture extremes in otherwise pine-dominated landscapes.

The generalized physiographic landforms of Puri and Vernon (1964) were further

subdivided into "ecoregions" to guide site selection and stratifieation. This ensured a

representative sample of physiographic environments throughout the area of study. I delineated

ecoregions based on homogeneity of geology, vegetation, soils, climate and physiography,

following several published works (Fenneman 1938, Puri and Vernon 1964, Davis 1967, Fernald

1981, Brooks 1982, Bailey et al. 1994, Griffith et al. 1994). There were a total of 19 ecoregions

in the study region. I present classification results relative to physiographic landforms, of which

ecoregions were subsets.

I stratified sampling by ecoregions and topographic-moisture conditions. Roughly equal

numbers of sites were selected per ecoregion depending on site availability and accessibility. To

the best of my ability, I selected three high quality sites in different locations within each

ecoregion. Ideally, each site contained an intact, continuous topographic-moisture gradient

supporting frequently burned native vegetation. Unfortunately, sites that satisfy this condition

are rare or absent in some regions, particularly those that lack large tracts of public land. Under

these conditions, I relaxed selection criteria to include: 1) sites that contained intact topographic-

moisture gradients, but lacked optimal fire history, and 2) sites with acceptable fire history but

lacking intact gradients. In the latter situation, I pieced together a representative topographic-

moisture gradient from several sites located in close proximity. Additional criteria were

considered in site selection: 1) little or no recent man-made ground disturbance, 2) absence of

invasive exotic species, 3) presence of native canopy and midstory tree composition and









structure, and 4) evidence of Gire within the previous Hyve years, and preferably a history of

frequent fires during the previous 50 years. In general, the integrity of the ground cover

vegetation was emphasized over structure of the tree canopy in selection evaluations. Candidate

sites were identified from various sources, including the Florida Natural Areas Inventory natural

community database (FNAI 2000a) and consultation with regional natural resource

professionals. Three sites (12 plots) were selected in South Georgia (within 20 miles of the

Florida state border). I assumed that vegetation of these sites were representative of Florida

pinelands in the same ecoregion. A total of 102 sites were selected (see Appendix A)

Field Methods

Once deemed suitable for sampling, a site (or a composite site) was delineated into three

or four topographic-moisture zones based on Hield observations. Sampling from a range of

topographic-moisture conditions maximized inclusive sampling of local vegetation associations

presumed to be associated with specific soil conditions. One 1000 m2 rectangular plot was

established in each zone such that the plot area encompassed an area of relatively homogenous

vegetation. The starting point of the long plot axis was randomly assigned. Usually the main

axis of the 50 x 20 m plot was oriented parallel to slope contours.

Vegetation sampling methodology followed the Carolina Vegetation Survey (CVS)

sampling protocol (Peet et al. 1998). The basic sampling unit was a 1000 m2 plOt (dimensions 50

x 20 m). Four 100-m2 "mOdules" were situated in each plot, each containing two sets of nested

sub-plots (0.01, 0.1, 1, and 10-m2). All vascular plant taxa were recorded as they were

encountered in the sequentially sampled nested sub-plots. I estimated the aerial cover of each

taxon in 100-m2 mOdules using cover classes: 1 = 0-1%, 2 = 1-2%, 3 = 2-5%, 4 = 5-10%, 5 =

10-25%, 6 = 25-50%, 7 = 50-75%, 8 = 75-95%, 9 = >95%. Mean cover estimates were









calculated from four module cover midpoints. Taxa encountered in the remaining 600-m2 plOt

area were tallied and assigned nominal cover estimates. In the 1000-m2 plOts, all woody stems >

1 cm and < 40 cm diameter at breast height (dbh) were tallied by species and 5 cm diameter

class. Stems > 40 cm dbh were measured and recorded individually. In plots with very sparse

woody vegetation, I sampled stems in a larger area (2000-m2) to obtain better estimates of stem

density and basal area.

All plots were sampled during the late summer though early winter (August-December).

Sampling flora in the late growing season increased my ability to identify the copious numbers

of graminoids and fall-flowering forbs typical of Southeastern pinelands. A total of 293 plots

were sampled over 4 years (2000 2004).

The maj ority of sampled taxa were identified to species or variety. Some taxa received

lower levels of taxonomic resolution due to problems with consistent field identification. Where

variation in taxonomic resolution existed, I used the lowest resolution necessary to ensure

consistency throughout the dataset. The term "species" is used to indicate the highest resolution

of identification, be it genus, species or variety. Nomenclature generally follows Kartesz (1999)

with a few exceptions. In field and herbarium plant identification I made frequent use of

(Godfrey and Wooten 1979, Godfrey and Wooten 1981, Clewell 1985, Godfrey 1988, Wunderlin

1998, Weakley 2002). Approximately 2500 voucher specimens were deposited in the University

of Florida herbarium in Gainesville, Florida.

Four surface soil samples were collected per plot. Each sample of approximately 250 g

was collected to 10 cm depth. Sub-soil samples were collected from a single point

approximately 50 cm below ground surface. Samples were dried and sent to Brookside Labs in

New Knoxville, Ohio for nutrient and textural analyses. Texture analysis determined










compositional percentages of sand, silt, and clay particles in the surface and sub-soil samples. In

addition, percent organic matter, pH, and exchangeable cations in ppm (Ca, Mg, K, Na) were

measured in surface soil samples.

Numerical Analysis

A matrix of species data was assembled from the 293 census plots hereafter referred to as

samples. Samples represent different topographic-moisture locations within sites. Pine species

(genus Pinus) were omitted from the species matrix, although other woody species were

retained. Species with fewer than three occurrences in were deleted from the final data matrix,

as rare species contribute little to calculations of inter-plot similarities (McCune and Grace

2002). The dimensions of the final response matrix were 293 samples x 575 species.

I transformed the species response matrix prior to multivariate analyses following the

guidelines of Legendre and Gallagher (2001) and McCune and Grace (2002). First, species

responses were relativized to maximum species cover values which tends to de-emphasize the

influence of common and abundant species. Then the species response matrix was transformed

using the Hellinger distance transformation. When used in conjunction with Euclidean distance

metrics this transformation improves representation of multidimensional data in low dimensional

space and avoids problems inherent to sample weighting (in chi-square based ordinations) in

addition to problems associated with using Euclidean distances with untransformed data

(Legendre and Gallagher 2001, Legendre et al. 2005).

I used a combination of ordination and cluster analyses to partition samples into

floristically similar groups. Specifically, I used non-hierarchical Euclidean-based K-means

cluster analysis to partition samples into a configuration that minimized within group sum of

squares relative to between group differences (Legendre and Legendre 1998). Partitions are









user-defined, so I used the "cascading K-means" function of the Vegan package (Oksanen et al.

2007) as implemented in R statistical software (R Development Core Team 2007). Cluster

analysis was run multiple times using various numbers of user-defined partitions (2 to 40

groups). I selected the number of partitions that maximized an optimization index, specifically

the "Simple Structure Index" (SSI). The SSI quantifies three elements of a partition model:

maximum difference of each species response between clusters, the sizes of the most contrasting

clusters and the deviation of species responses per cluster compared to its overall mean (Oksanen

et al. 2007).

The Einal partition model presented clusters of samples representing recognizable and

distinct floristic assemblages. I refer to these clusters as "associations". I graphically displayed

associations in a non-metric multidimensional scaling (NMS) ordination of Euclidean distances

derived from the Hellinger transformed species matrix. For this I used PC-ORD software,

version 5.0 (McCune and Mefford 1999).

Diagnostic species were recognized for each association, in terms of constancy and

fidelity. I used Indicator Species Analysis of Dufrene and Legendre (1997) implemented in PC-

ORD (McCune and Mefford 1999). The Indicator Value (IV) index quantifies a species' relative

frequency and abundance among associations. Indicator species were identified using Monte

Carlo randomization tests (McCune and Mefford 1999); the null hypothesis was that the

maximum IV among associations is no larger than would be expected by chance. Indicator

species were considered those with type I error < 0.05 in the IV randomization test.

From the species recognized as indicators for associations, I identified those with

restricted distributions in Florida. A species was identified as having "restricted range" if its

Florida distribution was limited to only one of three regions (Western Panhandle, Panhandle plus









North peninsula, or Central Peninsula), or if its entire range was limited to Florida. Species'

Florida distributions were categorized by visual inspection of on-line county range maps

available from the Institute of Systematic Botany Atlas of Florida website (Wunderlin and

Hansen 2004).

I compared soil characteristics and other community attributes among individual

associations, and between three higher level groups of associations (termed ecological "series").

Within ecological series, means and pairwise comparisons of response variables among

associations were analyzed with univariate ANOVA' s. In addition to soil variables, I compared

species richness (number of species /1000-m2 Sample) and basal area (m2/ha) between

associations. Response variables were transformed to improve normality of residual

distributions in each model. Count variables were log transformed, and logit transformations

were applied to proportion response variables (Tabachnick and Fidell 1996). I maintained a

Type I error of p < 0.01 for each pairwise comparison to reduce the overall Type I error

associated with each response variable. All ANOVA and post-hoc tests were performed using

SAS software, version 9 (SAS 2000).

Results

A total of 293 samples spanning the study region were included in the K-means cluster

analysis of mean species cover responses (Figure 3a). I identified 16 associations from the

optimal cluster solution. This partition yielded the second highest value of SSI (0.23, maximum

value = 1.0) among all partitions of 2 to 40 groups. Although the 28 group partition had a higher

SSI value (0.25), I chose the 16 group partition because it presented interpretable results with

relatively balanced cluster sizes, with no clusters containing fewer than four samples.









The 16 associations encompass a wide range of floristic variation over environmental

conditions. The primary gradient of variation, displayed by the first NMDS ordination axis,

concurs with a priori assigned topographic-moisture conditions (Figure 2). The correlation

between distances in ordination space (two dimensional NMS solution) versus distances in

original space was R2 = 0.83 (McCune and Mefford 1999, McCune and Grace 2002). The first

axis represents most of this variation (R2 = 0.54).

I categorized the 16 associations into three ecological series, which are superimposed on

the ordination diagram: Dry Uplands (D), Mesic Flatwoods (M), and Wetlands (W).

Associations were named using existing vernacular in plant community descriptions: sandhills,

clayhills and woodlands describe dry upland communities of varying canopy density and soil

texture; mesic flatwoods refer to pine savanna communities of poorly-drained flat terrain.

Occasionally to seasonally inundated wetlands are represented by various terms depending on

canopy density and moisture conditions, including wet flatwoods, wet prairies, and seepage

slopes (FNAI 1990, Myers and Ewel 1990, Peet and Allard 1993). Modifiers were added to

distinguish landscape and regional affinities.

One hundred and six species were categorized as having restricted ranges in Florida.

Eight species are endemic to Florida. The remaining 98 species have provincial distributions,

and are restricted to one of three regions in Florida: 1) Panhandle only, 2) Panhandle and north

peninusular Florida, and 3) peninsula only (Tables 4 and 5).

Associations are described below in terms of community aspect, soil characteristics, and

species composition. Throughout, the tables and appendix are referenced for the following: soil

and community attributes (Tables 2-1 and 2-2), common canopy and midstory woody species

(Table 2-3), indicator species of associations (Tables 2-4 and 2-5), and frequent and abundant










ground cover species (Appendix B). In addition, endemic and restricted-range indicator species

are indicated in Tables 2-4 and 2-5. Maps of plot locations are shown in Figure 2-2.

Physiographic and landscape attributes for associations are described, and follow the conventions

of Figure 2-1. Labels and cluster sizes are noted following association name. Association

descriptions are grouped into three major ecological series corresponding to Figure 2-3.

SERIES 1: Dry Uplands

Dry Uplands included six associations, which were categorized as sandhills, woodlands

or clayhills. The Dry Uplands associations were located within the Northern Highland and

Central Highland generalized land units, primarily within the Ridges and Uplands physiographic

landforms. In general, these associations occurred on ridgetops and upper slopes in areas with

topographic relief exceeding several meters. Soils of Dry Uplands were sandy and low in

organic content. Compared to Mesic Flatwoods and Wetlands series, Dry Uplands sand content

was high in surface soils and low in sub-soils. Soil pH was intermediate compared to other series

(Table 2-1).

The six Dry Upland associations exhibited geographic segregation relative to floristic

composition. The Ochlochnee River basin in the eastern Panhandle distinctly separated

associations of the Northern Highlands and Panhandle Coastal Lowlands from those of the

Central Highlands and peninsular Coastal Lowlands. Dry Uplands of the Northern Highlands

landform occurred on both Pliocene and Pleistocene deposits, including the Citronelle and

Torreya formations and the undifferentiated deposits of the lower Apalachicola basin (Puri and

Vernon 1964, Brown et al. 1990). East of the Ochlochnee River, Dry Upland associations

occurred primarily on Miocene and Pliocene deposits of the Central Highlands land unit,

specifically within the Ridges, Uplands and Slopes physiographic landforms.









Dry Uplands soil properties reflected those of Entisols and Ultisols, which are common

upland soil orders (Brown et al. 1990, Myers 1990, Myers 2000). Segregation of associations

coincides with soil clay and silt content. The well-developed Ultisols of the PANHANDLE

LONGLEAF PINE CLAYHILLS and PANHANDLE SILTY WOODLANDS had argillic sub-surface strata

enriched with clay and silt. Soil moisture availability is typical greater in these soils (Brown et

al. 1990, Brady and Weil 2000). The Dry Upland associations of the panhandle spanned a range

of soil texture composition. Conversely, Dry Upland associations of the peninsula did not

exhibit surface soil texture gradients, but variation was apparent in sub-soil silt and clay content

and organic content. All Dry Upland soils were similar in pH, with the exception of the

PANHANDLE LONGLEAF PINE CLAYHILLS association. In addition, fine-textured soil content was

positively correlated with species richness and canopy density. A description of individual

associations within Dry Uplands series follows.

PENINSULA XERIC SANDHILLS (22 plots, D3): This association is restricted to high

sandy ridges of the Central Highlands and Coastal Lowlands of the northern peninsula region

(Figure 2-2c). PENINSULA XERIC SANDHILLS soils consist of coarse sands with low

concentrations clay and silt. This association is species-poor compared to other Dry Upland

associations, although comparable to the PANHANDLE XERIC SANDHILLS further west.

Pine canopy of PENINSULA XERIC SANDHILLS was sparse. Longleaf pine (Pinus.

palustris) was the dominant canopy species (mean BA = 4.8 m2/ha), followed by turkey oak

(Quercus leavis; mean BA = 1.9 m2/ha). Common midstory species included turkey oak, sand

live oak (Q. geminata), saw palmetto (Serenoa repens) and bluej ack oak (Q. incana). Sand post

oak (Q. margarettiae), an oak common in other sandhill associations, was notably infrequent.









The most common herbaceous plants of PENINSULA XERIC SANDHILLS were grass and

forb species. Frequent grasses are wiregrass (Aristida beyrichiana), lopsided indiangrass

(Sorgha~strunt secundum), little bluestem (Schizachyrium scoparium var. stoloniferunt), and

eggleaf witchgrass (Dichanthelium ovale). The forbs silk grass (Pityopsis gra~ninifolium),

pineland pinweed (Lechea sessiliflora) and queens delight (Stillingia sylvatica) were common.

A few grasses were identified as indicator species, including pineywoods dropseed

(Sporobohes junceus), perennial sandgrass (Tripla~sis amnericana), and big threeawn

(Aristida condensata). The remaining indicator species were forb species common to xeric

habitats: Ware's hairsedge (Bulbostylis warei), coastal plain honeycombhead

(Balduina angustifolia), and pineland pinweed (Lechea sessiliflora). Coastal plain chaffhead

(Calrphephorus corynabosus) and wholly pawpaw (Asinzina incana: a small shrub) are indicator

species with ranges restricted to the peninsula. Two indicator species are legumes: eastern

milkpea (Galatia regularis) and scurf hoarypea (Tephrosia chyrsophylla). Legumes are

typically common to communities of finer-textured soils (James 2000).

PANHANDLE XERIC SANDHILLS (3 1 plots, D4): Sites of this association were restricted to

the Northern Highlands land unit (Figure 2-2b), primarily west of the Ochlochnee river basin.

PANHANDLE XERIC SANDHILLS were observed in two landscape contexts: 1) on sandy ridgetops

and upper slopes, and 2) as the dominant community of broad flat terrain with little apparent

topographic variation. I observed the latter situation on the broad continuous uplands of the

Citronelle formation in Eglin Air Force Base.

PANHANDLE XERIC SANDHILLS were similar in aspect to PENINSULA XERIC SANDHILLS.

Sparse canopies consist of scattered longleaf pines (P. pahtstris: mean BA = 7.9 m2/ha) and

turkey oak (Q. leavis: mean BA = 1.1 m2/ha). Midstory strata were dominated by turkey oaks,









bluejack oak (Q. incana) and sand live oak (Q. geminata). Unlike the PENINSULA XERIC

SANDHILLS, sand post oak (Q. margaretta), dwarf live oak (Q. minima), and dwarf huckleberry

(Gaylussacia dumosa) were common in PANHANDLE XERIC SANDHILLS.

Frequent species of PANHANDLE XERIC SANDHILLS included few grasses, most notably

little bluestem (Schizachyrium scoparium var. stoloniferum) and Elliotti's bluestem (A. gyrans

var. gyrans) with a low frequency of wiregrass (Aristid'a stricta). Herbaceous species of xeric

habitats distinguished PANHANDLE XERIC SANDHILLS ground cover. About a third of indicator

species have ranges restricted to the Panhandle, including piedmont gayfeather (Liatris

pauciJlora var. secunda), littleleaf milkpea (Galactia microphylla), Morh' s threeawn (Aristida'

morhii), Godfrey pineland hoarypea (Tephrosia morhii), royal snoutbean (Rhynchosia cytisoides)

and greater Florida spurge (Euphorbia florid ana).~dd~~~dd~~ The provincial herb Pityopsis aspera is

abundant and frequent (90% of plots). In contrast, P. aspera is absent from PENINSULA XERIC

SANDHILLS were P. gramninifolia is usually dominant.

NORTH FLORIDA SANDHILLS (3 1 plots, D2): Sites of this association occurred on

various landforms of the Coastal Lowlands and Central Highlands of the eastern panhandle and

northern peninsula (Figure 2-2c). NORTH FLORIDA SANDHILLS are usually found on ridgetops

and upper slopes. Soils were similar in textural composition to PENINSULA XERIC SANDHILLS

except they had higher silt content. In addition, they were very low in clay and organic matter.

Similar to other Dry Upland associations, NORTH FLORIDA SANDHILLS have canopies of

longleaf pines (mean BA = 8.6 m2/ha) with scattered upland oaks (most abundant: turkey oak,

mean BA = 1.2 m2/ha). The three common upland oaks dominate the midstory: turkey oak,

bluej ack oak (Q. incana) and sand post oak (Q. margaretta).









Common grasses of NORTH FLORIDA SANDHILLS are similar to those of PENINSULA XERIC

SANDHILLS: wiregrass, little bluestem, lopsided indiangrass, and eggleaf witchgrass. Other

frequent grasses include needleleaf witchgrass (D. angustfolium) and thin paspalum (Paspalum

setaceum). Many frequent species are low growing forbs, as are 14 of the 15 indicator species.

Three of these are legumes, including Florida ticktrefoil (Desmodium floridanum),~dddd~~~ddd~~~ dollarleaf

(Rhynchosia reniformis), and hairy lespedeza (Lespedeza hirta). None of the NORTH FLORIDA

SANDHILLS indicator species have restricted distributions. Species richness of NORTH FLORIDA

SANDHILLS is notably higher compared to the xeric Dry Uplands, and likely contributed to the

floristic segregation among these associations.

NORTH FLORIDA RICH WOODLANDS (11 plots, DI): This association includes longleaf

pine woodlands of mid- and lower slopes in the Central Highlands and Coastal Lowlands of the

northern peninsula (Figure 2-2c). These sites were usually adjacent to hardwood hammocks. All

NORTH FLORIDA RICH WOODLANDS sites were in or adj acent to vegetation zones identified as

"Hardwood hammocks" by Davis (1967). Most were located downslope of NORTH FLORIDA

SANDHILLS. NORTH FLORIDA RICH WOODLANDS soils had physical properties similar to the

three preceding Sandhills associations. However, they were distinguished by their very high

organic content and sub-surface clay content. These soil attributes suggest higher water retention

capacity (Brady and Weil 2000).

Canopy densities of NORTH FLORIDA RICH WOODLANDS were high relative to other Dry

Uplands associations (mean BA = 16. 1 m2/ha). Longleaf pine dominated canopies (mean BA =

8.5 m2/ha), but other subdominant pine species were present: slash pine (P. elliottii var. elliottii)

and loblolly pine (P. taeda; mean BAs = 1.3 and 1.2 m2/ha respectively). Sand live oak (Q.

geminata) and mockernut hickory (Carya alba) were canopy sub-dominants (mean BAs = 1.5









and 1.4 m2/ha respectively). Midstory strata were generally shrubbier compared to other Dry

Upland associations, dominated by saw palmetto (S. repens) followed by winged sumac (Rhus

coppelinum), mockernut hickory (C. alba) and two upland oaks (Q. geminate and Q.

margaretta).

Common herbaceous species of NORTH FLORIDA RICH WOODLANDS included grasses

typical of other upland associations, as well as some distinctive woodland forbs. Common

grasses were Elliott' s bluestem (A. gyrans var. gyrans), needleleaf witchgrass (D. angustfolium),

thin paspalum (P. setaceum), lopsided indiangrass (Sorgha~strum secundum), broomsedge

bluestem (A. viriginicus), and eggleaf witchgrass (D. ovale). Wiregrass (A. stricta) was present

in only about 50% of the plots, and was sparse compared to other Dry Upland associations.

Bracken fern (Pteridium aquilinum), laural greenbriar (Smilax laurifolia), whitetop aster

(Sericocarpus tortifolius), and lesser snakeroot (Ageratina aromatica) were ubiquitous

herbaceous species. Almost all indicator species were woodland forbs and infrequent grass

species, including seven grasses and five legumes. About half of the indicator species were

species with restricted ranges. Species richness of North Florida Rich Woodlands was

intermediate compared to other Dry Uplands (106 species/0.1 ha).

PANHANDLE LONGLEAF PINE CLAYHILLS (14 plots, DS): These sites are restricted to

the Northern Highlands land unit of the panhandle. PANHANDLE LONGLEAF PINE CLAYHILLS

inhabit the ridgetops and upper-slopes of dissected Pliocene and Miocene-aged sediments north

of the Cody Scarp (Figure 2-2b). In the western panhandle, this association occupied mid-slopes

in association with PANHANDLE XERIC SANDHILLS. The prominence of fine-textured sediments

distinguishes soils of PANHANDLE LONGLEAF PINE CLAYHILLS. Sub surface silt and clay content









was high compared to other Dry Uplands. Similarly, soil pH was higher than all other Dry

Uplands associations.

PANHANDLE LONGLEAF PINE WOODLANDS canopies were dense and dominated by

longleaf pine (mean BA = 10.9 m2/ha), with minor contributions of loblolly and shortleaf pines

(P. taeda and P. echinate; mean BAs = 1.2 and 0.4 m2/ha respectively). In addition to the

typical Dry Uplands oak species, shrubs of more mesic habitat occupy midstory strata such as

southern red oak (Q. falcata), running oak (Q. pumila), and Darrow' s blueberry (y. da~rrowii).

Species richness of PANHANDLE LONGLEAF PINE CLAYHILLS is exceedingly high. The

mean of 124.5 species/0. 1 ha is the highest of all Dry Uplands associations. Dense herbaceous

ground cover vegetation contained numerous forb and grass species. Wiregrass (Aristida

stricta, little bluestem (Schizachyrium scoparium var. stoloniferum) and narrowleaf witchgrass

(D. angustifolium) were ubiquitous. Indicator species included many legumes and composites

(members of the Asteraceae family). Ten out of 25 indicator species are legume species and

many of these are in the genus Desmodium. Eleven of the 25 indicator species have ranges that

were restricted to the panhandle or northern peninsula. Several bunch grass species with

restricted ranges were identified as indicators: big bluestem (A. gerardii), cutover muhly

(M~uhlenbergia cappilaris var. trichopodes), Carolina fluffgrass (7: carolinianus), yellow

indiangrass (Sorgha~strum nutans) and shortleaf skeletongrass (Gymnopogon brevifolius).

PANHANDLE SILTY WOODLANDS (22 plots; D6): This association occupied Pleistocene

and Miocene sediments of the Coastal Lowlands west of the Ochlochnee river basin (Figure 2-

2b). Most sites were located in the Apalachicola embayment region, and many occupied

Pleistocene and Holocene undifferentiated deposits of lowlands east of the Apalachicola river

(Puri and Vernon 1964, Florida Department of Environmental Protection 1998). Although









included in the Dry Uplands series, PANHANDLE SILTY WOODLANDS resembled Mesic Flatwoods

in landscape context. They inhabited side slopes and terraces of intermediate topography. Soils

of PANHANDLE SILTY WOODLANDS were high in silt and clay content. Notably, subsurface soils

had high silt and low organic content compared to other associations.

Longleaf pine dominated dense canopies of PANHANDLE SILTY WOODLANDS sites (mean

BA = 1 1.9 m2/ha). Other canopy species were infrequent (mean BA of all other species < 0.4

m2/ha). Upland oaks and other xeric midstory hardwoods were conspicuously absent. Low

growing evergreen shrub species typical of mesic habitats dominated the midstory strata,

including gallberry (Ilex glabra), running oak (Q. pumila), saw palmetto (Serenoa repens), and

dwarf live oak (Q. minima).

Although woody vegetation of PANHANDLE SILTY WOODLANDS resembled Mesic

Flatwoods, herbaceous ground cover was floristically similar to other associations in the Dry

Uplands series. Mean species richness of PANHANDLE SILTY WOODLANDS was relatively high,

comparable to NORTH FLORIDA SANDHILLS and NORTH FLORIDA RICH WOODLANDS associations.

Wiregrass (A. stricta), little bluestem (S. scoparium var. stoloniferum), narrowleaf witchgrass (D.

angustifolium), and cypress witchgrass (D. dicotomum var. tenue) were the most common grass

species. Forb species were frequent relative to grasses and shrubs. All indicator species were

herbaceous species; over half were members of Asteraceae or Fabaceae plant families. Eight of

14 indicator species have ranges restricted to the panhandle or north peninsula, including two

that are endemic to the Apalachicola region: pineland false sunflower (Phoebanthus tenuifolius)

and scareweed (Baptisia simplicifolia).









SERIES 2: Mesic Flatwoods

This series includes three associations that can be categorized as either Mesic Flatwoods

or Xeric Flatwoods according to FNAI conventions (FNAI 1990). Mesic Flatwoods associations

typically inhabited flat poorly drained regions of the panhandle and peninsula Coastal Lowlands,

and the peninsular Central Highlands (Figure 2-2d; Abrahamson and Hartnett 1990, FNAI 1990,

Myers 2000). Mesic Flatwoods appeared to be absent from the Northern Highlands landform.

Some geographic separation of Mesic Flatwoods associations was apparent, with complete

separation of NORTH FLORIDA MESIC FLATWOODS and CENTRAL FLORIDA MESIC

FLATWOODS/DRY PRAIRIES.

In general, Mesic Flatwoods soils are sandy and acidic. Mesic Flatwoods soils are

typically Spodosols, with acidic sands underlain by clayey or organic hardpans hindering water

percolation (FNAI 1990). Sub-surface clay content was consistently low in my Mesic Flatwoods

sites (although sample depths may have been to shallow to detect hardpans). I did not sample

sub-surface organic matter. Organic content of surface soils was high compared to Dry Uplands

and similar to Wetlands associations. Variation in soil texture between Mesic Flatwoods

associations was minimal, although the northern association tended to have higher surface soil

clay content. Descriptions of specific associations follow.

XERIC-MESIC FLATWOODS (36 plots; M1): XERIC-MESIC FLATWOODS generally

inhabited the Coastal Lowlands of the panhandle and peninsula (Figure 2-2d). They occupied

upper slopes of small sandy rises embedded in large expanses of Mesic Flatwoods vegetation. In

the few sites of the Central Highlands, this association occurred downslope of Dry Upland

vegetation.









Soils of XERIC-MESIC FLATWOODS were coarse and contained very small concentrations

of fine textured sediments. These differences distinguished XERIC-MESIC FLATWOODS SOilS

from NORTH FLORIDA MESIC FLATWOODS. Organic content of XERIC-MESIC FLATWOODS SOilS iS

similar to other Mesic Flatwoods associations, although soil pH is relatively high.

XERIC-MESIC FLATWOODS had sparse pine canopies and dense shrubby midstory strata.

Sparse longleaf pine formed the canopy (mean BA = 2.8 m2/ha). Slash pine (P. elliottii) was

sub-dominant (mean BA = 1.0 m2/ha). Saw palmetto (S. repens) was by far the most abundant

midstory shrub. Three upland "scrub" oaks were common in XERIC-MESIC FLATWOODS

midstories: sand live oak (Q. geminate), Chapman's oak (Q. chapmanii), and myrtle oak (Q.

myrtifolium). Notably absent were the upland oaks of Dry Upland associations. In addition,

evergreen shrub species of the heath family are common in Xeric-Mesic Flatwoods, including

fetterbush, Lyonia lucid; dwarf huckleberry, G. dumosa; lowbush blueberry y. myrsinites).

Mean species richness of XERIC-MESIC FLATWOODS sites was low due to sparse

herbaceous ground cover. Wiregrass (A. strict) was by far the most common herb, followed by

broomsedge bluestem (A. virginicus), hemlock witchgrass (D. sabulorum var. thinium) and

silkgrass (P. gramninifolium: not a grass, but a member of Asteraceae). Few indicator species

were identified for Xeric-Mesic Flatwoods. Of these, 3 (out of 7) were shrub species, including

two of the common scrub oaks and tarflower (Befaria racemosa). Three indicator species had

restricted peninsular ranges: tarflower, Chapman's goldenrod (Solidago odor var. chapmanii)

and shortleaf gayfeather (Liatris tenuifolia var. quadriflora ).

NORTH FLORIDA MESIC FLATWOODS (30 plots; M2): This association was observed in

the Coastal Lowlands land unit of the panhandle and peninsula. A few sites occurred in the

Central Highlands land unit of the peninsula (Figure 2-2d) where it occupied small areas









downslope of Dry Uplands. The northerly distribution of NORTH FLORIDA MZESIC FLATWOODS

separates it geographically from the CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES.

Typically, NORTH FLORIDA MESIC FLATWOODS occupied flat, poorly drained terrain of

Pleistocene origin. Soil clay content and pH was low compared to other Mesic Flatwoods

associations.

Overstory canopies of NORTH FLORIDA MESIC FLATWOODS were comparatively dense.

Longleaf pine was the most common canopy species (mean BA = 9.3 m2/ha), and slash pine was

sub-dominant (mean BA = 1.5 m2/ha). Midstory vegetation was generally low and sparse.

Common shrub species were gallberry (I. glabra), saw palmetto (S. repens), and runner oak (Q.

minima). These species formed patches of low growth in the understory, interspersed with thick

herbaceous ground cover. Runner oak and another woody sub-shrub (hairy wicky, Kalmia

hirsuta) were identified as indicator species.

NORTH FLORIDA MESIC FLATWOODS had intermediate species richness relative to other

Mesic Flatwoods. Common herbaceous species resembled those of XERIC-MESIC FLATWOODS:

wiregrass (A .stricta), broomsedge bluestem (A. virginicus), and silkgrass (P. gramninifolium).

Carolina yelloweyed grass (Xyris caroliniana) is the most frequent herbaceous species, as well as

an indicator species. Only four indicator species were recognized for NORTH FLORIDA MZESIC

FLATWOODS. Other than those already mentioned, these include Florida dropseed (Sporobolus

floridanus)~~dddd~~~ddd~~~ and dwarf live oak.

CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES (22 plots; M3): These sites

were restricted to the Coastal Lowlands land unit of the peninsula. CENTRAL FLORIDA MZESIC

FLATWOODS/DRY PRAIRIES typically occupied broad flat, poorly drained terrain with sediments

of Pleistocene origin (Figure 2-2d). They often surrounded XERIC-MZESIC FLATWOODS









communities present on slightly higher and drier ridges. Soil characteristics of CENTRAL

FLORIDA MESIC FLATWOODS/DRY PRAIRIES were similar to XERIC-MESIC FLAT WOODS, although

clay content was slightly higher in the sub-soil.

Canopies of CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES were either sparse or

absent. The latter condition distinguishes the dry prairies of Central Florida as described

elsewhere (FNAI 1990, Bridges 2006, Platt et al. 2006). I did not detect floristic differences

between sites with and without pine canopy. Where canopy was present, longleaf pine (P.

palustris) was dominant and the two slash pine varieties (P. elliottii var. elliotii and P. elliottii

var. densa) were infrequent (Mean BAs respectively: 2.0, 0.5, and 0.2 m2/ha). Midstory

vegetation was sparse in the frequently burned CENTRAL FLORIDA MZESIC FLATWOODS/DRY

PRAIRIES. Woody species were relegated to the understory, were saw palmetto (S. repens) was

abundant. Other understory shrubs included dwarf live oak (Q. minima), gallberry (I. glabra),

and fetterbush (L. lucid'a).

Grass species were the most common species, such as wiregrass (A. stricta), hemlock

witchgrass (D. sabulorum var. thinium), broomsedge bluestem (A. virginicus), bottlebrush

threeawn (A. spiciformis), and cypress witchgrass (D. chamnaelonche). The latter two grasses

were indicators of CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES. Two large bunchgrass

species restricted to peninsular Florida were conspicuous indicators: shortspike bluestem (A.

brachystachyus) and stoloniferous little bluestem (Schizachyrium stoloniferum). Graminoid

species comprise about a third of the indicator species (11 out of 27). Three indicator species

were shrubs and the remainder was forbs. No legumes, and only one member of Asteraceae

were identified as indicators. Five indicator species have ranges restricted to the Florida

Peninsula, including two endemics.









SERIES 3: Wetlands

Associations of the Wetlands ecological series encompassed a diverse array of floristic

variation and physiognomic settings, and spanned most of the study area. Most Wetlands

associations, particularly those of the Northern and Central Highlands land units, tended to

occupy small areas of lower slopes in regions with relatively high relief. Two associations

restricted to the Coastal Lowlands land unit (the PANHANDLE WET FLATWOODS/PRAIRIES and

PENINSULA WET FLATWOODS/PRAIRIES) inhabited poorly drained areas with little topographic

relief. Historically, large expanses of this vegetation extended across gradients of imperceptible

elevation change (Harper 1914, Myers 2000). Geographic segregation among Wetlands

associations is apparent. Three of seven associations are restricted to the panhandle and one

restricted to the central peninsula (Figure 2-2e, 2-2f)

Associations of the Wetlands series show considerable variation in physical soil

properties. Some associations were characterized by poorly drained and silty soils, where silt

content distinguished Wetlands from Mesic Flatwoods and Dry Upland series. Texture

differences were less apparent in sub-soil composition. Organic matter was high in all Wetlands

associations. Fine textured soils and surface organic matter were particularly abundant in

panhandle associations. Soil pH was variable among associations, ranging from 4.2 to 5.3.

Descriptions of specific associations follow.

MARGINAL PRAIRIES (11 plots; W1): This association included herbaceous dominated

vegetation of depressional wetland margins in the panhandle and northern peninsula (Figure 2-

2f). MARGINAL PRAIRIES occupied concentric zones surrounding shallow, seasonally inundated

depressions and were seasonally inundated during periods of high rainfall. These sites occurred









on the margins of various wetland types, including dome swamps, sandhill upland lakes, and

depression marshes (FNAI 1990).

MARGINAL PRAIRIES surface soils were sandy and acidic, and contained high levels of

organic matter. Sub-soil silt content was low but clay content was high. This may reflect the

presence of subsurface "hardpans" or "clay lens" that underlie depression marshes and dome

swamps (FNAI 1990).

Canopy vegetation of MARGINAL PRAIRIES was either absent, or was comprised of young

pond cypress and swamp tupelo (Ta-xodium a~scendens and Nyssa sylvatica var. biflora). The

latter condition may reflect "invasion" of saplings following fire suppression. Midstory

vegetation was sparse or absent. A few evergreen shrubs was sporadically present, the most

common were gallberry (I. glabra), titi (Cyrilla racemiflora) and sandweed (Hypericum

fasciculatum).

Herbaceous vegetation of MARGINAL PRAIRIES was low in aspect and diversity. Mean

species richness is the lowest of all Wetland associations (49.6 species/0. 1 ha). Few species had

high constancy across sites, reflected by the few species recognized as indicators. The most

frequent herbaceous species (> 80%) were five grasses and two weedy forbs: broomsedge

bluestem (A. virginicus), slender flattop goldenrod (Euthamia caroliniana), pale meadowbeauty

Rhexia mariana var. mariana), and maidencane (Panicum hemitomon). Wire grass (A. strict)

was absent from MARGINAL PRAIRIES. Maidencane, a species typical of inundated wetlands, was

an indicator species. The remaining five indicators included forbs and one shrub species

(buttonbush: Cephalan2thus occidentalis).

PENINSULA WET FLATWOODS/PRAIRIES (16 plots; W2): These inhabit the flat, poorly

drained regions of the Coastal Lowlands of peninsular Florida (Figure 2-2f). PENINSULA WET









FLATWOODS/PRAIRIES occupied barely perceptible lower slopes in association with NORTH

FLORIDA MESIC FLATWOODS. High water tables and seasonal inundation due to rainfall

contribute to wet conditions of PENINSULA WET FLATWOODS/PRAIRIES (FNAI 1990, Myers and

Ewel 1990). Soils were sandy and high in organic matter compared to other Wetlands

associations. Soil characteristics were particularly pronounced in sub-surface horizons, where

clay and silt content was low. Soil calcium concentration and pH was high compared to other

Wetlands associations.

PENINSULA WET FLATWOODS/PRAIRIES were variable in canopy structure and

composition. Pine canopies were either sparse or absent. Midstory vegetation was typically low

and sparse. The most common species were low growing shrubs: sandweed (Hypericum

fasciculatum), gallberry (I. glabra), and pond cypress (Taxodium a~scendens).

Floristic similarity of understory vegetation united PENINSULA WET

FLATWOODS/PRAIRIES sites as this association. All common species, and 95% of indicator

species were forbs and graminoids. Among frequent herbs, over 80% were recognized as

indicator species. Wetland herb species, such as water cowbane (Oxypolis fihformis), tenangle

pipewort (Eriocaulon decangulare), pineland rayless goldenrod (Bigelowia nudata), blue

maidencane (Amphicarpum muhlenbergianum) and Elliott' s yelloweyed grass (X. elliottii) were

diagnostic in their constancy and fidelity. Wiregrass was frequent in PENINSULA WET

FLATWOODS AND PRAIRIES, but was rivaled in abundance by other grass species, including

longleaf threeawn (A. palustris). None of the 40 indicator species have restricted ranges.

CALCAREOUS WET FLATWOODS (4 plots; W3): This association was an unusual wet

flatwoods assemblage that inhabited sites with shallow subsurface limestone. My small sample

size precludes much generalization; however, the four plots were floristically distinct (Figure 2-










2f). The sites occurred in two situations: 1) the coastal fringe of the Big Bend region of the

western peninsula, where marl is often immediately below the soil surface, and 2) as small

inclusions in Coastal Lowlands, embedded in large expanses of CENTRAL FLORIDA MZESIC

FLATWOODS.

Soil texture characteristics of CALCAREOUS WET FLATWOODS were similar to those of

PENINSULA WET FLATWOODS/PRAIRIES except that sub-surface clay content was higher.

CALCAREOUS WET FLATWOODS were basic and had exceedingly high calcium concentrations

relative to other Wetlands associations. This was consistent with the presence of shallow soils

overlying limestone outcrops.

Three of the four CALCAREOUS WET FLATWOODS sites had dense canopies of slash pine

(P. elliottii; mean BA = 11.15 m2/ha). Longleaf pine was absent. One plot, on Avon Park

Airforce Range, had no pine overstory. Sabal palms (Sabal palmetto) were present in all sites,

and comprised a significant canopy co-dominant (mean 1.88 m2/ha). Other hardwood species

were common in the midstory, including many that are typical of swamp vegetation: wax myrtle

(M~yrica cerifera), swamp bay (Persea palustris), red maple (Acer rubrum), saw palmetto (S.

repens), and sweetgum (Liquidambar~~~dddd~~~~ddd styraciflua).

Mean species richness of CALCAREOUS WET FLATWOODS was exceedingly high, rivaling

that of Dry Uplands on fine textured soils. Wiregrass (A. stricta) was infrequent. Dominant

grasses were redtop panicum (Panicum rigidulum var. rigidulum), sugarcane plumegrass

(Saccharum giganteum), cypress witchgrass (A. dichotomum var. nitidum), and switchgrass

(Panicum virgatum). Most common species and the majority of indicator species were forbs. In

addition, many graminoids were recognized as indicator species, particularly member of the









genus Rhynchospora. Sabal palm was the only non-herbaceous indicator species. Of 35

indicator species, only one (R. perplexa) has a restricted range.

NORTH FLORIDA SHRUBBY WET FLATWOODS (15 plots; W4): These sites occurred in

the Northern and Central Highlands, and Coastal Lowlands of the panhandle and northern

peninsula (Figure 2-2f). Most sites were east of the Ochlochnee River basin with the exception

of one site in the western panhandle. NORTH FLORIDA SHRUBBY WET FLATWOODS typically

occurred as small fringes along lower slopes, abutting wetland swamps. Textural composition of

NORTH FLORIDA SHRUBBY WET FLATWOODS soils did not distinguish them from other Wetlands

associations. However, surface soils were high in organic content and acidic.

NORTH FLORIDA SHRUBBY WET FLATWOODS have dense canopies of primarily slash pine

(P. elliottii var. elliottii; mean BA = 10.93 m2/ha). Longleaf and pond pine were minor

components (P. palustris and P. serotina; mean BA = 2.63 and 1.18 m2/ha respectively).

NORTH FLORIDA SHRUBBY WET FLATWOODS were floristically distinct from other

wetlands associations due shrubby species abundance in the midstory and understory strata.

Midstories were dominated by saw palmetto (S. repens), large gallberry (I. coriacea), and three

shrub species also recognized as indicators: swamp bay (P. palustris), fetterbush (L. lucid) and

coastal sweetpepperbush (Clethra alnifolia). The latter species is restricted in distribution to the

northern peninsula. Half of all indicator species were shrubs (10 out of 20).

Herbaceous cover in NORTH FLORIDA SHRUBBY WET FLATWOODS was sparse. The most

abundant herbs were purple bluestem (Andropogon glomeratus var. glaucopsis), cinnamon fern

(Osmunda cinnamnomea) and tenangle pipewort (E. decangulare). The few herbaceous

indicators are typical of wetland habitats, including hooded pitcherplant (Salrracenia minor),

bushy bluestem (A. glomeratus var. glomeratus, and A. glomeratus var. hirsutior), and sphagnum









moss (Sphagneticola spp.). Mean species richness of NORTH FLORIDA SHRUBBY WET

FLATWOODS was low, likely related to the paucity of herbaceous ground cover.

It may appear that this association represents a fire suppressed variant of wet flatwoods

formerly dominated by herbaceous vegetation. However, NORTH FLORIDA SHRUBBY WET

FLATWOODS appear compositionally distinct, and included sample sites with histories of recent

fire.

UPPER PANHANDLE WET FLATWOODS (7 plots; W5): These sites were restricted to

Miocene-aged sediments of the Northern Highlands north of the Cody Scarp in the panhandle

(Figure 2-2e). The number of samples comprising this association was small, which in part

reflects the rarity of this association. In these sites, UPPER PANHANDLE WET FLAT WOODS were

situated downslope of PANHANDLE LONGLEAF PINE CLAYHILLS or PANHANDLE XERIC

SANDHILLS. In general, this association occupied small areas of mid-slopes and flat terraces.

Fine-textured sediments distinguish soils of UPPER PANHANDLE WET FLAT WOODS. Clay

percentages of surface soils exceed that of all associations, and silt content was similarly high.

These characteristics also typify sub-surface soils. Soil organic matter of UPPER PANHANDLE

MESIC-WET FLATWOODS was low compared to other Wetlands associations.

UPPER PANHANDLE WET FLATWOODS have fairly sparse pine canopies comprised of three

pine species: longleaf (P. palustris), slash (P. elliottii) and loblolly (P. taeda) pines (mean BAs:

4.55, 3.31, and 2.22 m2/ha respectively). Pond pine was a minor component (P. serotina: mean

BA .84 m2/ha). Midstory vegetation was sparse, and shrubs were mainly relegated to the

understory in my frequently burned sites. Gallberry (I. glabra) was by far the most abundant

woody understory species, followed by running oak (Q. pumila), blue huckleberry (Gaylussacia

fr~ondosa var. nana) and Darrow' s blueberry (y. darrowii).~~ddd~~~dd~~~dd









UPPER PANHANDLE WET FLATWOODS resembled Mesic Flatwoods associations in aspect

and appearance. However, they were floristically similar to Mesic Flatwoods and Wetlands

associations. UPPER PANHANDLE MESIC-WET FLATWOODS were notable for their high species

richness, particularly of grass and forb species. Dominant grasses included some that are typical

upland species, such as little bluestem (S. scoparium var. stoloniferum), broomsedge bluestem

(A. virginicus), and Elliott's bluestem (A. gyrans var. gyrans). Other frequent grasses included

typical wetland species: toothache grass (Ctenium aromaticum), bushy bluestem (A. glomeratus

var. hirsutior), and arrowfeather threeawn (Aristida purpurascens var. virgata). Wiregrass was

absent. Two frequent grass species were also recognized as indicator species: warty panicgrass

(Panicum verrucosum) and (Dichanthelium consanguineum). Most indicator species were forbs

(19 out of 25). Four were members of the family Asteraceae, and three were legume species. A

large proportion of indicator species (32%) are species with ranges restricted to panhandle or

northern peninsula.

PANHANDLE WET FLATWOODS/PRAIRIES (16 plots; W6): This association occurred in

the Northern Highlands and Coastal Lowlands of the panhandle, west of the Ochlochnee River

basin (Figure 2-2e). Landscape and topographic context was variable; in hillier terrain

PANHANDLE WET FLATWOODS/PRAIRIES occupied narrow lower slopes and was associated with

Dry Upland sandhills and clayhills. In contrast, nearly treeless PANHANDLE WET

FLATWOODS/PRAIRIES of the Coastal Lowlands occupied large areas associated with NORTH

FLORIDA MESIC FLATWOODS. Examples of the latter are the wet prairies of the Apalachicola

National Forest (Clewell 1971). PANHANDLE WET FLATWOODS AND PRAIRIES soils were low in

sand and high in silt, in contrast to the PENINSULA WET PRAIRIES of Central Florida. Organic

content was among the lowest of Wetlands associations.









PANHANDLE WET FLATWOODS AND PRAIRIES had sparse or no canopy. A few sites had

sparse canopies of slash (P. elliottii) and longleaf (P. palustris) pines (mean BAs: 2.9 and 1.2

m2/ha respectively). Midstory vegetation was largely non-existent. Low growing gallberry (I.

glabra) is by far the most abundant understory woody species, followed by the wetland shrub,

buckwheat tree (Cliftonia monophylla).

Frequently burned PANHANDLE WET FLAT WOODS AND PRAIRIES had well-developed,

herbaceous dominated ground cover vegetation. Wiregrass (A. stricta) and toothache grass (C.

aromaticum) were ubiquitous dominant grasses; the latter was also identified as an indicator

species. Other common and indicator species included many forbs (27 out of 41 species),

including bristleleaf chaffhead (Carphephorus psuedoliatris), woolly sunbonnets (Chaptalia

tomentosa), coastalplain yelloweyed grass (Xyris amnbigua), and savanna meadowbeauty (Rhexia

alifan2us). Pinewoods bluestem (A. arctatus) was a distinctive and frequent bunch grass. Twenty

four (58%) of the indicator species have distributions restricted to panhandle or northern

peninsula.

PANHANDLE SEEPAGE SLOPES (5 plots; W7): These few sites were located in the

Northern Highlands and Coastal Lowlands of the western panhandle, situated on lower slopes

where soils were usually saturated from seepage (Figure 2-2e). This condition is thought to

result from water percolation through sandy soils underlain by impermeable clay or rock

hardpans (FNAI 1990). In my sites, PANHANDLE SEEPAGE SLOPES occurred downslope of the

drier PANHANDLE WET FLATWOODS AND PRAIRIES association, and farther downslope of Dry

Uplands.

Despite the putative existence of clay lenses in the subsoil, sub-soils of PANHANDLE

SEEPAGE SLOPES were very low in clay content. It is possible that my soil samples were not deep









enough to detect hardpans. Sub-soil silt was low compared to other panhandle Wetland

associations. Surface soils were silty, acidic and high in organic content (although small sample

size precluded statistical tests).

Sparse canopies of PANHANDLE SEEPAGE SLOPES consisted of slash (P. elliottii) and

longleaf (P. palustris) pines (mean BAs: 1.9 and 1.3 m2/ha respectively). Understory shrub

cover was relatively high and dominated by gallberry (I. glabra), evergreen bayberry (M~

caroliniensis) and large gallberry (I. coriacea). The latter two species were indicator species.

Species richness and herbaceous diversity of PANHANDLE SEEPAGE SLOPES is high.

Common grasses included several dominant bunch grasses: wiregrass (A. stricta, arrowfeather

threeawn (A. palustris), pinewoods bluestem (A. arctatus) and the wetland variant of Elliott' s

bluestem (A. gyrans var. stenophyllus). The latter three grass species were also recognized as

indicator species. Other frequent indicators of PANHANDLE SEEPAGE SLOPES included many

forbs and sedges, such as Texas tickseed (Coreopsis linifolia), largeleaf rosegentian (Sabatia

macrophylla), featherbristle beaksedge (R. oligantha), and giant whitetop (R. latifolia). A high

proportion of indicator species have ranges restricted to North and Panhandle Florida (26 out of

31 species). A few of these are locally abundant rhizomatous species: rush featherling (Pleea

tenuifolia), coastal false asphodel (Tofieldia racemosa), yellow pitcherplant (Salrracen~ia flava)

large beaksedge (R. macra), and featherbristle beaksedge (R. oligan2tha).

Discussion

Results from this study provide the first comprehensive community classification of

pyrogenic pinelands of Florida. Specifically, this classification describes variation among the

remnants of natural pineland habitat following the large reduction and fragmentation of a once

expansive pineland landscape. Conditions and landscape contexts of sites in this study were









dictated by the non-random distribution and management of natural areas in Florida, which is

related to the timing and configuration of human settlement and economic development (Kautz

and Cox 2006, Frost 2006). As such, this classification is not representative of pre-settlement

conditions of pineland variation and diversity, although it represents the best approximation

possible. Although most community associations fall within classifications previously suggested

for southeastern U. S. pinelands (FNAI 1990, Peet 2006) this system presents a much greater

refinement of recognizable species associations. The described associations are generally

defined by geographic region, physiographic landform, local topography, and soil characteristics,

providing additional guidance to their identification in the Hield.

Geographic segregation is pronounced in this floristic classification. Gradients in species

composition vary with well known environmental, climatic, and geologic gradients in Florida.

The Florida peninsula spans almost seven degrees in latitude, and excluding the Florida keys,

spans three bioclimatic life zones including the Warm Temperate Moist Forest of North Florida

and the Subtropical Moist Forest of extreme south Florida (Holdridge 1967, Myers 2000). Much

of the peninsula falls into a broad "transition zone" between the two. Variation in relative

proportions of temperate vs. tropical tree species with latitudinal gradient is a well-known

phenomenon (Wunderlin and Hansen 2000). Florida's complex recent geologic history also

underscores panhandle and peninsula differences, which may contribute to variation in

contemporary vegetation. Carbonate deposits of marine origin created the limestone platform of

the peninsula between 60 to 120 MYBP. In contrast, Miocene deposits of the panhandle were

mainly plastic sediments derived from Appalachian erosion and alluvial processes (Randazzo

and Jones 1997, Myers 2000). Until 12 to 30 MYBP the Suwannee Strait, an elongate negative

structure in southern Georgia and northeastern Florida, separated the two regions (Hull 1962,









Puri and Vernon 1964, Myers 2000). From the late Miocene to recent time, increased plastic

deposition, and a series of sea level fluctuations have influenced surface geology and soil

development, particularly in the peninsula (Randazzo and Jones 1997, Myers 2000). The distinct

panhandle-peninsula trend in floristic variation is correlated with many of these phenomenon,

including differences in geologic sedimentation, age of landforms, degree of isolation, and

climate variation associated with latitude.

To varying degrees, geographic patterns are similar within the three ecological series

(Dry Uplands, Mesic Flatwoods, and Wetlands). Eastern and western analogues of similar

edaphic and moisture conditions are nonetheless floristically different (e.g. the xeric sandhills of

peninsula vs. panhandle). The Dry Uplands associations display distinct separation east and west

of the Ochlochnee River basin. Although the separations are not as pronounced, similar east-

west divisions exist among the Wetlands associations. PANHANDLE WET FLATWOODS/PRAIRIES

and PANHANDLE SEEPAGE SLOPES of the western panhandle are floristically distinct from other

Wetlands associations. None of the Mesic Flatwoods associations is unique to the western

panhandle, although there is a north-south separation between NORTH FLORIDA MZESIC

FLATWOODS and CENTRAL FLORIDA MZESIC FLATWOODS/DRY PRAIRIES.

Geographic segregation of floristic groups is influenced by the prominence of plant taxa

with distributions restricted to a particular region of the State. Nearly a fifth of all taxa (18.4%

of 575) included in this analysis have restricted ranges, while a far smaller proportion (2.8 %) are

endemic to Florida. Most restricted range taxa reflect the familiar segregation between

panhandle and peninsula Florida, or the distinction between north (panhandle plus northern

peninsula) Florida and the central peninsula. Many northerly distributed species reach their

southern range limits in the north peninsula; their southern distributional limits closely










approximating the "warm temperate moist forest" bioclimate zone (Holdridge 1967, Myers

2000). Fewer species have Florida distributions either restricted to the panhandle west of the

Ochlochnee River basin, or the central Florida peninsula. The large number of taxa restricted to

the western panhandle (42 taxa) is consistent with other descriptions of endemism in regions

within the East Gulf Coastal Plain (125 endemic taxa reported by Sorrie and Weakley 2001,

Sorrie and Weakley 2006). Similarly, a large number of endemics have previously been

identified in the Florida peninsula (122 taxa: Sorrie and Weakley 2001, Sorrie and Weakley

2006), where I recorded 31 restricted range taxa. Because I omitted Lake Wales Ridge from the

sample region endemic species were likely under-represented in the associations of the Florida

peninsula.

Restricted range taxa were frequently selected as indicator species of associations. Not

surprisingly, indicator species of panhandle Dry Uplands and Wetlands associations included

many restricted-range species, and many of these are endemic to the East Gulf Coastal Plain as

reported by Sorrie and Weakley (2001). Taxa restricted to the Florida peninsula were prominent

indicators of the XERIC-MESIC FLATWOODS and CENTRAL FLORIDA MZESIC

FLATWOODS/PRAIRIES. No restricted range indicator species were selected for the single

Wetland association of restricted to the peninsula (PENINSULA WET FLATWOODS/PRAIRIES). In

contrast, the three Wetland associations restricted to the western panhandle were characterized

by numerous indicator species with restricted ranges (UPPER PANHANDLE WET FLATWOODS,

PANHANDLE WET FLATWOODS/PRAIRIES, and PANHANDLE SEEPAGE SLOPES).

Comparisons to Other Classifications

This classification of fire-adapted pinelands and associated communities resembles

community descriptions of the Florida Natural Areas Inventory (FNAI), both floristically and in









its description of landscape and edaphic conditions. However, my classification augments the

FNAI community descriptions in recognition of geographically related floristic variation. The

FNAI description of "Sandhill" corresponds to my three Dry Upland associations distinguished

by region and edaphic/moisture conditions (PANHANDLE XERIC SANDHILLS vs. PENINSULA XERIC

SANDHILLS and NORTH FLORIDA SANDHILLS). Likewise, three current associations (PANHANDLE

LONGLEAF PINE CLAYHILLS, PANHANDLE SILTY WOODLANDS and NORTH FLORIDA RICH

WOODLANDS) resemble the FNAI "Upland Pine Forest" community. FNAI reports that "Upland

Pine Forests" are restricted to the Miocene-aged rolling hills of extreme northern Florida (FNAI

1990). Two of the three aforementioned associations occur outside of the FNAI geographic and

landscape description, thus they lack FNAI analogues. My Mesic Flatwoods associations are

similar to FNAI community descriptions. However, the two Mesic Flatwoods associations were

segregated by region (NORTH FLORIDA MZESIC FLATWOODS VS. CENTRAL FLORIDA MESIC

FLAT WOODS AND DRY PRAIRIES), whereas the FNAI distinguishes mesic flatwoods by canopy

conditions ("Mesic Flatwoods" vs. "Dry Prairie"). Cross reference between this classification

and FNAI types is less clear for Wetlands associations. I recognized four Wetlands associations

that overlap (in whole or part) with the FNAI descriptions for "Wet Flatwoods", "Wet Prairie"

and "Marl Prairie" (the latter perhaps corresponds to CALCAREOUS WET FLATWOODS). The FNAI

description of "Bogs" may partially overlap with NORTH FLORIDA SHRUBBY WET FLATWOODS.

The PENINSULA WET PRAIRIES resemble FNAI' s "Wet Prairie", although mine is regionally

defined. The MARGINAL PRAIRIES association describes herbaceous vegetation associated with

two FNAI lacustrine communities: "Flatwoods/Prairie/Marsh Lake" and "Sandhill Upland

Lake". The FNAI "Seepage Slopes" closely resembles my PANHANDLE SEEPAGE SLOPES

association.










My classification and study area represents a subset of the regional treatment of Peet

(2007). The Peet classification includes a greater breadth of environmental conditions as well as

a larger geographic region. A comparison of the current community associations to associations

reported by Peet as present in Florida reveals differences between the two classifications with

regard to classification resolution. Geographic segregation is a primary trend in both treatments.

My Dry Uplands associations variously correspond to Peet' s "Xeric Sand Barrens/Uplands" and

"Subxeric Sandy Uplands" groups, which include three and six associations respectively (see

Table 2, Peet 2006). My PANHANDLE LONGLEAF PINE CLAYHILLS corresponds to two of Peet' s

"Silty Uplands" associations (types 3.4. 1 and 3.4.2, Peet 2007). The Peet analogues to NORTH

FLORIDA RICH WOODLANDS and PANHANDLE FLATWOODS/WOODLANDS are less obvious,

perhaps corresponding to other "Siltly Uplands" types. My three Mesic Flatwoods associations

have 14 counterparts in the Peet treatment. Likewise, my Wetlands associations correspond to

nine associations of Peet' s "Savannas and Seeps".

The subj ective nature of defining partitions likely contributes to differences between the

classification typologies. In this study I attempted to minimize subjectivity in sampling design

and numerical analysis. The distribution and management of Florida pineland natural areas is

non-random and largely influenced by natural area availability (Kautz and Cox 2006, Frost

2006). However, the sampling design stratified by ecoregion and moisture gradient, coupled

with large sample size, minimized bias associated with subj ective sample selection (Leps and

Smilauer 2007). The resulting classification describes variation in the contemporary

configuration Florida pinelands in a highly fragmented landscape. This work does not explicitly

describe natural variation of pre-settlement conditions. Furthermore, by using an optimization

index in conjunction with cluster analysis, I minimized subj activity associated with cluster









delineations (McCune and Grace 2002). My selection of cut-level in the cluster solution to

groups > 3 samples was subjective. However, this limited proliferation of associations, which is

highly dependent on sample size (Legendre and Legendre 1998, McCune and Grace 2002).

In conclusion, I developed a classification of fire-dependent pineland communities that is

as comprehensive as possible while remaining applicable for management and conservation

programs. The provision of indicator species, geographical distributions, and topographic

contexts of associations, in addition to their full species lists, will enable identification of

associations in the field. This classification system will assist in refining classifications of

associations in the greater than 9.6 million acres of non-submerged conservation lands owned

and/or managed by local, state, and federal agencies in Florida (over 25% of total land area;

FNAI 2007). Descriptions of associations and indicator species will also guide ecological

restoration efforts, by assisting the recognition of natural areas degraded from fire suppression or

other reasons. Quantitative descriptions based on existing high quality natural areas provide

templates for restoration goals and comparisons. Further land acquisitions via the Florida

Forever Program and other conservation efforts might benefit from this classification. The

present study presents a descriptive vegetation classification based on a comprehensive,

systematic, and quantitative inventory of fire-dependent pineland communities as they exist

today.



















Variable Mean SE Mean SE Mean SE
% sand A 93.13 b 0.67 96.18 0 0.84 88.49 a 0.97
% sand B 88.09 a 1.08 95.47 b 1.35 88.40 a 1.50
% silt A 4.29 b 0.49 2.52 a 0.61 8.39 C 0.71
% silt B 7.54 b 0.70 2.89 a 0.88 7.12 b 0.97
% clay A 2.58 b 0.36 1.29 a 0.45 3.11 b 0.52
% clay B 4.37 b 0.55 1.64 a 0.69 4.47 b 0.77
% org 2.86 a 0.35 4.82 b 0.45 5.69 b 0.51
pH 4.75 b 0.03 4.53 a 0.04 4.52 a 0.05


Table 2-1. Means and standard errors (SE) of soil and site variables by community series. Plot
number indicated in parentheses. Variables are labeled as in text. Significantly
different pairwise comparisons of means (p < 0.01) are denoted by dissimilar
superscripts and different shading.


Dry Uplands
(1 30)


Mesic Flatwoods


Wetlands
(64)









































Xeric-Mesic Flatw~oods 59.0 a 3.2 5.0 a 0.85 96.9 b 0.40 1.99 a 0.43 1.12 a 0.19 96.0 b 0.40 2.55 a 0.39 1.43 a 0.21 4.64 a 0.83 4.64 b 0.07 458.17 b 39.68

North Flosda Mesic 71.4 b 3.5 10.7 b 0.94 95.0 a 0.40 3.29 a 0.44 1.68 b 0.20 94.5 a 0.42 3.60 a 0.40 1.86 b 0.22 5.14 a 0.86 4.39 a 0.07 247.83 a 41.02

Peninsula Meslrares 72.9 b 4.1 2.8 a 1.09 96.7 ab 0.50 2.27 a 0.52 1.00 a 0.23 95.9 b 0.49 2.41 a 0.47 1.65 ab 0.26 4.65 a 1.01 4.55 ab 0.08 413.56 b 48.21


Table 2-2. Untransformed means and standard errors (SE) of soil and site variables by community association. Dissimilar

superscripts and shading indicate means that are significantly different (p < 0.01). Soil texture variables listed

separately for surface (A horizon) and sub-surface (B horizon). Rich = species number / 1000 m2; BA Basal
area m2/ha.


SURFACE SOILS (A)

SE
%sand sand % silt

96.9 b 1.58 1.68 a

94.9 b 1.35 3.5 ab

94.7 b 2.34 3.63 b

95.6 b 1.33 2.70 a

86.8 a 1.58 7.40 b

ES I 9 Id


SUBSOILS (B)


Association
Pninsusa Xeric

North Florida Sandhills
North Flonda Rich

Iahadle Xeric

PanhaaddesSilty

Panhandle Lngleaf


rich SE rich BA SE BA

68.5 a 3.4 8.1 a 1.01


SE silt % clay SE clay% sand sand
1.07 1.38 a 1.04 |96.97 a 2.76


% silt SE silt %clay SE cla l% org SE org pH SE pH Ca SE Ca
1.66 a 1.73 1.36 a 1.52 I4.01 b 0.56 4.62 a 0.07 570.57 b 65.58


1.53 a 0.89

1.68 a 1.54

1.67 a 0.88

5.8 b 1.04

a 9 I 10


95.16 a 2.36

90.25 ab 4.09

89.98 b 2.45

76.67 0 2.76

; I Sri 5 AG


2.96 a 0.48

6 an 0.83

1.18 a 0.47

2.02 a 0.56

:3 32 0 TO


4.76 a 0.07 469.17 b 56.16

4.58 a 0.11 307.50 b 97.27

4.69 a 0.04 152.01 a 55.25

4.78 a 0.06 174.38 a 65.58


96.2 b

106.5 b

75.1 a

92.1 b

124 5


11.4 ab

16.1 cd

9.5 a

12.8 bc


Marginal Pralrles
Calcareoss Wet


Fantin s/Wr ines
NorthFFlonda Shrubby


Ip er Panhandle Wet

Pnth Pdla les

Panhandle Seepage


89.2 ab 2.94 7.12 b

96.6 NA 0.62 2.47 NA

95.6 b 2.69 3.68 ab

87.8 a 2.81 8.43 b

80.9 a 3.52 11.05 b

84.5 a 2.40 12.76 bc


3.61 b 1.24

0.88 NA 0.24

0.70 a 1.13

3.73 b 1.18

3 03?. 1.48

2.77 ab 1.01


89.62 bc 3.61

93.13 NA 5.35

94.63 0 3.30

91.12 bc 3.45

75.50 a 4.32

82.73 ab 2.95


7.18 a 1.55 4.37 ab 0.05 212.33 a 47.32

5.97 NA 2.54 5 25 0.40 1133 25 ""563.53

6.17 a 1.42 4.64 b 0.11 424.63 b 88.86

6.88 a 1.48 4.22 a 0.12 202.11 a 51.13

4.23 a 1.85 4.62 b 0.09 183.25 a 23.67

3.46 a 1.27 4.53 b 0.07 124.73 a 25.93


49.6 a

1 11

69.7 b

65.7 ab



78.3 b


5.6 a 2.32

17.1 NA 7.50

2.4 a 1.92

15.7 bc 1.98

10.9 as 2.90

4.5 a 1.98

4.1 NA 2.20


3.62 a 2.51 6.75 0 1.96

3.07 NA 3.69 3.79 NA 2.93

3.77 a 2.28 1.59 a 1.79

6.11 ab 2.39 2.76 ab 1.87

13.52 0 3.00 10.97 0 2.34

12.07 bc 2.05 5.18 bc 1.60


86.5 NA 5.50 12.35 "^ 5.29 1.2 NA 0.36192.00 NA 6.18 6.56 NA 4.26 1.44 NA 3.3819.96 NA 5.92 4.28 NA 0.04 245.25 NA 59.50


90.8 NA 5.1


3.28 a 1.48 1.54 a 1.30

3.76 ab 2.57 5.97 bc 2.25

6.75 b 1.53 3.26 ab 1.35

17.00 0 1.73 6.32 c 1.52

IS5 32 2 I 13,13dI I 90































North Florida Sandhills (D2)
Bluej ack oak Quercus incana 6.72
Turkey oak Quercus laevis 6.18
Sand post oak Quercus nzargaretta 4.8 1

North Florida Rich Woodlands (DI)
Saw palmetto Serenoa repens 11.03
Winged sumac Rhus copallinunt 5.03
Mockernut hickory Carya alba 4.23
Dwarf waxmyrtle M~yrica cerifera var. puntila 3.01
Sand live oak Q~uercus gentinata 2.65
Sand post oak Quercus nzargaretta 2.43

Panhandle Longleaf Pine Clayhills (DS)
Sand post oak Quercus margaretta 5.71
Turkey oak Quercus laevis 4.89
Running oak Quercus puntila 4.46
Bluej ack oak Q~uercus incana 3.51
Darrow's blueberry Vacciniunt darrowiiddd~~~~ddd~~~ddd 3.13
Southern red oak Quercus falcata 2.97
Dwarf huckl eb erry Gayhtssacia duntosa 2.60
Winged sumac Rhus copallinunt 2.08


Table 2-3. Common woody shrubs of midstory and understory strata, listed by association.
Code in parentheses correspond to those in Figure 2-2. Mean cover in m2/100 m.
Peninsula Xeric Sandhills (D3)
Common Name Scientific Name Mean cover
Turkey oak Quercus laevis 14.20
Sand live oak Quercus gentinata 4.27
Saw palmetto Serenoa repens 1.91
Bluej ack oak Quercus incana 1.71


Panhandle Xeric Sandhills (D4)
Turkey oak Quercus laevis
Dwarf live oak Quercus nainita
Sand live oak Quercus gentinata
Saw palmetto Serenoa repens
Sand post oak Quercus nzargaretta
Dwarf huckl eb erry Gayhtssacia duntosa
Bluej ack oak Quercus incana


12.08
3.71
3.10
3.00
2.75
2.40
2.11









Table 2-3 continued.
Panhandle Silty Woodlands (D6)
Gallberry Ilex glabra 9.01
Running oak Quercus puntila 7.45
Saw palmetto Serenoa repens 6.19
Dwarf live oak Quercus naininta 6.04
Dwarf huckl eb erry Gayhtssacia duntosa 3.27
Shiny blueberry Vacciniunt myrsinites 2.59
Darrow's blueberry Vacciniunt darrowiiddd~~~~ddd~~~ddd 2.48
Blue huckleberry Ga~yhtssacia f~ondosa var. nana 1.93
Xeric-Mesic Flatwoods (M1)
Saw palmetto Serenoa repens 26.83
Sand live oak Quercus gentinata 9.92
Chapman's oak Quercus chapnzanii 7.90
Myrtle oak Quercus nzyrtifolia 5.35
Shiny blueberry Vacciniunt myrsinites 4.18
Fetterbush Lyonia hucid'a 3.73
Dwarf live oak Quercus naininta 3.68
North Florida Mesic Flatwoods (M2)
Gallberry Ilex glabra 19.49
Saw palmetto Serenoa repens 17.01
Dwarf live oak Quercus naininta 10.46
Shiny blueberry Vacciniunt myrsinites 6.56
Running oak Quercus puntila 5.93
Dwarf huckl eb erry Gayhtssacia duntosa 2.64
Fetterbush Lyonia hucid'a 2.45
Central Florida Mesic Flatwoods/Dry Prairies (M3)
Saw palmetto Serenoa repens 23.10
Dwarf live oak Quercus naininta 7.40
Gallberry Ilex glabra 6.72
Fetterbush Lyonia hucid'a 5.26
Shiny blueberry Vacciniunt myrsinites 3.84
Dwarf huckl eb erry Gayhtssacia duntosa 2.15










Table 2-3 continued.
Marginal Prairies (W1)
Gallberry
Titi
Sandweed
Swamp tupelo


Ilex glabra
Cyrilla racentiflora
Hypericunt fasciculatunt
Nyssa sylvatica var. biflora


3.55
3.24
3.09
1.94


4.3 0
3.48
2.33


5.38
2.75
2.28
2.22
2.06


6.41
5.73
4.43

3.27
3.15


19.27
4.52
1.71
1.61


8.39
3.24


13.00
4.55
1.78


Peninsula Wet Flatwoods/Prairies (W2)
S andweed Hypericunt fasciculatunt
Gallberry Hlex glabra
Pond cypress Taxodium a~scendens

Calcareous Wet Flatwoods (W3)
Wax myrtle M~yrica cerifera
Swamp bay Persea pahtstris
Saw palmetto Serenoa repens
Red maple Acer rubrunt
Sweetgum Liquidanabar~dddd~~~ddd~~~ styraci lua

North Florida Shrubby Wet Flatwoods (W4)
Saw palmetto Serenoa repens
Large gallberry Hlex coriacea
Fetterbush Lyonia hucida
Coastal
sweetpepperbush Clethra alnifolia
Sweetbay magnolia Magnolia virginiana

Upper Panhandle Wet Flatwoods (W5)
Gallberry Hlex glabra
Running oak Quercus punila
Blue huckleberry Ga~yhtssacia Jrondosa var. nana
Darrow's blueberry Vacciniunt darrowii d~~ddd~~~dd~~~dd

Panhandle Wet Flatwoods/Prairies (W6)
Gallberry Hlex glabra
Buckwheat tree Cliftonia nzonophylla

Panhandle Seepage Slopes (W7)
Gallberry Hlex glabra
Evergreen bayberry M~yrica caroliniensis
Large gallberry Hlex coriacea











Table 2-4. Indicator species of Dry Uplands and Mesic Flatwoods associations listed in
descending order of Indicator Value (IV). Superscripts indicate species with
restricted distributions in Florida: Iwestern Panhandle, 2nOrth Florida, central Florida
peninsula, 4Florida endemic.
Peninsula Xeric Sandhills (D3) North Florida Sandhills con't (D2)
Species IV Species IV
Bulbostylis warei 41.8 Heliantheinto carolinianton 31.9
Balduina (I,-.-;,I r, r;i. -le 40.3 Rhynchosia reniforinis 31.6
4ristida condensata 35.0 Physalis walteri 28.9
Lechea s.. l Il. ,, 33.4 Scutellaria ;,,ir.in,,;, thdi. s,1 28.7


Piriqueta cistoides ssp. caroliniana
Dvschoriste oblongifolia
4sclepias verticillata
Gvinnopogon ambiguus
Eupatoriuin glaucescens
Lespedeza hirta
Ruellia caroliniensis ssp. cibiosa
Croton (I, -I, I,,rnit. n, a r
Tragia urens

North Florida Rich Woodlands (D1)
Species
Erythrina herbacea
Dichanthelitan Je-;.. ,li,, var. Je-;
Eustachys floridana


33.2
31.0
29.0
26.8
23.3
23.2
22.9
22.8
22.6
21.3
20.1



IV
74.7
67.8
35.5
34.6
34.2
32.1
28.3
26.8
26.6
23.8
23.5
22.9
22.0
20.6



IV
38.4
33.8


28.6
28.4
26.9
26.9
25.0
25.0
23.8
23.3
23.3



IV
54.6
46.6
44.4
43.2
41.6
41.0
37.0
34.7
34.7
33.6
29.6
26.0
24.1
22.5
22.1
22.1
21.5
20.4
20.0


4simina incana 3
Triplasis americana
Opuntia htonifusa
Callisia graininea

Carphephorus corvinbosus '
Sporobolus junceus
Cnidoscolus stimulosus
Ouercus laevis
Galactia regulars
Tephrosia chrysophylla
Sisyrinchilan xerophylluin


Panhandle Xeric Sandhills (D4)

Species
Galactia inicrophylla
Euphorbia floridana
Liatris pauciflora var. secunda'a
Cyperus lupulinus ssp. lupulinus
Rhvnchosia cvtisoides *
Pitvopsis aspera '
Eriogonton toinentoston
Conunelina erecta
4ristida inohril 2
Stylisina patens ssp. patens
Liatris chapinanii
Tephrosia inohril '
Rhynchospora gravi
Bulbostylis ciliatifolia


North Florida Sandhills (D2)

Species
Desinodiuon loridanton
Palafoxia intearifolia


.. ,ank..x -


Galiton hispiduluon
Lactuca floridana
Cyperus plukenetil 2
Rhvnchosia cinerea 3'
4ristida lanosa
Tridens carolinianus
Sporobolus clandestinus
,i;,. ,tea,,l aroinatica 2
Galiton pilosuon
Vitis aestivalis
Centroseina arenicola
Clitoria inariana
Habenaria quinqueseta
4ristolochia serpentaria
Dichanthelitan conunutation var. ashei
Desinoditon glabellin 2













Panhandle Longleaf Pine Clayhills con't (D5)
Species IV
Gvmnopogon brevifolius 24.6
Eupatorium album 24.3
Dichanthelium sphaerocarpon 24.2
Stylodon carneus 24.0
Tridens carolinianus 24.0


Panhandle Silty Woodlands (D6)
Species IV
Symphvotrichum ad'natum 40.2
Baptisia simplicifolia '. 38.5
angelica d'entata 37.6
CI, la p -7'" mariana 32.1
Tragia smallii 30.7
Phoebanthus tenuifolius '. 29.6
Viola septemloba 29.4
Euphorbia curtis#i 28.4
Galactia erecta 27.6

4galinis divaricata 26.2
Helianthus radula 25.6
Dalea carnea var. gracilis 22.5
Crotalaria purshii 21.9
Sevmneria cassioides 20.4


Xeric-Mesic Flativoods (M1)
Species IV
Quercus chapmanii 38.1
Solidago od'ora var. chapmanii 3 32.9
Quercus myrtifolia 26.1
Galactia elliottil 25.6
Liatris tenuifolia var. quadriflora 3 22.1
Befaria racemosa 3 19.5
Rhvnchospora megalocarpa 17.1

North Florida Mesic Flativoods (M2)

Species IV
Alris caroliniana 23.7

Sporobolus floridanus 19.0
Ouercus minima 18.4
Kalmia hirsuta 17.5


Table 2-4 continued.


Panhandle Longleaf Pine Clayhills (D5)
Species
Rudbeckia hirta
4calypha gracilens
M~alus ow .-;,r s r;i.-le '
Vaccinium stamineum var. stamineum
Galactia volubilis
Quercus falcata
Carva alba
Ceanothus americanus
Desmodium ciliare
Desmodium lineatum
Toxicodendron pubescens 2
Desmodium viridiflorum
Prunus serotmna
Phlox floridana 2
Lesped'eza repens 2
Rhynchosia tomentosa 2
Euphorbia discoidalis
F, Ia-. .-st, spectabilis
Cornus florida
Strophostyles umbellata
Smilax smallii
Desmodium strictum
Gaura filipes
Sorghastrum nutans
Galium pilosum
ambrosia artemisiifolia
Eupatorium hyssopifolium 2
Clitoria mariana
Lobelia puberula
Vernonia cl,,-.-;, st, r;i. -le
4ristolochia serpentaria
4nd'ropogon gerardia '
Solidago od'ora var. od'ora
L~echea minor

,i;,. lr,, Ga aromatica 2
M~uhlenbergia capillaris var. trichopodes 2
Tephrosia spicata
Salvia azurea
Rubus cuneifolius


IV
76.0
63.3
55.9
55.4
54.4
48.2
47.1
42.7
42.7
42.7
42.7
39.7
39.0
38.2
37.0
36.3
35.0
34.1
32.2
3 1.1
30.7
30.6
30.5
29.5
29.1
28.6
28.6
27.4
27.3
27.3
26.9
26.7
26.5
26.1
26.0
25.5
25.5
25.4
24.8











Table 2-4 continued.


Central Florida Mesic Flatwoods/Dry Prairies (M3)
Species IV
Hypericuin reductuin 62.3
Polygala setacea 57.4
Eleocharis balchrinii 50.4
Rhexia nuttallii 43.0
Fiinbristylis puberula 41.5
4ristida spiciforinis 38.2
4simina reticulata 3' 37.8
Rhynchospora fernaldii 37.5
Xyris flabelliforinis 36.4
Lyonia fruticosa 34.9
Lechea torrevi 34.3
Lachnocaulon bevrichianuin 3 34.0
Dichanthelium chainaelonche 33.3
pI e; ''m"inur flavidulus 32.7
Xyris brevifolia 32.6
Polygala rugelii 3' 32.2
4sclepias pedicellata 31.9
4ristida purpurascens var. tenuispica 30.7
Oldenlandia uniflora 30.7
Ch'inopogon chapinanianus 3 29.0
Gratiola hispida 25.8
Drosera brevifolia 25.4
Schizachyrium stoloniferuin 3 24.5
Dichanthelium sabuloruin var. thiniuin 23.8
4nd'ropogon brachystachyus 3 22.6
Hypericuin tetrapetaluin 22.6
Lveodesinia achylla 21.7











Table 2-5. Indicator species of Wetlands associations. Indicator values (IV) listed in descending


order for each association.
Marginal Prairies (W1)
Species
Eupatoriuin leptophylhon
Panicuin heinitoinon
Ludl'igia suffr~uticosa
Cephalanthus occidentalis
Rhexia inariana var. mariana
Xvris dilient(,~,i\ var. curtiss# -


North Florida Shrubby Wet Flatwoods (W4)
Species
Per sea palustris
Osinunda cinnainoiea
4nd'ropogon gloineratus var. hirsutior
Nyssa biflora
Vacciniuin virgatumn
4nd'ropogon glaucopsis
4nd'ropogon gloineratus var. gloineratus
Viburnuin nuduin
Photinia pyrifolia


Rhexia virginica -
Rhexia petiolata
liex coriacea
Gordonia lasianthus
Sarracenia ininor
Vacciniton fuscation
Rhvnchospora fascicularis
Lyonia lucida
Care glaucescens
Clethra ahtifolia -


Superscripts same as Table 2-4.
Peninsula Wet Flatwoods/Prairies (con't)
IV Species
49.3 4nd'ropogon capillipes (wetland variant)
41.1 Coelorachis rugosa
33.5 4mphicarpuin iueblenbergianuin
30.7 Sabatia grandithirku c
28.8 4nd'ropogon gyrans var. stenophyllus
28.8 Fuirena scirpoidea
Hyptis alata
Hypericuin myrtifolium
IV Pluchea rosea
45.8 F, Is ; s elliotti
44.3 Scleria balat'inii
41.4 Bigelowia nud'ata
37.0 Rhvnchospora tracvi
36.1 Viola lanceolata
35.7 Eriocaulon decangulare
34.8 Centella erecta
32.3 Lobelia glandulosa
32.2 Rhvnchospora fihifolia
28.6 Eriocaulon compressuin
28.1 Pluchea foetida
26.7 Scleria inueblenbergii
26.4 Ludl'igia linearis
25.6 Eupatoriuin iohril
24.9 Scleria georgiana
24.3 Xyris dlirrclenti.\ var. floridana
24.0 Rhexia mariana var. exalbida
22.7 Heleniton pinnatifidumn
22.2 Iva microcephala
22.0 Schizachvrizon rhizoinatuon
Panician rigidulton var. pubescens


IV
43.9
43.7
43.1
42.2
40.0
40.0
38.4
37.7
36.7
36.6
36.4
36.2
33.4
32.1
31.4
30.1
29.7
28.0
26.9
25.8
25.1
25.0
24.6
24.1
23.4
22.8
22.1
21.3
21.2
21.1


Peninsula Wet Flatwoods/Prairies (W2)
Species IV
Oxypolis fihiforinis 81.1
Dichanthelitan erectifoliton 71.4
Proserpinaca pectinata 57.9
Gratiola rainosa 56.5
Coreopsis floridana 52.0
Ludl'igia linifolia 50.6
Xvris elliottii 48.7
Panictun tenerton 47.4
Hyperician fasciculation 44.8
4ristida palustris 44.6


Calcareous Wet Flatwoods (W3)
Species IV
4sclepias lanceolata 100.0
Panician rigidulton var. rigiduham 83.6
Heleniton pinnatifidian 77.9
Phyla nodiflora 70.7
Cirsiton nuttallii 67.9
Rhvnchospora colorata 65.6
Ludl'igia microcarpa 64.3
Sabal pabnetto 60.3
Dichanthelican dichotonnon var. nitidian 58.3











Table 2-5 continued.
Calcareous Wet Flatwoods (con't)
Species
Xyris jupicai
Cyperus polystachyos
Rhynchospora divergens
Hyptis alata
Erechtites hieraciifolia
Rhynchospora perplexa 2
Ludwigia curtissii
Saccharum giganteum
Parthenocissus quinquefolia
Cyperus haspan
Rhynchospora globularis
And'ropogon glomeratus var. pumilus
Rhynchospora microcarpa
Setaria parviflora
Mecardonia acuminata
Juncus roemerianus
Sacciolepis indica
Cladium mariscus ssp. jamaicense
Osmund'a regalis
Proserpinaca pectinata
Bidens mitis

Hypericum cistifolium
Lobelia glandulosa
Muhlenbergia capillaris var. capillaris
Mitreola sessilifolia
Iva microcephala


Upper Panhandle Wet Flatwoods (W5)
Species
Hypericum setosum
Pycnanthemum flexuosum '
Rhododend'ron canescenS 2
Dichanthelium consanguineum
Desmodium tenuifolium
Rhynchospora debilis 2
Hibiscus aculeatus 2

Elephantopus nud'atus 2
Solidago stricta
Agalinis georgiana
Lespedeza capitata'


Upper Panhandle Wet Flatwoods (con't)
Species
Diodia virginiana
Polygala nana
Helianthus to;,-,Ir ; s,17lat

Eupatorium leucolepiS 2
Andropogon glomeratus var. hirsutior
Linum medium
Tephrosia spicata
Gratiola pilosa
Crotalaria purshii
Panicum verrucosum
Rhexia mariana var. exalbida
Eupatorium rotundifolium
Gymnopogon brevifolius
Scutellaria integrifolia

Panhandle Wet Flatwoods/Prairies (W6)
Species
Coreopsis linifolia 2
Carphephorus pseudoliatris *
Helianthus heterophyllus'
Drosera filiformis *
Eurybia chapmanii 2
Lophiola aurea
Andropogon arctatus
Scleria pauciflora var. carohiniana
Xyris baldwiniana 2
Rhexia lutea 2
Ilex myrtifolia 2
Polygala cruciata
Smilax laurifolia
Eryngium integrifolium 2
Ple tnI OliG 1
Chaptalia tomentosa
Xyris ambigua
Sarracenia lava 2
Eupatorium leucolepiS 2
Pityopsis i,goastri,, '
Asclepias connivens
Gaylussacia mosieri 2
Rhynchospora latifolia


IV
29.8
29.2
27.9
27.2
23.8
23.6
23.5
23.4
23.3
23.1
23.1
22.1
21.7
21.5



IV
72.3
70.8
61
50
50
50
46.7
45.5
45.5
42.6
42.3
41.7
40
37.5
37.5
37.2
35.3
34.8
34.7
34.1
33.4
32.2
31.6











Table 2-5 continued
Panhandle Wet Flatwoods/Prairies (con't)
Species IV
Rhynchospora baldwinii 3 1.2
Tofieldia racemosa 2 31.2
Cliftonia monophylla 2 31.1
Morella caroliniensis 31.1
Dichanthelium leucothrix 31.0
Rhexia alifanus 30.7
Lobelia brevifolia 30.4
Anthaenantia rufa 2 30.2
Eurybia eryngiifolia 30.2
Erigeron vernus 30.1
Nyssa ursina 1,4 30.0
Rhynchospora chapmanii 29.7
Oxypolis ternata 28.2
Aletris sp. 27.5
And'ropogon mohrii 27.0
Ctenium aromaticum 27.0
Aristida simpliciflora 2 26.5
Sarracenia psittacina 2 25.8
Liatris spicata 25.2
Verbesina chapmanii 1,4 25.0

Panhandle Seepage Slopes (W7)
Species IV
Sabatia macrophylla 2 88.9
Rhynchospora iagrnia,, 2 81.9
Arnoglossum ovatum 76.8
Juncus trigonocarpus 68.9
Pleea tenuifolia 54.5
Rhynchospora macra 2 54.3
Symphyotrichum lateriflorum var. lateriflorum 54.3
Rhynchospora latifolia 50.0
Sarracenia leucophylla 49.4
Lophiola aurea 49.2
Xyris dilinttlli.\ var. dilientc~~i.1, 2 46.2
Panicum rigidulum var. combsii 45.5
Aristida palustris 44.2
Coreopsis linifolia 2 42.7
Eryngium integrifolium 2 39.9
And'ropogon arctatus 37.0
Sarracenia psittacina 2 36.8


Panhandle Seepage Slopes (con't)
Species IV
Morella caroliniensis 36.3
Hypericum brachyphyllum 34.1
Tofieldia racemosa 2 34.1
Rhexia lutea 2 33.5
Zigadenus glaberrimus' 33.4
Dichanthelium longiligulatum 33.1
Drosera filiformis 32.7
Oxypolis filiformis 32.0
Andropogon gyrans var. stenophyllus 31.3
Eleocharis tuberculosa 31.2
Fuirena squarrosa 2 31.2
Magnolia virginiana 31.1
Sarracenia lava 2 27.9
Balduina uniflora 2 27.5
Rhynchospora corniculata 27.5
Anthaenantia rufa 2 24.5
Xyris scabrifolia 24.5
Gaylussacia mosieri 2 23.0
Liatris spicata 22.2



















Physiographic Landform Types
SHighlands

SRidges, Uplands, Slopes

SLowlands, Gaps, Valleys, Plains



SCentral Highlands

SNorthern Highlands

SCoastal Lowlands aouthemn extent of
study area


Figure 2-1. Physiographic landforms modified from Puri and Vernon (1964). Colored shading
denotes three "generalized landforms" which separate Highlands from Lowlands, and
Northern Highlands (Clastic sediments) from Central Highlands (part of the carbonate
peninsular platform). Shaded regions indicate primary landforms denoted by
landform type. Approximate southern boundary of study region is shown.


















(b) Dry Uplands
SPanhandleXeric
Sand hills (D4)

O Panhandle Long leaf
Pine Clay hills (DS)


(a)All plot locations
SHistoric range
longleaf pine


Panhandle Slty
Vlodlands (D6)


(d) Mesic Flatwoods
North Florida Mesic I
Flat wood s (M2)

Xeric-Mesic
Flat wood s(M1)

Central Florida Mesic
Fl atwood s/ Dry Prai ries (M3)


(c) Dry Uplands
North Florida
Sandhills (D2)

North Florida Fich
Vlodlands (D1)

Fbninsula Xeric
Sand hills (D3)


Figure 2-2. Plot locations indicated by association. (a) Historic range of longleaf pine plus all
plot locations (yellow dots). Figures (b) through (f) show plot locations by
community association and primary landform types (Puri and Vernon 1964).
Community labels (in parentheses) correspond to Figure 2-3.






















(f ) at lan ds

A Marginal Rairies (W1)

North Florida Shrubby
We~t Flatwoods(W4)

Ran insula Wet
OFl atw ood s/ Rai ries (V2)

SCalcareous~let Flatwoods(VV3)


(e) At lands
Panhandle
Seepage Slopes (W7)

Pan hand le \et
Fl atwood s/ Rai ries (Vu~)

Upper Panhandle
We~t Flatwoods(W5)


Figure 2-2 continued.





u
NMS Axis 1


D5

m =.i

D1 .D2

,~~, ** "D4 !




DRY ,*
UPLANDS MR


v* O o


WETLANDS


MESIC
FLATWOODS


SW3


D1 North Florida Rich Woodlands (11)
D2 Noorth Florida Sandhills (31)
D3 Peninsula Xeric Sandhills (22)
D4 Xer c (31)
D5 Panhandle Longleaf Pine Clayhills (14)
D6 Panhandle Silty Woodlands (22)

M1 Xeric-Mesic Flatwoods (36)
M2 North Florida Mesic Flatwoods (30)
M3 Central Florida Mesic Flatwoods/Dry Prairies (22)


Peninsula Wet Flatwoods/Prairies (16)
Calcareous Wet Flatwoods (4)
North Florida Shrubby Wet Flatwoods (15)
Upper Panhandle Wet Flatwoods (7)
Panhandle Wet Flatwoods/Prairies (16)
Panhandle Seepage Slopes (5)


Figure 2-3. Two dimensional NMS ordination of 293 samples. Lines separate samples into three
community series, and colored symbols denote association from K-means cluster
analysis. Plot number per association noted in parentheses. Percent variation of
original distance matrix represented by NMS ordination: Axis 1 r2 0.54, Axis 2 r2
0.29. Stress = 14.78.


/ W5



4~ *
+=n e
o .W









CHAPTER 3
GEOGRAPHIC, ENVIRONMENTAL AND REGIONAL VARIATION IN FLORISTIC
COMPOSITION OF FLORIDA PYROGENIC PINELANDS

Introduction

Natural variability of plant communities is shaped by complex interactions of biotic and

abiotic factors. Models of "abiotic controls" (i.e. environmental control models) emphasize

influences of environmental gradients, resources limitations, and niche specialization (Whittaker

1956, Bray and Curtis 1957, Hutchinson 1957, Peet and Loucks 1977, Platt and Weis 1977,

Tilman 1994). In general, these models include local and regional processes which influence

species coexistence and distributions (e.g. niche assembly, limited resource availability,

environmental filters to species assembly, environmentally determined species pools). In

addition, natural disturbances are considered environmental influences, particularly those related

to density-independent processes (i.e. fire, hurricanes). Conversely, biotic control models of

community structure emphasize mechanisms unrelated to environmental determinants, e.g.

dispersal limitation, speciation and extinctions, competition, and herbivory. At its extreme,

biotic control theory states that community structure is governed strictly by dispersal limitation

and demographic stochasticity independent of local environmental influences (Hubbell and

Foster 1986, Bell 2001, Hubbell 2001).

The relative importance of biotic versus abiotic factors in structuring plant communities

is a subject of much recent debate (see Legendre et al. 2005). Recent models spatially explicit

models of community structure have suggested the relative importance of both (Borcard and

Legendre 1994, Okland et al. 2003, Tuomisto et al. 2003, Svenning et al. 2004). Spatial

autocorrelation in community composition is cited as evidence of biotic control models (Hubbell

and Foster 1986, Hubbell 2001, Condit et al. 2002). However, environmental variables

themselves may be spatially structured, and spatial autocorrelation of community structure may









be mis-interpreted if environmental-spatial relationships are not considered (Borcard and

Legendre 1994, Legendre and Legendre 1998, Legendre et al. 2005). In addition, inferred

environmental-species relationships may be biased in models that fail to include spatial trends

(Legendre and Fortin 1989). This underscores the need for spatially explicit models of species

composition and diversity, in which spatial autocorrelation and spatially structure environmental

variation can be quantified (Borcard et al. 1992, Legendre and Legendre 1998, Legendre et al.

2005). This approach allows formulation of hypotheses concerning the relative importance of

underlying mechanisms that influence community structure.

The relative influence of ecological determinates varies over different spatial and

temporal scales. Recent ecological theory supported by ecological observation suggests that,

with regard to relative importance as determinants of species composition and diversity, regional

factors and historic processes are comparable to local scale factors (Ricklefs 1987, Comell and

Lawton 1992, Collins et al. 2002). Processes related to historical biogeography, paleogeology

and recent land use history have been recognized as potential influencers of contemporary

community patterns (Okland et al. 2003, Graae et al. 2004, Svenning et al. 2004, Svenning and

Skov 2005). Regional influences of local species diversity may be a function of distinct species

pools (the "species pool effect"), in response to differential biogeographic and evolutionary

histories (Zobel 1992, 1997, Grace et al. 2000).

Natural disturbance and local environmental gradients influence local-scale community

structure of pyrogenic pineland vegetation of the Southeastemn Coastal Plain. Studies of

community composition, diversity, and species' response underscore the influence of

environmental determinants, including topography-moisture and soil properties (Kirkman et al.

2001, Drewa and Platt 2002a, Peet et al. 2003) and fire (Platt et al. 1988a, White et al. 1991, Platt










1999, Glitzenstein et al. 2003). Fire affects environmental conditions vis-a-vis regulation of

limiting resources such as available soil nutrients and light (Christensen 1977, Mitchell et al.

1999, Kirkman et al. 2001). In addition, fire affects competitive dynamics, particularly that

between woody and herbaceous vegetation (Streng et al. 1993). Biotic determinants of pineland

species coexistence are less well known in the Coastal Plain. However, there is some evidence

that dispersal limitation regulates plant diversity in temperate grasslands elsewhere.

Despite the relative wealth of research regarding local patterns and processes, little is

know about regional scale environmental-community relationships in Coastal Plain pinelands

and the interplay between regional and local scale relationships. Descriptions of grassland

diversity usually involve meso-scale regions < 1000 ha in size (Walker and Peet 1983, Tilman

1994, Grace et al. 2000, Weiher et al. 2004). In addition to high diversity linked to local

environmental gradients, Florida' s pyrogenic pinelands are notable for their regional floristic

diversity and high concentrations of endemic and restricted-range species (James 1961, Walker

and Peet 1983, Myers 1990, Peet and Allard 1993, Myers 2000, Sorrie and Weakley 2002).

These observations have lead to predictions of environmental controls of community

composition that include regional and local factors (Peet and Allard 1993, Grace and Pugesek

1997, Kirkman et al. 2001), as well as historical influences (Ricklefs 1987, Zobel 1992).

In this study, I present a model of variability in species composition and diversity of

pyrogenic pinelands. Specifically, environment-composition variation was analyzed in a

spatially explicit model, allowing quantification of environmental variation that is spatially

structured and spatially independent, plus the spatial component of variation that is unrelated to

measured environmental variables. This allowed formation of hypotheses concerning the

mechanisms controlling landscape vegetation patterns. Furthermore, I present hypotheses










regarding scales of influence of different environmental factors and historical processes. Finally,

I modified the spatially explicit model to include a generalized ecoregion model based on those

presently used in conservation efforts. The ecoregion model represents of regional differences of

biogeography and paleogeography. To develop these models, I used a large dataset of pineland

vegetation samples collected over a broad range of geography and environmental conditions.

Models based on these descriptive data were intended to generate hypotheses concerning

variation of relict pineland natural areas in a highly fragmented landscape of heterogeneous land

use.

Methods

Study Region

The study area includes the Florida Panhandle and most of central and northern

Peninsular Florida, approximating the current range of longleaf pine. This area includes roughly

nine million ha of the northern two-thirds of the state, extending approximately 480 km south

from the Georgia state border (approximately 310 00') to a southern boundary extending from

approximately 260 70' on the west coast to 280 80' on the east coast. This southern boundary

approximates southern extent of the "warm temperate moist forest" bioclimate zone (Holdridge

1967). The study region extends westward to approximately 87o 30' and eastward to the Atlantic

coast (approximately 800 00').

Florida is characterized by a humid subtropical climate. In general, mean temperature

and daily radiation increase with decreasing latitude. Mean annual maximum temperatures range

from 250 C in the western panhandle to 290 C in interior peninsular Florida, and minimum

temperatures and shortwave radiation vary likewise southward (13 tol7 o C, and 345 to 361

MJ/m2/day (Fernald 1981, Thornton et al. 1999). Average annual rainfall is highest in the









western panhandle (173 cm/year) and declines farther east and south, reaching its lowest point in

the central peninsula (approximately 124 cm/year; Fernald 1981). The distribution of rainfall

varies from northwest to southeast; winter and spring months are drier in peninsular Florida, with

more pronounced rainfall during the summer months. Rainfall is more evenly distributed

throughout the year in northwest Florida, with peaks during the late winter and summer months

(Chen and Gerber 1990).

The study region in Florida encompasses a wide range of edaphic conditions. Soils range

from drought coarse sands to poorly drained wetland mucks with high organic content. Entisols

are common in the well-drained uplands of panhandle and north Florida (Puri and Vernon 1964,

Myers 1990). Older and more weathered Ultisols and Alfisols are common in these regions and

are typically contain higher concentrations of Eine textured sediments such as clay and silt

(Myers and Ewel 1990, Myers 2000). Sandy, acidic spodosols are typical of upland woodlands

in coastal and peninsular regions. These infertile mineral soils have subsurface accumulations

of organic matter. Histosols with large accumulations of organic matter are common to poorly

drained wetlands (Brady and Weil 2000).

Vegetation and Environmental Data

Sample site selection and Hield methods are described in detail in Chapter 2. In brief, the

study area was stratified into 19 regions based on similarity of physiography, geology, soils,

climate, and historic vegetation maps (Fenneman 1938, Puri and Vernon 1964, Davis 1967,

Fernald 1981, Brooks 1982, Bailey et al. 1994, Griffith et al. 1994). Sample sites were selected

across regions in roughly equal numbers. Sample sites were subj ected to rigorous selection

criteria which precluded those without native pineland vegetation with recent history of Gire.









Sites degraded from anthropomorphic impact, fire suppression, and/or invasive species were

rej ected.

At each site, three or four zones were delineated relative to perceived gradients of

topographic-moisture conditions and change in plant species composition. A single 1000 m2 plOt

was randomly placed in each zone, and all vascular plant taxa were recorded as they were

encountered in a series of four nested sub-plots (plot areas: 0.01, 0. 1, 1, and 10 and 100 m2)

Aerial cover in the 100 m2 plOts was estimated by cover classes and averaged (by midpoint).

Species encountered in the remainder of the 1000 m2 area were assigned nominal cover

estimates.

All plots were sampled during the late summer though early winter (August-December)

over a four year period (2000 2004). Taxonomic nomenclature generally follows Kartesz

(1999). In Hield and herbarium plant identification I made frequent use of (Godfrey and Wooten

1979, Godfrey and Wooten 1981, Clewell 1985, Godfrey 1988, Wunderlin 1998, Weakley

2002). The vast maj ority of taxa were identified to species or variety; low resolution taxa

(family or genus) were omitted from analysis unless identification was consistent throughout the

dataset. The term "species" is used throughout to refer to the lowest recognized taxonomic

group.

Surface and sub-surface soil samples were collected for nutrient and texture analysis.

Four surface soil samples were collected from the upper 10 cm of mineral soil, and a single sub-

surface sample was collected from approximately 50 cm below surface. Dried samples were

analyzed at Brookside Labs in New Knoxville, Ohio. Nutrient analyses was performed via

Mehlich III extractant, an analytical procedure used for routine soil testing attempts to estimate

the amount of soil nutrients available to the plant during its growing season (Mehlich 1984).










Specific soil nutrient measurements were: total cation exchange capacity (meq/100g), pH,

estimated nitrogen release (N, ppm), extractable phosphorous (P, ppm), exchangeable cations in

ppm (Ca, Mg, K, Na), extractable micro-nutrients in ppm (B, Fe, Mn, Cu, Zn, Al), soluble sulfur

(S) and bulk density. Percent organic matter was determined by loss-on-ignition. Texture

analysis quantified percent sand, silt, and clay of surface and sub-soil samples.

Climate and elevation data were obtained for each geographic plot location. I

downloaded extrapolated weather parameters for specific locations from the DayMet

climatological model, available online (www.daymet.org). The Daymet model uses weather

station and elevation data to produce smoothed parameter estimates on a 1 km gridded surface

over the conterminous United States (Thornton et al. 1999). Daily parameter values were

available for an eighteen year period between 1980 and 1997. I calculated annual means and

standard deviations for the following: daily maximum air temperature, daily minimum air

temperature, daily average air temperature, total daily precipitation, and total daily shortwave

radiation. I downloaded elevation estimates for each geographic plot location from the HYDRO

1K North America DEM model webpage provided by the U.S. Geological Survey

(http://edc.usgs. gov/products/el evati on/gtopo30O/hydro/na~dem. html). Elevation values were

derived from a digital elevation model with 1 km resolution.

Numerical Data Assembly and Analysis

The response data matrix was assembled from species cover data from 270 samples.

Cover values for pine species (genus Pinus) were omitted from this the data matrix, although

other woody species cover values were retained. Species with fewer than three occurrences were

deleted (McCune and Grace 2002).










Species data were relativized to maximum species values and were transformed using a

Hellinger distance measure. When used in conjunction with Euclidean distance metrics and

linear ordination, the Hellinger transformation affects adequate representation of complex

species data without the problems associated with species weighting (i.e. chi-square distance

based methods; (Legendre and Gallagher 2001, Legendre et al. 2005).

I assembled four data matrices representing groups of potential explanatory variables for

statistical modeling. Collectively, these represent environmental and spatial explanatory factors.

The first, referred to as the edaphic varable matrix (EVM), initially included 24 soil descriptors

(listed above) and two variables describing local topography and elevation. Variable "topo" was

a subj ectively assigned descriptor of local topographic position relative to surrounding landscape

(value 1 to 4). The elevation ("elev") variable was the actual plot elevation derived from the

U.S.G.S. digital elevation model.

The second matrix was the climate variable matrix (CVM). It included means and

standard deviations calculated from the five DayMet parameter values (listed above). In

addition, I calculated means and standard deviations for total precipitation and daily shortwave

radiation for the growing season only (values from March 15 October 31). A total of 14

climate variables were included as potential explanatory descriptors in the initial CVG.

When necessary, soil, topographic, and climate environmental variables were

transformed to approximate normal distributions. Soil variables measured in ppm were log

transformed. Logit transformations were applied to proportional data (Tabachnick and Fidell

1996). Because of the varying scales and ranges of soils and climate variables, all EVM and

CVM variables were standardized to z-scores, expressed as standard deviations from the mean

(Tabachnick and Fidell 1996, Legendre and Legendre 1998).









The third matrix of potential explanatory variables included descriptors of spatial patterns

in the species data. The spatial variable matrix (SVM) described a trend surface response model

of geographic locations. Geographic coordinates of plot locations (X and Y, centered by mean)

were calculated from latitudes and longitudes superimposed on a geographic grid. Following

Bocard et al. (1992) and others (Borcard and Legendre 1994, Okland and Eilersten 1994,

Heikkinen and Birks 1996, Legendre and Legendre 1998), seven additional terms were included

in the initial SVM representing nine terms of a third-order polynomial regression of X and Y

coordinates:

Z = blX + b2Y + b3X2 + b4XY + bsY2 + b6X3 + b7X2Y + bsXY2 + b9 3

This approach allowed modeling of spatial trends that are more complex than linear gradients

(Legendre and Fortin 1989, Legendre and Legendre 1998).

The final explanatory matrix was based on a simple regional model of Florida

physiographic landforms. The regional variable matrix (RVM) categorized each plot location

into one of the four regions based on the generalized physiographic landforms of Puri and

Vernon (1964). In addition to the Northemn and Central Highlands, the Lowlands landform was

divided into the panhandle and peninsula regions (Figure 3-1). This regional delineation

approximates a general regionalized model based on several widely applied models of

Southeastern U.S. ecoregions (Omernik 1987, Bailey et al. 1994, Griffith et al. 1994, The Nature

Conservancy 2001). This model represents presumed regional differences in geologic and

evolutionary history that affect current patterns of spatial heterogeneity of pineland vegetation.

All environmental and spatial variables were individually screened for inclusion in their

respective variable matrix using the forward selection procedure and associated Monte Carlo

tests as implemented in CANOCO (Okland and Eilersten 1994, ter Braak and Smilauer 2002,









Leps and Smilauer 2003). Two forward selection tests were conducted. First, variables were

subj ected to forward selection in the context of a redundancy analysis (RDA) of a single variable

group. Second, forward selection was repeated in a partial redundancy analysis (pRDA) of each

variable group, with other environmental/spatial variables as covariables. Variables with p >

0.02 were excluded in subsequent canonical analyses corresponding to the model used for

selection. In this manner, explanatory variables with highest partial correlations with species

data were retained.

I applied a variation partitioning model to the species data, using the EVM, CVM, SVM

as explanatory variable matrices. Variation of the species data was decomposed into components

associated with pure and j oint contributions of explanatory factors. Specifically, I quantified

seven components of variation from six individual RDAs and pRDAs. The specific components

described fractions of total variation explained (TVE). These fractions are expressed in terms of

three pure factor effects and four interaction effects (following Cushman and McGarigal 2002).

Variation partitions correspond to fractions of reference diagram in Figure 3-2 as follows:

1. Pure ed'aphic effects: species variation explained by soils and topography variables, and
not related to climate and space variables (fraction a)

2. Pure climate effects: variation explained by CVM but not EVM and SVM (fraction b)

3. Pure spatial effects: variation explained by SVM but not EVM and CVM (fraction c)

4. Joint effects of edaphic and' spatial variables: variation j ointly explained by EVM and
SVM, but not related to CVM (fraction d)

5. Joint effects of edaphic and' climate variables: variation jointly explained by EVM and
CVM, but not related to SVM (fraction e)

6. Joint effects of climate and' spatial variables: variation j ointly explained by CVM and
SVM, but not related to EVM (fraction f)









7. Three way joint effects of edaphic, climate and' spatial variables: variation j ointly
explained by EVM, CVM, and SVM (fraction g)

The first variation partitioning model included EVM and CVM plus the polynomial spatial

matrix (POLY SVM). In the second model, the PCNM SVM replaced the POLY SVM.

Models of variation partitioning involved application of a series of constrained and partial

constrained canonical ordinations, as described by Borcard et al. 1992 and others (Borcard and

Legendre 1994, Okland and Eilersten 1994, Okland 2003). Significance of terms derived from

canonical ordination models were tested via Monte Carlo permutation tests (Peres-Neto et al.

2006). The null hypothesis tested was that of independence of species response data on the

values of the explanatory variables (Leps and Smilauer 2003). Fractions representing two and

three-way interactions and "unexplained" residual variation were calculated indirectly from

simple and partial terms; therefore, they were not statistically testable (Legendre and Legendre

1998, Peres-Neto et al. 2006). Variation partitioning and statistical tests were performed using

the vegan community ecology package version 1.8 for R software (Oksanen et al. 2007, R

Development Core Team 2007).

The relationships between floristic variation and abiotic gradients were assessed in the

context of the larger variation partition model. Constrained axes were tested via Monte Carlo

permutations for each of the following canonical ordinations: 1) RDA of EVM, 2) pRDA of

EVM after removal of CVM and POLY SVM effects, 3) RDA of CVM, and 4) pRDA of CVM

after removal of EVM and POLY SVM effects. Higher order canonical axes were tested using

pRDAs with lower order axes scores as covariables (Braak and Smilauer 2002, Leps and

Smilauer 2003). Significance of marginal effects of sequential canonical axes was assessed at p

< 0.02. Multiple simple correlations are presented as vector biplots superimposed on ordination

diagrams. Angles of the vectors denote direction of the highest correlation whereas vector









lengths correspond to strength of correlation. Significant correlations between canonical axes

and species richness (number species per 1000 m2 plOt area) are similarly presented. Canonical

axes scores derived from relevant RDA and pRDA ordinations are plotted against geographic

plot coordinates to visualize modeled trends in environmental and spatial variation. Individual

ordinations were run using CANOCO for Windows version 4.5 (ter Braak and Smilauer 2002).

Results

A total of 1009 taxa were identified from the 270 vegetation samples used in this study,

after omission of inconsistent and low resolution taxonomic identifications. After deletion of

infrequent entities, a total of 670 species were retained for analysis. Species richness ranged

widely in the samples, from 29 to 168 species/1000 m2 area. Similarly, vegetation samples

represent large variation in community composition, ranging from dry upland sandhills to

seasonally inundated wetland prairies. Community types and characteristic environmental

features are described in detail in preceding chapters.

About a quarter of the edaphic variables were omitted from inclusion in the EVM

explanatory matrix following the forward selection procedure in CANOCO (Okland and

Eilersten 1994, ter Braak and Smilauer 2002, Leps and Smilauer 2003). Of the original 26

potential explanatory variables, 14 soil property variables and two topographic variables were

retained for the RDA including the EVM explanatory factor only, with no covariables (see Table

3-2 for variable list). Thirteen soils and topographic variables were selected for the pRDA of

EVM, with CVM and SVM included as covariables. Three variables from the RDA set were

omitted in the pRDA set (Clay A, P, and Mn), while one was added (Al).

Of the initial 12 climate variables subj ected to the forward selection procedure, eight

were retained in the CVM for RDA ordination. These included descriptors of temperature,









radiation, and precipitation. Forward selection for the pRDA of CVM (with EVM and SVM

covariables) reduced the number of explanatory variables to only three (standard deviation and

total growing season precipitation, and standard deviation of daily shortwave radiation).

Similar forward selection procedures were applied to sets of spatial explanatory variables.

Of the nine terms of the initial polynomial trend regression, seven were retained for the SVM

(X2Y and Y3 were dropped).

Variation Partitioning Models

The total variation (aka "total inertia") of the Hellinger transformed species data set was

0.727, as expressed by the sum of all eigenvalues in an unconstrained principal component

analysis (PCA). The three explanatory matrices of the first variation partitioning model

accounted for 23% of the total variation (sum of all constrained eigenvalues / 0.727), leaving

77% of total inertia as "unexplained residual variation" (Figure 3-1). Although 23% is a small

portion of total inertia, this "total variation explained" (TVE) value is typical of community

variation studies (typically 20-50%; Okland and Eilersten 1994, Okland 1999). However,

following the advice of Okland (1999), TVE as a portion of total inertia should be interpreted

with extreme caution. The contribution of polynomial distortion (an artifact of eigenanalysis of

numerous response variables) was estimated as 30-70% of "unexplained variation" in simulation

studies of "noisy" data. The authors recommend presentation of variation fractions in terms of

proportions of TVE. This convention is followed hereafter.

Edaphic variables account for the largest component of species variation in the variation

partition model. Approximately 70% of TVE is related EVM, either as pure or interactions

effects: fraction a includes pure edaphic effects (48%), and fraction g, the three way interaction

effects (22%; Figure 3-1). The joint effects of edaphic variables with space (fraction b, 0%) and









with climate (fraction e, 0%) are too small to be included in the model. Variation attributed to

climate explanatory variables (pure or joint effects) is 44% of TVE (Figure 3-1). A small

percentage (9%) of TVE is pure climate effect (fraction b), and the remainder is divided between

space+climate joint effects (fraction f: 13%), and the three way interaction (fraction g). Similar

variation fractions are related to spatial trends with only 9% attributed to pure space effects.

Environmental Explanatory Variables

The primary complex gradient related to total edaphic variation (the EVM RDA; Figure

3-2) remains intact in the model of "pure" edaphic effects only, whereas higher order gradients

are diminished (Figure 3-3). The two-dimensional pRDA of EVM represents edaphic variation

after the removal of spatial structure (Figure 3-3 relative to fraction (a) of Figure 3-1). Similar to

the EVM RDA, axis one separates sites of lower slopes with organic soils from sandier sites on

higher topographic positions (first axis Figure 3-3 compared to Figure 3-2). Conversely, the soil

texture/nutrient gradient associated with geographic region (Axis 2 Figure 3-2) largely

disappeared in the pRDA (Figure 3-3), as did any evidence of regional separation of sites. The

second pRDA axis appears to represent a similar gradient of soil acidity and fertility similar to

the third RDA axis in Figure 3-2.

Geographic separation is dramatic in the ordination of climate variables related to species

variation. The CVM RDA represents all variation fractions associated with climate explanatory

factors (Figure 3-4 relative to fractions b+f+g of Figure 3-1). The first canonical RDA axis is a

gradient of all eight climate variables, related to temperature, daily radiation, and precipitation

(Figure 3-4). Peninsula sites are completely separated from panhandle sites along the first axes,

and they are characterized by higher mean annual temperatures and daily radiation. The

panhandle sites receive more annual rainfall that is more variables throughout the year and the









growing season. Species richness is also highly correlated with this climate gradient and

regional separation. The CVM RDA axis 2 explains about half again as much variation in

species data (Table 3-2), and is related to variation in precipitation (Table 3-1). Likewise, the

third canonical axis represents a gradient in precipitation (annual, growing season and standard

deviation) and radiation (daily mean and standard deviation; Table 3-1).

The influence of most climate explanatory variables disappears after removal of variation

related to spatial and edaphic factors. The CVM pRDA represents the variation fraction of

"pure" climate effects only, which is very small compared to total climate effects (bottom plot of

Figure 3-4 relative to fraction (b) of Figure 3-1). Not surprisingly, most of the climate variables

are not correlated with this small, non-spatially related variation fraction (and were eliminated in

the forward selection process prior to the pRDA). Regional site separation similarly disappears

in the pRDA (Figure 3-4). The first pRDA axis is related to total growing season precipitation

and variation in precipitation (Table 3-1). Axis two is a gradient of variation in radiation

(probably related to seasonality). Species richness is negatively correlated with pRDA axis 1,

although the relationship appears non-linear (Figure 3-4).

Mapped results from the simplified variation partition model of environmental and spatial

trends reveal regional patterns in species variation. The simple two part variation partition

model yielded the following fractions of TVE (figure not shown): pure environmental effects

(fraction a = 54%), pure spatial effects (fraction c = 9%), and joint effects of environment and

space (fraction b = 37%). These fractions resemble those of the initial three part variation

partition model; the spatial component (fraction c) is identical in both models. The contour map

corresponding to site scores from a RDA is constrained by all environmental and spatial

variables (fractions a+b+c = TVE) and displays regional distinctions (Figure 3-5). Scores from









the first and second canonical axes appear to distinguish the highlands physiographic landforms

in the upper panhandle and north peninsula from the coastal lowands (Figure 3-5a). Although it

represents a small portion of TVE (9%), the canonical axis site scores for variation explained by

pure space only display clear separation between panhandle and peninsula (Figure 3-5c).

Discussion

Based on the spatially explicit models of species-environment relationships, I propose

several interpretations regarding determinants of understory community structure of Florida

pyrogenic pinelands. First, environmental factors influence species composition and' diversity,

and' these controls are most prominent at local scales. The strongest gradient in species

composition is related to local topographic and edaphic features; about half of TVE is uniquely

correlated with these factors and not spatially structured (presumably representing small scale

gradients not captured by large scale spatial trends). Specifically, I presume that community

differentiation is concurrent with gradients of soil fertility and soil moisture, represented by total

N, organic matter, topographic position and soil texture\density variation. This gradient

separates dry upland vegetation from herbaceous dominated flatwoods and wetlands. Previous

community classification characterized vegetation along these primary gradients; sandhill

vegetation inhabits infertile sands (negatively associated with Axis 1, Figures 3-2 and 3-3) while

flatwoods and wetland vegetation are common in wetter, acidic sites with high nitrogen

availability and organic matter. The dominance of the soil fertility/moisture gradient is

demonstrated by large first axis eigenvalues, relative to those of higher order. Furthermore, this

local-scale gradient is consistently prominent after removal of variation effects of spatial trends

and climate.










Topographic position was the single most influential environmental variable of

community composition, and the most highly correlated with the local soil moisture-fertility

gradient. Understory community variation concurrent with local topographic gradients has been

noted elsewhere in the Southeastern Coastal Plain, both anecdotally and quantitatively (Walker

and Peet 1983, Bridges and Orzell 1989, Myers and Ewel 1990, Platt 1999, Kirkman et al. 2001,

Drewa and Platt 2002). In the current study, topographic position was a subj ective descriptor of

position along a local topographic gradient, regardless of absolute changes in elevation or slope

steepness. As revealed in the ordination models, topographic position covaries with soil texture

and fertility. However, in light of the large portion of variation attributed to topographic position

over and above that explained by other edaphic feature, this variable likely is a proxy for

unmeasured variation related to available soil moisture. Lower slope pineland communities

typically have higher perched water tables, more seasonal flooding and less soil leaching

(Abrahamson and Hartnett 1990, Myers and Ewel 1990, Kirkman et al. 2001).

Available nitrogen is related to local-scale variation in species composition in the current

study. In temperate grasslands elsewhere, nitrogen is a limiting resource and is related to

primary productivity and diversity (Seastedt et al. 1991, but see Turner et al. 1997). Soil

moisture and nitrogen availability are thought to be positively related in temperate forests

(Vitousek 1982). In this model, nitrogen is highly positively correlated with soil organic matter

and (presumably) soil moisture, and negatively related to soil density and coarseness. This is not

unexpected, as organic matter is a maj or source of nitrogen (Brady and Weil 2000). However,

these correlations contradict studies of nitrogen dynamics elsewhere in the Southeastern Coastal

Plain where nitrogen mineralization declined with increasing soil moisture, and/or was

negatively correlated with species diversity(Foster and Gross 1998, Wilson et al. 1999, Kirkman









et al. 2001). These studies involved correlation over single or few recently burned topographic-

moisture gradients, whereas the current study models environmental correlations and composite

gradients over a large geographic region. Available nitrogen varies with time since fire and fire

frequency (Christensen 1977). Despite attempts to "control" for recent fire history in site

selection, this may account for unmeasured variation in nitrogen (and other locally available

resources).

The second interpretation is that regional variation in pinelan2d community structure is

profound, and is related to variation of soil texture, soil nutrient availability, and climate.

Furthermore, regional variation of species composition is orthogonal to local variation. Spatial

trends in compositional variation are strongly dependent on latitude and longitude (X and Y

coordinates), and correspond to regional differences between the Florida panhandle and

peninsula. Eighty percent of variation related to climate variables is spatially structured,

reflecting distinct regional differences; the peninsula is hotter, less seasonal, and receives more

growing season rainfall, whereas panhandle rainfall is more evenly distributed throughout the

year (Fernald 1981, Chen and Gerber 1990). Regional differences in edaphic features represent

about half of spatially structure variation in species composition, and largely reflect soil texture

differences. Similar regional segregation concurrent with soil texture has been documented

elsewhere in the coastal plain, at both local and regional scales (Peet and Allard 1993, Dilustro et

al. 2002). Phosphorous and calcium are more abundant in soils of peninsular sites. This is not

surprising considering the presence of the carbonate Florida platform that underlies most of the

peninsula, and the presence of large amounts of phosphorite in some sediments of Pliestocene

origin (Puri and Vernon 1964, Brown et al. 1990).









The third interpretation is that residual spatially structured composition variation in is

related to regional differentiation in phytogeoguraphic distributions and endemism. Furthermore,

these patterns may reflect variations in biogeographic and evolutionary history, and/or recent

land use patterns. The unique fraction of space in the variation partition model represents a

small but inscrutable percentage of TVE (9%). This variation displays regional distinction

similar to that of the j oint effects of space and environmental factors (Figure 3-5). Interestingly,

the divide between peninsular central Florida vs. north and panhandle Florida coincides with

phytogeographic patterns. Many temperate plant species (woody and herbaceous) reach the

southern limit of their distributions in north peninsular Florida (see Chapter 2). Nearly a quarter

of the taxa included in the species-environment models have regionally restricted distributions in

Florida, and nearly 3% are endemic to one region (Chapter 2; James 1961, Sorrie and Weakley

2002). If the contribution of "pure space" does indeed manifest phytogeographic trends, this

suggests the influence of historical dynamics on contemporary patterns of species coexistence

(Ricklefs 1987). Florida's complex recent geologic history also underscores regional differences

between the panhandle and peninsula, including differential sources and timing of sediment

deposition and histories of sea level fluctuations (Randazzo and Jones 1997, Myers 2000). In

addition, there is evidence that the two regions were physically separated by the "Suwannee

Strait" for a period between 12 to 30 MRBP; an elongate negative structure extending across

southern Georgia and northeastern Florida (Hull 1962, Puri and Vemnon 1964, Myers 2000).

Differential land use patterns offer an alternative (but not mutually exclusive) explanation

for the compositional variation concurrent with regional segregation. Unfortunately this

observational study of floristic variation does not describe pre-settlement conditions, as this is no

longer possible because past land use and management created non-random "selections" of









natural areas in Florida. Further subj activity was introduced from the lack of random selection

of sites from an a priori "population" of natural areas. I attempted to minimize the latter problem

with a stratified sampling design and a large sample size (see Leps and Smilauer 2007).

However, any described variation of community structure is inherently confounded with recent

land use, particularly fire suppression and logging, as has been documented in other regions

(McIntyre and Lavorel 1994, Stohlgren et al. 1999, Vandvik and Birks 2002, Svenning and Skov

2005). Because open-range ranching was common in the central peninsula until recently, large

portions of this region continued to be frequently burned during the dormant season (Bridges

2006a, Bridges 2006b). Conversely, other regions of Florida suffered decades of Gire

suppression in the 20th century, with prescribed fire only recently introduced in selected natural

areas (e.g. the Big Bend and Marianna Lowlands regions, pers. obs.). Thus, regional differences

in recent fire regimes may contribute to unexplained compositional variation in the current

model. Effect may be direct (fire effects on resource availability, plant delectability) and indirect

(i.e. differential resource availability related to timber density and woody biomass).

A fourth interpretation is that grad'ients in composition are related' to grad'ients in species

richness, and' these are apparent at regional and' local scales. The variation in species richness

is high among study sites, ranging from 26 to 168 species/1000 m2. These herbaceous

dominated pineland communities are characterized by large numbers of small-statured species

present in low abundances. Community structure is influenced by the amount of "species

packing" at small scales. The richness gradient is most obvious at the regional scale, where

panhandle sites are consistently richer regardless of soil moisture/fertility conditions. Regional

influence on local diversity is a well documented phenomenon, and is attributed to the "species

pool" effect resulting from processes operating at multiple spatial and temporal scales (Zobel










1992, 1997, Collins et al. 2002). Interestingly, the richness gradient appears independent of the

primary local-scale gradient of soil moisture/fertility, contrary to observations in other grassland

ecosystems (Grace et al. 2000, Kirkman et al. 2001, Weiher et al. 2004).

After removal of variation associated with regional spatial trends, a richness gradient

persists and is weakly associated with soil pH, available nutrients and soil texture. This

secondary richness gradient is seemingly unrelated to regional segregation, and may reflect local

diversity patterns. Species richness associations with soil pH and calcium have been

documented in temperate grasslands and forests (Partel 2002, Palmer et al. 2003, Peet et al.

2003). Similar to Peet and Allard (2003), Florida pineland species richness is positively

correlated with pH and soil calcium, suggesting either larger pools of species adapted to basic

soils (regional "species pool" effect) or more favorable local conditions for plant colonization

and growth (local environmental effect). Alternatively, soil reaction is merely a proxy variable

for other unmeasured causative factors, such as competition for light or space. Density of woody

biomass increases rapidly on more fertile sites, which affects understory species richness vis-a-

vis competition for light and other resources(White et al. 1991, Streng et al. 1993, Grace and

Pugesek 1997, Palmer et al. 2003, Weiher et al. 2004). Fire encourages herbaceous growth,

colonization and diversity, in part, through control of woody competition (Drewa et al. 2002b).

Thus, the local variation in richness likely derives from a complex gradient of soil fertility and

di sturb ance.

The current model of community variation of Florida pineland vegetation underscores the

prominence of spatially structured and spatially independent environmental factors in shaping

community structure. Spatial structure in community structure and environment patterns is

common in studies of local to meso-scale variation (study region scale range approximately 10 -









1000 km2; Abrahamson and Hartnett 1990, Cushman and McGarigal 2002, Dilustro et al. 2002,

Graae et al. 2004,Svenning and Skov 2005, Laughlin and Abella 2007). By comparison, the

extent of my study region was orders of magnitude larger (approximately 137,000 km2)

Considering the relatively large region, it is somewhat surprising that the unique variation

fraction explained by space is small compared to total variation explained (9%). This suggests

the minor importance of biotic processes operating independently of environment. An

alternative explanation is that I failed to adequately model relevant spatial trends (e.g. small scale

spatial patterns). On the other hand, soils and topography appear very influential as determinants

of compositional variation supporting environmental control hypotheses. I posit that the relative

influence of environmental controls exceeds that of biological controls in species composition

and diversity of Florida pyrogenic pineland communities.

This model of pineland floristic variation supports hypotheses of regional influences on

local community structure and diversity. Regional differences in species richness and

composition exist, even after removal of regional environmental effects. Species composition

differs between the panhandle and peninsula sites with similar local environmental conditions.

This observations suggest influences of paleogeography and evolutionary history through

mediation of species pools (Ricklefs 1987, Zobel 1992, Zobel 1997), perhaps confounded with

trends in recent land use history (Graae et al. 2004, Graham et al. 2005, Laughlin et al. 2005,

Svenning and Skov 2005). Furthermore, the model suggests a hierarchical structure of

ecological determinants relative to the focal ecosystems, and that relative influences of

environmental factors are scale dependent.












Table 3-1: List of variables included in RDA and pRDA canonical ordination of variation partitioning models. Eigenvalue
indicates conditional correlation of single variables (with all other variables covariables). Correlation coefficients
listed for first three constrained RDA axes, and two constrained pRDA axes. Bold values indicated significant
correlations p < 0.05.


Eigen- RDA RDA RDA pRDA pRDA
value Al A2 A3 Al A2


Abbreviation Variable

Edaphic Variable~atrix (EV3)
Topo Relative position on slope (1-4)

Org Organic matter surface soil (%)

Sand A Sand in surface soil (%)

Sand B Sand in sub-soil (%/)

N Estimated total exractable nitrogen (ppm)

Density Bulk density (mg/m3)

Eley Elevation (m) from 1 km resolution DEM
coverage

Clay A Clay in surface soil (%)
pH pH surface soil
P Extractable phosphorous (ppm)
Ca Calcium (ppm)
B Boron (ppm)
Mn Manganese (ppm)
Fe Iron (ppm)
Al Aluminum (ppm)


0.08 0.79


0.05

-0.21

-0.52

-0.63

-0.16

0.09

0.50


0.50
0.26
-0.36
-0.39
0.07
0.54
0.29


0.06

0.28

-0.19

-0.08

0.25

-0.34

0.23


0.04
0.25
0.42
0.42
0.37
-0.34
0.04


0.82

0.41

-0.39

-0.15

0.43

-0.44





-0.27


-0.19
-0.33


-0.20
0.22


0.03

-0.03

-0.18

-0.24

-0.12

-0.03





0.43


0.12
0.42


0.14
-0.40


0.03

0.03

0.03

0.03

0.03

0.02


0.02
0.02
0.02
0.02
0.02
0.02
0.01
0.01



0.05

0.05

0.05


0.37

-0.25

-0.04

0.42

-0.41

-0.14


0.04
-0.25
-0.09
-0.12
-0.29
0.04
-0.14


Climate I ariable
Temp mean

Temp max
Srad GS std


Matrix (C~i 3
Mean annual daily temperature ( C)

Mean annual minimum temperature ( C)

Standard deviation mean growing season
shortwave radiation (MJ/nr'/day)


0.79

0.76

-0.78


0.08

0.15

-0.04


-0.13

-0.17

0.17


0.50














Table 3-1 continued.


Eigen- RDA RDA
value Al A2

0.04 -0.67 0.05


RDA
A3

0.32


0.36
-0.38
0.45
0.36


pRDA pRDA
AI A2


Abbreviation

Srad std


Variable

Standard deviation mean annual shortwave
radiation (MJ/m2/day)

Mean total annual precipitation (cm)
Mean daily shortwave radiation (MJ/m2/day)
Mean total growing season precipitation (cm)
Standard deviation of mean total growing season
precipitation (cm)


Prep_ann
Srad

Prep GS
Prep std


-0.50
0.38
0.30
-0.27


-0.33
0.16
-0.27
-0.36


0.47 0.16
0.48 0.33













Table 3-2: Results for Monte Carlo tests of canonical axes, for each of four ordinations.
Variation attributable to explanatory constraining variables indicated by the
culmulative percentage of variance (Cumulative % spp-env). F-ratio and p-value for
each test of canonical axis after partialling out variation attributable to lower
dimension axes.


Axis
eigenvalue
0.095
0.042
0.019
0.014


Culmulative
%o spp-eny
45.1
65.0
73.4
78.7


Canonical Model
RDA Edaphic variable matrix (14 vars)


Axis


2
3
4


F-ratio
26.67
12.29
5.73
4.38


p-value
0.002
0.002
0.002
0.002


pRDA Edaphic variable matrix (12 vars)



RDA Climate variable matrix (8 vars)






pRDA Climate variable matrix (4 vars)


0.081
0.012

0.055
0.028
0.017
0.010

0.010
0.005


57.6
66.2

42.8
64.4
77.4
85.2

39.6
72.7


26.06 0.002
3.83 0.004


15.27
7.97
4.88
2.93

3.60
1.87


0.002
0.002
0.002
0.006

0.002
0.016





E:~t 1


Fbtsidual


(a) 0.48 ()00
(e) 0.00


CVM



(g) 0.22 (b) 0.09
(f) 0.13



(c) 0.09 Variance not
SVM explained = 0.77





Figure 3-1: Venn diagrams of variation partition model. Ellipses represent three explanatory
matrices (see text). Shaded portions labeled with letters denote variation fractions
and proportion of TVE (see reference diagram top right).













115
O 110

00 100



Dest 0- 0 Top 7
o le o o o


0.4 4


0~ 0 4


-02 +


-02 t


O O


-08-0.6-04-02 0.0 0.2 0.4


0.6 0.8 1.0


RIDA Axi s 1


Figure 3-2:


Biplots of RDA ordination, constrained by edaphic variables (EVM). Top plot =
canonical axes 1 vs. 2; bottom plot = canonical axes 1 vs. 3. Vectors denote
individual soils and elevation variables, scales by direction and magnitude of
correlation with axes. Abbreviations same as Table 3-1. Symbols denote regional
locations (panhandle vs. peninsula). Top plot contours display significant
correlation of species richness with Axis 2 (r = 0.48).


Peninsula
0 Panhandle

O,
O

Bo yH Ca P 0, 0 0

a 00~ v Oo oo Topo








O 105

100


0.3 ..


0.0-


-0.1-


-0.2 -'


-0.3 --


Al "
OO
O


O 60


-0.4 -0.2 0.0


0.2

pRDA Axis 1


Figure 3-3:


Biplot of pRDA constrained by edaphic variables (EVM) with CVM and POLY
SVM covariables. Vectors show correlations of individual soils and elevation
variables with the first two canonical axes. Abbreviations listed in Table 3-1.
Symbols indicate regional locations (panhandle vs. peninsula). Contours display
significant correlation of species richness with Axis 2 (r = 0.59).


> Peninsula
O Panhandle
































I I I I I


0.6


0.4 4


0.2 4


Temp mean

*Temp max

L'


00
Srad std 18o
o -"


0.0 4


Srad GS std


-0.2 4


-0.2 0.0


0.2 0.4 0.6


RDAAxis 1


75


0.6 $


Prcp std


Prcp GS


-0.2 0.0 0.2


0.4 0.6 0.8


pRDAAxis 1


Figure 3-4:


Biplots of 1) RDA constrained by climate variables (CVM; top plot) and 2) pRDA
of CVM with EVM and POLY SVM covariables. Vectors show correlations of
individual climate variables with the first two canonical axes. Abbreviations listed
in Table 3-1. Symbols indicate regional locations (panhandle vs. peninsula).
Contours display significant correlation of species richness with Axis 1 (RDA: r =
-0.593, pRDA: r = -0.34). Note different scales for plots.


SPeninsula
o Panhandle





Canonical axis score

Plots a & b Plot c
1 < -0.2 < -0.4
S-0.1 -0.1
O 0.0 0.1
0 0.1 0.3
S> 0.2 > 0.6


Figure 3-5:


Contour maps derived from constrained ordination axis scores, displaying
geographic variation in variation parititions from the model of environmental-
compositional correlations. Plot (a) Axis 1 scores from RDA of all variation
components (environmental and spatial factors), corresponds to fractions a+b+c in
Figure 2-2. Plot (b) shows Axis 2 scores from the same RDA ordination. Plot (c)
displays Axis 1 scores from pRDA of spatial trends, after removal of environmental
factors (corresponds to fraction c).









CHAPTER 4
ECOLOGICAL RESTORATION OF A LONGLEAF PINE SAVANNA IN THE
SOUTHEASTERN COASTAL PLAIN

Introduction


Ecological restoration often involves rehabilitation of habitat structure to a semblance of

historic or "natural" conditions. It is assumed that restoration of physical habitat structure and

ecosystem function will prompt recovery or colonization of desirable native populations and

restoration of native diversity and composition (Palmer et al. 1997, Walker and Silletti 2006).

Community structure can be manipulated via reintroduction of natural processes (e.g., Gire)

and/or by artificiall" methods (mechanical or chemical treatments). Reintroduction of natural

disturbance regimes can variously influence community composition through mediation of

species recruitment and mortality and biotic interactions (Huston 1979, White 1979). Artificial

manipulation of community structure may expedite restoration of desired conditions, particularly

if remnant native populations remain on site, or if colonization is promoted (Palmer et al. 1997,

Walker and Silletti 2006). It is the goal of ecological restoration to induced temporal changes in

species composition and structure that resemble those of the desired conditions. Thus, it is

important to quantify succession following restoration, and compare that to reference conditions.

Longleaf pine savannas and woodlands native to the Southeastern Coastal Plain are

among the most imperiled ecosystems in North America (Croker 1987, Noss 1988, Frost 1993,

Means 1996, Platt 1999, Frost 2006). Native longleaf pinelands currently occupy less than 3

percent of their former range (Frost 1993, Outcalt and Sheffield 1996). Sites with vegetation

composition and structure similar to that of pre-settlement conditions are even rarer (Simberloff

1993, Varner et al. 2005). The rapid range reduction of longleaf pinelands coincided with

extensive logging, agricultural land use, and expanding rural settlement in the 19th and 20th









centuries (Croker 1987, Frost 1993). Virtually all old-growth longleaf pine was logged, and

much of the remaining land converted to pine plantations of slash (Pinus elliottii) or loblolly (P.

taeda) pines (Frost 1993). Other sites became overgrown with second growth pine and

hardwood species due to land fragmentation and suppression of natural fires (Platt 1999, Frost

2006).

Natural longleaf pine savannas of the Southeast are notable for their park-like stand

structure and exceedingly diverse ground cover vegetation. Under natural fire regimes,

monotypic stands of longleaf pine consist of patchily distributed cohorts of even-aged trees (Platt

et al. 1988b). Other trees and shrub species are largely relegated to the midstory and understory

strata and floristic diversity is concentrated in the herbaceous-dominated ground layer (Waldrop

et al. 1992, Peet and Allard 1993, Glitzenstein et al. 1995, Platt 1999, Drewa and Platt 2002). At

small scales, the ground cover vegetation of fire-maintained pinelands harbors exceedingly high

plant species diversity (Walker and Peet 1983, Bridges and Orzell 1989, Peet and Allard 1993).

Ground cover vegetation is comprised of large perennial bunch grasses, interspersed with smaller

and rarer grasses and forbs (Peet and Allard 1993, Platt 1999). An estimated 95 percent of

herbaceous ground cover species are perennials with adaptations for post-fire regeneration,

including rapid growth and sprouting, clonal growth, and obligate post-fire seeding (Platt 1999).

Post-fire growth of ground cover vegetation is rapid, with upwards of 100% biomass recovery

within 1 year (Oesterheld et al. 1999). Ground cover vegetation, coupled with highly flammable

longleaf pine needles, provides fine fuels necessary for ignition and spread of low intensity

ground fires (Robbins and Myers 1992, Streng et al. 1993, Platt 1999).

Most contemporary native longleaf pinelands are small and fragmented, and are no

longer subj ect to the natural processes to which constituent species are adapted. This condition










precludes the natural occurrence of frequent, low intensity fires that historically swept across the

landscape (Frost 1993, Simberloff 1993, Platt 1999, VanLear et al. 2005). Fire suppression of

remnant natural areas has induced shifts in species composition and community structure. Fire

intolerant pine and hardwood species colonize these sites and reduce herbaceous plant abundance

through shading and resource competition (Glitzenstein et al. 1995, Platt 1999, Provencher et al.

2000, VanLear et al. 2005). Fire suppressed longleaf pine communities are common in the

Southeast U. S. (Mehlman 1992, Brockway and Lewis 1997, Gilliam and Platt 1998, Varner et al.

2005). These sites may harbor persistent native plant populations (in dormant and active states),

but have suffered degradation of community structure in terms of shifts in species composition

and abundances and physical habitat.

Reintroduction of natural fire regimes can affect recovery of community structure and

species diversity in dry pinelands that have suffered moderate fire suppression (Brockway and

Outcalt 2000, Provencher et al. 2001, Kirkman et al. 2004, Walker and Silletti 2006). However,

little is know about restoration of pinelands that have suffered relatively long-term fire

suppression (>10 years). The persistence of native vegetation (perhaps in dormant states) is the

factor determining whether restoration can be accomplished by restoring natural environmental

conditions versus having to re-introduce native species, which is usually prohibitively difficult

and expensive (Seaman 1998, Walker and Silletti 2006, but see Cox et al. 2004).

Increases in woody biomass following fire suppression affect community structure and

function of native pinelands. Thick woody growth competes with herbaceous vegetation for

light and other resources, affecting succession, spatial heterogeneity, and species composition

(Brockway and Lewis 1997, Provencher et al. 2001, VanLear et al. 2005). Indirectly, woody

encroachment increases litter fuel loads, which may alter fire intensity and behavior to conditions









outside typical range of variability. Long periods of fire suppression and/or high intensity

wildfires may affect novel shifts in species composition and succession, or extirpation of native

populations (Varner et al. 2005).

Restoration success is highly dependent on initial site conditions. The degree of

departure from a desired restored condition (the "reference condition") guides the method and

intensity of restoration treatments (Fule et al. 1997, White and Walker 1997, Walker and Silletti

2006). Additionally, the reference condition provides a standard by which to evaluate site

restoration progress and success (White and Walker 1997). Despite problems with historic

differences in environment and stochasticity, and assumed "stasis" of a target condition (Palmer

et al. 1997), contemporary data from a disjunct but environmentally relevant is often the best

choice for a reference site ("same time, different place" sensu Walker and Silletti 2006).

Similarly, historical data may serve as reference site conditions, particularly if they are

proximate in geography and environment (Fule et al. 1997, Swetnam et al. 1999, Walker and

Silletti 2006).

Ecological restoration progress is often measured in terms of changes in overall species

richness or shifts in members of functional groups. Dynamics of these measures are assumed to

indicate changes in ecosystem production, function, and/or stability (Tilman 1996, Palmer et al.

1997). Species richness is often compared to reference sites to assess restoration progress

(White and Walker 1997, Provencher et al. 2001, Kirkman et al. 2004). However, increases in

species numbers do not necessarily imply congruence of successional trends or endpoints

between restored and reference sites (Walker and Silletti 2006). Explicit comparison of

compositional data provides more information about succession response to restoration

treatments.









The purpose of this study was to evaluate the effects native stand composition and

structure reconstruction have on the recovery of native ground cover vegetation. I hypothesized

that woody biomass reduction via mechanical treatment would release remaining understory

plant species from competition. Specifically, increased light availability from woody plant

reduction will favor growth and recovery of herbaceous plants. I predicted that reduction of

woody plant dominance via ecologically sensitive logging and reintroduction of native fire

regime would induce recovery of native ground cover vegetation resembling reference site

conditions. This prediction assumed a degree of resilience in natural temporal variability in plant

community succession. I measured the immediate and long-term effects on ground cover

composition following off-site pine removal (logging) and reintroduction of prescribed fire.

Methods

Study Site and Reference Sites

The Abita Creek Flatwoods Preserve (ACP) is located in the Gulf Coastal Plain region of

Southeast Louisiana (The Nature Conservancy 2001). The ACP is situated on a broad flat

terrace proximate to Abita Creek and its tributaries. The site encompasses Pleistocene deposits

of the Prairie and Citronelle terrace formations (The Nature Conservancy 2001), and is

characterized by level, poorly drained, fine silty loams dominated by the Stough soil series

(Fragiaquic Paleudults; Trahan et al. 1990).

Historical accounts of the ACP area describe a landscape of longleaf pine dominated

savannas and flatwoods prior to extensive logging of the early 1900's (Lockett 1870, Mohr 1898,

Penfound and Watkins 1937, Penfound 1944, Wahlenberg 1946). These pyrophytic

communities contained monotypic longleaf pine canopies, with trees up to 200-300 years old

interspersed with patches of younger pines, and lush ground cover vegetation of grasses and









other herbaceous species (Penfound and Watkins 1937, Penfound 1944). Pine savannas were

maintained by frequent (1-3 year interval) low intensity fires (Mohr 1898, Wahlenberg 1946,

Glitzenstein et al. 2003).

The Louisiana Field Office of The Nature Conservancy (TNC) acquired the 312 ha ACP

tract in 1996. The decision to purchase this tract was based in part on the presence of remnant

native ground cover vegetation and the presumption that vegetation could be restored to

approximate pre-settlement conditions. The site was never used for intensive agriculture. The

original pine canopy was logged in the 1930's; early aerial photographs show the area was

treeless in the early 1940's. Afterwards the site was burned infrequently, last burned circa 1980

before initiation of the current study (The Nature Conservancy 1997). As a result of fire

suppression, ACP was colonized by thick growth of tree and shrub species. By 1995, slash pine

(P. elliottii ) stands comprised the dominant ACP overstory (The Nature Conservancy 1997). A

few mature longleaf pines were present on site. Because regeneration of this species was

hindered by the competing thick shrub vegetation, no juvenile longleaf pines were present. The

midstory at ACP consisted of a nearly closed canopy of evergreen shrubs, including titi (Cyrilla

racemiflora), sweetbay magnolia (Magnolia virginiana), large leaf gallberry (Ilex coriacea),

swamp blackgum (Nyssa biflora), red maple (Acer rubrum), and gallberry (I. glabra).

Herbaceous ground cover vegetation was sparse and patchy, interspersed within the shrub

thickets.

I selected Lake Ramsey Wildlife Management Area (hereafter "Lake Ramsey") to serve

as a contemporary reference site for comparisons of ground cover composition. Lake Ramsey

contains native longleaf pine savanna vegetation that was frequently burned since the early

1990's and is considered a high quality example of regional native vegetation, resembling pre-









settlement conditions (Latimore Smith, pers. comm.). Lake Ramsey is approximately ten miles

west of ACP and is geologically and edaphically similar to ACP. The Louisiana Department of

Wildlife and Fisheries Natural Heritage Program established permanent vegetation monitoring

plots at Lake Ramsey and collected vegetation data annually between 1994 and 1999. The 1999

data are used in this study.

Secondly, I used species data from Penfound and Watkins (1937; hereafter "Penfound

data") as a historical reference condition of unaltered pine savanna vegetation in southeastern

Louisiana. The authors sampled ground cover vegetation of pine savannas and pine-cypress

communities in the mid-1930's, before and immediately after the original pine canopy was

logged. They recorded all ground cover species present in plots of fixed dimensions comparable

to that of the present study. Penfound' s "virgin longleaf pineland community" and "cutover

pine" sites (the "Knott tract... located some three miles East of Mandeville"; Penfound and

Watkins 1937) were within 15 kilometers of present-day ACP. Penfound's "slash pine-pond

cypress" community was also within 15 kilometers of my study area ("about 6 miles northeast of

Mandeville"). I deemed the Penfound data representative of "original" natural conditions of

these communities.

Restoration Treatments and Sampling Methods

My immediate goal at ACP was to restore the structure and composition of midstory and

overstory strata. My model was based on historical and contemporary descriptions of old-growth

longleaf pine stands. These accounts describe uneven-aged stands of longleaf pines with patches

of even-aged cohorts resulting from gap regeneration. Historical accounts were anecdotal

descriptions of Southeast Louisiana and surrounding regions (Lockett 1870, Mohr 1898,

Penfound and Watkins 1937, Penfound 1944, Wahlenberg 1946). Contemporary descriptions of









old growth longleaf pine stand structure and dynamics came from elsewhere in the Southeastern

Coastal Plain (Platt et al. 1988b, Platt et al. 1998; Platt and Rathbun 1993; Grace and Platt

1995a,b; Palik et al. 1997, 2003; Brockway and Outcalt 1998; McGuire et al. 2001).

A second goal was to restore longleaf pine canopy at ACP, and manage stands to promote

older-growth characteristics. Natural seed sources for longleaf pine regeneration were

unavailable. As such, I chose to remove canopy slash pine via commercial logging and plant

longleaf pine seedlings. I predicted logging would reduce woody biomass and reduce

competition for planted longleaf pine seedlings and recovering ground cover vegetation. In

addition, I predicted faster establishment of longleaf pine canopy would provide fuel and

promote fire behavior beneficial to natural community restoration (Robertson and Ostertag

2007). However, I was also concerned with potential negative impacts from soil disturbance

associated with the logging process (Greenberg et al. 1995).

Two restoration treatments were applied at ACP. The first was canopy removal via

commercial logging followed by prescribed fires (hereafter "logged+fire"). Five sample sites

were selected for monitoring ground cover response to the logged+fire treatments. The second

treatment consisted of prescribed fire only (hereafter "fire-only" treatment; four sites). Sample

sites were interspersed throughout the preserve, and treatments applied following a complete

randomized block design. All logged+fire sites were logged during the winter of 1997-1998

(Figure 4-1). A consulting forester worked with a small logging crew to remove trees using

skidders equipped with low-psi tires. In this manner, soil disturbance and damage to surrounding

trees were minimized. Logging was permitted only in dry conditions to minimize damage. All

merchantable canopy pines were removed in the logged+fire sites. Mean basal areas (woody

stems > 10 cm dbh) declined more than 90% after logging from 22.8 & 5.9 m2/ha in 1997 to 2.2 &









0.7 m2/ha in 1998. Over the same period, the mean basal area of the fire-only sites increased

slightly from 9.6 & 6.0 m2/ha in 1997 to 9.8 & 5.4 m2/ha in 1998.

All treatment sites were prescribed burned twice during the study period. The first burn

occurred in April, 2000, followed by the second in April 2003 (Figure 4-1). In a given year, all

study sites were burned in the same day under similar conditions. Basal area of woody stems >

10 cm dbh were little affected by the burns. Mean basal area of the logged+fire treatment was

already low following logging (2.2 & 0.7 m2/ha in 1998), and declined to 1.1 & 0.6 m2/ha in 2005.

Similarly, the fire-only mean basal areas declined slightly, from 9.8 & 5.4 m2/ha in 1998 to 8.6 &

4.3 m2/ha in 2005.

Vegetation sampling in Eixed permanent plots began in the summer-autumn of 1997,

before logging. Vegetation plots were sampled annually during the autumn months prior to plant

senescence after canopy removal (1998-2000). Autumn sampling facilitated plant identification

given the large proportion of plant species, especially grasses, that flower in the fall. The final

sample period was in the fall of 2005.

At each treatment site, I installed a single 0. 1-ha rectangular permanent plot (plot

dimensions: 20 m x 50 m, 1000-m2 total area). All vascular plant species were recorded in each

of four series of nested subplots ranging from 0.01-m2 to 100-m2 area. I estimated aerial cover

for each species recorded in 100-m2 Subplots using the following cover classes: 0-1%, 1-2%, 2-

5%, 5-10%, 10-25%, 25-50%, 50-75%, 75-95%, >95%. Additional taxa encountered in the

remaining 600 m2 were recorded and assigned a nominal cover estimate. Vegetation sampling

approximated the field methodology described by Peet et al. (1998).

In addition, I recorded plant species' cover in four 1-m2 Subplots per site to obtain more

accurate estimates of small scale cover. I overlaid a grid of 100 10 cm x 10 cm cells on each










subplot; each species present was assigned a value of 1-100, corresponding to the number of

cells in which it occurred. In this manner, species present in the 1-m2 Subplots received two

cover estimates: the first was the 1-m2 grid count, and the second was the cover class estimate

from the 100-m2 Sample area.

All woody stems > 2 cm diameter breast height (dbh) were tallied in the 1000-m2 Sample

area by species and size class (2-5 cm, then 5 cm size classes up to 40 cm dbh). For trees > 40

cm dbh actual diameter was recorded. Basal area (BA) was calculated using midpoint values per

size class.

I identified each plant taxon to the highest taxonomic resolution possible, which was

species for the maj ority of identifications. However, I assigned "low-resolution" identifiers

corresponding to genus or family for sterile or unidentifiable taxa. Low-resolution taxa were

included in species richness estimates, but omitted from datasets used for compositional

analyses.

Data Analysis

I used repeated measures models to compare trends of species richness and woody

species abundance between the two restoration treatments. My experimental design resembled a

Before-After Control-Impact (BACI) design, with two levels of treatment rather than one

treatment versus a control (Underwood 1994). I used mixed linear models consisting of fixed

and random effects, because of repeated measurements and my heterogeneous variance-

covariance estimates (Littell et al. 2000). Each model included two fixed effect factors:

treatment (TRT: logged+fire vs. fire-only) and time (YEAR: five levels corresponding to sample

years), and their interaction (TRT*YEAR). Random effects of mixed models included between-

year variances and within-year covariances of repeated measurements.









Within-subj ect measurements are often serially correlated in a predictable manner over

time. To accommodate this, I used the first three steps of the "four stage" method of Littell et al.

(2000) in determining mixed effects models for each response variable. First, I applied a

"saturated" model that included TRT and YEAR effects, plus pre-treatment basal area per 1000-

m2 area as a covariable (initial BA). The variation of initial BA was large, ranging from 0.96 to

3.93 m2/ha. Initial BA was omitted from subsequent models if it failed to explain significant

variation (p > 0.05 in the model including "unstructured" covariance parameter estimates).

Step two involved specifying models of covariance structure for each statistical model.

For this, the fixed effects portion of the model remained constant while I tested different

covariance structure models using Residual Maximum Likelihood (REML) computation in

PROC MIXED (SAS Version 8; Littell et al. 1996, Littell et al. 2000). I selected a covariance

structure for each statistical model which most closely approximated actual data with the fewest

parameter estimates. I compared the following model structures: compound symmetric (2

parameters: homogeneous variance and covariance), heterogeneous compound symmetric (6

parameters: heterogeneous variances and homogeneous covariance), Toeplitz (5 parameters:

homogeneous variance and heterogeneous covariance as functions of time lag), heterogeneous

Toeplitz (9 parameters: same as previous, with heterogeneous variance), first order

autoregressive (2 parameters: homogeneous variance and decreasing serial covariance dependent

on increasing time lag), first order heterogeneous autoregressive (6 parameters: same as previous

with heterogeneous variance), and antedependence (9 parameters: heterogeneous variances and

heterogeneous covariance relative to local serial autocorrelation, similar to Toeplitz structure).

The latter three covariance structures typically provide good fit to repeated measures data

(Kenward 1987, Littell et al. 2000). Each covariance model was compared to that of









unstructured covariance (each covariance parameter estimated independently) using Akaike's

information criterion (AIC) and likelihood-ratio tests with degrees of freedom equal to the

difference in number of estimated parameters (Littell et al. 1996, Littell et al. 2000). I selected

the most parsimonious covariance model with least deviation from the unstructured covariance

model in terms of parameter estimates and structure.

Finally, I used generalized least squares methods to test fixed effects on each of six

response variables. The first response variable was small woody stems per 1000-m2 (all stems >

2 cm but < 15 cm dbh). Next, I tested fixed effects on species richness, including counts by

lifeform group and at different sample scales (1-m2 and 1000-m2). I Separately analyzed 1) total

number of species, 2) number of graminoid species, which includes all true grasses (of the family

Poaceae) and morphologically similar species of the families Cyperaceae and Juncaceae, 3)

number of forb species, which include all non-graminoid herbaceous species, and 4) number of

woody species, including all non-herbaceous trees and shrubs. Significance tests were evaluated

using a conservative Type I error rate (p < 0.015) to avoid error inflation associated with

multiple tests. A separate error rate of p < 0.05 was applied to both models of TOTAL species

richness (1-m2 and 1000-m2). Most dependent variables were log transformed to improve

normality of residuals in the mixed linear models.

In each mixed linear model, the TRT*YEAR interaction was of primary interest because

this effect represents treatment effects over time. I assessed treatment effect in individual years

by using the "SLICE" option of PROC MIXED using SAS software Version 8 (SAS 2000).

Significance tests evaluate treatment effects for each year separately.

Species data from two sample scales were used to assemble species response matrices.

The 1000-m2 Species matrix included average cover class estimates from each of four years









(1997, 1998, 2000 and 2005; matrix dimensions = 36 plots x 217 species). The small scale

species matrix included mean percentage cover values from the four 1-m2 Small plots per site (45

plots x 121 species). Data from five years were included (1997-2000 and 2005). Species with

fewer than two occurrences in each dataset were deleted, as they contribute nothing to

calculations of inter-plot similarities (McCune and Grace 2002).

I applied the Hellinger distance transformation to the raw species cover data. When used

in conjunction with linear ordination methods, this transformation offers a better compromise

between linearity and resolution than do methods based on chi-square distances. This approach

avoids problems inherent to sample weighting, in addition to problems associated with using

Euclidean distances with untransformed data (Legendre and Gallagher 2001, Legendre et al.

2005). The 1-m2 Species data were log transformed prior to Hellinger transformation, whereas

the 1000-m2 data were not.

I used Redundancy Analysis (RDA) as a method for direct gradient analysis of species

compositional data. As a canonical ordination method, RDA directly relates species responses to

environmental factors. Sample scores are "constrained" as linear combinations of explanatory

variables, conceptually similar to linear (or multiple) regression. Treatment factors were

explanatory variables, identical to those used in univariate analyses.

I used RDA models similar to those described by Leps and Smilauer (2003) for analysis

of temporal compositional trends in a repeated measures experiment. I tested two null

hypotheses per response matrix: 1) there are no directional temporal changes in species

composition present in either or both restoration treatments (within subject effect), and 2)

temporal trends in composition changes are independent of treatments (between subj ect effect).

To test these, I varied constraining explanatory variables and covariables in each of two RDA' s.









In the first, YEAR effect and TRT*YEAR interaction were specified as constraining variables,

and Plot identity as a covariable. This corresponded to a model of YEAR effect only. The

second RDA was constrained by the TRT*YEAR interaction, and YEAR and Plot identity were

specified as covariables. This model corresponds to a test of interaction effects. Post-hoc

"contrasts" of treatment effects between specific years were performed using the same RDA

model as for hypothesis #2, with a species response data from the years of interest. Data

matrices were centered by species norms prior to ordinations. Scaling focused on inter-species

correlations to favor biplot interpretation (Leps and Smilauer 2003).

Significance of effects was tested with Monte Carlo permutation methods. Independence

of species data relative to the explanatory (constraining) variables was tested (McCune and

Grace 2002, Leps and Smilauer 2003). I used a restricted permutation configuration

corresponding to a split-plot design where permutations of repeated measurements were confined

within sample units (split-plots). Whole plots were permuted keeping within plot measurements

intact. All ordinations and permutation tests were performed with CANOCO (version 4.5) and

CanoDraw (version 4.0) software (Braak and Smilauer 2002). Small sample size and restricted

block design limited number of permutation configurations. Thus, to reduce the probability of a

Type II error, I selected a Type I error rate of p < 0. 10 for omnibus tests of main and interaction

effects and p < 0.05 for post-hoc contrasts.

Specific species most correlated with interaction effects (TRT*YEAR) were identified as

those with highest "fit" to the first canonical axes in RDA ordinations. The fit value for

individual species is the coefficient of determination corresponding to a regression of species

responses on sample scores on the first (canonical) axis. For each RDA, I selected the top 20-30

species with highest fit using the "lower axis minimum fit" inclusion option in CanoDraw (Braak









and Smilauer 2002). Species vectors in biplots represent magnitude and direction of the first axis

association.

Unconstrained ordinations were applied to species response matrices for illustrative

purposes. Solutions from principal components analysis (PCA: the unconstrained analogue to

RDA; Leps and Smilauer 2003) are presented to display successional trends of sample units.

Each PCA ordination was based on a cross-products matrix of inter-species correlations derived

from the Hellinger transformed data matrix. Compositional data were standardized by sample

norm (Leps and Smilauer 2003). I present two dimensional ordination solutions and report the

proportion of variance (in the species data) explained by ordination axes (McCune and Grace

2002).

Comparisons of ACP data to Reference Data

Species counts were transcribed from the Penfound data for comparison to ACP plots of

similar areas. Species richness was derived from non-overlapping subplots from the Penfound

data (areas = 1, 5, 10, 15, 20, 25 and 30 m2). Similarly, I tallied species from two overlapping

sample scales from the Lake Ramsey reference data (1-m2 and 10-m2 Sample areas; 11 plots).

Mean species numbers by ACP treatment were calculated for increasing sample areas: 1, 2, 10,

20, and 100-m2. Unlike the Penfound data, there was some overlap in ACP plot data due to the

nested plots.

I compared species area relationships between ACP restoration treatments in the final

sample year (2005). An ANCOVA model tested treatment effects (logged+fire vs. fire-only) on

species area relationships. Pre-treatment species counts (from 1997) and log sample area were

covariates in the model. Ninety-five percent confidence intervals were calculated for the two

species area curves by treatment type. A species area curve derived from the Penfound data is









displayed along with restoration treatment confidence intervals (statistical comparison is not

possible). Similarly, I visually compare species richness of Lake Ramsey data at two scales (1-

m2 and 10-m2)

I assembled species matrices from the Penfound and Lake Ramsey data in a manner

compatible with ACP taxonomy and sample scales. First, a presence-absence matrix of ACP and

Penfound species data was assembled from 30 m2 Sample areas (38 samples x 163 species). This

matrix contained data from four ACP sample years plus one sample from each of two Penfound

habitats sampled in late summer 1936 (the "cut-over pineland" and "pine-cypress community"

Penfound sites). Both habitats were described by the authors as open, herb-dominated

communities that burned frequently. Second, a similar presence-absence matrix was constructed

from ACP and Lake Ramsey species data from 10-m2 Sample areas. For this I used two 10-m2

sample areas per ACP (over 4 sample years; 72 total ACP samples), plus data from eleven 10-m2

Lake Ramsey plots sampled in October 1999.

I used unconstrained ordination to display temporal trends of ACP samples relative to the

Penfound and Lake Ramsey reference data. I applied non-metric multidimensional scaling

(NMS) ordination with an inter-sample distance matrix of Bray-Curtis coefficients derived from

the presence-absence species matrices. This method displays geographically disparate data

without constraints of explanatory factors (McCune and Grace 2002). Successional vectors

depict temporal trends in composition of the two restoration treatments relative to reference data.

Results

Trends in Species Richness and Woody Stems

Following logging, abundance of small woody stems (< 10 cm dbh) declined

precipitously (nearly 86%, from 192.4 & 67.9 stems / 0. 1 ha in 1997 to 27.2 & 9.7 stems in 1998)









and remained low in the logged+fire treatment throughout the study (Figure 4-2). Declines in

fire-only small stem counts occurred following the 2000 prescribed burn (dropping from a high

of 349.2 & 151.1 to 49.7 & 15.1 per 0. 1 ha). Repeated measures ANOVA of log transformed

small stem counts indicated significant YEAR and TRT*YEAR effects (Table 4-1). Treatment

differences are significant in the first two post-logging years (1998 and 1999) but disappear after

the first prescribed fire in 2000 (Figure 4-2).

Post-logging species richness increased in the logged+fire plots relative to the fire-only

plots (Figure 4-2). The main effect of YEAR was significant in an ANOVA of species richness

per 1000-m2 (Table 4-1). The TRT*YEAR interaction was significant (p = 0.047), indicating

different temporal changes in species richness by treatment. Treatment differences were greatest

in 1999 (two years post-logging), then diminished after prescribed burning in 2000 and 2005.

Similar trends were not apparent in the 1-m2 Sample data (Table 4-1).

Changes in numbers of graminoid species are responsible for treatment differences in

species richness. Individual ANOVA models showed significant temporal effects in graminoids

species richness (per 1000-m2) Only. The YEAR and TRT*YEAR effects were significant

(Table 4-1). Similar to overall trends, treatment differences in graminoid richness were

significant only in the second year post-logging (1999) and dissipated following prescribed fire

(Figure 4-3). Similar trends were not apparent in forb and woody species richness.

Trends in Species Composition

Initial changes in species compositional were pronounced in the logged+fire relative to

the fire-only treatment. Successional trajectories from 1997 to 1998 are greater in magnitude

and more uniform in direction for the logged+fire treatment, compared to the fire-only treatment.

Trends are more pronounced at the 1000-m2 than the 1-m2 Scale (Figure 4-4; PCA of 1-m2 data









not shown). The 1000-m2 PCA explained 36.3 percent of species variation in the first two

dimensions (first four eigenvalues = 0.22, 0.13, 0.10, 0.08). Permutation tests of treatment and

temporal effects support successional trends observed in PCA ordinations. The RDA of the pre-

and post-logging 1000-m2 Species data (1997 vs. 1998), constrained by TRT*YEAR interaction

effect, indicated differential species responses between treatments (Table 4-2; Figure 4-5).

Similarly, this contrast was significant in constrained ordination of the 1-m2 Species data (Table

4-2).

The magnitude and direction of species composition shifts became increasingly similar

between treatments over time. Successional trajectories between 1997 and 2005, represented by

the PCA of 1000-m2 Species data, are similar between logged+fire and fire-only treatments

(Figure 4-4). Permutation tests of YEAR and TRT*YEAR effects in RDA ordinations support

the observed pattern in succession. The constrained ordination of 1000-m2 Species data from all

years revealed significant YEAR effect, and TRT*YEAR interaction effects were marginally

significant (p = 0.07; Table 4-2 and Figure 4-6). These effects were similarly significant in

constrained ordinations of the 1-m2 Species data. Contrasts between first and last study years

only (1997 vs. 2005) showed no TRT*YEAR interaction effect in either the 1-m2 Or 1000-m2

species data (Table 4-2). As illustrated by the PCA of species data, successional trends from

1997 to 2005 were similar between treatments. Initial post-logging differences appear to have

diminished at the end of the study.

Initial response of herbaceous species to logging was pronounced. Logging triggered

increases in presence and abundance of many graminoid species and annual herbs. These

species were identified as those with highest correlations (of abundance data) with the first

constrained axis of the RDA of 1997 and 1998 data only. The first axis was constrained by









TRT*YEAR interaction (Figure 4-6; see Appendix A for species code legend). In RDA

solutions of 1-m2 and 1000-m2 Species data, there were more grasses, sedges, and forbs

associated with the logged+fire treatment plots. Eight logging responders are sedges (member of

the family Cyperaceae) and most of these are in the genus Rhynchospora. In addition,

depending on observation scale, there are 6 or 10 forbs and several grass species that responded

to logging. Many species that initially responded to logging were annuals, such as Scleria

muhlenbergia, Rhynchospora chapmanii, Eupatorium cappillifolium, Bidens mitis, BartoniaBBB~~~~BBB~~~BBB

paniculata, Drosera brevifolia, and Diodia teres. The response of an annual grass species,

Panicum verrecosum, was particularly pronounced at both sample scales. A few sub-dominant

perennial grasses responded quickly to woody removal, including Panicum rigidulum, Paspalum

floridanum,~~dddd~~~ddd~~~ and Anthaenantia rufa. In contrast, the few species associated with fire-only

treatment in the first post-logging year are mainly shrubs and vines.

Compositional differences between treatments persisted over most of the study period.

However, species associated with the logged+fire treatment over the study duration differed from

the initial responders. None of the longer-term responders were annuals. Most long-term

species associated with the logged+fire treatment were grasses, forbs, and a few sedges (at the

1000-m2 Scale: Figure 4-7). None were woody species. At the 1-m2 Scale, abundances of

bluestem grasses (Andropogon virginicus and A. cappilipes) increased in response to the

logged+fire treatment (Figure 4-7). Similarly, perennial sedges of the genus Rhynchospora (R.

elliotii, R. cephalan2thus, R. oligan2tha, and R. gracilis) increased in presence and abundance in

the logged+fire treatment. The latter species was the dominant non-grass monocot of ACP pine

savannas. Other post-logging responders included small statured perennial forbs, most having

over-wintering rosettes and member of the families Asteraceae and Xyridaceae. Few species









were associated with the fire-only treatment over the study period at either sample scale, and

these were mainly shrubs, vines, and forbs.

ACP Treatment Responses vs. Reference Conditions

Species-area relationships differed between restoration treatments at the end of the study

period (Figure 4-8). An ANCOVA of the 2005 sample data revealed a significant treatment

effect with initial BA included as a covariable in the model (TRT: F1,40 = 6.1, p = 0.018; initial

BA: F1,40 = 132.9, p < 0.0001). Area was significant (F1,40 = 47.39, p < 0.0001) but the

TRT*"AREA interaction was not (F1,40 = 0.49, p = 0.487).

Species richness of the logged+fire ACP plots (in 2005) exceeded that of Lake Ramsey

plots at the 10-m2 Sample area, although they were similar at the 1-m2 Scale. At the 10-m2 Scale,

mean and standard error of Lake Ramsey species counts falls below the 95 percent confidence

interval of the ACP logged+fire treatment (Figure 4-8). In contrast, species richness of the

Penfound data exceeds that of both ACP treatment sites at areas > 10-m2, exceeding ACP 95%

confidence intervals. I was unable to formally test differences in species richness between

Penfound and ACP data due to lack of replication and differences in sampling methods

(overlapping vs. non-overlapping plot layouts). However, the species-area pattern of the ACP

logged+fire treatment suggests recovery of ground cover richness approaching that of my

historic reference site.

Both of the ACP restoration treatments prompted species composition changes that

resembled reference site conditions. Compositional shifts were similar in direction but

apparently differed in magnitude between treatments. The largest temporal changes were in

plots with higher initial BA (Figures 3-8 and 3-9). The NMS ordination in Figure 4-8 compares

presence-absence species data from Lake Ramsey and ACP successional vectors (10-m2 plOt










size). Most ACP trajectories indicate directional shifts toward the reference composition.

Similar patterns were apparent in the NMS ordination of Penfound reference data plus ACP

successional vectors (Figure 3-9). Compositional shifts of the high initial BA plots appear most

pronounced along the first NMS axis, toward the Penfound data points. Smaller shifts of the

"Low initial BA" plots are directed toward the "pine-cypress" Penfound data point specifically

(Figure 3-9). In general, ACP treatments promoted compositional changes toward reference site

composition.

Discussion

The current study demonstrates dramatic recovery of an ecologically degraded pineland

plant community following restoration of natural forest structure and ecosystem processes.

These results underscore the historical importance of forest structure and fire regime in

maintaining this natural ecosystem, and the importance of timber management and prescribed

burning for restoration of similarly degraded pinelands. The recovery of understory herbaceous

vegetation was rapid, and resembled the quality of reference sites. Changes in ground cover

species composition were pronounced following reduction of midstory shrubby vegetation in

both ACP treatments, as the open aspect of pre-settlement conditions was restored. Although

direction of change appeared similar between logging treatments, rates of change appeared to be

accelerated by timber removal.

In this case, fire appeared to ultimately have a greater effect on ground cover species

composition than mechanical tree harvest. Following two prescribed burns, trends in species

richness and composition suggest convergence between mechanical restoration treatments,

despite persistent differences in canopy densities. The mechanical activity of logging

immediately reduced shrubby biomass, which prompted a flush of herbaceous growth and









diversity, which was subsequently sustained by prescribed Gire. However, the initial tree basal

area of fire-only sites was less than that of the logged+fire treatment. Within the range of my

overstory tree basal area among treatments following logging (1-10-m2/ha), competitive

interaction between canopy and understory vegetation was probably minimal compared to that of

midstory and understory interactions. Thus, it should be noted that the positive restorative

effects of Gire might be limited at higher levels of tree basal area. In the logged+fire treatment,

fire invoked woody biomass decline, subsequently resembling that of the fire-only treatment.

The trend toward similarity in woody biomass roughly coincided with convergence in

herbaceous plant richness and composition between treatments.

Fire mediated effects are most pronounced in the ground layer. Streng et al. (1993)

suggested that frequent fire promotes establishment of rarer species in an environment dominated

by long-lived perennials by freeing up space and resources available for colonization, and by

reducing competition from dominant grasses and woody species. Fire likely decreases

competition between understory woody and herbaceous plants for light and space (Platt et al.

1988a, Streng et al. 1993, Glitzenstein et al. 2003, Walker and Silletti 2006). Other restoration

studies report herbaceous vegetation recovery in response to fire plus hardwood reduction that

exceeded that expected from the chemical and mechanical treatments alone (Brockway and

Outcult 2000, Provencher et al. 2001). In the latter, fire alone prompted greater herbaceous

ground cover response than mechanical hardwood reduction in Florida longleaf pine sandhill

restoration (Provencher et al. 2001). Similar to my results, fire effects on ground cover

vegetation extend beyond reduction of canopy density (Platt et al. 1988a, Robbins and Myers

1992, Waldrop et al. 1992, Streng et al. 1993, Provencher et al. 2001).









The restoration treatments preferentially prompted responses of species that are

characteristic of pine savanna natural areas. All species that responded to restoration treatments

were native, and with few exceptions, were not ruderal generalists. Species that initially

responded to logging were characteristic savanna herbaceous species. The flush of "new"

species included many grasses (plant family Poaceae) and sedges (family Cyperaceae),

particularly small statured, rhizomotous species of the genus Rhynchospora. Longer-term

species responders were primarily sedges and perennial forbs. Plant populations likely persisted

at ACP during the period of fire suppression, either in the seed bank or in dormant vegetative

states. The initial flush of rhizomotous sedges and clonal grasses suggests long term persistence

in vegetative states followed by rapid growth in response to increased light and space. These

dormant lifeforms may not have been detected in pre-treatment sampling. Increases in species

richness and abundance have been noted in other studies of woody removal by mechanical and

chemical means (Greenberg et al. 1995, Harrington and Edwards 1999, Brockway and Outcalt

2000, Provencher et al. 2000, Provencher et al. 2001). In these studies, increased richness and

cover were attributed to soil disturbance as a direct effect of mechanical manipulations

(Greenberg et al. 1995, Harrington and Edwards 1999, Cox et al. 2004), in addition to indirect

effects of increased light, moisture, and space availability (Harrington and Edwards 1999,

Brockway and Outcult 2000, Provencher et al. 2000, Provencher et al. 2001). Most studies

indicate eventual increases in native species typical of the focal habitat, rather than increases in

exotic or ruderal species (but see Greenberg et al. 1995, Harrington and Edwards 1999).

The absence of ruderal additions to the local species pool affirms predictions that

ecologically sensitive logging would not cause novel trends in post-logging succession. The

number and abundance of non-ruderal annual species increased in response to logging, but









declined after the first year. Similarly, restoration of Florida xeric pinelands via mechanical

methods prompted an initial insurgence of native ruderal species (Greenberg et al. 1995,

Provencher et al. 2000), although the short duration of these studies precluded eventual detection

of decline. Similar to my findings, these studies documented no invasions of non-native species

that invoked novel succession.

Response of pine savanna herbaceous vegetation to woody reduction apparently differs

by life form type across moisture conditions. Restoration of xeric pineland communities of the

Gulf and Atlantic Coastal Plain regions (via mechanical, chemical, and prescribed burning

treatments) prompted greatest increases in forb richness and abundance (Brockway and Outcult

2000, Provencher et al. 2000, Provencher et al. 2001, Platt et al. 2006), whereas prescribed

burning of wet-mesic Atlantic coastal pine savannas simulated increases in grasses and sedges

(Walker and Peet 1983, Glitzenstein et al. 2003). Although I documented increases in all

herbaceous life form types following restoration treatments, grass and sedge increases were most

pronounced in my mesic to wet pine savanna site. These observations suggest similar restoration

treatments in different moisture conditions may trigger different compositional responses,

relative to the composition of the residual species pool and the differential loss of species groups

(Walker and Silletti 2006).

ACP Restoration Compared to Reference Model

Ground cover vegetation of ACP ultimately resembled that of my reference sites

following restoration of stand structure. Successional endpoints of ACP treatments were similar

to both the historic (Penfound) and contemporary (Lake Ramsey) reference data. Congruence

may be attributable to shifts in composition (constituent species and relative abundances) rather

than changes in species richness (species presence). Species richness increases were sustained in









the logged+fire treatment only, and these numbers approached those of the historic (Penfound)

reference condition. Similarly, species richness of fire suppressed Florida longleaf pine sandhills

increased following mechanical hardwood reduction coupled with fire, approaching and in some

cases exceeding that of the reference site (Provencher et al. 2001). The authors credit fire as the

dominant cause of increases in species richness and densities.

Succession in response to restoration of stand conditions and ecosystem function

depended largely on starting conditions. Compositional shifts were qualitatively similar between

treatments and over varying initial of canopy and midstory densities. However, the magnitude of

ground cover response depended on initial canopy and midstory densities, which ranged

considerably (approximately 10 22 m2/ha basal area and 24 756 stems (< 10 cm dbh) per 0. 1

ha). Restoration responses relative to starting condition have been observed by others (White

and Walker 1997, Walker and Silletti 2006), with densely wooded sites having largest vegetation

responses to woody biomass reductions.

Although studies of pineland restoration document significant changes in ground cover

vegetation (Greenberg et al. 1995, Harrington and Edwards 1999, Brockway and Outcult 2000,

Provencher et al. 2000, Provencher et al. 2001, Walker and Silletti 2006), most do not include

explicit and quantitative comparisons to reference models. Changes in species composition or

richness alone do not necessarily indicate restoration success, if the goal is to mimic some

historical or contemporary "natural" condition. It is possible to trigger novel or unintended

succession that may be difficult to detect without reference comparisons (Fule et al. 1997, White

and Walker 1997). Comparison of treatment responses to proximate and quantitative reference

data allowed me to assess progress and conclude that efforts have promoted vegetation recovery









approximating that of desired conditions. Furthermore, undesirable non-native or weedy species

were not introduced nor was undesirable succession invoked.

Management and Conservation Implications

Success of pineland restoration depends in large part on starting conditions and

ecological resiliency of the treatment site. To date, few studies in the Southeastern Coastal Plain

indicate the potential of plant community recovery, in terms of species composition and

diversity, resembling a desired condition (Harrington and Edwards 1999, Brockway and Outcalt

2000, Hedman et al. 2000, Provencher et al. 2001, Glitzenstein et al. 2003, Platt et al. 2006).

These studies indicate ground cover vegetation recovery following reduction of woody biomass

via various means, including mechanical and chemical methods, and prescribed burning

(Harrington and Edwards 1999, Brockway and Outcult 2000, Provencher et al. 2001). It is

important to examine ecological starting conditions represented in these restoration studies, and

compare these to the current study. Where available, approximate mean starting (or control)

densities of mechanically or chemically treated sites (in longitudinal or retrospective studies)

were as follows: 10.4 m2/ha basal area (Harrington and Edwards 1999), ~124 (oak) stems/0.1 ha

(Provencher et al. 2001), and ~14-18 m2/ha basal area and ~53-130 stems/0. 1 ha (Hedman et al.

2000). Canopy and midstory densities of the aforementioned studies appear to approximate

densities of typical second growth pine stands not subj ected to industrial plantation management

(basal areas < 20 m2/ha; see Robertson and Ostertag 2007). By comparison, my starting basal

areas of roughly 10-22 m2/ha were similar to those of Harrington and Edwards (1999) and

Hedman et al. (2000). However, initial stem densities of my treatment sites were high in

comparison to those reported, and my initial variance was also large (overall mean and standard

error is 241.4 + 81.7; range 47-781 stems/0. 1 ha). My higher stem density mean and variance










may be attributable to differences in moisture conditions and community type. Other restoration

study sites were upland woodlands or dry upland sandhills, whereas mine contained mesic and

wet pine savanna communities.

These results contribute to an overall model of restoration potential for Southeastern

Coastal Plain pinelands. Similar to other studies, mine indicates that ground cover recovery is

possible on degraded pineland site that has suffered fire suppression and/or fire regime alteration

over a period of 10-20 years, and the associated dense woody growth. Recovery is possible in

the absence of previous extensive soil disturbance associated with past agricultural or

silvicultural land uses. Little is known about the resiliency of pineland ground cover following

ground tilling, although Hedman et al. (1999) and Ostertag and Robertson (2006) found evidence

of persistent changes in succession in second growth pinelands on fallow Hields. Furthermore,

the current study shows plant community recovery is possible in a wetter site, with a species pool

adapted to different moisture and edaphic conditions than those previously examined.

Although reintroduction of native fire regime was arguably the most important

management prescription for ecological restoration, there were benefits to canopy removal via

commercial logging at ACP. Over time, it is likely that frequent fire alone would eventually

reduce woody vegetation in fire suppressed pinelands such as ACP. However, reduction of

woody vegetation in this manner may require decades of frequent and intense burning (Waldrop

et al. 1992, Glitzenstein et al. 1995, Olson and Platt 1995, Drewa et al. 2002). This may exceed

the time in which remnant populations of ground cover species are available for re-colonization

(Hedman et al. 2000, Walker and Silletti 2006). For this reason, expediting restoration of stand

structure using mechanical methods may enhance ecological restoration. In addition, financial










returns from commercial timber sale may offset restoration costs without introducing detrimental

effects.

These results, along with other restoration studies, demonstrate that ground cover

vegetation recovery is possible in degraded pine savannas over a range of starting conditions,

without resorting to artificial species reintroductions. Furthermore, results highlight the innate

resiliency of pineland groundcover plant communities. Life history adaptations of plant species

for dormancy may buffer populations in periods of atypical environmental conditions (i.e. Gire

suppression). Temporal rebounds of plant community composition suggest some degree of

successional "stability" within a range of Gire regime and stand structure variability.

However, succession in former pinelands that have been severely altered (ground tilling,

agricultural land uses, and fire suppression exceeding several decades) may exceed that range of

resiliency and require more intensive restoration treatments to achieve reference conditions

(Walker and Silletti 2006). Although re-seeding efforts were successful in former longleaf pine

sandhills (Seaman 1998, Cox et al. 2004), it is costly which limits its application to small areas

(Walker and Silletti 2006). Because of this, treatments of lower intensity and cost are desirable,

and appropriate for ecological restoration of large areas (Provencher et al. 2001). Fortunately,

restoration of Southeastern Coastal Plain pineland stands with mechanical woody reduction can

be labor and cost efficient, and yield favorable results if coupled with appropriate fire

management.











Table 4-1. ANOVA tables for models of species richness and stem numbers. Main, interaction,
and covariable effects listed in "Effect" column. "Slice" effects (treatment effects in
individual YEARS) indicated for first three models. The covariance structure and
number of parameters selected for each model are listed. P-values less than critical
values are shown in bold type.

Model Cov structure Effect Num df Den df F-value p-value
Number stems Hetero toeplitz Treatment 1 7.01 3.57 0.101


(1000 m2) (9 cov parameters)


Year
Trt*Year
1997: Trt effect
1998: Trt effect
1999: Trt effect
2000: Trt effect
2005: Trt effect


7.89
7.89
8.31
7.87
7.13
7.45
7.33


10.54
13.88
0.01
9.37
7.51
1.97
1.61


0.003
0.001
0.923
0.016
0.02
0.2
0.243


Total species
(1000 m2)


Antedependence Treatment


1 7.1 2.98 0.13


(9 cov parameters)


Year
Trt*Year
1997: Trt effect
1998: Trt effect
1999: Trt effect
2000: Trt effect
2005: Trt effect

Treatment
Year
Trt*Year
1997: Trt effect
1998: Trt effect
1999: Trt effect
2000: Trt effect
2005: Trt effect


4.81
3.64
0.23
3.89
8.96
6.18
4.59

2.56
3.79
7.82
4.73
3.82
8.81
3.99
1.99


0.022
0.047
0.644
0.089
0.02
0.042
0.069

0.149
0.014
0.001
0.045
0.068
0.009
0.063
0.178


Gram species
(1000 m2)


Autoregressive
(2 cov parameters)


7.95
27.3
27.3
16
16
16
16
16


Forb species
(1000 m2)


Woody species
(1000 m2)


Antedependence Treatment


1 7.03 0.39 0.551
4 8.71 2.23 0.148
4 8.71 1.4 0.311


(9 cov parameters)


Autoregressive
(2 cov parameters)


Year
Trt*Year

Treatment
Year
Trt*Year

Initial BA
Treatment
Year
Trt*Year


7.22
27.8
27.8

5.7
5.84
26.7
26.7


0.62
3.08
1.99

31.19
6.7
3.42
1.8


0.458
0.032
0.124

0.002
0.042
0.022
0.159


Total species Autoregressive


(1 m2)


(2 cov parameters)









Table 4-2. Results of Monte Carlo permutation tests from RDA constrained ordinations. Specific datasets subj ected to ordinations and
permutation tests are listed in left column (1000 and 1-m2 Species data), along with data matrix dimensions. The null
hypothesis tested is listed in the Model column. Significant p-values highlighted in bold text.

First
Sum all canonical
Dataset Matrix dimension Model eigenvalues eigenvalue F-ratio p-value
1000 m2 4 years 36 plots x217 spp No YEAR effect 0.369 0.057 4.62 0.05
No YEAR*TRT effect 0.341 0.022 1.71 0.07


1000 m2 '97 vs. '98 18 plots x 190 spp No YEAR*TRT effect 0. 141 0.044 3.2 0.02
1000 m2 '97 vs. '05 18 plots x 193 spp No YEAR*TRT effect 0.204 0.037 1.53 0.17


1 m2 5 years 45 plots x 121 spp No YEAR effect 0.308 0.034 4.22 0.01
No YEAR*TRT effect 0.279 0.016 2.09 0.02


1 m2 '97 vs. '98 18 plots x 98 spp No YEAR*TRT effect 0. 123 0.026 1.85 0.03


1 m2 '97 vs. '99 18 plots x 96 spp No YEAR*TRT effect 0. 138 0.065 1.62 0.05


1 m2 '97 vs. '05 18 plots x 99 spp No YEAR*TRT effect 0.209 0.032 1.26 0.17











ilr


Figure 4-1. ACP pictures: (a) pre-treatment in 1997, (b) immediately after logging in 1998, and (c) after logging and first
prescribed fire in 2000.


' ~(b)



































I I I I I
1997 1998 1999 2000 2005


Logging


1st Fire 2nd Fire


Year


Figure 4-2.


Least squares means (LS means) and standard errors of the
number of small stems (< 15 cm dbh) per 1000-m2 Sample.
Closed circles = fire-only plots, open circles = logged+fire
plots. Arrows indicate timing of specific restoration treatments.






























Graminoids only











Logging 1st Fire 2nd' Fire


1997 1998 1999 2000 2005
Year


All species


Figure 4-3.


Least square means (LS means) and standard errors of species
richness by treatment and year. Open circles = logged+fire
treatment; closed circles fire-only treatment. Top plot shows
total species richness 1000-m2 Sample area; bottom plot shows
means of graminoid species only. Timing of treatments
indicated by arrows.












1997 vs. 1998


5




0




.5


1997 vs2005


-1.0'-
-1.0


-0.5 0.0 0.5
PCA Axis 1


PCA ordination of ACP species data (1000-m2 Scale); first two
ordination axes displayed. Dots indicate 1997 pre-treatment
compositional data in ordination space. Successional trajectories
correspond to compositional shifts of data from individual plots:
red = logged+fire treatment, black = fire-only treatment. Top
plot shows changes between pre-treatment and first post-logged
years (1997 vs. 1998). Bottom plot shows shifts from pre-
treatment (1997) and after logging and fire (2005).


Figure 4-4.












11*1,


. 1 I
-1 .0 -0.5 0.0 0.5


PANVE


cu 0.2




0.0

-0.1


RHYCN
FBRLOBBR
BCLPX BIDMI
~vvo n\ EUPCA


TRADI


PANRI


NYSBI Fire only


Logged+fire


ANTRU Y''~l
EUPLERHF


PANSO


PANVE


Fire only


RIDA Axis 1


Figure 4-5.


Constrained RDA ordinations of pre-logged (1997) and first post-year
(1998) species data. First canonical axes are constrained by
TRT*YEAR interaction. Species highly correlated with first canonical
axes are displayed by vectors and codes (see Appendix C for species
names). Top plot displays RDA of 1000-m2 Species data; bottom plot
shows 1-m2 Species data. Note different scales of Axis 2.


Logged+fire















PINEL

SZIS-.1C RHOSP

POLR ,- ILEC.E



Logged+fire P-IJEr* Fire-only
C2 ::LPI.1 pI-1. EL





SCLPA


0.4 -



0.2 -


-0.4 t


0.4


ANCM


Logged+fire


Fire-only


-0.2


-0.4 -


-0.6


RHYEL


RHYGR


ANTRU


-1.0




Figure 4-6.


-0.5 0.0 0.5 1 .0
RDAAxis 1


Constrained RDA ordinations of ACP species data including 1997 pre-
treatment data. Top plot 1000-m2 data (excluding 1999 data), bottom
plot = 1-m2 Species data (all years). First canonical axes are
constrained by TRT*YEAR interaction indicated by treatment label in
red text. Species most highly correlated with first canonical axes are
displayed by red vectors and codes (see Appendix C). Vector length
and angles indicate strength and direction of correlation.














70

60

50

40

z30

20

10


1 10 100
Log (Area)




Figure 4-7. Number species per log sample area (m2). mean Species counts from
ACP treatments vs. Penfound and Lake Ramsey species richness.
Shaded regions show 95% confidence intervals of ACP treatment
means from 2005 samples; areas 1 2, 10, 20, and 100-m2. Light
shading = logged+fire; dark shading = fire-only. Dark line denotes
species numbers from Penfound "cut-over pineland" at areas 1 5, 10,
15, 20, and 25 m2. Red triangles show means and standard errors of
Lake Ramsey species richness (n 11) at areas 1 and 10-m2



















































Figure 4-8.


ACP vs. Lake Ramsey











High BA i CO





ACP vs. Penfou nd

High BA
Rne-CQpress









SLow BA






Successional trends of ACP species data compared to Penfound and
Lake Ramsey reference data (NMS ordination of presence-absence
data of comparable sample sizes see text). Red vectors depict
compositional shifts in the logged+fire treatment between 1997 and
2005; black vectors depicts shifts in fire-only treatment. Blue
symbols show relative positions of reference data in ordination
space. Bold dashed lines separate ACP samples with High and Low
initial BA.









CHAPTER 5
CONCLUSION

My dissertation presents a vegetation classification of pineland communities in Florida, a

model of the relationships between species composition and physical and spatial factors, and

description of temporal variation of pineland species composition in response to restoration of

fire regime and forest structure. The research revealed predictable spatial patterns of species

composition at both local and regional scales, and suggests resiliency of community composition

to temporary alterations to fire regime and timber density. Furthermore, the study of ecological

restoration suggests the resiliency of pineland vegetation, in that recovery approximating natural

conditions followed re-introduction of native fire regimes after a long period of fire suppression.

I presented a comprehensive vegetation classification based on floristic similarity, using

K-means cluster analysis and ordination methods. I recognized three ecological series

corresponding to idealized moisture conditions. These were further divided into 16 associations.

The series included Dry Uplands (6 associations), Mesic Flatwoods (3 associations), and

Wetlands (7 associations). Summary information described each community association relative

to species diversity, woody plant structure, diagnostic species, and environmental and

physiographic features. Floristic variation varied greatly with geographic segregation and

edaphic characteristics, particularly between the panhandle and peninsula regions of Florida.

Distinctions between community associations were related to the prominence of species with

restricted distributions, and to a lesser degree, endemic species. The floristic classification

presented here is comprehensive but applicable in the field, compared to other classification

based on regional flora.

The spatially explicit model of environmental-composition relationship revealed many of

the same patterns seen in the floristic classification. Most notably, the effects of geographic










separation were prominent in community structure, particularly the floristic distinctions between

the panhandle and peninsula. However, the effects of local environmental factors dominated

vegetation gradients, particularly variables related to topography and soil moisture. In the

context of all environmental factors, both regional and local environmental effects appear to

influence vegetation patterns. Effects of pure spatial structure were also evident, although these

made up a much smaller proportion of explained variance than environmental effects. Variation

associated with recent fire regime and timber stand structure may have contributed to

unexplained variance. I interpret the data as demonstrating a relatively strong control of

environmental factors on the distribution of pineland species, with biotic control mechanisms

and historical biogeography playing a lesser role.

The study of ground cover vegetation recovery following fire and stand structure

restoration suggests the resiliency of pineland community composition to atypical environmental

alterations. Temporal variation in pineland community structure is poorly understood relative to

spatial variation. However, longitudinal studies following restoration treatments shed light on

succession that is within the range of "natural" community stability. The study of ecological

restoration following a relatively long period of suppression of native fire regime demonstrates

that there is a certain degree of successional "stability" in pineland community structure.

Ecological restoration of a degraded Coastal Plain pineland remnant was successful in

terms of ground cover vegetation recovery that resembled reference site conditions. These

results suggest that recovery is possible on sites that have suffered moderate to severe fire

suppression but minimal ground disturbance (from cultivation). Ground cover species richness

was enhanced by overstory and midstory woody biomass reduction, mainly as increases in

detectable graminoid species. Furthermore, species richness and composition of restoration









treatments converged following two prescribed fires suggesting both the relative importance of

fire in ecological restoration and minimal adverse effects from mechanical logging. Pine

removal via carefully supervised mechanical logging does not appear to adversely affect savanna

vegetation recovery, and may expedite overall community restoration.

In sum, results of this observational and experimental research are relevant to

conservation and management of pyrogenic pineland communities of the Southeastern Coastal

Plain. The classification of Florida pineland communities based on compositional similarities

can be applied to inventory and restoration efforts in the State and surrounding regions. In

addition, the classification provides useful reference conditions for restoration, a basis for

quantifying regional variation in community variation, and specific diagnostic indicators for

community identification. The model of environmental-composition relationships contributes to

the overall understanding of environmental determinants of pineland vegetation, in addition to

quantifying ranges of environmental conditions correlated to compositional variation. The final

restoration study showed that pine savanna vegetation is resilient to moderate degradation related

to fire suppression, and can rebound following reconstruction of native stand structure. These

results add to the regional model of ecological restoration in the Southeast, specifically by

showing that restoration is achievable in wet pineland sites with increased alteration from fire

suppression.











APPENDIX A
LOCATIONS OF SAMPLE PLOTS AND SITES

Table A-1: Vegetation sample plots listed by code ("Plot"). Site code indicates Site containing
plots. The assigned community association ("Assoc") is indicated by code (see
Figure 2-3). Latitude and Longitude indicated in decimal degrees. Some specific
plot locations are not reported as per landowner request ("NR"). "Region" indicates
ecoregion delineation used for site stratification (see text). "Management Area"
indicates public land unit; private lands are noted. Abbreviations indicate
management unit types: SF = State Forest, AFB = Air Force Base, AFR = Air Force
Range, SP = State Park, WMA = Wildlife Management Area, NF = National Forest,
ARD = Apalachicola Ranger District, WRD = Wakulla Ranger District, NWFWMD
= Northwest Florida Water Management District, SWFWMD = Southwest Florida
Water Management District, DEP = Florida State Department of Environmental
Protection, NWR = National Wildlife Refuge, CA = Conservation Area (State), SRA
= State Recreation Area, TNC = The Nature Conservancy preserve.


Region
Northwest Uplands
Northwest Uplands
Northwest Uplands
Northwest Uplands
Northwest Uplands
Northwest Uplands
Northwest Uplands
Northwest Uplands
Northwest Uplands
Northwest Uplands
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
W Panhandle Gulf Coast
West Panhandle Sandhills
West Panhandle Sandhills


Management Area
Blackwater River SF
Blackwater River SF
Blackwater River SF
Blackwater River SF
Blackwater River SF
Blackwater River SF
Blackwater River SF
Blackwater River SF
Blackwater River SF
Blackwater River SF
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Pt. Washington SF
St. Joe Bufferlands (DEP)
St. Joe Bufferlands (DEP)
St. Joe Bufferlands (DEP)
St. Joe Bufferlands (DEP)
St. Joe Bufferlands (DEP)
Topsail SP
Topsail SP
Topsail SP
Topsail SP
Eglin AFB
Eglin AFB


Site
BWO1
BWO1
BWO1
BWO2
BWO3
BWO3
BWO3
BWO3
BWO1
BWO1
EGO8
EGO2
EGO2
EGO2
EGO2
PTO1
SJO1
SJO1
SJB01
SJO1
SJO1
TP01
TP01
TP01
TP01
EG01
EGO3


Plot
FLO92
FLO93
FLO94
FLO98
FLO99
FL100
FL275
FL276
FL284
FL285
FLO58
FL304
FLO41
FLO42
FLO46
FL280
FLO22
FL163
FL255
FL257
FL258
FLO77
FLO78
FLO79
FL293
FLO40
FLO43


Assoc
D4
W6
D6
D5
D4
W6
W7
W5
D5
D4
M2
M2
D4
W6
D4
D4
M2
W6
M1
W6
W6
M2
W6
M2
M2
D4
D4


Latitude
8639.023
8638.869
8638.835
8652.411
8656.497
8656.545
8656.577
8656.378
8639.132
8639.738
8625.442
8645.831
8645.827
8645.905
8646.216
8608.742
8517.839
8517.886
8517.558
8516.240
8515.401
8617.644
8617.739
8618.011
8617.609
8643.830
8647.440


Longitude
3054.668
3054.435
3054.417
3043.187
3050.387
3050.399
3050.434
3050.381
3054.441
3054.574
3027.510
3025.537
3025.486
3025.583
3025.085
3020.642
2942.753
2942.757
2942.183
2942.836
2943.102
3022.437
3022.383
3022.380
3022.307
3029.049
3027.046











Table A-1 continued


Region
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
West Panhandle Sandhills
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
Marianna Lowlands
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills


Management Area
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Eglin AFB
Apalachee WMA
Apalachee WMA
Apalachee WMA
Falling Waters SP
Falling Waters SP
Falling Waters SP
Rock Hill TNC
Rock Hill TNC
Rock Hill TNC
Rock Hill TNC
Rock Hill TNC
Rock Hill TNC
Three Rivers SP
Three Rivers SP
Apalachicola Bluffs TNC
Apalachicola Bluffs TNC
Apalachicola Bluffs TNC
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Econfina River NWFWMD


Site Plot Assoc Latitude Longitude
EGO3 FLO44 D4 8646.915 3027.165
EGO3 FLO45 W4 8646.786 3027.083
EGO4 FLO47 D4 8613.336 3036.250
EGO5 FLO48 D4 8611.709 3037.040
EGO5 FLO49 D6 8611.971 3036.542
EGO6 FLO50 D4 8651.788 3030.834
EGO7 FLO51 D4 8646.726 3027.923
EGO7 FLO52 D4 8646.742 3028.417
EGO7 FLO53 D4 8646.616 3028.509
EGO4 FLO54 M2 8613.419 3036.190
EGO4 FLO55 D6 8612.760 3035.829
EGO6 FLO56 D4 8651.927 3031.046
EGO7 FLO57 D4 8646.748 3028.740
EGO4 FLO95 W6 8613.417 3036.402
EGO5 FLO96 W7 8611.918 3036.500
EGO4 FLO97 W7 8613.460 3036.581
AP0 1 FLO76 D5 8457.250 3047.156
AP0 1 FLO90 D5 8457.504 3048.523
AP0 1 FLO91 D5 8457.343 3048.404
FWO 1 FLO74 D6 8531.682 3043.656
FWO 1 FLO75 D6 8531.796 3043.587
FWO 1 FLO85 D5 8531.530 3043.739
RHO 1 FLO60 D6 8529.065 3044.275
RHO 1 FLO65 W6 8529.151 3044.134
RHO 1 FLO89 D6 8529.375 3044.358
RHO 1 FL287 W5 8530.054 3044.665
RHO 1 FL288 W6 8530.190 3044.662
RHO 1 FL292 D5 8529.503 3044.369
TR0 1 FLO70 D5 8455.295 3044.163
TR0 1 FLO71 D5 8455.173 3044.162
AB01 FLO59 D4 8458.419 3027.440
ABO2 FLO61 D4 8458.438 3028.383
ABO2 FLO66 D4 8458.303 3028.538
WD0 1 FL0 11 D2 8415.779 3021.186
WD0 1 FL0 12 D2 8415.845 3021.163
WDO3 FL0 18 D2 8420.988 3019.288
WDO3 FL0 19 D2 8421.027 3019.361
WD01 FL229 W1 8416.194 3021.513
WDO3 FL230 W1 8421.040 3019.371
WD0 1 FL259 D2 8416.217 3021.525
WDO3 FL260 D2 8420.988 3019.333
ER0 1 FL269 D4 8532.023 3029.139











Table A-1 continued

Region
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
East Panhandle Sandhills
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Apalachicola Lowlands
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Tallahassee Red Hills
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands


Management Area
Econfina River NWFWMD
Econfina River NWFWMD
Private land
Private land
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Apalachicola NF: ARD
Pebble Hill (Private land)
Pebble Hill (Private land)
Pebble Hill (Private land)
Pebble Hill (Private land)
Private land
Private land
Private land
Private land
Private land
Private land
Torreya SP
Torreya SP
Wade Tract (Private land)
Wade Tract (Private land)
Wade Tract (Private land)
Wade Tract (Private land)
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD


Site Plot Assoc Latitude Longitude
ER0 1 FL270 D4 8531.764 3028.899
ER0 1 FL294 W1 8534.151 3027.401
DYO 1 FLO72 D4 8457.033 3026.435
DYO 1 FLO73 D6 8457.070 3026.424
AD0 1 FLO62 W6 8500.932 3003.534
ADO2 FLO63 D4 8459.533 3016.618
ADO2 FLO64 D6 8459.107 3016.757
AD0 1 FLO67 D6 8500.818 3003.595
ADO3 FLO68 D4 8458.989 3012.292
ADO3 FLO69 D6 8459.224 3012.360
ADO4 FLO80 D6 8505.324 3006.344
ADO4 FLO81 D6 8505.157 3006.033
ADO3 FLO82 W6 8459.254 3012.255
ADO2 FLO83 D6 8458.749 3016.388
ADO5 FLO84 D6 8501.290 3011.725
ADO5 FLO86 W6 8501.305 3011.662
ADO5 FLO87 D6 8500.861 3011.980
AD0 1 FLO88 W6 8500.194 3003.788
APO6 FL162 W7 8457.638 3002.444
APO6 FL164 M2 8457.496 3002.541
APO6 FL265 D6 8458.099 3002.054
AD0 1 FL266 D6 8500.904 3003.299
PHO 1 FL227 W1 8405.221 3045.918
PHO 1 FL228 W5 8405.213 3045.904
PHO 1 FL289 D5 8405.342 3045.849
PHO 1 FL290 D5 8405.508 3046.390
AV0 1 FLO28 D2 NR NR
AV0 1 FLO29 W5 NR NR
BE0 1 FLO31 W5 NR NR
BEO2 FLO37 D5 NR NR
BEO2 FLO3 8 W5 NR NR
BE0 1 FLO3 9 W4 NR NR
TYO 1 FLO30 D4 8457.027 3033.462
TYO 1 FLO36 D4 8457.129 3033.376
WTO 1 FL224 D5 8359.851 3045.724
WTO 1 FL226 W5 8359.742 3045.752
WTO 1 FL286 D5 8359.978 3045.688
WTO 1 FL291 W1 8400.672 3045.552
WDO2 FL0 17 D4 8429.542 3020.029
WDO4 FLO20 D4 8441.153 3015.381
WDO4 FLO21 M2 8441.207 3015.161
WDO5 FLO25 D4 8432.368 3016.952











Table A-1 continued


Region
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
Wakulla Lowlands
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
E Panhandle Gulf Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Coast
Big Bend Interior Lowlands
Big Bend Interior Lowlands
Big Bend Interior Lowlands
Big Bend Interior Lowlands
Big Bend Interior Lowlands
Big Bend Interior Lowlands


Management Area
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
Apalachicola NF: WRD
St. Joe Bufferlands (DEP)
St. Joe Bufferlands (DEP)
St. Joe Bufferlands (DEP)
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
St. Marks NWR: Panacea
Cedar Key Scrub SP
Cedar Key Scrub SP
Cedar Key Scrub SP
Cedar Key Scrub SP
Cedar Key Scrub SP
Lower Suwannee NWR
Lower Suwannee NWR
Lower Suwannee NWR
Lower Suwannee NWR
St. Marks NWR: St. Marks
St. Marks NWR: St. Marks
St. Marks NWR: St. Marks
Geothe SF
Goethe SF
Goethe SF
Goethe SF
Goethe SF
Goethe SF


Site Plot Assoc Latitude Longitude
WDO5 FLO26 D4 8432.287 3016.911
WDO6 FLO27 W4 8433.080 3013.853
WDO7 FLO32 D6 8433.938 3022.967
WDO7 FLO33 M2 8434.031 3023.016
WDO6 FLO34 M2 8433.130 3013.746
WDO5 FLO35 M2 8432.176 3013.861
WDO7 FL261 W4 8433.715 3022.721
WDO7 FL262 W4 8433.898 3022.858
WDO7 FL263 D6 8431.724 3021.255
WDO5 FL267 W4 8430.400 3017.106
WDO5 FL268 W4 8430.493 3017.197
SJO2 FL165 M1 8452.457 2946.138
SJO2 FL166 W7 8452.460 2946.223
SJO2 FL167 M2 8452.525 2946.190
SMO 1 FL00 1 D2 8425.094 3002.987
SMO 1 FLOO2 D6 8427.700 3002.281
SMO1 FLOO3 M2 8427.739 3002.233
SMO 1 FLOO4 W6 8427.761 3002.230
SMO 1 FLOO5 D2 8425.101 3002.886
SMO3 FLOO6 D2 8426.210 3002.846
SMO3 FLOO7 M2 8426.161 3002.904
SMO3 FLOO8 M2 8426.110 3002.889
SMO4 FLOO9 D6 8429.020 3002.591
SMO4 FL305 W6 8429.047 3002.529
CK0 1 FL199 M1 8259.824 2912.314
CK0 1 FL200 W3 8259.754 2912.307
CK0 1 FL201 M1 8259.680 2912.317
CK0 1 FL212 M1 8301.709 2912.251
CK0 1 FL295 M1 8258.871 2912.365
LSO1 FL238 M2 8301.263 2927.417
LSO1 FL239 M1 8301.153 2927.433
LSO2 FL240 M1 8311.944 2923.701
LSO2 FL241 M1 8311.969 2923.713
SMO6 FLO23 M2 8405.373 3007.873
SMO6 FLO24 W3 8405.479 3007.913
SMO6 FL283 M2 8409.138 3009.513
GOO4 FL299 M1 8237.265 2918.862
GOO2 FL119 M1 8235.430 2907.339
GOO2 FL120 M1 8236.329 2908.709
GOO2 FL121 W4 8236.277 2908.756
GOO3 FL122 W4 8236.443 2911.963
GOO3 FL123 M1 8237.461 2914.514











Table A-1 continued


Region
Big Bend Interior Lowlands
Big Bend Interior Lowlands
Big Bend Interior Lowlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
North Central Highlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
Interior Northeast Lowlands
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast
North Atlantic Coast


Management Area
St. Marks NWR: Wakulla
St. Marks NWR: Wakulla
St. Marks NWR: Wakulla
Ichetucknee SP
Manatee SP
Manatee SP
Manatee SP
Oleno SP
Oleno SP
Oleno SP
Oleno SP
Oleno SP
Oleno SP
River Rise SP
Twin Rivers SF
Twin Rivers SF
Twin Rivers SF
Twin Rivers SF
Twin Rivers SF
Twin Rivers SF
Twin Rivers SF
Jennings SF
Jennings SF
Jennings SF
Osceola NF
Osceola NF
Osceola NF
Osceola NF
Osceola NF
Osceola NF
Osceola NF
Favor Dykes SP
Favor Dykes SP
Favor Dykes SP
Heart Island CA
Heart Island CA
Heart Island CA
private land
Pumpkin Hill SP
Pumpkin Hill SP
Pumpkin Hill SP
Tigar Bay SF


Site Plot Assoc Latitude Longitude
SMO5 FL0 14 D2 8418.119 3007.903
SMO5 FL0 15 M2 8418.126 3007.803
SMO5 FL0 16 W4 8418.356 3007.606
IC0 1 FL168 D2 8246.013 2958.323
MA0 1 FL185 D1 8257.606 2929.961
MA0 1 FL186 D1 8257.919 2929.953
MA0 1 FL187 D1 8257.910 2929.422
OL0 1 FL148 D2 8234.150 2954.866
OL0 1 FL149 M1 8234.195 2954.837
OLO2 FL150 M1 8234.493 2954.978
OLO2 FL151 D1 8235.095 2955.080
OL0 1 FL235 M2 8234.198 2954.432
OL0 1 FL236 W1 8234.215 2954.623
RR0 1 FL161 D1 8238.055 2952.208
TR0 1 FL126 D2 8311.814 3029.308
TR0 1 FL127 D2 83 12.048 3029.460
TR0 1 FL128 D2 83 12.734 3030.330
TR0 1 FL129 D2 83 12.458 3030.175
TRO2 FL252 D2 8312.337 3022.682
TRO2 FL253 D1 83 12.452 3022.386
TRO2 FL254 D1 8312.361 3022.399
JE01 FL177 M2 8156.067 3010.410
JE01 FL178 W2 8156.087 3010.380
JE01 FL179 D2 8156.160 3010.816
OSO1 FL101 M2 8224.710 3014.328
OSO1 FL102 M2 8224.582 3014.134
OSO1 FL103 M2 8224.743 3014.261
OSO2 FL242 W4 8226.883 3011.485
OSO2 FL244 M2 8226.528 3011.540
OSO2 FL245 M1 8226.420 3011.883
OSO3 FL247 W4 8229.095 3017.087
FD0 1 FL309 M1 8116.487 2940.420
FD0 1 FL310 M1 8115.807 2940.653
FD0 1 FL313 W2 8116.588 2940.331
HIO 1 FL296 M1 8121.280 2911.812
HIO 1 FL297 W3 8123.827 2911.175
HIO 1 FL298 M3 8124.455 2910.867
HIO 1 FL301 D3 NR NR
PP0 1 FL311 M2 8116.559 2940.512
PP0 1 FL3 12 M1 8130.126 3028.824
PP0 1 FL315 W4 8130.536 3028.344
TB01 FL300 D3 8111.349 2914.129











Table A-1 continued


Region
North Atlantic Coast
North Atlantic Coast
Coastal Northeast Lowlands
Coastal Northeast Lowlands
Coastal Northeast Lowlands
Coastal Northeast Lowlands
Coastal Northeast Lowlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
West Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands


Management Area
Tigar Bay SF
Tigar Bay SF
Jennings SF
Jennings SF
Sinunons SF
Sinunons SF
Sinunons SF
Ashton (Private land)
Ashton (Private land)
Cross Florida Greenway SRA
Cross Florida Greenway SRA
Davidson Ranch TNC
Davidson Ranch TNC
Davidson Ranch TNC
Goethe SF
Goethe SF
Goethe SF
Goethe SF
Ross Prairie SF
Ross Prairie SF
Ross Prairie SF
San Felasco SP
San Felasco SP
San Felasco SP
San Felasco SP
Etoniah Creek SF
Etoniah Creek SF
Etoniah Creek SF
Etoniah Creek SF
Goldhead SP
Goldhead SP
Ocala NF
Ocala NF
Ocala NF
Ocala NF
Ocala NF
Ocala NF
Ocala NF
Ocala NF
Rock Springs SP
Rock Springs SP
Swisher-Ordway: Univ of FL


Site Plot Assoc Latitude Longitude
TB01 FL302 M3 8110.758 2913.184
TB01 FL303 M1 8110.287 2912.531
JEO2 FL180 M2 8155.906 3010.360
JEO2 FL181 D2 8155.888 3009.956
SIO 1 FL182 M2 8156.597 3047.671
SIO 1 FL183 D2 8156.999 3047.972
SIO 1 FL184 W4 8157.434 3046.913
ASO1 FL124 M1 8235.041 2932.284
ASO 1 FL125 D3 8234.781 2932.527
GWO 1 FL112 D3 8215.520 2903.090
GWO 1 FL113 D2 8215.197 2902.864
DR0 1 FL106 D3 8241.860 2944.586
DR0 1 FL107 D2 8241.710 2945.139
DR0 1 FL116 D1 8241.925 2944.969
GOO1 FL104 D3 8236.197 2921.450
GOO1 FL105 D2 8236.187 2921.480
GOO1 FL233 D3 8236.138 2921.452
GOO1 FL234 W1 8236.017 2921.580
RP0 1 FL114 D3 8217.955 2901.933
RP0 1 FL115 D2 8217.802 2901.821
RP0 1 FL232 D3 8217.527 2902.185
SF0 1 FL130 D3 8228.090 2944.222
SF0 1 FL131 D1 8227.762 2944.601
SF0 1 FL132 D1 8226.901 2943.903
SFO2 FL133 M2 8226.743 2942.961
ETO 1 FL117 D3 8152.401 2947.222
ETO 1 FL118 D3 8152.077 2947.064
ETO2 FL306 M1 8147.309 2943.782
ETO2 FL314 W2 8120.974 2737.601
GHO 1 FL175 D3 8157.421 2950.975
GHO 1 FL176 W1 8156.427 2949.631
OC0 1 FL139 D3 8148.342 2927.505
OC0 1 FL142 D2 8148.587 2927.381
OC0 1 FL144 D3 8148.596 2947.468
OC0 1 FL145 D3 8149.589 2927.472
OCO2 FL146 W4 8156.602 2909.451
OCO2 FL147 M1 8156.316 2909.687
OK0 1 FL307 D2 8154.813 2910.331
OK0 1 FL308 M3 8156.729 2909.302
RSO 1 FL206 M1 8127.770 2845.729
RSO 1 FL207 M1 8127.623 2845.946
OR0 1 FL152 D3 8159.907 2940.846











Table A-1 continued
Region
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
East Central Highlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
West Central Lowlands
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Kissimmee Basin
Southwest Central Lowlands
Southwest Central Lowlands


Management Area
Swisher-Ordway: Univ of FL
Swisher-Ordway: Univ of FL
Wekiwa Springs SP
Wekiwa Springs SP
Wekiwa Springs SP
Wekiwa Springs SP
Green Swamp WMA
Green Swamp WMA
Green Swamp WMA
Green Swamp WMA
Green Swamp WMA
Green Swamp WMA
Green Swamp WMA
Starkey Wildemess SWFWMD
Starkey Wildemess SWFWMD
Starkey Wildemess SWFWMD
Starkey Wildemess SWFWMD
Starkey Wildemess SWFWMD
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Avon Park AFR
Disney Wildemness TNC
Disney Wildemness TNC
Disney Wildemness TNC
Disney Wildemness TNC
Kississimee Prairie Preserve SP
Kississimee Prairie Preserve SP
Three Lakes WMA
Three Lakes WMA
Three Lakes WMA
Three Lakes WMA
Myakka River SP
Myakka River SP


Site
OR01
OR01
WE 1
WE 1
WE 1
WE 1
GSO1
GSO2
GSO2
GSO2
GSO2
GSO2
GSO1
SWO1
SWO1
SWO1
SWO2
SWO1
AV01
AV01
AV01
AV01
AVO2
AVO3
AVO2
AVO3
AVO3
AVO4
AVO5
AVO6
DWO1
DWO1
DWO1
DWO1
KIO1
KIO1
TL01
TL01
TL01
TL01
MYO 1
MYO2


Plot
FL153
FL154
FL135
FL137
FL141
FL143
FL155
FL156
FL157
FL158
FL159
FL160
FL174
FL169
FL170
FL171
FL172
FL173
FL188
FL189
FL190
FL191
FL192
FL193
FL196
FL197
FL198
FL219
FL220
FL221
FL134
FL136
FL138
FL140
FL194
FL195
FL202
FL203
FL204
FL205
FL208
FL209


Assoc
D3
W1
D3
M1
D3
M1
M3
D2
M1
W1
W1
D3
W2
M1
W2
M3
D3
W2
M3
W2
M3
M3
M1
W3
M3
W2
M3
M3
M3
M3
M1
M3
M3
M1
M3
W2
M3
W2
W2
M1
M1
M1


Latitude
8159.764
8200.641
8129.779
8129.866
8129.604
8129.792
8457.196
8208.258
8207.163
8207.115
8207.274
8207.151
8157.163
8235.709
8235.867
8235.840
8236.488
8235.853
8117.614
8117.347
8117.231
8117.292
8119.719
8112.519
8119.757
8112.691
8112.664
8112.629
8115.758
8119.176
8124.340
8124.281
8124.302
8124.378
8108.116
8106.414
8105.123
8105.205
8105.241
8104.352
8210.671
8217.429


Longitude
2940.761
2940.532
2844.307
2844.789
2843.923
2844.467
2822.252
2826.337
2826.372
2826.402
2826.118
2826.273
2821.828
2814.806
2814.418
2814.531
2815.010
2814.422
2740.173
2940.170
2740.268
2740.290
2744.531
2742.114
2744.752
2742.545
2742.536
2737.055
2743.773
2738.357
2804.012
2804.050
2804.020
2804.009
2735.082
2735.298
2758.670
2758.715
2758.735
2758.957
2714.092
2717.423











Table A-1 continued
Region
Southwest Central Lowlands
Southwest Central Lowlands
Southwest Central Lowlands
Southwest Central Lowlands
Southwest Central Lowlands
Southwest Highlands
Southwest Highlands
Southwest Highlands
Southwest Highlands
Caloosahatchie Lowlands
Caloosahatchie Lowlands
Caloosahatchie Lowlands
Caloosahatchie Lowlands
Caloosahatchie Lowlands


Management Area
Myakka River SP
Myakka River SP
Myakka River SP
Myakka River SP
Myakka River SP
Withlacoochee SF
Withlacoochee SF
Withlacoochee SF
Withlacoochee SF
Cecil Webb WMA
Cecil Webb WMA
Cecil Webb WMA
Cecil Webb WMA
Cecil Webb WMA


Site Plot Assoc Latitude Longitude


MYO3
MYO3
MYO4
MYO4
MYO3
WF 1
WF 1
WFO2
WFO2
CWO 1
CWO 1
CWO 1
CWO2
CWO2


FL210
FL211
FL213
FL214
FL215
FL108
FL109
FL110
FL111
FL216
FL217
FL218
FL222
FL223


8214.965
8214.926
8213.196
8212.229
8215.152
8225.963
8225.653
8224.071
8223.958
8156.349
8156.456
8155.595
8151.861
8151.948


2716.027
2716.119
2713.914
2713.572
2715.225
2847.513
2848.063
2843.674
2843.519
2652.335
2652.301
2653.341
2651.547
2651.597










APPENDIX B
LIST OF FREQUENT AND ABUNDANT SPECIES BY COMMUNITY ASSOCIATION

Table B-1: Species included that present in > 70-75 % of plots within an association, and > 0.2
m2 mean COVer. Freq percent frequency of occurrence, cover mean cover in m2
Number of plots per association indicated in parentheses


Peninsula Xeric Sandhills (22) Freq Cover
4ristida bevrichiana 100 39.80
Sorghastrunt secundum 100 2.99
Pitvopsis grantinifolia 100 2.52
Lechea ., I, .. ,, 100 0.44
Schizachyrium scoparium var. stoloniferunt 95 1.28
Dichanthelium ovale var. addisonii 95 0.80
stri,, .;,l sylvatica 95 0.72
Sporobolus junceus 91 1.14
Paspalunt setaceum 91 0.51
Cnidoscolus stimulosus 91 0.28
Bulbostylis ciliatifolia 86 0.84
4ndropogon ternarius 86 0.76
Smilax auriculata 86 0.59
Rhynchospora gravi 86 0.33
Tragia urens 86 0.33
Crotalaria rotundifolia 86 0.30
Balduina <,,t..-,Istr, r;le<, 82 0.50
4ndropogon gyrans var. gvrans 82 0.36
Tephrosia chrysophylla 77 1.11
Croton (I, .-I ,Iri,,. itie 77 0.48
Liatris tenuifolia var. tenuifolia 77 0.40
Scleria ciliata var. ciliata 77 0.30

Panhandle Xeric Sandhills (31) Freq Cover
Schizachyrium scoparium var. stoloniferum 100 4.38
Smilax auriculata 100 1.66
4ndropogon gyrans var. gvrans 100 1.26
Stylisina patens ssp. patens 100 0.45
Stylosanthes bi flora 94 0.3 1
Pitvopsis aspera 90 3.29
Bulbostylis ciliatifolia 90 1.16
Cyperus lupulinus ssp. hipulinus 90 0.38
Galactia inicrophylla 87 2.38
Sorghastruin secunduin 87 1.91
Eriogonuin toinentosuin 87 1.04
Rhynchospora gravi 87 0.58
Dichantheliuin cl,,-.-lr ;,s,Cle, a 87 0.52
4ndropogon virginicus 84 2.23
Solidago odora var. odora 84 1.56
Scleria ciliata var. ciliata 84 0.54
Conunelina erecta 84 0.35











Table B-1 continued
Panhandle Xeric Sandhills (continued) Freq Cover
Dichanthelium ovale var. addisonii 81 0.88
Sporobolus junceus 81 0.77
4ristida bevrichiana 77 19.06
Schizachvrium teneruin 77 1.19
Tragia urens 77 0.28
Croton (I, .-I ,Iri,,. itie 74 0.75
Liatris tenuifolia var. tenuifolia 74 0.36


North Florida Sandhills (31) Freq Cover
4ristida bevrichiana 100 33.06
Pitvopsis graininifolia 97 3.50
Dichanthelium ovale var. addisonii 97 0.96
Paspaluin setaceuin 97 0.79
Scleria ciliata var. ciliata 97 0.65
Tragia urens 97 0.42
Sorghastruin secunduin 94 2.73
Schizachyrium scoparium var. stoloniferuin 94 1.46
\r, ile .;es sylvatica 94 0.64
Rhynchosia reniforinis 94 0.44
Heliantheinui carolinianuin 94 0.30
Dichantheliuin cl,-.-;s,Clet, a;il, 90 0.72
Stylisina patens ssp. patens 90 0.33
Crotalaria rotundifolia 87 0.35
Dyschoriste oblongifolia 84 1.17
4nd'ropogon gvrans var. gvrans 84 0.92
Eupatoriuin compositifolium 84 0.59
Gvinnopogon ambiguus 84 0.45
Croton (I, .-I ,Irit.. ities 84 0.44
Rhynchospora gravi 84 0.38
Sporobolus junceus 81 0.89
Lechea ,ti Ir. 81 0.73
4nd'ropogon ternarius 81 0.63
Vernonia cl,,-I.- ;,s,Cle< 81 0.55
Sericocarpus tortifolius 81 0.44
Liatris tenuifolia var. tenuifolia 81 0.40
Swinphyotrichum concolor 81 0.36
Stylosanthes bi flora 8 1 0.21
Smilax auriculata 77 1.44
Lesped'eza hirta 77 0.39
Palafoxia integrifolia 77 0.38
Ruellia caroliniensis ssp. ciliosa 77 0.31
4ristolochia serpentaria 77 0.28
Hieracium gronovil 77 0.26
Elephantopus elatus 74 2.04
Solidago odora var. odora 74 0.50











Table B-1 continued
North Florida Sandhills (continued) Freq Cover
4ristida purpurascens var. purpurascens 74 0.37
4nd'ropogon virginicus 74 0.29
1;.i~ <,te Ga, arontatica 74 0.29

North Florida Rich Woodlands (11) Freq Cover
Pteridium aquilinum 100 4.72
Sorghastrunt secundum 100 4.00
1;.i~ <,te Ga, arontatica 100 1.24
Dichanthelium cl,-.-;s,Clet, a;il, 100 0.97
Paspalunt setaceum 100 0.75
4nd'ropogon gyrans var. gvrans 100 0.39
Smilax auriculata 100 0.36
Dichanthelium ovale var. addisonii 91 1.11
Scleria ciliata var. ciliata 9 1 0.77
Sericocarpus tortifolius 91 0.75
Houstonia procumbens 91 0.48
4nd'ropogon virginicus 91 0.40
4ristolochia serpentaria 82 0.3 1
Galium piloston 82 0.31
Cyperus plukenetil 82 0.27
Hypericuin hypericoides 82 0.18
Dvschoriste oblongifolia 73 1.61
Eupatoriuin compositifolium 73 1.49
Dichanthelium in e.:.s unit. var. Je-;.... ,li, 73 1.09
Dichanthelium aciculare 73 0.98
Pitvopsis grantinifolia 73 0.93
\r, ile .;es sylvatica 73 0.47
4ristida purpurascens var. purpurascens 73 0.42
4nd'ropogon ternarius 73 0.39
Heliantheinui carolinianton 73 0.38
Cyperus retrorsus 73 0.3 1
Rhynchosia reniforinis 73 0.3 1
Crotalaria rotundifolia 73 0.26
Hieracium gronovil 73 0.24

Panhandle Longleaf Pine Clavhills (14) Freq Cover
4ristida bevrichiana 100 22.72
Schizachyrium scoparium var. stoloniferuin 100 8.75
Solidago odora var. odora 100 4.46
Dichanthelium Elephantopus elatus 100 1.59
Vernonia cl,,-I.- ;,s,Cle< 100 0.82
Stylosanthes bi flora 100 0.50
Sericocarpus tortifolius 93 1.66
4nd'ropogon gvrans var. gvrans 93 1.13
Swinphyotrichum duinosuin var. duinosuin 93 1.09
Desinodium lineatuin 93 1.04











Table B-1 continued
Panhandle Longleaf Pine Clavhills (continued) Freq Cover
Lespedeza repens 93 0.38
Hieracium gronovil 93 0.29
Pteridium aquilinum 86 5.65
Rubus cuneifolius 86 1.19
Desmodium ciliare 86 1.12
Dichanthelium ovale var. addisonii 86 1.05
Ahthlenbergia capillaris var. trichopodes 86 1.00
Scleria ciliata var. ciliata 86 0.86
4ristida purpurascens var. purpurascens 86 0.73
4nd'ropogon virginicus 86 0.52
Synphyotrichunt ad'natum 86 0.47
Synphyotrichunt concolor 86 0.44
Mimosa microphylla 86 0.40
Eupatorium compositifolium 86 0.38
Liatris gracilis 86 0.34
Rudbeckia hirta 86 0.29
Pitvopsis grantinifolia 79 4.64
Schizachvrium tenerunt 79 2.27
Pitvopsis aspera 79 1.37
,;.. ,,r,, Ga arontatica 79 0.95
Chaniaecrista nictitans 79 0.78
Euphorbia discoidalis 79 0.58
(I, / gas. '" mariana 79 0.46
Smilax glauca 79 0.40
4calypha gracilens 79 0.27
4ristolochia serpentaria 79 0.27
Houstonia procumbens 79 0.26
Gvinnopogon ambiguus 79 0.23

Panhandle Silty Woodlands (22) Freq Cover
4ristida bevrichiana 100 33.64
Schizachyrium scoparium var. stoloniferuin 100 3.28
Dichanthelium dichotoinui var. tenue 100 1.15
Scleria ciliata var. ciliata 100 0.75
Synphyotrichuin ad'natuin 100 0.65
Dichanthelium el,,.-,Ier r;ile,,, 95 1.79
Tragia sinallii 95 0.47
Pitvopsis grantinifolia 91 5.36
4nd'ropogon gyrans var. gvrans 91 1.07
Sericocarpus tortifolius 91 0.73
Stylosanthes bi flora 91 0.3
(I, / gas. '" mariana 86 0.52
4nd'ropogon virginicus 86 0.41
Smilax auriculata 82 1.38
Lespedeza repens 82 0.27
Viola septeinloba 82 0.25
Pteridium aquilinuin 77 8.8










Table B-1 continued
Panhandle Silty Woodlands (continued) Freq Cover
Helianthus radula 77 3.11
Carphephorus odoratissimus 77 1.01
Galactia erecta 77 0.27


Xeric Mesic Flatwoods (36) Freq Cover


Aristida beyrichiana 92 18.24
Dichanthelium sabulorum var. thinium 86 0.75
Andropogon virginicus 86 0.69
Pityopsis graminifolia 83 0.85
Smilax auriculata 81 0.51
Pterocaulon virgatum 78 0.42
Bulbostylis ciliatifolia 75 0.67
Gratiola hispida 72 0.43

North Florida Mesic Flatwoods (30) Freq Cover
Aristida beyrichiana 97 14.93
Xyris caroliniana 93 0.79
Pityopsis graminifolia 87 1.64
Andropogon virginicus 87 1.07
Dichanthelium strigosum var. leucoblepharis 77 0.70
Pterocaulon virgatum 77 0.3 1
Pteridium aquilinum 73 3.05
Sericocarpus tortifolius 73 0.21
Dichanthelium sabulorum var. thinium 73 0.70


Central Florida Mesic Flatwoods/Dry Prairies (22) Freq Cover
Aristida beyrichiana 100 27.23
Dichanthelium sabulorum var. thinium 100 0.85
Andropogon virginicus 95 1.63
Pterocaulon virgatum 95 0.47
Aristida spiciformis 91 2.06
Pityopsis graminifolia 91 0.82
Dichanthelium chamaelonche 86 9.49
Xyris caroliniana 86 0.40
Paspalum setaceum 86 0.32
Drosera brevifolia 86 0.30
Polygala setacea 86 0.28
Euthamia tenuifolia var. tenuifolia 82 0.81
Oldenlandia uniflora 82 0.35
Fimbristylis puberula 82 0.30
Eleocharis baldwinii 77 0.67
Gratiola hispida 77 0.33


Marginal Priaries (11) Freq Cover
Andropogon virginicus 91 10.23
Euthamia tenuifolia var. tenuifolia 91 0.80
Panicum hemitomon 82 12.35






















































100
100
100
100
100
100
100
100
100
100
75
75
75
75
75
75
75
75


4.75
2.09
1.50
0.63
0.34
0.34
0.34
0.28
0.22
0.09
5.3 1
3.3 1
2.44
2.13
2.00
1.75
1.13
1.00


Peninsula Wet Flatwoods/Prairies (16) Freq Cover
Aristida palustris 75 6.44
Scleria muehlenbergii 75 5.38
Andropogon capillipes (wetland varient) 75 4.86
Fuirena scirpoidea 75 2.63
Panicum tenerum 75 1.98


Freq Cover


Table B-1 continued
Marginal Priaries (continued)
Rhexia mariana var. mariana
Axonopus furcatus
Eupatorium leptophyllum
Centella erecta
Andropogon capillipes (wetland varient)
Aristida purpurascens var. virgata

Peninsula Wet Flatwoods/Prairies (16)
Oxypolis fihiformis
Eriocaulon decangulare
Bigelowia nudata
F, ag.- u elliottii
Amphicarpum muchlenbergianum
Xyris elliottii
Andropogon gyrans var. stenophyllus
Aristida beyrichiana
Dichanthelium erectifolium
Centella erecta
Drosera brevifolia
Eupatorium mohrii


Fre
82
73
73
73
73
73


Fre
100
94
94
94
88
88
88
8 1
81
81
81
81


Cover
0.75
15.27
2.84
2.72
1.23
0.60


Cover
1.73
5.87
2.16
1.11
5.95
3.68
0.66
15.24
2.84
1.41
0.40
0.32


Calcareous Wet Flatwoods (4)
Sabal palmetto
Centella erecta
Hyptis alata
Saccharum giganteum
Helenium pinnatifidum
Lobelia glandulosa
Panicum rigidulum var. rigidulum
Rhynchospora globularis
Cirsium nuttallii
Asclepias lanceolata
Dichanthelium dichotomum var. nitidum
Panicum virgatum var. virgatum
Rhynchospora divergens
Dichanthelium caerulescens
Pluchea rosea
Ludwigia microcarpa
Rhynchospora colorata
Fuirena breviseta










Table B-1 continued
Calcareous Wet Flatwoods (continued) Freq Cover
Mikania scandens 75 0.97
Rubus trivialis 75 0.88
Scleria muehlenbergii 75 0.78
Hypericum cistifolium 75 0.75
Diodia virginiana 75 0.72
Rhynchospora perplexa 75 0.59
Dichanthelium strigosum var. glabrescens 75 0.53
Hypericum hypericoides 75 0.53
Andropogon capillipes (upland varient) 75 0.50
Berchemia scandens 75 0.50
Eustachys glauca 75 0.44
Scleria pauciflora 75 0.44
Phyla nodiflora 75 0.41
Cyperus polystachyos 75 0.38
Oxypolis fihformis 75 0.31
Proserpinaca pectinata 75 0.31
Smilax laurifolia 75 0.31
Andropogon glomeratus var. glomeratus 75 0.28
Mitreola petiolata 75 0.28
Toxicodendron radicans 75 0.28
Axonopus furcatus 75 0.25
Eleocharis flavescens 75 0.25
Mitreola sessihifolia 75 0.25
Vitis rotundifolia 75 0.25
Xyris jupicai 75 0.25

North Florida Shrubby Wet Flatwoods (15) Freq Cover
Andropogon glaucopsis 87 5.20
Osmunda cinnamomea 87 4.38
Eriocaulon decangulare 80 7.57
Smilax laurifolia 80 0.66
Xyris ambigua 80 0.39
Photinia pyrifolia 80 0.34
Andropogon glomeratus var. hirsutior 73 2.27
Rhynchospora fascicularis 73 2.09
Andropogon capillipes (upland varient) 73 0.43
Rhexia petiolata 73 0.23


Upper Panhandle Wet Flatwoods (7) Freq Cover
Schizachyrium scoparium var. stoloniferum 100 5.32
Pityopsis graminifolia 100 5.11
Pteridium aquilinum 100 4.68
Eupatorium rotundifolium 100 4.21
Andropogon virginicus 100 4.14
Panicum verrucosum 100 3.52
Rhexia alifanus 100 2.43
Helianthus to;-,Iar, ;ill,* 100 1.64










Table B-1 continued
Upper Panhandle Wet Flatwoods (continued) Freq Cover
Euthamia tenuifolia var. tenuifolia 100 1.32
Symphyotrichum dumosum var. dumosum 100 1.09
Smilax glauca 100 0.95
Solidago stricta 100 0.93
Diodia virginiana 100 0.50
Xyris caroliniana 100 0.48
Chamaecrista nictitans 100 0.34
Ctenium aromaticum 86 10.54
Dichanthelium dichotomum var. tenue 86 3.11
Andropogon glomeratus var. hirsutior 86 2.82
Aristida purpurascens var. virgata 86 2.32


Upper Panhandle Wet Flatwoods (7) Freq Cover
Panicum anceps var. rhizomatum 86 1.88
Chaptalia tomentosa 86 1.73
Dichanthelium strigosum var. leucoblepharis 86 1.59
Panicum virgatum var. virgatum 86 0.70
Desmodium tenuifolium 86 0.64
Bigelowia nudata 86 0.55
Hypericum crux-andreae 86 0.48
Andropogon gyrans var. gyrans 86 0.41
Dichanthelium consanguineum 86 0.38
Hypericum setosum 86 0.27
Gymnopogon brevifolius 86 0.25
Crotalaria purshii 86 0.23
Polygala nana 86 0.20
Rubus trivialis 86 0.20


Panhandle Wet Flatwoods/Prairies (16) Freq Cover
Aristida beyrichiana 100 50.51
Xyris ambigua 100 2.00
Rhexia alifanus 100 1.79
Smilax laurifolia 100 0.95
Ctenium aromaticum 94 9.88
Carphephorus pseudoliatris 94 1.38
Eriocaulon decangulare 88 4.75
Chaptalia tomentosa 88 1.21
Andropogon arctatus 81 5.31
Helianthus heterophyllus 81 2.47
Andropogon gyrans var. stenophyllus 8 1 1.43
Erigeron vernus 81 0.96
Coreopsis linifolia 81 0.74
Rhynchospora chapmanii 75 10.92
Bigelowia nu data 75 1.66
Muhlenbergia capillaris var. trichopodes 75 1.48
Rhynchospora plumosa 75 1.23
Rhynchospora baldwinii 75 0.95










Table B-1 continued
Panhandle Seepage Slopes (5) Freq Cover
Aristida beyrichiana 100 11.15
Scleria muehlenbergii 100 10.05
Rhynchospora i,gentrit,, 100 6.03
Aristida palustris 100 3.05
Andropogon gyrans var. stenophyllus 100 2.58
Eriocaulon decangulare 100 2.33
Smilax laurifolia 100 1.05
Bigelowia nudata 100 0.85
Coreopsis linifolia 100 0.75
Andropogon arctatus 100 0.70
Lobelia glandulosa 100 0.60
Rhynchospora latifolia 100 0.58
Oxypolis filiformis 100 0.50
Symphyotrichum dumosum var. dumosum 100 0.35
Rhexia alifanus 100 0.30
Sabatia macrophylla 100 0.15
Liatris spicata 80 9.28
Muhlenbergia capillaris var. trichopodes 80 6.70
Ctenium aromaticum 80 5.48
Pleea tenuifolia 80 1.03
Arnoglossum ovatum 80 0.73
Dichanthelium longiligulatum 80 0.65
Lophiola aurea 80 0.60
Paspalum praecox 80 0.50
Balduina uniflora 80 0.43
Lycopodiella appressa 80 0.43
Rhexia lutea 80 0.35
Drosera brevifolia 80 0.33
Erigeron vernus 80 0.30
Rubus trivialis 80 0.30
Juncus trigonocarpus 80 0.23
Rhexia petiolata 80 0.20
Eryngium integrifolium 80 0.20










APPENDIX C
MASTER LIST OF ABITA CREEK PRESERVE PLANT SPECIES.

Table C-1: All vascular plant species (and varieties) recorded at Abita Creek Preserve during
sample period 1997-2005. Code corresponds to labels on Figure 4-5 and 4-6.
"Type" indicates woody (W) or herbaceous (H). "Lifeform" indicates forb (F),
graminoid (G), and woody (W).

Code Species Type Lifeform
ACERU Acer rubrum W W
AGAOB Agalinus obtusifolia H F
AGASP Agalinus sp. H F
AGRPE 09.i~ r perennans H G
ALESP Aletris sp. H F
AM BAR Ambrosia artemisiifolia H F
AMOSP Amorpha sp. H F
ANDCA And'ropogon capillipes H G
ANDGL And'ropogon glomeratus H G
ANDGY And'ropogon gyrans var. gyrans H G
ANDMO And'ropogon mohrii H G
ANDPE And'ropogon perangustatus H G
ANDSP And'ropogon sp. H G
ANDVI And'ropogon virginicus H G
ANTRU Anthaenantia rufa H G
ANTVI Anthaenantia villosa H G
ARIPA Aristida palustris H G
ARIVI Aristida virgata H G
ARUTE Arundinaria gigantea ssp. tecta H G
ASCLO Asclepias sp. H F
ASCLO Asclepias longifolia H F
ASTAD Symphyotrichum ad'natum H F
ASTDU Symphyotrichum dumosum var. dumosum H F
AXOFI Axonopus fissifolius H G
BACHA Baccharis halimifolia W W
BALUN Balduwiana uniflora H F
BARPA Bartonia paniculata H F
BIDMI Bidens mitis H F
BIGCA Bignonia capreolata H F
BIGNU Bigelowia nud'ata H F
BOLSP Boltonia sp H F
BURSP Burmannia sp. H F
CACOV Cacalia ovata H F
CALAM Callicarpa americana W W
CARGL Carex glaucescens H C
CARPS Carphephorus pseudoliatris H F










Table C-1 continued
Code
CENTER
CEPOC
CHALA
CHAOR
CHATO
CHIVI
CLEAL
CLEDI
COERU
COETE
CORLI
CRASP
CROTO
CTEAR
CYPSP
CYRRA
DICDI
DICLA
DIOTE
DIOTE
DIOVI
DROBR
ELEMI
ELETU
ERARE
ERARE
EREHI
ERICO
ERIDE
ERIGI
ERIST
ERIVE
ERYIN
EUPCA
EUPLE
EUPRO
EUPSE
EUTLE
EUTTE
FRAPE
FUIBR
FUISP


Species
Centella erecta
Cephalanthus occidentalis
Chasmanthium laxum
Chasmanthium ornithorhynchum
Chaptalia tomentosa
Chionanthus virginicus
Clethra alnifolia
Cleistes divaricata
Coelorachis rugosa
Coelorachis tessellata
Coreopsis linifolia
Crataegus sp.
Croton sp.
Ctenium aromaticum
Cyperus compressus
Cyrilla racemiflora
Dichanthelium dichotomum
Rhynchospora latifolia
Diodia virginiana
Diodia teres
Diospyros virginiana
Drosera sp.
Eleocharis minima
Eleocharis tuberculosa
F a .~-t, r refracta
F a .~-t, r elliotii
Erechtites hieraciifolia
Eriocaulon compressum
Eriocaulon decangulare
Saccharum giganteum
Saccharum strictus
Erigeron vernus
Eryngium integrifolium
Eupatorium capillifolium
Eupatorium leucolepis
Eupatorium rotundifolium
Eupatorium semiserratum
Euthamia leptocephala
Euthamia tenuifolia var. tenuifolia
Fraxinus caroliniana
Fuirena breviseta
Fuirena sp.


H

H
H


H


H
H
H


H
H
H
W
H
H
H
H
W
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
W
H
H


Lifeform
F
W
G
G
F
W
W
F
G
G
F
W
F
G
C
W
G
C
F
F
W
F
C
C
G
G
F
F
F
G
G
F
F
F
F
F
F
F
F
W
G
G










Table C-1 continued
Code
GALMO
GAY MO
GELRA
GENSA
GRAPI
GRAPI
GRASP
GYMBR
HE LAN
HELHE
HELVE
HIBAC
HYPAL
HYPBR
HYPCI
HYPHI
HYPHY
HYPMU
HYPSE
HYPST
HYPWA
ILECO
ILEDE
ILEGL
ILEMY
ILEOP
ILEVO
IRIVI
ITEVI
JUNMA
JUNTR
LACCA
LECSP
LIASP
LIGSI
LINME
LINME
LIQST
LIRTU
LOBBR
LOBFL
LOBPU


Species
Gavhtssacia inosieri
Gavhtssacia duinosa
Gelseinium rankinii
Gentiana saponaria
Gratiola pilosa
Gratiola brevifolia
Gratiola sp.
Gvinnopogon brevifolius
Helianthus Helianthus heterophyllus
Heleniuin vernale
Hibiscus asculenta
Hyptis alata
Hypericuin brachyphylluin
Hypericuin cistifolium
Hypoxis sp.
Hypericuin hypericoides
Hypericuin iultiluin
Hypericuin setosuin
Hypericuin crux-andreae
Triadenuin virginicuin
Ilex coriacea
Ilex decidua
Ilex glabra
Ilex myrtifolia
Ilex opaca
Ilex voinitoria
Iris virginica
Itea virginica
Juncus inarginatus
Juncus trigonocarpus
Lachnanthes caroliana
Lechea sp.
Liatris spicata
Ligustruin sinense
Linuin medium
Linuin floridanuin
Liquidainhar a ,l,:r,l; me
Liriodendron tulipifera
Lobelia brevifolia
Lobelia floridana
Lobelia puberula


H
H
H
H
H
H
H
H
H
H

H

H
H
H
H
H
H
W
W
W





H
W
H
H
H
H
H
W
H
H
W
W
H
H
H


Lifeform
W
W
F
F
F
F
F
G
F
F
F
F
F
W
F
F
F
F
F
F
W
W
W
W
W
W
W
F
W
C
C
F
F
F
W
F
F
W
W
F
F
F










Table C-1 continued
Code
LOPAU
LUDGL
LUDHI
LUDHI
LUDLI
LUDSP
LUDVI
LYCAL
LYCVI
LYCVI
LYGJA
LYOLU
MAGGR
MAGVI
MALAN
MECAC
MITSE
MUHEX
MYRCE
MYRHE
NYSBI
OSMAM
OSMCI
OSMRE
OXYFI
PANAC
PANAN
PANAN
PANCO
PANEN
PANER
PANET
PANLE
PANLO
PANRI
PANSC
PANSO
PANSP
PANST
PANTE
PANVE
PANVI


Species
Lophiola aurea
Ludwigia glandulosa
Ludwigia pilosa
Ludwigia hirtella
Ludwigia linearis
Ludwigia sp.
Ludwigia virgata
Lycopodiella sp.
Lycopus virginicus
Lycopus rubellus var. co;,,-;,I r, illt a
Lygodium japonicum
Lyonia lucida
Magnolia gllrandirlora c
Magnolia virginiana
Malus cl,.-, or;:-1 Mecardonia acuminata
Mitreola sessihifolia
Muhlenbergia cappillaris var. tricopodes
Morella cerifera
Morella heterophylla
Nyssa biflora
Osmanthus americanus
Osmunda cinnamomea
Osmunda regalis
Oxypolis fihiformis
Dichanthelium acuminatum
Panicum anceps
Panicum hians
Dichanthelium consanguineum
Dichanthelium ensifolium
Dichanthelium erectifolium
Dichanthelium ensifolium var. tenue
Dichanthelium leaucothrix
Dichanthelium longiligulatum
Panicum rigidulum
Dichanthelium scabriusculum
Dichanthelium scoparium
Dichanthelium sp.
Dichanthelium strigosum
Panicum tenerum
Panicum verrucosum
Panicum virgatum


Tye
H
H
H
H
H
H
H
H
H
H



H



H
H



H






H
H
H
H
H
H
H
H
H
H
H
H
H
H
H


Lifeform
F
F
F
F
F
F
F
F
F
F
F
W
W
W
W
F
F
G
W
W
W
W
F
F
F
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G










Table C-1 continued
Code
PARQU
PASFL
PASPR
PASSE
PENSP
PERBO
PINEL
PINPA
PINTA
PITGR
PLURO
PLURO
POLL
POLRA
PROPE
PRUSE
PTEAQ
PYRAR
QUEFA
QUELA
QUENI
QUENI
QUEVI
RHEAL
RHELU
RHEMA
RHEPE
RHESP
RHEVI
RHOSP
RHUCO
RHUVE
RHYCA
RHYCE
RHYCH
RHYCN
RHYCO
RHYDE
RHYEL
RHYFI
RHYGB
RHYGL


Species
Parthenocissus quinquefolia
Paspalum floridanum
Paspalum praecox
Paspalum setaceum
Penstemon sp.
Persea borbonia
Pinus elliottii
Pinus palustris
Pinus taeda
Pityopsis graminifolia
Pluchea rosea
Pluchea foetida
Polygala lutea
Polygala ramosa
Proserpinaca pectinata
Prunus serotmna
Pteridium aquilinum
Photinia pyrifolia
Quercus falcata
Quercus laurifolia
Quercus nigra
Quercus laurifolia
Quercus virginiana
Rhexia alifanus
Rhexia lutea
Rhexia mariana var. mariana
Rhexia petiolata
Rhexia sp.
Rhexia virginiana
Rhododendron sp.
Rhus copallinum
Toxicodendron vernix
Rhynchospora chalarocephala
Rhynchospora cephalantha
Rhynchospora chapmanii
Rhynchospora corniculata
Rhynchospora compressa
Rhynchospora debilis
Rhynchospora elliottii
Rhynchospora fihifolia
Rhynchospora globularis
Rhynchospora glomerata


H
H
H



H





H
H
H
H
H





H
W






H




H
H
H




H
H
H
H


Lifeform
W
G
G
G
F
W
W
W
W
F
F
F
F
F
F
W
F
W
W
W
W
W
W
F
F
F
F
F
F
W
W
W
C
C
C
C
C
C
C
C
C
C










Table C-1 continued
Code
RHYGR
RHYIN
RHYOL
RHYPL
RHYPU
RHYRA
RHYSP
RUBUS
RUENO
SABSP
SABSP
SABSP
SAGLA
SALAZ
SAPSI
SARAL
SARPS
SCHSC
SCHTE
SCLCI
SCLGE
SCLHI
SCLMU
SCLPA
SCLPP
SCLTR
SCUIN
SETSP
SISAL
SMIBO
SMIGL
SMILA
SMIRO
SMISM
SOLOD
SOLRU
STOLA
STY AM
SYMTI
TEPON
TILUS
TOFRA


Species
Rhynchospora gracilenta
Rhynchospora inexpansa
Rhynchospora i,gentriter
Rhynchospora plumosa
Rhynchospora pusilla
Rhynchospora rariflora
Rhynchospora sp.
Rubus sp.
Ruellia noctiflora
Sabatia sp.
Sabatia difformis
Sabatia campanulata
Sagittaria lanceolata
Salvia azurea
Sapium sebiferum
Sarracenia alata
Sarracenia psittacina
Schizachyrium scoparium
Schizachyrium tenerum
Scleria ciliata var. ciliata
Scleria georgiana
Scleria hirtella
Scleria muhlenbergia
Scleria pauciflora var. caroliniana
Scleria pauciflora var. pauciflora
Scleria triglomerata
Scutellaria integrifolia
Setaria sp.
Sisyrinchium atlanticum
Smilax bona-nox
Smilax glauca
Smilax laurifolia
Smilax rotundifolia
Smilax smallii
Solidago odora
Solidago rugosa
Stokesia laevis
Styrax americanus
Symplocos tinctoria
Tephrosia onobrachyoides
Tillandsia usneoides
Tofieldia racemosa


Tye
H
H
H
H
H
H
H
H
H
H
H
H
H

H

H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H


H


H
H


Lifeform
C
C
C
C
C
C
C
F
F
F
F
F
F
F
W
F
F
G
G
C
C
C
C
C
C
C
F
G
F
F
F
F
F
F
F
F
F
W
W
F
F
F










Table C-1 continued
Code
TOXRA
TRADI
TRIAM
TRIVI
UTRIC
VACAR
VACEL
VIBDE
VIBNU
VIOLA
VIOPR
VITRO
WOOAR
XYR1AM
XYR1BA
XYR1CA
XYR1DI
XYR1LO
XYR1SM
XYR1SP
XYR1ST
XYR1ST
ZIGSP


Species
Toxicodendron radicans
Trachelospermum difforme
Tridens ambiguus
Triadenum virginicum
Utricularia juncea
Vaccinium arboreum
Vaccinium elliottii
Viburnum dentatum
Viburnum nudum
Viola lanceolata
Viola primulifolia
Vitis rotundifolia
Woodwardia areolata
Xyris ambigua
Xyris baldwiniana
Xyris caroliniana
Xyris d i ttbrm,,i.
Xyris louisianica
Xyris smalliana
Xyris sp.
Xyris iridifolia
Xyris sticta
Zigadenus sp.


H

H




H






H
H
W
H
H
H
H
H
H


Lifeform
W
F
G
W
F
W
W
W
W
F
F
W
F
F
F
F
F
F
F
F
F
F
F









LIST OF REFERENCES


Abella, S. R., V. B. Shelbume, and N. W. MacDonald. 2003. Multifactor classification of forest
landscape ecosystems of Jocassee Gorges, southern Appalachian Mountains, South
Carolina. Canadian Journal of Forest Research 33: 193 3-1946.

Abrahamson, W. G., and D. C. Hartnett. 1990. Flatwoods and dry prairies. Pages 103-149 in R.
L. Myers and J. J. Ewel, editors. Ecosystems of Florida. University of Central Florida
Press, Orlando, Florida, USA.

Adler, P. B., E. P. White, W. K. Lauenroth, D. M. Kaufman, A. Rassweiler, J. A. Rusak.
Evidence for a general species-time-area relationship. Ecology 86: 2032-2039.

Bailey, R. G., P. E. Avers, T. King, and W. H. McNab. 1994. Ecoregions and subregions of the
United States 1:7,500,000 (map) with supplementary table of map unit descriptions,
compiled and edited by W. H. McNab and R. G. Bailey. USDA Forest Service,
Washington D.C., USA.

Bell, G. 2001. Neutral macroecology. Science 293:2413-2418.

Bocard, D., L. Legendre, C. Avois-Jacquet, and H. Tuomisto. 2004. Dissecting the spatial
structure of ecological data at multiple scales. Ecology 85: 1826-1832.

Bocard, D., and P. Legendre. 2002. All-scale spatial analysis of ecological data by means of
principal coordinates of neighbour matrices. Ecological Modelling 153:51-68.

Borcard, D., and P. Legendre. 1994. Environmental control and spatial structure in ecological
communities: an example using oribatid mites (Acari, Orbatei). Environmental and
Ecological Statistics 1:1045-1055.

Borcard, D., P. Legendre, and P. Drapeau. 1992. Partialling out the spatial component of
ecological variation. Ecology 73:1045-1055.

Brady, N. C., and R. R. Weil. 2000. Elements of the nature and properties of soils, 12 edition.
Prentice-Hall Inc., Upper Saddle River, New Jersey, USA.

Bray, J. R., and J. T. Curtis. 1957. An ordination of the upland forest communities of southern
Wisconsin. Ecological Monographs 27:325-349.

Bridges, E. L. 2006a. Historical accounts of vegetation in the Kissimmee River dry prairie
landscape. in R. Noss, editor. Proceedings of the Florida Dry Prairie Conference.
University of Central Florida, Orlando, Florida, USA.









Bridges, E. L. 2006b. Landscape Ecology of Florida Dry Prairie in the Kissimmee River Region.
Pages 14-42 in R. Noss, editor. Land of Fire and Water. Proceedings of the Florida Dry
Prairie Conference. University of Central Florida, Orlando, Florida, USA.

Bridges, E. L., and S. L. Orzell. 1989. Longleaf pine communities of the West Gulf Coastal
Plain. Natural Areas Journal 9:246-263.

Brockway, D. G., and C. E. Lewis. 1997. Long-term effects of dormant-season prescribed fire on
plant community diversity, structure, and productivity in a longleaf pine-wiregrass
ecosystem. Forest Ecology and Management 96: 167-183.

Brockway, D. G., and K. W. Outcult. 2000. Restoring longleaf pine wiregrass ecosystems:
Hexazinone application enhances effects of prescribed fire. Forest Ecology and
Management 137:121-138.

Brooks, H. K. 1982. Guide to the Physiographic Divisions of Florida. IFAS Florida Cooperative
Extension Services, University of Florida, Gainesville, Florida, USA

Brown, R. B., E. L. Stone, and V. W. Carlisle. 1990. Soils. Pages 35-69 in R. L. Myers and J. J.
Ewel, editors. Ecosystems of Florida. University of Central Florida Press, Orlando,
Florida, USA.

Chen, E., and J. F. Gerber. 1990. Climate. Pages 11-34 in R. L. Myers and J. J. Ewel, editors.
Ecosystems of Florida. University of Central Florida Press, Orlando, Florida, USA.

Christensen, N. L. 1977. Fire and soil-plant nutrient relations in a pine-wiregrass savanna on the
coastal plain of North Carolina. Oecologia 31:1932-1939.

Cleland, D. T., J. B. Hart, G. E. Host, K. Pregitzer, and C. W. Ramm. 1993. Field guide:
ecological classification and inventory system of the Huron-Manistee National Forests.
U. S. Department of Agriculture Forest Service, Washington D.C., USA.

Clewell, A. F. 1971. The Vegetation of the Apalachicola National Forest: an Ecological
Perspective. U. S. Department of Agriculture Forest Service, Atlanta, Georgia, USA

Clewell, A. F. 1985. Guide to the Vascular Plants of the Florida Panhandle. Florida State
University Press, Tallahassee, Florida, USA.

Collins, S. L., S. M. Glenn, and J. M. Briggs. 2002. Effect of local and regional processes on
plant species richness in tallgrass prairie. Oikos 99:571-579.

Comer, P., D. Faber-Langendoen, R. Evans, S. Gawler, D. Josse, G. Kittel, S. Menard, M. Pyne,
M. Reid, K. Schulz, K. Snow, and J. Teague. 2003. Ecological Systems of the United
States: A Working Classification ofU.S. Terrestrial Systems. NatureServe, Arlington,
Virginia, USA.









Condit, R., N. Pitman, G. Leigh, J. Chave, J. Terborgh, R. B. Foster, P. Nunez, V. S. Aguilar, R.
Valencia, G. Villa, H. C. Muller-Landau, E. Loss, and S. P. Hubbell. 2002. Beta-
diversity in tropical forest trees. Science 295:666-669.

Cooper, A., T. McCann, and R. G. H. Bunce. 2006. The influence of sampling intensity on
vegetation classification and the implications for environmental management.
Environmental Conservation 33: 118-127.

Cornell, H. V., and J. H. Lawton. 1992. Species interactions, local and regional processes, and
limits to the richness of ecological communities: a theoretical perspective. Journal of
Animal Ecology 61:1-12.

Cox, A. C., D. R. Gordon, J. L. Slapcinsky, and G. S. Seamen. 2004. Understory restoration in
longleaf pine sandhills. Natural Areas Journal 24:4-14.

Croker, T. C. 1987. Longleaf Pine: A History of Man and a Forest. USDA Forest Service
Forestry Report R8-FR7. Asheville, North Carolina, USA.

Cushman, S. A., and K. McGarigal. 2002. Hierarchical, multi-scale decomposition of species-
environment relationships. Landscape Ecology 17:637-646.

Davis, J. H. 1967. General map of natural vegetation of Florida. Circular S-178. in. University
of Florida, Institute of Food and Agricultural Sciences, Gainesville, Florida, USA.

Defrene, M., and P. Legendre. 1997. Species assemblages and indicator species: the need for a
flexible asymmetrical approach. Ecological Monographs 67:345-366.

Dilustro, J. J., B. S. Collins, L. K. Duncan, and R. R. Sharitz. 2002. Soil texture, land-use
intensity, and vegetation of Fort Benning upland forest sites. Journal of the Torrey
Botanical Society 129:289-297.

Drewa, P. B., and W. J. Platt. 2002a. Community structure along elevation gradients in
southeastern longleaf pine savannas. Plant Ecology 160:61-78.

Drewa, P. B., W. J. Platt, and E. B. Moser. 2002b. Fire effects on resprouting of shrubs in
southeastern longleaf pine savannas. Ecology 83:755-767.

Fenneman, N. M. 1938. Physiography of eastern United States. McGraw-Hill Book Co., New
York, USA.

Fernald, E. A. 1981. Atlas of Florida. Florida State University Foundation, Tallahassee, Florida,
USA.

Florida Department of Environmental Protection. 1998. Surficial geology of Florida.. Florida
Geographic Data Library, Gainesville Florida, USA.









Florida Natural Areas Inventory. 1990. Guide to the natural communities of Florida. Florida
Department of Natural Resources, Tallahassee Florida, USA.

Florida Natural Areas Inventory. 2000a. Unpublished element occurrence data. Florida
Department of Natural Resources, Tallahassee, Florida, USA.

Florida Natural Areas Inventory. 2000b. Unpublished managed area data for Florida. Florida
Department of Natural Resources, Tallahassee Florida, USA.

Florida Natural Areas Inventory. 2007. Summary of Florida conservation lands. Florida
Department of Natural Resources, Tallahassee, Florida, USA.

Foster, B. L., and K. L. Gross. 1998. Species richness in a successional grassland: effects of
nitrogen enrichment and plant litter. Ecology 79:2593-2602.

Frost, C. 2006. History and future of the longleaf pine ecosystem. Pages 297-326 in S. Jose, E. J.
Jokela, and D. L. Miller, editors. The longleaf pine ecosystem: Ecology, silviculture, and
restoration. Springer, New York, USA.

Frost, C. C. 1993. Four centuries of changing landscape patterns in the longleaf pine ecosystem.
Pages 17-43 in S. M. Hermann, editor. The Longleaf Pine Ecosystem: Ecology,
Restoration, and management, Proceedings, 18th Tall Timbers Fire Ecology Conference.
Tall Timbers Research, Inc., Tallahassee, Florida, USA.

Fule, P. Z., W. W. Covington, and M. M. Moore. 1997. Determining reference conditions for
ecosystem management of southwestern ponderosa pine forests. Ecological Applications
7:895-908.

Gilliam, F. S., and W. J. Platt. 1998. Effects of long-tern fire exclusion on tree species
composition and stand structure in an old-growth Pinus palustris (Longleaf pine) forest.
Plant Ecology 0:1-12.

Glitzenstein, J. S., W. J. Platt, and S. D. R. 1995. Effects of fire regime and habitat on tree
dynamics in North Florida longleaf pine savannas. Ecological Monographs 65:441-476.

Glitzenstein, J. S., D. R. Streng, and D. D. Wade. 2003. Fire frequency effects on longleaf pine
(Pinus palustris) vegetation in South Carolina and Northeast Florida, USA. Natural Areas
Journal 23:22-37.

Godfrey, R. K. 1988. Tree, shrubs and woody vines of northern Florida and adj acent Georgia and
Alabama. University of Georgia Press, Athens, Georgia, USA.

Godfrey, R. K., and J. W. Wooten. 1979. Aquatic and Wetland Plants of Southeastern United
States: Monocotyledons. University of Georgia Press, Athens, Georgia, USA.

Godfrey, R. K., and J. W. Wooten. 1981. Aquatic and wetland plants of Southeastern United
States: Dicotyledons. University of Geogia Press, Athens, Georgia USA.










Goebel, P. C., B. J. Palik, L. K. Kirkman, M. B. Drew, L. West, and D. C. Pederson. 2001.
Forest ecosystems of a Lower Gulf Coastal Plain landscape: multifactor classification and
analysis. Journal of the Torrey Botanical Society 128:47-75.

Graae, B. J., R. H. Okland, P. M. Petersen, K. Jensen, and B. Fritzboger. 2004. Influence of
historical, geographical and environmental variables on understorey composition and
richness in Danish forests. Journal of Vegetation Science 15:465-474.

Grace, J. B., L. Allain, and C. Allen. 2000. Factors associated with plant species richness in a
coastal tall-grass prairie. Journal of Vegetation Science 11:443-452.

Grace, J. B., and B. H. Pugesek. 1997. A structural equation model of plant species richness and
its application to a coastal wetland. The American Naturalist 149:436-460.

Graham, C. H., T. B. Smith, and M. Languy. 2005. Current and historical factors influencing
patterns of species richness and turnover of birds in the Gulf of Guinea highlands. Journal
of Biogeography 32:1371-1384.

Greenberg, C. H., D. G. Neary, L. D. Harris, and S. P. Linda. 1995. Vegetation recovery
following high-intensity wildfire and silvicultural treatments in sand pine scrub.
American Midland Naturalist 133: 149-163.

Griffith, G. E., J. M. Omernik, C. W. Rohm, and S. M. Pierson. 1994. Florida regionalization
proj ect. U. S. Environmental Protection Agency, National Health and Environmental
Effects Research Laboratory, Corvallis, Oregon, USA.

Grime, J. P. 1979. Plant strategies and vegetation processes. Wiley, Chichester, U.K.

Grossman, D. H., D. Faber-Langendoen, A. S. Weakley, M. Anderson, P. Bourgeron, R.
Crawford, K. Goodin, S. Landaal, K. Metzler, K. D. Patterson, M. Pyne, M. Reid, and L.
Sneddon. 1998. International classification of ecological communities: terrestrial
vegetation of the United States. The Nature Conservancy, Arlington, Virginia, USA.

Hammond, E.H. 1964. Classes of land surface form in the 48 states, U.S.A. Annals of the
Association of American Geographers 54(1): map supplement.

Harper, R. M. 1914. Geography and vegetation of North Florida. Pages 163-437 in. Florida State
Geological Survey. Annual Report. Tallahassee, Florida, USA.

Harrington, T. B., and M. B. Edwards. 1999. Understory vegetation, resources availability, and
litterfall responses to pine thinning and woody vegetation control in longleaf pine
plantations. Canadian Journal of Forest Research 29: 105 5-1064.

Hedman, C. W., S. L. Grace, and S. E. King. 2000. Vegetation composition and structure of
southern coastal plain pine forests: an ecological comparison. Forest Ecology and
Management 134:233-247.











Heikkinen, R. K., and H. J. B. Birks. 1996. Spatial and environmental components of variation in
the distribution patterns of subarctic plant species at Kevo, N Finland a case study at the
meso-scale level. Ecography 19: 341-351.

Hix, D. M., and J. N. Pearcy. 1997. Forest ecosystems of the Marietta Unit, Wayne National
Forest, southeastern Ohio: multifactor classification and analysis. Canadian Journal of
Forest Research 27:1117-1131.


Hodgkins, E.J. 1965. Southeastern forest habitat regions based on physiography. Agricultural
Experiment Station, Auburn University, Forestry Department Series, No. 2. Auburn,
Alabama, USA.

Hodgkins, E.J., M. S. Golden and W.F. Miller. 1979. Forest habitat regions and types on a
photomorphic-physiographic basis: A guide to forest site classification in Alabama-
Mississippi. Southern Coop Series 210. Alabama Agriculture Experiment Station,
Auburn, Alabama, USA.

Holdridge, L. 1967. Life Zone Ecology. Tropical Science Center, San Jose, Costa Rica.

Hubbell, S. P. 2001. The unified neutral theory of biodiversity and biogeography. Princeton
University Press, Princeton, New Jersey, USA.

Hubbell, S. P., and R. B. Foster. 1986. Biology, chance, and history and the structure of tropical
rain forest tree communities. Pages 314-330 in J. Diamond and T. J. Case, editors.
Community Ecology. Harper & Row, New York, USA.

Hull, J. P. D. 1962. Cretaceous Suwannee strait, Georgia and Florida. AAPG Bulletin 46: 118-
122.

Huston, M. A. 1979. A general hypothesis of species diversity. American Naturalist 113:81-101.

Hutchinson, G. E. 1957. Concluding remarks. Cold Spring Harbor Symposia on Quantitative
Biology 22:415-427.

James, M. M. 2000. Legumes in loamy soil communities of the Carolina Sandhills: their natural
distributions and performance of seeds and seedlings along complex ecological gradients.
University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA.

James, C. W. 1961. Endemism in Florida. Brittonia 13:225-244.

Kautz, R. S., and J. A. Cox. 2001. Strategic habitats for biodiversity conservation in Florida.
Conservation Biology 15:55-77.

Kenward, M. G. 1987. A method for comparing profiles of repeated measures. Applied Statistics
36:296-308.









Kirkman, L. K., K. L. Coffey, R. J. Mitchell, and E. B. Moser. 2004. Ground cover recovery
patterns and life-history traits: implications for restoration obstacles and opportunities in
a species-rich savanna. Journal of Ecology 92:409-421.

Kirkman, L. K., R. J. Mitchell, R. C. Helton, and M. B. Drew. 2001. Productivity an species
richness across an environmental gradient in a fire-dependent ecosystem. American
Journal of Botany 88:2119-2128.

Landers, J. L., D. H. V. Lear, and W. D. Boyer. 1995. The longleaf pine forests of the Southeast:
requiem or renaissance? Journal of Forestry 93:39-44.

Laughlin, D. C., and S. R. Abella. 2007. Abiotic and biotic factors explain independent gradients
of plant community composition in ponderosa pine forests. Ecological Modelling
205:231-240.

Laughlin, D. C., J. D. Bakker, and P. Z. Fule. 2005. Understorey plant community structure in
lower montane and subalpine forests, Grand Canyon National Park, USA. Journal of
Biogeography 32:2083-2102.

Legendre, P., D. Borcard, and P. R. Peres-Neto. 2005. Analyzing beta diversity: partioning the
spatial variation of community composition data. Ecological Monographs 75: 435-450

Legendre, P., and M. Fortin. 1989. Spatial pattern and ecological analysis. Vegetatio 80:107-138.

Legendre, P., and E. D. Gallagher. 2001. Ecologically meaningful transformations for ordination
of species data. Oecologia 129:271-280.

Legendre, P., and L. Legendre. 1998. Numerical Ecology, 2 edition. Elsevier Science,
Amsterdam.

Leps, J., and P. Smilauer. 2003. Multivariate analysis of ecological data using CANOCO.
Cambridge University Press, Cambridge, UK.

Leps, J., and P. Smilauer. 2007. Subjectively sampled vegetation data: don't throw out the baby
with the bath water. Folia Geobotanica 42: 169-178.

Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D. Wolfinger. 1996. SAS System for Mixed
Model. SAS Institute Inc., Cary, North Carolina, USA.

Littell, R. C., J. Pendergast, and R. Natarajan. 2000. Tutorial in Biostatistics: Modeling
covariance structure in the analysis of repeated measures data. Statistics in Medicine
19:1793-1819.

Lockett, S. H. 1870. Louisiana as it is. Louisiana State University Press, Baton Rouge,
Louisiana, USA.









Martin, W. H., S. G. Boyce, and A. C. Echternacht, editors. 1993. Biodiversity of the
Southeastern United States: Upland Terrestrial Communities. Wiley, New York, USA.

McCune, B., and J. Grace. 2002. Multivariate Analysis of Ecological Communities. MjM
Software, Gleneden Beach, Oregon, USA.

McCune, B., and M. J. Mefford. 1999. PC-ORD. Multivariate Analysis of Ecological Data
version 5.0. MjM Software, Gleneden Beach, Oregon, USA.

McIntyre, S., and S. Lavorel. 1994. How environmental and disturbance factors influence species
composition in temperate Australian grasslands. Journal of Vegetation Science 5:373-
384.

Means, D. B. 1996. Longleaf pine forests, going going. Pages 210-229 in M. E. Davis, editor.
Eastern Old-growth Forests. Island Press, Washington D.C., USA.

Mehlich, A. 1984. Mehlich 3 soil test extraction modification ofMehlich 2 extractant.
Communications in Soil Science and Plant Analysis 15:1409-1416.

Mehlman, D. W. 1992. Effects of fire on plant community composition of North Florida second
growth pineland. Bulletin of the Torrey Botanical Club 119:376-383

Mitchell, R. J., L. K. Kirkman, S. D. Pecot, C. A. Wilson, B. J. Palik, and L. R. Boring. 1999.
Patterns and controls of ecosystem function in longeaf pine wiregrass savannas. I.
Aboveground net primary productivity. Canadian Journal of Forest Research 29:743-751.

Myers, R. L. 1990. Scrub and high pine. Pages 150-193 in R. L. Myers and J. J. Ewel, editors.
Ecosystems of Florida. University of Central Florida Press, Orlando, Florida, USA.

Mohr, C. 1898. The Timber Pines of the Southern United States. Government Printing Office,
Washington D. C., USA.

Myers, R. L. 2000. Physical setting. in R. P. Wunderlin and B. F. Hansen, editors. Flora of
Florida: Pteridophytes and gymnosperms. The University Press of Florida, Gainesville,
Florida, USA.

Myers, R. L., and J. J. Ewel, editors. 1990. Ecosystems of Florida. University Press of Florida,
Gainesville, Florida, USA.

Nekola, J. C., and P. S. White. 1999. The distance decay of similarity in biogeography and
ecology. Journal of Biogeography 26:867-878.

Noel, J. M., W. J. Platt, and E. B. Moser. 1998. Characteristics of old- and second-growth stand
of longleaf pine (Pinus palustris) in the Gulf coastal region of the U. S.A. Conservation
Biology 12:533-548.









Noss, R. F. 1988. The longleaf pine landscape of the Southeast: almost gone and almost
forgotten. Endangered Species UPDATE 5:1-8.

Oesterheld, M. J., J. Loreti, M. Semmartin, and J. M. Paruelo. 1999. Grazing, Gire, and climate
effects on primary productivity of grasslands and savannas. Pages 287-306 in L. R.
Walker, editor. Ecosystems of Disturbed Ground. Elsevier, New York, USA.

Okland, R. H. 1999. On the variation explained by ordination and constrained ordination axes.
Journal of Vegetation Science 10: 13 1-136.

Okland, R. H. 2003. Partitioning the variation in a plot-by-species data matrix that is related to n
sets of explanatory variables. Journal of Vegetation Science 14:693-700.

Okland, R. H., and O. Eilersten. 1994. Canonical correspondence analysis with variation
partitioning: some comments and an application. Journal of Vegetation Science 5: 117-
126.

Okland, R. H., K. Rydgren, and T. Okland. 2003. Plant species composition of boreal spruce
swamp forests: closed doors and windows of opportunity. Ecology 84: 1909-1919.

Oksanen, J., R. Kindt, P. Legendre, and R. B. O'Hara. 2007. vegan: Community Ecology
Package version 1.8-6.

Olson, M. S., and W. J. Platt. 1995. Effects of habitat and growing season fires on resprouting of
shrubs in longleaf pine savannas. Vegetatio 119: 101-1 18.

Omernik, J. M. 1987. Ecoregions of the Conterminous United States. Map (scale 1:7,500,000).
Annals of the Association of American Geographers 77: 118-125.

Outcalt, K. W., and R. M. Sheffield. 1996. The longleaf pine forest: trends and current
conditions. Resource Bulletin SRS-9, USDA Forest Service Southern Research Station.

Ostertag, T.E. and K.M. Robertson. 2006. A comparison of native versus old-Hield vegetation in
upland pinelands managed with frequent fire, south Georgia, USA. Tall Timbers Fire
Ecology Conference Proceedings, 23, in press.

Palmer, M. A., R. F. Ambrose, and N. L. Poff. 1997. Ecological theory and community
restoration ecology. Restoration Ecology 5:291-300.

Palmer, M. W., J. R. Arevalo, M. C. Cobo, and P. G. Earls. 2003. Species richness and soil
reaction in a northeastern Oklahoma landscape. Folia Geobotanica 38:381-389.

Partel, M. 2002. Local plant diversity patterns and evolutionary history at the regional scale.
Ecology 83:2361-2366.









Peet, R. K. 2006. Ecological classification of the longleaf pine woodlands. Pages 5 1-94 in S.
Jose, E. J. Jokela, and D. L. Miller, editors. The longleaf pine ecosystem: Ecology,
silviculture, and restoration. Springer, New York, USA.

Peet, R. K., and D. J. Allard. 1993. Longleaf pine vegetation of the Southern Atlantic and
Eastern Gulf Coast regions: A preliminary classification. Tall Timbers Fire Ecology
Conference Proceedings 18:45-81.

Peet, R. K., J. D. Fridley, and J. M. Gramling. 2003. Variation in species richness and species
pool size across a pH gradient in forests of the southern Blue Ridge mountains. Folia
Geobotanica 38:391-401.

Peet, R. K., and O. L. Loucks. 1977. A gradient analysis of southern Wisconsin forests. Ecology
58:485-499.

Peet, R. K., T. R. Wentworth, and P. S. White. 1998. A flexible, multipurpose method for
recording vegetation composition and structure. Castanea 63:262-274.

Penfound, W. T. 1944. Plant distribution in relation to the geology of Louisiana. The
Proceedings of the Louisiana Academy of Sciences 8:25-34.

Penfound, W. T., and A. G. Watkins. 1937. Phytosociological studies in the pinelands of
Southeastern Louisiana. American Midland Naturalist 18:661-682.

Peres-Neto, P. R., P. Legendre, S. Dray, and D. Borcard. 2006. Variation partitioning of species
data matrices: estimation and comparison of fractions. Ecology 87:2614-2625.

Platt, W. J. 1999. Southeastern pine savannas. in R. C. Anderson, J. S. Fralish, and J. M. Baskin,
editors. Savannas, barrens, and rock outcrop plant communities of North America.
Cambridge University Press, Cambridge, UK.

Platt, W. J., S. C. Carr, M. Reilly, and J. Fahr. 2006. Pine savanna overstory influences on
ground cover biodiversity. Applied Vegetation Science 9:37-50.

Platt, W. J., G. W. Evans, and M. M. Davis. 1988a. Effects of fire season on flowering forbs and
shrubs in longleaf pine forests. Oecologia 76:353-368.

Platt, W. J., G. W. Evans, and S. L. Rathbun. 1988b. The population dynamics of a long-lived
conifer (Pinus palustris). American Naturalist 131:491-525

Platt, W. J., J. M. Huffman, and M. G. Slocum. 2006. Fire regimes and trees in Florida dry
prairie landscapes. Pages 3-13 in R. Noss, editor. Land of Fire and Water. Proceedings of
the Florida Dry Prairie Conference. University of Central Florida, Painter, DeLeon
Springs, Florida, USA.

Platt, W. J., and I. M. Weis. 1977. Resource partitioning and competition within a guild of
fugitive prairie plants. The American Naturalist 111:479-513.











Provencher, L., B. J. Herring, D. R. Gordon, H. L. Rodgers, K. E. M. Galley, G. W. Tanner, J. L.
Hardesty, and L. A. Brennan. 2001. Effects of hardwood reduction techniques on
longleaf pine sandhill vegetation in Northwest Florida. Restoration Ecology 9: 13-27.

Provencher, L., B. J. Herring, D. R. Gordon, H. L. Rodgers, G. W. Tanner, L. A. Brennan, and J.
L. Hardesty. 2000. Restoration of Northwest Florida sandhills through harvest of invasive
Pinus clause. Restoration Ecology 8:175-185.

Puri, H. S., and R. O. Vernon. 1964. Summary of the geology of Florida and a guidebook to the
classic exposures. Fla. Geol. Sury. Spec. Publ. 5, Tallahassee, Florida, USA.

R Development Core Team (2007). R: A language and environment for statistical computing. R
foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL
http://www.R-proj ect. org.

Randazzo, A. F., and D. S. Jones, editors. 1997. The Geology of Florida. University Press of
Florida, Gainesville, Florida, USA

Rebertus, A. J., G. B. Williamson, and E. B. Moser. 1989. Longleaf pine pyrogenicity and turkey
oak mortality in Florida xeric sandhills. Ecology 70:60-70.

Ricklefs, R. E. 1987. Community diversity: relative roles of local and regional processes.
Science 235: 167-171.

Robbins, L. E., and R. L. Myers. 1992. Seasonal effects of prescribed burning in Florida: a
review. Miscellaneous Publication Number 8, Tall Timbers Research, Inc., Tallahassee,
Florida, USA.

SAS Institute Inc. 2000. SAS OnlineDoc, Version 8, Cary, North Carolina, USA.

Seaman, G. 1998. A longleaf pine sandhill restoration in northwest Florida. Restoration and
Management Notes 16:46-50.

Seastedt, T. R., J. M. Briggs and D. J. Gibson. 1991. Controls of nitrogen limitation in tallgrass
prairie. Oecologia 87:72-79.

Simberloff, D. 1993. Species-area and fragmentation effects on old-growth forests: prospects for
longleaf pine communities. Proc. Tall Timbers Fire Ecology Conference 18:227-263.

Sorrie, B. A., and A. S. Weakley. 2006. Conservation of the endangered Pinus palustris
ecosystem based on Coastal Plain centres of plant endemism. Applied Vegetation Science
9:59-66.

Sorrie, B. A., and A. S. Weakley. 2002. Coastal plain vascular plant endemics: phytogeographic
patterns. Castanea 66:50-82.









Stohlgren, T. J., L. D. Schell, and B. V. Heuvel. 1999. How grazing and soil quality affect native
and exotic plant diversity in Rocky mountain grasslands. Ecological Applications 9:45-
64.

Streng, D. R., J. S. Glitzenstein, and W. J. Platt. 1993. Evaluation effects of season of burn in
longleaf pine forests: a critical literature review and some results from an ongoing long-
term study. Tall Timbers Fire Ecology Conference No. 18, Tallahassee, Florida, USA.

Svenning, J., and F. Skov. 2005. The relative roles of environment and history as controls of tree
species composition and richness in Europe. Journal of Biogeography 32:1019-1033.

Svenning, J. C., D. A. Kinner, R. F. Stallard, B. M. J. Engelbrecht, and S. J. Wright. 2004.
Ecological determinism in plant community structure across a tropical forest landscape.
Ecology 85:2526-2538.

Swetnam, T. W., C. D. Allen, and J. L. Betancourt. 1999. Applied historical ecology: using the
past to manage for the future. Ecological Applications 9: 1189-1206.

Tabachnick, B. G., and L. S. Fidell. 1996. Using multivariate statistics, 3 edition. HarperCollins
College Publishers Inc., New York, USA.

ter Braak, C. J. F., and P. Smilauer. 2002. CANOCO reference manual and CanoDraw for
Windows user's guide: software for Canonical Community Ordination (version 4.5).
Microcomputer Power, Ithaca, New York, USA.

The Nature Conservancy 2001. East Gulf Coastal Plain Ecoregional Plan. The Nature
Conservancy, Arlinton, Virginia, USA.

The Nature Conservancy. 1997. Abita Creek Flatwoods Preserve Site Conservation Plan.
Unpublished report. The Nature Conservancy Louisiana Field Office, Baton Rouge,
Louisiana, USA.

Thornton, P. E., S. W. Running, and M. A. White. 1999. Generating surfaces of daily
meteorological variables over large regions of complex terrian. Journal of Hydrology
190:214-251.

Tilman, D. 1996. Biodiversity: population versus ecosystem stability. Ecology 77:350-363.

Tilman, D. 1994. Competition and biodiversity in spatially structured habitats. Ecology 75:2-16.

Trahan, L., J.J. Bradley, and L.Morris. 1990. Soil survey of St. Tammany Parish, Louisiana.
USDA Soil Conservation Service. Baton Rouge, Louisiana, USA

Tuomisto, H., K. Ruokolainen, and M. Yli-Halla. 2003. Dispersal, environment, and floristic
variation of western Amazonian forests. Science 299:241-244.









Turner, C. L., J. M. Blair, R. J. Schartz, and J. C. Neel. 1997. Soil N and plant responses to fire,
topography, and supplemental N in tallgrass prairie. Ecology 78: 1832-1843.

Underwood, A. J. 1994. On beyond BACI: sampling designs that might reliably detect
environmental disturbances. Ecological Applications 4:3-15.

Vandvik, V., and H. J. B. Birks. 2002. Partitioning floristic variance in Norwegian upland
grasslands into within-site and between-site components: are there patterns determined by
environment or by land-use? Plant Ecology 162:233-245.

VanLear, D. H., W. D. Carroll, P. R. Kapeluck, and R. Johnson. 2005. History and restoration of
the longleaf pine-grassland ecosystem: implications for species at risk. Forest Ecology
and Management 211.

Varner, J. M., D. R. Gordon, F. E. Putz, and J. K. Hiers. 2005. Restoring fire to long-unburned
Pinus palustris ecosystems: novel fire effects and consequences for long-unbumed
ecosystems. Restoration Ecology 13:536-544.

Vitousek, P. M. 1982. Nutrient cycling and nutrient use efficiency. American Naturalist 119:553-
572.

Walker, J., and R. K. Peet. 1983. Composition and species diversity of pine-wiregrass savannas
of the Green Swamp, North Carolina. Vegetatio 55:163-179.

Walker, J. L., and A. M. Silletti. 2006. Restoring the ground layer of longleaf pine ecosystems.
Pages 297-326 in S. Jose, E. J. Jokela, and D. L. Miller, editors. The longleaf pine
ecosystem: Ecology, silviculture, and restoration. Springer, New York, USA.

Wahlenberg, W. G. 1946. Longleaf pine: its use, ecology, regeneration, protection, growth and
management. Charles Lathrop Pack Forestry Foundation, Washington, D.C, USA.

Waldrop, T. A., D. L. White, and S. M. Jones. 1992. Fire regimes for pine-grassland
communities in the southeastern United States. Forest Ecology and Management 47:195-
210.

Ward, D., editor. 1979. Rare and endangered biota of Florida. Volume 5. Plants. University
Presses of Florida, Gainesville, Florida, USA.

Weakley, A. S. 2002. Flora of the Carolinas, Virginia, Georgia and surrounding areas.
Unpublised manuscript. University of North Carolina Herbarium, University of North
Carolina at Chapel Hill, Chapel Hill, North Carolina, USA.

Webb, S. D. 1990. Historical biogeography. Pages 70-102 in R. L. Myers and J. J. Ewel, editors.
Ecosystems of Florida. University of Central Florida Press, Orlando, Florida, USA

Weiher, E., S. Forbes, T. Schauwecker, and J. B. Grace. 2004. Multivariate control of plant
species richness and community biomass in blackland prairie. Oikos 106:151-157.











White, D. L., W. T. A., and J. S. M. 1991. Forty years of prescribed burning on the Santee fire
plots: effects on understory vegetation. in S. C. Nodvin and T. A. Waldrop, editors. Fire
and the environment: ecological and cultural perspectives. USDA, Forest Service,
Southeastern Forest Experiment Station, Asheville, North Carolina, USA.

White, P. S. 1979. Pattern, process, and natural disturbance in vegetation. Botanical Review:229-
299.

White, P. S., and J. L. Walker. 1997. Approximating nature's variation: selecting and using
reference information in restoration ecology. Restoration Ecology 5:338-349.

Whittaker, R. H. 1967. Gradient analysis of vegetation. Biological Reviews 42:207-264.

Whittaker, R. H. 1962. Classification of natural communities. Botanical Review 28: 1-239.

Whittaker, R. H. 1956. Vegetation of the Great Smoky Mountains. Ecological Monographs 26: 1-
80.

Wilson, C. A., R. J. Mitchell, J. J. Hendricks, and L. R. Boring. 1999. Patterns and controls of
ecosystem function in longleaf pine wiregrass savannas. II. Nitrogen dynamics.
Canadian Joumnal of Forest Research 29:752-760

Wiser, S. K., R. K. Peet, and P. S. White. 1996. High-elevation rock outcrop vegetation of the
Southern Appalachian mountains. Journal of Vegetation Science 7:703-722.

Wunderlin, R. P. 1998. Guide to the Vascular Plants of Florida. University Press of Florida,
Gainesville, Florida, USA.

Wunderlin, R. P., and B. F. Hansen. 2004. Atlas of Florida Vascular Plants
(http ://www.plantatlas.usf. edu). Institute for Systematic Botany, University of South
Florida, Tampa, Florida, USA.

Wunderlin, R. P., and B. F. Hansen, editors. 2000. Flora of Florida. The University Press of
Florida, Gainesville, Florida, USA.

Zobel, M. 1992. Plant-species coexistence the role of historical, evolutionary and ecological
factors. Oikos 65:314-320.

Zobel, M. 1997. The relative role of species pools in determining plant species richness: an
alternative explanation of species coexistence? Trends in Ecology and Evolution 12:266-
269.









BIOGRAPHICAL SKETCH

Susan Carr was born and raised in Gainesville, Florida. She graduated from the University of

Florida in 1982 with a Bachelor of Science degree in botany. After college, Susan worked in the

fields of land management and conservation, including employment with the Florida Natural

Areas Inventory, U.S. Forest Service and The Nature Conservancy. She returned to graduate

school in the mid-1990s and earned a Master of Science in plant ecology from Louisiana State

University. Following a long period of field data collection in Florida and employment with the

University of North Carolina in Chapel Hill, Susan j oined the Department of Wildlife Ecology

and Conservation at the University of Florida in 2005.





PAGE 1

1 FLORISTIC AND ENVIRONMENTAL VARIATI ON OF PYROGENIC PINELANDS IN THE SOUTHEASTERN COASTAL PLAIN: DESC RIPTION, CLASSIFICATION, AND RESTORATION By SUSAN CATHERINE CARR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Susan Carr

PAGE 3

3 To Mike, my husband and great love

PAGE 4

4 ACKNOWLEDGMENTS I thank the members of my committee, Dr Debbie Miller, Dr. Doria Gordon, and Dr. Wiley Kitchens for their patience and support. I especially thank my advisors, Dr. George Tanner and Dr. Kevin Robertson. They were rea lly there for me, and I am grateful for their support. Many people were involved in every stage of this project, including initial information gathering, site selection, vegeta tion sampling, data entry and mana gement, and data analysis. For assistances with the initial ecoregion delineati on and site selection pro cess, I thank Dr. Bruce Means, Dr. Bill Platt, Wilson Baker, Andy VanLoan, Ann Johns on, Carolyn Kindell, Dr. Louis Provencher, and Brenda Herri ng. Many employees of state, federal, and private land management agencies provided invaluable assi stance with logistical matters ranging from permitting paperwork to site selection and access. These include:, Dr. Dennis Hardin, Charlie Pederson, Dr. Ann Cox, Ace Haddick, Tom Servi ss, Bobby Cahal, Vince Morris, Scott Crosby, Dan Pearson, John McKenzie, Cr aig Parenteau, Ginger Morgan, Ro sie Mulholland, Alice Bard, Ken Alvarez, Terry Hingtgen, Bobby Hattaway, D onna Watkins, Mark Latch, Roy Ogles, Carla Jean Ogles, Dr. Jean Huffman, Louise Kirn, Dr. Guy Aglin, Jim Ruhl, Lorraine Miller, Kevin Love, Greg Seamon, Monica Folk, Sandy Woiak, B ee Pace, Lynn Askins, Kevin Hiers, Steve Orzell, Edwin Bridges, Amanda Stevens, Jerr y Pitts, Kristen Wood, Ra y and Patricia Ashton, Raymond Bass, and Gary Maxwell. I thank those that helped with field data collection, and data entry, management and analysis, including Dr. Joel Gramling, Dr. A nn Johnson, Kevin Hiers, Brian Mealor, Steve Orzell, Edwin Bridges, John Brubaker, Dr. Jeff Glitsenzen, Dr. Donna Streng, Maynard Hiss, Dr. Brian Strom, Dr. Jean Huffman, Dr. Bill Platt, Dr. Robert Peet and Deb Cupples. I extend

PAGE 5

5 special gratitude to Jessica Kapl an, Christine Carlson, Brian Stro m, and Dr. Joel Gramling for their long hours toi ling in the field an d at the computer. I am grateful to the faculty and staff of the University of Florida Herbarium, for allowing me to use their facilities a nd exploit their knowledge includ ing Kent Perkins, Dr. Norris Williams, Trudy Lindler and Dr. Walter Judd. I especially thank Richard Abbot, Patrick McMillian, and Brenda Herring for their help with plant identifi cations. I sampled vegetation on the Ordway-Swisher Biological Station which is owned and managed by the University of Florida. Thanks to Steve Coates for allowing acce ss and helping with site selection. I thank the staff of the Department of Wildlife Ecology an d Conservation for their acceptance and support during my tenure as a graduate student at UF. Elaine Culpepper, Dana Tomasevic, and Delores Tillman helped with final dissertation preparation and presentation. The Florida portion of this work was funded by the Florida Fish and Wildlife Conservation Commission. I tha nk Dr. Robert Peet, the project Principal Investigator, for providing this research opportunity. The Abita Preserve re search was funded by the Louisiana Field Office of The Nature Conservancy. I than k the TNC employees who initiated the project and helped along the way: Nelwyn McInnis, La timore Smith, Richard Martin, Judy Teague, David Moore, and David Baker. Finally, I thank my parents, Drs. Thomas a nd Glenna Carr. They taught me the value of knowledge and scientific inquir y. Lastly, I thank my husband Mike Hoganson. Without his love, support, and tolerance, I neve r would have finished this proj ect. He is my biggest fan.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4 LIST OF TABLES ................................................................................................................ ...........8 LIST OF FIGURES ............................................................................................................... ..........9 ABSTRACT ...................................................................................................................... .............10 CHAPTER 1 INTRODUCTION ................................................................................................................ ...13 2 VEGETATION CLASSIFICATION OF FLORIDA’S PYROGE NIC PINELANDS ...........19 Introduction .................................................................................................................. ..... 19 Methods....................................................................................................................... .......23 Study Area ...................................................................................................................2 3 Selection of Sample Sites.............................................................................................25 Field Methods ..............................................................................................................27 Numerical Analysis ......................................................................................................29 Results ...................................................................................................................... ..........31 Series 1: Dry Uplands ..................................................................................................33 Series 2: Mesic Flatwoods ...........................................................................................41 Series 3: Wetlands........................................................................................................45 Discussion .................................................................................................................... ......53 Comparisons to Other Classifications ..........................................................................56 3 GEOGRAPHIC, ENVIRONMENTAL AND REGIONAL VARIATION IN FLORISTIC COMPOSITION OF FLORIDA PYROGENIC PINELANDS ...............................................75 Introduction ................................................................................................................. .......75 Methods....................................................................................................................... .......78 Study Region ................................................................................................................ 78 Vegetation and Environmental Data ............................................................................79 Numerical Data Assembly and Analysis .....................................................................81 Results ....................................................................................................................... .........86 Variation Partitioning Models ......................................................................................87 Environmental Explanatory Variables .........................................................................88 Discussion ................................................................................................................... .......90 4 ECOLOGICAL RESTORATION OF A LONGLEAF PINE SA VANNA IN THE SOUTHEASTERN COASTAL PLAIN ................................................................................105 Introduction .................................................................................................................. ....105 Methods....................................................................................................................... .....109 Study Site and Reference Sites ..................................................................................109

PAGE 7

7 Restoration Treatment s and Sampling Methods ........................................................111 Data Analysis .............................................................................................................11 4 Comparisons of ACP data to Reference Data ............................................................119 Results ....................................................................................................................... .......120 Trends in Species Richness and Woody Stems .........................................................120 Trends in Species Composition .................................................................................121 ACP Treatment Responses vs. Reference Conditions ...............................................124 Discussion .................................................................................................................... ....125 ACP Restoration Compared to Reference Model ......................................................128 Management and Conservation Implications .............................................................130 5 CONCLUSION ................................................................................................................. .....143 APPENDIX A LOCATIONS OF FLORI DA VEGETATION PLOTS .........................................................146 B LIST OF FREQUENT AND ABUN DANT SPECIES BY COMMUNITY ASSOCIATION .................................................................................................................. ...154 C ABITA CREEK PRESERVE PLANT SPECIES ..................................................................163 LIST OF REFERENCES ............................................................................................................ .170 BIOGRAPHICAL SKETCH .......................................................................................................184

PAGE 8

8 LIST OF TABLES Table page 2-1 Means and standard errors for soil and site variables by community series ......................60 2-2 Means and standard errors of soil and site variables by comm unity association. .............61 2-3 Common woody shrubs in midstory a nd understory strata by association ........................62 2-4 Indicator species of Dry uplands and Mesic flatwoods associations .................................65 2-5 Indicator species of Wetland associations .........................................................................68 3-1 List of variables included in RDA and partial RDA ordinations .......................................97 3-2 Monte Carlo tests of canonical axes for all RDA and partial RDA ordinations ................99 4-1 ANOVA tables for models of species richness and stem density ....................................133 4-2 Results of Monte Carlo permut ation tests for RDA ordinations ......................................134

PAGE 9

9 LIST OF FIGURES Figure page 2-1 Physiographic landforms of Florida ...................................................................................71 2-2 Florida plot locat ions by associations ................................................................................72 2-3 Non-metric multidimensional ordi nation of Florida species data ......................................74 3-1 Venn diagrams of va riation partition model ....................................................................100 3-2 Biplots of RDA ordination c onstrained by edaphic variables .........................................101 3-3 Biplot of partial RDA ordinati on constrained by edaphic variables ................................102 3-4 Biplots of RDA and pRDA c onstrained by climate variables .........................................103 3-5 Contour maps derived from constrained or dination axis scores displaying geographic variation in variation par titions from the model of environmental-compositional correlations .................................................................................................................. .....104 4-1 Abita Creek Preserve : (a) pre-treatment in 1997, (b) immediately after logging in 1998, and (c) after logging and fi rst prescribed fire in 2000 ...........................................135 4-2 Least squares means and sta ndard errors of small stems/1000-m2 sample ......................136 4-3 Least square means and standard errors of species richness by treatment and year ........137 4-4 PCA ordination of ACP species data (1000-m2 scale).....................................................138 4-5 Constrained RDA ordinations of prelogged vs. first post-year species data ..................139 4-6 Constrained RDA ordinations of ACP species: all sample years ....................................140 4-7 Number species per log sample area (m2): mean species counts from ACP treatments vs. Penfound and Lake Ramsey species richness. ..........................................141 4-8 Successional trends of ACP species da ta compared to Penfound and Lake Ramsey reference site data ........................................................................................................... .142

PAGE 10

10 Abstract of Dissertation Presented to the Graduate School of the Un iversity of Florida in Partial Fulfillment of the Requirements for th e Degree of Doctor of Philosophy FLORISTIC AND ENVIRONMENTAL VARIATI ON OF PYROGENIC PINELANDS IN THE SOUTHEASTERN COASTAL PLAIN: DESC RIPTION, CLASSIFICATION, AND RESTORATION By Susan Catherine Carr December 2007 Chair: George Tanner Cochair: Kevin M. Robertson Major: Wildlife Ecology and Conservation Until recent times, the landscape of north a nd central Florida was dominated by firedependent pineland savanna vegetation with sparse canopies of longleaf pine ( Pinus palustris ). Economic development coupled with fire suppr ession lead to the dr astic declin e in the distribution and integrity of thes e natural communities. I present a vegetation classification of natural pineland communities in this highly fr agmented landscape based on data collected over large gradients of environmental and geological va riation. I collected fiel d data that quantified species composition and abundance from 293 plots (from 103 sites) distributed throughout the northern two thirds of Florida. After omission of species that occu rred in < 3% of plots, a total of 677 plant species were used in numerical analyses. I deve loped a vegetation classification based on floristic similarity using K-means cl uster and indicator species analyses. Three ecological series were described corresponding to idealized moisture conditions. These were further divided into 16 species associations. Floristic variat ion was related to geographic separation between the panhandle and peninsula re gions of Florida. I hypothesized that the numerous plant species that have limited di stributional ranges contribute to compositional

PAGE 11

11 patterns. Similar geographic trends were apparent in a model of compositional variation related to environment and spatial variation. Local environmental factors, including location on a topographic/moisture gradient and soil fertility, were important correlates of local floristic variation. Regional variation wa s correlated with soil texture and nutrient availability. A much greater proportion of the expl ained variance was provided by e nvironmental variables than by pure spatial variables. The mode l revealed that both regional factors (climate, edaphic, and geographic) and local factors (topographic position, soil chemistry) were correlates of with floristic variation. In addition to spatial variation, natural pineland communities under go temporal variation in response to periodic fires and changes in timber stand structur e. Central questions regarding ecological restoration of Coastal Plain pinela nds are: how resilient are these communities following anthropogenic alterations? Will ecologi cal restoration affect vegetation succession within the range of “nat ural” temporal variation? I studied ground cove r vegetation response to removal of woody biomass and reintr oduction of natural fire regimes as it related to a program of ecological restoration in a degrad ed pine savanna remnant. Tr eatment plots were thinned for timber or else un-thinned as a control. Prescribed fire was applied at tw o subsequent times, and changes in species composition were monitored ov er an eight year peri od. Species richness was enhanced by mechanical woody reduction in the fi rst two years, compared to sites that were burned only (not logged). This response largel y reflected increases in detectable graminoid species. However, species richness of treatment s converged within eight years, following two prescribed fires 3 years apart. Species co mposition responded similarly, converging between treatments over time. Succession was toward pre-settlement conditi ons, as suggested by comparisons to reference sites a nd historical data. Community composition appears to be robust

PAGE 12

12 to temporary alterations in fire regime and change s in timber stand structure within the range of conditions studied. Spatial variation in speci es composition of pineland communities may be relatively stable over time. W oody biomass reduction via carefu l mechanical logging does not appear to adversely affect pi neland vegetation recovery, and may expedite overall community restoration.

PAGE 13

13 CHAPTER 1 INTRODUCTION Pinelands of the Southeastern Coastal Pl ain are exceptional, both for their overall biodiversity and degree of biotic endemism. The combination of large climatic gradients, long growing seasons, variable geology and large speci es pools creates a prime environment for high local and landscape scale floristi c variation. Large compositional variation has been documented both across geographic and local gradients (Peet and Allard 1993, Bridges and Orzell 1989). Such high degrees of alpha, beta and gamma diversity (sensu Whittaker 1967) belie the exceptional habitat specialization and regionalization of many sp ecies of the Coastal Plain. Seemingly imperceptible topographic-moisture grad ients coincide with al most complete changes in plant species composition (Pee t and Allard 1993). At larger scales, there is evidence of regionalization of biota concurrent with geology, physiography, soils, and historical biogeography. Accordingly, biogeographers have recognized and delineated distinct “ecoregions” in Florida based on differences in environmental conditions and vegetation patterns (Davis 1967, Brooks 1982). In addition to biodiversity variation related to local and landscape gradients, levels of endemism are exceptionally high in the Southeas tern Coastal Plain, and in Florida pinelands specifically. Sorrie and Weak ley (2002) report over 1600 taxa of plants endemic to the Southeastern Coastal Plain. In addition to wide -ranging Coastal Plain endemics, many “narrow” endemic species inhabit very rest ricted geographic regions and/or habitats. Florida is notable both for the number of narrow endemic species it harbors, and the number of “centers of endemism” (sensu Sorrie and Weakley 2002) locate d within the State, pa rticularly relative to plant species. In addition, Flor ida is home to over 2500 native plant species, many of which are restricted in range of distribution, or habitat specificity (Wunderlin 2000).

PAGE 14

14 Pine savannas and woodlands native to th e Southeastern Coasta l Plain are among the most imperiled ecosystems in North Amer ica (Walker and Peet 1983, Croker 1987, Noss 1988, Frost 1993, Peet and Allard 1993). Although th ey once dominated the landscape, native pinelands now occupy less than three percent of their former range (Frost 1993, Outcalt and Sheffield 1996). Of this, an even smaller ar ea contains vegetation composition and structure similar to that of pre-settlement conditions (Simberloff 1993). The rapid range reduction of longleaf pinelands coincided with extensive logging, agricultural land use, and expanding rural settlement in the 19th and 20th centuries (Croker 1987, Frost 1993). Most contemporary native pinelands are sma ll and fragmented, and are no longer subject to the natural processes under which constituent species evolve d. Most notably, this precludes the natural occurrence of frequent low intensity fires that historically swept across the landscape (Frost 1993, Simberloff 1993, Platt 1999, VanLear et al. 2005). Fire suppression of longleaf pine natural areas has contributed to large scale species replacement, as less fire tolerant pines and hardwoods invade these pyrogenic communities (Glitzenstein et al. 1995, Platt 1999, Provencher et al. 2000, VanLear et al. 2005). In the absence of frequent fire, thick growths of woody plants compete with herbaceous vegetation for light and other resources, affecting succession and community structure (Brockway and Lewis 1997, Pr ovencher et al. 2001, VanL ear et al. 2005). Vegetation classification pl ays a key role in many areas of conservation, land management and scientific research. Classifica tion of vegetation delimits the number of relevant natural communities to provide a conceptual fram ework for understanding the natural variation. The process of delimitation is subjective by nature. Much of this subjectivity resides in deciding which data to use, what quantitat ive methods to use, and how to in terpret the resul ting solutions.

PAGE 15

15 Some vegetation classifications of large landscapes also explicitly incorporate information about geography (Peet and Allard 1993, Newell and Peet 1998, Wimberly and Spies 2001). To date, classification systems develope d specifically for Southeastern pineland communities have been quantitatively rigorous but local in scope, or wide-ranging but qualitative. Examples of the fo rmer are ecological classifications of vegetation are limited to a specific management areas, usually on the scale of several thousand hectares (Carter et al. 1999, Grace et al. 1999, Goebel et al. 2001, Abella and Shelburne 2004). Additionally, many vegetation descriptions of Southeastern plan t communities only describe woody species (e.g., Harcombe et al. 1993, others), thus missing import ant floristic “information” residing in the ground flora (e.g. Bridges and Orzel l 1989, Peet and Allard 1993, De Coster et al. 1999, Schmitz et al. 2002, Drewa et al. 2002). Classifications that do includ e explicit descriptions of herbaceous vegetation are generally subjective a nd anecdotal in nature. The exception is the quantitative treatment of Peet and Allard (1993) which includes a regi onal classification of pineland vegetation of the Coastal Plai n emphasizing ground cover vegetation. Traditionally, ecologists have studied the di stribution of plant species according to environmental factors (Bray and Curtis 1957, Peet 1978, Newell and Peet 1998). However, recent studies have underscored the need for spatially explicit models of environmentalcomposition variation (Legendre and Fortin 1989). Sp atial trends are relevant in such models for three reasons: 1) failure to account for spatially auto -correlated response da ta leads to biased interpretations of environmental effects, 2) environmental determinants of vegetation composition may be spatially structured, and 3) spatial autocorrel ation independent of environment suggests other control mechanisms of community composition (Legendre 2005).

PAGE 16

16 Little is know about temporal variation in composition of pyrogenic pineland vegetation, particularly compared to spatial variation. Stud ies of grasslands of ot her regions suggest that temporal variation in species distributions is la rge compared to species-area relationships (Adler et al. 2005). Longitudal studies of pineland community structure are rare and generally address successional responses to specific treatments. Na tural pinelands temporar ily altered by unnatural fire regimes and forest structur e also provide opportunities to study the resiliency of pineland vegetation to such alterations by quantifying co mmunity responses to re storation of natural conditions (Walker and White 2006). A better unde rstanding of temporal changes under typical and degraded conditions will contribute to the applied models of pineland restoration, as well as the models of “natural” variati on used by conservationists. I present a two-step process of vegetation classification and description of Florida pyrogenic pineland flora. First, I classify pi neland communities based on vegetation data alone (primarily herbaceous ground cover vegetation). Second, I present a model of environmentalcomposition correlations in a spatially explicit contex t. Finally, I present re sults of an ecological restoration program of a degrad ed pineland remnant, and interpret vegetation responses in the context of life history traits. The classification of pyrogenic Florida pine land vegetation is ba sed on 293 vegetation plots (58.6 ha total) collected over a broad range of environm ental conditions throughout the range of longleaf pine in panhandle and peninsul ar Florida. The classification, derived from floristic data alone, is presented as a system that can be used in land survey and management. Sixteen community associations are described by environmental characteristics, diagnostic and indicator species, general appearance and landsca pe context. I discus s floristic differences

PAGE 17

17 between community associati on, and how range-restricted a nd endemic taxa influence community variation. A spatially explicit model of environmental and historical determinants is presented with regard to the composition and diversity of pyrogen ic pineland vegetation. Environmental factors included edaphic, topographic, and climate variab les, presumed to be operating at different spatial and temporal scales. Va riation related to pur e spatial autocorrelat ion is hypothesized to be indicative of biotic processes (not related to environmental determinants). Biogeographic patterns were assessed by testing an “ecoregi on” hypothesis of regionalization of community variation. Significant environmental-composition correlations were used to generate hypotheses regarding controls of co mmunity variation in Co astal Plain pinelands. An experimental and long itudinal study of ecological restoration underscored the resiliency of pyrogenic pineland groundcover plant communities. Changes mediated by restoration treatments affected succession toward de sired reference conditions Furthermore, this study suggested stability in succession even in at ypical conditions of long fire-free intervals. Temporal dynamics in pineland plant communities is quite variable (as are spatial trends), and it is hypothesized that life history adaptations of typical plant sp ecies buffet the community over a range of atypical environmental conditions. As part of the restoration program, tree stand st ructure and frequent fi re were restored in a degraded pineland savanna remnant to resemble pre-settlement conditions. I measured the effects of two restoration treatments on co mposition and diversity of native ground cover vegetation. Restoration was meas ured as changes in composition relative to that of reference sites which represented desired re stored conditions. The larger question involved resiliency of a specific pineland community to temporal changes in fire regime and timber stand structure.

PAGE 18

18 From that, it may be concluded that pyroge nic pineland communities in general might be relatively stable over time and over a range of conditions. Secondly, the study demonstrated resiliency of native pineland vegetation following decades of man-induced fire suppression and contributes to predictions of restoration success relati ve to starting conditions.

PAGE 19

19 CHAPTER 2 A VEGETATION CLASSIFI CATION OF FLORIDA’S PYROGENIC PINELANDS Introduction Fire-dependent pineland vege tation once dominated the land scape of the Southeastern Coastal Plain, ranging from southern Virginia south to the tip of Florida a nd westward to eastern Texas. Frequent fires perpetuated the open as pect of pine savannas and woodlands, promoting development of species-rich herb aceous ground cover vegetation. It is estimated that prior to European settlement of the Gulf and lower Atlantic Coastal Plain regions, fire return intervals in upland pinelands averaged once per 2-3 years (Martin et al. 1993, Ols on and Platt 1995, Platt 1999, Glitzenstein et al. 2003). Fo llowing disruption of fire re gimes, these communities are rapidly colonized by fire-intol erant woody growth, prompting dr astic alteration of community composition and dynamics (Glitzenstein et al. 1995, Platt 1999, Glitzenstein et al. 2003). Economic development removed native pineland vege tation from much of its former range in the Coastal Plain, particularly from the finer-texture d soils that readily suppo rt agriculture (Frost 1993, Frost 2006). Native longleaf pinelands curre ntly occupy less than th ree percent of their former range (Frost 1993, Outcalt and Sheffiel d 1996). Even rarer are Coastal Plain pineland communities managed with fire regimes that mimic those of pre-settlement conditions (Simberloff 1993, Varner et al. 2005). Fire-dependent pineland communities Flor ida are exceptional both for their overall biodiversity and the degree of biotic endemism Over 1600 plant taxa are endemic to the Southeastern Coastal Plain, and over 250 of these are endemic or near-endemic to Florida (Ward 1979, Kautz and Cox 2001, Sorrie and Weakley 2001, Sorrie and Weakley 2006). The Florida peninsula has a complex geologic history of inun dation and land expansion related to sea level change and glaciation. Ancient is lands isolated during sea level ri se gave rise to many endemic

PAGE 20

20 species of contemporary highlands and ridge pr ovinces and other regions served as glacial “refugia” (Webb 1990). Florida is notable for the number of “centers of endemism” (sensu Sorrie and Weakley 2002) located in the State. More than 2500 plant sp ecies are native to Florida, representing a mixture of temperate and tropical specie s that changes with latitude (Holdridge 1967, Ward 1979, Wunderlin 1998). The combination of large clim atic gradients, long growi ng seasons, variable geology and large species pools in Florida creates a prime e nvironment for exceptional floristic variation at local and landscape scales. Florida has the th ird richest flora of a ll States (Wunderlin and Hansen 2000). Plant species richness of Florid a pinelands are among the highest recorded at small scales (Walker and Peet 1983, Peet 2006). In addition, subtle topographic-moisture gradients can harbor almost complete turnover in plant species composition (Bridges and Orzell 1989, Abrahamson and Hartnett 1990, Peet and Allard 1993, Platt 1999). Such a high degree of “beta” and “gamma” diversity (sensu Whittake r 1962, 1967) belies the exceptional habitat specialization and regionalizati on of many pineland species. On a landscape scale, there is evidence of aggregation in floris tic and community similarity asso ciated with specific regions. Accordingly, Florida “ecoregions” have been re cognized and delineated based on similarity of edaphic, geologic, physiognomic, and vegeta tive features (Puri and Vernon 1964, Davis 1967, Brooks 1982, Brown et al. 1990). Floristic classification system s provide a conceptual framew ork for understanding natural variation across environmental and geographic grad ients. Such systems are widely applied in ecological inventory, conservati on, and management (1990, Grossm an et al. 1998, Comer et al. 2003). To be useful in the field, a vegetation classification should provi de detailed information regarding frequent and abundant species, as we ll as those that are diagnostic of specific

PAGE 21

21 associations (i.e., “indicator species”; Defrene and Legendre 1997). Ideally, a classification would also describe relevant environmental attr ibutes, including typical ranges of variation. A comprehensive account of floristic types and variation could aide ecological restoration programs by providing a range of reference conditions and guidi ng land conservation priorities (White and Walker 1997, Walker and Silletti 2006). Vegetation classification based on quantitative data is very much dependent on sampling design, intensity, and breadth (Nekola and Wh ite 1999, Cooper et al. 2006). Random and areaproportionate sampling designs are often not pract ically possible in large regions containing fragmented landscapes with variable natura l conditions and mixed land ownership, land use history, and degree of public acce ss. However, subjective bias can be minimized by application of a stratified sampling design which promotes balanced sampling intensity and effort across gradients of interest (Leps and Smilauer 2007). Such a design may not yield an unbiased representation of variation of pre-settlement natural vege tation, but may facilitate a representative sample of contemporary natural vegetation in a highly modified landscape (e.g. most of Florida). Classification systems differ in many respec ts, including geographic and environmental scope, and type and quality of input data. Many vegetation classifications are strictly qualitative and descriptive (FNAI 1990, Grossman et al. 199 8, Comer et al. 2003), although widely used for community classification and cons ervation policy guidelines in Flor ida. These works are based on expert accounts of floris tic variation over a large re gion. Conversely, quantitative classifications typically incorporate site sp ecific vegetation data. Depending on program objectives, abiotic environmental at tributes are either explicitly included in the classification or are presented as descriptors or explanatory factors of floristica lly defined types. “Ecosystem

PAGE 22

22 classification” and “ecological landtype phases” t ypify the former approach (Cleland et al. 1993, Hix and Pearcy 1997, Goebel et al. 2001, Abella et al. 2003). Regional ve getation classifications that include all or part of Flor ida are of the latter type, base d on quantitative data of species abundances collected using sta ndardized sampling methodology (Peet and Allard 1993, Peet 2006). In the present study, abio tic variables are descriptors of community classifications, including soil properties and geol ogy. The quantitative delineation of floristic data approach has several advantages: 1) it encourages objectiv ity in classification partitioning, 2) it allows a posteri examination of relationships between abiotic variables and community types, which can be useful for inventory and predictive modeling; 3) it may uncover “unexplained” gradients of floristic variation, stimulating generation of hypot heses regarding determin ants of biodiversity (McCune and Grace 2002, Leps and Smilauer 2003, Legendre et al. 2005, Leps and Smilauer 2007). I present a quantitative classification of fi re-adapted pineland vegetation of northern and central Florida. The study region includes the entire historic range of longleaf pine in Florida. My focus was the classification of natural commun ities: i.e. frequently burned (at least 2-3 times over the past two decades) vegetation of pine lands and associated communities relatively unaltered by soil disturbance or severe fire s uppression. My objective wa s to characterize plant communities based on floristic assemblages alone, followed by descriptions of geographic distribution, topographic context, and soil characteristics. Co mmunity descriptions include identification of dominant and diagnostic plan t species, facilitating eas y field recognition of characteristic vegetation. My sampling design, coupled with an objective approach to cluster analysis, yielded a comprehensive yet manageable classification of 16 asso ciations. I describe edaphic and landscape features th at are useful for field identif ication, such as soil texture

PAGE 23

23 attributes and landscape context. Furthermore, I describe geog raphic and environmental trends in floristic similarity among pinela nd associations as they relate to distribution and identification of community types. Methods Study Area The study area included the entire Florida Pa nhandle and most of central and northern Peninsular Florida. This ar ea extends south from the Stat e border to a southern boundary extending from roughly 26 70’ latitude on the west coast to 28 80’ on the east coast (Figure 21). This area roughly coincides w ith the current range of longleaf pi ne in Florida (Figure 2-2(a)). This range is thought to repres ent the historic longleaf pine range in Florida (Platt 1999 and references within), although ther e is some evidence that histor ic distribution extended farther south. The southern boundary also approximates the southern extent of the “warm temperate moist forest” bioclimate zone, separating it from the “subtropical moist forest” zone (Holdridge 1967). Three generalized land units of Puri a nd Vernon (1964) subdivide the Florida study region according to common geologic history. Th ese generalized land units describe geographic regions: 1) Northern Highlands, 2) Central Highlands, and 3) Co astal Lowlands (Figure 2-1). These are further subdivided according to physiogr aphic landforms, which describe major soil types, geology and prevailing landscape feat ures (Puri and Vernon 1964, Myers 1990). These are 1) Highlands; 2) Ridges, Hills, Inclines a nd Slopes; and 3) Lowlands, Gaps, and Valleys. The Northern Highlands of the upper panha ndle lie north of a prominent ancient Pleistocene shoreline known as the Cody Scarp (Myers a nd Ewel 1990). This region is distinguished by broad expanses of conti nuous highlands. The We stern and Tallahassee

PAGE 24

24 Highlands, New Hope and Grand Ridges, and Marianna Lowlands landforms comprise the Northern Highlands land unit (Puri and Vernon 1964). The first two have dissected topography and clastic sediments of mainly Appalachian or igin from the Miocene epoch (20 to 5 million years before present; Puri and Vernon 1964, Br own et al. 1990, Myers 2000). The Marianna Lowlands landform contains outcrops of Eo cene and Oligocene carbonates in a low lying anticline (Puri and Vernon 1964, Brown et al. 1990). Although lower than the first two landforms, it is higher than the Coastal Lowlands and is generally well-drained owing to sandy soils shallowly overlying limerock perforated by sink holes (Brown et al. 1990). Ultisols are common upland soils of Northland Highlands, although Entisols typify Citronelle Formation uplands in the Western Highland portion as well as the sandy uplands of central panhandle Ridges. The Central Highlands land unit contai ns discontinuous highl ands of the central Peninsular ridge system amid lower and fla tter landforms (Figure 2-1). The former are landforms of the Ridges, Uplands, and Slopes and Highlands types while th e latter are Lowlands, Gaps, Valleys and Plains (Puri and Vernon 1964). The Central Highlands and the Northern Highl ands approximate the emergent portion of the Wicomico shoreline, an early Pleistocene shor eline of high sea level. This region was once an integrated highland that has since been pa rtitioned by erosion and solution (Puri and Vernon 1964). The Ridges and Uplands of the peninsula arose from ancient shorelines, dune systems, barrier islands, and associated terraces (Puri and Vernon 1964). Larger ridge systems of the Central Highlands include the Brookville, Dela nd, Trail, Mount Dora and Lake Wales Ridge physiographic landforms, and major Uplands incl ude Sumter, Polk, Marion, Duval and Lake landforms. Soils are mainly coarse, excessively drained Entisols and loamy Ultisols. Soils of

PAGE 25

25 Lowlands landforms are typica lly Spodosols underlain by limest one of the Florida peninsula platform (Brown et al., 1990). The Coastal Lowlands land unit includes the southern tier of the panhandle below the Cody Scarp, in addition to the coastal regions of th e peninsula (Figure 2-1). Much of this region has been subjected several marine inundations during the Late Miocene to the Early Pliocene (Puri and Vernon 1964, Webb 1990). Most of the Co astal Lowlands region contains Lowlands, Gaps, Valleys, and Plains physiographic landforms. These are broad plains with little relief, containing poorly drained S podosols (Brown et al. 1990). Selection of Sample Sites The focus of this study was fire-dependent plant communities of Florida containing herbaceous-dominated ground cover vegetation. Th is included many types of pine woodlands and savannas, variously labele d pine flatwoods, sandhills, hi gh pine, piney woods, mesic flatwoods, wet flatwoods, and scrubby (or xeric) fl atwoods. Also included were fire-dependent herbaceous dominated communities associated with pinelands, such as prairies, bogs, lake margins, and seepage slopes. These communities are naturally characterized by frequent, lowintensy fires in which herbaceous vegetation and litter provide the dominant fuel matrix (Platt 1999). I omitted scrub and maritime pinelands of Central Florida and coastal regions, which are typically characterized by crown fi res in the shrub or tree layers and have relatively longer firefree intervals (Myers and Ewel 1990). Although the Florida range of longleaf pine is the large scale region of interest of this study, descriptions of pyrogenic communities were not restricted to l ongleaf pine dominated sites. The geographic and hab itat scope of this study included all pineland and associated communities within the longleaf pi ne range of Florida. Sites lacking pine overstory were

PAGE 26

26 included in the study based on their similarity in ecosystem processes and herbaceous ground cover structure and diversity to pine-dominated s ites. Such sites often represented topographicmoisture extremes in otherwise pine-dominated landscapes. The generalized physiographic landforms of Puri and Vernon (1964) were further subdivided into “ecoregions” to guide site se lection and stratification. This ensured a representative sample of physiographic environm ents throughout the area of study. I delineated ecoregions based on homogeneity of geology, ve getation, soils, climate and physiography, following several published works (Fenneman 1938, Puri and Vernon 1964, Davis 1967, Fernald 1981, Brooks 1982, Bailey et al. 1994, Griffith et al 1994). There were a total of 19 ecoregions in the study region. I present classification result s relative to physiograp hic landforms, of which ecoregions were subsets. I stratified sampling by ecoregions and topogr aphic-moisture conditions. Roughly equal numbers of sites were selected per ecoregion depe nding on site availability and accessibility. To the best of my ability, I selected three high qual ity sites in different locations within each ecoregion. Ideally, each site contained an inta ct, continuous topographic-moisture gradient supporting frequently burned native vegetation. Unfortunately, sites that satisfy this condition are rare or absent in some regions, particularly th ose that lack large tracts of public land. Under these conditions, I relaxed selection criteria to include: 1) sites that contained intact topographicmoisture gradients, but lacked optimal fire histor y, and 2) sites with acceptable fire history but lacking intact gradients. In th e latter situation, I pieced togeth er a representative topographicmoisture gradient from several sites located in close proximity. Additional criteria were considered in site selection: 1) little or no recent man-made ground disturbance, 2) absence of invasive exotic species, 3) presence of na tive canopy and midstory tree composition and

PAGE 27

27 structure, and 4) evidence of fi re within the previous five year s, and preferably a history of frequent fires during the previ ous 50 years. In general, th e integrity of the ground cover vegetation was emphasized over structure of the tr ee canopy in selection ev aluations. Candidate sites were identified from various sources, including the Florida Natural Areas Inventory natural community database (FNAI 2000a) and consul tation with regional natural resource professionals. Three sites (12 pl ots) were selected in South Georgia (within 20 miles of the Florida state border). I assumed that vegetation of these sites we re representative of Florida pinelands in the same ecoregion. A total of 102 sites were selected (see Appendix A) Field Methods Once deemed suitable for sampling, a site (or a composite site) was delineated into three or four topographic-moisture zones based on field observations. Sampling from a range of topographic-moisture conditions maximized inclus ive sampling of local vegetation associations presumed to be associated with specific soil conditions. One 1000 m2 rectangular plot was established in each zone such that the plot area encompassed an area of relatively homogenous vegetation. The starting point of the long plot axis was randoml y assigned. Usually the main axis of the 50 x 20 m plot was orient ed parallel to sl ope contours. Vegetation sampling methodology followed th e Carolina Vegetation Survey (CVS) sampling protocol (Peet et al. 1998). The basic sampling unit was a 1000 m2 plot (dimensions 50 x 20 m). Four 100-m2 “modules” were situated in each plot each containing two sets of nested sub-plots (0.01, 0.1, 1, and 10-m2). All vascular plant taxa were recorded as they were encountered in the sequentially sampled nested s ub-plots. I estimated the aerial cover of each taxon in 100-m2 modules using cover classes: 1 = 01%, 2 = 1-2%, 3 = 2-5%, 4 = 5-10%, 5 = 10-25%, 6 = 25-50%, 7 = 50-75%, 8 = 75-95%, 9 = >95%. Mean cover estimates were

PAGE 28

28 calculated from four module cover midpoints. Taxa encountered in the remaining 600-m2 plot area were tallied and assigned nomin al cover estimates. In the 1000-m2 plots, all woody stems > 1 cm and < 40 cm diameter at breast height (dbh) were tallie d by species and 5 cm diameter class. Stems > 40 cm dbh were measured and reco rded individually. In plots with very sparse woody vegetation, I sampled stems in a larger area (2000-m2) to obtain better estimates of stem density and basal area. All plots were sampled during the late summ er though early winter (August-December). Sampling flora in the late growing season increas ed my ability to identify the copious numbers of graminoids and fall-flowering forbs typical of Southeastern pinelands. A total of 293 plots were sampled over 4 years (2000 – 2004). The majority of sampled taxa were identified to species or variety. Some taxa received lower levels of taxonomic resoluti on due to problems with consiste nt field identification. Where variation in taxonomic resolution existed, I used the lowest resolution necessary to ensure consistency throughout the dataset. The term “sp ecies” is used to indicat e the highest resolution of identification, be it genus, speci es or variety. Nomenclature generally follows Kartesz (1999) with a few exceptions. In field and herbarium plant identification I made frequent use of (Godfrey and Wooten 1979, Godfrey and Woot en 1981, Clewell 1985, Godfrey 1988, Wunderlin 1998, Weakley 2002). Approximately 2500 voucher specimens were deposited in the University of Florida herbarium in Gainesville, Florida. Four surface soil samples were collected per plot. Each sample of approximately 250 g was collected to 10 cm depth. Sub-soil sa mples were collected from a single point approximately 50 cm below ground surface. Samples were dried and sent to Brookside Labs in New Knoxville, Ohio for nutrient and textural analyses. Texture analysis determined

PAGE 29

29 compositional percentages of sand, silt, and clay part icles in the surface and sub-soil samples. In addition, percent organic matter, pH, and exchang eable cations in ppm (Ca, Mg, K, Na) were measured in surface soil samples. Numerical Analysis A matrix of species data was assembled from the 293 census plots hereafter referred to as samples. Samples represent different topographic-mo isture locations within sites. Pine species (genus Pinus ) were omitted from the species matrix, although other woody species were retained. Species with fewer than three occurren ces in were deleted from the final data matrix, as rare species contribute little to calculati ons of inter-plot similarities (McCune and Grace 2002). The dimensions of the final response matrix were 293 samples x 575 species. I transformed the species response matrix pr ior to multivariate analyses following the guidelines of Legendre and Gallagher (2001) and McCune and Grace (2002). First, species responses were relativized to maximum species cover values which tends to de-emphasize the influence of common and abundant species. Then the species response matrix was transformed using the Hellinger distance transf ormation. When used in conjunc tion with Euclidean distance metrics this transformation improves representati on of multidimensional data in low dimensional space and avoids problems inherent to sample we ighting (in chi-square based ordinations) in addition to problems associated with using Eu clidean distances with untransformed data (Legendre and Gallagher 2001, Legendre et al. 2005). I used a combination of ordination and clus ter analyses to partition samples into floristically similar groups. Specifically, I used non-hierarch ical Euclidean-based K-means cluster analysis to partition samples into a c onfiguration that minimized within group sum of squares relative to between group differences (Legendre and Legendre 19 98). Partitions are

PAGE 30

30 user-defined, so I used the “cascading K-means” function of the Vegan package (Oksanen et al. 2007) as implemented in R statistical software (R Development Core Team 2007). Cluster analysis was run multiple times using various numbers of user-defined partitions (2 to 40 groups). I selected the number of partitions that maximized an optimization index, specifically the “Simple Structure Index” (SSI). The SSI quantifies three elements of a partition model: maximum difference of each species response between clusters, the sizes of the most contrasting clusters and the deviation of speci es responses per cluster compared to its overall mean (Oksanen et al. 2007). The final partition model presented clusters of samples representing recognizable and distinct floristic assemblages. I refer to these cl usters as “associations”. I graphically displayed associations in a non-metric multidimensional sc aling (NMS) ordination of Euclidean distances derived from the Hellinger transformed species matrix. For this I used PC-ORD software, version 5.0 (McCune and Mefford 1999). Diagnostic species were recognized for each association, in terms of constancy and fidelity. I used Indicator Speci es Analysis of Dufrene and Legendre (1997) implemented in PCORD (McCune and Mefford 1999). The Indicator Va lue (IV) index quantifies a species’ relative frequency and abundance among associations. I ndicator species were identified using Monte Carlo randomization tests (McCune and Meffo rd 1999); the null hypothesis was that the maximum IV among associations is no larger than would be expected by chance. Indicator species were considered thos e with type I error < 0.05 in the IV randomization test. From the species recognized as indicators for associations, I identified those with restricted distributions in Florida. A species wa s identified as having “r estricted range” if its Florida distribution was limited to only one of th ree regions (Western Panhandle, Panhandle plus

PAGE 31

31 North peninsula, or Central Peninsula), or if its entire range was limited to Florida. Species’ Florida distributions were cat egorized by visual inspection of on-line county range maps available from the Institute of Systematic Botany Atlas of Florida website (Wunderlin and Hansen 2004). I compared soil characteri stics and other community attributes among individual associations, and between three higher level groups of associations (terme d ecological “series”). Within ecological series, means and pairwi se comparisons of response variables among associations were analyzed with univariate ANOVA’s. In addition to soil variables, I compared species richness (numbe r of species /1000-m2 sample) and basal area (m2/ha) between associations. Response variables were transformed to improve normality of residual distributions in each model. Count variables were log transformed, and logit transformations were applied to proportion res ponse variables (Tabachnick and Fidell 1996). I maintained a Type I error of p < 0.01 for each pairwise comparison to reduce the overall Type I error associated with each response variable. A ll ANOVA and post-hoc tests were performed using SAS software, version 9 (SAS 2000). Results A total of 293 samples spanning the study regi on were included in the K-means cluster analysis of mean species cover responses (Figur e 3a). I identified 16 associations from the optimal cluster solution. This pa rtition yielded the second high est value of SSI (0.23, maximum value = 1.0) among all partitions of 2 to 40 group s. Although the 28 group partition had a higher SSI value (0.25), I chose the 16 group partition b ecause it presented interpretable results with relatively balanced cluster sizes, with no clus ters containing fewer than four samples.

PAGE 32

32 The 16 associations encompass a wide range of floristic variati on over environmental conditions. The primary gradient of variati on, displayed by the first NMDS ordination axis, concurs with a priori assigned topographic-moisture cond itions (Figure 2). The correlation between distances in ordination space (two di mensional NMS solution) versus distances in original space was R2 = 0.83 (McCune and Mefford 1999, McCune and Grace 2002). The first axis represents most of this variation (R2 = 0.54). I categorized the 16 associations into three ecological series, which are superimposed on the ordination diagram: Dry Uplands (D), Mesic Flatwoods (M), and Wetlands (W). Associations were named using existing vernacular in plant comm unity descriptions: sandhills, clayhills and woodlands describe dry upland communities of varying canopy density and soil texture; mesic flatwoods refer to pine savanna communities of poorly-drained flat terrain. Occasionally to seasonally inundated wetlands are represented by various terms depending on canopy density and moisture conditions, includ ing wet flatwoods, wet prairies, and seepage slopes (FNAI 1990, Myers and Ewel 1990, Peet and Allard 1993). Modifiers were added to distinguish landscape and regional affinities. One hundred and six species were categorized as having restricted ranges in Florida. Eight species are endemic to Florida. The re maining 98 species have provincial distributions, and are restricted to one of th ree regions in Florida: 1) Panha ndle only, 2) Panhandle and north peninusular Florida, and 3) peni nsula only (Tables 4 and 5). Associations are described below in terms of community aspect, soil characteristics, and species composition. Throughout, the tables and appendix are refere nced for the following: soil and community attributes (Tables 2-1 and 22), common canopy and midstory woody species (Table 2-3), indicator species of associations (Tables 2-4 an d 2-5), and frequent and abundant

PAGE 33

33 ground cover species (Appendix B). In addition, e ndemic and restricted-range indicator species are indicated in Tables 2-4 a nd 2-5. Maps of plot locati ons are shown in Figure 2-2. Physiographic and landscape attributes for associ ations are described, a nd follow the conventions of Figure 2-1. Labels and cluster sizes are noted following associati on name. Association descriptions are grouped into three major ecological series corresponding to Figure 2-3. SERIES 1: Dry Uplands Dry Uplands included six associations, which were categorized as sandhills, woodlands or clayhills. The Dry Uplands associations were located wi thin the Northern Highland and Central Highland generalized land units, primarily within the Ridges and Uplands physiographic landforms. In general, these associations occu rred on ridgetops and uppe r slopes in areas with topographic relief exceeding several meters. Soils of Dry Uplands were sandy and low in organic content. Compared to Mesic Flatwoods and Wetlands series, Dry Uplands sand content was high in surface soils and low in sub-soils. Soil pH was intermediate compared to other series (Table 2-1). The six Dry Upland associations exhibited ge ographic segregation relative to floristic composition. The Ochlochnee River basin in th e eastern Panhandle di stinctly separated associations of the Northern Highlands and Panhandle Coastal Lowlands from those of the Central Highlands and peninsular Coastal Lowla nds. Dry Uplands of the Northern Highlands landform occurred on both Pliocene and Pleist ocene deposits, including the Citronelle and Torreya formations and the undifferentiated depos its of the lower Apalachicola basin (Puri and Vernon 1964, Brown et al. 1990). East of the Ochlochnee River, Dry Upland associations occurred primarily on Miocene and Pliocene de posits of the Centra l Highlands land unit, specifically within the Ridges, Upland s and Slopes physiographic landforms.

PAGE 34

34 Dry Uplands soil properties reflected those of Entisols and Ultisols, which are common upland soil orders (Brown et al. 1990, Myers 1990, Myers 2000). Segregation of associations coincides with soil clay and silt content. The well-developed Ultisols of the PANHANDLE LONGLEAF PINE CLAYHILLS and PANHANDLE SILTY WOODLANDS had argillic sub-surface strata enriched with clay and silt. Soil moisture availa bility is typical greater in these soils (Brown et al. 1990, Brady and Weil 2000). The Dry Upland a ssociations of the pa nhandle spanned a range of soil texture composition. Conversely, Dry Up land associations of the peninsula did not exhibit surface soil texture gradie nts, but variation was apparent in sub-soil silt and clay content and organic content. All Dry Upland soils we re similar in pH, with the exception of the PANHANDLE LONGLEAF PINE CLAYHILLS association. In addition, fi ne-textured soil content was positively correlated with species richness and canopy density. A description of individual associations within Dry Uplands series follows. PENINSULA XERIC SANDHILLS (22 plots, D3): This asso ciation is restricted to high sandy ridges of the Central Highlands and Coasta l Lowlands of the northern peninsula region (Figure 2-2c). PENINSULA XERIC SANDHILLS soils consist of coarse sands with low concentrations clay and silt. This associati on is species-poor compared to other Dry Upland associations, although comparable to the PANHANDLE XERIC SANDHILLS further west. Pine canopy of PENINSULA XERIC SANDHILLS was sparse. Longleaf pine ( Pinus. palustris ) was the dominant canopy species (mean BA = 4.8 m2/ha), followed by turkey oak ( Quercus leavis ; mean BA = 1.9 m2/ha). Common midstory speci es included turkey oak, sand live oak ( Q. geminata ), saw palmetto ( Serenoa repens ) and bluejack oak ( Q. incana ). Sand post oak ( Q. margarettiae ), an oak common in other sandhill asso ciations, was notably infrequent.

PAGE 35

35 The most common herbaceous plants of PENINSULA XERIC SANDHILLS were grass and forb species. Frequent grasses are wiregrass ( Aristida beyrichiana ), lopsided indiangrass ( Sorghastrum secundum ), little bluestem ( Schizachyrium scoparium var. stoloniferum ), and eggleaf witchgrass ( Dichanthelium ovale ). The forbs silkgrass ( Pityopsis graminifolium ), pineland pinweed ( Lechea sessiliflora ) and queens delight ( Stillingia sylvatica ) were common. A few grasses were identified as indicat or species, including pineywoods dropseed ( Sporobolus junceus ), perennial sandgrass ( Triplasis americana) and big threeawn ( Aristida condensata ). The remaining indicator species were forb species common to xeric habitats: Ware’s hairsedge ( Bulbostylis warei) coastal plain honeycombhead ( Balduina angustifolia) and pineland pinweed ( Lechea sessiliflora ). Coastal plain chaffhead ( Carphephorus corymbosus ) and wholly pawpaw ( Asimina incana : a small shrub) are indicator species with ranges restricted to the peninsul a. Two indicator species are legumes: eastern milkpea ( Galatia regularis) and scurf hoarypea ( Tephrosia chyrsophylla ). Legumes are typically common to communities of finer-textured soils (James 2000). PANHANDLE XERIC SANDHILLS (31 plots, D4): Sites of this association were restricted to the Northern Highlands land unit (Figure 2-2b), primarily west of the Ochlochnee river basin. PANHANDLE XERIC SANDHILLS were observed in two landscape contexts: 1) on sandy ridgetops and upper slopes, and 2) as the dominant community of broad flat terrain with little apparent topographic variation. I observed the latter situation on the broa d continuous uplands of the Citronelle formation in Eglin Air Force Base. PANHANDLE XERIC SANDHILLS were similar in aspect to PENINSULA XERIC SANDHILLS. Sparse canopies consist of scattered longleaf pines ( P. palustris : mean BA = 7.9 m2/ha) and turkey oak ( Q. leavis : mean BA = 1.1 m2/ha). Midstory strata were dominated by turkey oaks,

PAGE 36

36 bluejack oak ( Q. incana ) and sand live oak ( Q. geminata ). Unlike the PENINSULA XERIC SANDHILLS, sand post oak ( Q. margaretta ), dwarf live oak ( Q. minima ), and dwarf huckleberry ( Gaylussacia dumosa ) were common in PANHANDLE XERIC SANDHILLS. Frequent species of PANHANDLE XERIC SANDHILLS included few grasses, most notably little bluestem ( Schizachyrium scoparium var. stoloniferum ) and Elliotti’s bluestem ( A. gyrans var. gyrans ) with a low frequency of wiregrass ( Aristida stricta ). Herbaceous species of xeric habitats distinguished PANHANDLE XERIC SANDHILLS ground cover. About a third of indicator species have ranges restricted to the Panhandle, including pi edmont gayfeather ( Liatris pauciflora var secunda) littleleaf milkpea ( Galactia microphylla) Morh’s threeawn ( Aristida morhii) Godfrey pineland hoarypea ( Tephrosia morhii) royal snoutbean ( Rhynchosia cytisoides) and greater Florida spurge ( Euphorbia floridana) The provincial herb Pityopsis aspera is abundant and frequent (90% of plots). In contrast, P. aspera is absent from PENINSULA XERIC SANDHILLS were P. graminifolia is usually dominant. NORTH FLORIDA SANDHILLS (31 plots, D2): Sites of this association occurred on various landforms of the Coastal Lowlands and Central Highlands of the eastern panhandle and northern peninsula (Figure 2-2c). NORTH FLORIDA SANDHILLS are usually found on ridgetops and upper slopes. Soils were sim ilar in textural composition to PENINSULA XERIC SANDHILLS except they had higher silt content. In addition, they were very low in clay and organic matter. Similar to other Dry Upland associations, NORTH FLORIDA SANDHILLS have canopies of longleaf pines (mean BA = 8.6 m2/ha) with scattered upland oa ks (most abundant: turkey oak, mean BA = 1.2 m2/ha). The three common upland oaks dominate the midsto ry: turkey oak, bluejack oak ( Q. incana ) and sand post oak ( Q. margaretta ).

PAGE 37

37 Common grasses of NORTH FLORIDA SANDHILLS are similar to those of PENINSULA XERIC SANDHILLS: wiregrass, little bluestem, lopsided indiangrass, and eggleaf witchgrass. Other frequent grasses include needleleaf witchgrass ( D. angustfolium ) and thin paspalum ( Paspalum setaceum ). Many frequent species are low growing fo rbs, as are 14 of the 15 indicator species. Three of these are legumes, in cluding Florida ticktrefoil ( Desmodium floridanum ), dollarleaf ( Rhynchosia reniformis ), and hairy lespedeza ( Lespedeza hirta ). None of the NORTH FLORIDA SANDHILLS indicator species have restricted distributi ons. Species richness of NORTH FLORIDA SANDHILLS is notably higher compared to the xeric Dry Uplands, and likely contributed to the floristic segregation among these associations. NORTH FLORIDA RICH WOODLANDS (11 plots, D1): This association includes longleaf pine woodlands of midand lowe r slopes in the Central Highlands and Coastal Lowlands of the northern peninsula (Figure 2-2c). These sites we re usually adjacent to hardwood hammocks. All NORTH FLORIDA RICH WOODLANDS sites were in or adjacent to vegetation zones identified as “Hardwood hammocks” by Davis (1967). Mo st were located downslope of NORTH FLORIDA SANDHILLS. NORTH FLORIDA RICH WOODLANDS soils had physical properties similar to the three preceding Sandhills associations. However, they were distinguished by their very high organic content and sub-surface clay content. These soil attribut es suggest higher water retention capacity (Brady and Weil 2000). Canopy densities of NORTH FLORIDA RICH WOODLANDS were high relative to other Dry Uplands associations (mean BA = 16.1 m2/ha). Longleaf pine domi nated canopies (mean BA = 8.5 m2/ha), but other subdominant pine sp ecies were present: slash pine ( P. elliottii var. elliottii ) and loblolly pine ( P. taeda ; mean BAs = 1.3 and 1.2 m2/ha respectively). Sand live oak ( Q. geminata ) and mockernut hickory ( Carya alba ) were canopy sub-dominants (mean BAs = 1.5

PAGE 38

38 and 1.4 m2/ha respectively). Midstory strata were generally shrubbi er compared to other Dry Upland associations, dominated by saw palmetto ( S. repens ) followed by winged sumac ( Rhus coppelinum ), mockernut hickory ( C. alba ) and two upland oaks ( Q. geminata and Q. margaretta ). Common herbaceous species of NORTH FLORIDA RICH WOODLANDS included grasses typical of other upland associa tions, as well as some distin ctive woodland forbs. Common grasses were Elliott’s bluestem ( A. gyrans var. gyrans ), needleleaf witchgrass ( D. angustfolium ), thin paspalum ( P. setaceum ), lopsided indiangrass ( Sorghastrum secundum ), broomsedge bluestem ( A. viriginicus ), and eggleaf witchgrass ( D. ovale ). Wiregrass ( A. stricta ) was present in only about 50% of the plots, and was sparse compared to other Dry Upland associations. Bracken fern ( Pteridium aquilinum ), laural greenbriar ( Smilax laurifolia ), whitetop aster ( Sericocarpus tortifolius) and lesser snakeroot ( Ageratina aromatica) were ubiquitous herbaceous species. Almost all indicator speci es were woodland forbs and infrequent grass species, including seven grasses and five legumes. About half of the indicator species were species with restricted ranges. Species ri chness of North Florida Rich Woodlands was intermediate compared to other Dr y Uplands (106 species/0.1 ha). PANHANDLE LONGLEAF PINE CLAYHILLS (14 plots, D5): These sites are restricted to the Northern Highlands land unit of the panhandle. PANHANDLE LONGLEAF PINE CLAYHILLS inhabit the ridgetops and upper-sl opes of dissected Pliocene and Miocene-aged sediments north of the Cody Scarp (Figure 2-2b). In the western panhandle, this associa tion occupied mid-slopes in association with PANHANDLE XERIC SANDHILLS. The prominence of fine-textured sediments distinguishes soils of PANHANDLE LONGLEAF PINE CLAYHILLS. Sub surface silt and clay content

PAGE 39

39 was high compared to other Dry Uplands. Sim ilarly, soil pH was higher than all other Dry Uplands associations. PANHANDLE LONGLEAF PINE WOODLANDS canopies were dense and dominated by longleaf pine (mean BA = 10.9 m2/ha), with minor contributions of loblolly and shortleaf pines ( P. taeda and P. echinata ; mean BAs = 1.2 and 0.4 m2/ha respectively). In addition to the typical Dry Uplands oak species, sh rubs of more mesic habitat o ccupy midstory strata such as southern red oak ( Q. falcata) running oak ( Q. pumila ), and Darrow’s blueberry ( V. darrowii ). Species richness of PANHANDLE LONGLEAF PINE CLAYHILLS is exceedingly high. The mean of 124.5 species/0.1 ha is the highest of all Dry Uplands associations. Dense herbaceous ground cover vegetation contai ned numerous forb and grass species. Wiregrass ( Aristida stricta), little bluestem ( Schizachyrium scoparium var stoloniferum ) and narrowleaf witchgrass ( D. angustifolium ) were ubiquitous. Indicator species included many legumes and composites (members of the Asteraceae family). Ten out of 25 indicato r species are legume species and many of these are in the genus Desmodium Eleven of the 25 indicator species have ranges that were restricted to the panhandle or northern peninsula. Several bunc h grass species with restricted ranges were identified as indicators: big bluestem ( A. gerardii ), cutover muhly ( Muhlenbergia cappilaris var. trichopodes), Carolina fluffgrass (T. carolinianus), yellow indiangrass ( Sorghastrum nutans ) and shortleaf skeletongrass (Gymnopogon brevifolius). PANHANDLE SILTY WOODLANDS (22 plots; D6): This asso ciation occupied Pleistocene and Miocene sediments of the Coastal Lowlands west of the Ochlochnee river basin (Figure 22b). Most sites were located in the Apal achicola embayment region, and many occupied Pleistocene and Holocene undiffere ntiated deposits of lowlands east of the Apalachicola river (Puri and Vernon 1964, Florida Department of Environmental Protection 1998). Although

PAGE 40

40 included in the Dry Uplands series, PANHANDLE SILTY WOODLANDS resembled Mesic Flatwoods in landscape context. They inhabited side slopes and terraces of intermed iate topography. Soils of PANHANDLE SILTY WOODLANDS were high in silt and clay content. Notably, subsurface soils had high silt and low organic content compared to other associations. Longleaf pine dominanted dense canopies of PANHANDLE SILTY WOODLANDS sites (mean BA = 11.9 m2/ha). Other canopy species were infreque nt (mean BA of all other species < 0.4 m2/ha). Upland oaks and other xeric midstory hardwoods were conspicuously absent. Low growing evergreen shrub species typical of mesic habitats dominated the midstory strata, including gallberry ( Ilex glabra ), running oak ( Q. pumila ), saw palmetto (Serenoa repens ), and dwarf live oak ( Q. minima ). Although woody vegetation of PANHANDLE SILTY WOODLANDS resembled Mesic Flatwoods, herbaceous ground cover wa s floristically similar to ot her associations in the Dry Uplands series. Mean species richness of PANHANDLE SILTY WOODLANDS was relatively high, comparable to NORTH FLORIDA SANDHILLS and NORTH FLORIDA RICH WOODLANDS associations. Wiregrass ( A. stricta ), little bluestem ( S. scoparium var. stoloniferum ), narrowleaf witchgrass ( D. angustifolium ), and cypress witchgrass ( D. dicotomum var. tenue ) were the most common grass species. Forb species were freque nt relative to grasses and shr ubs. All indicator species were herbaceous species; over ha lf were members of Asteraceae or Fabaceae plant families. Eight of 14 indicator species have ranges restricted to the panhandle or north peninsula, including two that are endemic to the Apalachicol a region: pineland false sunflower ( Phoebanthus tenuifolius) and scareweed ( Baptisia simplicifolia ).

PAGE 41

41 SERIES 2: Mesic Flatwoods This series includes three asso ciations that can be categori zed as either Mesic Flatwoods or Xeric Flatwoods according to FNAI conventio ns (FNAI 1990). Mesic Flatwoods associations typically inhabited flat poorly drained regions of the panhandle and peninsula Coastal Lowlands, and the peninsular Central Highlands (Fi gure 2-2d; Abrahamson and Hartnett 1990, FNAI 1990, Myers 2000). Mesic Flatwoods appeared to be ab sent from the Northern Highlands landform. Some geographic separation of Mesic Flatwoods associations was apparent, with complete separation of NORTH FLORIDA MESIC FLATWOODS and CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES. In general, Mesic Flatwoods soils are sa ndy and acidic. Mesic Flatwoods soils are typically Spodosols, with acidic sands underlain by clayey or organic hardpans hindering water percolation (FNAI 1990). Sub-surf ace clay content was consistently low in my Mesic Flatwoods sites (although sample depths may have been to sh allow to detect hardpans). I did not sample sub-surface organic matter. Orga nic content of surface soils was high compared to Dry Uplands and similar to Wetlands associations. Variat ion in soil texture between Mesic Flatwoods associations was minimal, alt hough the northern association tended to have higher surface soil clay content. Descriptions of specific associations follow. XERIC-MESIC FLATWOODS (36 plots; M1): XERIC-MESIC FLATWOODS generally inhabited the Coastal Lowlands of the panhandle and peninsula (Figure 2-2d). They occupied upper slopes of small sandy rises embedded in larg e expanses of Mesic Flatwoods vegetation. In the few sites of the Central Highlands, this association occurred dow nslope of Dry Upland vegetation.

PAGE 42

42 Soils of XERIC-MESIC FLATWOODS were coarse and contained very small concentrations of fine textured sediments. These differences distinguished XERIC-MESIC FLATWOODS soils from NORTH FLORIDA MESIC FLATWOODS. Organic content of XERIC-MESIC FLATWOODS soils is similar to other Mesic Flatwoods associat ions, although soil pH is relatively high. XERIC-MESIC FLATWOODS had sparse pine canopies and dense shrubby midstory strata. Sparse longleaf pine form ed the canopy (mean BA = 2.8 m2/ha). Slash pine ( P. elliottii ) was sub-dominant (mean BA = 1.0 m2/ha). Saw palmetto ( S. repens ) was by far the most abundant midstory shrub. Three upland “scrub” oaks were common in XERIC-MESIC FLATWOODS midstories: sand live oak ( Q. geminata ), Chapman’s oak ( Q. chapmanii ), and myrtle oak ( Q. myrtifolium ). Notably absent were the upland oaks of Dry Upland associations. In addition, evergreen shrub species of th e heath family are common in Xeric-Mesic Flatwoods, including fetterbush, Lyonia lucida ; dwarf huckleberry, G. dumosa ; lowbush blueberry V. myrsinites ). Mean species richness of XERIC-MESIC FLATWOODS sites was low due to sparse herbaceous ground cover. Wiregrass ( A. stricta ) was by far the most common herb, followed by broomsedge bluestem ( A. virginicus ), hemlock witchgrass ( D. sabulorum var. thinium) and silkgrass ( P. graminifolium: not a grass, but a member of Asteraceae ). Few indicator species were identified for Xeric-Mesic Flatwoods. Of thes e, 3 (out of 7) were shrub species, including two of the common scrub oaks and tarflower ( Befaria racemosa ). Three indicator species had restricted peninsular ranges: ta rflower, Chapman’s goldenrod ( Solidago odora var. chapmanii ) and shortleaf gayfeather ( Liatris tenuifolia var. quadriflora ). NORTH FLORIDA MESIC FLATWOODS (30 plots; M2): This as sociation was observed in the Coastal Lowlands land unit of the panhandle and peninsula. A few sites occurred in the Central Highlands land unit of the peninsula (Figure 2-2d) where it occupied small areas

PAGE 43

43 downslope of Dry Uplands. The northerly distribution of NORTH FLORIDA MESIC FLATWOODS separates it geographically from the CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES. Typically, NORTH FLORIDA MESIC FLATWOODS occupied flat, poorly drained terrain of Pleistocene origin. Soil clay content and pH was low compared to other Mesic Flatwoods associations. Overstory canopies of NORTH FLORIDA MESIC FLATWOODS were comparatively dense. Longleaf pine was the most comm on canopy species (mean BA = 9.3 m2/ha), and slash pine was sub-dominant (mean BA = 1.5 m2/ha). Midstory vegetation wa s generally low and sparse. Common shrub species were gallberry ( I. glabra ), saw palmetto ( S. repens ), and runner oak ( Q. minima ). These species formed patches of low grow th in the understory, in terspersed with thick herbaceous ground cover. Runner oak and another woody sub-shrub (hairy wicky, Kalmia hirsuta ) were identified as indicator species. NORTH FLORIDA MESIC FLATWOODS had intermediate species richness relative to other Mesic Flatwoods. Common herbaceous species resembled those of XERIC-MESIC FLATWOODS: wiregrass ( A .stricta ), broomsedge bluestem ( A. virginicus ), and silkgrass ( P. graminifolium ). Carolina yelloweyed grass ( Xyris caroliniana ) is the most frequent he rbaceous species, as well as an indicator species. Only four indi cator species were recognized for NORTH FLORIDA MESIC FLATWOODS. Other than those already mentione d, these include Florida dropseed ( Sporobolus floridanus ) and dwarf live oak. CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES (22 plots; M3): These sites were restricted to the Coastal Lowl ands land unit of the peninsula. CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES typically occupied broad flat, poo rly drained terrain with sediments of Pleistocene origin (Figure 2-2d). They often surrounded XERIC-MESIC FLATWOODS

PAGE 44

44 communities present on slightly higher and dr ier ridges. Soil characteristics of CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES were similar to XERIC-MESIC FLATWOODS, although clay content was slightly higher in the sub-soil. Canopies of CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES were either sparse or absent. The latter condition distinguishes the dry prairies of Centra l Florida as described elsewhere (FNAI 1990, Bridges 2006, Platt et al. 2006). I did not detect floristic differences between sites with and without pine canopy. Where canopy was presen t, longleaf pine ( P. palustris ) was dominant and the two slash pine varieties ( P. elliottii var. elliotii and P. elliottii var. densa ) were infrequent (Mean BAs respectively: 2.0, 0.5, and 0.2 m2/ha). Midstory vegetation was sparse in the frequently burned CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES. Woody species were relegated to the understory, were saw palmetto ( S. repens ) was abundant. Other understory shr ubs included dwarf live oak ( Q. minima ), gallberry ( I. glabra ), and fetterbush ( L. lucida ). Grass species were the most comm on species, such as wiregrass ( A. stricta ), hemlock witchgrass ( D. sabulorum var thinium) broomsedge bluestem ( A. virginicus ), bottlebrush threeawn ( A. spiciformis) and cypress witchgrass ( D. chamaelonche ). The latter two grasses were indicators of CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES. Two large bunchgrass species restricted to peninsular Florida were conspicuous indi cators: shortspike bluestem ( A. brachystachyus ) and stoloniferous little bluestem ( Schizachyrium stoloniferum ). Graminoid species comprise about a third of the indicator species (11 out of 27). Three indicator species were shrubs and the remainder was forbs. No legumes, and only one member of Asteraceae were identified as indicators. Five indicator species have ranges restricted to the Florida Peninsula, including two endemics.

PAGE 45

45 SERIES 3: Wetlands Associations of the Wetlands ecological series encompassed a divers e array of floristic variation and physiognomic settings, and spanne d most of the study area. Most Wetlands associations, particularly thos e of the Northern and Central Highlands land units, tended to occupy small areas of lower slopes in regions wi th relatively high relie f. Two associations restricted to the Coastal Lowlands land unit (the PANHANDLE WET FLATWOODS/PRAIRIES and PENINSULA WET FLATWOODS/PRAIRIES) inhabited poorly drained areas with little topographic relief. Historically, large expa nses of this vegetation extended across gradients of imperceptible elevation change (Harper 1914, Myers 2000). Geographic segregation among Wetlands associations is apparent. Thr ee of seven associations are rest ricted to the panhandle and one restricted to the central pe ninsula (Figure 2-2e, 2-2f) Associations of the Wetlands series show considerable variation in physical soil properties. Some associations were characterized by poorly drai ned and silty soils, where silt content distinguished Wetlands from Mesic Flatwoods and Dry Upland series. Texture differences were less apparent in sub-soil comp osition. Organic matter was high in all Wetlands associations. Fine textured soils and surface organic matter we re particularly abundant in panhandle associations. Soil pH was variable among associa tions, ranging from 4.2 to 5.3. Descriptions of specifi c associations follow. MARGINAL PRAIRIES (11 plots; W1): This associat ion included herbaceous dominated vegetation of depressional wetl and margins in the panhandle a nd northern peninsula (Figure 22f). MARGINAL PRAIRIES occupied concentric zones surr ounding shallow, seasonally inundated depressions and were seasonally in undated during periods of high ra infall. These sites occurred

PAGE 46

46 on the margins of various wetland types, includi ng dome swamps, sandhi ll upland lakes, and depression marshes (FNAI 1990). MARGINAL PRAIRIES surface soils were sandy and acidic, and contained high levels of organic matter. Sub-soil silt content was low but clay content was high. This may reflect the presence of subsurface “hardpans” or “clay lens ” that underlie depression marshes and dome swamps (FNAI 1990). Canopy vegetation of MARGINAL PRAIRIES was either absent, or was comprised of young pond cypress and swamp tupelo ( Taxodium ascendens and Nyssa sylvatica var. biflora ). The latter condition may reflect “i nvasion” of saplings followi ng fire suppression. Midstory vegetation was sparse or absent. A few evergreen shrubs was s poradically present, the most common were gallberry ( I. glabra ), titi ( Cyrilla racemiflora ) and sandweed ( Hypericum fasciculatum ). Herbaceous vegetation of MARGINAL PRAIRIES was low in aspect and diversity. Mean species richness is the lowest of all Wetland associations (49.6 sp ecies/0.1 ha). Few species had high constancy across sites, refl ected by the few species recogni zed as indicators. The most frequent herbaceous species (> 80%) were fi ve grasses and two weedy forbs: broomsedge bluestem ( A. virginicus ), slender flattop goldenrod ( Euthamia caroliniana ), pale meadowbeauty Rhexia mariana var. mariana ), and maidencane ( Panicum hemitomon ). Wiregrass ( A. stricta ) was absent from MARGINAL PRAIRIES. Maidencane, a species typical of inundated wetlands, was an indicator species. The remaining five i ndicators included forbs and one shrub species (buttonbush: Cephalanthus occidentalis ). PENINSULA WET FLATWOODS/PRAIRIES (16 plots; W2): These inhabit the flat, poorly drained regions of the Coastal Lowlands of peninsular Florida (Figure 2-2f). PENINSULA WET

PAGE 47

47 FLATWOODS/PRAIRIES occupied barely perceptible lowe r slopes in association with NORTH FLORIDA MESIC FLATWOODS. High water tables and seas onal inundation due to rainfall contribute to wet conditions of PENINSULA WET FLATWOODS/PRAIRIES (FNAI 1990, Myers and Ewel 1990). Soils were sandy and high in or ganic matter compared to other Wetlands associations. Soil characteristics were partic ularly pronounced in sub-surface horizons, where clay and silt content was low. Soil calcium c oncentration and pH was high compared to other Wetlands associations. PENINSULA WET FLATWOODS/PRAIRIES were variable in canopy structure and composition. Pine canopies were either sparse or absent. Midstory vegetation was typically low and sparse. The most common species were low growing shrubs: sandweed ( Hypericum fasciculatum ), gallberry ( I. glabra ), and pond cypress ( Taxodium ascendens ). Floristic similarity of un derstory vegetation united PENINSULA WET FLATWOODS/PRAIRIES sites as this association. All co mmon species, and 95% of indicator species were forbs and graminoids. Among fr equent herbs, over 80% were recognized as indicator species. Wetland herb species, such as water cowbane ( Oxypolis filiformis ), tenangle pipewort ( Eriocaulon decangulare ), pineland rayless goldenrod ( Bigelowia nudata ), blue maidencane ( Amphicarpum muhlenbergianum ) and Elliott’s yelloweyed grass ( X. elliottii ) were diagnostic in their constancy and fide lity. Wiregrass was frequent in PENINSULA WET FLATWOODS AND PRAIRIES, but was rivaled in abundance by other grass species, including longleaf threeawn ( A. palustris ). None of the 40 indicator sp ecies have restricted ranges. CALCAREOUS WET FLATWOODS (4 plots; W3): This association was an unusual wet flatwoods assemblage that inhabited sites with shallow subsurface limestone. My small sample size precludes much generalization; however, the f our plots were floristica lly distinct (Figure 2-

PAGE 48

48 2f). The sites occurred in two situations: 1) the coastal fringe of th e Big Bend region of the western peninsula, where marl is often immedi ately below the soil surface, and 2) as small inclusions in Coastal Lowlands, embedded in large expanses of CENTRAL FLORIDA MESIC FLATWOODS. Soil texture characteristics of CALCAREOUS WET FLATWOODS were similar to those of PENINSULA WET FLATWOODS/PRAIRIES except that sub-surface clay content was higher. CALCAREOUS WET FLATWOODS were basic and had exceedingly high calcium concentrations relative to other Wetlands associations. This was consistent with the presence of shallow soils overlying limestone outcrops. Three of the four CALCAREOUS WET FLATWOODS sites had dense canopies of slash pine ( P. elliottii ; mean BA = 11.15 m2/ha). Longleaf pine was absent. One plot, on Avon Park Airforce Range, had no pine overstory. Sabal palms ( Sabal palmetto ) were present in all sites, and comprised a significant canopy co-dominant (mean 1.88 m2/ha). Other hardwood species were common in the midstory, including many that are typical of swamp vegetation: wax myrtle ( Myrica cerifera ), swamp bay ( Persea palustris ), red maple ( Acer rubrum ), saw palmetto ( S. repens ), and sweetgum ( Liquidambar styraciflua ). Mean species richness of CALCAREOUS WET FLATWOODS was exceedingly high, rivaling that of Dry Uplands on fine textured soils. Wiregrass ( A. stricta ) was infrequent. Dominant grasses were redtop panicum ( Panicum rigidulum var. rigidulum ), sugarcane plumegrass (Saccharum giganteum), cypress witchgrass ( A. dichotomum var. nitidum ), and switchgrass ( Panicum virgatum ). Most common species and the majority of indicator species were forbs. In addition, many graminoids were r ecognized as indicator species, particularly member of the

PAGE 49

49 genus Rhynchospora Sabal palm was the only non-herb aceous indicator species. Of 35 indicator species, only one ( R. perplexa ) has a restricted range. NORTH FLORIDA SHRUBBY WET FLATWOODS (15 plots; W4): These sites occurred in the Northern and Central Highlands, and Coas tal Lowlands of the panhandle and northern peninsula (Figure 2-2f). Most sites were east of the Ochlochnee River basin with the exception of one site in the western panhandle. NORTH FLORIDA SHRUBBY WET FLATWOODS typically occurred as small fringes along lower slopes, ab utting wetland swamps. Textural composition of NORTH FLORIDA SHRUBBY WET FLATWOODS soils did not distinguish them from other Wetlands associations. However, surface soils we re high in organic content and acidic. NORTH FLORIDA SHRUBBY WET FLATWOODS have dense canopies of primarily slash pine ( P. elliottii var elliottii ; mean BA = 10.93 m2/ha). Longleaf and pond pine were minor components ( P. palustris and P. serotina ; mean BA = 2.63 and 1.18 m2/ha respectively). NORTH FLORIDA SHRUBBY WET FLATWOODS were floristically distinct from other wetlands associations due shr ubby species abundance in the mids tory and understory strata. Midstories were dominated by saw palmetto ( S. repens ), large gallberry ( I. coriacea ), and three shrub species also recognized as indicators: swamp bay ( P. palustris ), fetterbush ( L. lucida ) and coastal sweetpepperbush ( Clethra alnifolia ). The latter species is re stricted in distribution to the northern peninsula. Half of all indica tor species were shrubs (10 out of 20). Herbaceous cover in NORTH FLORIDA SHRUBBY WET FLATWOODS was sparse. The most abundant herbs were purple bluestem ( Andropogon glomeratus var. glaucopsis) cinnamon fern ( Osmunda cinnamomea ) and tenangle pipewort ( E. decangulare ). The few herbaceous indicators are typical of wetland ha bitats, including hooded pitcherplant ( Sarracenia minor ), bushy bluestem ( A. glomeratus var. glomeratus, and A. glomeratus var. hirsutior), and sphagnum

PAGE 50

50 moss ( Sphagneticola spp.). Mean species richness of NORTH FLORIDA SHRUBBY WET FLATWOODS was low, likely related to the pa ucity of herbaceous ground cover. It may appear that this asso ciation represents a fire suppr essed variant of wet flatwoods formerly dominated by herbaceous vegetation. However, NORTH FLORIDA SHRUBBY WET FLATWOODS appear compositionally distinct, and include d sample sites with histories of recent fire. UPPER PANHANDLE WET FLATWOODS (7 plots; W5): These sites were restricted to Miocene-aged sediments of the Northern Highl ands north of the Cody Scarp in the panhandle (Figure 2-2e). The number of samples comprising this association was small, which in part reflects the rarity of this a ssociation. In these sites, UPPER PANHANDLE WET FLATWOODS were situated downslope of PANHANDLE LONGLEAF PINE CLAYHILLS or PANHANDLE XERIC SANDHILLS. In general, this association occupied sma ll areas of mid-slopes a nd flat terraces. Fine-textured sediments distinguish soils of UPPER PANHANDLE WET FLATWOODS. Clay percentages of surface soils exceed that of all a ssociations, and silt content was similarly high. These characteristics also typify sub-surface soils. Soil organic matter of UPPER PANHANDLE MESIC-WET FLATWOODS was low compared to other Wetlands associations. UPPER PANHANDLE WET FLATWOODS have fairly sparse pine canopies comprised of three pine species: longleaf ( P. palustris ), slash ( P. elliottii ) and loblolly ( P. taeda ) pines (mean BAs: 4.55, 3.31, and 2.22 m2/ha respectively). Pond pine was a minor component ( P. serotina : mean BA .84 m2/ha). Midstory vegetation was sparse, a nd shrubs were mainly relegated to the understory in my frequently burned sites. Gallberry ( I. glabra ) was by far the most abundant woody understory species, followed by running oak ( Q. pumila ), blue huckleberry ( Gaylussacia frondosa var. nana ) and Darrow’s blueberry ( V. darrowii ).

PAGE 51

51 UPPER PANHANDLE WET FLATWOODS resembled Mesic Flatwoods associations in aspect and appearance. However, they were floristic ally similar to Mesic Flatwoods and Wetlands associations. UPPER PANHANDLE MESIC-WET FLATWOODS were notable for their high species richness, particularly of grass and forb species. Dominant grasses included some that are typical upland species, such as little bluestem ( S. scoparium var. stoloniferum ), broomsedge bluestem ( A. virginicus ), and Elliott’s bluestem ( A. gyrans var. gyrans ). Other frequent grasses included typical wetland species: toothache grass ( Ctenium aromaticum ), bushy bluestem ( A. glomeratus var. hirsutior ), and arrowfeather threeawn ( Aristida purpurascens var. virgata ). Wiregrass was absent. Two frequent grass spec ies were also recognized as i ndicator species: warty panicgrass ( Panicum verrucosum ) and ( Dichanthelium consanguineum). Most indicator species were forbs (19 out of 25). Four were members of the family Asteraceae and three were legume species. A large proportion of indicator species (32%) are sp ecies with ranges restricted to panhandle or northern peninsula. PANHANDLE WET FLATWOODS/PRAIRIES (16 plots; W6): This association occurred in the Northern Highlands and Coastal Lowlands of the panhandle, west of the Ochlochnee River basin (Figure 2-2e). Landscape and topographi c context was variable; in hillier terrain PANHANDLE WET FLATWOODS/PRAIRIES occupied narrow lower slopes and was associated with Dry Upland sandhills and clayhills. In contrast, nearly treeless PANHANDLE WET FLATWOODS/PRAIRIES of the Coastal Lowlands occupied large areas associated with NORTH FLORIDA MESIC FLATWOODS. Examples of the latter are the wet prairies of the Apalachicola National Forest (Clewell 1971). PANHANDLE WET FLATWOODS AND PRAIRIES soils were low in sand and high in silt, in contrast to the PENINSULA WET PRAIRIES of Central Florida. Organic content was among the lowest of Wetlands associations.

PAGE 52

52 PANHANDLE WET FLATWOODS AND PRAIRIES had sparse or no canopy. A few sites had sparse canopies of slash ( P. elliottii ) and longleaf ( P. palustris ) pines (mean BAs: 2.9 and 1.2 m2/ha respectively). Midstory vegetation was la rgely non-existent. Low growing gallberry ( I. glabra ) is by far the most abundant understory woody species, followed by the wetland shrub, buckwheat tree ( Cliftonia monophylla ). Frequently burned PANHANDLE WET FLATWOODS AND PRAIRIES had well-developed, herbaceous dominated ground cover vegetation. Wiregrass ( A. stricta ) and toothache grass ( C. aromaticum ) were ubiquitous dominant gras ses; the latter was also id entified as an indicator species. Other common and indicator species included many forbs (27 out of 41 species), including bristleleaf chaffhead ( Carphephorus psuedoliatris ), woolly sunbonnets ( Chaptalia tomentosa ), coastalplain yelloweyed grass ( Xyris ambigua ), and savanna meadowbeauty ( Rhexia alifanus ). Pinewoods bluestem ( A. arctatus ) was a distinctive and frequent bunch grass. Twenty four (58%) of the indicator species have dist ributions restricted to panhandle or northern peninsula. PANHANDLE SEEPAGE SLOPES (5 plots; W7): These few sites were located in the Northern Highlands and Coastal Lowlands of th e western panhandle, situated on lower slopes where soils were usually saturated from seepage (Figure 2-2e). This condition is thought to result from water percolation through sandy soils underlain by impermeable clay or rock hardpans (FNAI 1990). In my sites, PANHANDLE SEEPAGE SLOPES occurred downslope of the drier PANHANDLE WET FLATWOODS AND PRAIRIES association, and fart her downslope of Dry Uplands. Despite the putative existence of clay le nses in the subso il, sub-soils of PANHANDLE SEEPAGE SLOPES were very low in clay content. It is possible that my soil samples were not deep

PAGE 53

53 enough to detect hardpans. Sub-soil silt was low compared to other panhandle Wetland associations. Surface soils were silty, acidic and high in organic content (although small sample size precluded statis tical tests). Sparse canopies of PANHANDLE SEEPAGE SLOPES consisted of slash ( P. elliottii ) and longleaf ( P. palustris ) pines (mean BAs: 1.9 and 1.3 m2/ha respectively). Understory shrub cover was relatively high and dominated by gallberry ( I. glabra ), evergreen bayberry ( M. caroliniensis ) and large gallberry ( I. coriacea ). The latter two species were indicator species. Species richness and herbaceous diversity of PANHANDLE SEEPAGE SLOPES is high. Common grasses included several domi nant bunch grasses: wiregrass ( A. stricta ), arrowfeather threeawn ( A. palustris ), pinewoods bluestem ( A. arctatus ) and the wetland variant of Elliott’s bluestem (A. gyrans var. stenophyllus ). The latter three grass species were also recognized as indicator species. Other frequent indicators of PANHANDLE SEEPAGE SLOPES included many forbs and sedges, such as Texas tickseed ( Coreopsis linifolia ), largeleaf rosegentian ( Sabatia macrophylla ), featherbris tle beaksedge ( R. oligantha ), and giant whitetop ( R. latifolia ). A high proportion of indicator species ha ve ranges restricted to North and Panhandle Florida (26 out of 31 species). A few of these are locally abunda nt rhizomatous species: rush featherling ( Pleea tenuifolia ), coastal false asphodel ( Tofieldia racemosa ), yellow pitcherplant ( Sarracenia flava ), large beaksedge ( R. macra ), and featherbristle beaksedge ( R. oligantha ). Discussion Results from this study provide the first comprehensive community classification of pyrogenic pinelands of Florida. Specifically, this classificat ion describes variation among the remnants of natural pineland habitat following the large reduction and fr agmentation of a once expansive pineland landscape. Conditions and land scape contexts of sites in this study were

PAGE 54

54 dictated by the non-random distribution and manage ment of natural areas in Florida, which is related to the timing and confi guration of human settlement and economic development (Kautz and Cox 2006, Frost 2006). As such, this classifica tion is not representati ve of pre-settlement conditions of pineland variati on and diversity, although it repres ents the best approximation possible. Although most community associations fa ll within classifications previously suggested for southeastern U.S. pinelands (FNAI 1990, Peet 2006) this system presents a much greater refinement of recognizable spec ies associations. The describe d associations are generally defined by geographic region, physio graphic landform, local topogra phy, and soil characteristics, providing additional guidance to thei r identification in the field. Geographic segregation is pronounced in this floristic classifi cation. Gradients in species composition vary with well known environmental, cl imatic, and geologic gradients in Florida. The Florida peninsula spans almost seven degrees in latitude, and excluding the Florida keys, spans three bioclimatic life zones including the Warm Temperate Moist Forest of North Florida and the Subtropical Moist Forest of extreme s outh Florida (Holdridge 1967, Myers 2000). Much of the peninsula falls into a broad “transiti on zone” between the two. Variation in relative proportions of temperate vs. tropical tree species with latitudinal gradient is a well-known phenomenon (Wunderlin and Hansen 2000). Flor ida’s complex recent geologic history also underscores panhandle and peninsula differences which may contribut e to variation in contemporary vegetation. Carbonate deposits of marine origin crea ted the limestone platform of the peninsula between 60 to 120 MYBP. In c ontrast, Miocene deposits of the panhandle were mainly clastic sediments derived from Appalach ian erosion and alluvial processes (Randazzo and Jones 1997, Myers 2000). Until 12 to 30 MYBP the Suwannee Strait, an elongate negative structure in southern Georgia and northeastern Florida, sepa rated the two regions (Hull 1962,

PAGE 55

55 Puri and Vernon 1964, Myers 2000). From the late Miocene to recent ti me, increased clastic deposition, and a series of sea level fluctuat ions have influenced surface geology and soil development, particularly in the peninsula (R andazzo and Jones 1997, Myers 2000). The distinct panhandle-peninsula trend in fl oristic variation is correlated with many of these phenomenon, including differences in geol ogic sedimentation, age of landf orms, degree of isolation, and climate variation associ ated with latitude. To varying degrees, geographic patterns are similar within the three ecological series (Dry Uplands, Mesic Flatwoods, and Wetlands). Eastern and wester n analogues of similar edaphic and moisture conditions are nonetheless flor istically different (e. g. the xeric sandhills of peninsula vs. panhandle). The Dry Uplands associat ions display distinct se paration east and west of the Ochlochnee River basin. Although the sepa rations are not as pronounced, similar eastwest divisions exist among the Wetlands associations. PANHANDLE WET FLATWOODS/PRAIRIES and PANHANDLE SEEPAGE SLOPES of the western panhandle are flor istically distinct from other Wetlands associations. None of the Mesic Flatwoods associations is unique to the western panhandle, although there is a nor th-south separation between NORTH FLORIDA MESIC FLATWOODS and CENTRAL FLORIDA MESIC FLATWOODS/DRY PRAIRIES. Geographic segregation of flor istic groups is influenced by the prominence of plant taxa with distributions restricted to a particular region of the State. Nearly a fifth of all taxa (18.4% of 575) included in this analysis have restricted ranges, while a far smaller proportion (2.8 %) are endemic to Florida. Most restricted range taxa reflect the familiar segregation between panhandle and peninsula Florida, or the distinct ion between north (panhandle plus northern peninsula) Florida and the central peninsula. Many northerly distributed species reach their southern range limits in the north peninsula; their southern distri butional limits closely

PAGE 56

56 approximating the “warm temporate moist fore st” bioclimate zone (Holdridge 1967, Myers 2000). Fewer species have Florida distributions either restricted to the panhandle west of the Ochlochnee River basin, or the cent ral Florida peninsula. The larg e number of taxa restricted to the western panhandle (42 taxa) is consistent with other descript ions of endemism in regions within the East Gulf Coasta l Plain (125 endemic taxa re ported by Sorrie and Weakley 2001, Sorrie and Weakley 2006). Similarly, a large nu mber of endemics have previously been identified in the Florida peninsula (122 ta xa: Sorrie and Weakley 2001, Sorrie and Weakley 2006), where I recorded 31 restricted range taxa. Because I omitted Lake Wales Ridge from the sample region endemic species were likely under-re presented in the associ ations of the Florida peninsula. Restricted range taxa were fre quently selected as indicator species of associations. Not surprisingly, indicator species of panhandle Dry Uplands and Wetlands associations included many restricted-range species, and many of these are endemic to the East Gulf Coastal Plain as reported by Sorrie and Weakley (2001) Taxa restricted to the Fl orida peninsula were prominent indicators of the XERIC-MESIC FLATWOODS and CENTRAL FLORIDA MESIC FLATWOODS/PRAIRIES. No restricted range indicator sp ecies were selected for the single Wetland association of restricted to the peninsula (PENINSULA WET FLATWOODS/PRAIRIES). In contrast, the three Wetland associat ions restricted to the wester n panhandle were characterized by numerous indicator species with restricted ranges (UPPER PANHANDLE WET FLATWOODS, PANHANDLE WET FLATWOODS/PRAIRIES, and PANHANDLE SEEPAGE SLOPES). Comparisons to Other Classifications This classification of fire-adapted pinelands and associated communities resembles community descriptions of the Florida Natural Ar eas Inventory (FNAI), both floristically and in

PAGE 57

57 its description of landscape and edaphic conditions However, my classification augments the FNAI community descriptions in recognition of geographically re lated floristic variation. The FNAI description of “Sandhill” corresponds to my three Dry Upland associations distinguished by region and edaphic/moisture conditions (PANHANDLE XERIC SANDHILLS vs. PENINSULA XERIC SANDHILLS and NORTH FLORIDA SANDHILLS). Likewise, three current associations (PANHANDLE LONGLEAF PINE CLAYHILLS, PANHANDLE SILTY WOODLANDS and NORTH FLORIDA RICH WOODLANDS) resemble the FNAI “Upland Pine Forest ” community. FNAI re ports that “Upland Pine Forests” are restricted to the Miocene-aged rolling hills of extreme northern Florida (FNAI 1990). Two of the three aforemen tioned associations occur outside of the FNAI geographic and landscape description, thus they lack FNAI anal ogues. My Mesic Flatwoods associations are similar to FNAI community descriptions. Howeve r, the two Mesic Flatwoods associations were segregated by region (NORTH FLORIDA MESIC FLATWOODS vs. CENTRAL FLORIDA MESIC FLATWOODS AND DRY PRAIRIES), whereas the FNAI distinguishes mesic flatwoods by canopy conditions (“Mesic Flatwoods” vs. “Dry Prairie”). Cross reference between this classification and FNAI types is less clear for Wetlands associa tions. I recognized four Wetlands associations that overlap (in whole or part) with the FNAI de scriptions for “Wet Flatwoods”, “Wet Prairie” and “Marl Prairie” (the latte r perhaps corresponds to CALCAREOUS WET FLATWOODS). The FNAI description of “Bogs” may partially overlap with NORTH FLORIDA SHRUBBY WET FLATWOODS. The PENINSULA WET PRAIRIES resemble FNAI’s “Wet Prairi e”, although mine is regionally defined. The MARGINAL PRAIRIES association describes herbaceous vegetation associated with two FNAI lacustrine communities: “Flatwoods /Prairie/Marsh Lake” and “Sandhill Upland Lake”. The FNAI “Seepage Slopes” closely resembles my PANHANDLE SEEPAGE SLOPES association.

PAGE 58

58 My classification and study area represents a subset of the regiona l treatment of Peet (2007). The Peet classification in cludes a greater breadth of envi ronmental conditions as well as a larger geographic region. A comparison of the cu rrent community associat ions to associations reported by Peet as present in Florida reveals di fferences between the two classifications with regard to classification resoluti on. Geographic segregation is a pr imary trend in both treatments. My Dry Uplands associations variously correspo nd to Peet’s “Xeric Sand Barrens/Uplands” and “Subxeric Sandy Uplands” groups, which include thre e and six associations respectively (see Table 2, Peet 2006). My PANHANDLE LONGLEAF PINE CLAYHILLS corresponds to two of Peet’s “Silty Uplands” associations (types 3.4.1 a nd 3.4.2, Peet 2007). The Peet analogues to NORTH FLORIDA RICH WOODLANDS and PANHANDLE FLATWOODS/WOODLANDS are less obvious, perhaps corresponding to other “Siltly Uplands” type s. My three Mesic Flatwoods associations have 14 counterparts in the Peet treatment. Likewise, my Wetlands a ssociations correspond to nine associations of Peet’s “Savannas and Seeps”. The subjective nature of defining partitions likely contributes to differences between the classification typologies. In th is study I attempted to minimi ze subjectivity in sampling design and numerical analysis. The distribution and ma nagment of Florida pine land natural areas is non-random and largely influenced by natural area availability (Kautz and Cox 2006, Frost 2006). However, the sampling design stratified by ecoregion and moisture gradient, coupled with large sample size, minimized bias associat ed with subjective sample selection (Leps and Smilauer 2007). The resulting classification describes variation in the contemporary configuration Florida pinelands in a highly fragmented landscape. This work does not explicitly describe natural variation of pre-settlement c onditions. Furthermore, by using an optimization index in conjunction with cluster analysis, I mi nimized subjectivity associated with cluster

PAGE 59

59 delineations (McCune and Grace 2002). My selec tion of cut-level in the cluster solution to groups > 3 samples was subjective. However, this limited proliferation of associations, which is highly dependent on sample size (Legendre a nd Legendre 1998, McCune and Grace 2002). In conclusion, I developed a classification of fire-dependent pineland communities that is as comprehensive as possible while remaining applicable for management and conservation programs. The provision of indicator speci es, geographical distri butions, and topographic contexts of associations, in addition to their full species lists, will enable identification of associations in the field. This classification sy stem will assist in refining classifications of assocations in the greater than 9.6 million ac res of non-submerged conservation lands owned and/or managed by local, state, and federal agen cies in Florida (over 25 % of total land area; FNAI 2007). Descriptions of associations a nd indicator species will also guide ecological restoration efforts, by assisting the recognition of natu ral areas degraded from fire suppression or other reasons. Quantitative de scriptions based on existing hi gh quality natural areas provide templates for restoration goals and comparisons Further land acquisitions via the Florida Forever Program and other conservation efforts mi ght benefit from this classification. The present study presents a desc riptive vegetation classificati on based on a comprehensive, systematic, and quantitative inve ntory of fire-dependent pinela nd communities as they exist today.

PAGE 60

60 Table 2-1. Means and stardard errors (SE) of so il and site variables by community series. Plot number indicated in parentheses. Variab les are labeled as in text. Significantly different pairwise comparisons of mean s (p < 0.01) are denoted by dissimilar superscripts and different shading. Dry Uplands Mesic Flatwoods Wetlands (130) (99) (64) Variable Mean SE Mean SE Mean SE % sand A 93.13 b 0.67 96.18 c 0.84 88.49 a 0.97 % sand B 88.09 a 1.08 95.47 b 1.35 88.40 a 1.50 % silt A 4.29 b 0.49 2.52 a 0.61 8.39 c 0.71 % silt B 7.54 b 0.70 2.89 a 0.88 7.12 b 0.97 % clay A 2.58 b 0.36 1.29 a 0.45 3.11 b 0.52 % clay B 4.37 b 0.55 1.64 a 0.69 4.47 b 0.77 % org 2.86 a 0.35 4.82 b 0.45 5.69 b 0.51 pH 4.75 b 0.03 4.53 a 0.04 4.52 a 0.05

PAGE 61

Table 2-2. Untransformed means and starda rd errors (SE) of soil and site variab les by community association. Dissimilar superscripts and shading indicate mean s that are significantly di fferent (p < 0.01). Soil te xture variables listed separately for surface (A horizon) and sub-surfa ce (B horizon). Rich = species number / 1000 m2; BA = Basal area m2/ha. SURFACE SOILS (A)SUBSOILS (B) Association richSE richBASE BA% sand SE sand% siltSE silt % claySE clay% sand SE sand% siltSE silt % claySE clay% orgSE orgpHSE pHCaSE Ca Peninsula Xeric Sandhills68.5 a3.4 8.1 a1.01 96.9 b1.58 1.68 a1.07 1.38 a1.04 96.97 a2.76 1.66 a1.731.36 a1.52 4.01 b0.56 4.62 a0.07 570.57 b65.58North Florida Sandhills96.2 b2.8 11.4 ab0.87 94.9 b1.35 3.5 ab0.92 1.53 a0.89 95.16 a2.36 3.28 a1.481.54 a1.30 2.96 a0.48 4.76 a0.07 469.17 b56.16North Florida Rich Woodlands106.5 b4.8 16.1 cd1.44 94.7 b2.34 3.63 b1.59 1.68 a1.54 90.25 ab4.09 3.76 ab2.575.97 bc2.25 6.50 c0.83 4.58 a0.11 307.50 b97.27Panhandle Xeric Xandhills75.1 a2.8 9.5 a0.85 95.6 b1.33 2.70 a0.90 1.67 a0.88 89.98 b2.45 6.75 b1.533.26 ab1.35 1.18 a0.47 4.69 a0.04 152.01 a55.25Panhandle Silty Woodlands92.1 b3.4 12.8 bc1.01 86.8 a1.58 7.40 b1.07 5.8 b1.04 76.67 c2.76 17.00 c1.736.32 c1.52 2.02 a0.56 4.78 a0.06 174.38 a65.58Panhandle Longleaf Pine Clayhills124.5 c4.2 16.8 d1.27 86.5 a1.98 9.14 c1.34 4.39 b1.30 71.54 c3.46 15.32 c2.1713.13 d1.90 3.32 b0.70 5.10 b0.11 387.21 b82.21Xeric-Mesic Flatwoods59.0 a3.2 5.0 a0.85 96.9 b0.40 1.99 a0.43 1.12 a0.19 96.0 b0.40 2.55 a0.39 1.43 a0.21 4.64 a0.83 4.64 b0.07 458.17 b39.68North Florida Mesic Flatwoods71.4 b3.5 10.7 b0.94 95.0 a0.40 3.29 a0.44 1.68 b0.20 94.5 a0.42 3.60 a0.40 1.86 b0.22 5.14 a 0.86 4.39 a0.07 247.83 a41.02Peninsula Mesic Flatwoods/Dry Prairies72.9 b4.1 2.8 a1.09 96.7 ab0.50 2.27 a0.52 1.00 a0.23 95.9 b0.49 2.41 a0.47 1.65 ab0.26 4.65 a1.01 4.55 ab0.08 413.56 b48.21Marginal Prairies49.6 a6.6 5.6 a2.32 89.2 ab2.94 7.12 b2.43 3.61 b1.24 89.62 bc3.61 3.62 a2.51 6.75 c1.96 7.18 a1.55 4.37 ab0.05 212.33 a47.32Calcareous Wet Flatwoods125.0 NA18.4 17.1 NA7.50 96.6 NA0.62 2.47 NA3.74 0.88 NA0.24 93.13 NA5.35 3.07 NA3.69 3.79 NA2.93 5.97 NA2.54 5.28 NA0.40 1133.25 N A 563.53Peninsula Wet Flatwoods/Prairies69.7 b6.0 2.4 a1.92 95.6 b2.69 3.68 ab2.22 0.70 a1.13 94.63 c3.30 3.77 a2.28 1.59 a1.79 6.17 a1.42 4.64 b0.11 424.63 b88.86North Florida Shrubby Wet Flatwoods65.7 ab6.3 15.7 bc1.98 87.8 a2.81 8.43 b2.32 3.73 b1.18 91.12 bc3.45 6.11 ab2.39 2.76 ab1.87 6.88 a1.48 4.22 a0.12 202.11 a51.13Upper Panhandle Wet Flatwoods126.4 c7.9 10.9 ac2.90 80.9 a3.52 11.05 b2.90 8.08 c1.48 75.50 a4.32 13.52 c3.00 10.97 c2.34 4.23 a1.85 4.62 b0.09 183.25 a23.67Panhandle Wet Flatwoods/Prairies78.3 b5.4 4.5 a1.98 84.5 a2.40 12.76 bc1.98 2.77 ab1.01 82.73 ab2.95 12.07 bc2.05 5.18 bc1.60 3.46 a1.27 4.53 b0.07 124.73 a25.93Panhandle Seepage Slopes90.8 NA5.1 4.1 NA2.20 86.5 NA5.50 12.35 N A 5.29 1.2 NA0.36 92.00 NA6.18 6.56 NA4.26 1.44 NA3.38 9.96 NA5.92 4.28 NA0.04 245.25 NA59.50 61

PAGE 62

62 Table 2-3. Common woody shrubs of midstory and understory st rata, listed by association. Code in parentheses correspond to thos e in Figure 2-2. Mean cover in m2/100 m. Peninsula Xeric Sandhills (D3) Common Name Scientific Name Mean cover Turkey oak Quercus laevis 14.20 Sand live oak Quercus geminata 4.27 Saw palmetto Serenoa repens 1.91 Bluejack oak Quercus incana 1.71 Panhandle Xeric Sandhills (D4) Turkey oak Quercus laevis 12.08 Dwarf live oak Quercus minima 3.71 Sand live oak Quercus geminata 3.10 Saw palmetto Serenoa repens 3.00 Sand post oak Quercus margaretta 2.75 Dwarf huckleberry Gaylussacia dumosa 2.40 Bluejack oak Quercus incana 2.11 North Florida Sandhills (D2) Bluejack oak Quercus incana 6.72 Turkey oak Quercus laevis 6.18 Sand post oak Quercus margaretta 4.81 North Florida Rich Woodlands (D1) Saw palmetto Serenoa repens 11.03 Winged sumac Rhus copallinum 5.03 Mockernut hickory Carya alba 4.23 Dwarf waxmyrtle Myrica cerifera var. pumila 3.01 Sand live oak Quercus geminata 2.65 Sand post oak Quercus margaretta 2.43 Panhandle Longleaf Pine Clayhills (D5) Sand post oak Quercus margaretta 5.71 Turkey oak Quercus laevis 4.89 Running oak Quercus pumila 4.46 Bluejack oak Quercus incana 3.51 Darrow's blueberry Vaccinium darrowii 3.13 Southern red oak Quercus falcata 2.97 Dwarf huckleberry Gaylussacia dumosa 2.60 Winged sumac Rhus copallinum 2.08

PAGE 63

63 Table 2-3 continued. Panhandle Silty Woodlands (D6) Gallberry Ilex glabra 9.01 Running oak Quercus pumila 7.45 Saw palmetto Serenoa repens 6.19 Dwarf live oak Quercus minima 6.04 Dwarf huckleberry Gaylussacia dumosa 3.27 Shiny blueberry Vaccinium myrsinites 2.59 Darrow's blueberry Vaccinium darrowii 2.48 Blue huckleberry Gaylussacia frondosa var nana 1.93 Xeric-Mesic Flatwoods (M1) Saw palmetto Serenoa repens 26.83 Sand live oak Quercus geminata 9.92 Chapman's oak Quercus chapmanii 7.90 Myrtle oak Quercus myrtifolia 5.35 Shiny blueberry Vaccinium myrsinites 4.18 Fetterbush Lyonia lucida 3.73 Dwarf live oak Quercus minima 3.68 North Florida Mesic Flatwoods (M2) Gallberry Ilex glabra 19.49 Saw palmetto Serenoa repens 17.01 Dwarf live oak Quercus minima 10.46 Shiny blueberry Vaccinium myrsinites 6.56 Running oak Quercus pumila 5.93 Dwarf huckleberry Gaylussacia dumosa 2.64 Fetterbush Lyonia lucida 2.45 Central Florida Mesic Flatwoods/Dry Prairies (M3) Saw palmetto Serenoa repens 23.10 Dwarf live oak Quercus minima 7.40 Gallberry Ilex glabra 6.72 Fetterbush Lyonia lucida 5.26 Shiny blueberry Vaccinium myrsinites 3.84 Dwarf huckleberry Gaylussacia dumosa 2.15

PAGE 64

64 Table 2-3 continued. Marginal Prairies (W1) Gallberry Ilex glabra 3.55 Titi Cyrilla racemiflora 3.24 Sandweed Hypericum fasciculatum 3.09 Swamp tupelo Nyssa sylvatica var biflora 1.94 Peninsula Wet Flatwoods/Prairies (W2) Sandweed Hypericum fasciculatum 4.30 Gallberry Ilex glabra 3.48 Pond cypress Taxodium ascendens 2.33 Calcareous Wet Flatwoods (W3) Wax myrtle Myrica cerifera 5.38 Swamp bay Persea palustris 2.75 Saw palmetto Serenoa repens 2.28 Red maple Acer rubrum 2.22 Sweetgum Liquidambar styraciflua 2.06 North Florida Shrubby Wet Flatwoods (W4) Saw palmetto Serenoa repens 6.41 Large gallberry Ilex coriacea 5.73 Fetterbush Lyonia lucida 4.43 Coastal sweetpepperbush Clethra alnifolia 3.27 Sweetbay magnolia Magnolia virginiana 3.15 Upper Panhandle Wet Flatwoods (W5) Gallberry Ilex glabra 19.27 Running oak Quercus pumila 4.52 Blue huckleberry Gaylussacia frondosa var nana 1.71 Darrow's blueberry Vaccinium darrowii 1.61 Panhandle Wet Flatwoods/Prairies (W6) Gallberry Ilex glabra 8.39 Buckwheat tree Cliftonia monophylla 3.24 Panhandle Seepage Slopes (W7) Gallberry Ilex glabra 13.00 Evergreen bayberry Myrica caroliniensis 4.55 Large gallberry Ilex coriacea 1.78

PAGE 65

65 Table 2-4. Indicator species of Dry Uplands and Mesic Flatw oods associations listed in descending order of Indicator Value (IV). Superscripts indicate species with restricted distributions in Florida: 1western Panhandle, 2north Florida, 3central Florida peninsula, 4Florida endemic. Peninsula Xeric Sandhills (D3) North Florida Sandhills con't (D2) Species IV Species IV Bulbostylis warei 41.8 Helianthemum carolinianum 31.9 Balduina angustifolia 40.3 Rhynchosia reniformis 31.6 Aristida condensata 35.0 Physalis walteri 28.9 Lechea sessiliflora 33.4 Scutellaria multiglandulosa 28.7 Asimina incana 3 33.2 Piriqueta cistoides ssp. caroliniana 28.6 Triplasis americana 31.0 Dyschoriste oblongifolia 28.4 Opuntia humifusa 29.0 Asclepias verticillata 26.9 Callisia graminea 26.8 Gymnopogon ambiguus 26.9 Carphephorus corymbosus 3 23.3 Eupatorium glaucescens 25.0 Sporobolus junceus 23.2 Lespedeza hirta 25.0 Cnidoscolus stimulosus 22.9 Ruellia caroliniensis ssp. ciliosa 23.8 Quercus laevis 22.8 Croton argyranthemus 23.3 Galactia regularis 22.6 Tragia urens 23.3 Tephrosia chrysophylla 21.3 Sisyrinchium xerophyllum 20.1 North Florida Rich Woodlands (D1) Species IV Panhandle Xeric Sandhills (D4) Erythrina herbacea 54.6 Species IV Dichanthelium oligosanthes var. oligosanthes 2 46.6 Galactia microphylla 74.7 Eustachys floridana 44.4 Euphorbia floridana 1 67.8 Galium hispidulum 43.2 Liatris pauciflora var. secunda 1 35.5 Lactuca floridana 41.6 Cyperus lupulinus ssp. lupulinus 34.6 Cyperus plukenetii 2 41.0 Rhynchosia cytisoides 1 34.2 Rhynchosia cinerea 3,4 37.0 Pityopsis aspera 1 32.1 Aristida lanosa2 34.7 Eriogonum tomentosum 28.3 Tridens carolinianus2 34.7 Commelina erecta 26.8 Sporobolus clandestinus 33.6 Aristida mohrii 2 26.6 Ageratina aromatica 2 29.6 Stylisma patens ssp. patens 23.8 Galium pilosum 26.0 Liatris chapmanii 23.5 Vitis aestivalis 24.1 Tephrosia mohrii 1 22.9 Centrosema arenicola3 22.5 Rhynchospora grayi 22.0 Clitoria mariana 22.1 Bulbostylis ciliatifolia 20.6 Habenaria quinqueseta 22.1 Aristolochia serpentaria 21.5 North Florida Sandhills (D2) Dichanthelium commutatum var. ashei 20.4 Species IV Desmodium glabellum 2 20.0 Desmodium floridanum 38.4 Palafoxia integrifolia 33.8

PAGE 66

66 Table 2-4 continued. Panhandle Longleaf Pine Clayhills (D5) Panhandle Longleaf Pine Clayhills con't (D5) Species IV Species IV Rudbeckia hirta 76.0 Gymnopogon brevifolius 24.6 Acalypha gracilens 63.3 Eupatorium album 24.3 Malus angustifolia 1 55.9 Dichanthelium sphaerocarpon 2 24.2 Vaccinium stamineum var. stamineum 55.4 Stylodon carneus 24.0 Galactia volubilis 54.4 Tridens carolinianus 2 24.0 Quercus falcata 48.2 Carya alba 47.1 Panhandle Silty Woodlands (D6) Ceanothus americanus 42.7 Species IV Desmodium ciliare 42.7 Symphyotrichum adnatum 40.2 Desmodium lineatum 42.7 Baptisia simplicifolia 1,4 38.5 Toxicodendron pubescens 2 42.7 Angelica dentata 1 37.6 Desmodium viridiflorum 39.7 Chrysopsis mariana 32.1 Prunus serotina 39.0 Tragia smallii 2 30.7 Phlox floridana 2 38.2 Phoebanthus tenuifolius 1,4 29.6 Lespedeza repens 2 37.0 Viola septemloba 29.4 Rhynchosia tomentosa 2 36.3 Euphorbia curtisii 2 28.4 Euphorbia discoidalis 1 35.0 Galactia erecta 2 27.6 Eragrostis spectabilis 34.1 Agalinis divaricata 2 26.2 Cornus florida 32.2 Helianthus radula 25.6 Strophostyles umbellata 31.1 Dalea carnea var. gracilis 1 22.5 Smilax smallii 30.7 Crotalaria purshii 21.9 Desmodium strictum 30.6 Seymeria cassioides 20.4 Gaura filipes 2 30.5 Sorghastrum nutans 29.5 Xeric-Mesic Flatwoods (M1) Galium pilosum 29.1 Species IV Ambrosia artemisiifolia 28.6 Quercus chapmanii 38.1 Eupatorium hyssopifolium 2 28.6 Solidago odora var. chapmanii 3 32.9 Clitoria mariana 27.4 Quercus myrtifolia 26.1 Lobelia puberula 27.3 Galactia elliottii 25.6 Vernonia angustifolia 27.3 Liatris tenuifolia var. quadriflora 3 22.1 Aristolochia serpentaria 26.9 Befaria racemosa 3 19.5 Andropogon gerardii 1 26.7 Rhynchospora megalocarpa 17.1 Solidago odora var. odora 26.5 Lechea minor 26.1 North Florida Mesic Flatwoods (M2) Ageratina aromatica 2 26.0 Species IV Muhlenbergia capillaris var. trichopodes 2 25.5 Xyris caroliniana 23.7 Tephrosia spicata 25.5 Sporobolus floridanus 2 19.0 Salvia azurea 25.4 Quercus minima 18.4 Rubus cuneifolius 24.8 Kalmia hirsuta 2 17.5

PAGE 67

67 Table 2-4 continued. Central Florida Mesic Flat woods/Dry Prairies (M3) Species IV Hypericum reductum 62.3 Polygala setacea 57.4 Eleocharis baldwinii 50.4 Rhexia nuttallii 43.0 Fimbristylis puberula 41.5 Aristida spiciformis 38.2 Asimina reticulata 3,4 37.8 Rhynchospora fernaldii 37.5 Xyris flabelliformis 36.4 Lyonia fruticosa 34.9 Lechea torreyi 34.3 Lachnocaulon beyrichianum 3 34.0 Dichanthelium chamaelonche 33.3 Syngonanthus flavidulus 32.7 Xyris brevifolia 32.6 Polygala rugelii 3,4 32.2 Asclepias pedicellata 31.9 Aristida purpurascens var. tenuispica 30.7 Oldenlandia uniflora 30.7 Gymnopogon chapmanianus 3 29.0 Gratiola hispida 25.8 Drosera brevifolia 25.4 Schizachyrium stoloniferum 3 24.5 Dichanthelium sabulorum var. thinium 23.8 Andropogon brachystachyus 3 22.6 Hypericum tetrapetalum 22.6 Lygodesmia aphylla 21.7

PAGE 68

68 Table 2-5. Indicator species of Wetlands associations. Indicator values (IV) listed in descending order for each association. Superscripts same as Table 2-4. Marginal Prairies (W1) Penins ula Wet Flatwoods/Prairies (con't) Species IV Species IV Eupatorium leptophyllum 49.3 Andropogon capillipes (wetland variant) 43.9 Panicum hemitomon 41.1 Coelorachis rugosa 43.7 Ludwigia suffruticosa 33.5 Amphicarpum muehlenbergianum 43.1 Cephalanthus occidentalis 30.7 Sabatia grandiflora 42.2 Rhexia mariana var. mariana 28.8 Andropogon gyrans var. stenophyllus 40.0 Xyris difformis var. curtissii 2 28.8 Fuirena scirpoidea 40.0 Hyptis alata 38.4 North Florida Shrubby Wet Flatwoods (W4) Hypericum myrtifolium 37.7 Species IV Pluchea rosea 36.7 Persea palustris 45.8 Eragrostis elliottii 36.6 Osmunda cinnamomea 44.3 Scleria baldwinii 36.4 Andropogon glomeratus var. hirsutior 41.4 Bigelowia nudata 36.2 Nyssa biflora 37.0 Rhynchospora tracyi 33.4 Vaccinium virgatum 36.1 Viola lanceolata 32.1 Andropogon glaucopsis 35.7 Eriocaulon decangulare 31.4 Andropogon glomeratus var. glomeratus 34.8 Centella erecta 30.1 Viburnum nudum 32.3 Lobelia glandulosa 29.7 Photinia pyrifolia 32.2 Rhynchospora filifolia 28.0 Sphagneticola sp. 28.6 Eriocaulon compressum 26.9 Rhexia virginica 2 28.1 Pluchea foetida 25.8 Rhexia petiolata 26.7 Scleria muehlenbergii 25.1 Ilex coriacea 26.4 Ludwigia linearis 25.0 Gordonia lasianthus 25.6 Eupatorium mohrii 24.6 Sarracenia minor 24.9 Scleria georgiana 24.1 Vaccinium fuscatum 24.3 Xyris difformis var. floridana 23.4 Rhynchospora fascicularis 24.0 Rhexia mariana var. exalbida 22.8 Lyonia lucida 22.7 Helenium pinnatifidum 22.1 Carex glaucescens 22.2 Iva microcephala 21.3 Clethra alnifolia 2 22.0 Schizachyrium rhizomatum 21.2 Panicum rigidulum var. pubescens 21.1 Peninsula Wet Flatwoods/Prairies (W2) Species IV Calcareous Wet Flatwoods (W3) Oxypolis filiformis 81.1 Species IV Dichanthelium erectifolium 71.4 Asclepias lanceolata 100.0 Proserpinaca pectinata 57.9 Panicum rigidulum var. rigidulum 83.6 Gratiola ramosa 56.5 Helenium pinnatifidum 77.9 Coreopsis floridana 52.0 Phyla nodiflora 70.7 Ludwigia linifolia 50.6 Cirsium nuttallii 67.9 Xyris elliottii 48.7 Rhynchospora colorata 65.6 Panicum tenerum 47.4 Ludwigia microcarpa 64.3 Hypericum fasciculatum 44.8 Sabal palmetto 60.3 Aristida palustris 44.6 Dichanthelium dichotomum var. nitidum 58.3

PAGE 69

69 Table 2-5 continued. Calcareous Wet Flatwoods (con't) Upper Panhandle Wet Flatwoods (con't) Species IV Species IV Xyris jupicai 55.6 Diodia virginiana 29.8 Cyperus polystachyos 54.1 Polygala nana 29.2 Rhynchospora divergens 46.8 Helianthus angustifolius 27.9 Hyptis alata 44.4 Eupatorium leucolepis 2 27.2 Erechtites hieraciifolia 42.3 Andropogon glomeratus var. hirsutior 23.8 Rhynchospora perplexa 2 41.2 Linum medium 23.6 Ludwigia curtissii 41.1 Tephrosia spicata 23.5 Saccharum giganteum 40.9 Gratiola pilosa 23.4 Parthenocissus quinquefolia 40.6 Crotalaria purshii 23.3 Cyperus haspan 40.4 Panicum verrucosum 23.1 Rhynchospora globularis 38.8 Rhexia mariana var. exalbida 23.1 Andropogon glomeratus var. pumilus 37.4 Eupatorium rotundifolium 22.1 Rhynchospora microcarpa 37.3 Gymnopogon brevifolius 21.7 Setaria parviflora 35.6 Scutellaria integrifolia 21.5 Mecardonia acuminata 34.9 Juncus roemerianus 34.1 Panhandle Wet Flat woods/Prairies (W6) Sacciolepis indica 34.1 Species IV Cladium mariscus ssp. jamaicense 31.6 Coreopsis linifolia 2 72.3 Osmunda regalis 29.8 Carphephorus pseudoliatris 1 70.8 Proserpinaca pectinata 29.7 Helianthus heterophyllus 1 61 Bidens mitis 29.2 Drosera filiformis 1 50 Hypericum cistifolium 28.3 Eurybia chapmanii 2 50 Lobelia glandulosa 27.8 Lophiola aurea 50 Muhlenbergia capillaris var. capillaris 27 Andropogon arctatus 46.7 Mitreola sessilifolia 26.7 Scleria pauciflora var. caroliniana 45.5 Iva microcephala 26.1 Xyris baldwiniana 2 45.5 Rhexia lutea 2 42.6 Upper Panhandle Wet Flatwoods (W5) Ilex myrtifolia 2 42.3 Species IV Polygala cruciata 41.7 Hypericum setosum 61.4 Smilax laurifolia 40 Pycnanthemum flexuosum 1 57.1 Eryngium integrifolium 2 37.5 Rhododendron canescens 2 53.5 Pleea tenuifolia 1 37.5 Dichanthelium consanguineum 52.3 Chaptalia tomentosa 37.2 Desmodium tenuifolium 39.1 Xyris ambigua 35.3 Rhynchospora debilis 2 38.7 Sarracenia flava 2 34.8 Hibiscus aculeatus 2 35.9 Eupatorium leucolepis 2 34.7 Elephantopus nudatus 2 34.7 Pityopsis oligantha 1 34.1 Solidago stricta 32.6 Asclepias connivens 33.4 Agalinis georgiana 1 31.6 Gaylussacia mosieri 2 32.2 Lespedeza capitata 1 30 Rhynchospora latifolia 31.6

PAGE 70

70 Table 2-5 continued Panhandle Wet Flatwoods/Prairies (con't) Panhandle Seepage Slopes (con't) Species IV Species IV Rhynchospora baldwinii 31.2 Morella caroliniensis 36.3 Tofieldia racemosa 2 31.2 Hypericum brachyphyllum 34.1 Cliftonia monophylla 2 31.1 Tofieldia racemosa 2 34.1 Morella caroliniensis 31.1 Rhexia lutea 2 33.5 Dichanthelium leucothrix 31.0 Zigadenus glaberrimus 1 33.4 Rhexia alifanus 30.7 Dichanthelium longiligulatum 33.1 Lobelia brevifolia 1 30.4 Drosera filiformis 1 32.7 Anthaenantia rufa 2 30.2 Oxypolis filiformis 32.0 Eurybia eryngiifolia 1 30.2 Andropogon gyrans var. stenophyllus 31.3 Erigeron vernus 30.1 Eleocharis tuberculosa 31.2 Nyssa ursina 1,4 30.0 Fuirena squarrosa 2 31.2 Rhynchospora chapmanii 29.7 Magnolia virginiana 31.1 Oxypolis ternata 1 28.2 Sarracenia flava 2 27.9 Aletris sp. 27.5 Balduina uniflora 2 27.5 Andropogon mohrii 1 27.0 Rhynchospora corniculata 27.5 Ctenium aromaticum 27.0 Anthaenantia rufa 2 24.5 Aristida simpliciflora 2 26.5 Xyris scabrifolia 1 24.5 Sarracenia psittacina 2 25.8 Gaylussacia mosieri 2 23.0 Liatris spicata 25.2 Liatris spicata 22.2 Verbesina chapmanii 1,4 25.0 Panhandle Seepage Slopes (W7) Species IV Sabatia macrophylla 2 88.9 Rhynchospora oligantha 2 81.9 Arnoglossum ovatum 76.8 Juncus trigonocarpus 1 68.9 Pleea tenuifolia 1 54.5 Rhynchospora macra 2 54.3 Symphyotrichum lateriflorum var. lateriflorum 54.3 Rhynchospora latifolia 50.0 Sarracenia leucophylla 1 49.4 Lophiola aurea 49.2 Xyris difformis var. difformis 2 46.2 Panicum rigidulum var. combsii 45.5 Aristida palustris 44.2 Coreopsis linifolia 2 42.7 Eryngium integrifolium 2 39.9 Andropogon arctatus 37.0 Sarracenia psittacina 2 36.8

PAGE 71

71 Figure 2-1. Physiographic landforms modified from Puri and Vernon (1964). Colored shading denotes three “generalized landforms” which separate Hi ghlands from Lowlands, and Northern Highlands (Clastic sediments) fr om Central Highlands (part of the carbonate peninsular platform). Shaded regions indicate primary landforms denoted by landform type. Approximate southe rn boundary of study region is shown. Highlands Ridges, Uplands, Slopes Lowlands, Gaps,Valleys, PlainsPhysiographic LandformTypes Central Highlands Northern Highlands Coastal LowlandsGeneralized Locations of LandformsSouthernextent of study area

PAGE 72

72 Figure 2-2. Plot locations indicated by association. (a) Histor ic range of longleaf pine plus all plot locations (yellow dots). Figures (b) through (f) show plot locations by community association and primary la ndform types (Puri and Vernon 1964). Community labels (in parenthese s) correspond to Figure 2-3. ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( !( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( !( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (! ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( !( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( !( ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( !( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( !( ( ( ( ( ( ( ( ( ( ( ( (( ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ( ( ( ( ( ( ( ( ( ( ( ( ( ( !( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (b)DryUplands PanhandleXeric Sandhills (D4) PanhandleLongleaf PineClayhills (D5) PanhandleSilty Woodlands (D6)(d)MesicFlatwoodsNorthFloridaMesic Flatwoods (M2) Xeric-Mesic Flatwoods (M1) CentralFloridaMesic Flatwoods/DryPrairies (M3) (c)DryUplands NorthFlorida Sandhills (D2) NorthFloridaRich Woodlands (D1) PeninsulaXeric Sandhills (D3) (a)All plot locations Historicrange longleaf pine

PAGE 73

73 Figure 2-2 continued. ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ) ) ) ) ) # # # # # # # ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (" ) ) ) )# # # # # # # # # # # *! ((f)Wetlands (e)Wetlands Panhandle SeepageSlopes(W7) PanhandleWet Flatwoods/Prairies(W6) UpperPanhandle WetFlatwoods(W5) MarginalPrairies(W1) NorthFloridaShrubby WetFlatwoods(W4) PeninsulaWet Flatwoods/Prairies(W2) CalcareousWetFlatwoods(W3)

PAGE 74

74 Figure 2-3. Two dimensional NMS ordination of 293 samples. Lines separate samples into three community series, and colored symbols denote association from K-means cluster analysis. Plot number per association noted in parentheses. Percent variation of original distance matrix represen ted by NMS ordination: Axis 1 r2 = 0.54, Axis 2 r2 = 0.29. Stress = 14.78. NMS Axis 1 N M S A x i s 2 DRY UPLANDS WETLANDS MESIC FLATWOODS D1North Florida RichWoodlands (11) (31) D3Peninsula Xeric Sandhills (22) D4Panhandle Xe ric Sandhills (31) D5Panhandle Longleaf Pine Clayhills (14) D6Panhandle SiltyWoodlands (22) M1Xeric-Mesic Flatwoods (36) M2North Florida Mesic Flatwoods (30) M3Central Florida Mesic Flatwoods/Dry Prairies (22) W1Marginal Prairies (11) W2PeninsulaWet Flatwoods/Prairies (16) W3CalcareousWet Flatwoods (4) W4North Florida ShrubbyWet Flatwoods (15) W5Upper PanhandleWet Flatwoods (7) W6PanhandleWet Flatwoods/Prairies (16) W7Panhandle Seepage Slopes (5)

PAGE 75

75 CHAPTER 3 GEOGRAPHIC, ENVIRONMENTAL AND RE GIONAL VARIATION IN FLORISTIC COMPOSITION OF FLORIDA PYROGENIC PINELANDS Introduction Natural variability of plant communities is sh aped by complex interact ions of biotic and abiotic factors. Models of “abiotic controls ” (i.e. environmental control models) emphasize influences of environmental gradients, resources limitations, and niche specialization (Whittaker 1956, Bray and Curtis 1957, Hutchinson 1957, P eet and Loucks 1977, Platt and Weis 1977, Tilman 1994). In general, these models incl ude local and regional pr ocesses which influence species coexistence and distri butions (e.g. niche assembly, li mited resource availability, environmental filters to species assembly, envi ronmentally determined species pools). In addition, natural disturbances are considered environmental influenc es, particularly those related to density-independent processes (i.e. fire, hurricanes). Converse ly, biotic control models of community structure emphasize mechanisms unrel ated to environmental determinants, e.g. dispersal limitation, speciation and extinctions, competition, and herbivory. At its extreme, biotic control theory states that community st ructure is governed strict ly by dispersal limitation and demographic stochasticity independent of local environmental influences (Hubbell and Foster 1986, Bell 2001, Hubbell 2001). The relative importance of biotic versus abio tic factors in structuring plant communities is a subject of much recent deba te (see Legendre et al. 2005). R ecent models spatially explicit models of community structure have suggested the relative importance of both (Borcard and Legendre 1994, Okland et al. 2003, Tuomisto et al. 2003, Svenning et al. 2004). Spatial autocorrelation in community com position is cited as ev idence of biotic control models (Hubbell and Foster 1986, Hubbell 2001, Condit et al. 200 2). However, environmental variables themselves may be spatially structured, and spa tial autocorrelation of community structure may

PAGE 76

76 be mis-interpreted if environmental-spatial relationships are not c onsidered (Borcard and Legendre 1994, Legendre and Legendre 1998, Lege ndre et al. 2005). In addition, inferred environmental-species relationships may be biased in models that fail to include spatial trends (Legendre and Fortin 1989). This underscores the need for spatia lly explicit models of species composition and diversity, in which spatial autoco rrelation and spatially structure environmental variation can be quantified (Borcard et al 1992, Legendre and Legendr e 1998, Legendre et al. 2005). This approach allows formulation of hypotheses concerning the relative importance of underlying mechanisms that infl uence community structure. The relative influence of ecological determin ates varies over di fferent spatial and temporal scales. Recent ecological theory suppo rted by ecological observation suggests that, with regard to relative importance as determinan ts of species composition and diversity, regional factors and historic processes are comparable to local scale factors (Ricklefs 1987, Cornell and Lawton 1992, Collins et al. 2002). Processes relate d to historical biogeography, paleogeology and recent land use history have been recognized as potential influencers of contemporary community patterns (Okland et al. 2003, Graae et al. 2004, Svenning et al. 2004, Svenning and Skov 2005). Regional influences of local species di versity may be a functi on of distinct species pools (the “species pool effect”), in response to differential biogeographic and evolutionary histories (Zobel 1992, 19 97, Grace et al. 2000). Natural disturbance and local environmental gradients influence local-scale community structure of pyrogenic pineland vegetation of the Southeastern Coastal Plain. Studies of community composition, diversity, and species ’ response underscore the influence of environmental determinants, including topography-m oisture and soil properties (Kirkman et al. 2001, Drewa and Platt 2002a, Peet et al. 2003) and fi re (Platt et al. 1988a, White et al. 1991, Platt

PAGE 77

77 1999, Glitzenstein et al. 2003). Fi re affects environmental cond itions vis--vis regulation of limiting resources such as availa ble soil nutrients and light (Chr istensen 1977, Mitchell et al. 1999, Kirkman et al. 2001). In addition, fire a ffects competitive dynamics, particularly that between woody and herbaceous vegetation (Streng et al. 1993). Biotic determinants of pineland species coexistence are less well known in the Coas tal Plain. However, there is some evidence that dispersal limitation regulates plant divers ity in temperate grasslands elsewhere. Despite the relative wealth of research rega rding local patterns and processes, little is know about regional scale environmental-community relationships in Coastal Plain pinelands and the interplay between regiona l and local scale relationships. Descriptions of grassland diversity usually involve meso-scale regions < 1000 ha in size (Walker and Peet 1983, Tilman 1994, Grace et al. 2000, Weiher et al. 2004). In addition to high diversity linked to local environmental gradients, Florida’s pyrogenic pine lands are notable for their regional floristic diversity and high concentrations of endemic and restricted-ra nge species (James 1961, Walker and Peet 1983, Myers 1990, Peet and Allard 1993, Myers 20 00, Sorrie and Weakley 2002). These observations have lead to predictions of environmental controls of community composition that include regional and local fact ors (Peet and Allard 1993, Grace and Pugesek 1997, Kirkman et al. 2001), as we ll as historical influences (Ricklefs 1987, Zobel 1992). In this study, I present a model of variabi lity in species composition and diversity of pyrogenic pinelands. Specifically, environmen t-composition variation was analyzed in a spatially explicit model, allowing quantification of environmental variation that is spatially structured and spatially independent, plus the spatial component of variation that is unrelated to measured environmental variables. This allowed formation of hypotheses concerning the mechanisms controlling landscape vegetation patterns. Furthe rmore, I present hypotheses

PAGE 78

78 regarding scales of influence of different environmental factors and historical processes. Finally, I modified the spatially explicit model to incl ude a generalized ecoregion model based on those presently used in conservation efforts. The ecore gion model represents of regional differences of biogeography and paleogeography. To develop thes e models, I used a large dataset of pineland vegetation samples collected over a broad rang e of geography and environmental conditions. Models based on these descriptive data were intended to generate hypotheses concerning variation of relict pineland natu ral areas in a highly fragmented landscape of heterogeneous land use. Methods Study Region The study area includes the Florida Panhandl e and most of central and northern Peninsular Florida, approximating the current ra nge of longleaf pine. This area includes roughly nine million ha of the northern two-thirds of the state, extending approximately 480 km south from the Georgia state border (approximately 31 00’) to a southern boundary extending from approximately 26 70’ on the west coast to 28 80’ on the east coast. This southern boundary approximates southern extent of the “warm temper ate moist forest” bioclimate zone (Holdridge 1967). The study region extends westward to approxi mately 87 30’ and eastward to the Atlantic coast (approximately 80 00’). Florida is characterized by a humid subtropical climate. In general, mean temperature and daily radiation increase w ith decreasing latitude. Mean a nnual maximum temperatures range from 25 C in the western panhandle to 29 C in interior peninsular Florida, and minimum temperatures and shortwave radiation vary li kewise southward (13 to17 C, and 345 to 361 MJ/m2/day (Fernald 1981, Thornton et al. 1999). Average annual rainfall is highest in the

PAGE 79

79 western panhandle (173 cm/year) and declines fart her east and south, reaching its lowest point in the central peninsula (approximately 124 cm/year ; Fernald 1981). The distribution of rainfall varies from northwest to southeas t; winter and spring months are drie r in peninsular Florida, with more pronounced rainfall during the summer months Rainfall is more evenly distributed throughout the year in northwest Florida, with peaks during the late wi nter and summer months (Chen and Gerber 1990). The study region in Florida encompasses a wide range of edaphic conditions. Soils range from droughty coarse sands to poorly drained wetla nd mucks with high organic content. Entisols are common in the well-drained uplands of panhandle and north Florida (Puri and Vernon 1964, Myers 1990). Older and more weat hered Ultisols and Alfisols are common in these regions and are typically contain higher concentrations of fine textured sediments such as clay and silt (Myers and Ewel 1990, Myers 2000). Sandy, acidic spodosols are typical of upland woodlands in coastal and peninsular regions. These infertile mineral soils have subsurface accumulations of organic matter. Histosols with large accumu lations of organic matter are common to poorly drained wetlands (Brady and Weil 2000). Vegetation and Environmental Data Sample site selection and field methods are described in detail in Chapter 2. In brief, the study area was stratified into 19 regions based on similarity of physiography, geology, soils, climate, and historic vegetation maps (Fenneman 1938, Puri and Vernon 1964, Davis 1967, Fernald 1981, Brooks 1982, Bailey et al. 1994, Griffith et al. 1994). Sample sites were selected across regions in roughly equal numbers. Sample sites were subjected to rigorous selection criteria which precluded those w ithout native pineland vegetation w ith recent history of fire.

PAGE 80

80 Sites degraded from anthropomorphic impact, fire suppression, and/or in vasive species were rejected. At each site, three or four zones were delin eated relative to perceived gradients of topographic-moisture conditions and change in plant species composition. A single 1000 m2 plot was randomly placed in each zone and all vascular plant taxa were recorded as they were encountered in a series of four nested sub-plots (plot areas: 0.01, 0.1, 1, and 10 and 100 m2). Aerial cover in the 100 m2 plots was estimated by cover classes and averaged (by midpoint). Species encountered in the remainder of the 1000 m2 area were assigned nominal cover estimates. All plots were sampled during the late su mmer though early winter (August-December) over a four year period (2000 – 2004). Taxonomic nomenclature generally follows Kartesz (1999). In field and herbarium pl ant identification I made freque nt use of (Godfrey and Wooten 1979, Godfrey and Wooten 1981, Clewell 1985, Godfrey 1988, Wunderlin 1998, Weakley 2002). The vast majority of taxa were identifie d to species or variet y; low resolution taxa (family or genus) were omitted from analysis unl ess identification was consistent throughout the dataset. The term “species” is used throughout to refer to the lowest recognized taxonomic group. Surface and sub-surface soil samples were coll ected for nutrient and texture analysis. Four surface soil samples were collected from the upper 10 cm of mineral soil, and a single subsurface sample was collected from approximately 50 cm below surface. Dried samples were analyzed at Brookside Labs in New Knoxville, Oh io. Nutrient analyses was performed via Mehlich III extractant, an analyt ical procedure used for routine soil testing attempts to estimate the amount of soil nutrients available to the pl ant during its growing s eason (Mehlich 1984).

PAGE 81

81 Specific soil nutrient measurements were: tota l cation exchange capa city (meq/100g), pH, estimated nitrogen release (N, ppm), extractabl e phosphorous (P, ppm), exchangeable cations in ppm (Ca, Mg, K, Na), extractabl e micro-nutrients in ppm (B, Fe, Mn, Cu, Zn, Al), soluble sulfur (S) and bulk density. Percent organic matter was determined by loss-on-ignition. Texture analysis quantified percent sand, silt, and clay of su rface and sub-soil samples. Climate and elevation data were obtained for each ge ographic plot location. I downloaded extrapolated weather parameters for specific locations from the DayMet climatological model, available online ( www.daymet.org ). The Daymet model uses weather station and elevation data to produce smoothed parameter estimates on a 1 km gridded surface over the conterminous United States (Thornton et al. 1999). Daily parameter values were available for an eighteen year period between 1980 and 1997. I calculated annual means and standard deviations for the following: daily maximum air temperature, daily minimum air temperature, daily average air temperature, total daily precipi tation, and total daily shortwave radiation. I downloaded eleva tion estimates for each geographic plot location from the HYDRO 1K North America DEM model webpage pr ovided by the U.S. Geological Survey ( http://edc.usgs.gov/products/e levation/gtopo30/hydro/na_dem.html ). Elevation values were derived from a digital elevati on model with 1 km resolution. Numerical Data Assembly and Analysis The response data matrix was assembled from species cover data from 270 samples. Cover values for pine species (genus Pinus ) were omitted from this the data matrix, although other woody species cover values were retained. Species with fewer than three occurrences were deleted (McCune and Grace 2002).

PAGE 82

82 Species data were relativized to maximum sp ecies values and were transformed using a Hellinger distance measure. When used in co njunction with Euclidean distance metrics and linear ordination, the He llinger transformation affects adequate representation of complex species data without the problems associated with species weigh ting (i.e. chi-square distance based methods; (Legendre and Gallagher 2001, Legendre et al. 2005). I assembled four data matrices representing groups of potential expl anatory variables for statistical modeling. Collectively, these represent environmental and spatial explanatory factors. The first, referred to as the edaphic varable matrix (EVM), initially included 24 soil descriptors (listed above) and two variable s describing local topography and el evation. Variable “topo” was a subjectively assigned descriptor of local t opographic position relative to surrounding landscape (value 1 to 4). The elevation (“elev”) variable was the actual plot el evation derived from the U.S.G.S. digital el evation model. The second matrix was the climate variable matrix (CVM). It included means and standard deviations calculated from the five DayMet parameter values (listed above). In addition, I calculated means and standard deviat ions for total precipita tion and daily shortwave radiation for the growing season only (values fr om March 15 – October 31). A total of 14 climate variables were included as potential explanatory descriptors in the initial CVG. When necessary, soil, topographic, and climate environmental variables were transformed to approximate normal distributions Soil variables measured in ppm were log transformed. Logit transformations were appl ied to proportional data (Tabachnick and Fidell 1996). Because of the varying scales and ranges of soils and climate variables, all EVM and CVM variables were standardized to z-scores, expressed as standard deviations from the mean (Tabachnick and Fidell 1 996, Legendre and Legendre 1998).

PAGE 83

83 The third matrix of potential e xplanatory variables included de scriptors of spatial patterns in the species data. The spatial variable matrix (SVM) described a trend surface response model of geographic locations. Geographi c coordinates of plot locations (X and Y, centered by mean) were calculated from latitudes and longitudes superimposed on a geographic grid. Following Bocard et al. (1992) and othe rs (Borcard and Legendre 1994, Okland and Eilersten 1994, Heikkinen and Birks 1996, Legendre and Legendre 1998), seven additional terms were included in the initial SVM representing nine terms of a third-order polynomial regression of X and Y coordinates: Z = b1X + b2Y + b3X2 + b4XY + b5Y2 + b6X3 + b7X2Y + b8XY2 + b9Y3 This approach allowed modeling of spatial trends that are more complex than linear gradients (Legendre and Fortin 1989, Le gendre and Legendre 1998). The final explanatory matrix was based on a simple regional model of Florida physiographic landforms. The regional variable matrix (RVM) categorized each plot location into one of the four regions based on the ge neralized physiographic la ndforms of Puri and Vernon (1964). In addition to the Northern and Central Highla nds, the Lowlands landform was divided into the panhandle and peninsula regions (Fi gure 3-1). This regional delineation approximates a general regionalized model ba sed on several widely applied models of Southeastern U.S. ecoregions (Omernik 1987, Baile y et al. 1994, Griffith et al. 1994, The Nature Conservancy 2001). This model represents pres umed regional differences in geologic and evolutionary history that affect current patterns of spatial heterogeneity of pi neland vegetation. All environmental and spatial variables were individually screened fo r inclusion in their respective variable matrix using the forward se lection procedure and a ssociated Monte Carlo tests as implemented in CANOCO (Okland and Eilersten 1994, ter Braak and Smilauer 2002,

PAGE 84

84 Leps and Smilauer 2003). Two forward selection te sts were conducted. First, variables were subjected to forward selection in the context of a redundancy analys is (RDA) of a single variable group. Second, forward selection wa s repeated in a partial redunda ncy analysis (pRDA) of each variable group, with other envir onmental/spatial variables as cova riables. Variables with p > 0.02 were excluded in subsequent canonical an alyses corresponding to the model used for selection. In this manner, expl anatory variables with highest partial corre lations with species data were retained. I applied a variation partitioning model to th e species data, using the EVM, CVM, SVM as explanatory variable matrices Variation of the species data was decomposed into components associated with pure and joint contributions of explanatory factors. Sp ecifically, I quantified seven components of variation fr om six individual RDAs and pRDAs. The specific components described fractions of total vari ation explained (TVE). These frac tions are expressed in terms of three pure factor effects and f our interaction effects (followi ng Cushman and McGarigal 2002). Variation partitions correspond to fractions of reference diagra m in Figure 3-2 as follows: 1. Pure edaphic effects : species variation explained by so ils and topography variables, and not related to climate an d space variables (fraction a ) 2. Pure climate effects : variation explained by CVM but not EVM and SVM (fraction b ) 3. Pure spatial effects : variation explained by SVM but not EVM and CVM (fraction c ) 4. Joint effects of edaphic and spatial variables: variation jointly explained by EVM and SVM, but not relate d to CVM (fraction d ) 5. Joint effects of edaphic and climate variables: variation jointly explained by EVM and CVM, but not relate d to SVM (fraction e ) 6. Joint effects of climat e and spatial variables: variation jointly explained by CVM and SVM, but not relate d to EVM (fraction f )

PAGE 85

85 7. Three way joint effects of edaphi c, climate and sp atial variables: variation jointly explained by EVM, CVM, and SVM (fraction g ) The first variation partitioning model include d EVM and CVM plus the polynomial spatial matrix (POLY SVM). In the second model, the PCNM SVM replaced the POLY SVM. Models of variation partitioning involved application of a series of constrained and partial constrained canonical ordinations, as described by Borcard et al. 1992 and others (Borcard and Legendre 1994, Okland and Eilersten 1994, Okland 2003) Significance of terms derived from canonical ordination models were tested via Monte Carlo permuta tion tests (Peres-Neto et al. 2006). The null hypothesis tested was that of in dependence of species response data on the values of the explanatory variables (Leps and Smilauer 2003). Fractions representing two and three-way interactions and “unexplained” residua l variation were calcul ated indirectly from simple and partial terms; therefore, they were not statistically testable (Legendre and Legendre 1998, Peres-Neto et al. 2006). Variation partitio ning and statistical tests were performed using the vegan community ecology package version 1 .8 for R software (Oksanen et al. 2007, R Development Core Team 2007). The relationships between floristic variation and abiotic gradients were assessed in the context of the larger va riation partition model. Constrained axes were tested via Monte Carlo permutations for each of the following canonica l ordinations: 1) RDA of EVM, 2) pRDA of EVM after removal of CVM and POLY SVM effects, 3) RDA of CVM, and 4) pRDA of CVM after removal of EVM and POLY SVM effects. Higher order canonical axes were tested using pRDAs with lower order axes scores as c ovariables (Braak and Smilauer 2002, Leps and Smilauer 2003). Significance of marginal effects of sequential canonical axes was assessed at p < 0.02. Multiple simple correlations are presente d as vector biplots s uperimposed on ordination diagrams. Angles of the vectors denote direct ion of the highest corr elation whereas vector

PAGE 86

86 lengths correspond to strength of correlation. Significant correlations between canonical axes and species richness (number species per 1000 m2 plot area) are similarly presented. Canonical axes scores derived from rele vant RDA and pRDA ordinations are plotted agai nst geographic plot coordinates to visualize m odeled trends in environmental a nd spatial varia tion. Individual ordinations were run using CANOCO for Windo ws version 4.5 (ter Braak and Smilauer 2002). Results A total of 1009 taxa were identified from the 270 vegetation samples used in this study, after omission of inconsistent and low resoluti on taxonomic identifications. After deletion of infrequent entities, a total of 670 species were retained for analysis. Species richness ranged widely in the samples, fr om 29 to 168 species/1000 m2 area. Similarly, vegetation samples represent large variation in community co mposition, ranging from dry upland sandhills to seasonally inundated wetland prairi es. Community types and ch aracteristic environmental features are described in de tail in preceding chapters. About a quarter of the edaphic variables were omitted from inclusion in the EVM explanatory matrix following the forward se lection procedure in CANOCO (Okland and Eilersten 1994, ter Braak and Smilauer 2002, Leps and Smilauer 2003). Of the original 26 potential explanatory variables, 14 soil property variables and two topographic variables were retained for the RDA including the EVM explanator y factor only, with no covariables (see Table 3-2 for variable list). Thirteen soils and t opographic variables were selected for the pRDA of EVM, with CVM and SVM included as covariable s. Three variables from the RDA set were omitted in the pRDA set (Clay A, P, and Mn), while one was added (Al). Of the initial 12 climate variables subjected to the forward selection procedure, eight were retained in the CVM for RDA ordination. These included descriptors of temperature,

PAGE 87

87 radiation, and precipita tion. Forward selection for the pRDA of CVM (with EVM and SVM covariables) reduced the number of explanatory variables to only three (standard deviation and total growing season precipitation, and standard deviation of daily shortwave radiation). Similar forward selection procedures were applie d to sets of spatial explanatory variables. Of the nine terms of the initial polynomial tr end regression, seven were retained for the SVM (X2Y and Y3 were dropped). Variation Partitioning Models The total variation (aka “total inertia”) of the Hellinger transformed species data set was 0.727, as expressed by the sum of all eigenvalue s in an unconstrained principal component analysis (PCA). The three explanatory matrices of the first variation partitioning model accounted for 23% of the total variation (sum of all constrained eige nvalues / 0.727), leaving 77% of total inertia as “unexplained residual va riation” (Figure 3-1). Although 23% is a small portion of total inertia, this “total variation ex plained” (TVE) value is typical of community variation studies (typically 20-50%; Okland and Eilersten 1994, Okland 1999). However, following the advice of Okland (1999), TVE as a portion of total inertia should be interpreted with extreme caution. The contribution of polynomia l distortion (an artifact of eigenanalysis of numerous response variables) was estimated as 30-70% of “unexplained va riation” in simulation studies of “noisy” data. The aut hors recommend presentation of va riation fractions in terms of proportions of TVE. This conve ntion is followe d hereafter. Edaphic variables account for the largest com ponent of species variation in the variation partition model. Approximately 70% of TVE is related EVM, eith er as pure or interactions effects: fraction a includes pure edaphic effects (48%), and fraction g, the three way interaction effects (22%; Figure 3-1). The jo int effects of edaphic variables with space (fraction b, 0%) and

PAGE 88

88 with climate (fraction e, 0%) are too small to be included in th e model. Variation attributed to climate explanatory variables (pure or joint e ffects) is 44% of TVE (Figure 3-1). A small percentage (9%) of TVE is pure climate effect ( fraction b), and the remainder is divided between space+climate joint effects (fraction f: 13%), and the three way interaction (fraction g). Similar variation fractions are related to spatial trends with only 9% attrib uted to pure space effects. Environmental Explanatory Variables The primary complex gradient related to to tal edaphic variation (the EVM RDA; Figure 3-2) remains intact in the model of “pure” edap hic effects only, whereas higher order gradients are diminished (Figure 3-3). The two-dimensi onal pRDA of EVM represents edaphic variation after the removal of spatial struct ure (Figure 3-3 relative to fracti on (a) of Figure 3-1). Similar to the EVM RDA, axis one separates sites of lower slopes with organic soils from sandier sites on higher topographic positions (first axis Figure 3-3 compared to Figur e 3-2). Conversely, the soil texture/nutrient gradient asso ciated with geographic regi on (Axis 2 Figure 3-2) largely disappeared in the pRDA (Figure 33), as did any evidence of regi onal separation of sites. The second pRDA axis appears to represent a similar gr adient of soil acidity and fertility similar to the third RDA axis in Figure 3-2. Geographic separation is dramatic in the ordination of climat e variables related to species variation. The CVM RDA represents all variation fractions associ ated with climate explanatory factors (Figure 3-4 relative to fractions b+f+g of Figure 3-1). The first canonical RDA axis is a gradient of all eight climate vari ables, related to temperature, daily radiation, and precipitation (Figure 3-4). Peninsula sites are completely se parated from panhandle s ites along the first axes, and they are characterized by higher mean a nnual temperatures and daily radiation. The panhandle sites receive more annual rainfall that is more variables throughout the year and the

PAGE 89

89 growing season. Species richness is also highly correlated with this climate gradient and regional separation. The CVM RDA axis 2 explai ns about half again as much variation in species data (Table 3-2), and is related to variation in precipita tion (Table 3-1). Likewise, the third canonical axis represents a gradient in pr ecipitation (annual, grow ing season and standard deviation) and radiation (d aily mean and standard de viation; Table 3-1). The influence of most climate explanatory va riables disappears after removal of variation related to spatial and edaphic factors. The CVM pRDA represen ts the variation fraction of “pure” climate effects only, which is very small co mpared to total climate effects (bottom plot of Figure 3-4 relative to fraction (b ) of Figure 3-1). Not surprisingl y, most of the climate variables are not correlated with this sma ll, non-spatially related variation fraction (and were eliminated in the forward selection process prio r to the pRDA). Regional site separation similarly disappears in the pRDA (Figure 3-4). The first pRDA axis is related to total growing season precipitation and variation in precipitation (Tab le 3-1). Axis two is a gradie nt of variation in radiation (probably related to seasonality). Species richness is negatively correlated with pRDA axis 1, although the relationship appear s non-linear (Figure 3-4). Mapped results from the simplified variation partition model of envi ronmental and spatial trends reveal regional patterns in species variation. The simp le two part variation partition model yielded the following fractions of TVE (figure not shown): pure environmental effects (fraction a = 54%), pure spatial e ffects (fraction c = 9%), and jo int effects of environment and space (fraction b = 37%). These fractions resemb le those of the initial three part variation partition model; the spatial component (fraction c) is identical in both models. The contour map corresponding to site scores from a RDA is constrained by all environmental and spatial variables (fractions a+b+c = TVE) and displays regional distincti ons (Figure 3-5). Scores from

PAGE 90

90 the first and second canonical axes appear to di stinguish the highlands physiographic landforms in the upper panhandle and north peninsula from the coastal lowands (Figure 3-5a). Although it represents a small portion of TVE (9%), the canonical axis site scores for variation explained by pure space only display clear separation betw een panhandle and penins ula (Figure 3-5c). Discussion Based on the spatially explicit models of species-environment relationships, I propose several interpretations regarding determinants of understory community structure of Florida pyrogenic pinelands. First, environmental factors influence sp ecies composition and diversity, and these controls are most prominent at local scales. The strongest gradient in species composition is related to local topographic and ed aphic features; about ha lf of TVE is uniquely correlated with these factors and not spatially structured (presumably representing small scale gradients not captured by large scale spatial tr ends). Specifically, I presume that community differentiation is concurrent with gradients of so il fertility and soil moisture, represented by total N, organic matter, topographic position and soil texture\density variation. This gradient separates dry upland vegetation from herbaceous dominated flatwoods and wetlands. Previous community classification characterized vegeta tion along these primary gradients; sandhill vegetation inhabits infertile sands (negatively associated with Ax is 1, Figures 3-2 and 3-3) while flatwoods and wetland vegetation are common in wetter, acidic sites with high nitrogen availability and organic matter. The dominan ce of the soil fertility/moisture gradient is demonstrated by large first axis eigenvalues, relativ e to those of higher order. Furthermore, this local-scale gradient is consistently prominent after removal of variation effects of spatial trends and climate.

PAGE 91

91 Topographic position was the single most in fluential environmental variable of community composition, and the most highly correlat ed with the local soil moisture-fertility gradient. Understory community variation concurrent with local topographic gradients has been noted elsewhere in the Southeastern Coastal Pl ain, both anecdotally and quantitatively (Walker and Peet 1983, Bridges and Orzell 1989, Myers and Ewel 1990, Platt 1999, Kirkman et al. 2001, Drewa and Platt 2002). In the current study, top ographic position was a subjective descriptor of position along a local topographic gr adient, regardless of absolute changes in elevation or slope steepness. As revealed in the ordination mode ls, topographic position covaries with soil texture and fertility. However, in light of the large por tion of variation attributed to topographic position over and above that explained by other edaphic feature, this va riable likely is a proxy for unmeasured variation related to available soil moisture. Lower slope pineland communities typically have higher perched water tables, more seasonal flooding and less soil leaching (Abrahamson and Hartnett 1990, Myers a nd Ewel 1990, Kirkman et al. 2001). Available nitrogen is related to local-scale variation in species composition in the current study. In temperate grasslands elsewhere, nitr ogen is a limiting resource and is related to primary productivity and diversity (Seastedt et al. 1991, but see Turner et al. 1997). Soil moisture and nitrogen availability are thought to be positively related in temperate forests (Vitousek 1982). In this model, nitrogen is highly positivel y correlated with soil organic matter and (presumably) soil moisture, and negatively relate d to soil density and co arseness. This is not unexpected, as organic matter is a major source of nitrogen (Brady and Weil 2000). However, these correlations contradict studi es of nitrogen dynamics elsewher e in the Southeastern Coastal Plain where nitrogen mineralization declined with increasing soil moisture, and/or was negatively correlated with species diversity(Fo ster and Gross 1998, Wilson et al. 1999, Kirkman

PAGE 92

92 et al. 2001). These studies involved correlation over single or few recently burned topographicmoisture gradients, whereas the current study m odels environmental correlations and composite gradients over a large geographic region. Available nitrogen varies with time since fire and fire frequency (Christensen 1977). Despite attempts to “control” for recent fire history in site selection, this may account for unmeasured variat ion in nitrogen (and other locally available resources). The second interpretation is that regional variation in pinela nd community structure is profound, and is related to varia tion of soil texture, soil nutr ient availability, and climate Furthermore, regional variation of species composition is orthogonal to local variation Spatial trends in compositional variation are strongly dependent on latitude and longitude (X and Y coordinates), and correspond to regional diffe rences between the Florida panhandle and peninsula. Eighty percent of va riation related to climate variables is spatially structured, reflecting distinct regional differences; the peni nsula is hotter, less seasonal, and receives more growing season rainfall, whereas panhandle rainfall is more even ly distributed throughout the year (Fernald 1981, Chen and Gerber 1990). Regiona l differences in edaphic features represent about half of spatially structure variation in sp ecies composition, and largely reflect soil texture differences. Similar regional segregation concurrent with so il texture has been documented elsewhere in the coastal plain, at both local and regional scales (P eet and Allard 1993, Dilustro et al. 2002). Phosphorous and calcium are more abundant in soils of peninsular sites. This is not surprising considering the presence of the carbonate Florida platfo rm that underlies most of the peninsula, and the presence of large amounts of phosphorite in so me sediments of Pliestocene origin (Puri and Vernon 1964, Brown et al. 1990).

PAGE 93

93 The third interpretation is that residual spatially structured composition variation in is related to regional differentiation in phytogeographic distributions and endemism Furthermore, these patterns may reflect variations in biogeogr aphic and evolutionary history, and/or recent land use patterns. The unique fraction of space in the variation partition model represents a small but inscrutable percentage of TVE (9%). This variation displa ys regional distinction similar to that of the joint effects of space and environmental factors (Figure 3-5). Interestingly, the divide between peninsular central Florida vs. north and panhandle Florida coincides with phytogeographic patterns. Many temperate pl ant species (woody and herbaceous) reach the southern limit of their distributions in north peni nsular Florida (see Chapter 2). Nearly a quarter of the taxa included in the species-environment mode ls have regionally restri cted distributions in Florida, and nearly 3% are endemic to one region (Chapter 2; Jame s 1961, Sorrie and Weakley 2002). If the contribution of “pure space” does i ndeed manifest phytogeographic trends, this suggests the influence of historical dynamics on contemporary patterns of species coexistence (Ricklefs 1987). Florida’s complex recent geologic history also underscores regional differences between the panhandle and penins ula, including differential so urces and timing of sediment deposition and histories of sea level fluctuations (Randazzo a nd Jones 1997, Myers 2000). In addition, there is evidence that the two regi ons were physically separated by the “Suwannee Strait” for a period between 12 to 30 MRBP; an elongate negative structure extending across southern Georgia and north eastern Florida (Hul l 1962, Puri and Vernon 1964, Myers 2000). Differential land use patterns offer an alternat ive (but not mutually exclusive) explanation for the compositional variation concurrent with regional segregation. Unfortunately this observational study of floristic variation does not de scribe pre-settlement c onditions, as this is no longer possible because past land use and ma nagement created non-random “selections” of

PAGE 94

94 natural areas in Florida. Further subjectivity was introduced from the lack of random selection of sites from an a priori “populat ion” of natural areas. I attemp ted to minimize the latter problem with a stratified sampling design and a larg e sample size (see Leps and Smilauer 2007). However, any described variation of community structure is inherently confounded with recent land use, particularly fire suppression and l ogging, as has been documented in other regions (McIntyre and Lavorel 1994, St ohlgren et al. 1999, Vandvik and Birks 2002, Svenning and Skov 2005). Because open-range ranching was common in the central peninsula until recently, large portions of this region continue d to be frequently burned dur ing the dormant season (Bridges 2006a, Bridges 2006b). Conversely, other region s of Florida suffered decades of fire suppression in the 20th century, with prescribed fire only r ecently introduced in selected natural areas (e.g. the Big Bend and Marianna Lowlands re gions, pers. obs.). Thus, regional differences in recent fire regimes may contribute to une xplained compositional variation in the current model. Effect may be direct (f ire effects on resource availability, plant delectabilit y) and indirect (i.e. differential resource availability re lated to timber density and woody biomass). A fourth interpretation is that gradients in composition are related to gradients in species richness, and these are apparent at regional and local scales The variation in species richness is high among study sites, rangi ng from 26 to 168 species/1000 m2. These herbaceous dominated pineland communities are characterized by large numbers of small-statured species present in low abundances. Community structur e is influenced by the amount of “species packing” at small scales. The richness gradient is most obvious at th e regional scale, where panhandle sites are consistently richer regardle ss of soil moisture/fertil ity conditions. Regional influence on local diversity is a well documented phenomenon, and is attributed to the “species pool” effect resulting from pro cesses operating at multiple spatia l and temporal scales (Zobel

PAGE 95

95 1992, 1997, Collins et al. 2002). Interestingly, the ri chness gradient appears independent of the primary local-scale gradient of soil moisture/fertil ity, contrary to observati ons in other grassland ecosystems (Grace et al. 2000, Kirkman et al. 2001, Weiher et al. 2004). After removal of variation associated with regional spatial trends, a richness gradient persists and is weakly associated with soil pH available nutrients and soil texture. This secondary richness gradient is s eemingly unrelated to regional se gregation, and may reflect local diversity patterns. Species richness associa tions with soil pH and calcium have been documented in temperate grasslands and forest s (Partel 2002, Palmer et al. 2003, Peet et al. 2003). Similar to Peet and Allard (2003), Fl orida pineland species richness is positively correlated with pH and soil calcium, suggesting e ither larger pools of species adapted to basic soils (regional “species pool” e ffect) or more favorable local conditions for plant colonization and growth (local environmental effect). Alternativ ely, soil reaction is me rely a proxy variable for other unmeasured causative factors, such as competition for light or space. Density of woody biomass increases rapidly on more fertile sites, which affects understory species richness vis-avis competition for light and other resources (White et al. 1991, Streng et al. 1993, Grace and Pugesek 1997, Palmer et al. 2003, Weiher et al. 2004). Fire encourages herbaceous growth, colonization and diversity, in part, through cont rol of woody competition (Drewa et al. 2002b). Thus, the local variation in ric hness likely derives from a comple x gradient of soil fertility and disturbance. The current model of community variation of Florida pineland vegetation underscores the prominence of spatially structured and spatially independent environmen tal factors in shaping community structure. Spatial structure in co mmunity structure and environment patterns is common in studies of local to meso-scale variat ion (study region scale range approximately 10 –

PAGE 96

96 1000 km2; Abrahamson and Hartnett 1990, Cushman a nd McGarigal 2002, Dilustro et al. 2002, Graae et al. 2004,Svenning and Skov 2005, Laughlin and Abella 2007). By comparison, the extent of my study region was orders of magnitude larger (a pproximately 137,000 km2). Considering the relatively large region, it is somewhat surprising that the unique variation fraction explained by space is small compared to total variation explaine d (9%). This suggests the minor importance of biotic processes ope rating independently of environment. An alternative explanation is that I failed to adequately model rele vant spatial trends (e.g. small scale spatial patterns). On the other hand, soils and top ography appear very influe ntial as determinants of compositional variation supporting environmental control hypotheses. I posit that the relative influence of environmental controls exceeds that of biological controls in species composition and diversity of Florida pyrogenic pineland communities. This model of pineland floristic variation supports hypotheses of regional influences on local community structure and diversity. Regional differenc es in species richness and composition exist, even after removal of regiona l environmental effects. Species composition differs between the panhandle and peninusula site s with similar local environmental conditions. This observations suggest influences of pa leogeography and evoluti onary history through mediation of species pools (Ricklefs 1987, Zobel 1992, Zobel 1997), perhaps confounded with trends in recent land use hi story (Graae et al. 2004, Graham et al. 2005, Laughlin et al. 2005, Svenning and Skov 2005). Furthermore, the m odel suggests a hierarchical structure of ecological determinants relative to the focal ecosystems, and that relative influences of environmental factors are scale dependent.

PAGE 97

Table 3-1: List of variables included in RDA and pRDA canonical ordination of va riation partitioning models. Eigenvalue indicates conditional correlation of single variables (with a ll other variables covariables). Correlation coefficients listed for first three constrained RDA axes, and two cons trained pRDA axes. Bold va lues indicated significant correlations p < 0.05. Abbreviation Variable Eigenvalue RDA A1 RDA A2 RDA A3 pRDA A1 pRDA A2 Edaphic Variable Matrix (EVM) Topo Relative position on slope (1-4) 0.08 0.79 0.05 0.06 0.82 0.03 Org Organic matter surface soil (%) 0.03 0.37 -0.21 0.28 0.41 -0.03 Sand A Sand in surface soil (%) 0.03 -0.25 -0.52 -0.19 -0.39 -0.18 Sand B Sand in sub-soil (%) 0.03 -0.04 -0.63 -0.08 -0.15 -0.24 N Estimated total exractable nitrogen (ppm) 0.03 0.42 -0.16 0.25 0.43 -0.12 Density Bulk density (mg/m3) 0.03 -0.41 0.09 -0.34 -0.44 -0.03 Elev Elevation (m) from 1 km resolution DEM coverage 0.02 -0.14 0.50 0.23 Clay A Clay in surface soil (%) 0.02 0.04 0.50 0.04 pH pH surface soil 0.02 -0.25 0.26 0.25 -0.27 0.43 P Extractable phosphorous (ppm) 0.02 -0.09 -0.36 0.42 Ca Calcium (ppm) 0.02 -0.12 -0.39 0.42 -0.19 0.12 B Boron (ppm) 0.02 -0.29 0.07 0.37 -0.33 0.42 Mn Manganese (ppm) 0.02 0.04 0.54 -0.34 Fe Iron (ppm) 0.01 -0.14 0.29 0.04 -0.20 0.14 Al Aluminum (ppm) 0.01 0.22 -0.40 Climate Variable Matrix (CVM) Temp mean Mean annual daily temperature ( C) 0.05 0.79 0.08 -0.13 Temp max Mean annual minimum temperature ( C) 0.05 0.76 0.15 -0.17 Srad GS std Standard deviation mean growing season shortwave radiation (MJ/m2/day) 0.05 -0.78 -0.04 0.17 0.02 0.50 97

PAGE 98

Table 3-1 continued. Abbreviation Variable Eigenvalue RDA A1 RDA A2 RDA A3 pRDA A1 pRDA A2 Srad std Standard deviation mean annual shortwave radiation (MJ/m2/day) 0.04 -0.67 0.05 0.32 Prcp_ann Mean total annual precipitation (cm) 0.03 -0.50 -0.33 0.36 Srad Mean daily shortwave radiation (MJ/m2/day) 0.02 0.38 0.16 -0.38 Prcp GS Mean total growing season precipitation (cm) 0.02 0.30 -0.27 0.45 0.47 0.16 Prcp std Standard deviation of mean total growing season precipitation (cm) 0.02 -0.27 -0.36 0.36 0.48 0.33 98

PAGE 99

99 Table 3-2: Results for Monte Carlo tests of canonical axes, for each of four ordinations. Variation attributable to explanatory constraining variables indicated by the culmulative percentage of variance (Cumula tive % spp-env). F-ratio and p-value for each test of canonical axis after partiall ing out variation attr ibutable to lower dimension axes. Canonical Model Axis Axis eigenvalue Culmulative % spp-env F-ratio p-value RDA Edaphic variable matrix (14 vars) 1 0.095 45.1 26.67 0.002 2 0.042 65.0 12.29 0.002 3 0.019 73.4 5.73 0.002 4 0.014 78.7 4.38 0.002 pRDA Edaphic variable matrix (12 vars) 1 0.081 57.6 26.06 0.002 2 0.012 66.2 3.83 0.004 RDA Climate variable matrix (8 vars) 1 0.055 42.8 15.27 0.002 2 0.028 64.4 7.97 0.002 3 0.017 77.4 4.88 0.002 4 0.010 85.2 2.93 0.006 pRDA Climate variable matrix (4 vars) 1 0.010 39.6 3.60 0.002 2 0.005 72.7 1.87 0.016

PAGE 100

100 Figure 3-1: Venn diagrams of va riation partition model. Ellips es represent three explanatory matrices (see text). Shaded portions labe led with letters denote variation fractions and proportion of TVE (see refere nce diagram top right). (a) 0.48 (b) 0.09 (c) 0.09 (g) 0.22 (f) 0.13 Variance not explained = 0.77CVMEVM SVM(d) 0.00 (e) 0.00 d b ae c g fResidual

PAGE 101

101 Figure 3-2: Biplots of RDA ordination, constrained by edaphi c variables (EVM). Top plot = canonical axes 1 vs. 2; bottom plot = ca nonical axes 1 vs. 3. Vectors denote individual soils and eleva tion variables, scales by direction and magnitude of correlation with axes. Abbreviations same as Table 3-1. Symbols denote regional locations (panhandle vs. peninsula). Top plot contours display significant correlation of species richne ss with Axis 2 (r = 0.48). 60 65 70 75 80 85 90 95 100 105 110 115 -0.8-0.6-0.4-0.20.00.20.40.60.81.0 -0.4 -0.2 0.0 0.2 0.4 0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 Elev Topo Org Sand A Clay A Sand B pH N P Density Ca B Fe MnRDAAxis 1RD A A xis 2 RDAAxis 3Elev Topo Org Sand A Clay A Sand B pH N P Density Ca B Fe MnPeninsula Panhandle

PAGE 102

102 Figure 3-3: Biplot of pRDA constrained by edaphic vari ables (EVM) with CVM and POLY SVM covariables. Vectors show correla tions of individual soils and elevation variables with the first two canonical axes Abbreviations listed in Table 3-1. Symbols indicate regional lo cations (panhandle vs. peni nsula). Contours display significant correlation of species ri chness with Axis 2 (r = 0.59). -0.6-0.4-0.20.00.20.40.60.8 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 pRDAAxis 1pRD A Axis 2 Peninsula Panhandle 60 65 70 75 80 85 90 95 100 105 Topo Org Sand A Sand B pH N Density Ca B Fe Al

PAGE 103

103 Figure 3-4: Biplots of 1) R DA constrained by climate variable s (CVM; top plot) and 2) pRDA of CVM with EVM and POLY SVM covari ables. Vectors show correlations of individual climate variables with the first two canonical axes. Abbreviations listed in Table 3-1. Symbols indicate regional locations (panhandle vs. peninsula). Contours display significant co rrelation of species richness with Axis 1 (RDA: r = -0.593, pRDA: r = -0.34). Note diffe rent scales for plots. -0.6-0.4-0.20.00.20.40.6 -0.4 -0.2 0.0 0.2 0.4 0.6Temp mean Temp max Prcp ann Srad Srad GS std Prcp std Srad std 70 75 80 85 90 95 100 105 110 115 120 40 45 50 55 60 60 65 65 70 70 75 75 75 80 80 80 85 85 85 90 90 95-0.2 0.0 0.2 0.4 0.6 -0.4 -0.4-0.20.00.20.40.6 0.8 0.8 Peninsula Panhandle RDAAxis 1RDAAxis 2pRDAAxis 1pRD A Axis 2Prcp GS(a) (b) Srad std Prcp std Prcp GS

PAGE 104

104 Figure 3-5: Contour maps de rived from constrained ordinati on axis scores, displaying geographic variation in vari ation parititions from the model of environmentalcompositional correlations. Plot (a) Axis 1 scores from RDA of all variation components (environmental and spatial fact ors), corresponds to fractions a+b+c in Figure 2-2. Plot (b) shows Axis 2 scores from the same RDA ordination. Plot (c) displays Axis 1 scores from pRDA of spa tial trends, after removal of environmental factors (corresponds to fraction c). Plotb:Axis2sitescores Fractions(a+b+c) Environment+Space Plotc:Axis1sitescores Fraction(c) Spaceonly Plota:Axis1sitescores Fractions(a+b+c) Environment+Space <-0.2<-0.4 -0.1-0.1 0.00.1 0.10.3 >0.2>0.6 Plotsa&bPlotc Canonical axis score

PAGE 105

105 CHAPTER 4 ECOLOGICAL RESTORATION OF A LO NGLEAF PINE SAVANNA IN THE SOUTHEASTERN COASTAL PLAIN Introduction Ecological restoration often invo lves rehabilitation of habita t structure to a semblance of historic or “natural” conditions. It is assumed that restoration of physical habitat structure and ecosystem function will prompt recovery or co lonization of desirable native populations and restoration of native diversity and composition (Palmer et al 1997, Walker and Silletti 2006). Community structure can be ma nipulated via reintroduction of natural processes (e.g., fire) and/or by “artificial” methods (mechanical or ch emical treatments). Reintroduction of natural disturbance regimes can variously influen ce community composition through mediation of species recruitment and mortality and biotic in teractions (Huston 1979, Wh ite 1979). Artificial manipulation of community structure may expedite restoration of desired conditions, particularly if remnant native populations remain on site, or if colonization is promoted (Palmer et al. 1997, Walker and Silletti 2006). It is the goal of ecological restoration to induced temporal changes in species composition and structure that resemble those of the desired conditions. Thus, it is important to quantify succession following restora tion, and compare that to reference conditions. Longleaf pine savannas and woodlands native to the Southeastern Coastal Plain are among the most imperiled ecosystems in North America (Croker 1987, Noss 1988, Frost 1993, Means 1996, Platt 1999, Frost 2006). Native longleaf pinelands currently occupy less than 3 percent of their former range (Frost 1993, Outcalt and Sheffiel d 1996). Sites with vegetation composition and structure similar to that of pr e-settlement conditions are even rarer (Simberloff 1993, Varner et al. 2005). The rapid range reduc tion of longleaf pinela nds coincided with extensive logging, agricultural land use, a nd expanding rural settlement in the 19th and 20th

PAGE 106

106 centuries (Croker 1987, Frost 1993). Virtually all old-growth longleaf pine was logged, and much of the remaining land converted to pine plantations of slash ( Pinus elliottii ) or loblolly ( P. taeda ) pines (Frost 1993). Other sites became overgrown with second growth pine and hardwood species due to land fragmentation and s uppression of natural fires (Platt 1999, Frost 2006). Natural longleaf pine savanna s of the Southeast are notable for their park-like stand structure and exceedingly diverse ground cover vegetation. Under natural fire regimes, monotypic stands of longleaf pine consist of patch ily distributed cohorts of even-aged trees (Platt et al. 1988b). Other trees and sh rub species are largely relegated to the midstory and understory strata and floristic diversity is concentrated in the herbaceous-dominated ground layer (Waldrop et al. 1992, Peet and Allard 1993, Glitzenstein et al. 1995, Platt 1999, Drewa and Platt 2002). At small scales, the ground cover vegetation of fire -maintained pinelands ha rbors exceedingly high plant species diversity (Walke r and Peet 1983, Bridges and Orze ll 1989, Peet and Allard 1993). Ground cover vegetation is comprised of large pere nnial bunch grasses, inte rspersed with smaller and rarer grasses and forbs (Peet and Allard 1993, Platt 1999). An estimated 95 percent of herbaceous ground cover species are perennials w ith adaptations for post-fire regeneration, including rapid growth and sprouting, clonal grow th, and obligate post-fir e seeding (Platt 1999). Post-fire growth of ground cover vegetation is rapid, with upwards of 100% biomass recovery within 1 year (Oesterheld et al. 1999). Ground cover vegetation, coupled with highly flammable longleaf pine needles, provides fine fuels nece ssary for ignition and spread of low intensity ground fires (Robbins and Myers 1992, Streng et al. 1993, Platt 1999). Most contemporary native longleaf pinela nds are small and fragmented, and are no longer subject to the natural processes to which constituent species are adapted. This condition

PAGE 107

107 precludes the natural occurrence of frequent, low intensity fires th at historically swept across the landscape (Frost 1993, Simberloff 1993, Platt 1999, Va nLear et al. 2005). Fire suppression of remnant natural areas has induced shifts in spec ies composition and community structure. Fire intolerant pine and hardwood species colonize these sites and reduce herbaceous plant abundance through shading and resource competition (Glit zenstein et al. 1995, Pla tt 1999, Provencher et al. 2000, VanLear et al. 2005). Fire suppressed lo ngleaf pine communities are common in the Southeast U.S. (Mehlman 1992, Brockway and Lewis 1997, Gilliam and Platt 1998, Varner et al. 2005). These sites may harbor persistent native plant populations (in dorma nt and active states), but have suffered degradation of community struct ure in terms of shifts in species composition and abundances and physical habitat. Reintroduction of natural fire regimes can a ffect recovery of community structure and species diversity in dry pinela nds that have suffered moderate fire suppression (Brockway and Outcalt 2000, Provencher et al. 2001, Kirkman et al. 2004, Walker and Silletti 2006). However, little is know about restoration of pinelands that have suffe red relatively long-term fire suppression (>10 years). The persis tence of native vegeta tion (perhaps in dormant states) is the factor determining whether restoration can be accomplished by restoring natural environmental conditions versus having to re-introduce native species, which is usually prohibitively difficult and expensive (Seaman 1998, Walker and Silletti 2006, but see Cox et al. 2004). Increases in woody biomass following fire suppression affect community structure and function of native pinelands. Thick woody growth competes with herbaceous vegetation for light and other resources, affecting succession, spatial heterogeneity, and species composition (Brockway and Lewis 1997, Provencher et al. 2 001, VanLear et al. 2005). Indirectly, woody encroachment increases litter fuel loads, which ma y alter fire intensity a nd behavior to conditions

PAGE 108

108 outside typical range of variab ility. Long periods of fire s uppression and/or high intensity wildfires may affect novel shifts in species co mposition and succession, or extirpation of native populations (Varner et al. 2005). Restoration success is highly dependent on initial site conditi ons. The degree of departure from a desired restored condition (the “reference condition”) guides the method and intensity of restoration treatment s (Fule et al. 1997, White and Wa lker 1997, Walker and Silletti 2006). Additionally, the reference condition prov ides a standard by which to evaluate site restoration progress and success (White and Walker 1997). Despite problems with historic differences in environment and st ochasticity, and assumed “stasis” of a target condition (Palmer et al. 1997), contemporary data fr om a disjunct but environmentall y relevant is often the best choice for a reference site (“same time, diffe rent place” sensu Walker and Silletti 2006). Similarly, historical data may serve as referenc e site conditions, particularly if they are proximate in geography and environment (Ful e et al. 1997, Swetnam et al. 1999, Walker and Silletti 2006). Ecological restoration progress is often measur ed in terms of change s in overall species richness or shifts in members of functional groups Dynamics of these measures are assumed to indicate changes in ecosystem production, functi on, and/or stability (Tilman 1996, Palmer et al. 1997). Species richness is often compared to reference sites to asse ss restoration progress (White and Walker 1997, Provencher et al. 2001, Ki rkman et al. 2004). However, increases in species numbers do not necessarily imply congr uence of successional trends or endpoints between restored and reference sites (Walke r and Silletti 2006). Explicit comparison of compositional data provides more informati on about succession response to restoration treatments.

PAGE 109

109 The purpose of this study was to evaluate the effects native stand composition and structure reconstruction have on the recovery of native ground c over vegetation. I hypothesized that woody biomass reduction via mechanical tr eatment would release remaining understory plant species from competition. Specifically, increased light availability from woody plant reduction will favor growth and recovery of herb aceous plants. I predicted that reduction of woody plant dominance via ecol ogically sensitive logging and reintroduction of native fire regime would induce recovery of native gr ound cover vegetation resembling reference site conditions. This prediction assumed a degree of resi lience in natural temporal variability in plant community succession. I measured the imme diate and long-term effects on ground cover composition following off-site pine removal (l ogging) and reintroduction of prescribed fire. Methods Study Site and Reference Sites The Abita Creek Flatwoods Preserve (ACP) is lo cated in the Gulf Coastal Plain region of Southeast Louisiana (The Nature Conservancy 2001). The ACP is situated on a broad flat terrace proximate to Abita Creek and its tributarie s. The site encompasses Pleistocene deposits of the Prairie and Citronelle terrace formations (The Na ture Conservancy 2001), and is characterized by level, poorly drained, fine silty loams dominated by the Stough soil series (Fragiaquic Paleudults; Tr ahan et al. 1990). Historical accounts of the ACP area describe a landscape of longleaf pine dominated savannas and flatwoods prior to extensive logg ing of the early 1900’s (Lockett 1870, Mohr 1898, Penfound and Watkins 1937, Penfound 1944, Wa hlenberg 1946). These pyrophytic communities contained monotypic longleaf pine canopies, with trees up to 200-300 years old interspersed with patches of younger pines, a nd lush ground cover vege tation of grasses and

PAGE 110

110 other herbaceous species (P enfound and Watkins 1937, Penfound 1944). Pine savannas were maintained by frequent (1-3 year interval) low intensity fires (Mohr 1898, Wahlenberg 1946, Glitzenstein et al. 2003). The Louisiana Field Office of The Nature Conservancy (TNC) acquired the 312 ha ACP tract in 1996. The decision to pur chase this tract was based in pa rt on the presence of remnant native ground cover vegetation and the presump tion that vegetation c ould be restored to approximate pre-settlement conditions. The site was never used for intensive agriculture. The original pine canopy was logged in the 1930’s; early aerial ph otographs show the area was treeless in the early 1940’s. Afterwards the site was bur ned infrequently, last burned circa 1980 before initiation of the current study (The Natu re Conservancy 1997). As a result of fire suppression, ACP was colonized by thick growth of tree and shrub species. By 1995, slash pine ( P. elliottii ) stands comprised the dominant ACP overs tory (The Nature Conservancy 1997). A few mature longleaf pines were present on site Because regeneration of this species was hindered by the competing thick sh rub vegetation, no juvenile longlea f pines were present. The midstory at ACP consisted of a nearly closed canopy of evergreen shrubs, including titi ( Cyrilla racemiflora ), sweetbay magnolia ( Magnolia virginiana ), large leaf gallberry ( Ilex coriacea ), swamp blackgum ( Nyssa biflora ), red maple ( Acer rubrum ), and gallberry ( I. glabra ). Herbaceous ground cover vegetation was sparse and patchy, interspersed within the shrub thickets. I selected Lake Ramsey Wildlife Management Area (hereafter “Lake Ramsey”) to serve as a contemporary reference site for comparisons of ground cover composition. Lake Ramsey contains native longleaf pine sa vanna vegetation that was freque ntly burned since the early 1990’s and is considered a high quality example of regional native vegetation, resembling pre-

PAGE 111

111 settlement conditions (Latimore Smith, pers. comm. ). Lake Ramsey is approximately ten miles west of ACP and is geologically and edaphically similar to ACP. The Louisiana Department of Wildlife and Fisheries Natural Heritage Program established permanen t vegetation monitoring plots at Lake Ramsey and collected vegeta tion data annually between 1994 and 1999. The 1999 data are used in this study. Secondly, I used species data from Penf ound and Watkins (1937; hereafter “Penfound data”) as a historical reference condition of unaltered pine savanna vegetation in southeastern Louisiana. The authors sampled ground cover ve getation of pine savannas and pine-cypress communities in the mid-1930’s, before and imme diately after the original pine canopy was logged. They recorded all ground co ver species present in plots of fixed dimensions comparable to that of the present study. Penfound’s “v irgin longleaf pineland community” and “cutover pine” sites (the “Knott tract… located some three miles East of Mandeville”; Penfound and Watkins 1937) were within 15 kilometers of present-day ACP. Penfound’s “slash pine-pond cypress” community was also w ithin 15 kilometers of my study area (“about 6 miles northeast of Mandeville”). I deemed the Penfound data represen tative of “original” natural conditions of these communities. Restoration Treatments and Sampling Methods My immediate goal at ACP was to restore th e structure and compos ition of midstory and overstory strata. My model was based on historic al and contemporary descri ptions of old-growth longleaf pine stands. These accounts describe un even-aged stands of longleaf pines with patches of even-aged cohorts resulting from gap regene ration. Historical accounts were anecdotal descriptions of Southeast Louisiana a nd surrounding regions (Lockett 1870, Mohr 1898, Penfound and Watkins 1937, Penfou nd 1944, Wahlenberg 1946). Cont emporary descriptions of

PAGE 112

112 old growth longleaf pine stand st ructure and dynamics came from elsewhere in the Southeastern Coastal Plain (Platt et al. 1988b, Platt et al. 1998; Platt and Rathbun 1993; Grace and Platt 1995a,b; Palik et al. 1997, 2003; Brockway a nd Outcalt 1998; McGu ire et al. 2001). A second goal was to restore longleaf pine canopy at ACP, and manage stands to promote older-growth characteristics. Natural seed sources for l ongleaf pine regeneration were unavailable. As such, I chose to remove canopy slash pine via commercial logging and plant longleaf pine seedlings. I predicted l ogging would reduce woody biomass and reduce competition for planted longleaf pine seedlings and recovering ground cover vegetation. In addition, I predicted faster establishment of longleaf pine canopy would provide fuel and promote fire behavior benefici al to natural community restor ation (Robertson and Ostertag 2007). However, I was also concerned with pot ential negative impacts from soil disturbance associated with the logging proce ss (Greenberg et al. 1995). Two restoration treatments were applied at ACP. The first was canopy removal via commercial logging followed by prescribed fires (h ereafter “logged+fire”). Five sample sites were selected for monitoring ground cover response to the logged+fire treatments. The second treatment consisted of prescribed fire only (hereafter “fire-only” treatment; four sites). Sample sites were interspersed throughout the preserve, and treatments applied following a complete randomized block design. All logged+fire si tes were logged during the winter of 1997-1998 (Figure 4-1). A consulting fore ster worked with a small loggi ng crew to remove trees using skidders equipped with low-psi tires. In this manner, soil disturbance and damage to surrounding trees were minimized. Logging was permitted only in dry conditions to minimize damage. All merchantable canopy pines were removed in the logged+fire sites. Mean basal areas (woody stems > 10 cm dbh) declined more th an 90% after logging from 22.8 5.9 m2/ha in 1997 to 2.2

PAGE 113

113 0.7 m2/ha in 1998. Over the same period, the mean basal area of the fire -only sites increased slightly from 9.6 6.0 m2/ha in 1997 to 9.8 5.4 m2/ha in 1998. All treatment sites were prescribed burned twice during the study period. The first burn occurred in April, 2000, followed by the second in April 2003 (Figure 4-1). In a given year, all study sites were burned in the same day under si milar conditions. Basal area of woody stems > 10 cm dbh were little affected by the burns. M ean basal area of the logged+fire treatment was already low following logging (2.2 0.7 m2/ha in 1998), and declined to 1.1 0.6 m2/ha in 2005. Similarly, the fire-only mean basal areas declined slightl y, from 9.8 5.4 m2/ha in 1998 to 8.6 4.3 m2/ha in 2005. Vegetation sampling in fixed permanent plots began in the summer-autumn of 1997, before logging. Vegetation plots were sampled a nnually during the autumn months prior to plant senescence after canopy removal (1998-2000). Autumn sampling faci litated plant identification given the large proportion of plant species, especially grasses, that flower in the fall. The final sample period was in the fall of 2005. At each treatment site, I installed a singl e 0.1-ha rectangular permanent plot (plot dimensions: 20 m x 50 m, 1000-m2 total area). All vascular plan t species were recorded in each of four series of nested subplots ranging from 0.01-m2 to 100-m2 area. I estimated aerial cover for each species recorded in 100-m2 subplots using the following c over classes: 0-1%, 1-2%, 25%, 5-10%, 10-25%, 25-50%, 50-75%, 75-95%, >95%. Additional taxa encountered in the remaining 600 m2 were recorded and assigned a nominal cover estimate. Vegetation sampling approximated the field methodology desc ribed by Peet et al. (1998). In addition, I recorded plant species’ cover in four 1-m2 subplots per site to obtain more accurate estimates of sm all scale cover. I overlaid a grid of 100 10 cm x 10 cm cells on each

PAGE 114

114 subplot; each species present was assigned a va lue of 1-100, corresponding to the number of cells in which it occurred. In this manner, species present in the 1-m2 subplots received two cover estimates: the first was the 1-m2 grid count, and the second was the cover class estimate from the 100-m2 sample area. All woody stems > 2 cm diameter breast he ight (dbh) were ta llied in the 1000-m2 sample area by species and size class (2-5 cm, then 5 cm size classes up to 40 cm dbh). For trees > 40 cm dbh actual diameter was recorded. Basal area (BA) was calculated using midpoint values per size class. I identified each plant tax on to the highest taxonomic resolution possible, which was species for the majority of identifications. Ho wever, I assigned “low-r esolution” identifiers corresponding to genus or family for sterile or unidentifiable taxa. Low-resolution taxa were included in species richness estimates, but omitted from datasets used for compositional analyses. Data Analysis I used repeated measures models to compare trends of species richness and woody species abundance between the tw o restoration treatments. My experimental design resembled a Before-After Control-Impact (BACI) design, with two levels of treatment rather than one treatment versus a control (Unde rwood 1994). I used mixed linear models consisting of fixed and random effects, because of repeated m easurements and my heterogeneous variancecovariance estimates (Littell et al. 2000). E ach model included two fixed effect factors: treatment (TRT: logged+fire vs. fi re-only) and time (YEAR: five levels corresponding to sample years), and their interaction (T RT*YEAR). Random effects of mixed models included betweenyear variances and within-year covariances of repeated measurements.

PAGE 115

115 Within-subject measurements are often seria lly correlated in a predictable manner over time. To accommodate this, I used the first three steps of the “four stage” method of Littell et al. (2000) in determining mixed effects models fo r each response variable. First, I applied a “saturated” model that included TRT and YEAR effects, plus pr e-treatment basal area per 1000m2 area as a covariable (initial BA). The variation of initial BA was large, ranging from 0.96 to 3.93 m2/ha. Initial BA was omitted from subsequent models if it failed to explain significant variation (p > 0.05 in the model including “unstructured” covari ance parameter estimates). Step two involved specifying models of cova riance structure for each statistical model. For this, the fixed effects portion of the model remained constant while I tested different covariance structure models using Residual Maximum Likelihood (REML) computation in PROC MIXED (SAS Version 8; Littell et al. 19 96, Littell et al. 2000). I selected a covariance structure for each statistical model which most cl osely approximated actual data with the fewest parameter estimates. I compared the following model structures: compound symmetric (2 parameters: homogeneous variance and covari ance), heterogeneous compound symmetric (6 parameters: heterogeneous variances and homogen eous covariance), Toeplitz (5 parameters: homogeneous variance and heteroge neous covariance as functions of time lag), heterogeneous Toeplitz (9 parameters: same as previous, with heterogeneous variance), first order autoregressive (2 parameters: homogeneous vari ance and decreasing serial covariance dependent on increasing time lag), first order heterogeneous au toregressive (6 parameters: same as previous with heterogeneous variance), and antedependenc e (9 parameters: heterogeneous variances and heterogeneous covariance relative to local serial autocorrelation, similar to Toeplitz structure). The latter three covariance stru ctures typically provide good fit to repeated measures data (Kenward 1987, Littell et al. 2000). Each c ovariance model was compared to that of

PAGE 116

116 unstructured covariance (each covariance parame ter estimated independently) using Akaike’s information criterion (AIC) and likelihood-ratio tests with degrees of freedom equal to the difference in number of estimated parameters (Lit tell et al. 1996, Littell et al. 2000). I selected the most parsimonious covariance model with least deviation from the unstructured covariance model in terms of parameter estimates and structure. Finally, I used generalized l east squares methods to test fixed effects on each of six response variables. The first response variable was small woody stems per 1000-m2 (all stems > 2 cm but < 15 cm dbh). Next, I tested fixed effects on species richness, including counts by lifeform group and at different sample scales (1-m2 and 1000-m2). I separately analyzed 1) total number of species, 2) number of graminoid species which includes all true grasses (of the family Poaceae ) and morphologically similar species of the families Cyperaceae and Juncaceae 3) number of forb species, which include all non-g raminoid herbaceous specie s, and 4) number of woody species, including all non-herbaceous trees a nd shrubs. Significance tests were evaluated using a conservative Type I error rate (p < 0.015) to avoid error infl ation associated with multiple tests. A separate error rate of p < 0.05 was applied to both models of TOTAL species richness (1-m2 and 1000-m2). Most dependent variables were log transformed to improve normality of residuals in the mixed linear models. In each mixed linear model, the TRT*YEAR in teraction was of primary interest because this effect represents treatment effects over time. I assessed treatment ef fect in individual years by using the “SLICE” option of PROC MIXED using SAS software Version 8 (SAS 2000). Significance tests evaluate treatment e ffects for each year separately. Species data from two sample scales were us ed to assemble species response matrices. The 1000-m2 species matrix included average cover class estimates from each of four years

PAGE 117

117 (1997, 1998, 2000 and 2005; matrix dimensions = 36 plots x 217 species). The small scale species matrix included mean percenta ge cover values from the four 1-m2 small plots per site (45 plots x 121 species). Data from five years were included (1997-2000 and 2005). Species with fewer than two occurrences in each dataset were deleted, as they contribute nothing to calculations of inter-plot simila rities (McCune and Grace 2002). I applied the Hellinger distance transformation to the raw spec ies cover data. When used in conjunction with linear ordination methods, th is transformation offers a better compromise between linearity and resolution than do methods based on chi-squa re distances. This approach avoids problems inherent to sample weighting, in addition to problems associated with using Euclidean distances with untransformed data (Legendre and Gallagher 2001, Legendre et al. 2005). The 1-m2 species data were log transformed pr ior to Hellinger transformation, whereas the 1000-m2 data were not. I used Redundancy Analysis (RDA) as a method for direct gradient analysis of species compositional data. As a canonical ordination met hod, RDA directly relates species responses to environmental factors. Sample scores are “cons trained” as linear combinations of explanatory variables, conceptually similar to linear (or multiple) regression. Treatment factors were explanatory variables, identical to th ose used in univariate analyses. I used RDA models similar to those describe d by Leps and Smilauer (2003) for analysis of temporal compositional trends in a repeat ed measures experiment. I tested two null hypotheses per response matrix: 1) there are no directional temporal changes in species composition present in either or both restorati on treatments (within subject effect), and 2) temporal trends in composition ch anges are independent of treatment s (between subject effect). To test these, I varied constrai ning explanatory variables and covari ables in each of two RDA’s.

PAGE 118

118 In the first, YEAR effect and TRT*YEAR interac tion were specified as c onstraining variables, and Plot identity as a covariable. This corresponded to a model of YEAR effect only. The second RDA was constrained by the TRT*YEAR inte raction, and YEAR and Plot identity were specified as covariables. This model corresponds to a test of interacti on effects. Post-hoc “contrasts” of treatment effects between specif ic years were performed using the same RDA model as for hypothesis #2, with a species response data from th e years of interest. Data matrices were centered by species norms prior to ordinations. Scaling focused on inter-species correlations to favor biplot interp retation (Leps and Smilauer 2003). Significance of effects was tested with M onte Carlo permutation methods. Independence of species data relative to the explanatory (c onstraining) variables wa s tested (McCune and Grace 2002, Leps and Smilauer 2003). I used a restricted permutation configuration corresponding to a split-plot design where permutat ions of repeated measurements were confined within sample units (split-plots). Whole plots were permuted keeping within plot measurements intact. All ordinations and permutation tests were performed with C ANOCO (version 4.5) and CanoDraw (version 4.0) software (Braak and Smilau er 2002). Small sample size and restricted block design limited number of perm utation configurations. Thus, to reduce the probability of a Type II error, I selected a Type I error rate of p < 0.10 for omnibus tests of main and interaction effects and p < 0.05 for post-hoc contrasts. Specific species most correlated with interact ion effects (TRT*YEAR) were identified as those with highest “fit” to the first canonical axes in RDA ordinations. The fit value for individual species is the coe fficient of determination corresp onding to a regression of species responses on sample scores on the first (canonic al) axis. For each RDA, I selected the top 20-30 species with highest fit using the “lower axis mi nimum fit” inclusion option in CanoDraw (Braak

PAGE 119

119 and Smilauer 2002). Species vectors in biplots represent magnitude a nd direction of the first axis association. Unconstrained ordinations were applied to species response matrices for illustrative purposes. Solutions from principal components analysis (PCA: the unconstrained analogue to RDA; Leps and Smilauer 2003) are presented to di splay successional trends of sample units. Each PCA ordination was based on a cross-products matrix of inte r-species correlations derived from the Hellinger transformed data matrix. Co mpositional data were standardized by sample norm (Leps and Smilauer 2003). I present two di mensional ordination solutions and report the proportion of variance (in the species data) e xplained by ordination axes (McCune and Grace 2002). Comparisons of ACP data to Reference Data Species counts were transcribed from the Penfound data for comparison to ACP plots of similar areas. Species richness was derive d from non-overlapping subplots from the Penfound data (areas = 1, 5, 10, 15, 20, 25 and 30 m2). Similarly, I tallied species from two overlapping sample scales from the Lake Ramsey reference data (1-m2 and 10-m2 sample areas; 11 plots). Mean species numbers by ACP treatment were calculated for increasing sample areas: 1, 2, 10, 20, and 100-m2. Unlike the Penfound data, there was some overlap in ACP plot data due to the nested plots. I compared species area relationships between ACP restoration treatments in the final sample year (2005). An ANCOVA model tested tr eatment effects (logged+fire vs. fire-only) on species area relationships. Pr e-treatment species counts (from 1997) and log sample area were covariates in the model. Nine ty-five percent confidence interval s were calculated for the two species area curves by treatment type. A species area curve de rived from the Penfound data is

PAGE 120

120 displayed along with restoration treatment confid ence intervals (statistical comparison is not possible). Similarly, I visually compare species ri chness of Lake Ramsey data at two scales (1m2 and 10-m2). I assembled species matrices from the Penfound and Lake Ramsey data in a manner compatible with ACP taxonomy and sample scales. First, a presence-absence matrix of ACP and Penfound species data was assembled from 30 m2 sample areas (38 samples x 163 species). This matrix contained data from four ACP sample years plus one sample from each of two Penfound habitats sampled in late summ er 1936 (the “cut-over pineland” and “pine-cypress community” Penfound sites). Both habitats were descri bed by the authors as open, herb-dominated communities that burned frequently. Second, a si milar presence-absence matrix was constructed from ACP and Lake Ramsey species data from 10-m2 sample areas. For this I used two 10-m2 sample areas per ACP (over 4 sample years; 72 to tal ACP samples), plus data from eleven 10-m2 Lake Ramsey plots sampled in October 1999. I used unconstrained ordination to display tem poral trends of ACP samples relative to the Penfound and Lake Ramsey reference data. I applied non-metric multidimensional scaling (NMS) ordination with an inter-sample distance matrix of Bray -Curtis coefficients derived from the presence-absence species matrices. This method displays geographically disparate data without constraints of explanatory factors (McCune and Grace 2002). Successional vectors depict temporal trends in compositi on of the two restoration treatments relative to reference data. Results Trends in Species Richness and Woody Stems Following logging, abundance of small woody stems (< 10 cm dbh) declined precipitously (nearly 86%, from 192.4 67.9 stems / 0.1 ha in 1997 to 27.2 9.7 stems in 1998)

PAGE 121

121 and remained low in the logged+fire treatment throughout the study (Figure 4-2). Declines in fire-only small stem counts occurred followi ng the 2000 prescribed burn (dropping from a high of 349.2 151.1 to 49.7 15.1 per 0.1 ha). Repeated measures ANOVA of log transformed small stem counts indicated significant YEAR an d TRT*YEAR effects (Tab le 4-1). Treatment differences are significant in the first two post-loggi ng years (1998 and 1999) but disappear after the first prescribed fire in 2000 (Figure 4-2). Post-logging species richness increased in the logged+fire plots relative to the fire-only plots (Figure 4-2). The main effect of YEAR was significant in an ANOVA of species richness per 1000-m2 (Table 4-1). The TRT*YEAR interac tion was significant (p = 0.047), indicating different temporal changes in sp ecies richness by treatment. Tr eatment differences were greatest in 1999 (two years post-logging), then diminish ed after prescribed burning in 2000 and 2005. Similar trends were not apparent in the 1-m2 sample data (Table 4-1). Changes in numbers of grami noid species are responsible fo r treatment differences in species richness. Individual ANOV A models showed significant tem poral effects in graminoids species richness (per 1000-m2) only. The YEAR and TRT*YEAR effects were significant (Table 4-1). Similar to overall trends, trea tment differences in graminoid richness were significant only in the second year post-logging (1999) and dissipated following prescribed fire (Figure 4-3). Similar trends were not a pparent in forb and woody species richness. Trends in Species Composition Initial changes in species compositional were pronounced in the logged+fire relative to the fire-only treatment. Succe ssional trajectories from 1997 to 1998 are greater in magnitude and more uniform in direction for the logged+fire tr eatment, compared to the fire-only treatment. Trends are more pronounced at the 1000-m2 than the 1-m2 scale (Figure 4-4; PCA of 1-m2 data

PAGE 122

122 not shown). The 1000-m2 PCA explained 36.3 percent of speci es variation in the first two dimensions (first four eigenvalues = 0.22, 0.13, 0.10, 0.08). Permutation tests of treatment and temporal effects support successional trends obser ved in PCA ordinations. The RDA of the preand post-logging 1000-m2 species data (1997 vs. 1998), cons trained by TRT*YEAR interaction effect, indicated differential species responses between treatments (Tab le 4-2; Figure 4-5). Similarly, this contrast was significant in constrained ordination of the 1-m2 species data (Table 4-2). The magnitude and direction of species com position shifts became increasingly similar between treatments over time. Successional tr ajectories between 1997 and 2005, represented by the PCA of 1000-m2 species data, are similar between l ogged+fire and fire-only treatments (Figure 4-4). Permutation tests of YEAR a nd TRT*YEAR effects in RDA ordinations support the observed pattern in succession. The constrained ordination of 1000-m2 species data from all years revealed significant YEAR effect, and TR T*YEAR interaction effects were marginally significant (p = 0.07; Table 4-2 a nd Figure 4-6). These effects we re similarly significant in constrained ordinations of the 1-m2 species data. Contrasts betw een first and last study years only (1997 vs. 2005) showed no TRT*YEAR in teraction effect in either the 1-m2 or 1000-m2 species data (Table 4-2). As illustrated by the PCA of species data, successional trends from 1997 to 2005 were similar between treatments. In itial post-logging differences appear to have diminished at the end of the study. Initial response of herba ceous species to logging was pronounced. Logging triggered increases in presence and a bundance of many graminoid species and annual herbs. These species were identified as those with highest correlations (of abundance data) with the first constrained axis of the RDA of 1997 and 1998 da ta only. The first axis was constrained by

PAGE 123

123 TRT*YEAR interaction (Figure 4-6; see Appe ndix A for species code legend). In RDA solutions of 1-m2 and 1000-m2 species data, there were more grasses, sedges, and forbs associated with the logged+fire treatment plots. Eight logging responders are sedges (member of the family Cyperaceae ) and most of these are in the genus Rhynchospora In addition, depending on observation scale, th ere are 6 or 10 forbs and seve ral grass species that responded to logging. Many species that initially responded to loggi ng were annuals, such as Scleria muhlenbergia Rhynchospora chapmanii Eupatorium cappillifolium Bidens mitis Bartonia paniculata Drosera brevifolia and Diodia teres The response of an annual grass species, Panicum verrecosum was particularly pronounced at both sa mple scales. A few sub-dominant perennial grasses responded quick ly to woody removal, including Panicum rigidulum Paspalum floridanum and Anthaenantia rufa In contrast, the few species associated with fire-only treatment in the first post-logging year are mainly shrubs and vines. Compositional differences between treatments persisted over most of the study period. However, species associated with the logged+fire treatment over the study duration differed from the initial responders. None of the longer-term responders we re annuals. Most long-term species associated with the logge d+fire treatment were grasses, forbs, and a few sedges (at the 1000-m2 scale: Figure 4-7). None were woody species. At the 1-m2 scale, abundances of bluestem grasses ( Andropogon virginicus and A. cappilipes ) increased in response to the logged+fire treatment (Figur e 4-7). Similarly, perenni al sedges of the genus Rhynchospora ( R. elliotii R. cephalanthus R. oligantha and R. gracilis ) increased in presence and abundance in the logged+fire treatment. The latter species was the dominant non-gra ss monocot of ACP pine savannas. Other post-logging responders include d small statured perennial forbs, most having over-wintering rosettes and member of the families Asteraceae and Xyridaceae Few species

PAGE 124

124 were associated with the fireonly treatment over the study period at either sample scale, and these were mainly shrubs, vines, and forbs. ACP Treatment Responses vs. Reference Conditions Species-area relationships differe d between restorati on treatments at th e end of the study period (Figure 4-8). An ANCOVA of the 2005 sa mple data revealed a significant treatment effect with initial BA included as a covariable in the model (TRT: F1,40 = 6.1, p = 0.018; initial BA: F1,40 = 132.9, p < 0.0001). Area was significant (F1,40 = 47.39, p < 0.0001) but the TRT*AREA interaction was not (F1,40 = 0.49, p = 0.487). Species richness of the logged+fire ACP plot s (in 2005) exceeded that of Lake Ramsey plots at the 10-m2 sample area, although they were similar at the 1-m2 scale. At the 10-m2 scale, mean and standard error of Lake Ramsey speci es counts falls below the 95 percent confidence interval of the ACP logged+fire treatment (Figur e 4-8). In contrast, sp ecies richness of the Penfound data exceeds that of bot h ACP treatment sites at areas 10-m2, exceeding ACP 95% confidence intervals. I was unable to formally test differences in species richness between Penfound and ACP data due to lack of replic ation and differences in sampling methods (overlapping vs. non-overlapping pl ot layouts). However, the sp ecies-area pattern of the ACP logged+fire treatment suggests recovery of ground cover richness approaching that of my historic reference site. Both of the ACP restoration treatments prompted species composition changes that resembled reference site conditions. Compositio nal shifts were similar in direction but apparently differed in magnitude between treatm ents. The largest temporal changes were in plots with higher initial BA (Figures 3-8 and 3-9). The NM S ordination in Figure 4-8 compares presence-absence species data from Lake Ramsey and ACP successional vectors (10-m2 plot

PAGE 125

125 size). Most ACP trajectories i ndicate directional shifts toward the reference composition. Similar patterns were apparent in the NMS ordination of Penfound reference data plus ACP successional vectors (Figure 3-9). Compositional sh ifts of the high initial BA plots appear most pronounced along the first NMS axis, toward the Pe nfound data points. Smaller shifts of the “Low initial BA” plots ar e directed toward the “pine-cypress” Penfound da ta point specifically (Figure 3-9). In general, AC P treatments promoted compositional changes toward reference site composition. Discussion The current study demonstrates dramatic reco very of an ecologically degraded pineland plant community following restoration of natural forest structure and ecosystem processes. These results underscore the hist orical importance of forest st ructure and fire regime in maintaining this natural ecosystem, and the im portance of timber management and prescribed burning for restoration of similarly degraded pine lands. The recovery of understory herbaceous vegetation was rapid, and resemble d the quality of reference site s. Changes in ground cover species composition were pronounced following re duction of midstory shrubby vegetation in both ACP treatments, as the open aspect of pr e-settlement conditions was restored. Although direction of change appeared sim ilar between logging treatments, ra tes of change appeared to be accelerated by timber removal. In this case, fire appeared to ultimately have a greater effect on ground cover species composition than mechanical tree harvest. Foll owing two prescribed bur ns, trends in species richness and composition suggest convergence be tween mechanical restoration treatments, despite persistent differences in canopy dens ities. The mechanical activity of logging immediately reduced shrubby biomass, which pr ompted a flush of herbaceous growth and

PAGE 126

126 diversity, which was subsequently sustained by prescribed fire. However, the initial tree basal area of fire-only sites was less th an that of the logged+fire treat ment. Within the range of my overstory tree basal area among tr eatments following logging (1-10-m2/ha), competitive interaction between canopy and understory vegetati on was probably minimal compared to that of midstory and understory interactions. Thus, it should be noted that the positive restorative effects of fire might be limited at higher levels of tree basal area. In the logged+fire treatment, fire invoked woody biomass decline, subsequently resembling that of the fire-only treatment. The trend toward similarity in woody biomass roughly coincided with convergence in herbaceous plant richness and composition between treatments. Fire mediated effects are most pronounced in the ground layer. Streng et al. (1993) suggested that frequent fire prom otes establishment of rarer speci es in an environment dominated by long-lived perennials by freeing up space and resources available for colonization, and by reducing competition from dominant grasses an d woody species. Fire likely decreases competition between understory woody and herbaceous plants for light and space (Platt et al. 1988a, Streng et al. 1993, Glitzenstein et al. 2003, Walker and Silletti 2006). Other restoration studies report herbaceous vegetation recovery in response to fire plus hardwood reduction that exceeded that expected from the chemical a nd mechanical treatments alone (Brockway and Outcult 2000, Provencher et al. 2001). In the la tter, fire alone prompted greater herbaceous ground cover response than mechanical hardwood reduction in Florida longleaf pine sandhill restoration (Provencher et al. 2001). Similar to my results fire effects on ground cover vegetation extend beyond reduction of canopy dens ity (Platt et al. 1988a, Robbins and Myers 1992, Waldrop et al. 1992, Streng et al 1993, Provencher et al. 2001).

PAGE 127

127 The restoration treatments preferentially prompted res ponses of species that are characteristic of pine savanna natural areas. Al l species that responded to restoration treatments were native, and with few exceptions, were not ruderal generalists. Species that initially responded to logging were characteristic savann a herbaceous species. The flush of “new” species included many grasses (plant family Poaceae ) and sedges (family Cyperaceae ), particularly small statured, rh izomotous species of the genus Rhynchospora. Longer-term species responders were primarily sedges and pere nnial forbs. Plant populations likely persisted at ACP during the period of fire suppression, either in the seed bank or in dormant vegetative states. The initial flush of rhizomotous sedges and clonal grasses suggest s long term persistence in vegetative states followed by rapid growth in response to increased light and space. These dormant lifeforms may not have been detected in pre-treatment sampling. Increases in species richness and abundance have been noted in ot her studies of woody removal by mechanical and chemical means (Greenberg et al. 1995, Ha rrington and Edwards 1999, Brockway and Outcalt 2000, Provencher et al. 2000, Provenche r et al. 2001). In these stud ies, increased richness and cover were attributed to soil disturbance as a direct effect of mechanical manipulations (Greenberg et al. 1995, Harrington and Edwards 1999, Cox et al. 2004), in addition to indirect effects of increased light, mo isture, and space availability (Harrington and Edwards 1999, Brockway and Outcult 2000, Provencher et al. 2000, Provencher et al. 2001). Most studies indicate eventual increases in na tive species typical of the focal ha bitat, rather than increases in exotic or ruderal species (but see Greenbe rg et al. 1995, Harrington and Edwards 1999). The absence of ruderal additions to the lo cal species pool affirms predictions that ecologically sensitive logging would not cause n ovel trends in post-logging succession. The number and abundance of non-ruderal annual spec ies increased in res ponse to logging, but

PAGE 128

128 declined after the first year. Similarly, restor ation of Florida xeric pi nelands via mechanical methods prompted an initial insurgence of native ruderal species (Gre enberg et al. 1995, Provencher et al. 2000), although th e short duration of these studies precluded eventual detection of decline. Similar to my findings, these stud ies documented no invasions of non-native species that invoked novel succession. Response of pine savanna herbaceous vege tation to woody reduction apparently differs by life form type across moisture conditions. Restoration of xeric pineland communities of the Gulf and Atlantic Coastal Plain regions (via mechanical, chemical, and prescribed burning treatments) prompted greatest increases in fo rb richness and abundance (Brockway and Outcult 2000, Provencher et al. 2000, Provenc her et al. 2001, Platt et al. 2006), whereas prescribed burning of wet-mesic Atlantic co astal pine savannas simulated in creases in grasses and sedges (Walker and Peet 1983, Glitzenstein et al. 2003). Although I documented increases in all herbaceous life form types following restoration treatments, grass and sedge increases were most pronounced in my mesic to wet pine savanna site. These observati ons suggest similar restoration treatments in different moisture conditions may trigger different compositional responses, relative to the composition of th e residual species pool and the di fferential loss of species groups (Walker and Silletti 2006). ACP Restoration Compared to Reference Model Ground cover vegetation of ACP ultimately resembled that of my reference sites following restoration of stand structure. Succes sional endpoints of ACP treatments were similar to both the historic (Penfound) and contemporary (Lake Ramsey) reference data. Congruence may be attributable to shifts in composition (c onstituent species and rela tive abundances) rather than changes in species richness (species presence) Species richness incr eases were sustained in

PAGE 129

129 the logged+fire treatment only, and these numbers approached those of th e historic (Penfound) reference condition. Similarly, species richness of fire suppressed Florida longleaf pine sandhills increased following mechanical hardwood reduction coupled with fire, approaching and in some cases exceeding that of the refere nce site (Provencher et al. 2001). The authors credit fire as the dominant cause of increases in sp ecies richness and densities. Succession in response to restoration of stand conditions and ecosystem function depended largely on starting conditions. Compositional shifts were qualitatively similar between treatments and over varying initia l of canopy and midstory densities However, the magnitude of ground cover response depended on initial canopy and midstory densities, which ranged considerably (approximately 10 – 22 m2/ha basal area and 24 – 756 stems (< 10 cm dbh) per 0.1 ha). Restoration responses relative to star ting condition have been observed by others (White and Walker 1997, Walker and Sillet ti 2006), with densely wooded sites having largest vegetation responses to woody biomass reductions. Although studies of pineland restoration doc ument significant changes in ground cover vegetation (Greenberg et al. 1995, Harringt on and Edwards 1999, Brockway and Outcult 2000, Provencher et al. 2000, Provencher et al. 2001, Walker and Silletti 2006), most do not include explicit and quantitative comparisons to referenc e models. Changes in species composition or richness alone do not necessarily indicate restoration success, if the goal is to mimic some historical or contemporary “nat ural” condition. It is possible to trigger novel or unintended succession that may be difficult to detect withou t reference comparisons (Fule et al. 1997, White and Walker 1997). Comparison of treatment resp onses to proximate and quantitative reference data allowed me to assess progr ess and conclude that efforts ha ve promoted vegetation recovery

PAGE 130

130 approximating that of desired conditions. Furt hermore, undesirable non-native or weedy species were not introduced nor was unde sirable succession invoked. Management and Conservation Implications Success of pineland restoration depends in large part on starting conditions and ecological resiliency of the treatme nt site. To date, few studies in the Southeastern Coastal Plain indicate the potential of pl ant community recovery, in te rms of species composition and diversity, resembling a desired condition (Harrington and Edwards 1999, Brockway and Outcalt 2000, Hedman et al. 2000, Provencher et al. 2001, Glitzenstein et al. 2003, Platt et al. 2006). These studies indicate ground c over vegetation recovery follow ing reduction of woody biomass via various means, including mechanical and chemical methods, and prescribed burning (Harrington and Edwards 1999, Brockway and Ou tcult 2000, Provencher et al. 2001). It is important to examine ecological starting conditions represented in these restoration studies, and compare these to the current study. Where availa ble, approximate mean starting (or control) densities of mechanically or chemically treate d sites (in longitudinal or retrospective studies) were as follows: 10.4 m2/ha basal area (Harrington and Edwards 1999), ~124 (oak) stems/0.1 ha (Provencher et al. 2001), and ~14-18 m2/ha basal area and ~53-130 stems/0.1 ha (Hedman et al. 2000). Canopy and midstory densities of the af orementioned studies appear to approximate densities of typical second grow th pine stands not subjected to industrial plantation management (basal areas < 20 m2/ha; see Robertson and Ostertag 2007) By comparison, my starting basal areas of roughly 10-22 m2/ha were similar to those of Harrington and Edwards (1999) and Hedman et al. (2000). However, initial stem de nsities of my treatment sites were high in comparison to those reported, and my initial vari ance was also large (overall mean and standard error is 241.4 81.7; range 47-781 stems/0.1 ha). My higher stem density mean and variance

PAGE 131

131 may be attributable to differences in moisture conditions and community t ype. Other restoration study sites were upland woodlands or dry upland sandhills, wher eas mine contained mesic and wet pine savanna communities. These results contribute to an overall model of restoration poten tial for Southeastern Coastal Plain pinelands. Similar to other studies mine indicates that ground cover recovery is possible on degraded pineland site that has suffered fire suppression and/or fire regime alteration over a period of 10-20 years, and the associated dense woody growth. Recovery is possible in the absence of previous extens ive soil disturbance associated with past agricultural or silvicultural land uses. Little is known about the resiliency of pine land ground cover following ground tilling, although Hedman et al. (1999) and Ostertag and Robertson (2006) found evidence of persistent changes in succession in second gr owth pinelands on fallow fields. Furthermore, the current study shows plant commun ity recovery is possible in a we tter site, with a species pool adapted to different moisture and edaphic cond itions than those previously examined. Although reintroduction of native fire regi me was arguably the most important management prescription for ecological restora tion, there were benefits to canopy removal via commercial logging at ACP. Over time, it is likel y that frequent fire alone would eventually reduce woody vegetation in fire suppressed pinela nds such as ACP. However, reduction of woody vegetation in this manner ma y require decades of frequent and intense burning (Waldrop et al. 1992, Glitzenstein et al. 1995, Olson and Platt 1995, Drewa et al. 2002). This may exceed the time in which remnant populations of ground cover species are available for re-colonization (Hedman et al. 2000, Walker and Silletti 2006). Fo r this reason, expediting restoration of stand structure using mechanical methods may enhance ecological restoration. In addition, financial

PAGE 132

132 returns from commercial timber sale may offset restoration costs without introducing detrimental effects. These results, along with other restorati on studies, demonstrate that ground cover vegetation recovery is possible in degraded pi ne savannas over a range of starting conditions, without resorting to artif icial species reintroductions. Furtherm ore, results highlight the innate resiliency of pineland groundcover plant communities. Life history adaptati ons of plant species for dormancy may buffer populations in periods of atypical environmental conditions (i.e. fire suppression). Temporal rebounds of plant comm unity composition suggest some degree of successional “stability” within a range of fire regime and stand structure variability. However, succession in former pinelands that have been severely altered (ground tilling, agricultural land uses, and fire suppression exceed ing several decades) may exceed that range of resiliency and require more inte nsive restoration treatments to achieve reference conditions (Walker and Silletti 2006). Although re-seeding efforts were successful in former longleaf pine sandhills (Seaman 1998, Cox et al. 2004), it is costly which limits its application to small areas (Walker and Silletti 2006). Because of this, treatme nts of lower intensity and cost are desirable, and appropriate for ecological re storation of large areas (Prove ncher et al. 2001). Fortunately, restoration of Southeastern Coastal Plain pinela nd stands with mechanical woody reduction can be labor and cost efficient, and yield favorable results if coupled with appropriate fire management.

PAGE 133

133 Table 4-1. ANOVA tables for models of species ri chness and stem number s. Main, interaction, and covariable effects listed in “Effect” column. “Slice” e ffects (treatment effects in individual YEARS) indicated for first thr ee models. The covariance structure and number of parameters selected for each m odel are listed. P-values less than critical values are shown in bold type. ModelCov structureEffectNum dfDen dfF-valuep-value Number stemsHetero toeplitz Treatment17.013.570.101 (1000 m2) (9 cov parameters)Year47.8910.54 0.003 Trt*Year47.8913.88 0.001 1997: Trt effect18.310.010.923 1998: Trt effect17.879.37 0.016 1999: Trt effect17.137.51 0.02 2000: Trt effect17.451.970.2 2005: Trt effect17.331.610.243 Total species Antedependence Treatment17.12.980.13 (1000 m2) (9 cov parameters)Year49.54.81 0.022 Trt*Year49.53.64 0.047 1997: Trt effect170.230.644 1998: Trt effect173.890.089 1999: Trt effect178.96 0.02 2000: Trt effect176.180.042 2005: Trt effect174.590.069 Gram speciesAutoregressiveTreatment17.952.560.149 (1000 m2) (2 cov parameters)Year427.33.79 0.014 Trt*Year427.37.82 0.001 1997: Trt effect1164.730.045 1998: Trt effect1163.820.068 1999: Trt effect1168.81 0.009 2000: Trt effect1163.990.063 2005: Trt effect1161.990.178 Forb speciesAntedependenceTreatment17.030.390.551 (1000 m2) (9 cov parameters)Year48.712.230.148 Trt*Year48.711.40.311 Woody species AutoregressiveTreatment17.220.620.458 (1000 m2)(2 cov parameters)Year427.83.08 0.032 Trt*Year427.81.990.124 Total speciesAutoregressiveInitial BA15.731.19 0.002 (1 m2) (2 cov parameters)Treatment15.846.7 0.042 Year426.73.42 0.022 Trt*Year426.71.80.159

PAGE 134

Table 4-2. Results of Monte Carlo permutati on tests from RDA constrained ordinations. Specific datasets subjected to ordinatio ns and permutation tests are listed in left column (1000 and 1-m2 species data), along with data matrix dimensions. The null hypothesis tested is listed in the Model column. Significant p-values high lighted in bold text. DatasetMatrix dimension Model Sum all eigenvalues First canonical eigenvalue F-ratiop-value 1000 m2 4 years 36 plots x 217 spp No YEAR effect0.3690.057 4.62 0.05 No YEAR*TRT effect0.3410.022 1.71 0.07 1000 m2 '97 vs. '98 18 plots x 190 spp No YEAR*TRT effect0.1410.044 3.2 0.02 1000 m2 '97 vs. '05 18 plots x 193 spp No YEAR*TRT effect0.2040.037 1.530.17 1 m2 5 years 45 plots x 121 spp No YEAR effect0.3080.034 4.22 0.01 No YEAR*TRT effect0.2790.016 2.09 0.02 1 m2 '97 vs. '98 18 plots x 98 spp No YEAR*TRT effect0.1230.026 1.85 0.03 1 m2 '97 vs. '99 18 plots x 96 spp No YEAR*TRT effect0.1380.065 1.620.05 1 m2 '97 vs. '05 18 plots x 99 spp No YEAR*TRT effect0.2090.032 1.260.17 134

PAGE 135

Figure 4-1. ACP pictures: (a) pre-treatment in 1997, (b) immediately after logging in 1998, and (c) af ter logging and first prescribed fire in 2000. (a) (b) (c) 135

PAGE 136

136 Figure 4-2. Least squares means (LS m eans) and standard errors of the number of small stems (< 15 cm dbh) per 1000-m2 sample. Closed circles = fire-only plot s, open circles = logged+fire plots. Arrows indicate timing of specific restoration treatments. Logging 1st Fire2nd Fire

PAGE 137

137 Figure 4-3. Least square means (LS mean s) and standard errors of species richness by treatment and year. Open circles = logged+fire treatment; closed circles = fire-o nly treatment. Top plot shows total species richness 1000-m2 sample area; bottom plot shows means of graminoid species only. Timing of treatments indicated by arrows. Logging 1st Fire 2nd Fire

PAGE 138

138 Figure 4-4. PCA ordination of ACP species data (1000-m2 scale); first two ordination axes displayed. Dots indicate 1997 pre-treatment compositional data in ordination space. Successional trajectories correspond to compositional shifts of data from individual plots: red = logged+fire treatment, black = fire-only treatment. Top plot shows changes between pre-treatment and first post-logged years (1997 vs. 1998). Bottom pl ot shows shifts from pretreatment (1997) and after l ogging and fire (2005). -1.0-0.50.00.51.0-1.0-0.5 0.0 0.5 1.0 -1.0-0.5 0.0 0.5 1.0 P CA A x i s 2PCAAxis 11997vs. 1998 1997vs. 2005

PAGE 139

139 Figure 4-5. Constrained RDA or dinations of pre-logged ( 1997) and first post-year (1998) species data. First ca nonical axes are constrained by TRT*YEAR interaction. Species highly correlated with first canonical axes are displayed by vectors and codes (see Appendix C for species names). Top plot displays RDA of 1000-m2 species data; bottom plot shows 1-m2 species data. Note diffe rent scales of Axis 2. 0.2-0.1 0.0 0.1 0.2 0.3 0.4 0.5 ANTRU BIDMI ELETU EUPCA EUPLE FUIBR LOBBRNYSBIPANRI PANSO POLRA RHUVE RHYCN RHYFI SCLMU SCLPA TOXRA TRADI XYRAM AXOFI ERYIN PANVE SCLPP DROBR PLURO ANDVI RHYGL RHYPU XYRLO Fire only Logged+fire-1.0-0.50.00.51.0 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 ANDMO BARPA BIGNU CACOV CARPS COETE DROBR ELETU FUIBR GAYMO GELRA LOBBR NYSBI PANTE PANVE PERBO PROPE RHYCE RHYCH RHYCN RHYFI RHYOL SCLPA SCUIN TOXRA VIOPR VITROLogged+fire Fire onlyPASFLRDA Axis 1R D A A x i s 2R D A A x i s 2

PAGE 140

140 Figure 4-6. Constrained R DA ordinations of ACP speci es data including 1997 pretreatment data. Top plot = 1000-m2 data (excluding 1999 data), bottom plot = 1-m2 species data (all years) First canonical axes are constrained by TRT*YEAR interacti on indicated by treatment label in red text. Species most highly corre lated with first canonical axes are displayed by red vectors and code s (see Appendix C). Vector length and angles indicate strength and direction of correlation. -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 AGASP ANDCA ANDGL ANDVI ARIPA ASCLO CARPS COETE EUTTE HYPBR ILEDE LUDGL LUDSP PANET PANRI PINEL PLURO POLRA RHEPE RHOSP RHUVE RHYEL SABSP SCLPA SISAL SOLRU -1.0-0.50.00.51.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 ANDCA ANDVI ANTRU ASTDU CENER CLEAL COERU CRASP EUPSE HYPBR HYPCI HYPHI HYPST ILEVO PANEN POLRA PROPE RHYCE RHYCH RHYEL RHYGR RHYOL RUBUS SABSP SMILA SMIRO SPHAG STYAM XYRAMRDAAxis 1R D A A x i s 2Logged+fire Logged+fire Fire-only Fire-only

PAGE 141

141 Figure 4-7. Number species per log sample area (m2): mean species counts from ACP treatments vs. Penfound and Lake Ramsey species richness. Shaded regions show 95% confiden ce intervals of ACP treatment means from 2005 samples; areas = 1, 2, 10, 20, and 100-m2. Light shading = logged+fire; dark shadi ng = fire-only. Dark line denotes species numbers from Penfound “cut-over pineland” at areas = 1, 5, 10, 15, 20, and 25 m2. Red triangles show mean s and standard errors of Lake Ramsey species richness (n = 11) at areas = 1 and 10-m2. 0 10 20 30 40 50 60 70 80 1 100 10 Log (Area)N umb e r s p e c i e s

PAGE 142

142 Figure 4-8. Successional trends of ACP sp ecies data compared to Penfound and Lake Ramsey reference data (NMS ordination of presence-absence data of comparable sample sizes – see text). Red vectors depict compositional shifts in the logged+ fire treatment between 1997 and 2005; black vectors depicts shifts in fire-only treatment. Blue symbols show relative positions of reference data in ordination space. Bold dashed lines separate ACP samples with High and Low initial BA. HighBA LowBAACPvs. LakeRamsey HighBA LowBA Pine-Cypress Cutover LLACPvs.Penfound

PAGE 143

143 CHAPTER 5 CONCLUSION My dissertation presents a vegetation classifi cation of pineland communities in Florida, a model of the relationships betw een species composition and physic al and spatial factors, and description of temporal variati on of pineland species composition in response to restoration of fire regime and forest structure. The research revealed predictable spatial patterns of species composition at both local and regional scales, an d suggests resiliency of community composition to temporary alterations to fire regime and timb er density. Furthermore, the study of ecological restoration suggests the resiliency of pineland vegetation, in that recovery approximating natural conditions followed re-introduction of native fire regimes after a l ong period of fire suppression. I presented a comprehensive vegetation classifi cation based on floristic similarity, using K-means cluster analysis and ordination methods. I recogniz ed three ecological series corresponding to idealized moisture conditions. These were furthe r divided into 16 associations. The series included Dry Uplands (6 associations), Mesic Flatwoods (3 associations), and Wetlands (7 associations). Summary information described each community association relative to species diversity, woody plant structure, diagnostic species, and environmental and physiographic features. Floristic variation varied greatly w ith geographic segregation and edaphic characteristics, particularly between th e panhandle and peninusula regions of Florida. Distinctions between community associations were related to the prominence of species with restricted distributions, and to a lesser degree, endemic species The floristic classification presented here is comprehensive but applicable in the field, co mpared to other classification based on regional flora. The spatially explicit model of environmenta l-composition relationshi p revealed many of the same patterns seen in the floristic classifi cation. Most notably, th e effects of geographic

PAGE 144

144 separation were prominent in community structure, particularly the floristic distinctions between the panhandle and peninsula. However, the e ffects of local environm ental factors dominated vegetation gradients, particularly variables re lated to topography and so il moisture. In the context of all environmental factors, both regi onal and local environmental effects appear to influence vegetation patterns. Effects of pure sp atial structure were also evident, although these made up a much smaller proportion of explained va riance than environmental effects. Variation associated with recent fire regime and timbe r stand structure may have contributed to unexplained variance. I interpre t the data as demonstrating a relatively str ong control of environmental factors on the distribution of pi neland species, with biot ic control mechanisms and historical biogeogra phy playing a lesser role. The study of ground cover vegetation recove ry following fire and stand structure restoration suggests the resiliency of pineland community composition to atypical environmental alterations. Temporal variation in pineland comm unity structure is poorly understood relative to spatial variation. However, longitudinal studies following restoration treatments shed light on succession that is within the range of “natural ” community stability. The study of ecological restoration following a relatively long period of suppression of na tive fire regime demonstrates that there is a certain degree of successional “stability” in pi neland community structure. Ecological restoration of a degraded Coasta l Plain pineland remnant was successful in terms of ground cover vegetation recovery that resembled reference site conditions. These results suggest that recovery is possible on sites that have suff ered moderate to severe fire suppression but minimal ground disturbance (fro m cultivation). Ground cover species richness was enhanced by overstory and midstory woody biomass reduction, mainly as increases in detectable graminoid species. Furthermore, species richness and composition of restoration

PAGE 145

145 treatments converged following two prescribed fires suggesting both the relative importance of fire in ecological restoration and minimal adverse effects from mechanical logging. Pine removal via carefully supervised mechanical logg ing does not appear to a dversely affect savanna vegetation recovery, and may expedite overall community restoration. In sum, results of this ob servational and experimental research are relevant to conservation and management of pyrogenic pinela nd communities of the Southeastern Coastal Plain. The classification of Florida pineland communities based on compositional similarities can be applied to inventory and restoration effo rts in the State and surrounding regions. In addition, the classification provi des useful reference conditions for restoration, a basis for quantifying regional variation in community variation, and specific diagnostic indicators for community identification. The model of envir onmental-composition relati onships contributes to the overall understanding of environmental determ inants of pineland vegetation, in addition to quantifying ranges of environmental conditions co rrelated to compositional variation. The final restoration study showed that pine savanna vegetati on is resilient to modera te degradation related to fire suppression, and can rebound following recons truction of native stand structure. These results add to the regional model of ecological restoration in the Sout heast, specifically by showing that restoration is achievable in wet pi neland sites with increased alteration from fire suppression.

PAGE 146

146 APPENDIX A LOCATIONS OF SAMPLE PLOTS AND SITES Table A-1: Vegetation sample plot s listed by code (“Plot”). Site code indicates Site containing plots. The assigned community associat ion (“Assoc”) is indicated by code (see Figure 2-3). Latitude and Longitude indica ted in decimal degrees. Some specific plot locations are not reported as per la ndowner request (“NR”). “Region” indicates ecoregion delineation used for site stratification (see text). “Management Area” indicates public land unit; private land s are noted. Abbreviations indicate management unit types: SF = State Forest AFB = Air Force Base, AFR = Air Force Range, SP = State Park, WMA = Wildlife Ma nagement Area, NF = National Forest, ARD = Apalachicola Ranger District, WRD = Wakulla Ranger District, NWFWMD = Northwest Florida Water Management District, SWFWMD = Southwest Florida Water Management District, DEP = Florid a State Department of Environmental Protection, NWR = National Wildlife Refuge CA = Conservation Area (State), SRA = State Recreation Area, TNC = Th e Nature Conservancy preserve. Region Management Area Site Plot Assoc Latitude Longitude Northwest Uplands Blackwater River SF BW01 FL092 D4 8639.023 3054.668 Northwest Uplands Blackwater River SF BW01 FL093 W6 8638.869 3054.435 Northwest Uplands Blackwater River SF BW01 FL094 D6 8638.835 3054.417 Northwest Uplands Blackwater River SF BW02 FL098 D5 8652.411 3043.187 Northwest Uplands Blackwater River SF BW03 FL099 D4 8656.497 3050.387 Northwest Uplands Blackwater River SF BW03 FL100 W6 8656.545 3050.399 Northwest Uplands Blackwater River SF BW03 FL275 W7 8656.577 3050.434 Northwest Uplands Blackwater River SF BW03 FL276 W5 8656.378 3050.381 Northwest Uplands Blackwater River SF BW01 FL284 D5 8639.132 3054.441 Northwest Uplands Blackwater River SF BW01 FL285 D4 8639.738 3054.574 W Panhandle Gulf Coast Eglin AFB EG08 FL058 M2 8625.442 3027.510 W Panhandle Gulf Coast Eglin AFB EG02 FL304 M2 8645.831 3025.537 W Panhandle Gulf Coast Eglin AFB EG02 FL041 D4 8645.827 3025.486 W Panhandle Gulf Coast Eglin AFB EG02 FL042 W6 8645.905 3025.583 W Panhandle Gulf Coast Eglin AFB EG02 FL046 D4 8646.216 3025.085 W Panhandle Gulf Coast Pt. Washington SF PT01 FL280 D4 8608.742 3020.642 W Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJ01 FL022 M2 8517.839 2942.753 W Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJ01 FL163 W6 8517.886 2942.757 W Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJB01 FL255 M1 8517.558 2942.183 W Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJ01 FL257 W6 8516.240 2942.836 W Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJ01 FL258 W6 8515.401 2943.102 W Panhandle Gulf Coast Topsail SP TP01 FL077 M2 8617.644 3022.437 W Panhandle Gulf Coast Topsail SP TP01 FL078 W6 8617.739 3022.383 W Panhandle Gulf Coast Topsail SP TP01 FL079 M2 8618.011 3022.380 W Panhandle Gulf Coast Topsail SP TP01 FL293 M2 8617.609 3022.307 West Panhandle Sandhills Eglin AFB EG01 FL040 D4 8643.830 3029.049 West Panhandle Sandhills Eglin AFB EG03 FL043 D4 8647.440 3027.046

PAGE 147

147 Table A-1 continued Region Management Area Site Plot Assoc Latitude Longitude West Panhandle Sandhills Eglin AFB EG03 FL044 D4 8646.915 3027.165 West Panhandle Sandhills Eglin AFB EG03 FL045 W4 8646.786 3027.083 West Panhandle Sandhills Eglin AFB EG04 FL047 D4 8613.336 3036.250 West Panhandle Sandhills Eglin AFB EG05 FL048 D4 8611.709 3037.040 West Panhandle Sandhills Eglin AFB EG05 FL049 D6 8611.971 3036.542 West Panhandle Sandhills Eglin AFB EG06 FL050 D4 8651.788 3030.834 West Panhandle Sandhills Eglin AFB EG07 FL051 D4 8646.726 3027.923 West Panhandle Sandhills Eglin AFB EG07 FL052 D4 8646.742 3028.417 West Panhandle Sandhills Eglin AFB EG07 FL053 D4 8646.616 3028.509 West Panhandle Sandhills Eglin AFB EG04 FL054 M2 8613.419 3036.190 West Panhandle Sandhills Eglin AFB EG04 FL055 D6 8612.760 3035.829 West Panhandle Sandhills Eglin AFB EG06 FL056 D4 8651.927 3031.046 West Panhandle Sandhills Eglin AFB EG07 FL057 D4 8646.748 3028.740 West Panhandle Sandhills Eglin AFB EG04 FL095 W6 8613.417 3036.402 West Panhandle Sandhills Eglin AFB EG05 FL096 W7 8611.918 3036.500 West Panhandle Sandhills Eglin AFB EG04 FL097 W7 8613.460 3036.581 Marianna Lowlands Apalachee WMA AP01 FL076 D5 8457.250 3047.156 Marianna Lowlands Apalachee WMA AP01 FL090 D5 8457.504 3048.523 Marianna Lowlands Apalachee WMA AP01 FL091 D5 8457.343 3048.404 Marianna Lowlands Falling Waters SP FW01 FL074 D6 8531.682 3043.656 Marianna Lowlands Falling Waters SP FW01 FL075 D6 8531.796 3043.587 Marianna Lowlands Falling Waters SP FW01 FL085 D5 8531.530 3043.739 Marianna Lowlands Rock Hill TNC RH01 FL060 D6 8529.065 3044.275 Marianna Lowlands Rock Hill TNC RH01 FL065 W6 8529.151 3044.134 Marianna Lowlands Rock Hill TNC RH01 FL089 D6 8529.375 3044.358 Marianna Lowlands Rock Hill TNC RH01 FL287 W5 8530.054 3044.665 Marianna Lowlands Rock Hill TNC RH01 FL288 W6 8530.190 3044.662 Marianna Lowlands Rock Hill TNC RH01 FL292 D5 8529.503 3044.369 Marianna Lowlands Three Rivers SP TR01 FL070 D5 8455.295 3044.163 Marianna Lowlands Three Rivers SP TR01 FL071 D5 8455.173 3044.162 East Panhandle Sandhills Apalachicola Bluffs TNC AB01 FL059 D4 8458.419 3027.440 East Panhandle Sandhills Apalachicola Bluffs TNC AB02 FL061 D4 8458.438 3028.383 East Panhandle Sandhills Apalachicola Bluffs TNC AB02 FL066 D4 8458.303 3028.538 East Panhandle Sandhills Apalachicola NF: WRD WD01 FL011 D2 8415.779 3021.186 East Panhandle Sandhills Apalachicola NF: WRD WD01 FL012 D2 8415.845 3021.163 East Panhandle Sandhills Apalachicola NF: WRD WD03 FL018 D2 8420.988 3019.288 East Panhandle Sandhills Apalachicola NF: WRD WD03 FL019 D2 8421.027 3019.361 East Panhandle Sandhills Apalachicola NF: WRD WD01 FL229 W1 8416.194 3021.513 East Panhandle Sandhills Apalachicola NF: WRD WD03 FL230 W1 8421.040 3019.371 East Panhandle Sandhills Apalachicola NF: WRD WD01 FL259 D2 8416.217 3021.525 East Panhandle Sandhills Apalachicola NF: WRD WD03 FL260 D2 8420.988 3019.333 East Panhandle Sandhills Econfina River NWFWMD ER01 FL269 D4 8532.023 3029.139

PAGE 148

148 Table A-1 continued Region Management Area Site Plot Assoc Latitude Longitude East Panhandle Sandhills Econfina River NWFWMD ER01 FL270 D4 8531.764 3028.899 East Panhandle Sandhills Econfina River NWFWMD ER01 FL294 W1 8534.151 3027.401 East Panhandle Sandhills Private land DY01 FL072 D4 8457.033 3026.435 East Panhandle Sandhills Private land DY01 FL073 D6 8457.070 3026.424 Apalachicola Lowlands Apalachicola NF: ARD AD01 FL062 W6 8500.932 3003.534 Apalachicola Lowlands Apalachicola NF: ARD AD02 FL063 D4 8459.533 3016.618 Apalachicola Lowlands Apalachicola NF: ARD AD02 FL064 D6 8459.107 3016.757 Apalachicola Lowlands Apalachicola NF: ARD AD01 FL067 D6 8500.818 3003.595 Apalachicola Lowlands Apalachicola NF: ARD AD03 FL068 D4 8458.989 3012.292 Apalachicola Lowlands Apalachicola NF: ARD AD03 FL069 D6 8459.224 3012.360 Apalachicola Lowlands Apalachicola NF: ARD AD04 FL080 D6 8505.324 3006.344 Apalachicola Lowlands Apalachicola NF: ARD AD04 FL081 D6 8505.157 3006.033 Apalachicola Lowlands Apalachicola NF: ARD AD03 FL082 W6 8459.254 3012.255 Apalachicola Lowlands Apalachicola NF: ARD AD02 FL083 D6 8458.749 3016.388 Apalachicola Lowlands Apalachicola NF: ARD AD05 FL084 D6 8501.290 3011.725 Apalachicola Lowlands Apalachicola NF: ARD AD05 FL086 W6 8501.305 3011.662 Apalachicola Lowlands Apalachicola NF: ARD AD05 FL087 D6 8500.861 3011.980 Apalachicola Lowlands Apalachicola NF: ARD AD01 FL088 W6 8500.194 3003.788 Apalachicola Lowlands Apalachicola NF: ARD AP06 FL162 W7 8457.638 3002.444 Apalachicola Lowlands Apalachicola NF: ARD AP06 FL164 M2 8457.496 3002.541 Apalachicola Lowlands Apalachicola NF: ARD AP06 FL265 D6 8458.099 3002.054 Apalachicola Lowlands Apalachicola NF: ARD AD01 FL266 D6 8500.904 3003.299 Tallahassee Red Hills Pebble Hill (Private land) PH01 FL227 W1 8405.221 3045.918 Tallahassee Red Hills Pebble Hill (Private land) PH01 FL228 W5 8405.213 3045.904 Tallahassee Red Hills Pebble Hill (Private land) PH01 FL289 D5 8405.342 3045.849 Tallahassee Red Hills Pebble Hill (Private land) PH01 FL290 D5 8405.508 3046.390 Tallahassee Red Hills Private land AV01 FL028 D2 NR NR Tallahassee Red Hills Private land AV01 FL029 W5 NR NR Tallahassee Red Hills Private land BE01 FL031 W5 NR NR Tallahassee Red Hills Private land BE02 FL037 D5 NR NR Tallahassee Red Hills Private land BE02 FL038 W5 NR NR Tallahassee Red Hills Private land BE01 FL039 W4 NR NR Tallahassee Red Hills Torreya SP TY01 FL030 D4 8457.027 3033.462 Tallahassee Red Hills Torreya SP TY01 FL036 D4 8457.129 3033.376 Tallahassee Red Hills Wade Tract (Private land) WT01 FL224 D5 8359.851 3045.724 Tallahassee Red Hills Wade Tract (Private land) WT01 FL226 W5 8359.742 3045.752 Tallahassee Red Hills Wade Tract (Private land) WT01 FL286 D5 8359.978 3045.688 Tallahassee Red Hills Wade Tract (Private land) WT01 FL291 W1 8400.672 3045.552 Wakulla Lowlands Apalachicola NF: WRD WD02 FL017 D4 8429.542 3020.029 Wakulla Lowlands Apalachicola NF: WRD WD04 FL020 D4 8441.153 3015.381 Wakulla Lowlands Apalachicola NF: WRD WD04 FL021 M2 8441.207 3015.161 Wakulla Lowlands Apalachicola NF: WRD WD05 FL025 D4 8432.368 3016.952

PAGE 149

149 Table A-1 continued Region Management Area Site Plot Assoc Latitude Longitude Wakulla Lowlands Apalachicola NF: WRD WD05 FL026 D4 8432.287 3016.911 Wakulla Lowlands Apalachicola NF: WRD WD06 FL027 W4 8433.080 3013.853 Wakulla Lowlands Apalachicola NF: WRD WD07 FL032 D6 8433.938 3022.967 Wakulla Lowlands Apalachicola NF: WRD WD07 FL033 M2 8434.031 3023.016 Wakulla Lowlands Apalachicola NF: WRD WD06 FL034 M2 8433.130 3013.746 Wakulla Lowlands Apalachicola NF: WRD WD05 FL035 M2 8432.176 3013.861 Wakulla Lowlands Apalachicola NF: WRD WD07 FL261 W4 8433.715 3022.721 Wakulla Lowlands Apalachicola NF: WRD WD07 FL262 W4 8433.898 3022.858 Wakulla Lowlands Apalachicola NF: WRD WD07 FL263 D6 8431.724 3021.255 Wakulla Lowlands Apalachicola NF: WRD WD05 FL267 W4 8430.400 3017.106 Wakulla Lowlands Apalachicola NF: WRD WD05 FL268 W4 8430.493 3017.197 E Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJ02 FL165 M1 8452.457 2946.138 E Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJ02 FL166 W7 8452.460 2946.223 E Panhandle Gulf Coast St. Joe Bufferlands (DEP) SJ02 FL167 M2 8452.525 2946.190 E Panhandle Gulf Coast St. Marks NWR: Panacea SM01 FL001 D2 8425.094 3002.987 E Panhandle Gulf Coast St. Marks NWR: Panacea SM01 FL002 D6 8427.700 3002.281 E Panhandle Gulf Coast St. Marks NWR: Panacea SM01 FL003 M2 8427.739 3002.233 E Panhandle Gulf Coast St. Marks NWR: Panacea SM01 FL004 W6 8427.761 3002.230 E Panhandle Gulf Coast St. Marks NWR: Panacea SM01 FL005 D2 8425.101 3002.886 E Panhandle Gulf Coast St. Marks NWR: Panacea SM03 FL006 D2 8426.210 3002.846 E Panhandle Gulf Coast St. Marks NWR: Panacea SM03 FL007 M2 8426.161 3002.904 E Panhandle Gulf Coast St. Marks NWR: Panacea SM03 FL008 M2 8426.110 3002.889 E Panhandle Gulf Coast St. Marks NWR: Panacea SM04 FL009 D6 8429.020 3002.591 E Panhandle Gulf Coast St. Marks NWR: Panacea SM04 FL305 W6 8429.047 3002.529 Big Bend Coast Cedar Key Scrub SP CK01 FL199 M1 8259.824 2912.314 Big Bend Coast Cedar Key Scrub SP CK01 FL200 W3 8259.754 2912.307 Big Bend Coast Cedar Key Scrub SP CK01 FL201 M1 8259.680 2912.317 Big Bend Coast Cedar Key Scrub SP CK01 FL212 M1 8301.709 2912.251 Big Bend Coast Cedar Key Scrub SP CK01 FL295 M1 8258.871 2912.365 Big Bend Coast Lower Suwannee NWR LS01 FL238 M2 8301.263 2927.417 Big Bend Coast Lower Suwannee NWR LS01 FL239 M1 8301.153 2927.433 Big Bend Coast Lower Suwannee NWR LS02 FL240 M1 8311.944 2923.701 Big Bend Coast Lower Suwannee NWR LS02 FL241 M1 8311.969 2923.713 Big Bend Coast St. Marks NWR: St. Marks SM06 FL023 M2 8405.373 3007.873 Big Bend Coast St. Marks NWR: St. Marks SM06 FL024 W3 8405.479 3007.913 Big Bend Coast St. Marks NWR: St. Marks SM06 FL283 M2 8409.138 3009.513 Big Bend Interior Lowlands Geothe SF GO04 FL299 M1 8237.265 2918.862 Big Bend Interior Lowlands Goethe SF GO02 FL119 M1 8235.430 2907.339 Big Bend Interior Lowlands Goethe SF GO02 FL120 M1 8236.329 2908.709 Big Bend Interior Lowlands Goethe SF GO02 FL121 W4 8236.277 2908.756 Big Bend Interior Lowlands Goethe SF GO03 FL122 W4 8236.443 2911.963 Big Bend Interior Lowlands Goethe SF GO03 FL123 M1 8237.461 2914.514

PAGE 150

150 Table A-1 continued Region Management Area Site Plot Assoc Latitude Longitude Big Bend Interior Lowlands St. Marks NWR: Wakulla SM05 FL014 D2 8418.119 3007.903 Big Bend Interior Lowlands St. Marks NWR: Wakulla SM05 FL015 M2 8418.126 3007.803 Big Bend Interior Lowlands St. Marks NWR: Wakulla SM05 FL016 W4 8418.356 3007.606 North Central Highlands Ichetucknee SP IC01 FL168 D2 8246.013 2958.323 North Central Highlands Manatee SP MA01 FL185 D1 8257.606 2929.961 North Central Highlands Manatee SP MA01 FL186 D1 8257.919 2929.953 North Central Highlands Manatee SP MA01 FL187 D1 8257.910 2929.422 North Central Highlands Oleno SP OL01 FL148 D2 8234.150 2954.866 North Central Highlands Oleno SP OL01 FL149 M1 8234.195 2954.837 North Central Highlands Oleno SP OL02 FL150 M1 8234.493 2954.978 North Central Highlands Oleno SP OL02 FL151 D1 8235.095 2955.080 North Central Highlands Oleno SP OL01 FL235 M2 8234.198 2954.432 North Central Highlands Oleno SP OL01 FL236 W1 8234.215 2954.623 North Central Highlands River Rise SP RR01 FL161 D1 8238.055 2952.208 North Central Highlands Twin Rivers SF TR01 FL126 D2 8311.814 3029.308 North Central Highlands Twin Rivers SF TR01 FL127 D2 8312.048 3029.460 North Central Highlands Twin Rivers SF TR01 FL128 D2 8312.734 3030.330 North Central Highlands Twin Rivers SF TR01 FL129 D2 8312.458 3030.175 North Central Highlands Twin Rivers SF TR02 FL252 D2 8312.337 3022.682 North Central Highlands Twin Rivers SF TR02 FL253 D1 8312.452 3022.386 North Central Highlands Twin Rivers SF TR02 FL254 D1 8312.361 3022.399 Interior Northeast Lowlands Jennings SF JE01 FL177 M2 8156.067 3010.410 Interior Northeast Lowlands Jennings SF JE01 FL178 W2 8156.087 3010.380 Interior Northeast Lowlands Jennings SF JE01 FL179 D2 8156.160 3010.816 Interior Northeast Lowlands Osceola NF OS01 FL101 M2 8224.710 3014.328 Interior Northeast Lowlands Osceola NF OS01 FL102 M2 8224.582 3014.134 Interior Northeast Lowlands Osceola NF OS01 FL103 M2 8224.743 3014.261 Interior Northeast Lowlands Osceola NF OS02 FL242 W4 8226.883 3011.485 Interior Northeast Lowlands Osceola NF OS02 FL244 M2 8226.528 3011.540 Interior Northeast Lowlands Osceola NF OS02 FL245 M1 8226.420 3011.883 Interior Northeast Lowlands Osceola NF OS03 FL247 W4 8229.095 3017.087 North Atlantic Coast Favor Dykes SP FD01 FL309 M1 8116.487 2940.420 North Atlantic Coast Favor Dykes SP FD01 FL310 M1 8115.807 2940.653 North Atlantic Coast Favor Dykes SP FD01 FL313 W2 8116.588 2940.331 North Atlantic Coast Heart Island CA HI01 FL296 M1 8121.280 2911.812 North Atlantic Coast Heart Island CA HI01 FL297 W3 8123.827 2911.175 North Atlantic Coast Heart Island CA HI01 FL298 M3 8124.455 2910.867 North Atlantic Coast private land HI01 FL301 D3 NR NR North Atlantic Coast Pumpkin Hill SP PP01 FL311 M2 8116.559 2940.512 North Atlantic Coast Pumpkin Hill SP PP01 FL312 M1 8130.126 3028.824 North Atlantic Coast Pumpkin Hill SP PP01 FL315 W4 8130.536 3028.344 North Atlantic Coast Tigar Bay SF TB01 FL300 D3 8111.349 2914.129

PAGE 151

151 Table A-1 continued Region Management Area Site Plot Assoc Latitude Longitude North Atlantic Coast Tigar Bay SF TB01 FL302 M3 8110.758 2913.184 North Atlantic Coast Tigar Bay SF TB01 FL303 M1 8110.287 2912.531 Coastal Northeast Lowlands Jennings SF JE02 FL180 M2 8155.906 3010.360 Coastal Northeast Lowlands Jennings SF JE02 FL181 D2 8155.888 3009.956 Coastal Northeast Lowlands Simmons SF SI01 FL182 M2 8156.597 3047.671 Coastal Northeast Lowlands Simmons SF SI01 FL183 D2 8156.999 3047.972 Coastal Northeast Lowlands Simmons SF SI01 FL184 W4 8157.434 3046.913 West Central Highlands Ashton (Private land) AS01 FL124 M1 8235.041 2932.284 West Central Highlands Ashton (Private land) AS01 FL125 D3 8234.781 2932.527 West Central Highlands Cross Florida Greenway SRA GW01 FL112 D3 8215.520 2903.090 West Central Highlands Cross Florida Greenway SRA GW01 FL113 D2 8215.197 2902.864 West Central Highlands Davidson Ranch TNC DR01 FL106 D3 8241.860 2944.586 West Central Highlands Davidson Ranch TNC DR01 FL107 D2 8241.710 2945.139 West Central Highlands Davidson Ranch TNC DR01 FL116 D1 8241.925 2944.969 West Central Highlands Goethe SF GO01 FL104 D3 8236.197 2921.450 West Central Highlands Goethe SF GO01 FL105 D2 8236.187 2921.480 West Central Highlands Goethe SF GO01 FL233 D3 8236.138 2921.452 West Central Highlands Goethe SF GO01 FL234 W1 8236.017 2921.580 West Central Highlands Ross Prairie SF RP01 FL114 D3 8217.955 2901.933 West Central Highlands Ross Prairie SF RP01 FL115 D2 8217.802 2901.821 West Central Highlands Ross Prairie SF RP01 FL232 D3 8217.527 2902.185 West Central Highlands San Felasco SP SF01 FL130 D3 8228.090 2944.222 West Central Highlands San Felasco SP SF01 FL131 D1 8227.762 2944.601 West Central Highlands San Felasco SP SF01 FL132 D1 8226.901 2943.903 West Central Highlands San Felasco SP SF02 FL133 M2 8226.743 2942.961 East Central Highlands Etoniah Creek SF ET01 FL117 D3 8152.401 2947.222 East Central Highlands Etoniah Creek SF ET01 FL118 D3 8152.077 2947.064 East Central Highlands Etoniah Creek SF ET02 FL306 M1 8147.309 2943.782 East Central Highlands Etoniah Creek SF ET02 FL314 W2 8120.974 2737.601 East Central Highlands Goldhead SP GH01 FL175 D3 8157.421 2950.975 East Central Highlands Goldhead SP GH01 FL176 W1 8156.427 2949.631 East Central Highlands Ocala NF OC01 FL139 D3 8148.342 2927.505 East Central Highlands Ocala NF OC01 FL142 D2 8148.587 2927.381 East Central Highlands Ocala NF OC01 FL144 D3 8148.596 2947.468 East Central Highlands Ocala NF OC01 FL145 D3 8149.589 2927.472 East Central Highlands Ocala NF OC02 FL146 W4 8156.602 2909.451 East Central Highlands Ocala NF OC02 FL147 M1 8156.316 2909.687 East Central Highlands Ocala NF OK01 FL307 D2 8154.813 2910.331 East Central Highlands Ocala NF OK01 FL308 M3 8156.729 2909.302 East Central Highlands Rock Springs SP RS01 FL206 M1 8127.770 2845.729 East Central Highlands Rock Springs SP RS01 FL207 M1 8127.623 2845.946 East Central Highlands Swisher-Ordway: Univ of FL OR01 FL152 D3 8159.907 2940.846

PAGE 152

152 Table A-1 continued Region Management Area Site Plot Assoc Latitude Longitude East Central Highlands Swisher-Ordway: Univ of FL OR01 FL153 D3 8159.764 2940.761 East Central Highlands Swisher-Ordway: Univ of FL OR01 FL154 W1 8200.641 2940.532 East Central Highlands Wekiwa Springs SP WE01 FL135 D3 8129.779 2844.307 East Central Highlands Wekiwa Springs SP WE01 FL137 M1 8129.866 2844.789 East Central Highlands Wekiwa Springs SP WE01 FL141 D3 8129.604 2843.923 East Central Highlands Wekiwa Springs SP WE01 FL143 M1 8129.792 2844.467 West Central Lowlands Green Swamp WMA GS01 FL155 M3 8457.196 2822.252 West Central Lowlands Green Swamp WMA GS02 FL156 D2 8208.258 2826.337 West Central Lowlands Green Swamp WMA GS02 FL157 M1 8207.163 2826.372 West Central Lowlands Green Swamp WMA GS02 FL158 W1 8207.115 2826.402 West Central Lowlands Green Swamp WMA GS02 FL159 W1 8207.274 2826.118 West Central Lowlands Green Swamp WMA GS02 FL160 D3 8207.151 2826.273 West Central Lowlands Green Swamp WMA GS01 FL174 W2 8157.163 2821.828 West Central Lowlands Starkey Wilderness SWFWMD SW01 FL169 M1 8235.709 2814.806 West Central Lowlands Starkey Wilderness SWFWMD SW01 FL170 W2 8235.867 2814.418 West Central Lowlands Starkey Wilderness SWFWMD SW01 FL171 M3 8235.840 2814.531 West Central Lowlands Starkey Wilderness SWFWMD SW02 FL172 D3 8236.488 2815.010 West Central Lowlands Starkey Wilderness SWFWMD SW01 FL173 W2 8235.853 2814.422 Kissimmee Basin Avon Park AFR AV01 FL188 M3 8117.614 2740.173 Kissimmee Basin Avon Park AFR AV01 FL189 W2 8117.347 2940.170 Kissimmee Basin Avon Park AFR AV01 FL190 M3 8117.231 2740.268 Kissimmee Basin Avon Park AFR AV01 FL191 M3 8117.292 2740.290 Kissimmee Basin Avon Park AFR AV02 FL192 M1 8119.719 2744.531 Kissimmee Basin Avon Park AFR AV03 FL193 W3 8112.519 2742.114 Kissimmee Basin Avon Park AFR AV02 FL196 M3 8119.757 2744.752 Kissimmee Basin Avon Park AFR AV03 FL197 W2 8112.691 2742.545 Kissimmee Basin Avon Park AFR AV03 FL198 M3 8112.664 2742.536 Kissimmee Basin Avon Park AFR AV04 FL219 M3 8112.629 2737.055 Kissimmee Basin Avon Park AFR AV05 FL220 M3 8115.758 2743.773 Kissimmee Basin Avon Park AFR AV06 FL221 M3 8119.176 2738.357 Kissimmee Basin Disney Wilderness TNC DW01 FL134 M1 8124.340 2804.012 Kissimmee Basin Disney Wilderness TNC DW01 FL136 M3 8124.281 2804.050 Kissimmee Basin Disney Wilderness TNC DW01 FL138 M3 8124.302 2804.020 Kissimmee Basin Disney Wilderness TNC DW01 FL140 M1 8124.378 2804.009 Kissimmee Basin Kississimee Prairie Preserve SP KI01 FL194 M3 8108.116 2735.082 Kissimmee Basin Kississimee Prairie Preserve SP KI01 FL195 W2 8106.414 2735.298 Kissimmee Basin Three Lakes WMA TL01 FL202 M3 8105.123 2758.670 Kissimmee Basin Three Lakes WMA TL01 FL203 W2 8105.205 2758.715 Kissimmee Basin Three Lakes WMA TL01 FL204 W2 8105.241 2758.735 Kissimmee Basin Three Lakes WMA TL01 FL205 M1 8104.352 2758.957 Southwest Central Lowlands Myakka River SP MY01 FL208 M1 8210.671 2714.092 Southwest Central Lowlands Myakka River SP MY02 FL209 M1 8217.429 2717.423

PAGE 153

153 Table A-1 continued Region Management Area Site Plot Assoc Latitude Longitude Southwest Central Lowlands Myakka River SP MY03 FL210 M3 8214.965 2716.027 Southwest Central Lowlands Myakka River SP MY03 FL211 W2 8214.926 2716.119 Southwest Central Lowlands Myakka River SP MY04 FL213 W2 8213.196 2713.914 Southwest Central Lowlands Myakka River SP MY04 FL214 M3 8212.229 2713.572 Southwest Central Lowlands Myakka River SP MY03 FL215 M3 8215.152 2715.225 Southwest Highlands Withlacoochee SF WF01 FL108 D2 8225.963 2847.513 Southwest Highlands Withlacoochee SF WF01 FL109 D1 8225.653 2848.063 Southwest Highlands Withlacoochee SF WF02 FL110 D2 8224.071 2843.674 Southwest Highlands Withlacoochee SF WF02 FL111 D2 8223.958 2843.519 Caloosahatchie Lowlands Cecil Webb WMA CW01 FL216 M3 8156.349 2652.335 Caloosahatchie Lowlands Cecil Webb WMA CW01 FL217 W2 8156.456 2652.301 Caloosahatchie Lowlands Cecil Webb WMA CW01 FL218 W2 8155.595 2653.341 Caloosahatchie Lowlands Cecil Webb WMA CW02 FL222 M3 8151.861 2651.547 Caloosahatchie Lowlands Cecil Webb WMA CW02 FL223 W2 8151.948 2651.597

PAGE 154

154 APPENDIX B LIST OF FREQUENT AND ABUNDANT SPE CIES BY COMMUNITY ASSOCIATION Table B-1: Species included that present in > 70-75 % of plots within an association, and > 0.2 m2 mean cover. Freq = per cent frequency of occurrence, cover = mean cover in m2. Number of plots per associat ion indicated in parentheses Peninsula Xeric Sandhills (22) Freq Cover Aristida beyrichiana 100 39.80 Sorghastrum secundum 100 2.99 Pityopsis graminifolia 100 2.52 Lechea sessiliflora 100 0.44 Schizachyrium scoparium var. stoloniferum 95 1.28 Dichanthelium ovale var. addisonii 95 0.80 Stillingia sylvatica 95 0.72 Sporobolus junceus 91 1.14 Paspalum setaceum 91 0.51 Cnidoscolus stimulosus 91 0.28 Bulbostylis ciliatifolia 86 0.84 Andropogon ternarius 86 0.76 Smilax auriculata 86 0.59 Rhynchospora grayi 86 0.33 Tragia urens 86 0.33 Crotalaria rotundifolia 86 0.30 Balduina angustifolia 82 0.50 Andropogon gyrans var. gyrans 82 0.36 Tephrosia chrysophylla 77 1.11 Croton argyranthemus 77 0.48 Liatris tenuifolia var. tenuifolia 77 0.40 Scleria ciliata var. ciliata 77 0.30 Panhandle Xeric Sandhills (31) Freq Cover Schizachyrium scoparium var. stoloniferum 100 4.38 Smilax auriculata 100 1.66 Andropogon gyrans var. gyrans 100 1.26 Stylisma patens ssp. patens 100 0.45 Stylosanthes biflora 94 0.31 Pityopsis aspera 90 3.29 Bulbostylis ciliatifolia 90 1.16 Cyperus lupulinus ssp. lupulinus 90 0.38 Galactia microphylla 87 2.38 Sorghastrum secundum 87 1.91 Eriogonum tomentosum 87 1.04 Rhynchospora grayi 87 0.58 Dichanthelium angustifolium 87 0.52 Andropogon virginicus 84 2.23 Solidago odora var. odora 84 1.56 Scleria ciliata var. ciliata 84 0.54 Commelina erecta 84 0.35

PAGE 155

155 Table B-1 continued Panhandle Xeric Sandhills (continued) Freq Cover Dichanthelium ovale var. addisonii 81 0.88 Sporobolus junceus 81 0.77 Aristida beyrichiana 77 19.06 Schizachyrium tenerum 77 1.19 Tragia urens 77 0.28 Croton argyranthemus 74 0.75 Liatris tenuifolia var. tenuifolia 74 0.36 North Florida Sandhills (31) Freq Cover Aristida beyrichiana 100 33.06 Pityopsis graminifolia 97 3.50 Dichanthelium ovale var. addisonii 97 0.96 Paspalum setaceum 97 0.79 Scleria ciliata var. ciliata 97 0.65 Tragia urens 97 0.42 Sorghastrum secundum 94 2.73 Schizachyrium scoparium var. stoloniferum 94 1.46 Stillingia sylvatica 94 0.64 Rhynchosia reniformis 94 0.44 Helianthemum carolinianum 94 0.30 Dichanthelium angustifolium 90 0.72 Stylisma patens ssp. patens 90 0.33 Crotalaria rotundifolia 87 0.35 Dyschoriste oblongifolia 84 1.17 Andropogon gyrans var. gyrans 84 0.92 Eupatorium compositifolium 84 0.59 Gymnopogon ambiguus 84 0.45 Croton argyranthemus 84 0.44 Rhynchospora grayi 84 0.38 Sporobolus junceus 81 0.89 Lechea sessiliflora 81 0.73 Andropogon ternarius 81 0.63 Vernonia angustifolia 81 0.55 Sericocarpus tortifolius 81 0.44 Liatris tenuifolia var. tenuifolia 81 0.40 Symphyotrichum concolor 81 0.36 Stylosanthes biflora 81 0.21 Smilax auriculata 77 1.44 Lespedeza hirta 77 0.39 Palafoxia integrifolia 77 0.38 Ruellia caroliniensis ssp. ciliosa 77 0.31 Aristolochia serpentaria 77 0.28 Hieracium gronovii 77 0.26 Elephantopus elatus 74 2.04 Solidago odora var. odora 74 0.50

PAGE 156

156 Table B-1 continued North Florida Sandhills (continued) Freq Cover Aristida purpurascens var. purpurascens 74 0.37 Andropogon virginicus 74 0.29 Ageratina aromatica 74 0.29 North Florida Rich Woodlands (11) Freq Cover Pteridium aquilinum 100 4.72 Sorghastrum secundum 100 4.00 Ageratina aromatica 100 1.24 Dichanthelium angustifolium 100 0.97 Paspalum setaceum 100 0.75 Andropogon gyrans var. gyrans 100 0.39 Smilax auriculata 100 0.36 Dichanthelium ovale var. addisonii 91 1.11 Scleria ciliata var. ciliata 91 0.77 Sericocarpus tortifolius 91 0.75 Houstonia procumbens 91 0.48 Andropogon virginicus 91 0.40 Aristolochia serpentaria 82 0.31 Galium pilosum 82 0.31 Cyperus plukenetii 82 0.27 Hypericum hypericoides 82 0.18 Dyschoriste oblongifolia 73 1.61 Eupatorium compositifolium 73 1.49 Dichanthelium oligosanthes var. oligosanthes 73 1.09 Dichanthelium aciculare 73 0.98 Pityopsis graminifolia 73 0.93 Stillingia sylvatica 73 0.47 Aristida purpurascens var. purpurascens 73 0.42 Andropogon ternarius 73 0.39 Helianthemum carolinianum 73 0.38 Cyperus retrorsus 73 0.31 Rhynchosia reniformis 73 0.31 Crotalaria rotundifolia 73 0.26 Hieracium gronovii 73 0.24 Panhandle Longleaf Pine Clayhills (14) Freq Cover Aristida beyrichiana 100 22.72 Schizachyrium scoparium var. stoloniferum 100 8.75 Solidago odora var. odora 100 4.46 Dichanthelium angustifolium 100 1.65 Elephantopus elatus 100 1.59 Vernonia angustifolia 100 0.82 Stylosanthes biflora 100 0.50 Sericocarpus tortifolius 93 1.66 Andropogon gyrans var. gyrans 93 1.13 Symphyotrichum dumosum var. dumosum 93 1.09 Desmodium lineatum 93 1.04

PAGE 157

157 Table B-1 continued Panhandle Longleaf Pine Clayhills (continued) Freq Cover Lespedeza repens 93 0.38 Hieracium gronovii 93 0.29 Pteridium aquilinum 86 5.65 Rubus cuneifolius 86 1.19 Desmodium ciliare 86 1.12 Dichanthelium ovale var. addisonii 86 1.05 Muhlenbergia capillaris var. trichopodes 86 1.00 Scleria ciliata var. ciliata 86 0.86 Aristida purpurascens var. purpurascens 86 0.73 Andropogon virginicus 86 0.52 Symphyotrichum adnatum 86 0.47 Symphyotrichum concolor 86 0.44 Mimosa microphylla 86 0.40 Eupatorium compositifolium 86 0.38 Liatris gracilis 86 0.34 Rudbeckia hirta 86 0.29 Pityopsis graminifolia 79 4.64 Schizachyrium tenerum 79 2.27 Pityopsis aspera 79 1.37 Ageratina aromatica 79 0.95 Chamaecrista nictitans 79 0.78 Euphorbia discoidalis 79 0.58 Chrysopsis mariana 79 0.46 Smilax glauca 79 0.40 Acalypha gracilens 79 0.27 Aristolochia serpentaria 79 0.27 Houstonia procumbens 79 0.26 Gymnopogon ambiguus 79 0.23 Panhandle Silty Woodlands (22) Freq Cover Aristida beyrichiana 100 33.64 Schizachyrium scoparium var. stoloniferum 100 3.28 Dichanthelium dichotomum var. tenue 100 1.15 Scleria ciliata var. ciliata 100 0.75 Symphyotrichum adnatum 100 0.65 Dichanthelium angustifolium 95 1.79 Tragia smallii 95 0.47 Pityopsis graminifolia 91 5.36 Andropogon gyrans var. gyrans 91 1.07 Sericocarpus tortifolius 91 0.73 Stylosanthes biflora 91 0.3 Chrysopsis mariana 86 0.52 Andropogon virginicus 86 0.41 Smilax auriculata 82 1.38 Lespedeza repens 82 0.27 Viola septemloba 82 0.25 Pteridium aquilinum 77 8.8

PAGE 158

158 Table B-1 continued Panhandle Silty Woodlands (continued) Freq Cover Helianthus radula 77 3.11 Carphephorus odoratissimus 77 1.01 Galactia erecta 77 0.27 Xeric Mesic Flatwoods (36) Freq Cover Aristida beyrichiana 92 18.24 Dichanthelium sabulorum var. thinium 86 0.75 Andropogon virginicus 86 0.69 Pityopsis graminifolia 83 0.85 Smilax auriculata 81 0.51 Pterocaulon virgatum 78 0.42 Bulbostylis ciliatifolia 75 0.67 Gratiola hispida 72 0.43 North Florida Mesic Flatwoods (30) Freq Cover Aristida beyrichiana 97 14.93 Xyris caroliniana 93 0.79 Pityopsis graminifolia 87 1.64 Andropogon virginicus 87 1.07 Dichanthelium strigosum var. leucoblepharis 77 0.70 Pterocaulon virgatum 77 0.31 Pteridium aquilinum 73 3.05 Sericocarpus tortifolius 73 0.21 Dichanthelium sabulorum var. thinium 73 0.70 Central Florida Mesic Flatwoods/D ry Prairies (22) Freq Cover Aristida beyrichiana 100 27.23 Dichanthelium sabulorum var. thinium 100 0.85 Andropogon virginicus 95 1.63 Pterocaulon virgatum 95 0.47 Aristida spiciformis 91 2.06 Pityopsis graminifolia 91 0.82 Dichanthelium chamaelonche 86 9.49 Xyris caroliniana 86 0.40 Paspalum setaceum 86 0.32 Drosera brevifolia 86 0.30 Polygala setacea 86 0.28 Euthamia tenuifolia var. tenuifolia 82 0.81 Oldenlandia uniflora 82 0.35 Fimbristylis puberula 82 0.30 Eleocharis baldwinii 77 0.67 Gratiola hispida 77 0.33 Marginal Priaries (11) Freq Cover Andropogon virginicus 91 10.23 Euthamia tenuifolia var. tenuifolia 91 0.80 Panicum hemitomon 82 12.35

PAGE 159

159 Table B-1 continued Marginal Priaries (continued) Freq Cover Rhexia mariana var. mariana 82 0.75 Axonopus furcatus 73 15.27 Eupatorium leptophyllum 73 2.84 Centella erecta 73 2.72 Andropogon capillipes (wetland varient) 73 1.23 Aristida purpurascens var. virgata 73 0.60 Peninsula Wet Flatwoods/Prairies (16) Freq Cover Oxypolis filiformis 100 1.73 Eriocaulon decangulare 94 5.87 Bigelowia nudata 94 2.16 Eragrostis elliottii 94 1.11 Amphicarpum muehlenbergianum 88 5.95 Xyris elliottii 88 3.68 Andropogon gyrans var. stenophyllus 88 0.66 Aristida beyrichiana 81 15.24 Dichanthelium erectifolium 81 2.84 Centella erecta 81 1.41 Drosera brevifolia 81 0.40 Eupatorium mohrii 81 0.32 Peninsula Wet Flatwoods/Prairies (16) Freq Cover Aristida palustris 75 6.44 Scleria muehlenbergii 75 5.38 Andropogon capillipes (wetland varient) 75 4.86 Fuirena scirpoidea 75 2.63 Panicum tenerum 75 1.98 Calcareous Wet Flatwoods (4) Freq Cover Sabal palmetto 100 4.75 Centella erecta 100 2.09 Hyptis alata 100 1.50 Saccharum giganteum 100 0.63 Helenium pinnatifidum 100 0.34 Lobelia glandulosa 100 0.34 Panicum rigidulum var. rigidulum 100 0.34 Rhynchospora globularis 100 0.28 Cirsium nuttallii 100 0.22 Asclepias lanceolata 100 0.09 Dichanthelium dichotomum var. nitidum 75 5.31 Panicum virgatum var. virgatum 75 3.31 Rhynchospora divergens 75 2.44 Dichanthelium caerulescens 75 2.13 Pluchea rosea 75 2.00 Ludwigia microcarpa 75 1.75 Rhynchospora colorata 75 1.13 Fuirena breviseta 75 1.00

PAGE 160

160 Table B-1 continued Calcareous Wet Flatwoods (continued) Freq Cover Mikania scandens 75 0.97 Rubus trivialis 75 0.88 Scleria muehlenbergii 75 0.78 Hypericum cistifolium 75 0.75 Diodia virginiana 75 0.72 Rhynchospora perplexa 75 0.59 Dichanthelium strigosum var. glabrescens 75 0.53 Hypericum hypericoides 75 0.53 Andropogon capillipes (upland varient) 75 0.50 Berchemia scandens 75 0.50 Eustachys glauca 75 0.44 Scleria pauciflora 75 0.44 Phyla nodiflora 75 0.41 Cyperus polystachyos 75 0.38 Oxypolis filiformis 75 0.31 Proserpinaca pectinata 75 0.31 Smilax laurifolia 75 0.31 Andropogon glomeratus var. glomeratus 75 0.28 Mitreola petiolata 75 0.28 Toxicodendron radicans 75 0.28 Axonopus furcatus 75 0.25 Eleocharis flavescens 75 0.25 Mitreola sessilifolia 75 0.25 Vitis rotundifolia 75 0.25 Xyris jupicai 75 0.25 North Florida Shrubby Wet Flatwoods (15) Freq Cover Andropogon glaucopsis 87 5.20 Osmunda cinnamomea 87 4.38 Eriocaulon decangulare 80 7.57 Smilax laurifolia 80 0.66 Xyris ambigua 80 0.39 Photinia pyrifolia 80 0.34 Andropogon glomeratus var. hirsutior 73 2.27 Rhynchospora fascicularis 73 2.09 Andropogon capillipes (upland varient ) 73 0.43 Rhexia petiolata 73 0.23 Upper Panhandle Wet Flatwoods (7) Freq Cover Schizachyrium scoparium var. stoloniferum 100 5.32 Pityopsis graminifolia 100 5.11 Pteridium aquilinum 100 4.68 Eupatorium rotundifolium 100 4.21 Andropogon virginicus 100 4.14 Panicum verrucosum 100 3.52 Rhexia alifanus 100 2.43 Helianthus angustifolius 100 1.64

PAGE 161

161 Table B-1 continued Upper Panhandle Wet Flatwoods (continued) Freq Cover Euthamia tenuifolia var. tenuifolia 100 1.32 Symphyotrichum dumosum var. dumosum 100 1.09 Smilax glauca 100 0.95 Solidago stricta 100 0.93 Diodia virginiana 100 0.50 Xyris caroliniana 100 0.48 Chamaecrista nictitans 100 0.34 Ctenium aromaticum 86 10.54 Dichanthelium dichotomum var. tenue 86 3.11 Andropogon glomeratus var. hirsutior 86 2.82 Aristida purpurascens var. virgata 86 2.32 Upper Panhandle Wet Flatwoods (7) Freq Cover Panicum anceps var. rhizomatum 86 1.88 Chaptalia tomentosa 86 1.73 Dichanthelium strigosum var. leucoblepharis 86 1.59 Panicum virgatum var. virgatum 86 0.70 Desmodium tenuifolium 86 0.64 Bigelowia nudata 86 0.55 Hypericum crux-andreae 86 0.48 Andropogon gyrans var. gyrans 86 0.41 Dichanthelium consanguineum 86 0.38 Hypericum setosum 86 0.27 Gymnopogon brevifolius 86 0.25 Crotalaria purshii 86 0.23 Polygala nana 86 0.20 Rubus trivialis 86 0.20 Panhandle Wet Flatwoods/Prairies (16) Freq Cover Aristida beyrichiana 100 50.51 Xyris ambigua 100 2.00 Rhexia alifanus 100 1.79 Smilax laurifolia 100 0.95 Ctenium aromaticum 94 9.88 Carphephorus pseudoliatris 94 1.38 Eriocaulon decangulare 88 4.75 Chaptalia tomentosa 88 1.21 Andropogon arctatus 81 5.31 Helianthus heterophyllus 81 2.47 Andropogon gyrans var. stenophyllus 81 1.43 Erigeron vernus 81 0.96 Coreopsis linifolia 81 0.74 Rhynchospora chapmanii 75 10.92 Bigelowia nudata 75 1.66 Muhlenbergia capillaris var. trichopodes 75 1.48 Rhynchospora plumosa 75 1.23 Rhynchospora baldwinii 75 0.95

PAGE 162

162 Table B-1 continued Panhandle Seepage Slopes (5) Freq Cover Aristida beyrichiana 100 11.15 Scleria muehlenbergii 100 10.05 Rhynchospora oligantha 100 6.03 Aristida palustris 100 3.05 Andropogon gyrans var. stenophyllus 100 2.58 Eriocaulon decangulare 100 2.33 Smilax laurifolia 100 1.05 Bigelowia nudata 100 0.85 Coreopsis linifolia 100 0.75 Andropogon arctatus 100 0.70 Lobelia glandulosa 100 0.60 Rhynchospora latifolia 100 0.58 Oxypolis filiformis 100 0.50 Symphyotrichum dumosum var. dumosum 100 0.35 Rhexia alifanus 100 0.30 Sabatia macrophylla 100 0.15 Liatris spicata 80 9.28 Muhlenbergia capillaris var. trichopodes 80 6.70 Ctenium aromaticum 80 5.48 Pleea tenuifolia 80 1.03 Arnoglossum ovatum 80 0.73 Dichanthelium longiligulatum 80 0.65 Lophiola aurea 80 0.60 Paspalum praecox 80 0.50 Balduina uniflora 80 0.43 Lycopodiella appressa 80 0.43 Rhexia lutea 80 0.35 Drosera brevifolia 80 0.33 Erigeron vernus 80 0.30 Rubus trivialis 80 0.30 Juncus trigonocarpus 80 0.23 Rhexia petiolata 80 0.20 Eryngium integrifolium 80 0.20

PAGE 163

163 APPENDIX C MASTER LIST OF ABITA CREEK PRESERVE PLANT SPECIES. Table C-1: All vascular plant sp ecies (and varieties) recorded at Abita Creek Preserve during sample period 1997-2005. Code corresponds to labels on Figure 4-5 and 4-6. “Type” indicates woody (W) or herbaceous (H ). “Lifeform” indicates forb (F), graminoid (G), and woody (W). Code Species Type Lifeform ACERU Acer rubrum W W AGAOB Agalinus obtusifolia H F AGASP Agalinus sp. H F AGRPE Agrostis perennans H G ALESP Aletris sp. H F AMBAR Ambrosia artemisiifolia H F AMOSP Amorpha sp. H F ANDCA Andropogon capillipes H G ANDGL Andropogon glomeratus H G ANDGY Andropogon gyrans var. gyrans H G ANDMO Andropogon mohrii H G ANDPE Andropogon perangustatus H G ANDSP Andropogon sp. H G ANDVI Andropogon virginicus H G ANTRU Anthaenantia rufa H G ANTVI Anthaenantia villosa H G ARIPA Aristida palustris H G ARIVI Aristida virgata H G ARUTE Arundinaria gigantea ssp. tecta H G ASCLO Asclepias sp. H F ASCLO Asclepias longifolia H F ASTAD Symphyotrichum adnatum H F ASTDU Symphyotrichum dumosum var. dumosum H F AXOFI Axonopus fissifolius H G BACHA Baccharis halimifolia W W BALUN Balduwiana uniflora H F BARPA Bartonia paniculata H F BIDMI Bidens mitis H F BIGCA Bignonia capreolata H F BIGNU Bigelowia nudata H F BOLSP Boltonia sp H F BURSP Burmannia sp. H F CACOV Cacalia ovata H F CALAM Callicarpa americana W W CARGL Carex glaucescens H C CARPS Carphephorus pseudoliatris H F

PAGE 164

164 Table C-1 continued Code Species Type Lifeform CENER Centella erecta H F CEPOC Cephalanthus occidentalis W W CHALA Chasmanthium laxum H G CHAOR Chasmanthium ornithorhynchum H G CHATO Chaptalia tomentosa H F CHIVI Chionanthus virginicus W W CLEAL Clethra alnifolia W W CLEDI Cleistes divaricata H F COERU Coelorachis rugosa H G COETE Coelorachis tessellata H G CORLI Coreopsis linifolia H F CRASP Crataegus sp. W W CROTO Croton sp. H F CTEAR Ctenium aromaticum H G CYPSP Cyperus compressus H C CYRRA Cyrilla racemiflora W W DICDI Dichanthelium dichotomum H G DICLA Rhynchospora latifolia H C DIOTE Diodia virginiana H F DIOTE Diodia teres H F DIOVI Diospyros virginiana W W DROBR Drosera sp. H F ELEMI Eleocharis minima H C ELETU Eleocharis tuberculosa H C ERARE Eragrostis refracta H G ERARE Eragrostis elliotii H G EREHI Erechtites hieraciifolia H F ERICO Eriocaulon compressum H F ERIDE Eriocaulon decangulare H F ERIGI Saccharum giganteum H G ERIST Saccharum strictus H G ERIVE Erigeron vernus H F ERYIN Eryngium integrifolium H F EUPCA Eupatorium capillifolium H F EUPLE Eupatorium leucolepis H F EUPRO Eupatorium rotundifolium H F EUPSE Eupatorium semiserratum H F EUTLE Euthamia leptocephala H F EUTTE Euthamia tenuifolia var. tenuifolia H F FRAPE Fraxinus caroliniana W W FUIBR Fuirena breviseta H G FUISP Fuirena sp. H G

PAGE 165

165 Table C-1 continued Code Species Type Lifeform GALMO Gaylussacia mosieri W W GAYMO Gaylussacia dumosa W W GELRA Gelsemium rankinii H F GENSA Gentiana saponaria H F GRAPI Gratiola pilosa H F GRAPI Gratiola brevifolia H F GRASP Gratiola sp. H F GYMBR Gymnopogon brevifolius H G HELAN Helianthus angustifolius H F HELHE Helianthus heterophyllus H F HELVE Helenium vernale H F HIBAC Hibiscus asculenta H F HYPAL Hyptis alata H F HYPBR Hypericum brachyphyllum W W HYPCI Hypericum cistifolium H F HYPHI Hypoxis sp. H F HYPHY Hypericum hypericoides H F HYPMU Hypericum multilum H F HYPSE Hypericum setosum H F HYPST Hypericum crux-andreae H F HYPWA Triadenum virginicum W W ILECO Ilex coriacea W W ILEDE Ilex decidua W W ILEGL Ilex glabra W W ILEMY Ilex myrtifolia W W ILEOP Ilex opaca W W ILEVO Ilex vomitoria W W IRIVI Iris virginica H F ITEVI Itea virginica W W JUNMA Juncus marginatus H C JUNTR Juncus trigonocarpus H C LACCA Lachnanthes caroliana H F LECSP Lechea sp. H F LIASP Liatris spicata H F LIGSI Ligustrum sinense W W LINME Linum medium H F LINME Linum floridanum H F LIQST Liquidambar styraciflua W W LIRTU Liriodendron tulipifera W W LOBBR Lobelia brevifolia H F LOBFL Lobelia floridana H F LOBPU Lobelia puberula H F

PAGE 166

166 Table C-1 continued Code Species Type Lifeform LOPAU Lophiola aurea H F LUDGL Ludwigia glandulosa H F LUDHI Ludwigia pilosa H F LUDHI Ludwigia hirtella H F LUDLI Ludwigia linearis H F LUDSP Ludwigia sp. H F LUDVI Ludwigia virgata H F LYCAL Lycopodiella sp. H F LYCVI Lycopus virginicus H F LYCVI Lycopus rubellus var. angustifolius H F LYGJA Lygodium japonicum H F LYOLU Lyonia lucida W W MAGGR Magnolia grandiflora W W MAGVI Magnolia virginiana W W MALAN Malus angustifolia W W MECAC Mecardonia acuminata H F MITSE Mitreola sessilifolia H F MUHEX Muhlenbergia cappillaris var. tricopodes H G MYRCE Morella cerifera W W MYRHE Morella heterophylla W W NYSBI Nyssa biflora W W OSMAM Osmanthus americanus W W OSMCI Osmunda cinnamomea H F OSMRE Osmunda regalis H F OXYFI Oxypolis filiformis H F PANAC Dichanthelium acuminatum H G PANAN Panicum anceps H G PANAN Panicum hians H G PANCO Dichanthelium consanguineum H G PANEN Dichanthelium ensifolium H G PANER Dichanthelium erectifolium H G PANET Dichanthelium ensifolium var. tenue H G PANLE Dichanthelium leaucothrix H G PANLO Dichanthelium longiligulatum H G PANRI Panicum rigidulum H G PANSC Dichanthelium scabriusculum H G PANSO Dichanthelium scoparium H G PANSP Dichanthelium sp. H G PANST Dichanthelium strigosum H G PANTE Panicum tenerum H G PANVE Panicum verrucosum H G PANVI Panicum virgatum H G

PAGE 167

167 Table C-1 continued Code Species Type Lifeform PARQU Parthenocissus quinquefolia W W PASFL Paspalum floridanum H G PASPR Paspalum praecox H G PASSE Paspalum setaceum H G PENSP Penstemon sp. H F PERBO Persea borbonia W W PINEL Pinus elliottii W W PINPA Pinus palustris W W PINTA Pinus taeda W W PITGR Pityopsis graminifolia H F PLURO Pluchea rosea H F PLURO Pluchea foetida H F POLLU Polygala lutea H F POLRA Polygala ramosa H F PROPE Proserpinaca pectinata H F PRUSE Prunus serotina W W PTEAQ Pteridium aquilinum H F PYRAR Photinia pyrifolia W W QUEFA Quercus falcata W W QUELA Quercus laurifolia W W QUENI Quercus nigra W W QUENI Quercus laurifolia W W QUEVI Quercus virginiana W W RHEAL Rhexia alifanus H F RHELU Rhexia lutea H F RHEMA Rhexia mariana var. mariana H F RHEPE Rhexia petiolata H F RHESP Rhexia sp. H F RHEVI Rhexia virginiana H F RHOSP Rhododendron sp. W W RHUCO Rhus copallinum W W RHUVE Toxicodendron vernix W W RHYCA Rhynchospora chalarocephala H C RHYCE Rhynchospora cephalantha H C RHYCH Rhynchospora chapmanii H C RHYCN Rhynchospora corniculata H C RHYCO Rhynchospora compressa H C RHYDE Rhynchospora debilis H C RHYEL Rhynchospora elliottii H C RHYFI Rhynchospora filifolia H C RHYGB Rhynchospora globularis H C RHYGL Rhynchospora glomerata H C

PAGE 168

168 Table C-1 continued Code Species Type Lifeform RHYGR Rhynchospora gracilenta H C RHYIN Rhynchospora inexpansa H C RHYOL Rhynchospora oligantha H C RHYPL Rhynchospora plumosa H C RHYPU Rhynchospora pusilla H C RHYRA Rhynchospora rariflora H C RHYSP Rhynchospora sp. H C RUBUS Rubus sp. H F RUENO Ruellia noctiflora H F SABSP Sabatia sp. H F SABSP Sabatia difformis H F SABSP Sabatia campanulata H F SAGLA Sagittaria lanceolata H F SALAZ Salvia azurea H F SAPSI Sapium sebiferum W W SARAL Sarracenia alata H F SARPS Sarracenia psittacina H F SCHSC Schizachyrium scoparium H G SCHTE Schizachyrium tenerum H G SCLCI Scleria ciliata var. ciliata H C SCLGE Scleria georgiana H C SCLHI Scleria hirtella H C SCLMU Scleria muhlenbergia H C SCLPA Scleria pauciflora var. caroliniana H C SCLPP Scleria pauciflora var. pauciflora H C SCLTR Scleria triglomerata H C SCUIN Scutellaria integrifolia H F SETSP Setaria sp. H G SISAL Sisyrinchium atlanticum H F SMIBO Smilax bona-nox H F SMIGL Smilax glauca H F SMILA Smilax laurifolia H F SMIRO Smilax rotundifolia H F SMISM Smilax smallii H F SOLOD Solidago odora H F SOLRU Solidago rugosa H F STOLA Stokesia laevis H F STYAM Styrax americanus W W SYMTI Symplocos tinctoria W W TEPON Tephrosia onobrachyoides H F TILUS Tillandsia usneoides H F TOFRA Tofieldia racemosa H F

PAGE 169

169 Table C-1 continued Code Species Type Lifeform TOXRA Toxicodendron radicans W W TRADI Trachelospermum difforme H F TRIAM Tridens ambiguus H G TRIVI Triadenum virginicum W W UTRIC Utricularia juncea H F VACAR Vaccinium arboreum W W VACEL Vaccinium elliottii W W VIBDE Viburnum dentatum W W VIBNU Viburnum nudum W W VIOLA Viola lanceolata H F VIOPR Viola primulifolia H F VITRO Vitis rotundifolia W W WOOAR Woodwardia areolata H F XYRAM Xyris ambigua H F XYRBA Xyris baldwiniana H F XYRCA Xyris caroliniana H F XYRDI Xyris difformis H F XYRLO Xyris louisianica H F XYRSM Xyris smalliana H F XYRSP Xyris sp. H F XYRST Xyris iridifolia H F XYRST Xyris sticta H F ZIGSP Zigadenus sp. H F

PAGE 170

170 LIST OF REFERENCES Abella, S. R., V. B. Shelburne, and N. W. Ma cDonald. 2003. Multifactor classification of forest landscape ecosystems of Jo cassee Gorges, southern Appalachian Mountains, South Carolina. Canadian Journal of Forest Research 33 :1933-1946. Abrahamson, W. G., and D. C. Hartnett. 1990. Flatwoods and dry prairies. Pages 103-149 in R. L. Myers and J. J. Ewel, editors. Ecosystems of Florida. University of Central Florida Press, Orlando, Florida, USA. Adler, P. B., E. P. White, W. K. Lauenroth, D. M. Kaufman, A. Rassweiler, J. A. Rusak. Evidence for a general species-t ime-area relationship. Ecology 86 : 2032-2039. Bailey, R. G., P. E. Avers, T. King, and W. H. McNab. 1994. Ecoregions and subregions of the United States 1:7,500,000 (map) with supplemen tary table of map unit descriptions, compiled and edited by W. H. McNab and R. G. Bailey. USDA Forest Service, Washington D.C., USA. Bell, G. 2001. Neutral macroecology. Science 293 :2413-2418. Bocard, D., L. Legendre, C. Avois-Jacquet, and H. Tuomisto. 2004. Dissecting the spatial structure of ecological data at multiple scales. Ecology 85 :1826-1832. Bocard, D., and P. Legendre. 2002. All-scale spatia l analysis of ecological data by means of principal coordinates of neighbour matrices. Ecological Modelling 153 :51-68. Borcard, D., and P. Legendre. 1994. Environmental control and spatial st ructure in ecological communities: an example using oribatid mites (Acari, Orbatei). Environmental and Ecological Statistics 1 :1045-1055. Borcard, D., P. Legendre, and P. Drapeau. 1992. Partialling out the spatial component of ecological variation. Ecology 73 :1045-1055. Brady, N. C., and R. R. Weil. 2000. Elements of the nature and propertie s of soils, 12 edition. Prentice-Hall Inc., Upper Saddl e River, New Jersey, USA. Bray, J. R., and J. T. Curtis. 1957. An ordinati on of the upland forest communities of southern Wisconsin. Ecological Monographs 27 :325-349. Bridges, E. L. 2006a. Histori cal accounts of vegetation in th e Kissimmee River dry prairie landscape. in R. Noss, editor. Proceedings of the Florida Dry Prairie Conference. University of Central Florid a, Orlando, Florida, USA.

PAGE 171

171 Bridges, E. L. 2006b. Landscape Ecology of Florid a Dry Prairie in the Kissimmee River Region. Pages 14-42 in R. Noss, editor. Land of Fire and Wa ter. Proceedings of the Florida Dry Prairie Conference. University of Ce ntral Florida, Orlando, Florida, USA. Bridges, E. L., and S. L. Orzell. 1989. Longleaf pine communities of the West Gulf Coastal Plain. Natural Areas Journal 9 :246-263. Brockway, D. G., and C. E. Lewis. 1997. Long-term effects of dormant-season prescribed fire on plant community diversity, structure, and productivity in a longleaf pine-wiregrass ecosystem. Forest Ecology and Management 96 :167-183. Brockway, D. G., and K. W. Outcult. 2000. Rest oring longleaf pine wiregrass ecosystems: Hexazinone application enhances effects of prescribed fire. Forest Ecology and Management 137 :121-138. Brooks, H. K. 1982. Guide to the Ph ysiographic Divisions of Flor ida. IFAS Florida Cooperative Extension Services, University of Fl orida, Gainesville, Florida, USA Brown, R. B., E. L. Stone, and V. W. Carlisle. 1990. Soils. Pages 35-69 in R. L. Myers and J. J. Ewel, editors. Ecosystems of Florida. Univ ersity of Central Florida Press, Orlando, Florida, USA. Chen, E., and J. F. Gerber. 1990. Climate. Pages 11-34 in R. L. Myers and J. J. Ewel, editors. Ecosystems of Florida. University of Ce ntral Florida Press, Orlando, Florida, USA. Christensen, N. L. 1977. Fire and soil-plant nutrient relations in a pine-wiregrass savanna on the coastal plain of North Carolina. Oecologia 31 :1932-1939. Cleland, D. T., J. B. Hart, G. E. Host, K. Pregitzer, and C. W. Ramm. 1993. Field guide: ecological classification and i nventory system of the Hur on-Manistee National Forests. U. S. Department of Agriculture Fo rest Service, Wash ington D.C., USA. Clewell, A. F. 1971. The Vegetation of the Ap alachicola National Forest: an Ecological Perspective. U. S. Department of Agricult ure Forest Service, Atlanta, Georgia, USA Clewell, A. F. 1985. Guide to the Vascular Pl ants of the Florida Panhandle. Florida State University Press, Tallahassee, Florida, USA. Collins, S. L., S. M. Glenn, and J. M. Briggs 2002. Effect of local and regional processes on plant species richness in tallgrass prairie. Oikos 99 :571-579. Comer, P., D. Faber-Langendoen, R. Evans, S. Gawl er, D. Josse, G. Kittel, S. Menard, M. Pyne, M. Reid, K. Schulz, K. Snow, and J. Teague. 2003. Ecological Systems of the United States: A Working Classification of U.S. Te rrestrial Systems. NatureServe, Arlington, Virginia, USA.

PAGE 172

172 Condit, R., N. Pitman, G. Leigh, J. Chave, J. Ter borgh, R. B. Foster, P. Nunez, V. S. Aguilar, R. Valencia, G. Villa, H. C. Muller-Landau, E. Losos, and S. P. Hubbell. 2002. Betadiversity in tropical forest trees. Science 295 :666-669. Cooper, A., T. McCann, and R. G. H. Bunce. 2006. The influence of sampling intensity on vegetation classification and the impli cations for environmental management. Environmental Conservation 33 :118-127. Cornell, H. V., and J. H. Lawton. 1992. Species in teractions, local and regional processes, and limits to the richness of ecological communitie s: a theoretical perspective. Journal of Animal Ecology 61 :1-12. Cox, A. C., D. R. Gordon, J. L. Slapcinsky, a nd G. S. Seamon. 2004. Understory restoration in longleaf pine sandhills. Natural Areas Journal 24 :4-14. Croker, T. C. 1987. Longleaf Pine: A History of Man and a Forest. USDA Forest Service Forestry Report R8-FR7 Asheville, North Carolina, USA. Cushman, S. A., and K. McGari gal. 2002. Hierarchical, multi-s cale decomposition of speciesenvironment relationships. Landscape Ecology 17 :637-646. Davis, J. H. 1967. General map of natural vegetation of Florid a. Circular S-178. in University of Florida, Institute of Food and Agricultur al Sciences, Gainesville, Florida, USA. Defrene, M., and P. Legendre. 1997. Species assemb lages and indicator species: the need for a flexible asymmetrical appr oach. Ecological Monographs 67 :345-366. Dilustro, J. J., B. S. Collins, L. K. Dunca n, and R. R. Sharitz. 2002. Soil texture, land-use intensity, and vegetation of Fort Benning upl and forest sites. Journal of the Torrey Botanical Society 129 :289-297. Drewa, P. B., and W. J. Platt. 2002a. Comm unity structure along elevation gradients in southeastern longleaf pine savannas. Plant Ecology 160 :61-78. Drewa, P. B., W. J. Platt, and E. B. Moser. 2002b. Fire effects on resp routing of shrubs in southeastern longleaf pine savannas. Ecology 83 :755-767. Fenneman, N. M. 1938. Physiography of easte rn United States. McGraw-Hill Book Co., New York, USA. Fernald, E. A. 1981. Atlas of Florida. Florida St ate University Foundation, Tallahassee, Florida, USA. Florida Department of Environmental Protecti on. 1998. Surficial geology of Florida.. Florida Geographic Data Library, Ga inesville Florida, USA.

PAGE 173

173 Florida Natural Areas Inventory. 1990. Guide to the natural communities of Florida. Florida Department of Natural Resources, Tallahassee Florida, USA. Florida Natural Areas Inventory. 2000a. U npublished element occurrance data. Florida Department of Natural Resources, Tallahassee, Florida, USA. Florida Natural Areas Inventor y. 2000b. Unpublished managed area data for Florida. Florida Department of Natural Resources, Tallahassee Florida, USA. Florida Natural Areas Inventory. 2007. Summar y of Florida conser vation lands. Florida Department of Natural Resources, Tallahassee, Florida, USA. Foster, B. L., and K. L. Gross. 1998. Species ri chness in a successional grassland: effects of nitrogen enrichment and plant litter. Ecology 79 :2593-2602. Frost, C. 2006. History and future of the longleaf pine ecosystem. Pages 297-326 in S. Jose, E. J. Jokela, and D. L. Miller, editors. The longl eaf pine ecosystem: Ecology, silviculture, and restoration. Springer, New York, USA. Frost, C. C. 1993. Four centuries of changing la ndscape patterns in the longleaf pine ecosystem. Pages 17-43 in S. M. Hermann, editor. The Longleaf Pine Ecosystem: Ecology, Restoration, and management, Proceedings, 18t h Tall Timbers Fire Ecology Conference. Tall Timbers Research, Inc., Ta llahassee, Florida, USA. Fule, P. Z., W. W. Covington, and M. M. M oore. 1997. Determining reference conditions for ecosystem management of southwestern ponderosa pine forests. Ecological Applications 7 :895-908. Gilliam, F. S., and W. J. Pla tt. 1998. Effects of long-tern fire exclusion on tree species composition and stand structure in an old-growth Pinus palustris (Longleaf pine) forest. Plant Ecology 0 :1-12. Glitzenstein, J. S., W. J. Platt, and S. D. R. 1995. Effects of fire regime and habitat on tree dynamics in North Florida longleaf pi ne savannas. Ecological Monographs 65 :441-476. Glitzenstein, J. S., D. R. Streng, and D. D. Wade. 2003. Fire frequency effects on longleaf pine (Pinus palustris) vegetation in South Carolin a and Northeast Florida, USA. Natural Areas Journal 23 :22-37. Godfrey, R. K. 1988. Tree, shrubs and woody vines of northern Florida and adjacent Georgia and Alabama. University of Georgia Press, Athens, Georgia, USA. Godfrey, R. K., and J. W. W ooten. 1979. Aquatic and Wetland Plants of Southeastern United States: Monocotyledons. University of Georgia Press, Athens, Georgia, USA. Godfrey, R. K., and J. W. W ooten. 1981. Aquatic and wetland plants of Southeastern United States: Dicotyledons. Univ ersity of Geogia Press, Athens, Georgia USA.

PAGE 174

174 Goebel, P. C., B. J. Palik, L. K. Kirkman, M. B. Drew, L. West, and D. C. Pederson. 2001. Forest ecosystems of a Lower Gulf Coastal Plain landscape: multifactor classification and analysis. Journal of the Torrey Botanical Society 128 :47-75. Graae, B. J., R. H. Okland, P. M. Petersen, K. Jensen, and B. Fritzboger. 2004. Influence of historical, geographical and environmenta l variables on understorey composition and richness in Danish forests. Journal of Vegetation Science 15 :465-474. Grace, J. B., L. Allain, and C. Allen. 2000. Factor s associated with plant species richness in a coastal tall-grass prairie. J ournal of Vegetation Science 11 :443-452. Grace, J. B., and B. H. Pugesek. 1997. A structur al equation model of pl ant species richness and its application to a coastal we tland. The American Naturalist 149 :436-460. Graham, C. H., T. B. Smith, and M. Languy. 2005. Current and historical factors influencing patterns of species richness and turnover of birds in the Gulf of Guinea highlands. Journal of Biogeography 32 :1371-1384. Greenberg, C. H., D. G. Neary, L. D. Harri s, and S. P. Linda. 1995. Vegetation recovery following high-intensity wildfire and silvic ultural treatments in sand pine scrub. American Midland Naturalist 133 :149-163. Griffith, G. E., J. M. Omernik, C. W. Rohm and S. M. Pierson. 1994. Florida regionalization project. U.S. Environmental Protection Ag ency, National Health and Environmental Effects Research Laboratory, Corvallis, Oregon, USA. Grime, J. P. 1979. Plant strategies and vege tation processes. Wiley, Chichester, U.K. Grossman, D. H., D. Faber-Langendoen, A. S. Weakley, M. Anderson, P. Bourgeron, R. Crawford, K. Goodin, S. Landaal, K. Metzler, K. D. Patterson, M. Pyne, M. Reid, and L. Sneddon. 1998. International classification of ecological communities: terrestrial vegetation of the United States. The Natu re Conservancy, Arlington, Virginia, USA. Hammond, E.H. 1964. Classes of land surface form in the 48 states, U.S.A. Annals of the Association of American Geographers 54(1): map supplement. Harper, R. M. 1914. Geography and vegeta tion of North Florida. Pages 163-437 in Florida State Geological Survey. Annual Report. Tallahassee, Florida, USA. Harrington, T. B., and M. B. Edwards. 1999. Unde rstory vegetation, resources availability, and litterfall responses to pine thinning and woody vegetation control in longleaf pine plantations. Canadian Jour nal of Forest Research 29 :1055-1064. Hedman, C. W., S. L. Grace, and S. E. Ki ng. 2000. Vegetation composition and structure of southern coastal plain pine forests: an ecological co mparison. Forest Ecology and Management 134 :233-247.

PAGE 175

175 Heikkinen, R. K., and H. J. B. Birks. 1996. Spatia l and environmental com ponents of variation in the distribution patterns of suba rctic plant species at Kevo, N Finland a case study at the meso-scale level. Ecography 19 : 341-351. Hix, D. M., and J. N. Pearcy. 1997. Forest ecosy stems of the Marietta Unit, Wayne National Forest, southeastern Ohio: multifactor classification and analysis. Canadian Journal of Forest Research 27 :1117-1131. Hodgkins, E.J. 1965. Southeastern forest habi tat regions based on physiography. Agricultural Experiment Station, Auburn University, Fo restry Department Series, No. 2. Auburn, Alabama, USA. Hodgkins, E.J., M.S. Golden and W.F. Miller. 1979. Forest habitat regions and types on a photomorphic-physiographic basi s: A guide to forest site classification in AlabamaMississippi. Southern Coop Series 210. Alabama Agriculture Experiment Station, Auburn, Alabama, USA. Holdridge, L. 1967. Life Zone Ecology. Tropica l Science Center, San Jose, Costa Rica. Hubbell, S. P. 2001. The unified neutral theo ry of biodiversity and biogeography. Princeton University Press, Princeton, New Jersey, USA. Hubbell, S. P., and R. B. Foster. 1986. Biology, ch ance, and history and th e structure of tropical rain forest tree communities. Pages 314-330 in J. Diamond and T. J. Case, editors. Community Ecology. Harper & Row, New York, USA. Hull, J. P. D. 1962. Cretaceous Suwannee st rait, Georgia and Florida. AAPG Bulletin 46 :118122. Huston, M. A. 1979. A general hypothesis of species diversity. American Naturalist 113 :81-101. Hutchinson, G. E. 1957. Concluding remarks. Co ld Spring Harbor Symposia on Quantitative Biology 22 :415-427. James, M. M. 2000. Legumes in loamy soil comm unities of the Carolina Sandhills: their natural distributions and performance of seeds and seedlings along complex ecological gradients. University of North Carolina at Chapel Hill, Chapel Hill, No rth Carolina, USA. James, C. W. 1961. Endemism in Florida. Brittonia 13 :225-244. Kautz, R. S., and J. A. Cox. 2001. Strategic habita ts for biodiversity conservation in Florida. Conservation Biology 15 :55-77. Kenward, M. G. 1987. A method for comparing profile s of repeated measures Applied Statistics 36 :296-308.

PAGE 176

176 Kirkman, L. K., K. L. Coffey, R. J. Mitchell, and E. B. Moser. 2004. Ground cover recovery patterns and life-history traits: implications for restoration obstacles and o pportunities in a species-rich savanna. Journal of Ecology 92 :409-421. Kirkman, L. K., R. J. Mitchell, R. C. Helt on, and M. B. Drew. 2001. Productivity an species richness across an environmental gradient in a fire-dependent ecosystem. American Journal of Botany 88 :2119-2128. Landers, J. L., D. H. V. Lear, and W. D. Boyer. 1995. The longleaf pine forests of the Southeast: requiem or renaissance? Journal of Forestry 93 :39-44. Laughlin, D. C., and S. R. Abella. 2007. Abiotic a nd biotic factors explain independent gradients of plant community compositi on in ponderosa pine fore sts. Ecological Modelling 205 :231-240. Laughlin, D. C., J. D. Bakker, and P. Z. Fule 2005. Understorey plant community structure in lower montane and subalpine forests, Gr and Canyon National Park, USA. Journal of Biogeography 32 :2083-2102. Legendre, P., D. Borcard, and P. R. Peres-Net o. 2005. Analyzing beta diversity: partioning the spatial variation of community composition data. Ecological Monographs 75 : 435-450 Legendre, P., and M. Fortin. 1989. Spatial pa ttern and ecological analysis. Vegetatio 80 :107-138. Legendre, P., and E. D. Gallagher. 2001. Ecologica lly meaningful transformations for ordination of species data. Oecologia 129 :271-280. Legendre, P., and L. Legendre. 1998. Nume rical Ecology, 2 edition. Elsevier Science, Amsterdam. Leps, J., and P. Smilauer. 2003. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, Cambridge, UK. Leps, J., and P. Smilauer. 2007. Subjectively samp led vegetation data: don't throw out the baby with the bath water. Folia Geobotanica 42 :169-178. Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D. Wolfi nger. 1996. SAS System for Mixed Model. SAS Institute Inc., Cary, North Carolina, USA. Littell, R. C., J. Pendergast, and R. Natara jan. 2000. Tutorial in Biostatistics: Modeling covariance structure in the analysis of rep eated measures data. Statistics in Medicine 19 :1793-1819. Lockett, S. H. 1870. Louisiana as it is. Louisiana State Un iversity Press, Baton Rouge, Louisiana, USA.

PAGE 177

177 Martin, W. H., S. G. Boyce, and A. C. Ec hternacht, editors. 1993. Biodiversity of the Southeastern United States: Upland Terrest rial Communities. Wiley, New York, USA. McCune, B., and J. Grace. 2002. Multivariate Analysis of Ecological Communities. MjM Software, Gleneden Beach, Oregon, USA. McCune, B., and M. J. Mefford. 1999. PC-ORD. Multivariate Analysis of Ecological Data version 5.0. MjM Software, Gl eneden Beach, Oregon, USA. McIntyre, S., and S. Lavorel. 1994. How environmen tal and disturbance fact ors influence species composition in temperate Australian gra sslands. Journal of Vegetation Science 5 :373384. Means, D. B. 1996. Longleaf pine forests, going going. Pages 210-229 in M. E. Davis, editor. Eastern Old-growth Forests. Isla nd Press, Washington D.C., USA. Mehlich, A. 1984. Mehlich 3 soil test extracti on modification of Me hlich 2 extractant. Communications in Soil Science and Plant Analysis 15 :1409-1416. Mehlman, D. W. 1992. Effects of fire on plant community composition of North Florida second growth pineland. Bulletin of the Torrey Botanical Club 119 :376-383 Mitchell, R. J., L. K. Kirkman, S. D. Pecot, C. A. Wilson, B. J. Palik, and L. R. Boring. 1999. Patterns and controls of ecosystem function in longeaf pine wi regrass savannas. I. Aboveground net primary productivity. Cana dian Journal of Forest Research 29 :743-751. Myers, R. L. 1990. Scrub and high pine. Pages 150-193 in R. L. Myers and J. J. Ewel, editors. Ecosystems of Florida. University of Ce ntral Florida Press, Orlando, Florida, USA. Mohr, C. 1898. The Timber Pines of the Southern United States. Govern ment Printing Office, Washington D. C., USA. Myers, R. L. 2000. Physical setting. in R. P. Wunderlin and B. F. Hansen, editors. Flora of Florida: Pteridophytes and gymnosperms. The University Press of Florida, Gainesville, Florida, USA. Myers, R. L., and J. J. Ewel, editors. 1990. Ecosystems of Florida. University Press of Florida, Gainesville, Florida, USA. Nekola, J. C., and P. S. White. 1999. The distan ce decay of similarity in biogeography and ecology. Journal of Biogeography 26 :867-878. Noel, J. M., W. J. Platt, and E. B. Moser. 1998. Characteristics of oldand second-growth stand of longleaf pine (Pinus palustris ) in the Gulf coastal region of the U.S.A. Conservation Biology 12 :533-548.

PAGE 178

178 Noss, R. F. 1988. The longleaf pine landscape of the Southeast: almost gone and almost forgotten. Endangered Species UPDATE 5 :1-8. Oesterheld, M. J., J. Loreti, M. Semmartin, and J. M. Paruelo. 1999. Grazing, fire, and climate effects on primary productivity of gra sslands and savannas. Pages 287-306 in L. R. Walker, editor. Ecosystems of Dist urbed Ground. Elsevier, New York, USA. Okland, R. H. 1999. On the variation explained by ordination and constrained ordination axes. Journal of Vegetation Science 10 :131-136. Okland, R. H. 2003. Partitioning the va riation in a plot-by-species data matrix that is related to n sets of explanatory variables. Journal of Vegetation Science 14 :693-700. Okland, R. H., and O. Eilersten. 1994. Canonical correspondence analysis with variation partitioning: some comments and an app lication. Journal of Vegetation Science 5 :117126. Okland, R. H., K. Rydgren, and T. Okland. 2003. Plant species composition of boreal spruce swamp forests: closed doors a nd windows of opportunity. Ecology 84 :1909-1919. Oksanen, J., R. Kindt, P. Legendre, and R. B. O'Hara. 2007. vegan: Community Ecology Package version 1.8-6. Olson, M. S., and W. J. Platt. 1995. Effects of ha bitat and growing season fires on resprouting of shrubs in longleaf pine savannas. Vegetatio 119 :101-118. Omernik, J. M. 1987. Ecoregions of the Cont erminous United States. Map (scale 1:7,500,000). Annals of the Association of American Geographers 77 :118-125. Outcalt, K. W., and R. M. Sheffield. 1996. Th e longleaf pine fores t: trends and current conditions. Resource Bulletin SRS-9, USDA Fore st Service Southern Research Station. Ostertag, T.E. and K.M. Robertson. 2006. A compar ison of native versus old-field vegetation in upland pinelands managed with frequent fire south Georgia, USA. Tall Timbers Fire Ecology Conference Proceedings, 23, in press Palmer, M. A., R. F. Ambrose, and N. L. Poff. 1997. Ecological theory and community restoration ecology. Restoration Ecology 5 :291-300. Palmer, M. W., J. R. Arevalo, M. C. Cobo, and P. G. Earls. 2003. Species richness and soil reaction in a northeastern Oklahom a landscape. Folia Geobotanica 38 :381-389. Partel, M. 2002. Local plant diversity patterns an d evolutionary history at the regional scale. Ecology 83 :2361-2366.

PAGE 179

179 Peet, R. K. 2006. Ecological classification of the longleaf pine w oodlands. Pages 51-94 in S. Jose, E. J. Jokela, and D. L. Miller, editors. The longleaf pi ne ecosystem: Ecology, silviculture, and restorati on. Springer, New York, USA. Peet, R. K., and D. J. Allard. 1993. Longleaf pi ne vegetation of the S outhern Atlantic and Eastern Gulf Coast regions: A preliminar y classification. Tall Timbers Fire Ecology Conference Proceedings 18 :45-81. Peet, R. K., J. D. Fridley, and J. M. Gram ling. 2003. Variation in spec ies richness and species pool size across a pH gradient in forests of the southern Blue Ridge mountains. Folia Geobotanica 38 :391-401. Peet, R. K., and O. L. Loucks. 1977. A gradient an alysis of southern Wisconsin forests. Ecology 58 :485-499. Peet, R. K., T. R. Wentworth, and P. S. White. 1998. A flexible, multipurpose method for recording vegetation compositi on and structure. Castanea 63 :262-274. Penfound, W. T. 1944. Plant distri bution in relation to the geology of Louisiana. The Proceedings of the Louisiana Academy of Sciences 8 :25-34. Penfound, W. T., and A. G. Watkins. 1937. Phytos ociological studies in the pinelands of Southeastern Louisiana. Am erican Midland Naturalist 18 :661-682. Peres-Neto, P. R., P. Legendre, S. Dray, and D. Borcard. 2006. Variation partitioning of species data matrices: estimation and comparison of fractions. Ecology 87 :2614-2625. Platt, W. J. 1999. Southeastern pine savannas. in R. C. Anderson, J. S. Fralish, and J. M. Baskin, editors. Savannas, barrens, and rock outcrop plant communities of North America. Cambridge University Press, Cambridge, UK. Platt, W. J., S. C. Carr, M. Reilly, and J. Fahr. 2006. Pine savanna overstory influences on ground cover biodiversity. A pplied Vegetation Science 9: 37-50. Platt, W. J., G. W. Evans, and M. M. Davis. 1988a. Effects of fire season on flowering forbs and shrubs in longleaf pine forests. Oecologia 76 :353-368. Platt, W. J., G. W. Evans, and S. L. Ra thbun. 1988b. The population dyna mics of a long-lived conifer ( Pinus palustris ). American Naturalist 131 :491-525 Platt, W. J., J. M. Huffman, and M. G. Sloc um. 2006. Fire regimes and trees in Florida dry prairie landscapes. Pages 3-13 in R. Noss, editor. Land of Fire and Water. Proceedings of the Florida Dry Prairie Conference. Universi ty of Central Florida, Painter, DeLeon Springs, Florida, USA. Platt, W. J., and I. M. Weis. 1977. Resource partitioning and competiti on within a guild of fugitive prairie plants. The American Naturalist 111 :479-513.

PAGE 180

180 Provencher, L., B. J. Herring, D. R. Gordon, H. L. Rodgers, K. E. M. Galley, G. W. Tanner, J. L. Hardesty, and L. A. Brennan. 2001. Eff ects of hardwood reduction techniques on longleaf pine sandhill vegetation in Nort hwest Florida. Restoration Ecology 9 :13-27. Provencher, L., B. J. Herring, D. R. Gordon, H. L. Rodgers, G. W. Tanner, L. A. Brennan, and J. L. Hardesty. 2000. Restoration of Northwest Fl orida sandhills through harvest of invasive Pinus clausa Restoration Ecology 8 :175-185. Puri, H. S., and R. O. Vernon. 1964. Summary of the geology of Florida and a guidebook to the classic exposures. Fla. Geol. Surv. Spec. Publ. 5, Tallahassee, Florida, USA. R Development Core Team (2007). R: A language and environment for statistical computing. R foundation for Statistical Computing, Vi enna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. Randazzo, A. F., and D. S. Jones, editors. 1997. The Geology of Florida. University Press of Florida, Gainesville, Florida, USA Rebertus, A. J., G. B. Williamson, and E. B. Moser. 1989. Longleaf pine pyrogenicity and turkey oak mortality in Florida xeric sandhills. Ecology 70 :60-70. Ricklefs, R. E. 1987. Community diversity: rela tive roles of local and regional processes. Science 235 :167-171. Robbins, L. E., and R. L. Myers. 1992. Seasonal effects of prescribed burning in Florida: a review. Miscellaneous Publication Number 8, Tall Timbers Research, Inc., Tallahassee, Florida, USA. SAS Institute Inc. 2000. SAS OnlineDoc, Version 8, Cary, North Carolina, USA. Seaman, G. 1998. A longleaf pine sandhill restoration in northwest Florida. Restoration and Management Notes 16 :46-50. Seastedt, T. R., J. M. Briggs and D. J. Gibs on. 1991. Controls of nitroge n limitation in tallgrass prairie. Oecologia 87 :72-79. Simberloff, D. 1993. Species-area a nd fragmentation effects on old-gr owth forests: prospects for longleaf pine communities. Proc. Ta ll Timbers Fire Ecology Conference 18 :227-263. Sorrie, B. A., and A. S. Weakley. 2006. Cons ervation of the endangered Pinus palustris ecosystem based on Coastal Plain centres of plant endemism. Applied Vegetation Science 9 :59-66. Sorrie, B. A., and A. S. Weakley. 2002. Coastal plain vascular plant endemics: phytogeographic patterns. Castanea 66 :50-82.

PAGE 181

181 Stohlgren, T. J., L. D. Schell, and B. V. He uvel. 1999. How grazing and soil quality affect native and exotic plant diversity in Rocky mount ain grasslands. Ecological Applications 9 :4564. Streng, D. R., J. S. Glitzenstein, and W. J. Pl att. 1993. Evaluation effects of season of burn in longleaf pine forests: a critic al literature review and some results from an ongoing longterm study. Tall Timbers Fire Ecology Conf erence No. 18, Tallahassee, Florida, USA. Svenning, J., and F. Skov. 2005. The re lative roles of environment and history as controls of tree species composition and richness in Europe. Journal of Biogeography 32 :1019-1033. Svenning, J. C., D. A. Kinner, R. F. Stallard, B. M. J. Engelbrecht, and S. J. Wright. 2004. Ecological determinism in plant community st ructure across a tropical forest landscape. Ecology 85 :2526-2538. Swetnam, T. W., C. D. Allen, and J. L. Betancourt. 1999. Applie d historical ecology: using the past to manage for the future. Ecological Applications 9 :1189-1206. Tabachnick, B. G., and L. S. Fidell. 1996. Using multivariate statistics, 3 edition. HarperCollins College Publishers Inc., New York, USA. ter Braak, C. J. F., and P. Smilauer. 2002. CANOCO reference manual and CanoDraw for Windows user's guide: software for Canoni cal Community Ordination (version 4.5). Microcomputer Power, It haca, New York, USA. The Nature Conservancy 2001. East Gulf Coas tal Plain Ecoregional Plan. The Nature Conservancy, Arlinton, Virginia, USA. The Nature Conservancy. 1997. Abita Creek Flat woods Preserve Site C onservation Plan. Unpublished report. The Nature Conservanc y Louisiana Field Office, Baton Rouge, Louisiana, USA. Thornton, P. E., S. W. Running, and M. A. White. 1999. Generating surfaces of daily meteorological variables over large regions of complex terrian. Journal of Hydrology 190 :214-251. Tilman, D. 1996. Biodiversity: populati on versus ecosystem stability. Ecology 77 :350-363. Tilman, D. 1994. Competition and biodiversity in spatially structured habitats. Ecology 75 :2-16. Trahan, L., J.J. Bradley, and L.Morris. 1990. Soil survey of St. Tammany Parish, Louisiana. USDA Soil Conservation Service. Baton Rouge, Louisiana, USA Tuomisto, H., K. Ruokolainen, and M. Yli-Halla 2003. Dispersal, environment, and floristic variation of western Amaz onian forests. Science 299 :241-244.

PAGE 182

182 Turner, C. L., J. M. Blair, R. J. Schartz, and J. C. Neel. 1997. Soil N and plant responses to fire, topography, and supplemental N in tallgrass prairie. Ecology 78 :1832-1843. Underwood, A. J. 1994. On beyond BACI: sampli ng designs that might reliably detect environmental disturbances. Ecological Applications 4 :3-15. Vandvik, V., and H. J. B. Birks. 2002. Parti tioning floristic variance in Norwegian upland grasslands into within-site and between-site components: are there patterns determined by environment or by land-use? Plant Ecology 162 :233-245. VanLear, D. H., W. D. Carroll, P. R. Kapeluc k, and R. Johnson. 2005. History and restoration of the longleaf pine-grassland ecosystem: implicat ions for species at risk. Forest Ecology and Management 211 Varner, J. M., D. R. Gordon, F. E. Putz, and J. K. Hiers. 2005. Restoring fire to long-unburned Pinus palustris ecosystems: novel fire effects an d consequences for long-unburned ecosystems. Restoration Ecology 13 :536-544. Vitousek, P. M. 1982. Nutrient cycling and nut rient use efficiency. American Naturalist 119 :553572. Walker, J., and R. K. Peet. 1983. Composition and species diversity of pine-wiregrass savannas of the Green Swamp, Nort h Carolina. Vegetatio 55 :163-179. Walker, J. L., and A. M. Silletti. 2006. Restor ing the ground layer of longleaf pine ecosystems. Pages 297-326 in S. Jose, E. J. Jokela, and D. L. Miller, editors. Th e longleaf pine ecosystem: Ecology, silviculture, and rest oration. Springer, New York, USA. Wahlenberg, W. G. 1946. Longleaf pine: its us e, ecology, regeneration, protection, growth and management. Charles Lathrop Pack Forestry Foundation, Washington, D.C, USA. Waldrop, T. A., D. L. White, and S. M. J ones. 1992. Fire regimes for pine-grassland communities in the southeastern United States. Forest Ecology and Management 47 :195210. Ward, D., editor. 1979. Rare and endangered biota of Florida. Volume 5. Plants. University Presses of Florida, Gain esville, Florida, USA. Weakley, A. S. 2002. Flora of the Carolinas, Virginia, Georgia and surrounding areas. Unpublised manuscript. University of North Carolina Herbarium, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA. Webb, S. D. 1990. Historical biogeography. Pages 70-102 in R. L. Myers and J. J. Ewel, editors. Ecosystems of Florida. University of Ce ntral Florida Press, Orlando, Florida, USA Weiher, E., S. Forbes, T. Schauwecker, and J. B. Grace. 2004. Multivariate control of plant species richness and community biom ass in blackland prairie. Oikos 106 :151-157.

PAGE 183

183 White, D. L., W. T. A., and J. S. M. 1991. Forty years of prescribed burning on the Santee fire plots: effects on understory vegetation. in S. C. Nodvin and T. A. Waldrop, editors. Fire and the environment: ecological and cultura l perspectives. USDA, Forest Service, Southeastern Forest Experiment Stat ion, Asheville, North Carolina, USA. White, P. S. 1979. Pattern, process, and natural disturbance in vegetati on. Botanical Review:229299. White, P. S., and J. L. Walker. 1997. Approximati ng nature's variation: selecting and using reference information in restor ation ecology. Restoration Ecology 5 :338-349. Whittaker, R. H. 1967. Gradient analys is of vegetation. Biological Reviews 42 :207-264. Whittaker, R. H. 1962. Classification of natural communities. Botanical Review 28 :1-239. Whittaker, R. H. 1956. Vegetation of the Great Smoky Mountains. Ecological Monographs 26 :180. Wilson, C. A., R. J. Mitchell, J. J. Hendricks and L. R. Boring. 1999. Pa tterns and controls of ecosystem function in longleaf pine wire grass savannas. II. Nitrogen dynamics. Canadian Journal of Forest Research 29 :752-760 Wiser, S. K., R. K. Peet, and P. S. White. 1996. High-elevation rock outcrop vegetation of the Southern Appalachian mountains. Journal of Vegetation Science 7 :703-722. Wunderlin, R. P. 1998. Guide to the Vascular Plants of Florida. University Press of Florida, Gainesville, Florida, USA. Wunderlin, R. P., and B. F. Hansen. 2004. Atlas of Florida Vascular Plants (http://www.plantatlas.usf.edu). Institute fo r Systematic Botany, University of South Florida, Tampa, Florida, USA. Wunderlin, R. P., and B. F. Hansen, editors. 200 0. Flora of Florida. The University Press of Florida, Gainesville, Florida, USA. Zobel, M. 1992. Plant-species coexistence the ro le of historical, evol utionary and ecological factors. Oikos 65 :314-320. Zobel, M. 1997. The relative role of species poo ls in determining plant species richness: an alternative explanation of species coexis tence? Trends in Ecology and Evolution 12 :266269.

PAGE 184

184 BIOGRAPHICAL SKETCH Susan Carr was born and raised in Gainesville, Florida. She gra duated from the University of Florida in 1982 with a Bachelor of Science degree in botany. Afte r college, Susan worked in the fields of land management and conservation, including employment with the Florida Natural Areas Inventory, U.S. Forest Service and The Na ture Conservancy. She returned to graduate school in the mid-1990s and earne d a Master of Science in pl ant ecology from Louisiana State University. Following a long period of field data collection in Fl orida and employment with the University of North Carolina in Chapel Hill, Susan joined the Depart ment of Wildlife Ecology and Conservation at the Univer sity of Florida in 2005.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101108_AAAABT INGEST_TIME 2010-11-08T19:23:44Z PACKAGE UFE0021711_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 53363 DFID F20101108_AABNDK ORIGIN DEPOSITOR PATH carr_s_Page_143.pro GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
a116e2865dc1936f874206ecc961142c
SHA-1
e743a67ad807610537077a72a63539e5623196b5
50889 F20101108_AABNCW carr_s_Page_056.pro
0d151709ca9f7b084914010adf16d8a3
8f3a1301384a9a29e4243beb424405628f42d0ea
63554 F20101108_AABMYE carr_s_Page_061.jpg
3a72a6f8859e2af34148ecd24fe54d43
1222a9b104b9fd99be1cf35ec589ed918f75483e
25271604 F20101108_AABMXQ carr_s_Page_072.tif
6a7c6fb0025a28636f86a2b01d3598fc
f690706a08ac4ac044b8a8509ad5b4174c3b8fed
53839 F20101108_AABLUO carr_s_Page_102.jpg
a9e5302d16fb876c02d0d381f51c51b6
3d1679753067f649f0d9d9ef74ddb0abc928002e
1053954 F20101108_AABLTZ carr_s_Page_020.tif
2b18f740b2c8a02df88098b812ca59d8
8f3918bdf5adea7d081726e0fc7b407671738f7a
25269 F20101108_AABMAI carr_s_Page_178.QC.jpg
23f0c67af5fcd1830a5ebf191c83ea37
ab2faab9b33d495a8adf3a8a367e68d4fbac57b8
7087 F20101108_AABLVC carr_s_Page_091thm.jpg
230426c0c453a5871d8f8508b386dd94
a6379ee0d5d5e98ddcf6fcecf8121437bf5f8e28
833 F20101108_AABNEA carr_s_Page_060.txt
7ef4cbc32b9906fdc432ce225d22a382
d4f18c5a08e85ed2bee803a95f37b44704449d65
82365 F20101108_AABNDL carr_s_Page_150.pro
af2588fd299522c0a7d09173ca48db08
1d54cae6b108089c995f600c64dfcb2a3e4c51f4
19301 F20101108_AABNCX carr_s_Page_060.pro
dc351139612de7782b2def84fc01e93e
98f50aa73de73b3a5a88446c5363ad660088204e
63774 F20101108_AABMYF carr_s_Page_062.jpg
b6e2a6798d62d0bca303326a032d0301
b9e17766b35c2e7dd0bb2d4a8721c7264195e93e
76470 F20101108_AABMXR carr_s_Page_035.jpg
e87d7f17592ae2e0030cfd02da3a148f
1973461dfa6bc1fcb2aa7e44d809a62738661d3a
455 F20101108_AABLUP carr_s_Page_136.txt
e42b4361a231637fe43edf4193f91fda
aee6433fcedc3b4a1d8197797228aafcde2e5293
125540 F20101108_AABMAJ carr_s_Page_180.jp2
e0367a811b884116a8dbb3e382a2885d
6e6ac67a7d5ebc7fa9f34406b0c2080127a27b69
24988 F20101108_AABLVD carr_s_Page_111.QC.jpg
5ce002fb7571a1030d2da43cfd4cfdcd
9dad3eb85d74cf08beed5a7654ed06056ceeb17a
939 F20101108_AABNEB carr_s_Page_072.txt
f8e8ec70809345182d5d374f09a632bd
375c8b2624cf90f80211b4662036d75ce7a39610
84472 F20101108_AABNDM carr_s_Page_152.pro
03aed6499e7d1570b05e2a13f2110ada
15453812e0587b3972f88e82c1fef15960c5bdd3
38232 F20101108_AABNCY carr_s_Page_062.pro
ae55863e22c14abc4aeceff70e0f50ae
76f33d64a28f082ebc51eccfb51b1c8ec627bbac
69007 F20101108_AABMYG carr_s_Page_065.jpg
381e7dd2f463e12f6eee0e620e18d7a9
4ce6f5c547f77d8362b5f77507ad92f90be82a8c
209783 F20101108_AABMXS UFE0021711_00001.mets FULL
f84f6466eb788f5b22af0da6033d5cb6
cc5fc82a49fd110a41b5ca05a098fc886098b50b
24888 F20101108_AABLUQ carr_s_Page_021.QC.jpg
f06492ce6835363198f684d0780ef5db
cc3d0c60358abfddc94a539cb54228f2acd29af8
2078 F20101108_AABMAK carr_s_Page_015.txt
88f5fef428287da7b183576c421c6bf4
577eb5c94fd208304e9d280aef4f29aefed13b43
3561 F20101108_AABLVE carr_s_Page_097thm.jpg
7cb4929cf75c08c8390eb564fe74a5cb
c20835dc2c8dd5d2c5ceed9935e04f071498bb4b
1954 F20101108_AABNEC carr_s_Page_078.txt
bc8e8efdfd59d7ac79cb1fd7f327f579
a37c804ad24413ff386c679351a4c37948a128d7
57289 F20101108_AABNDN carr_s_Page_170.pro
6e0a9924cc1142d10f139c915c2b3ff1
44c3a4d836ddf2cec2a76fe513759c8208c2af1c
33857 F20101108_AABNCZ carr_s_Page_064.pro
983e8d4bfa4fa6fbd4be73b520759c89
448bd8aecbf36360283e3693423e32bd5924850f
62794 F20101108_AABMYH carr_s_Page_066.jpg
ed4326d730b5944a8c558d8f64d32359
f0e00587cb83d8378e9fa7c120f6ca7991842e44
74784 F20101108_AABLUR carr_s_Page_072.jpg
565dcd1ff435294c5b415259c1909b31
9aa2be095aedf228c27c6a3dca8411544cac19c9
F20101108_AABMBA carr_s_Page_076.tif
ad47711c89929431d8e4ea1b2df80785
e1dfd59f1431c852459b86556283ac25d50e4e05
15905 F20101108_AABMAL carr_s_Page_071.pro
36f6336b7148d2730e5fb2869e40c38d
44e181bd56a1c5796b4567e4c822932c9d79eec2
73026 F20101108_AABLVF carr_s_Page_052.jpg
d2b61edf412d298d74a7f24f193b0662
457e33a35d223adf2e2a1775c39196fcce99c500
2205 F20101108_AABNED carr_s_Page_089.txt
b4c9b6340de76012a45195bcedc81b70
3b46a63ef2663d76c37b079720e0f72f8a2d83b5
60585 F20101108_AABNDO carr_s_Page_172.pro
b2d418aa2de721fc617ee022261128be
53564c6d2298e30d4e9d816146b009a7278cacdd
72797 F20101108_AABMYI carr_s_Page_068.jpg
0dfec1fbf0eb7d34a4a8b0677136969e
c1f517370155421d7f36369b9fb8fd39dc1ca93b
25504 F20101108_AABLUS carr_s_Page_177.QC.jpg
679a8bc4424e8aa19cedc793187c2ab7
29fc4369439f605b2a7e589500baaff88edcb368
50988 F20101108_AABMBB carr_s_Page_111.pro
d60d31767dc6c335945eab159221b80d
bb7aa070f8e91cf3262c84071cb9eb09e0878016
1051977 F20101108_AABMAM carr_s_Page_025.jp2
e7cde7acdb8d09b2acd0927b9a5a14b8
554e91c274eeb205d24a799ad7900aa32099531f
81533 F20101108_AABLVG carr_s_Page_159.jp2
f871f655a0525d0aa67f1d62b2e4e34e
4d6f28485568cf7838b9a95ca877a5db734d2fe4
67834 F20101108_AABNDP carr_s_Page_173.pro
4353aea6e1234678ac84abea4846e7e2
9a2841f70532659e751ca958b86537cc4d8be348
53930 F20101108_AABMYJ carr_s_Page_070.jpg
61402d90ae4710f98dc15821247bc046
8393a30163d4b8098be30db55e3070c49f890f3c
74060 F20101108_AABMXV carr_s_Page_004.jpg
ed2d74fd5334177853b9afa03c797d77
f1bdb859b471de771743081d6465e16d1aefb644
F20101108_AABLUT carr_s_Page_055.tif
9bc04308a54e37d98b7c1f3938ea2ca9
79bd21770c941fff150592841c6322e004addd5e
F20101108_AABMAN carr_s_Page_172.tif
e061586388d54668f83ba219bfce4a19
1f8ffebfd2dd60799b975daec8e0352ac3befb00
42568 F20101108_AABLVH carr_s_Page_155.pro
f7fb02b0d367c3c19e0c4e30f6e9d740
d73f122c93b79ff8cf152beee532fd8bf3090fef
2043 F20101108_AABNEE carr_s_Page_092.txt
96492866998b6c75f163f1ffadf41353
69cc3ac2f4edfb6c4b594fc551baa8e64411f957
59059 F20101108_AABNDQ carr_s_Page_176.pro
2a0cdb087cef4899097d46a6b79f6f33
7b82e566062efbbb18f3cc7b99d19da15dbd6ebc
50193 F20101108_AABMYK carr_s_Page_071.jpg
9c8e81ca2271e1b9defe5ed4b43c2c35
957b521236f0c83736d75b8d420dd3623b3f6a11
21604 F20101108_AABMXW carr_s_Page_012.jpg
f58ee2c42f2e350e00b35a7726f6af09
e16b47513fc28e4b6f4127d8a9e43e8900774163
533686 F20101108_AABLUU carr_s_Page_138.jp2
68d27693cbe25d5c2ddda2b601544f46
50f8089c085c9e03ce547ff2205ed2e828cc0472
140653 F20101108_AABMBC carr_s_Page_173.jp2
ed032cedec201db0d9f2bf47715f8fa7
d6f204a5611b99474b2bb38abab2bbc90dc3fe7d
2466 F20101108_AABMAO carr_s_Page_180.txt
d88cf263db93c6e375fbc8c9d33b963a
df9244b105d1cc411570461d258c2bd5ed101371
24112 F20101108_AABLVI carr_s_Page_120.QC.jpg
d7dd62a03731eaaa42708adf72e262bb
8bcb38fd6ff4f776cf871c8087c2fefe9087fa56
2241 F20101108_AABNEF carr_s_Page_094.txt
03878a5aa208aea2f5c6bb6d3244c0bd
236962b563be497d56cc3c5c66caa038cfaf2d95
92 F20101108_AABNDR carr_s_Page_002.txt
1d379bc2a0421fe866458e61bba025ab
7aa6ce1ff7cf446a64a3a3171da51e934a9683d4
58693 F20101108_AABMYL carr_s_Page_074.jpg
63c94b3b5a94f31b950edbd33b5ece4b
c998c1d04bf709abb1b810737121e5eda2299454
75632 F20101108_AABMXX carr_s_Page_013.jpg
7bb9a76bd5a9f4fc6519e1c7ff2818a4
4f9cac0aa112d34e1148ff339837de917c2c3fd6
118914 F20101108_AABLUV carr_s_Page_130.jp2
02fad9bd7a1babe5a601a1aa374845d3
7fafe26ac243ec79469c518709e2c7fb193f568c
93185 F20101108_AABMBD carr_s_Page_070.jp2
90e0b32de6129f3560c941565bf4dfe7
71f3a7ae14ebe5156c3c8ebaf45854c79ea44834
25078 F20101108_AABMAP carr_s_Page_175.QC.jpg
84283d6e6d2b20dcd6077cbf40507c0e
3ac79556d51aa22bb6422ca55c73a17f90e934b5
2065 F20101108_AABLVJ carr_s_Page_129.txt
f31ca709accab2683df22e56285712d6
f411049cb6bb9ea646612fd33fbe3ebe8b082765
609 F20101108_AABNEG carr_s_Page_098.txt
ec4c2be0a6aeb3f28ef18c91d41dfa7a
fb0ce504982b59e2dd9e71f02823c3e09eb10cdc
91 F20101108_AABNDS carr_s_Page_003.txt
734fa8ea0ad2098bc69d903013369279
f51cb52b0a9c43dd4023be46b3e382862a5091bd
102187 F20101108_AABMZA carr_s_Page_151.jpg
ebd9d3fcb64c1057fbcff440d07a4174
9af81d927bcfad318e0d59dbfaea080387aeb4c3
77707 F20101108_AABMYM carr_s_Page_076.jpg
27fca70afc6983a0253be0efe23b7342
fd7cffff10d2238d1488153939ab5b5dca25ed3f
41537 F20101108_AABMBE carr_s_Page_145.pro
2a17adbaf9e89841d01b0befc3bef77f
9c6a9ba1cdc5e7f49ac7ddf3d5f8f8481a9e932e
F20101108_AABMAQ carr_s_Page_151.tif
5d310546cdb7a8780b7dc48b57673a45
7f6ee6f68b4574a0331d237dc82f11271bbd9a60
2725 F20101108_AABLVK carr_s_Page_179.txt
ca936055cb6017dc2266ab20050804df
8991a0bd02a902be2a32bbd5b84a7ac9c6d10a48
616 F20101108_AABNEH carr_s_Page_100.txt
0af2ea45a795f30fe6d3788f8de14b92
6b8c95b9b9853e34448f4bff4f374911060dd688
1790 F20101108_AABNDT carr_s_Page_005.txt
3940c6406314e38f61a2b6b882191603
b5e4fc3dbb4ae6837d9c6aa41f382639d22964f1
48386 F20101108_AABMZB carr_s_Page_155.jpg
0880173574da3c8fda783b71c46ef191
42c799e24c8f3fb0729b4c831c71fe494fd541ff
79119 F20101108_AABMYN carr_s_Page_077.jpg
3dd4e8c60d5bb80a9358146963a44588
c73faa3777b9ff34d1b39a00685566588257590c
20623 F20101108_AABMXY carr_s_Page_018.jpg
46ebf61eb62a1ad8707865a136b0d62e
a0397f8c5734b9bb80dc16947a697c979bce26b2
2234 F20101108_AABLUW carr_s_Page_091.txt
9e54988abe475793b66b258a2ef8d070
e179570b0c7efbf1984d35fa73792ac353278891
77328 F20101108_AABMBF carr_s_Page_055.jpg
7c0159917abc552a2972b7cc9256e394
e37a7b01db25c20bb80fbdb83fba228a89b2afdf
7283 F20101108_AABMAR carr_s_Page_179thm.jpg
188978bad30d34634a313ad8ac2d244c
83a19d295538b756ccd3be8aaaf7920011f469ef
F20101108_AABLVL carr_s_Page_144.tif
eab385aaa3b03dd891bdcced551906d8
a52e0d4bfac8e3666444f690509b4c88d4d45244
1876 F20101108_AABNEI carr_s_Page_114.txt
275e482453c4d15cd17e9848b7c2fdd8
3d628bb2b622a00709b238972a4d782e974ec00e
2168 F20101108_AABNDU carr_s_Page_011.txt
c2c909b852b7790698342fb52deff266
1dcdf18b181c74de6d53841bd3570ab1b0a51653
50200 F20101108_AABMZC carr_s_Page_158.jpg
4b5a2209dfc7ae4c818b18d8a0eab58d
fb28a69d9f50fc8e8c44e5b7d494ef178f809a2e
72929 F20101108_AABMYO carr_s_Page_092.jpg
f1c2e583bb90f5b5db68aa80b44481dd
6c007a909a83b4822e36ad2e555471a4ce33a9ad
73828 F20101108_AABMXZ carr_s_Page_028.jpg
da80b2d651450ef811652cff9349d876
14c287af1e450648ae27749c0072c0ac564776ab
2654 F20101108_AABLUX carr_s_Page_001thm.jpg
4757967548c54cb83f48dd508d5eefdd
4ddd7b698a4b888b9af294d729e1b8f37f02f8ea
120089 F20101108_AABMBG carr_s_Page_011.jp2
4622d9e014517a56cbb16b43be8a8c33
b28c31ce741cf06bd24b486e0c4b738f378bf774
1166 F20101108_AABLWA carr_s_Page_003.pro
6f7949113fe36a5fbcb2be543a12336c
dfdd7b1115243861fdb0f4f3d68658c09926e190
65663 F20101108_AABMAS carr_s_Page_134.jp2
a7e789b721dc5b592ddb6b85210a29b9
88254a45f45b037954e7d0b115c2c2f560544b1f
2045 F20101108_AABLVM carr_s_Page_110.txt
571183d1fa8abd000fe84f13e64d32d3
336aee3292ecc912f243f88da072219c14e51e0a
2146 F20101108_AABNEJ carr_s_Page_118.txt
3557207b415e985976b725eff4c90beb
b4b6d01b04db9aa6446f5d488ea100f5baf39d0d
2128 F20101108_AABNDV carr_s_Page_014.txt
33921d929799c5a6532882e66127d685
b61934f1981c8f6fc02ad69e2cc524413397c0f3
50378 F20101108_AABMZD carr_s_Page_159.jpg
ddaf0f4f27707612003169e16517a48b
4b73fe9b4f92308aa1f5c5b14dfc2a83650ea3a1
80123 F20101108_AABMYP carr_s_Page_094.jpg
c6a531e70a665235c73530a828961b45
3fc97244abc1179c37f700658225da7e866ab238
F20101108_AABLUY carr_s_Page_056.tif
8c983b1d9506cb6739aa05e4dc168a1d
f315673cfa271578273b158a2574409894df9d93
F20101108_AABMBH carr_s_Page_145.tif
68ca762a366eda3bc559af6bfc6c8bcf
b121ea1f69ef031b3bca4820b5c4ed3bb0ad5fb4
51703 F20101108_AABLWB carr_s_Page_161.jpg
aa8ec4ed1cee702f5890bb1fad2db029
12e3f97d5a7b56aa69d697b2ab907d6e9c8a7c4e
1271 F20101108_AABMAT carr_s_Page_074.txt
e922fe8ba42e69885571448e25543d58
94203e0966f282583181866f786fc4758b54fcda
6107 F20101108_AABLVN carr_s_Page_010thm.jpg
7339ae92676cfb3c18aef73d1e3fc743
5542036dbebed1b17a6a0c6f36518c4aa7ef6579
1977 F20101108_AABNEK carr_s_Page_120.txt
49137fd6005f5e0cf1ca2f0fa9256c02
b3adb6bb626859c9804b334129488f318365065c
2018 F20101108_AABNDW carr_s_Page_033.txt
d3f38fa8f9dd575c737c2d7b7b176b48
07e41fc3e444a2b1aa754cf33eb5243a1a96834c
80840 F20101108_AABMZE carr_s_Page_170.jpg
1e8ff38362be2d8548fcb60939017d64
7c9210e88db208a14a610f4c90f9d7bb094bd0f7
71377 F20101108_AABMYQ carr_s_Page_096.jpg
67591ca178a01b3a96be1e7e5cfae7d4
ac1a73fd2259cc4e43e95ba8abae7bef7297c8fb
19087 F20101108_AABLUZ carr_s_Page_169.pro
8ce6985ff04037d9c04c60b191c93f87
081d261685432067060f7d5f6721c22410805e0a
7251 F20101108_AABMBI carr_s_Page_150thm.jpg
f58d414241d68f17cbdc50a8741969b4
e5d19c02bf2f53f6d6a07c8e0cb1fbc52573d241
6904 F20101108_AABLWC carr_s_Page_026thm.jpg
a41e98439088858eaacad1eab9c87b63
b44e92dae9da43019236bad7eb9c0fbbaec9fe09
64966 F20101108_AABMAU carr_s_Page_068.pro
07f76682b96b26b99b68d5d4308c63e9
ca46b8abc73b66297fb7be64b814e26c8d5ed46f
113598 F20101108_AABLVO carr_s_Page_028.jp2
ab5b3e8718d79b0d6c1d428576351080
06d03617e51997a10717ea5c6f05a0e8e8a1b704
20576 F20101108_AABNFA carr_s_Page_101.QC.jpg
cbc7b278c24039f03835980377c8841a
7e501ad8279db45b39452779d2a6e4099330fd3b
2064 F20101108_AABNEL carr_s_Page_122.txt
4dd8c6520ee495499cb1dfaf72c8c231
5e9e9d7ed6d2a0179f944c639b719917eb51bd66
2108 F20101108_AABNDX carr_s_Page_042.txt
463e673340315f0d90e46c962d3d5533
da16b462728db8dfe25b62ff592b83e0310ab790
81100 F20101108_AABMZF carr_s_Page_175.jpg
10e000b0ed8a7f9921f43c3e9c744ea6
893c6b4c3650907f5691ef75fa25aeef808c7ea7
38673 F20101108_AABMYR carr_s_Page_097.jpg
0e16d95d538431edffe4424846e21b55
acc822cb872845ccf33c48a9d679ed2a8450d308
6936 F20101108_AABMBJ carr_s_Page_115thm.jpg
6bca9c42a782bcffb00b8ec4d6a654b6
bedb7040300e53cf8d1636a0a5fa6b247386f53a
4814 F20101108_AABLWD carr_s_Page_064thm.jpg
128002c5a3305dc912e9fb8918a37417
4294dd8ff2a81b8b5b40d8b16e2927e11dffa62b
15820 F20101108_AABMAV carr_s_Page_163.QC.jpg
9ab0b5f2ef9477a78f5112dba2f5dbac
bcbb7d390798391d062b6bb0c652f76b5dbb8db1
22407 F20101108_AABLVP carr_s_Page_005.QC.jpg
17b5a91609944eef69b6f5761954a281
b7d165e184415688258812197a0a856c3a633c71
17913 F20101108_AABNFB carr_s_Page_104.QC.jpg
e58d9e16ec1d50add49c3ccc548eb07d
261bc08c34033d64da163f1cb3c3e0de57dd0ab5
1998 F20101108_AABNEM carr_s_Page_124.txt
7c8e356debe8c3857471cf57acfe509a
320ce14dcf1554e5e6102c8eee82e398310ccdfc
2135 F20101108_AABNDY carr_s_Page_055.txt
02bebbc7bfae6166bce161ef18068ced
7ba8941358a25501e0b6a76702b4a43e30adac98
83210 F20101108_AABMZG carr_s_Page_180.jpg
b94deb13a1a17a4e5572d58279136955
0604c36bfd4e9a80c8a3856583e0cd5af8aa680b
17302 F20101108_AABMYS carr_s_Page_098.jpg
992b346deee71e55c464f24ced2fb58f
3eb3f2c1be6ba405d238da1477408d3a4101b553
24373 F20101108_AABMBK carr_s_Page_029.QC.jpg
44d55fa902db3f614502d6982535ec17
6042237c9a56690ce9c475c3329e402f783a6e7d
6559 F20101108_AABLWE carr_s_Page_124thm.jpg
551a009ab683cffb2d8432637fddd6fc
a1e0c14de7185d95e4987a3951ab25cb9c73c71c
1824 F20101108_AABMAW carr_s_Page_041.txt
96668e723341b2a430bbe1b70d7c1267
9ca3e3242c70efc280d26c205435ebf5bd975b55
18340 F20101108_AABLVQ carr_s_Page_006.QC.jpg
68fb1b2a5de2882eab7e65ed02ab4870
efe1e5fd8bf78d1a0fd51d84fa633c79d7589c1a
24856 F20101108_AABNFC carr_s_Page_121.QC.jpg
cccc32c7945bdb15672bbde8f2f237fd
d0af1ba83b22740716d546cc445f4fad6d7acbad
1414 F20101108_AABNEN carr_s_Page_134.txt
6614ff15ccc8ebfcc0cea82489a8e1ca
1bfb96c8b7bc33ff136314d31dc50c49dd4f700c
2013 F20101108_AABNDZ carr_s_Page_056.txt
b9ccdac01792dbd174207622da03f135
7be04a302cd87965c9ffc0c781920ad77f456269
86246 F20101108_AABMZH carr_s_Page_181.jpg
1430e4ea3cc293ba26357d1104324158
80aff3c0e67d48a89e1398d56035725c6c5a2f3b
76056 F20101108_AABMYT carr_s_Page_121.jpg
a78a86c7ecef8292d19960af53227d35
e9fb02b0bc77ca0a61f1f845ac0096be94dcadea
6839 F20101108_AABMBL carr_s_Page_112thm.jpg
66347e4f3b082d13e1f5aaf505608c35
c8bb840e4baa662c7b79b6cee63ca260204a13d3
F20101108_AABLWF carr_s_Page_162.tif
7546fb1bd45d884ad0873e2c8d6edda9
ca81658a40b0c62f6a1972e26a646a87a4efd234
1746 F20101108_AABMAX carr_s_Page_159.txt
f482d7c0a9c5097b6c33c604e33081e7
d15f4ebc125a4dfb0b7ed14739d980812caec5b8
26032 F20101108_AABLVR carr_s_Page_022.QC.jpg
1de3a0837554266684182f328c5033b2
7cc34efc27fbc585bc3fce693cb03f9df36121c3
26040 F20101108_AABMCA carr_s_Page_058.QC.jpg
8564a2362b6386ee9992b046e367c31f
6154a93c13a329648d2f5d100e440f0e61c56635
25057 F20101108_AABNFD carr_s_Page_027.QC.jpg
e1027e8feab4b10cee0ba6cba826578d
12c852b6492a678e43dde52a0ef1c79fff488134
1025 F20101108_AABNEO carr_s_Page_140.txt
5b186505c0caed1de336256cc5e6ddf0
df412a584b55fcdd8596aa220a95606b7c205b0e
93178 F20101108_AABMZI carr_s_Page_182.jpg
c3107105976f51a43c9ad7f3c80f372d
159adc3d3f163399dc249eeee4fc09c40b66e805
75077 F20101108_AABMYU carr_s_Page_128.jpg
c567892ac6ab4dd0c9e0a4f6e1161c8e
cd76e3b6f4e4442bc1abc9a55db00821c7b5157c
78712 F20101108_AABMBM carr_s_Page_057.jpg
87f6a5c9226236f990bcd6e108e0f4b2
0cc5689cd3cf2352905a1f82e18c9a555a570e79
6812 F20101108_AABLWG carr_s_Page_019thm.jpg
e88a5f4206ab8e45c475d97af6472fb6
8b78e9e3d21dbc449d2cc8c22f0b4df37fd96559
66955 F20101108_AABMAY carr_s_Page_179.pro
1a48dd9fa69559bc5e5a4adac78d0f3e
2bd741944e2bf83cddc79953009a1912be694bcd
80547 F20101108_AABLVS carr_s_Page_091.jpg
833a4cf88fbb399a613d476e2494589a
674907dfd5a36a0d9f1bffa4ac50950e012abbbd
2386 F20101108_AABMCB carr_s_Page_012thm.jpg
0149d11b098a69cda66e97f94e4185e0
cf37c0e5d8f3344a87745e1c09ca5003ed4d235c
26371 F20101108_AABNFE carr_s_Page_093.QC.jpg
7545129894200ae1a79a75e8fed7c0a7
1c58c1df6f5940c2332eaac1c968f4db227b95e3
3294 F20101108_AABNEP carr_s_Page_148.txt
b4234014d7df5329a6ad2245585f1b3a
a05b44145b23b4b2232a839e09d7c5e883269142
33823 F20101108_AABMZJ carr_s_Page_184.jpg
040e24dbc80db297587fd77f0881bc1f
e965b6294b03872fb015a4266554eb3c7894cf27
48229 F20101108_AABMYV carr_s_Page_140.jpg
d78b8f0fc20d7e869d3a39b48f33306c
04f3a48397cbcfa01a9781b92c4673b04e3cfe64
1034696 F20101108_AABMBN carr_s_Page_101.jp2
da3e67ba2e0895b43a4c4e9d59879c04
3c6eb219d0ecb5eed0a5a4a90ec87f8d79f487fd
48596 F20101108_AABLWH carr_s_Page_120.pro
05425cababb47d8f6d4189e4aca071f0
8f9fec1d01d5037c24375f0e83c0b5cf9d35895b
F20101108_AABMAZ carr_s_Page_164.tif
d20be903adb7544b65429751371f7312
eff2ea3292303b81ffc3b8b4d66826cee35e23ef
49256 F20101108_AABLVT carr_s_Page_080.pro
884bef41e03f943f2b9b72e4af9f7810
de664e5b909d07c61b30d7877731c4203eede18d
27791 F20101108_AABMCC carr_s_Page_182.QC.jpg
7ab31ff884e74f8ffdb010e958ff09fa
0320ffdc3d3af196fd3dcd7159017ce88d2b740c
2422 F20101108_AABNEQ carr_s_Page_175.txt
afbbf749541c757f0bb575ffe5e868d4
7fc2a8b890f70c24eea06aaca09e101c104a0ad8
97261 F20101108_AABMZK carr_s_Page_005.jp2
1419ccdbffa69254a1d83869a4703632
579bc2d31ac90dfdd8682000e542c1ab2ce969f9
42959 F20101108_AABMYW carr_s_Page_142.jpg
3650900d6319ce67f976ddf2f235cac7
b3776de8ae794f68f585416e4d8362c458252353
78110 F20101108_AABMBO carr_s_Page_088.jpg
49d1dc760e243404a4a3f2063ddcf283
9b450cc110c7c05850d127819b89bd678203105b
76830 F20101108_AABLWI carr_s_Page_051.jpg
00d0fb8d0dc6545a32b2cddf6b3b7d07
78372dd35bb2dd1a3b27c7c4b63cf0064e23bf6d
24137 F20101108_AABLVU carr_s_Page_036.QC.jpg
8707c2f089e23e67999747e4a81750b5
445eafe63a36d7d4de4435fbcebf5f65bf829e2e
6433 F20101108_AABNFF carr_s_Page_041thm.jpg
694e30a3bbf2b36850b4d4bfacf077e2
ae65c37b63067ab74da4abe808fa5b55163d96b3
2502 F20101108_AABNER carr_s_Page_177.txt
16e10a1f3ecbb5d0d0e9fe9643d4a860
de6a75bece7a320480de9ecdfda90b5de5b3f93d
887439 F20101108_AABMZL carr_s_Page_008.jp2
4492731f8d1d8ad4008dfe9b708562d7
c4029742628725276e4521000a3043af46c4ac47
60848 F20101108_AABMYX carr_s_Page_145.jpg
be7263731f60607545458a8bca107774
7f4efa61909ce77e2c4da91fca15fc564ae670cd
13857 F20101108_AABMBP carr_s_Page_140.QC.jpg
e4e38f5e66451449c8d2e8bd05272d55
5ec92e2562c11240f0b8cdb0e93c0af5066feaf1
12131 F20101108_AABLWJ carr_s_Page_097.QC.jpg
093febfc5937fc3199357a3b5d37a957
4a6d980c9185aa99128d72932a730f55b0b1a99f
47684 F20101108_AABLVV carr_s_Page_023.pro
6e93f2f4fa412834108313302a05c5ca
be3ed3a127dc58b13b92e9448a24f611c7830812
51804 F20101108_AABMCD carr_s_Page_024.pro
a2090604f29147a5983772808d2e1c17
7bdbcc2ad8065012e81e696dc0c123c7daa39892
4965 F20101108_AABNFG carr_s_Page_103thm.jpg
fb26ef4ed9d49e3d74eedbcc15ba494d
43ac97414c5eede3bb12a18d9437b2da74bf2ba3
2769 F20101108_AABNES carr_s_Page_182.txt
bb470e13f1b9e45296c79f2d40d6e962
ee03dcb64c8f0bf00678bef4422b4a7da69d5fb7
1051980 F20101108_AABMZM carr_s_Page_009.jp2
8a525dfd1eb7ecea59232e7ba4018027
173cd7cef0058954230214b5fa5a05b0a0f7d98c
108505 F20101108_AABMYY carr_s_Page_146.jpg
a63980e5ad6694bbe31012b5fbfb7f47
55c386e84d1d46a709c78402a378b84e683ed7cb
69767 F20101108_AABMBQ carr_s_Page_080.jpg
80243d983805a44632a9eb9b6186ecec
d29b0235d1ee0cb8257e9ae4565e9935f027a320
107975 F20101108_AABLWK carr_s_Page_048.jp2
a79a234feb6996606dd2fc6a210c8c6a
04b1e4f37842966f04d8450146e5ed02f5acebfc
1997 F20101108_AABLVW carr_s_Page_025.txt
d4df1b03efa686b2f912e94eb87fd673
d1af29753fea363bcd58cd6d1d5499f9b3ff30d4
53272 F20101108_AABMCE carr_s_Page_064.jpg
5cc727466197690c9ff17ca9f83b4108
cf262f47c9b5936d58c0ea684fd1ed191b7bbe91
6597 F20101108_AABNFH carr_s_Page_092thm.jpg
b78f63947bbb9f542c3db5c0680f22c4
c4e2cd88d140dc5ec4e3a1d36e44a41fe4efe00c
2213 F20101108_AABNET carr_s_Page_183.txt
a0465291373d4f06e9116d20c7f85e5b
a4b0f748f07dddb157170a95d75439d61ab28631
114835 F20101108_AABMZN carr_s_Page_015.jp2
cefeb036de55ec6a60f907772d4b8cb3
7e0dbb0141594784822fa88872358bcbca9f0689
119404 F20101108_AABMBR carr_s_Page_089.jp2
0a0bf6b28e3e0b806cb461ceae90d36b
4086e710b3828deee6c32e2f17b3dae992077334
50385 F20101108_AABLWL carr_s_Page_090.pro
82a18dde4cff63a93421bc117d1625b4
ab8e53b5d0a01beb33e9626c0f8e8115a4bfde4b
104237 F20101108_AABMCF carr_s_Page_147.jpg
2cd8ea8fb697d029762a18e428eb2af5
486dbd140986651228bf3b7be56b81144d274729
3352 F20101108_AABNFI carr_s_Page_162thm.jpg
72a608385209cfd275b1cfc57d981f21
4603674f4ec1b0b236a0cf6309615569241aac6b
25518 F20101108_AABNEU carr_s_Page_107.QC.jpg
8eee71f7b29d3f6cacceb5d3b38e13aa
2d63806a0df879019184b1b95f160ff10ae6bcc7
113102 F20101108_AABMZO carr_s_Page_036.jp2
0988b4bd1099af205bed9f312308387b
2811ffc1e0aeb15426288a22fe3c8c06d27a2896
101516 F20101108_AABMYZ carr_s_Page_150.jpg
7759efb7e4ca21eb29155a974bf28848
1d73cfc16b630db49470dbdfd6ac83e40664a3be
F20101108_AABLXA carr_s_Page_143.tif
2414d9d5fdfdd96e76f8afafd4a8025a
f1c566edcdffed6f8c13f54b57adf25e4781f25f
1090 F20101108_AABMBS carr_s_Page_162.txt
558ee563d4b09ba52145b7f7bca02489
d36cb1ec2cfeb77cd5e5638a13245e2f6a8b7f80
F20101108_AABLWM carr_s_Page_059.tif
c58a8b8d44b7941062ad96cc3dcabd99
dcd56d36652ebe82de122b1d8d289bbbb05ed239
73680 F20101108_AABLVX carr_s_Page_113.jpg
b2228cdd749e7de72eb14e9a2f1ad1e0
654fba814bca29d210c2624851720f7483bcc304
6439 F20101108_AABMCG carr_s_Page_096thm.jpg
c0762bfd8d7e94a2c283db58c52448de
522ee992de7cd9ee0053d414b0502f606a241781
4416 F20101108_AABNFJ carr_s_Page_071thm.jpg
a526b1e4772497102c532af76e0214ef
2e349c5d623f3dd49957044e7a90a56614a00881
24107 F20101108_AABNEV carr_s_Page_125.QC.jpg
a9547d37a1d5519ee938bcab496b49c2
f4698e5de4374a8df9c54109e350d63b25cc90e0
112544 F20101108_AABMZP carr_s_Page_038.jp2
69aa7891669415f0d7448ec0e1652180
d5c26ace34d386a6f94c0ecd8d39d5071d9c959d
2052 F20101108_AABLXB carr_s_Page_057.txt
7df3b3f96d85eda0b9e37d190289a163
bd5383289e3755b8baf7300371f13e865cf4321a
F20101108_AABMBT carr_s_Page_101.tif
cbc341adbd7f0f31efbaa97377d07677
1ec02dbdd621f7f29c8ebd19810fdcb3d391b4a8
2094 F20101108_AABLWN carr_s_Page_086.txt
650cc89186adee765886fdec32d8b1fc
74cfaf320cc0155edeb73404c6a51f812c3ce996
2154 F20101108_AABLVY carr_s_Page_026.txt
688bb43a738e4bbd0674512fd104fd05
e84c74cbe4ee5af8624ecf8ee30aaa66b4318147
105801 F20101108_AABMCH carr_s_Page_148.jpg
e59569a0121f2b7bdf070e9b7e5f7072
cbbeaf616c1d89dc1a3b20b8d1363ac0ed44ccd8
25534 F20101108_AABNFK carr_s_Page_130.QC.jpg
b8d31d822b2620eb9d68e64d6641a513
2ce120b0eeaa287ae09dbb523e4662051afa5f7d
25377 F20101108_AABNEW carr_s_Page_057.QC.jpg
97ba6a168baacf6cea23a30b4290c21c
b53f274d849efd32eae91df64f218c1d25a8b1fe
116076 F20101108_AABMZQ carr_s_Page_047.jp2
efabcc578bcd59b49524ea4a5812065f
71e2e9799c5fb6eebb898ec6d1892e28117ac62d
1652 F20101108_AABLXC carr_s_Page_145.txt
676e26f87e1408bcf05dc86ea2ef6997
8b1ee3d12359cbb9754833040479ddf769f708c3
6558 F20101108_AABMBU carr_s_Page_031thm.jpg
f8fcda870997d7c8bd3e9e002f6994dd
d512a16035d371c820ad8f1235c2d04cddd7f526
14231 F20101108_AABLWO carr_s_Page_098.pro
cee45b424a2084d0763fbed5e9107dde
1bf924f62e896c934a68e83d9182fee2727729bb
53799 F20101108_AABLVZ carr_s_Page_183.pro
fb39deea44878ca263d0d71d9c4414fa
6ea5e515745f056c4dc648a913da2fc9a9649d17
F20101108_AABMCI carr_s_Page_039.tif
62498712dbdb6a93724ffecceb8b22b9
eb7ecc9193fff679e4ad002fa7c7dc6ae40098e2
6617 F20101108_AABNGA carr_s_Page_030thm.jpg
44f7221020c00af8f0cc7b4c1a811c1f
ebfb364889a0762c0242b0dba830803907690e69
23007 F20101108_AABNFL carr_s_Page_048.QC.jpg
dbf09887dcfb30093a3e0c667ca9b9f7
f11758893a91d8e4c54a84215043f1010a57ead6
6504 F20101108_AABNEX carr_s_Page_040thm.jpg
c6551ae44995a1b190e098d381bf2dcb
5ff717008dca5455c5d6ee63c801d51bdb5bbffd
117027 F20101108_AABMZR carr_s_Page_053.jp2
220dc7cb818203eee88a62cc262932a6
b9a857320209a42f35e18a909983b0160c589210
3205 F20101108_AABLXD carr_s_Page_151.txt
db20bf223c71a72883168c174bcf7ef0
520ab70e98a4f751300b0b1f33eebd6d8c002a81
2082 F20101108_AABMBV carr_s_Page_027.txt
850072597152c6be3ac2d2afdfa73e82
d738482a72aacee6e8363f9cbafb2086ca7b00f7
F20101108_AABLWP carr_s_Page_152.tif
a8fdefb71dd0370d618975acc8301f62
b583c98f50a08275d0d906f5d654cfffe2f557e9
24830 F20101108_AABMCJ carr_s_Page_082.QC.jpg
6d64bea4eb3ba0f09ca77cb840190e75
fbf81a50bb5e21ab30d67e16fe5a4bba2b7633c7
24408 F20101108_AABNGB carr_s_Page_034.QC.jpg
a051f0e19f3daccb4b779a3f1eb079e7
e305ef6f60783f536b6cd49452300b84665f9a47
12050 F20101108_AABNFM carr_s_Page_138.QC.jpg
076beba7f0925eeeae95aebf1928421e
e513a88a5ecff678fa9eaa1cda7812a761f0115b
6755 F20101108_AABNEY carr_s_Page_017thm.jpg
0d27527103ddde4f0ec2d3ac463307e6
489d6bd25f577ea7ee7ab7dbacf5bf127fd1743b
117545 F20101108_AABMZS carr_s_Page_055.jp2
07322b0eac2bf20b69813c142db361f0
96d30b6194ad14117400d5d46963482f7ecec4d5
55400 F20101108_AABLXE carr_s_Page_058.pro
bdef701cc4ce276991ce975a64cfabe7
ec6b9e6f20d584bfc8d57c97bfa46a5397e79122
120212 F20101108_AABMBW carr_s_Page_042.jp2
2bc88aee26e6766ba370ea976b5edf5b
302d15ee5e147aa2ed317f3219999b2c38789aa2
115548 F20101108_AABLWQ carr_s_Page_128.jp2
84ecc201d9a28fcba354fa441ee50b7b
bf191d99ff66bf8726a88978318ed5a8ae8ed307
1051955 F20101108_AABMCK carr_s_Page_007.jp2
e946707dba611300d87bc7a1d354c11a
b05b3ff4f891379d8e69c71e4ea9597e4cd08daf
6624 F20101108_AABNGC carr_s_Page_036thm.jpg
55c6a4352a020c5d9b655e41fb4e41ec
9fdafdc4ca6087ec3bf30dffddfc2f423f02a015
6865 F20101108_AABNFN carr_s_Page_131thm.jpg
4cb3292bd1fd74284a4af47fbf4a6165
05c051591db9b1b92d80df0c43336ad0b007bebf
24198 F20101108_AABNEZ carr_s_Page_043.QC.jpg
583f9bfd3bfe14ce083d5205e91d0b30
5a681165c53cfd0dcc10576dfd820b27aea18a31
78173 F20101108_AABMZT carr_s_Page_063.jp2
e28774bc37949efddeb4e92638333816
1404c3eb92aea457ff7324dc275e54a6db566bcd
5597 F20101108_AABLXF carr_s_Page_132thm.jpg
02566713d06e7c9b6db666af1a089ece
e0b88a0812130cdd9acf7b008edabf31f1041bd7
87428 F20101108_AABMBX carr_s_Page_161.jp2
0075d907718dc9ea4d82e0907f41a48a
e99f19cae85655aec60133f50d3ac94d7092937c
5452 F20101108_AABLWR carr_s_Page_069thm.jpg
8ee6a3f78ad95ed88bd22c12f800296e
a9edab8828edd9ff355853fee95d2e7f546e5327
7442 F20101108_AABMDA carr_s_Page_148thm.jpg
d4bb4ddf94dbb54cd631aaccbd95f916
22534e39613857fcd038198f088e5c94a1a3e36c
1589 F20101108_AABMCL carr_s_Page_062.txt
7c7d62b3f631b0a20d0f2b0280cd8ec7
8952757e19c4a68ebe278a697c0c3b2e006040c8
25585 F20101108_AABNGD carr_s_Page_042.QC.jpg
3a750f4502ac3791e12c67aa8edb0992
5eddb825b174c2eafcc194f2dec4dcf5e3283031
26271 F20101108_AABNFO carr_s_Page_054.QC.jpg
e3bd59bb4d2150151b42556288cdebf4
0d68721fcdc7a251a0bd87a45b22ac717a3bc51a
131042 F20101108_AABMZU carr_s_Page_068.jp2
af347758b4bbb72f440e6bacd2a92e69
133fd8b6aa07feb586a96df1ac764e4a0766b006
53093 F20101108_AABLXG carr_s_Page_117.pro
1b7872fb6e12e20015ae8096904c0d62
fedbba74e4563d057ab93cfc2560175d3579559c
2175 F20101108_AABMBY carr_s_Page_022.txt
02a270fd4f0e2e3205b08de1d5056fe1
17e7d359172f5d4557eede616a724c37dd71e5e0
124507 F20101108_AABLWS carr_s_Page_123.jp2
f72e65d4d4e0dfdb40de670eec08094f
c2eae73609132cb66988e369b966551b32e490b7
15525 F20101108_AABMDB carr_s_Page_063.QC.jpg
cbcfddbce8d06c901c86c0af2512f3cb
4ed8b89c4603d2a847c0d31c2482903332f97b90
F20101108_AABMCM carr_s_Page_133.tif
c19a7c20d9437b9fd1409ce13c2b9728
b28edac03b33eea02ad1a3dc42dc5f22827c2777
6659 F20101108_AABNGE carr_s_Page_043thm.jpg
59d48fb26b00a03bd3f4b96df9425aac
215bd06eaa1fdd972c75adc319f77afa1ddb095c
1376 F20101108_AABNFP carr_s_Page_003thm.jpg
68775a035220b46a428267171d00223a
951db72c3a53e7920e5ada24e4d220f9792fcf4f
746467 F20101108_AABMZV carr_s_Page_071.jp2
690579321514c5ca53273bcc06c9d207
217c7dfee6ac19f8cdcccd0c728f22de443e33a0
120337 F20101108_AABLXH carr_s_Page_112.jp2
8079da11941b664dacb5c26404d03177
9988b63aec7ad40192cc554233a453f18b7c575a
44904 F20101108_AABMBZ carr_s_Page_097.pro
0dbbdf27e5ee5148fbd94692bae919dd
952de9f372494b1e4b3af39c22fb9fc82a0c1759
6934 F20101108_AABLWT carr_s_Page_181thm.jpg
dfec455d6c65c5c9d128467d26645fa2
4882127403a02822611e3d8e7410a4198f414352
103061 F20101108_AABMDC carr_s_Page_010.jp2
12f47fa707e109945d6df0e53aea43d0
f4951e60747b120d180cf007329dcf91aaf4cc70
12016 F20101108_AABMCN carr_s_Page_100.QC.jpg
bfdc0123c3e4b57b34ba6d1f9d732b22
6722d058ae012d290e5ced60e63e92f19350d717
24797 F20101108_AABNGF carr_s_Page_047.QC.jpg
13c68522b234cfe610f1f5f6b072d9ac
4462e688c6f4a5b8b76597f7fb2eb4e40126fc99
15245 F20101108_AABNFQ carr_s_Page_156.QC.jpg
f2b3ee0feeffa1ad09318c89cbe1d8d9
9ec4b1125fe6cfcc4dad07111e7c2a0003fb5c00
117170 F20101108_AABMZW carr_s_Page_075.jp2
494b1cbe1375b23c1dcc676235413c4d
19e3f7afd6596f7bfc7c303bdd0c27db3a56b822
105109 F20101108_AABLXI carr_s_Page_152.jpg
1d1c8dbe7dfe725e41c054ab2c0f6f6e
cdd83e93b9970439183c3a9c13ce14047b2cd242
25153 F20101108_AABLWU carr_s_Page_030.QC.jpg
eac4ba5e632d25cf440ef22cc76081dc
2cf8e10f8875780db5c5113649efa8467a60fc84
84205 F20101108_AABMDD carr_s_Page_149.pro
607ad8f8de0bacc60dfca2c7342e15c7
e720d9389081673b730c89bee42e65a1faae6da1
4380 F20101108_AABMCO carr_s_Page_161thm.jpg
a82a5e6b132958cfd12e3fa9a6704e6f
58da1ce12854b7de5edc6bf65662b54df9056798
26209 F20101108_AABNFR carr_s_Page_123.QC.jpg
5a5cb66a66bf36bd73cf4c34f22c22eb
d2fbcce01eab401210d886276b9846c87a53ea87
120613 F20101108_AABMZX carr_s_Page_077.jp2
3bd0631cea22ee0d58aa8ce3bfcdc388
27d3271f3da17cb1f328a091b3fe185c6a345c59
119838 F20101108_AABLXJ carr_s_Page_069.jp2
6b9486cf9420c9f6e407fa120514dfec
923d84b077718657d3dde1328a384cff9b016765
F20101108_AABLWV carr_s_Page_139.tif
21a74314a0b402c4b7a7a6ee1077dd0d
70a245a7ac776f4b56e4719881a2db0c963d0e44
85027 F20101108_AABMCP carr_s_Page_148.pro
fac73adc0a52a8ac84e3782468a7e82a
82adb4ac43226351fd5c00ee115834788ff3f16f
25416 F20101108_AABNGG carr_s_Page_049.QC.jpg
949ba722f486b1c7013af9e7e98592eb
fd072e3222847946684695ace56c2994c3664f09
271747 F20101108_AABNFS UFE0021711_00001.xml
f686368cede39dc4b990a4a4752e23fe
d46546094bbd6882f196619d259e63e0a981f310
115887 F20101108_AABMZY carr_s_Page_082.jp2
9a12f78778c7a543ee98b61fb7617a85
f654784f2f7a48501438ac3409de0fceffe5d6e7
5520 F20101108_AABLXK carr_s_Page_098.QC.jpg
32a795d8bc6a9408b417f2d5f942a35d
db84bf98e51525dd3efbed5677d7ef61ebcbd5d4
123411 F20101108_AABLWW carr_s_Page_175.jp2
7394654992a55baa8d69e4e3c2b4271c
9d7e1764221e810ec9a2dac7ebde8da5ffcf2bf7
111313 F20101108_AABMDE carr_s_Page_125.jp2
24468c013e5a4d81ae141c48c94a0034
67ac66f5ffe4461be600f6a73ec7ae5acba90ad7
2482 F20101108_AABMCQ carr_s_Page_172.txt
09293dde53beec7c315ba02c0d788652
8ba5ebbd06388539d8aaeb7b53224a1c46b5bf01
6823 F20101108_AABNGH carr_s_Page_051thm.jpg
aec4e86354b4941be3e0c0c98d91635f
c0bf508163326f01e681666b6c118fcb022e6419
7650 F20101108_AABNFT carr_s_Page_001.QC.jpg
af75e361a37b309fedaadb0c6e85af05
cb77b954cf1a4df14d85a776aea88ce501aae817
114111 F20101108_AABMZZ carr_s_Page_086.jp2
b8589004641a3d574eec5bf24335cb75
17ffa124afb618915920099e528cb7ee9df3e007
49490 F20101108_AABLXL carr_s_Page_160.jpg
6813d14549757506bf2650423ad2b1ab
985182b2181e41ebc349d164e529f3c99da440d6
F20101108_AABLWX carr_s_Page_121.tif
182df22497c0c7e0f11af383bc7a02a6
e783f157a69e8fb91a46fc580387e6f395de2467
54603 F20101108_AABMDF carr_s_Page_118.pro
f0a907e2f8d72ffc17412504ad0c0ce0
8d6d607b16120ddfec7dbb32b98cd720ea83c2d1
F20101108_AABMCR carr_s_Page_074.tif
100349fe803532267d6429bd1fa25c43
d640851fe39ac44ad96674c3665eefa131c69c2c
6493 F20101108_AABNGI carr_s_Page_052thm.jpg
d02ec1aa64821677b0afddbe7ad9a0bc
e58eb74beeb15164e01e2023463f600a34a2725b
18053 F20101108_AABNFU carr_s_Page_009.QC.jpg
ece044cb7630421ae0f20f25c61dbcdf
13b731dfb92188a397eea59ea508325f364325df
3707 F20101108_AABLXM carr_s_Page_141thm.jpg
689d214959de48390feac32cb4d11ca8
2342ff4ac37d8815ea0b4e25a825a81df3a766df
1908 F20101108_AABMDG carr_s_Page_098thm.jpg
b7333ab8a825ed4487848295ec439243
fc88a9acbff11c31353503bcca39ed8ceee1f41a
6896 F20101108_AABLYA carr_s_Page_027thm.jpg
9e2c1abf1b96c4d684656f864ab8b6c9
6f64f12f1f154ade8a9e4bae77108152f6a8c5cf
1927 F20101108_AABMCS carr_s_Page_096.txt
9c6d0f896ce1c8a8830dcfcf6ca8441d
4e31596944cfee6c1c26baf5e488c50053aa5063
24589 F20101108_AABNGJ carr_s_Page_053.QC.jpg
455301707fe08c90fd7cecfafe6af352
d5e6f90921840d22e3fc3a49a6ba41dbbb1738d5
6732 F20101108_AABNFV carr_s_Page_013thm.jpg
f89d23b6db29cdc64a5f584ee69fdad5
21f1f9ec9685c24a1de065c66d4a2b3794d5b69f
89458 F20101108_AABLXN carr_s_Page_179.jpg
2b02d24f22a90274a830ee505df7fe93
59d46330b6df778ff05fb0d1208e3a79e5ce027f
76343 F20101108_AABLWY carr_s_Page_024.jpg
b5a07ddcbda76ed14b2c268c7b569702
a8108c7ceba90feac8660127f9d97f7f87b3e324
75237 F20101108_AABMDH carr_s_Page_030.jpg
44affc1d4ea823d70e2ef6544ffab87f
2647f499cde247ccc7bbfae129613da89475bb1d
6919 F20101108_AABLYB carr_s_Page_021thm.jpg
c10f7ce46cd2e85996c78bce46c2d50e
0bdb335141b2b58ea0da60e183859b5364c377c9
6916 F20101108_AABMCT carr_s_Page_057thm.jpg
f2021f884d7dfb9eb1a94973478e79aa
aa559944df547ca44e0a9cef9fbe1bfac589c151
4266 F20101108_AABNGK carr_s_Page_061thm.jpg
c2e9d5e35fa54070698189b615074015
238f12f468572c99532632b6961d7ae66578f260
6884 F20101108_AABNFW carr_s_Page_014thm.jpg
c5cecb673ab8e36ce81ae4a23e545c33
9868c07db7eb32c5856e9a5347f49cc27fe8fe74
919 F20101108_AABLXO carr_s_Page_067.txt
be470a8d264804be8249e58a19b78889
ef53cab5a2f1f3839210f56c4b352bd524c4b4a9
4392 F20101108_AABLWZ carr_s_Page_159thm.jpg
63298c5bdba1bd053ab396d6f2945948
99506bcc2bb94b8f32b8e0a119f70254dd21a659
F20101108_AABMDI carr_s_Page_122.tif
3d40f0bf5537bd9a7a7039564e48b690
c66bee4d8cde7375c111ffd399ee92b704a0fa98
F20101108_AABLYC carr_s_Page_183.tif
09a8022fb31be217132153ae34918e09
8691e05bc740a41ee9f7508283de15fc4c241eea
2201 F20101108_AABMCU carr_s_Page_115.txt
dc504a25ca37464a0f5d3b82225369df
905a03e879ce5318a5a08520583244154fbe7c61
25113 F20101108_AABNHA carr_s_Page_116.QC.jpg
6e52fc65ccaa98a24b94f656e91579c6
db0308a230e23cb53e639cad8e1277486946cadf
18850 F20101108_AABNGL carr_s_Page_066.QC.jpg
b4bed1e3f61e44c949f1d0b78d73e4d6
9a1b9d002c7debbf866845127b6bca37c999cca5
6725 F20101108_AABNFX carr_s_Page_015thm.jpg
64160391718b6b08bdf347257f98c02e
4954fa1275e8b197bc8efb1675005d2161691bea
107388 F20101108_AABLXP carr_s_Page_149.jpg
e11af6cf246f796ace043a483480140a
83216454781ddf969e6e9163eb8ac6fd1a6a5d0d
40933 F20101108_AABMDJ carr_s_Page_160.pro
8f41666e5b70c8f29db63efbabacb57c
c00949ae7bc6d9204cc1ad52a4148211efd48434
75084 F20101108_AABLYD carr_s_Page_110.jpg
43b4a3481180e254161c40e0b25711dd
26f81490c88ab7d34799368e2e26f21f2865abca
67529 F20101108_AABMCV carr_s_Page_114.jpg
d0718ecc98f4fd11ec791ba80beb8bbc
3305ca67269d8a0586fc4b71f09989ebcc6251fd
6647 F20101108_AABNHB carr_s_Page_120thm.jpg
ee12004d5bd11a61c5a7dfbb2a66a6dc
778becc66b95339f191e62487744913f3f439df6
20705 F20101108_AABNGM carr_s_Page_068.QC.jpg
288b3679f2a44247a82b1e82205f0b7a
5fd1799c66d155df099b5f667d0ca936561484e0
6773 F20101108_AABNFY carr_s_Page_018.QC.jpg
d87d6358552c8d264a4c52ed4142af61
8fe816b3bfa2d5989f07b27a3851e813b607042d
25899 F20101108_AABLXQ carr_s_Page_077.QC.jpg
1006d79cf55b1f3425c0c109839b1a48
54dcdfc26e7dbb360af36be6eab5f44f18d90de1
123036 F20101108_AABMDK carr_s_Page_094.jp2
7de5d8a13da00ef34eebd707d3a6bd72
4d539c3513bcbf91d7be898f466fab8958d931bb
473798 F20101108_AABLYE carr_s_Page_141.jp2
b5f02a4c1228cf75ffc4e44110cfa61b
8b1cd32872695042ef07a5e381df11db893e1caf
83233 F20101108_AABMCW carr_s_Page_176.jpg
e892d96e89d42232e1fd41210240ef1d
1df029d91ff94c4d2482745664315acff015141f
6648 F20101108_AABNHC carr_s_Page_121thm.jpg
0df3f75e415c8d9b43f2207c7ec8eed7
fc805956f2bfe8d59c7a54a0386a0a8b16b309c5
4456 F20101108_AABNGN carr_s_Page_070thm.jpg
0409cf8ce4d26725942162901d250291
a0505beccc799d17e9880d5c8408456ab7a0a87b
25366 F20101108_AABNFZ carr_s_Page_019.QC.jpg
21e4c91839bc0af51e6004f1f21a7d56
b02161ccc7d2e52816fe095ea86c9f12a11a49d8
47321 F20101108_AABMEA carr_s_Page_031.pro
41ddcdfa93982b61b6bc7e383444433f
cc5cfb9ff2a4090f6201a76cd7d9f98bcf28677b
24311 F20101108_AABMDL carr_s_Page_015.QC.jpg
89eb330fa092eae25cf94cffb51b815b
ed5189cc3a30fa95c602cae92f06dfb7d60403a2
22433 F20101108_AABLYF carr_s_Page_010.QC.jpg
07901a8d1f8a0d6ffba1a6aa77d832a3
c81051f946e244e196a86693c9a634f4531ec81f
F20101108_AABMCX carr_s_Page_017.tif
4d249adce94c334955bcc3e1817f8699
48d4b50483cb54529587e93002fd3767a2e23117
25697 F20101108_AABLXR carr_s_Page_172.QC.jpg
c0e179c988080012a78c3e89fe12d438
ea65a0a107b88f1cb82981e205593339be301912
6906 F20101108_AABNHD carr_s_Page_123thm.jpg
4108ce5c4ccf8095a419abf0a04e469c
84f77dd20441a2fb31d68abc347c180a52cc244a
12770 F20101108_AABNGO carr_s_Page_073.QC.jpg
c3c9cd7155a22a01b93cfd730bbfc458
810f43019c154b85d3de3fc2feebbbbd5fd108b7
75089 F20101108_AABMEB carr_s_Page_034.jpg
f49ae536ecd2b105cbf17ffc051c3b6b
5fc3e1336c680c6cbb2c3a782d6489def0e1e974
76946 F20101108_AABMDM carr_s_Page_095.jpg
e9fa06277f24b9e9b93ecd9fd1f05498
feaf38bd4a32424b7b876ee5a451126fc6080ac3
F20101108_AABLYG carr_s_Page_059.pro
91e66d33db9cbd1e5c819ff1b2441698
b5d35f9f327b9fca8c3bc0bc0db52da5614a619f
3063 F20101108_AABMCY carr_s_Page_060thm.jpg
6f22b9fc9e52432f9dd8f257852fe77d
bde464071b0d4ae79d57b88f10e80ad259db8a74
2041 F20101108_AABLXS carr_s_Page_028.txt
02b00af2dc3c7ed506d8ae098a84d97a
5e59cade625ddedd6077719fd441945268cc58d0
19044 F20101108_AABNHE carr_s_Page_133.QC.jpg
c57992a6dd72e0b866578b51a5d3da79
3cd12062e12612e2a8ebebb408ebe6c417ee2839
6619 F20101108_AABNGP carr_s_Page_080thm.jpg
66e419c073491f9880cef6b2c3ab0bf3
e958653fd85794e6ae6c328a04b2b8fdc4225c1c
77434 F20101108_AABMEC carr_s_Page_011.jpg
fdf30ffbdd120a4af632d51b376a718d
6844c87f2e865a0d0dc19c1ee0771f2ee87f0c38
115694 F20101108_AABMDN carr_s_Page_110.jp2
deb3d6c63a4869908c3c87d07990b0fd
d2f9569012c798cc88130235ee7129227e880f51
49244 F20101108_AABLYH carr_s_Page_164.jpg
20477b0b8a9048c0b5e0e0856c8ec2a6
45a1077f183d87a1893dbe8076d6a8124e8f11e6
52290 F20101108_AABMCZ carr_s_Page_057.pro
98b5990b00ca155bf53d24f9f4e57702
d62d2a6ae9c3d6586afeb516a6ff9955a53ce598
77648 F20101108_AABLXT carr_s_Page_130.jpg
26d1b360b860936946ce5f568f8de5af
7c61760a27e8a634dfe946ca0788a2bf1a32a7e7
10768 F20101108_AABNHF carr_s_Page_134.QC.jpg
2ed91fb5eec9c275f8f5f04cf3073aef
72a30f2aa40f95a812c171279c96600a7baaac45
24915 F20101108_AABNGQ carr_s_Page_081.QC.jpg
566f0a7b4dc74403b53449cfc3701c57
fd735c901c45e113f8d3652b30655e07e5733837
117105 F20101108_AABMED carr_s_Page_057.jp2
92aff7ed3ffa155d83d2eb25e909f777
08952e387fc517ab2a5edc7d2cf4e9eac518053a
7487 F20101108_AABMDO carr_s_Page_149thm.jpg
43bad90e9face51c9ce5e179eafdf25e
4a5f936e346d26eb675946c9faa715b86a293863
6788 F20101108_AABLYI carr_s_Page_076thm.jpg
ba9a83efc2a51e51adc2463facfce551
15ee965adefdb443b92e6b7bbe6fac373b4acd06
1514 F20101108_AABLXU carr_s_Page_168.txt
3a7ef4c60d1f61735305f25bf07950cb
2900359c468dc39b8d069714d523bed0e72f04c6
3516 F20101108_AABNHG carr_s_Page_137thm.jpg
61dea169bae0c89b95757520cbc5dae3
afe7444afa0cf76348a132c80cc46ef7f98d0bcb
7055 F20101108_AABNGR carr_s_Page_085thm.jpg
c019c2d6fb14da5969f2c76b645c98d6
04f2a10cb691c0ebb7cc2c7fb671d10ac558f1b2
85005 F20101108_AABMEE carr_s_Page_177.jpg
9f8b311d5c5054cadf6a4183f446b28c
613c714ffffa6333098d49ee0a2898761262b58d
74000 F20101108_AABMDP carr_s_Page_043.jpg
8e3a20388d4bdee63a6c75a9e4baca7a
94a8155692d701b967fccb2f1c7d03e260a71142
75886 F20101108_AABLYJ carr_s_Page_122.jpg
73b112a16f8bef52e3c3f05ad5192055
1e8c2fc97877879a43c734fa84b685c24f72aa46
76920 F20101108_AABLXV carr_s_Page_026.jpg
01236f121e1b6080c207dc257bf91b41
13015b3c16c2fc82cd8824fc9cee536f9a978925
6607 F20101108_AABNGS carr_s_Page_086thm.jpg
8d28877c3ff6df6d469a974d6511d75e
71d86c12f872d6f76ae3f98a7a96c546789c4156
48734 F20101108_AABMDQ carr_s_Page_119.pro
a32624e59eaaa3085abcc89a5a63d1d2
6ee5c1433307b1055e33d01a8dc962f560eab2ac
20303 F20101108_AABLYK carr_s_Page_132.QC.jpg
3bb2fce5e61a206364c3766435f5d7b0
a7b6e7ae2ce298e7cd870f8446dd9e8feb99bc27
50517 F20101108_AABLXW carr_s_Page_124.pro
f1a99e381dc69c5deb1aab449d0acc02
32e1d96524fb2b47e66c901e80fcfa0033a863e4
4228 F20101108_AABNHH carr_s_Page_139thm.jpg
3d201c571356b1f08caa6b6f1a6cc1b7
fbc078ce384cbc15360d14e2495c23fd06fd6a6c
25724 F20101108_AABNGT carr_s_Page_088.QC.jpg
99a92d351298a2de477ab8aff300fda3
12235b69cfe300a6722e0857151b1e0b2958cf2f
F20101108_AABMEF carr_s_Page_093.tif
f39d63603e7e2d9d1588cb022ba7b8fe
930c2b4bd7ddb050084894154662ffff77271e9e
3022 F20101108_AABMDR carr_s_Page_002.QC.jpg
20cd3881dcf301cdd481eded1a273d9f
12285786bb9aa47d01a21a894b9be75b6566b6d3
85093 F20101108_AABLYL carr_s_Page_097.jp2
723d0d103b5163621832424db7e8d149
c4898e485e4f3aefe17a98398b76be428bea746b
5556 F20101108_AABLXX carr_s_Page_065thm.jpg
24a7bf53e627c63af4c9b71a6b378a7d
a63c57f36a58f51e65cf445fa9953e2748a416db
13371 F20101108_AABNHI carr_s_Page_142.QC.jpg
49d3ed60e69cce635334bbcbaa19765e
506ab82821399f0af31b79f5a6576fa1e1b36364
23395 F20101108_AABNGU carr_s_Page_096.QC.jpg
9a65fdd456a35161427e75c45debd5f5
44c0ef0e677eec40ab1d905d7cd6e07dd004d157
F20101108_AABMEG carr_s_Page_112.tif
f8ed0b5f3d979a9ed20da78916a77feb
42dd35b6ff8ee286ba3cd2d006e0bc4d928f9497
43304 F20101108_AABLZA carr_s_Page_157.pro
c8c5c7b90597758d130926c8b7f6c393
8ec2f0a4e8d1704e21802eb2564be31864389a76
28438 F20101108_AABMDS carr_s_Page_150.QC.jpg
4baf0f5b000a613773190c1c50069a6d
1873d67634c149d26ac8a89f831315e58e6270c7
F20101108_AABLYM carr_s_Page_129.tif
2c41cf7f7359d24202fa38cd1b76d185
7d232b5708bb97bb749dd53451fb9caa4a70b17e
34872 F20101108_AABLXY carr_s_Page_008.jpg
73ebcde799cac3780dbe992ba9ca42ce
bcf3ebd1d6b2e83f1c63d7d523c7fb50ad685e66
4115 F20101108_AABNHJ carr_s_Page_142thm.jpg
8b8523f75274c3f54134c916b4edf66d
d013ee0aab628bde4453df0e906ea5e115b97f2f
4185 F20101108_AABNGV carr_s_Page_100thm.jpg
8b4274ce0d94924ecdf1b15524971d50
83c6ab9974e44914252f8f731538c2d2024ea247
4507 F20101108_AABMEH carr_s_Page_158thm.jpg
9289e1ad424949aca90874d1110b5cff
cdf8cdbcf7bf886e4e8a61848b60499b305127ee
113072 F20101108_AABLZB carr_s_Page_033.jp2
4048be76b2b5af85a9c3de92c3bf606a
d196c6d41a588232e7b980c94cd6c7405cdf80b7
5794 F20101108_AABMDT carr_s_Page_068thm.jpg
c91d55aa797d344c167589d0d9f6be1f
133bf61a32b3477ac16e17be0c09434ed2efb912
92974 F20101108_AABLYN carr_s_Page_154.jp2
c95994bef5fee0cb6476b01efbe9868b
b2fd7a07e7259eed6b95f46cff02d7ffa84bd9fb
7307 F20101108_AABNHK carr_s_Page_146thm.jpg
f1f6cebb5ff9c4ac8bae3e5bdadc71c1
f1e924d4d855bb855158ed7055f6db54809cb0d5
5894 F20101108_AABNGW carr_s_Page_101thm.jpg
f13a0eb3efb0d4c20be346de9775613d
9721383c2a193c238da4ab3c1e5f77e5b9395ac5
29442 F20101108_AABMEI carr_s_Page_149.QC.jpg
c3c0d10754b434e2b5f5aa746a29df9b
36b20059117d5ccc37b7d9715c00d72f22044136
6974 F20101108_AABLZC carr_s_Page_025thm.jpg
675e4e00d1fec47ebe01c6656fa679c3
20f99b17a919b9c61a5d1e82807578128e0eb698
55935 F20101108_AABMDU carr_s_Page_123.pro
3ed2c660b709ab35a9db1772ca036020
1b21ecf264d8eeb9604eeaa6ccd4c17728f0fdee
92260 F20101108_AABLYO carr_s_Page_145.jp2
b605d0501192e63deeb1efd571f014e8
414ba3cdf6b191e049f689c69355504a370f777c
39544 F20101108_AABLXZ carr_s_Page_138.jpg
241adb7dd24a655a5dacd6cd1ac4825f
7fc9560e76e017e80c0b35686e91c235d0a446ba
7341 F20101108_AABNHL carr_s_Page_147thm.jpg
709e455b1dc4bb6970d6f1e0efb902aa
3aab46bae72e0ad5bde1ae34f50e7521ff974a68
7021 F20101108_AABNGX carr_s_Page_107thm.jpg
8edbb951212f246e01c63ca87bea6e50
a4e147cbe5c77a95e30af3825ed4ed06bde19d4e
F20101108_AABMEJ carr_s_Page_106.tif
960228dd28ab82080800cd465f073bac
f74c348e685d33d325d11e9f7e4b79267d5d3a8a
587157 F20101108_AABLZD carr_s_Page_142.jp2
51e80e857ff43d27dc12a831e1b773ac
720a514b7319f6fde735f09e1aaf8a834cf4c6c0
74698 F20101108_AABMDV carr_s_Page_056.jpg
9565f227dda852bf569dc282d4acc148
743695df6ec91b5bd912ad2bb1a71a240843e646
F20101108_AABLYP carr_s_Page_177.tif
772bc8e3d6c531b12495955ba87cb4b1
b23262c2de3d469c916f641739904245ebeb2421
28942 F20101108_AABNHM carr_s_Page_152.QC.jpg
68bb891d91a316560e794b346ed39aa0
27aa428931923a0550013ad66a0c335e6dfa7e98
6521 F20101108_AABNGY carr_s_Page_108thm.jpg
50095e42401c98c89e52f80df9db1561
7c87b35583180174c2ec0fb48c715797fd4efd8f
2033 F20101108_AABMEK carr_s_Page_109.txt
746668650683362f186121793cb1afa2
b1ce96e2a5a83362c76775c5ac22b6f6e719dba1
1051959 F20101108_AABLZE carr_s_Page_081.jp2
3c870a33b4cce5449980c54dc6b26832
c6708cce9103a03301455841e162531b5902157f
6989 F20101108_AABMDW carr_s_Page_093thm.jpg
f4b9b5a33a6c064621f86d95f292fd60
f0e822dfb85e86c88b10457f23455ee6c61b03e8
884 F20101108_AABLYQ carr_s_Page_103.txt
37152c895519a0e4ff7d3d115803d8bb
67100cadc6fc02174a67a5b8e17262da8fbe4350
3560 F20101108_AABNHN carr_s_Page_153thm.jpg
79c93ff1539ba86f14e0b0bb836579b7
032f404ef419b0cf0a4c4297d8905812bb03f52d
23912 F20101108_AABNGZ carr_s_Page_109.QC.jpg
802e5975577ff5cc0a9cbd3730f4febf
bd2868c6dee59083ac48f327a2e84eab3eabbd75
1961 F20101108_AABMEL carr_s_Page_079.txt
e9fa304ff9564ff26f22566dca305459
c247231b642dc3509e6c24e1dfedb45c8febe325
F20101108_AABLZF carr_s_Page_149.tif
508c8b3571ffc9d4a183e2c150300419
b359f69bc3717ff5a47ce40e405b3bb83cd72bd2
F20101108_AABMDX carr_s_Page_123.tif
84d7445c239bc44c0695d7db49587bca
97522f173fcf78e0ffa3191d54cf285c31edd23a
24044 F20101108_AABLYR carr_s_Page_124.QC.jpg
f8d6eb2a79095fb47be8eb8b364c5dc8
18d12c4f9e83b6123b6193934a1d6b77522fdd4c
24944 F20101108_AABMFA carr_s_Page_035.QC.jpg
b1f4e6a99717a3c97931e6e2cd0a4486
1d006b9c6fa9caf6298259e4f9c9d6ecc2cc9e6b
13504 F20101108_AABNHO carr_s_Page_165.QC.jpg
edc69c247ef9bfe49a3cabac2e0b9208
7d2a940ccb29eec046c4fd1b46e8b1ee61b09e0b
2021 F20101108_AABMEM carr_s_Page_097.txt
5a5781201b6828a0e40f5dc9e8052583
b3a0ec2d379199d2e2a780b9ed78f3ad8e5c5507
6950 F20101108_AABLZG carr_s_Page_127thm.jpg
3542e04a8cb0f38ebcd4e431381bc8fc
72a9d86d4953a78a34b60dcda583b08d79b52bf4
2191 F20101108_AABMDY carr_s_Page_123.txt
796cbae1a8593613e75872a38726efc5
77e94770ff761325aa43ac89dd051ccea13597c4
69479 F20101108_AABLYS carr_s_Page_040.jpg
603d56d304e703b8635f1f7b6695ffc9
15328ba5d238939380a24866d036bc9e1413a6bc
69569 F20101108_AABMFB carr_s_Page_078.jpg
295d1a431e3ac19c48ade803acaeda99
a0e1146819ffd0a1dfe46b444f911f4fa903e6f3
14484 F20101108_AABNHP carr_s_Page_166.QC.jpg
2c8e83cded601fbcd49019231a35fc4c
8e8ac4b5a37f0d092f8795baa45fefa285742833
2149 F20101108_AABMEN carr_s_Page_112.txt
c40f96bdf1e3bdbf038a3962ab57a3f1
42b76ba6d1c15049ee5dd91b1a58c46bca06d372
69123 F20101108_AABLZH carr_s_Page_041.jpg
6d33b062654a8c877c5daba0bd9d39f6
e20f31f61697f5477d5abeda52e5cbec7a1a089e
25626 F20101108_AABMDZ carr_s_Page_011.QC.jpg
2fc24b7957532da409f50517def25e3d
cbc739ac80b0f681f84059ca1196356467f39ea7
1275 F20101108_AABLYT carr_s_Page_063.txt
1bd8fafed68e2b141fedf93d7c0ebeae
d5f4a3654992d97024c93d9af48452156a0cbce5
F20101108_AABMFC carr_s_Page_045.tif
6cda6ebc9053404c50f9dc8860ff4457
c5b9edb00e6aed85483ba43a5219375dcdf2996d
14171 F20101108_AABNHQ carr_s_Page_168.QC.jpg
1d1e942f28e014fef4d506403819d5c5
74b3f53d746b669a0324c8b8f01f757cdda1f376
39392 F20101108_AABLZI carr_s_Page_099.jpg
3a1c8456752b2074a23559045dcbcdd0
dc5a521fab4f182e47aae572264ab8c59662248c
6955 F20101108_AABLYU carr_s_Page_144thm.jpg
22b9d601d8aecd4234cee1bf027f0294
033fed221691d2a8a9bd8c26587f26ad60f66cd1
115179 F20101108_AABMFD carr_s_Page_017.jp2
6561287e88af16cdcad80bbb7f567ebe
4554ad4d1ec08b9e4094d49c5c5f15619e0d4588
25243 F20101108_AABMEO carr_s_Page_144.QC.jpg
b14f9be12bf34c282e1528aebf28d2d0
ceb018e30f117fcc03054c78d98bac188c82dc4a
23715 F20101108_AABNHR carr_s_Page_170.QC.jpg
6cc3220e6bc96847aaac94452b5e59d7
64e0cc5c944fd5f7b0ac06c6621bf773f2918c44
26108 F20101108_AABLZJ carr_s_Page_127.QC.jpg
d3bc3d5e323ea9851ee3b7a381694ac2
c2c91bc89175680b4183356dc8a1b7bb4a48d7a9
5239 F20101108_AABLYV carr_s_Page_066thm.jpg
fef58b65be2ea1a0c8f76495903f54e1
2a510eb6bb3abf56b903773bf6c72934dcead44f
113039 F20101108_AABMFE carr_s_Page_183.jp2
05d547a3d716b6f75ec355c7f9401ff2
22cb816811f485c81e375ffc92f564923381442d
F20101108_AABMEP carr_s_Page_166.tif
320f0fc6e308fb50e712399a4bfc243a
536890a66a2011b3f838cff8c8b40f6189a3283a
F20101108_AABNHS carr_s_Page_170thm.jpg
248a384cc3109c557153fb7188ab1040
945335c649d23d8a65bccd007c2b9962daa93166
2119 F20101108_AABLZK carr_s_Page_116.txt
98f8503e0aa3344ae27cf8cd51350eb7
714e13b3349a9e5f0e5f3e40ed547f410b1586b3
122302 F20101108_AABLYW carr_s_Page_085.jp2
b957cdeaa35bd96dd0003f423c8f9f7b
fad6ea4cef80982c284e81c44140c855bcf973bc
6818 F20101108_AABMFF carr_s_Page_083thm.jpg
df46ff28f9786546064c13acaa86ab8f
6cb10bf44e3254184b6be6ea0e3c1a4341a4f563
1950 F20101108_AABMEQ carr_s_Page_038.txt
71a4a8158e0848dd53d7876d015d7153
9425b80887227edd3725c6c395046e3caa643391
7321 F20101108_AABNHT carr_s_Page_171thm.jpg
c447d0681bc06b6f9af51d59eae81acc
59b9dfcb8d27ccbf8cc1cc97dd2fc64185620fe3
56122 F20101108_AABLZL carr_s_Page_089.pro
f40de0249a5c269255528d2737cba1be
f8cf44602c0322f189aba217ac28a0df9cf341e1
72901 F20101108_AABLYX carr_s_Page_038.jpg
5ca4a08f92c4ac4d8dd600dc512db1ca
91d4bd8dc7da491523376721e9ea2ae1cb9e955d
28050 F20101108_AABMER carr_s_Page_134.pro
7e8eaf3b241753a4d8e416100733c4e5
299353531f5800579f54ffc8c8e390b11b0065fb
7305 F20101108_AABNHU carr_s_Page_182thm.jpg
6b6ea34209f9e6407aa106276ca97e11
727ccfad3b30fec0d9d8e21ea51d7e9b289c3583
23854 F20101108_AABLZM carr_s_Page_050.QC.jpg
938e5d6a4e94b18b0f90e8673ce958b8
3ea5ebc95aa81ad1e19d4ed782d2ee0978887a04
F20101108_AABLYY carr_s_Page_035.tif
f6690cd198e65910c0ba090c55187a15
b6228e614a5aff1dc03fc870fd2468e5370e6c66
F20101108_AABMFG carr_s_Page_070.tif
779e3517652b7efdb617e822175cfeb5
5f4a4ca6780d68653f776dc45bcaa8294c883dd1
F20101108_AABMES carr_s_Page_176.tif
981f2bfb03d972851fc5037b1cf27151
1a60887d72efe0c3379b024ff46692129ca54a83
23174 F20101108_AABNHV carr_s_Page_183.QC.jpg
b79e1b4c2eec31cb863be024af77f882
b4bd4c6cdef25bc0391635156850cba42ae75017
77575 F20101108_AABLZN carr_s_Page_020.jpg
e5775a83d1eab3ac737e348b05372404
faf308e3c55be2d93d31644be4be393b29598b29
F20101108_AABLYZ carr_s_Page_042.tif
5c8757e20b4cb8ffa865925389933065
07941e22cc6d3b3a742e8022d77440b5eb71c6f0
45739 F20101108_AABMFH carr_s_Page_041.pro
aa4bfcc35fd46c9206d2f767574de982
76287ce64574818758e85803ff4506d46e4aa67d
F20101108_AABMET carr_s_Page_054.tif
b97bb4a33d5044fb53d9ef42ebfaef6b
7af644250cd171edfa5ca7213640f151f2dfc963
3367 F20101108_AABNHW carr_s_Page_184thm.jpg
1cb56ca8d5f45c2801f39d9a67df55dc
b62ca31d7652dad768d7efbe0894a5a16ed9ea9f
11829 F20101108_AABLZO carr_s_Page_099.QC.jpg
e5d6ef745d2d7e3abec9e3762382bd3f
f0f5fcba330583ff27fd4a28489a46edae3eadaf
43326 F20101108_AABMFI carr_s_Page_007.jpg
9b797d61dd6f23901fd114d509fb7e9b
5b7cdc59bda9cab8cfebb84cbd80645cb37673f0
48945 F20101108_AABMEU carr_s_Page_052.pro
7390a748caf4cc258f2f07cc898a1e39
532956f10dab51a2f7bce58b8058953c75dd50e3
121986 F20101108_AABLZP carr_s_Page_115.jp2
6d71ae74b71444711de1a33e94e6d47f
1729651de13e3aad0d4905dece27e255a62b7cbd
10476 F20101108_AABMFJ carr_s_Page_008.QC.jpg
2c017a94c2bde67b288cfc0e3f682e9f
31a8a60f66174fd3b4a4c5f5db05dc6851eb406f
71879 F20101108_AABMEV carr_s_Page_045.jpg
a4744ef6b6295f4a13480deaf650b491
8a230e9872ee9258f79b706b06e819536a6c975d
7099 F20101108_AABLZQ carr_s_Page_075thm.jpg
e96e28b8cacb2eb346753ad3d105463e
ca2b91f420b54c0186a093b5d4599e8f7848b42b
6565 F20101108_AABMFK carr_s_Page_033thm.jpg
325981425a987d268273376368061a13
94df499fdca1c0999daf6b88f4f2eddeb24a50c9
25265604 F20101108_AABMEW carr_s_Page_135.tif
c4d0b15f201b988558c6a006ef541e90
0a7c9ce0d0baf76e80190195afb33cff3927cdfc
25257 F20101108_AABLZR carr_s_Page_117.QC.jpg
4ac4693102e98d1371521685607b7003
c7f288ac6f921a9f695a94f23b593781aa3a0495
2088 F20101108_AABMGA carr_s_Page_017.txt
f3bcbf58c4eed37f8a564dcc1bd9da15
b4e8e38fd499f72d16bdb848650882289d74baed
6671 F20101108_AABMFL carr_s_Page_079thm.jpg
70bfba2d4c31f913a57ae4de86b30217
01bacd8391e5746680b9deeb4a5e6837cf783013
533835 F20101108_AABMEX carr_s_Page_100.jp2
5727269d5550b4cd807dbe67151428c9
59da9e8eacaf7be886138a4a65f6dcaf95c94d76
92692 F20101108_AABLZS carr_s_Page_059.jp2
d8bf3921da5a25ef5eedb32a23009232
983b986a2a82b56143f26d6b0cd290fb63046c5e
7025 F20101108_AABMGB carr_s_Page_176thm.jpg
bbda19080405cfba9cb7683a25a96d28
0b77ee42b32faf9249fda38009be20846cec8148
F20101108_AABMFM carr_s_Page_126.tif
e17631d70a9dd99361f177854facdeff
bfb519d918b009db61b03cf56a9f4681b8acc101
14170 F20101108_AABMEY carr_s_Page_167.QC.jpg
b8e6d540bfdc45e689e2a9414a32ddd3
ee7132a0a74ce24d6df213e6bfb82c7a42b8ec38
4493 F20101108_AABLZT carr_s_Page_166thm.jpg
d37925aa9723ef57c7b51a657edca545
df14d35d39b0272cdcd2f00b6ab9951f3d8fd615
1991 F20101108_AABMGC carr_s_Page_010.txt
7632284fa4a48f503f0a6e86cf5166d5
a0b1d3892bb40e8632b4539d9a1a2a2f709c140e
F20101108_AABMFN carr_s_Page_110.tif
8fafdca155f650563289ba52abfec3ca
b9dd09a24ed2d9deb0d722c71e4707d4bc91d5ab
F20101108_AABMEZ carr_s_Page_049.txt
c8d16979810f9f4c7f612fbc9b137620
1a818eeb7757b1147058522ea6e40c10a5379936
108328 F20101108_AABLZU carr_s_Page_045.jp2
271c09aa93b48b174442d07580c484ea
2c9fddeb7b06622d593da72cfb192b6433138301
7038 F20101108_AABMGD carr_s_Page_106thm.jpg
a4342b34d368d1e5cd45e50c05d84e7a
ab1ad82d114feecc305aec90db885afde7c546f4
19278 F20101108_AABMFO carr_s_Page_184.pro
e0c1260af6e0822a410d657ef5143e99
0db1f1d109526fcaf48004979c7878c69961d36b
F20101108_AABLZV carr_s_Page_099.tif
663fbd70896aae659b23f49f416aea4b
c842e1805a68f808871dcaa0d3639dfe295becd3
2111 F20101108_AABMGE carr_s_Page_021.txt
2c7aece7c0ccd8a4f70d970dbc0a8a0b
f7e3576defbe1c09e9baa95d15c9c0ed133554d0
77720 F20101108_AABMFP carr_s_Page_075.jpg
730ba37306da7413d0b66259e9f6c576
d60bc648d18bd64bb7f3ab88579b047cd397e4dc
113388 F20101108_AABLZW carr_s_Page_024.jp2
5c1d933c9a89dca1717c0a9d3cbe884d
1d59c117fc14e71b244217b6c7c0035fd051468a
5466 F20101108_AABMGF carr_s_Page_059thm.jpg
1363f04b32cb594a335871d099fdca6c
8cffa74279793195d0409ad71d748643de861d93
75533 F20101108_AABMFQ carr_s_Page_053.jpg
a2743b5d354fb8ec22e6018fd6a38fd1
03fc1d3bfa9169ef37118f30570afc392d19000d
55539 F20101108_AABLZX carr_s_Page_022.pro
a8b4950769149dc6a83974c12cbd2c46
b4f53940d02c1de82892ffe8e712e9f74824ab82
107284 F20101108_AABMGG carr_s_Page_078.jp2
ef2e280236b15ef40f83b4ceca8647c6
9ef3cc4ed5412946a4cd5e4eedf4f2b92a8efeb5
52485 F20101108_AABMFR carr_s_Page_128.pro
061ca13b079c5a686a295a792ebae0c8
86ba8a7361856265a28c572590a7afa993cb0032
22967 F20101108_AABLZY carr_s_Page_078.QC.jpg
f2c6201325eaea0336478f122fcae209
eb9b57f078347ee6425bd7085a52fe95e6e0ba51
73264 F20101108_AABMFS carr_s_Page_108.jpg
f89d1b29c059fd318db869b4ea2b98a4
d8a7c8fd747f1c12d7ef4f8d44cc80a293cd8e6d
75178 F20101108_AABLZZ carr_s_Page_084.jpg
155bf2c673a2b33f71b304dd1ee8157d
353b5c0b26f87c34b748841894b10ea0dd4c51d8
93600 F20101108_AABMGH carr_s_Page_173.jpg
e65fc5d84f41c74806fe7e2e5c58bd7b
066b6803cc1fe2ab6e0dbaa0a1f15292f3a1e9f5
113405 F20101108_AABMFT carr_s_Page_056.jp2
8b59a28caefcfee1f2d2b0ac614a7116
9c4b7db5eb52d41fe6f0cc30161752d0b40a9a08
118800 F20101108_AABMGI carr_s_Page_020.jp2
0b3e9a8021fc7c15a63ac9b4fe26350e
fc29243cb58d6bac40bb1312f42bb6b5dfae50a4
1545 F20101108_AABMFU carr_s_Page_164.txt
e9f2cdb3caff4daee6c329ec45c1816e
778574373a287f0aaf4057cd1bf447263dd3353d
F20101108_AABMGJ carr_s_Page_043.txt
7984c0e4d593d85978b9bf3159034930
e5d9ba6c8bdc3e12e3cfbe70e63daf91f346a3d4
1922 F20101108_AABMFV carr_s_Page_046.txt
b013dcf21f18d82e6cae2e277fea0b48
010472ae72ecc85ab20f746421f346e012b68c7e
7061 F20101108_AABMGK carr_s_Page_177thm.jpg
f6f2f431043dd3cb427f48010191be4b
12df4c6c5eb6de3ff97768667ff0523790176b3d
54503 F20101108_AABMFW carr_s_Page_163.jpg
1d3bb92c17c7d1a145616a4a26d4ffbd
12616f0eaa63eb85d0b2ae8653006df678674208
F20101108_AABMGL carr_s_Page_174.tif
3ec0991724939521da8c13738854f3f8
ae24d496e69840a730dbf67b18858b695792e6dd
118944 F20101108_AABMFX carr_s_Page_144.jp2
d2bc59e0e31b9b0615e0dbd6043cf56e
106dfe78957bed09227e2f6ce25083e9ffb20de3
F20101108_AABMHA carr_s_Page_127.tif
2a51acd596ea94c9c45cadd00720e459
c14530dbc7ec9acb3b85bd21bd69a54d0c735d8a
74891 F20101108_AABMGM carr_s_Page_027.jpg
e799d1c718fc7394dcbd2bb84b163a61
86e088e549e8bf9e7d635431520cc35140f38122
4747 F20101108_AABMFY carr_s_Page_154thm.jpg
8ab4855786ff4c25bf5c027a4bd1f6d3
c10595a2779c97e7d3dd6a29b5461f07bf442025
F20101108_AABMHB carr_s_Page_107.tif
6233a50e166677a3a293762ee613720c
3803dd56ce0e43dc7398bca8fb9a6764aae1249e
6822 F20101108_AABMGN carr_s_Page_024thm.jpg
a5735e2b880a81254557d13d66272165
1de1bceb68291b718ce94b16617768bb2cd373b2
1176 F20101108_AABMFZ carr_s_Page_153.txt
be6dbff64b36ae42d9cac832cc3a6a69
dd953df9d1a9fd5d2dcface9a09b82b9faf30c7a
24497 F20101108_AABMHC carr_s_Page_004.QC.jpg
0a4f7cc973cd1bb24f5189d8fe6082e0
31431fa712b338ea6c42099c4bea6e856be645e2
55298 F20101108_AABMGO carr_s_Page_011.pro
04d447e433a263d0a440f2a8fb3984f0
f065d55ed05f2f1f5795603c04eaf6fb90286493
77393 F20101108_AABMHD carr_s_Page_014.jpg
1da75901036ec8c123315af0f1917322
68583189470de90dbb6f3ce378c5e1ac006ccc16
661 F20101108_AABMGP carr_s_Page_104.txt
6889acdfad0de3599afdf8ab131cc5e1
e55412a0f1b6ba9e67476b99776ab1c50222b2d3
F20101108_AABMHE carr_s_Page_084.tif
e4e8af3df0e864cc0bf13c9086c31812
c03bbbdc1d25939ebbfdeb02169f273df7bd1aa0
15888 F20101108_AABMGQ carr_s_Page_102.QC.jpg
988f9b0479f43a9f1391e306d204b8be
81ba16af92c6b94fc4c464d4ab665f78aff73d1a
F20101108_AABMHF carr_s_Page_128.txt
02364562e42c295fc0ab51fd0aa544bb
2a9cf783e90e86eca90b5da4e51da5ba3d83f077
F20101108_AABMGR carr_s_Page_175.tif
ab1618afdd8f3c3f64297cc733e471ce
2732625673d5e91150ffe06f218a275faffb6497
615290 F20101108_AABMHG carr_s_Page_073.jp2
a4059449617fbb3c7740a6257bfc063e
3d7929d2b33b5f44633c43c3c8455faa42deaafd
F20101108_AABMGS carr_s_Page_075.tif
fe36a01938678017376793463ef7ae75
0dd9fea2e74ec99921cb43cd008080702dee5d72
2916 F20101108_AABMHH carr_s_Page_006.txt
c4bc390156935f2bfb3d069361a0d6b4
94357fe64eccf79d697ea78e3a6ee8ab4cb3560f
1987 F20101108_AABMGT carr_s_Page_154.txt
2bdc1e27bdb3a11a44c987438675ec54
286f984fbd8d95f1f2c5c4ca3490bb88a357bd87
159522 F20101108_AABMGU carr_s_Page_152.jp2
93933cc78ef6b1a97d909d0e20053794
bfa7dffd5ff5c9f834ee2cd7e754599d808d53b8
6383 F20101108_AABMHI carr_s_Page_072thm.jpg
287516778b44f12cf13fbdaf8d351843
5db85b6a8c679f35734ff5f8bf0668955c5a7eba
71875 F20101108_AABMGV carr_s_Page_006.jpg
30c11e8a7d0952aca9376b34c4177485
84fa4b913ea4adf2b4500d78e5296915a91f06ea
1051943 F20101108_AABMHJ carr_s_Page_072.jp2
409bb7499b7e5bdcfabaf15c3b7a9387
eb41acca0ff0c2e65710392f9dc5b6660381e5b9
23133 F20101108_AABMGW carr_s_Page_079.QC.jpg
de90f95d8f323215f3cce716c61230e4
faa88b45d2bb87d63a9b2e8aa4347f6af1fde435
F20101108_AABMHK carr_s_Page_063.tif
df446b565012b44703c9e5055ba79592
81e2b163d2d666d3a9d91aa5b0cfa3ae91bd83e4
F20101108_AABMGX carr_s_Page_178.tif
1cbf3b601d166c28635bbce9b1602204
3c9b8cd74a126a184105742183518cbe11b84449
664 F20101108_AABMIA carr_s_Page_138.txt
94a78eb81752d81a42bd5050514b5f4c
7be12e4d5e62b69d4c26e3a93b4b7abd4c3e7683
48513 F20101108_AABMHL carr_s_Page_067.jp2
2298d69934ccfa83db8c583bf490f427
079b5f8fdd8988a8f552b063b5ff851a3b7179b7
2093 F20101108_AABMGY carr_s_Page_082.txt
6dd8dd2bb75cdeda68bb3cdaf3f57391
f6bfa1017b67f0d557e2df06f40e0318c269365d
21943 F20101108_AABMIB carr_s_Page_072.QC.jpg
3eb19759babe4c300bcb4f2408a16ac5
c4d6a0c4a9aa611fa9c0ff5d457c8d6589ba62b4
6868 F20101108_AABMHM carr_s_Page_111thm.jpg
09a2a5dbcb5565f077f6bcad8e24cad4
29dbde0d70b61d49ed089c45424e5b0844b14842
139465 F20101108_AABMGZ carr_s_Page_182.jp2
0bcb52ae9e3b000f0828d4558a99d561
3c031bbb297ba3f352452daa9a28a1d83fe57592
F20101108_AABMIC carr_s_Page_117.tif
fec00983e5c1fbc8dae76a5c12ccf951
d5d018087a98a733637fd915325616f2efcabf1a
52374 F20101108_AABMHN carr_s_Page_126.pro
27cb7103d643ee42053ed63e0bbe35db
791e2a9db9619c6f7bf7ecb9acb26f63e74b2451
37216 F20101108_AABMID carr_s_Page_168.pro
0e86e4c99978a5cad5bfdfe65c2416ea
6c129608b3b0efe7f724e61aaff0b85a58a082d5
120671 F20101108_AABMHO carr_s_Page_058.jp2
42199281fde90c75fbaa08240d80dbd3
3c470c73977c46b58c80d0a46ecec5b1d1fda691
F20101108_AABMIE carr_s_Page_130.tif
346c74d2191f0ddd554573d31056ed73
d452d386ea0a18adec1d4b19e42ac2c243205b34
F20101108_AABMHP carr_s_Page_091.tif
fa00cc777a1c0ad1f294cfb7cae42e1a
a267118a61d6837136fd8dbb4e1ddb89bbe2bf77
805 F20101108_AABMIF carr_s_Page_169.txt
438b7d62808f936bdf5bf9c59c88ed0a
87ec5d4883dd050a99d418eef8da321f14a76475
F20101108_AABMHQ carr_s_Page_084.txt
34ff5c370c380495173378f13788005b
c140f7cd35b9a9153c134b33e1b258ed27735d17
79131 F20101108_AABMIG carr_s_Page_127.jpg
e4a8a3b3bae0a97c15391db15cad5074
65ffbda25d41e95dda15b62ddf168d6700563d4d
4415 F20101108_AABMHR carr_s_Page_167thm.jpg
a95eae69897b640ce1065b6e5141d927
2b14ef41dc6a3ac4918fd92b5472848242c644d3
77513 F20101108_AABMIH carr_s_Page_037.jpg
8acd1ccab30bed35260c8eadcd3f3c13
ff12f055ff113f45c6941444b120dd65ab5aa867
F20101108_AABMHS carr_s_Page_081.tif
1422930b3b251dd6c9d0d8ab0e79233d
4f910b7e53197d6179b553b70af3cf212aa06beb
6645 F20101108_AABMII carr_s_Page_090thm.jpg
06817af9fb94ca02069d3dc5e61b8740
94007f8a5bb83e1b7a2e7ff8697b0f6eee594966
24924 F20101108_AABMHT carr_s_Page_008.pro
9dae080561daa9b13e043ca2ce6a5ab7
ae8f6ef263004fad3062ff02f229debd4496f4d4
24090 F20101108_AABMHU carr_s_Page_038.QC.jpg
b0e12f989a44f4d242f1a9eac71223bd
15338926ae690f84fddaa335134500d7df20fc4e
70532 F20101108_AABMIJ carr_s_Page_046.jpg
164da5fd247457b55df51ab083a9a0f7
d46c04daa2099cf337d79eb4a7dd801a7303dfa5
112875 F20101108_AABMHV carr_s_Page_092.jp2
72d4faeacbbb029038d507b43f17bddf
bf041d22c21a9a082e2951322ccdba58758531fd
6760 F20101108_AABMIK carr_s_Page_105thm.jpg
f6e4872591991fa6da3bf2881d3d9fd5
de0d58d67245518cf0024dc217838dddffb3063e
2076 F20101108_AABMHW carr_s_Page_087.txt
83d7fd70b00b6a83a6e3db3a88bd6a76
c01d33efca0227d7f39af533e520f76ed8006cf7
118400 F20101108_AABMJA carr_s_Page_039.jp2
ccfc0990707af4fbc67e72232acae8ba
a521de6cda7af2af985acdd4a57e09eac352672b
22857 F20101108_AABMIL carr_s_Page_119.QC.jpg
3857f0f5da15765a11023c5691b8aa5c
246d967ae0c7051809511c29c14583aca72395a9
6735 F20101108_AABMHX carr_s_Page_048thm.jpg
5046bc7c7ec2c5204952d8f9d1044222
80033141e00bf44d2c11f45f3a60647d9c62ee15
25254 F20101108_AABMJB carr_s_Page_051.QC.jpg
5ceaac8c528ff966bbfbdad5b5a4f103
42c0cbaf0bc1379a3864a34ac40870fc65c12d3d
2060 F20101108_AABMIM carr_s_Page_126.txt
07799e59e8686567d2ee48eb35a9ddc5
af94eb3246e49b571ee00c914b162e89fa5a87be
F20101108_AABMHY carr_s_Page_080.tif
edb3b9eecd6f2d23fa21dcaf88fd885e
a71f1d63695aecb2c410c3354918b9a577bd71e9
F20101108_AABMJC carr_s_Page_147.tif
7246fedf8ccd1291f7e4d23a89f2716c
dc9ae93670a12bd658a38fe12848b751cef63f28
50039 F20101108_AABMIN carr_s_Page_125.pro
ee78ced976a3deb4d616678d7bc8fecc
9910c2a1e655a23e0b9b8e8d337d7c985c1b2f29
49504 F20101108_AABMHZ carr_s_Page_043.pro
65362536975f5fc86a877341488ba8bf
147f0861017cacca7765b3dd617b95461c0d0818
F20101108_AABMJD carr_s_Page_173.tif
ddd63edd5ba7d2ce18261110472149e8
b4bd3214bc9fc53189529a96f257fc2e96f2ddb5
6951 F20101108_AABMIO carr_s_Page_012.QC.jpg
59dbc571c007e87ae13834dfd17bebdd
aa1d059b8e030e4de6a065a055a0e6cb658cb3c7
6438 F20101108_AABMJE carr_s_Page_183thm.jpg
76bb7f083fada2d86fba15226639a96f
ab7bcebea52196ca79c7c347d8fcc53b7137de2e
53855 F20101108_AABMIP carr_s_Page_014.pro
de6856a4854d757ca0c11e7974085d85
6151d45c96eee6c85723048e1c987edb14c54a4f
38088 F20101108_AABMJF carr_s_Page_166.pro
9799a97beb2dcaddef3f1e3e156b5ac1
8c0188b91098cf561f017765d57464da8a9d8ea0
F20101108_AABMIQ carr_s_Page_052.tif
6cbe67553c2a448e8d12855afb4e98a7
248f0902e73215fc30b486dd56000210950202c0
701464 F20101108_AABMJG carr_s_Page_139.jp2
84ac55bb51255de2d113f82fab04038c
4524961dacf79b224e1ae4b0cda8281b600c498c
79435 F20101108_AABMIR carr_s_Page_081.jpg
6d98d3b5d6184ca3d780d4b0d2d4ebec
0663668873ecb733437854746571a04eb300703e
2075 F20101108_AABMJH carr_s_Page_030.txt
b0ee60659fe83d41ae21cd2f28d6e700
52f9ed09dd0c390708e0b90d2d75a9e2a44230a7
114020 F20101108_AABMIS carr_s_Page_131.jp2
8207a705f5036b6d47c7f36a6ed1c2a3
0eff690a11e205b09f1899ca124b6c5c6a00938f
F20101108_AABMJI carr_s_Page_163.tif
1822da3a016151d90244974c2db0ffcf
434888424c7d377e98ae541f582bac2cf0536e3c
2026 F20101108_AABMIT carr_s_Page_133.txt
e31e0edb751e821004a502b01f63db54
e34081228b5599e237ca946b038611e3c725667d
54676 F20101108_AABMJJ carr_s_Page_112.pro
f5a3e104417d8e4951261cb1b1d60a7c
40c901ea6aa07d71dd73e77cc8c5c6792a0037aa
F20101108_AABMIU carr_s_Page_116.tif
5a9d6e1af5ea620e414cb7bd2dc089dd
5bf9379d860807d5093ea6f5b1418e6c353a9ece
49842 F20101108_AABMIV carr_s_Page_157.jpg
d2f334d78279a58d29e4829cc0bc161a
62549b933c263651568703f19882b7c715137ccc
28626 F20101108_AABMJK carr_s_Page_074.pro
27b76a282cc5e82874d26abf6097b8e7
521a0f75fe7c524decc0643bf902c722166a175c
35431 F20101108_AABMIW carr_s_Page_060.jpg
94d9acb05e69c96c8db3ecffdcaab602
8284a8832ea496a1ad5b1a762169dd8a8806e150
45016 F20101108_AABMIX carr_s_Page_154.pro
413415d4d7e0e0e6161413c05cdc9f91
db1d001cdc2cdbb75509c282cfa6ba45ead2caaf
F20101108_AABMKA carr_s_Page_077thm.jpg
5d817db2b1f49c49ce1457e1ac5ad837
5e2f9c3d98329657d7d96de5b66088907f9842ba
24528 F20101108_AABMJL carr_s_Page_110.QC.jpg
bf90bf0be2a2911a1dbe2f9a6808d303
b63926634ddc792759650d8b5915f6e22893495c
46803 F20101108_AABMIY carr_s_Page_165.jpg
b8b6d11f23f0fd9b183d9198d2ddfe8e
32f972fa7281e57e6107fe8e22a015b790433aae
1448 F20101108_AABMKB carr_s_Page_167.txt
69d8b2a776a40814588bc471449e6549
165a75409fd9b2b13a8046f5ceb95aa18f28a351
17336 F20101108_AABMJM carr_s_Page_074.QC.jpg
586c2a76a491e41339be5ee275bd2833
f97d6ee4f2c6de168ce0d20548eb030764e111d7
119569 F20101108_AABMIZ carr_s_Page_106.jp2
8650afe5fd7bc241681761468c9ccdaf
1d831c03c35269983f8c44c02bd7a4f21bd09cee
6614 F20101108_AABMKC carr_s_Page_053thm.jpg
dda45a97669851f7c013ca4d63c57f43
8bf5943b9b9b1566622e95560e2b9e595570e078
F20101108_AABMJN carr_s_Page_089.tif
ab1d2b58495ffe6344a9aaa9356b7f17
78b67311243c85fde1c40280cc8b03f3c7c15e92
6798 F20101108_AABMKD carr_s_Page_016thm.jpg
e65377bd9052e9ed3740342e864f6332
75d03b32f045f693f84f2987bf29c6a071be7fdb
25674 F20101108_AABMJO carr_s_Page_118.QC.jpg
968590cdb9004283760f31ac9405140b
d804436dfb967fe02b246aa8bd9c6d5fcff45d8a
42445 F20101108_AABMKE carr_s_Page_073.jpg
424564de58cac623c33e39aac2d49738
13afd064ba231243411c1d756f91aea07e20a237
35723 F20101108_AABMJP carr_s_Page_165.pro
196648c205a6c0331bde605b0712f19c
2f4dd2ac62d53d44d92e2f2fb77289c29cf3eced
70567 F20101108_AABMKF carr_s_Page_119.jpg
78201a8a92ea288547955ca49695f418
f27377e8721b81642a495010e79db43cfa27a4b2
3411 F20101108_AABMJQ carr_s_Page_007thm.jpg
4f272e5e1b84c83d558c025c87710f1f
f85af59e8bc1dbb6eb12751b114c5db6ee5b9b89
24583 F20101108_AABMKG carr_s_Page_131.QC.jpg
10b5b89f3832c0f0e2ec48be1ca80baf
a19bf2d2766f29087a18d4e20373e9715daac090
F20101108_AABMJR carr_s_Page_029.tif
dee433736b7e7169766c7e72cdca0d96
665d253d605bbc4ffcc22b7876ff81d7961524f2
1054428 F20101108_AABMKH carr_s_Page_098.tif
20aedef954045bcb1c79a84917a30ff1
5243b14681aa1928ecd579c8708ce53e06af3102
26381 F20101108_AABMJS carr_s_Page_091.QC.jpg
c1e9859cdd997937ac0161cece99ddee
ff9fa1c4c7c48b8c654d23d9829376edb9a260d7
66996 F20101108_AABMKI carr_s_Page_005.jpg
d28dfab304d9ad878dc25d28442f09f7
6cbf7a6c5407b9a5d9d60bb4ab23c4e083b4a48e
22964 F20101108_AABMJT carr_s_Page_067.pro
5e2277dfa3c524261267ba8f6079f2f0
44d320db06258c910a9b2655b77fc3bdae7886b5
F20101108_AABMKJ carr_s_Page_120.tif
9fd8f9b8fe6d0d797edb965dae78957d
0883eae5aa3d0cf2fea33d9c1f370d269fb2d09e
148400 F20101108_AABMJU carr_s_Page_150.jp2
ae13befafa4ece8a278f2ae18db0a7a6
444d915bc5ff5630d06fa0ec64c3a9bd2cc328f1
13706 F20101108_AABMKK carr_s_Page_139.QC.jpg
4b59f430a92e9db7bb0b6339c5ff4fde
e0de0fd478f00f1f3f67715a108b6709e5f6d86b
F20101108_AABMJV carr_s_Page_179.tif
1278b4ef68dd0818ce1a3b4b99053ef9
f1e8d81062b2d79252dd58ffbfa4ca5c47849d54
F20101108_AABMJW carr_s_Page_158.tif
0e960a7b497c879cd547aa863fce8126
48ad24d0ed3e74926c5a944763bc368618e55946
6809 F20101108_AABMLA carr_s_Page_128thm.jpg
c0415970c611f955f5f7cc17caa824d8
2affe5edb00b956c1d5f45f42bacc2fc2ecfbf89
58403 F20101108_AABMKL carr_s_Page_066.pro
7b3b06755e2a6a83ef37f894ba0ba26b
2ad6cf77f678525bfe1b88792c14b00fb240be28
F20101108_AABMJX carr_s_Page_024.tif
7ccafcd4fb9b5b7624ab6ef073a77925
b0b71eafa40f93993a82e85ca0ec76bb0ebd07ad
F20101108_AABMLB carr_s_Page_109.tif
f7b00cabce50fa758adabcf5219a019d
950043a120eb05233722f17b1c386255b551eea2
162097 F20101108_AABMKM carr_s_Page_149.jp2
2f3e2191df312c0b5c1839850fd48390
d6549283956b8b87b3f958d668eebd95c11e9b75
1051975 F20101108_AABMJY carr_s_Page_006.jp2
b2a31ed9a677794c82adf4ad41b29b0f
f85d88381b28f111322340b554a059dd1be20653
8506 F20101108_AABMLC carr_s_Page_136.QC.jpg
4cc5bcc3b62584e6d0c99b62df39bdec
772c3db50f093a3627b50c26f522810082d484b3
67003 F20101108_AABMKN carr_s_Page_006.pro
593ca4025b5eb1e54e861e152c37668d
40bfde8873c7f24f8239fb018bce9d2801cfe5b7
6640 F20101108_AABMJZ carr_s_Page_029thm.jpg
3280be5cc8cb4afedc748fe9fb3b6ed3
216027a6dde31a96b0882283e4075631836f3de8
6987 F20101108_AABMLD carr_s_Page_054thm.jpg
624c1e64f675c0d81974494ce46bf000
d6fed50a0e533b8ecf92c8c447d88a9f7f93c3be
11018 F20101108_AABMKO carr_s_Page_162.QC.jpg
4642c3e8368c9bc6cbd2d1b74bb509f6
f8fc2cab577fd0d88e50b18b6daf216712842ce2
F20101108_AABMLE carr_s_Page_037.tif
0870c68222c73ea6c908fd428a2aa72c
c968437d5e2b19073546a7c986b0aae518901a88
53700 F20101108_AABMKP carr_s_Page_021.pro
e0521683071487040c181ebae7b845ff
c4da81b891eedec08fb9e696df99b3b9691c6732
25266 F20101108_AABMLF carr_s_Page_180.QC.jpg
8d10f9cd0ce31f297eeb391909697ce4
87789ea3c0d68cc403f0c2223acbab9a9be48c3c
115600 F20101108_AABMKQ carr_s_Page_129.jp2
34b7d316fed654fba0a121470270f004
b43b9d2800741a7662a06680c6ec856f08562d03
113384 F20101108_AABMLG carr_s_Page_066.jp2
06057e28b2b102fb94c647fa7d40d75d
bd4ff92fbb2244159a7cebed211d08663846dd57
48744 F20101108_AABMKR carr_s_Page_168.jpg
8d5bbe61f2a385434af53bb3aebf2fc9
267113e6fad44140e9ffedbad666b0bf45b2ca6f
27696 F20101108_AABMLH carr_s_Page_174.QC.jpg
d9a1a116124fcec885a7004201cb45ba
be4370f2413d4c66ce4c45842afc9c8f7ee0cf5f
811397 F20101108_AABMKS carr_s_Page_140.jp2
dbe561410aab439fe3289c47c57453a7
47fdfbb51acfc6e2b08fd9494109413d683c0416
F20101108_AABMLI carr_s_Page_061.tif
e48c8b2657dc56a628f0b74b046d9461
a885af396030d53bc05c9adb84c595a40263fe8e
42488 F20101108_AABMKT carr_s_Page_132.pro
726494cdfba053e9f7e38077dc2b4dbd
1053ced4b503365b0a73e375eca79092bdaed162
F20101108_AABMLJ carr_s_Page_022.tif
0a39820d824b4d5724bf55ed26f3a65b
de8f75393cf785132f50857d0a2fda7a6dbfc3be
F20101108_AABMKU carr_s_Page_096.tif
328897aa2d46dc7f80565012e9bee2a4
28e491a2ccda2ab6a4a6e9be8a6bf259998c0fa0
F20101108_AABMLK carr_s_Page_094.tif
5e51f747bfb917e6a9bc1ae638f46ccb
3516b8b589d18e889703d12e8a0568d9d0b13b2b
107609 F20101108_AABMKV carr_s_Page_096.jp2
816680857e449ebfe52790bd88637e13
2d7c39cae4bfbc2c30d57f4ba69aa532e5ef614c
2758 F20101108_AABMLL carr_s_Page_173.txt
81659d742b494980d07aa02f6b28a5f7
0be91bff23c04d16bd36b2261132207aae194b38
7234 F20101108_AABMKW carr_s_Page_151thm.jpg
df4aab3c9d8665b61b4bd50297d291e5
0a0b035e4e7faec5009ffef0885297c3be2ec6a5
55500 F20101108_AABMKX carr_s_Page_144.pro
adb732bf520bf00b9cd00f936cdc829b
3bf4cfa6a7aeee0ea7cf3b7f2380848384534940
2020 F20101108_AABMMA carr_s_Page_047.txt
d53a3ea3c1f1f05079c73b590ca16d60
032deae41ebc5e29f814d044bd72a3eff45dd247
40791 F20101108_AABMLM carr_s_Page_141.jpg
febeaff9f55ad229029c0bd03ec86374
a891faf091859ef550532bc44052d601ca747ca1
37751 F20101108_AABMKY carr_s_Page_164.pro
be98dd6f0acce3d9e314eb0d76ddd58e
ae8a24d8c6d9d92b16dd19d0d3f9aafe4ab2aac8
25005 F20101108_AABMMB carr_s_Page_122.QC.jpg
ae48d211ba14586a6b8d9b78e4096580
2b17f0649d01faebee5616d5e926073590433c61
108253 F20101108_AABMLN carr_s_Page_080.jp2
fca442e194bd5ac9d04de6abf9bfe58e
1e172c478d24d7d16e505b9d817eb9cc5d8e43e9
52717 F20101108_AABMKZ carr_s_Page_082.pro
71545258970d568848bd4aafac5a69b9
0f88ea31a5defb1cce90293df709c40bf2e83f1d
F20101108_AABMMC carr_s_Page_137.tif
383f4e19087d40cabfcb80c9219543c7
890572986d7079be0273a6cfa55927d3dafbb5c1
1895 F20101108_AABMLO carr_s_Page_163.txt
77639963d0b8131b1d2e6cb37d8e94af
71dddc908feb5a552bee47ac985a91a16e721496
2062 F20101108_AABMMD carr_s_Page_083.txt
205c3dea4539e7dc452b05e4cae87763
442ad42f813759a4665dde173b97395e771b0910
75079 F20101108_AABMLP carr_s_Page_111.jpg
ea3dc33050b02ff5bf4ef785e7b28280
b6df1d7cd8d5456f0f9c8158a771791df5b9a591
120273 F20101108_AABMME carr_s_Page_022.jp2
dad03830fb18d51468fa2742e419f11f
89f5920cb632abac88a9a3d55ea5078383531c34
F20101108_AABMLQ carr_s_Page_138.tif
da577bb9bd02754711961924567e4cdb
09bbb7fbf75279509b8b5779a387fce6d82cfd85
111714 F20101108_AABMMF carr_s_Page_109.jp2
4378013b9d98b0c005fc246920b8c148
fb3aa17437490b47662a5cf55aca26edcf73cb4f
7012 F20101108_AABMLR carr_s_Page_175thm.jpg
c1ab44483a3f8aebefe880d9ec28c665
48ce48cc5f6130ae8653d8ca60622b7915ae087f
6353 F20101108_AABMMG carr_s_Page_078thm.jpg
131942b35ab99ca4ec6a8fcafb598ff8
4e493750d378972adff568d3a1e59209c66a3473
2130 F20101108_AABMLS carr_s_Page_130.txt
b74d47c87ca26cfd5c16cd3e3c8cf135
d8d5da99ea9b64ba39a21bc4c4694e58959ca5ae
6694 F20101108_AABMMH carr_s_Page_095thm.jpg
1ff5ee9f27d3198753320fcf1a5bb1c4
f3b2d815d11deca372ed03badd432ada067ff60a
76312 F20101108_AABMLT carr_s_Page_021.jpg
5f3f1ab063b498c47e6e38e9b7f0bed1
446c7006756b96f2aefa28741bb0ab6ee7c6a249
2179 F20101108_AABMMI carr_s_Page_058.txt
e70a336175b6cafb96237e488a7ddad5
3c078f4451b54f51664bbad59a4f95fd93bcd49b
25559 F20101108_AABMLU carr_s_Page_106.QC.jpg
92553ab24f7ab0d3b8ab37c3766bf747
b318369d75e819b1033f4dcea82a686fa0f1ea83
15053 F20101108_AABMMJ carr_s_Page_158.QC.jpg
d7e47dbd4131ba976f6d82eb370aa59e
8cf83e401def76944def346a301b300d2e15c703
52912 F20101108_AABMLV carr_s_Page_017.pro
d6654a8caa4421df4570d8b29b303e12
9dd0529e6ed159ff7f31f136cb8b95bbfab0e261
2188 F20101108_AABMMK carr_s_Page_107.txt
a59c9851c1af05a5994195348c18d883
3f91aa28d66363d4e3708c98cc1d58f0d753f413
24460 F20101108_AABMLW carr_s_Page_113.QC.jpg
e9ff93ada71bef5367e729696b495e63
66c4952e1ff72cd73ef598025ab00f3516b54d12
89627 F20101108_AABMML carr_s_Page_062.jp2
147c32d27bd66791b51465c94a826c69
cbfe9f9e549bcff4ab4bd5e3f952bde7baa6ab6c
6912 F20101108_AABMLX carr_s_Page_118thm.jpg
1c0d8919a365ca23c34393f5dc86ceeb
e9501b831ed9d0b1431fb78a639507b5cf223a6e
2032 F20101108_AABMNA carr_s_Page_090.txt
ec98a95ccc395453233e023b35ce629e
ed3cf9dc9ea996c577bfd313a0228a51ccafa040
7371 F20101108_AABMMM carr_s_Page_173thm.jpg
20f28f01127799fe89dc9d209508eb81
d660692af5edce11032cb2e1f0bf7e5b763b4013
80806 F20101108_AABMLY carr_s_Page_093.jpg
81b18bdd54084807ab7278e106ad1e8f
75f419c382358762bdfe34d178e602f55c92a1c8
24992 F20101108_AABMNB carr_s_Page_176.QC.jpg
50d643265341a158714687b0d925656c
1342e1b9469e65313facef682b533b39fddeb67e
6924 F20101108_AABMLZ carr_s_Page_058thm.jpg
3ab13f6a8113232b8687208b2463c2f7
5a01a7eb148b03e60d16d9cb58ffe968baee210a
2189 F20101108_AABMNC carr_s_Page_105.txt
997fc70dffe9121f94b5d9df8597a36e
88f538dd35e74f53e2d8af09a6622cd9eaf232d9
4196 F20101108_AABMMN carr_s_Page_155thm.jpg
35320b389e658ef1893b3c1554818a9b
4950c7583c33a3bdc31bcb9bc34455a174887a0d
48681 F20101108_AABMND carr_s_Page_045.pro
5dfbc68a1762eee86b70cbb50bae5677
b843cbd53ad890d9c5b84cbca475ddf79c0f5973
20268 F20101108_AABMMO carr_s_Page_065.QC.jpg
f98b03ed8d6203b1b33657561fe77a0d
a3c1a72792685d1f975ea9975641c68de005ac53
121990 F20101108_AABMNE carr_s_Page_054.jp2
900f94cd81aa428d67b0e43b0cf8e4a7
8f8b6269d554f66cb90dd09a9727c989cc58b4cd
1629 F20101108_AABMMP carr_s_Page_007.txt
fecaa1dbe96945c33b371deacbafd3b5
befc2302654d736592281b0b718e4fbf74ddccff
61812 F20101108_AABMNF carr_s_Page_132.jpg
8a72c93db24179adc8f085c9d541edfb
bf6bab50e922e580eafa9811d307a9bc9cb69cd2
23803 F20101108_AABMMQ carr_s_Page_092.QC.jpg
3a433e727b23d3489aaa182438436e36
b73b32b30a8ba536d1887967d4629c3cb9b6d0d1
52819 F20101108_AABLKE carr_s_Page_037.pro
a393a8229b5d8ba5cc3fdcff89f6661d
5a9c0e1139586f3e2e640a4c3b0a63a4e3daba35
3266 F20101108_AABMNG carr_s_Page_149.txt
f537f16a174a510c71409883688dfd6b
bd5fdee83048a00ef7424f83097e451abbc868a2
35367 F20101108_AABMMR carr_s_Page_137.jpg
d4f44ff9db407d6c429308b0f63b215d
6614696763e1d3bad443a9169f60df3d3807f55e
74781 F20101108_AABLKF carr_s_Page_086.jpg
51cc7401d792e0accefcb1fbdb19a5af
89821230a3a28cde9ccc57ced5389a083fc4f63a
73214 F20101108_AABMNH carr_s_Page_101.jpg
6c6d86f72f77e4bffef660b11208dc1e
c46a7f2d1a4fa8bb0e981256f7430ac1be6c5412
2080 F20101108_AABMMS carr_s_Page_035.txt
c78cba67cb37f9a58750f7d01c31376e
ad3f6110c09d9c3e5358859a1ff7615a1fe46a5d
9968 F20101108_AABLKG carr_s_Page_003.jpg
54bdd53c1e641457f97647614c4d0341
43ec95e0e08526f3257de7811a0e535b335c21f5
2095 F20101108_AABMNI carr_s_Page_117.txt
9ec4c5574f77668be140677e60b549f5
1940648ff38b08f8c01647aefb42671e563cb0cb
77330 F20101108_AABMMT carr_s_Page_019.jpg
649ab7b656ccd1570fcb0e9d2bb48b2f
a7314011fe9355e0c4d6b501da59407a5299cc9b
7092 F20101108_AABLKH carr_s_Page_089thm.jpg
7c14d7e077ed9a030c27efb315e73f12
df0b7ea70d1692f5bbe9d37f7a05bb9f3019e140
2832 F20101108_AABMNJ carr_s_Page_136thm.jpg
18db5ddca885a1d3a52136116c5bbe72
4ce29440678c22706198becb38e0e909411c2eab
52601 F20101108_AABMMU carr_s_Page_087.pro
0a8fc7df3f660da7f80389607de2b666
f35acca8b661b808fe212f03c4402ca5691500bf
11155 F20101108_AABLKI carr_s_Page_184.QC.jpg
376c7888b47b1848f19383ade1ad7e27
869c68f53d533136bd68ebc3ac872695688a86b1
1031 F20101108_AABMNK carr_s_Page_101.txt
fbf55845886bec0ed070c0a7b246d5fc
927082f9f156973790df4da9dc9c1b82d5a45642
26299 F20101108_AABMMV carr_s_Page_012.jp2
fe2ba06a8f3d5e0ab92606d0a7693f2a
1366357791fb412066fffd11da82f333bf0de9d1
756 F20101108_AABLKJ carr_s_Page_139.txt
c15d8918703056e168510b2d5772c076
461d45c4729d46365f8e8028559ee554d48ea47b
71771 F20101108_AABMNL carr_s_Page_079.jpg
955b4064e1a00d26e27b46fbafb58864
31885c64de2de7c1a282d97651b07f5b7dd67425
798 F20101108_AABMMW carr_s_Page_184.txt
c3255c1bf02acf01e084b8a054aac039
f5ebed36ecbee8ba10657fd07c9613cbfbce8b51
80067 F20101108_AABMOA carr_s_Page_054.jpg
6870cd864ab18b9eee13b9ae0c99150d
da2fbc43097770d5deb7c4c09b0e335d83ef2360
84291 F20101108_AABLKK carr_s_Page_172.jpg
ac27c5d3c843634ee7d02ac982edb0ad
af9d8392ae4e7b6c94a05ef71b7b3deb333cec7c
23630 F20101108_AABMNM carr_s_Page_108.QC.jpg
431f141a1edb58fdd795f54f9563e3c5
736b7d59be96555f5c947c4eb213f2382bc5bedc
51158 F20101108_AABMMX carr_s_Page_033.pro
76c0b746c2c25b636027d3e321d83543
936cd5e8086656dff17e9d08e0eaff72afcc498b
112528 F20101108_AABMOB carr_s_Page_029.jp2
585241d721b662df753f72aa836e8221
05bf820744361c096cffb1ce5cba24c5d43161bb
3714 F20101108_AABLKL carr_s_Page_135thm.jpg
3a976db68c4b2d5374b1522519470950
910b768cbc1fad032fb08f32af3d24bc74743462
2182 F20101108_AABMNN carr_s_Page_009.txt
da0c6e4cd2f4b3be8cc55c71dfbc67b7
8d2a907006541ea99a2964efd56e643eb3511fe1
72675 F20101108_AABMMY carr_s_Page_124.jpg
2481c4237b83c620e7e631388d098aca
5efe8ac5bd10f5193239274ca06fa82d3c8ee759
79042 F20101108_AABMOC carr_s_Page_022.jpg
4526a633df050f698a9c1ce3f8ca9e20
89da28cfd393e60950428d07c4ba583065eb1736
115062 F20101108_AABMMZ carr_s_Page_019.jp2
904a267b40e8b1d99a2e9ccfbc2c3117
6e682e81b91c16dd045c4d62feaee311efa0a4b2
F20101108_AABLLA carr_s_Page_050thm.jpg
ce333283b6a65adf94b2e5497aef26dc
97675244e8469d5c2a019e702e6c0e117a2ce841
26000 F20101108_AABMOD carr_s_Page_115.QC.jpg
22d6d409d9273639b41b1d615140b911
6c37787226674a22dad57e9fdaf40bb2c16bc015
50324 F20101108_AABLKM carr_s_Page_036.pro
e490b3bedcf0c15b7efa054c7c809fd2
67c12e9121cac230ce16202ef80232e4f042117c
51090 F20101108_AABMNO carr_s_Page_047.pro
ab78bc91cef8f499b902f645e0e11017
b8db8bd64093d00ff5a82dfbb12b5dee16dea518
F20101108_AABLLB carr_s_Page_081thm.jpg
6d37f06cd5c6c5fdd1b44495e3d259a7
7ef33d2e74303775fe27658fd436495874d3cdf3
23120 F20101108_AABMOE carr_s_Page_045.QC.jpg
604aa8949992a7c3c73698d6564416f7
d8d803a23eed594f812db65a80c6dd0130cf052e
48401 F20101108_AABLKN carr_s_Page_133.pro
c7ea8396aecea66737154551be1dc9c9
ac8d78caeee7202f713efd4cb7674c94e03cfda0
F20101108_AABMNP carr_s_Page_065.tif
20227955a21ed9c4178d1e7bbc60adaa
82e895b692f1066bda195bbaeb4de10668b364c3
4278 F20101108_AABLLC carr_s_Page_140thm.jpg
d257311e52f5e4299c4e70c1f2e2efb1
91af0a49509dfdbd9ea72435c0336e214027768b
75795 F20101108_AABMOF carr_s_Page_047.jpg
47887e90613affbff511e33969a3a21b
95dcbb4dcf3fcf1574b3f011958bc617febe1eb5
104574 F20101108_AABLKO carr_s_Page_041.jp2
38eb098f566eadc750b483f92b888881
ed08695c2d4a75c260fa7bd68870cb1e08a482fe
52670 F20101108_AABMNQ carr_s_Page_015.pro
788676808bb707bf32cd489c46be019c
695eb9018e5dc8b002c5ff3e30c4e465afcda9b2
78662 F20101108_AABLLD carr_s_Page_123.jpg
e990e32c9bd0b5deffa18331d6b92d45
e847fc3c9c5522df31dcdc7faa1c9cd10ec6cdbd
55882 F20101108_AABMOG carr_s_Page_154.jpg
24b438446744682e75eb23ca473c6e6a
3379999c2561af57fc33997d0de872bd6c5e60bf
24391 F20101108_AABLKP carr_s_Page_044.QC.jpg
51b74018b6e8826f64cf737604230788
99f4f6e06402a19d85accbfb9558598532ba7b5f
52456 F20101108_AABMNR carr_s_Page_016.pro
6a14d7127fcfff4d7d54f79763028a92
4a55bfe427c18fdb4bd539f25d204e6fd670fc56
106893 F20101108_AABLLE carr_s_Page_040.jp2
c183c0d579204966415599fb65344c0b
f7ee49f71435d4bec9b38ea9282de1d6aac15edd
F20101108_AABMOH carr_s_Page_155.tif
de9cff75518c8ba08011c8de64a03eca
5943d17232921ba625f33fa69a762da2191fa940
70547 F20101108_AABLKQ carr_s_Page_174.pro
b70520d0126654528a65ef6dc58efea6
3aef888a4f93105062e0a3beea79d9520bda691d
F20101108_AABMNS carr_s_Page_079.tif
9f49db41915628c87bdc1106a2b02982
2f17b3cd6a1bfc1679c49c0a332e2976db4d7bbc
79518 F20101108_AABLLF carr_s_Page_112.jpg
0fd0ea8f7ca6643dc4bc9e4ca2e17c23
c2e224503d2e0621087adce392201118336bc9c9
1563 F20101108_AABMOI carr_s_Page_166.txt
6b05636a19a8663375fd0151ca846be1
a9fd2a28d9f3d2212876bbe063c2db5ec3452c5d
F20101108_AABLKR carr_s_Page_100.tif
0b690fd2125c594e19ec06d28192038e
0f41cedb1eee191a2b90eae396603eeab7908763
1700 F20101108_AABMNT carr_s_Page_132.txt
7f6e3b11e72f5ef2ae7d0692ee0ba6d0
4feb6403258d322499ec6e4f11ab5dbb6876ad5f
25390 F20101108_AABLLG carr_s_Page_025.QC.jpg
5a927fdf5cb5c29c041ed2266c264201
732b57455007d05e14f1e1421ad10dc426f3998b
56042 F20101108_AABLKS carr_s_Page_077.pro
8c1b827da0232c9c6e5c826160e8f054
9813f0744536df48f2db2013388ade2970240b00
F20101108_AABMNU carr_s_Page_044.tif
514a4088b71fd457aab69307e79847b6
db8bca71aba7be184894d68eb0fa05e90f988270
48376 F20101108_AABLLH carr_s_Page_078.pro
0515f70ad9858f1348a475422b8cc3f1
bde81a6612b097ea90b059bef867bfb2fadb700d
73832 F20101108_AABMOJ carr_s_Page_036.jpg
876c78486b0fb3ed0f89936b811e4c86
b70235aed1ef7543dfea339eec1d20cae6e25a52
6642 F20101108_AABMOK carr_s_Page_113thm.jpg
03a5fad103e6910034f6aefbfd89aeac
c819005172d7e4878e0eea17fedeac7217868a49
F20101108_AABLKT carr_s_Page_029.txt
f7e048e04b9024388a06ae5ff455ac1f
b0d579933f1c7341eb54397d91534a26870baac5
9509 F20101108_AABMNV carr_s_Page_169.QC.jpg
6192cc9ac01f614e3fd42dcf6b2dec86
69bc784b3e36e651ac84fcc78df72a19bcbbb59a
F20101108_AABLLI carr_s_Page_020.QC.jpg
52fcc0ff0c4106a9faaf04fa6544fe80
b21dd954e72495c329b7c67a75332d83ae07fe89
17608 F20101108_AABMOL carr_s_Page_061.QC.jpg
8a497c5b0aac2304bb0b2f4ff4f902e4
65fc6106e5f84c633c8be8496a242039db51ad60
F20101108_AABLKU carr_s_Page_171.tif
69c24a3e7aaa7a727ce3908e4d3c18c9
6291e332e610774610969d9d6d8996ddeaecf771
61080 F20101108_AABMNW carr_s_Page_104.jpg
d0ae7003284ce3c75703aac08737c637
2f2a8842fe1f4c2be8e9158d0e98739a276780b9
F20101108_AABLLJ carr_s_Page_069.tif
3d6376ab36d54368969370408cc33307
ce8565f856e72d8a2c6839e5a45b494c1294bfe3
F20101108_AABMPA carr_s_Page_059.txt
3056685a0e13823d1567ea4c508cccd0
8ecc62aaf6dda8463766488034a2cb07fe48de6d
2183 F20101108_AABMOM carr_s_Page_144.txt
f134234beb9756f7052a9a14638887a9
e1b770dce4fee0e94ed25ed3d4b1367546b948ad
4180 F20101108_AABLKV carr_s_Page_157thm.jpg
002aeb2135898e08ced30ef51ea1154e
1acfdb79f1bacd473fd3d75ec15db02552325a0e
24518 F20101108_AABMNX carr_s_Page_126.QC.jpg
69dda267c2505b099fad89ac4feb38a7
f4cdccd34efe7630766778e5b45107f996790345
61588 F20101108_AABLLK carr_s_Page_065.pro
758d1287b673b77ed72220c7e0765fe8
3ddc5d1028fc157f1d3db015ce32b81c5ba1ef1f
74081 F20101108_AABMPB carr_s_Page_044.jpg
5e69c2b9a19ab40f980f16832fe72d5c
814c172780d6d9ba8594ef91bf778c1e805d0957
6316 F20101108_AABMON carr_s_Page_005thm.jpg
956bad5bbfe76c90bb3f6721a19251ea
9fee364e619d7b1bd1d92d8ed7408f44942e686c
81761 F20101108_AABLKW carr_s_Page_155.jp2
0e96bd6b255b73f7fafdc771d2101205
6b07003f559cfeae262abe661ddca36fee2845ae
20356 F20101108_AABMNY carr_s_Page_103.pro
c99608ecd566de111ba1510672321cbd
3039a30841924fe9a2bfbb835e110f4134eb1426
6866 F20101108_AABLLL carr_s_Page_116thm.jpg
078c70f140b9b6fe712642e31772f629
a696dcaec67af327162e1a8cbce9b61e605a206b
F20101108_AABMPC carr_s_Page_049thm.jpg
a3cb6ffeb2e0913afff57647402edd5f
87378befc3d80a718afeaccfca8bb0ff4df72715
55020 F20101108_AABMOO carr_s_Page_085.pro
a3cca8a35dc97f1234d571e3164dc187
419a7e1a7a75a1e911729311e6bf314a8a66f7d3
52722 F20101108_AABLKX carr_s_Page_121.pro
429d9b8f292b473f619479cdb19bfe5b
2f45a5231ba9a92195b4b3cb70eaaf0874f7af0c
47243 F20101108_AABMNZ carr_s_Page_114.pro
f8e5c5f7f0e24fc35aa6abda122db8f4
cf35ad0a31b3a6e595ec18ea0e6949d4e79e5c00
60675 F20101108_AABLMA carr_s_Page_162.jp2
19511311b23b88f4583f2522c6886541
72db82627dfb4b8246e190f3b4b0bc5f93634eef
54722 F20101108_AABLLM carr_s_Page_076.pro
442f5946a097db55d50990d3eb08ac90
f90b64d0262f465476cd6387f4587bbf33f830e5
124722 F20101108_AABMPD carr_s_Page_172.jp2
dd9ffc03ab6a8e973f0368aa2e65b4ca
1d81f0aa2cb65561ab57ff916823661efee2dc82
52629 F20101108_AABLKY carr_s_Page_156.jpg
e53e8a547fd804913ea198998d67ba93
016f8453a1d287b33b1b89d690b34ea6e3ae119e
1613 F20101108_AABLMB carr_s_Page_160.txt
38cefd391beb8bd6d1c610a951e3d44b
2983f10f964e75236f55e24f0c9a4919a3e33e0a
52834 F20101108_AABMPE carr_s_Page_105.pro
252a3fcdb7bbd3870d3004aefc9575f1
5d171eeca7bfddff6b7ccc47dfbc560a9856e55b
29103 F20101108_AABMOP carr_s_Page_153.pro
05ca8fa710ce119e5b4e369dadc17abb
762abf616ebcbe496a6aa6669e10685cd7088280
F20101108_AABLKZ carr_s_Page_118.tif
d59929c543d8ee425655ca496cee071f
7059ecfb9bc084b7dc401ef8ed07837bbd01d87d
45881 F20101108_AABLMC carr_s_Page_135.jpg
cee9d54330dee56cbcd30a1ef07b2a86
fb6854f9dd376b61e623b7b12efabdb651390af7
62143 F20101108_AABLLN carr_s_Page_181.pro
92dc4dcfdfa8fa8e4ed05b0cfad12c55
deb243037762fc0c6a7c26f61447fdc0770805ed
F20101108_AABMPF carr_s_Page_036.tif
c5db9af43b99302243d1daa365726663
71de1a131d2d772ea48a333d0c4c817a07f54d8a
104166 F20101108_AABMOQ carr_s_Page_031.jp2
27423fa8c0dc3cdaefbe91b5317a31f0
a353be808a2c5a839bc220f7e84e750d5fa0a19e
20184 F20101108_AABLMD carr_s_Page_059.QC.jpg
77288a4ca2d49f1a2a71e5d1d13eb479
054a7271bb163688f972b64e07035bfa332c9bdd
9642 F20101108_AABLLO carr_s_Page_002.jpg
d0e9b953f87e68c1bd4ac969baa826e4
2c104169362cc62bfd04ee0a8b149c020d21f4cc
2534 F20101108_AABMPG carr_s_Page_181.txt
f423ccbec17082839ab9a94d0afa302c
6f16b03fc16daf6f9658f040855cf1b2acc124e7
1051932 F20101108_AABMOR carr_s_Page_135.jp2
f65b7addc40c55ea6e23f266a9e7a385
3fde6df024021795f1942856ddaaa05d9050417e
24844 F20101108_AABLME carr_s_Page_024.QC.jpg
7935a89eaf8b9e157738cb355e19efad
a56f1ce5dfeaa148cae94f44f46078dd3ef8d80d
4361 F20101108_AABLLP carr_s_Page_165thm.jpg
a36d24fea51d2734d7b55849245f1655
4d1ad0fc4eb27d9386a74115c2850d193828af29
F20101108_AABMPH carr_s_Page_124.tif
09938b059953b57f94549aef795acaa7
9badd73875c14a9c7bf93a6a9e3e92415b1e00c7
3196 F20101108_AABMOS carr_s_Page_150.txt
9eec70efdf3bb8707aeff85eb7320515
2c262131a82fab1f0144c4d6ca5506cbb882d749
6469 F20101108_AABLMF carr_s_Page_045thm.jpg
3fbd1febb3beec26d3d9784e35212e3b
5fecdd3645e71686bc2240c07111a8d9486d4bd5
26392 F20101108_AABLLQ carr_s_Page_094.QC.jpg
d874b688dd6d44ce3ef279a17f44d2b4
b016686d94a89a2046b7d5e1b404ac34ea3ab220
23653 F20101108_AABMPI carr_s_Page_090.QC.jpg
7ef1bc153548ae3deeafed34c1e98c56
4c6b6413503575eaf2d8ef2873ddda768af02f13
56022 F20101108_AABMOT carr_s_Page_115.pro
27173dedb2fc9197058b10159f750931
8ab2a5332076a285a4a8b4cd38bc2bb63017b48a
4377 F20101108_AABLMG carr_s_Page_168thm.jpg
730b9e40ab708644b1fe59bfadb4f9aa
f42d5d15e9e018671d9f16e0c66ad438afd6d0fb
29282 F20101108_AABLLR carr_s_Page_148.QC.jpg
9a6f4cb80405b68efe1c32c0b880bec3
0c9aaf824270ddd18a9b234da7cbf49eec34d9bb
41513 F20101108_AABMPJ carr_s_Page_159.pro
10941606463ac5ad3b8b5e023f79a63f
853708878d95780fcaa7e2e7af25cebe14e410e6
1776 F20101108_AABMOU carr_s_Page_070.txt
d02a8f0b7c241367f075af6fcea13e5e
b8c5dd4f67f2aa9cf74df60ac48a4cb994669340
82403 F20101108_AABLMH carr_s_Page_147.pro
d8ab3fed3219de9615338065f49fdbd2
d57c517ddb2e753ac0fa424869d8d6f8548143dc
75661 F20101108_AABLLS carr_s_Page_087.jpg
bf01744ef003aacd54baa349b41895b8
63090edc7457316eece2780ce1a32ede1ece9c88
75956 F20101108_AABMPK carr_s_Page_131.jpg
705b5ce83c7974112d57573e34a90c5b
80894d5f987ce5199e365cf374cfd536d7ad0041
23212 F20101108_AABMOV carr_s_Page_023.QC.jpg
6147c8119dcf1ff4aedc96acb5388e1e
772f1b4e886feac7dc11547e12c113ba4d1b4c9a
80257 F20101108_AABLMI carr_s_Page_085.jpg
0cd601da6c34e3263bb23d1a40b7f9da
baf9ef2043d78adb321143859a751cc757869391
6638 F20101108_AABLLT carr_s_Page_109thm.jpg
14827632251537c52ca7ec7aab63b95b
3795dd468f78e7fc7150d8e7aea1d77ef71be969
75067 F20101108_AABMPL carr_s_Page_033.jpg
a17301f08000ca73b54dbe5167de0af4
78f72e70bfde63f353e169c67ff76174bc8f61c2
F20101108_AABMOW carr_s_Page_167.tif
961ee36b4f9235050c5ecbd53557cd9e
d93a2b4638b8820bf2b3ecaf167150ee29614cd2
2870 F20101108_AABLMJ carr_s_Page_174.txt
0cbd4b039af50fd5d3303974c2aeeec1
7a4cc15175af0b862fbe5d43ba8336f8a276c4c6
57274 F20101108_AABLLU carr_s_Page_094.pro
5c741e3b83d7812f513c33f83d0587c1
4fb7100de1633d829a573112d1f1da734cbb2afa
2251 F20101108_AABMQA carr_s_Page_093.txt
194c21dd137492eda039786f3f893171
800b92d04b744f146aa4d75d1973effa5a9d0c4a
51902 F20101108_AABMPM carr_s_Page_092.pro
53647553366784fa738058c4a44a5afe
ac5f2d9600e19258d3ee456bda231cebf16193c5
25037 F20101108_AABMOX carr_s_Page_014.QC.jpg
b785d2e2907b1c5e081b0e11ad0bd322
1f05d60ff81b45f608fc82109536fa8ad5019444
54938 F20101108_AABLMK carr_s_Page_088.pro
173ec2e9af0f8974b4b432a3d08aa871
293def10755cd5f6538a1c54f5a64125a092597e
66210 F20101108_AABLLV carr_s_Page_171.pro
549e2d6a06be7c48bf10cf8c13423adb
d01c15a7c38750e16c06b853b576e2756e872c65
17929 F20101108_AABMQB carr_s_Page_062.QC.jpg
01eb59314bf3e770242d93b9ac5fdc5b
7e5e71fc23cbdef1f07dd8413f46c629ada52343
F20101108_AABMPN carr_s_Page_126thm.jpg
deee430bebf8e682e9213dcaef295765
0ea9d9fe527dbe682ff2790c172445b1c930942c
1988950 F20101108_AABMOY carr_s.pdf
7fabc0bcac649ced2493f330caf0b253
a17f0e0367fa3569030146e38b748a5733a332fd
F20101108_AABLML carr_s_Page_090.tif
dbba41720de5c8890b941e200c439a8a
e5fadef137aff6bc1985c7c779e0794ce59104e2
70047 F20101108_AABLLW carr_s_Page_023.jpg
a053d99703eb33208fad6699fe0ebdee
9d4e5aea0a7470a2e3c1ae1ffbc233426d66a563
23848 F20101108_AABMQC carr_s_Page_016.QC.jpg
5293e309ae3f5c362df83972ffef85d9
f13c5467a9518a245df26cd60fdfba9a10e47d9e
3383 F20101108_AABMPO carr_s_Page_061.txt
3c8bb173622e0a042098077684393610
9b80c7428dedc1f9c9cab04048e94d533a34bc75
76649 F20101108_AABMOZ carr_s_Page_117.jpg
7a6a825e9886a66bdda7a38771136097
1f9c7afc729de9ef7adc66e943d826f2591c3b67
119247 F20101108_AABLNA carr_s_Page_118.jp2
f2654a7ce037cf14ad2f98bc507ba91e
2c1a6aa8817bddd47d513f884d72f3f4c0890e79
115208 F20101108_AABLMM carr_s_Page_027.jp2
6d3c7f007e7dd53fedaf1e0c3732e028
f0cd4625d5c33021ca1e86a00376c91fcf823c3c
1912 F20101108_AABLLX carr_s_Page_081.txt
837f2621fe25e88e1cbae90d69ae2ec4
2b3098235f53f3fede1d20e2fd069f27d4256377
1051952 F20101108_AABMQD carr_s_Page_061.jp2
1076d6967ad05f4d198402eca98dd275
31fd773802902dff784816cddcb8ca43f4cc7e2a
2037 F20101108_AABMPP carr_s_Page_024.txt
93494f81a785774e720ce773c0cd8e39
ba51336f9b8a783e835d537bc6a329756a32f090
60634 F20101108_AABLNB carr_s_Page_178.pro
1ff54fe1bb562a0903d1bdab0a3d9f9e
c353852be4a09f7e4d7d203c6033ac2503f93d3c
6836 F20101108_AABLMN carr_s_Page_020thm.jpg
cecc262d9094a2b3f431e4d9feab32ae
6c3c8514cbb4606f1866b754e9da91bf65c397d4
115549 F20101108_AABLLY carr_s_Page_083.jp2
fd2a8ff5593db6e05ef3c0a5aea7c3a8
eec2e41021ffe6cdf087a9c8135f8f0c03309d32
2141 F20101108_AABMQE carr_s_Page_106.txt
899e42fcdcdd0ef304aa30448494b0bb
5d4e12330b522d4c761a36fcc5f8366a727de6fb
24118 F20101108_AABLNC carr_s_Page_056.QC.jpg
591689023c805a173b7eaab0d8ba3d61
a21297f17f71ae29d7a1f214b34762433164f0d6
F20101108_AABLLZ carr_s_Page_047.tif
f74cb2d56a476923a61fcb24fe47c2ee
0911f8cf2461b5ddc4e24f9edbf51d9e11560af6
7504 F20101108_AABMQF carr_s_Page_174thm.jpg
564d4cc01e22613b190c2b8b48d80875
02b08e64566294fa14aa1db78face55f5348b227
120378 F20101108_AABMPQ carr_s_Page_107.jp2
ec0c61c9f5bf0286d2811d3a3131be4c
fede5269d2827bea9bb310914860d46a83bd696a
2415 F20101108_AABLND carr_s_Page_066.txt
2abfc90d3b996c28c35c04181a0586e7
872e1898a3906eedf431800d851adf670591e2de
52870 F20101108_AABLMO carr_s_Page_049.pro
623fbc8dc92a036e5423a3165f7ca0ef
12ec030d5696b01bff617916bd78974efb52d822
23927 F20101108_AABMQG carr_s_Page_084.QC.jpg
061434d13687db1acc5c8ecac16a1d5f
8ad0317814e54ee742fb66096412a2b817f0a8b2
3385 F20101108_AABMPR carr_s_Page_134thm.jpg
e13e00546043fe76cb8b6d2fa5e6352f
dfb089787a625cfe37d7f947e1cff1172daacc52
F20101108_AABLNE carr_s_Page_078.tif
1d7142d69e31db88a98568bd5b66f7b2
492324404037da5f65ffe0ee3a6a0f8df221f34d
F20101108_AABLMP carr_s_Page_153.tif
af413cd32d705bac501c7153629e9357
9c44020e838ace94f8eabbf28de779253c39c36c
22601 F20101108_AABMQH carr_s_Page_031.QC.jpg
5616860f3f5d64ac861ba61642e552f1
5470ccf55b0bb18ec3cff1d142710b405dfdf9fd
3141 F20101108_AABMPS carr_s_Page_008thm.jpg
a8ee19478555cf97315c715be5eae856
45a4e31f62911fcbbd01baa38e059880d40ccf43
41861 F20101108_AABLNF carr_s_Page_161.pro
275b1fc10544c699d507f3c2db123b1b
dfa5f4495f9b0e58fe984712a3ec9fb2c290d255
11745 F20101108_AABLMQ carr_s_Page_141.QC.jpg
1e76d49df64ca36b79866d8b3512a2b4
9c3daf960ca877976ff50d6785da2e83bfb7ff0a
F20101108_AABMQI carr_s_Page_064.tif
446fb407346a686fc023085f8698b3af
4d0a0808ca3348c04059e4484d8b268f2256f3f3
69414 F20101108_AABMPT carr_s_Page_048.jpg
85191a44fea3be3115d408d732e6059b
53ccf30bacd7ed71008c63255a5412210df60b8e
76710 F20101108_AABLNG carr_s_Page_039.jpg
25e5306918c5e993f1332dce07a0e2c7
a1c06d28aa7d159a8c0c61214d60fba8afe66d19
75983 F20101108_AABLMR carr_s_Page_129.jpg
d075f478fd4949ce17e00af898438625
60021a35929630147bd3db35de572828a74f2226
F20101108_AABMQJ carr_s_Page_113.tif
a097130f4f5d25e013b301db60f8a840
4792e0514e5014d464d5018cba232fe2b6d863ba
25657 F20101108_AABMPU carr_s_Page_076.QC.jpg
cd4cc417ca52c37dcba26ba44e383ac0
06a20c2a4ae95cf75b1af1322d25bb2e2ba095f1
85320 F20101108_AABLNH carr_s_Page_163.jp2
5dd0da08b23f6528dcc9b0c9d7992b8c
76afbfe64f4427f76fc900d1e05f646a9451add8
53149 F20101108_AABLMS carr_s_Page_042.pro
977c9fb314b57e54399f466542c6dfa6
b04e698f2047565abe3cfa3d7b14a8204e3e96e2
1989 F20101108_AABMQK carr_s_Page_004.txt
a3695e45681c89b4bd4eb1b8a0c15a45
d96bdc13ba940a152050b2b29bf8443f7a9ff69f
F20101108_AABMPV carr_s_Page_032.txt
548a10570aef2e069c1164987924817d
bc7d8284fb24d8a7839507afd5724086931d401c
104335 F20101108_AABLNI carr_s_Page_023.jp2
ac1ac1c1eb5334a801a958d9e1e270b8
b94c13dd7799bbb86faa26b39ecbe4d06cdca278
1928 F20101108_AABLMT carr_s_Page_119.txt
12b546bc8dbf2b96704e7415a12125bf
74a2fd83cb0947183a2351d690d642a95fab6091
52847 F20101108_AABMQL carr_s_Page_039.pro
4e055bf0a8e8eeba8d0b392e4c65b338
412f9d3750fce94149469949bd93f0865107b64c
1051893 F20101108_AABMPW carr_s_Page_074.jp2
5c50f44b048c1a0cbdc1d7541fb5eb29
2a759b96eff2ace6ce07dd86705161cce1120fde
92118 F20101108_AABLNJ carr_s_Page_171.jpg
7e584c2ce4b9553bcb00563b4f69d288
83192290ecd12413249611a0d08c80d5496dc864
F20101108_AABLMU carr_s_Page_049.tif
9860109d65d8e387f03f153b706636eb
f70bf77ebca1e8b356b05ba9d195056805ad6e45
2203 F20101108_AABMRA carr_s_Page_075.txt
228d92e08b79b0804c2e1e684ef4706c
8bf7b24fd6f5410cc0335d28da90ea0a3cbf2213
2199 F20101108_AABMQM carr_s_Page_127.txt
e387332d7970d959f692a6fa6e388678
dce372817a5fde45df85492b00264e42d3fe9655
51016 F20101108_AABMPX carr_s_Page_034.pro
15c91794650fb23b6413758531a4aabd
eccce736b2fc3f85559528020e60ca3ff8ac7c78
12343 F20101108_AABLNK carr_s_Page_153.QC.jpg
0feca313e4adf2121118b5b5495d261d
96fc20d08d62139dd56044bda5939617d8a4e4dd
31814 F20101108_AABLMV carr_s_Page_067.jpg
9dc5bd38fcde0a304b05a4b1b2398d6f
6433153e1ca29b1955f857eb96d350cc580ebdd4
73475 F20101108_AABMRB carr_s_Page_050.jpg
578c28236aa3a8afeac01abfd28d8630
ae6d787970894cb51d0277ae514804f82f4a6a57
94094 F20101108_AABMQN carr_s_Page_132.jp2
a3a787c1d0fbee37299b635dcb9461da
88b7308196641feb1df4443fa69bbdf0ab589cf4
23988 F20101108_AABMPY carr_s_Page_028.QC.jpg
8535b99efbc1c4cd59fa7d6f5510d140
c57bc2afc1c31af451e23e2f819d674bd1ad1c24
F20101108_AABLNL carr_s_Page_047thm.jpg
2045ea40deddaa34f9523a03193d12c3
91576ccbf74501c8824f2143048ec0fce249dd00
69093 F20101108_AABLMW carr_s_Page_010.jpg
7dd4d2d5c4b4917eeef30bf8427c2fb6
372fd0a1306aa3157ccc3b74b54bb09b473fb156
6807 F20101108_AABMRC carr_s_Page_034thm.jpg
eeef50bda7469fd7313e8668b0ccb879
ebd1949aff1288c83e767a500cee5bb621686ff8
F20101108_AABMQO carr_s_Page_140.tif
0fde366028216a7e8e1764cfea453a27
e4f08d51dba786f02c15c9e0328dbe47fffff284
2017 F20101108_AABMPZ carr_s_Page_125.txt
e1339f7ae38ad86b3028c8b72dbf9852
0a68cf7ea9122965a9c1c2673c346f0a2ba51e7a
118998 F20101108_AABLNM carr_s_Page_035.jp2
2c4e2c1a7d7bd187bf02263f55676101
d2d717df1d26edb1bcb36bf5b3007506c613e4d6
4983 F20101108_AABLMX carr_s_Page_006thm.jpg
e3a8c0663d872927083866648eba7ebd
b2c391fda3785f56b13333fe008cc149dadd6b75
76238 F20101108_AABLOA carr_s_Page_061.pro
f0c9accd840bb226eb6fac3f9abbebb1
929ffc9f5fda149232b2565c621f68da859dff18
6124 F20101108_AABMRD carr_s_Page_003.jp2
3ae68f2f1cec19d5d22e7212c05fdc3c
c4bbbc14b768f8a3692fc414c079c8133f0ca73b
24993 F20101108_AABMQP carr_s_Page_039.QC.jpg
88de68fa8aa4787f981cb9561e3d780c
fc7861cf0e4255b383edb341c92b1dc2278ef6c1
3668 F20101108_AABLNN carr_s_Page_099thm.jpg
29f2fad775faa11c5369e24a653533eb
385e5bf8cacdb11752d263a82a92b5824846e86c
F20101108_AABLMY carr_s_Page_092.tif
3040e3457d3340873601788648e841f4
83a5aba1aab36b74f47dfb5bb273d8ac88a430a6
F20101108_AABLOB carr_s_Page_148.tif
ae5960e2edecae246b220eae4327fd2c
54b616ebb880ede3a575fba9cb3a0265e83734a0
15312 F20101108_AABMRE carr_s_Page_141.pro
e3ef34b8b72d5b8bcd7bd1acfa647360
ab97bec89ff037e890b0a0037242ba0fd619c77b
F20101108_AABMQQ carr_s_Page_114.tif
1d6393cbf68f941b0605520e4287b855
bbcd0dfdc762af5c305617c91a7f3666c9bd80ac
650 F20101108_AABLNO carr_s_Page_102.txt
d9f724e7b361ae4edaa3a07824aad211
21e621a513486cbaca568d1e488a31a7ddb18d79
29308 F20101108_AABLMZ carr_s_Page_146.QC.jpg
4c2ca8ffd990fb6fe0812670b5829298
a4bdaefcb81a4726fb529b372a6df65578399e29
2160 F20101108_AABLOC carr_s_Page_069.txt
ffbd735317f48cde0cedcebab186bb6e
be26706cf8b9a474ea599d85990f95a83f27793b
18038 F20101108_AABMRF carr_s_Page_139.pro
020ddc3aed5bacacfa4138920046090a
b47711121213215496c2d5630e405d7717473b83
6947 F20101108_AABLOD carr_s_Page_130thm.jpg
cecebe1db0b50fc88438f51f3729fb5c
78ef44d23e29b20157cb0efeaae27e1eac0da49f
161641 F20101108_AABMRG carr_s_Page_146.jp2
59eb2dbec9f3aa21c0628d6c7ef07dc0
d19d7f31208cdebc82fbcf8f9bcd02a7125e937b
F20101108_AABMQR carr_s_Page_168.tif
735e3343af317d13bf0a7afaedb10ae2
11cb60db94159e52347175dde59bc25effb35ae4
110050 F20101108_AABLNP carr_s_Page_108.jp2
4a795d978cbdc02ffb07679dd9452c28
93d4cf7e3da499c804d82b50c739e8ef89197afe
46680 F20101108_AABLOE carr_s_Page_040.pro
d817d9ba022deed93d3cbe6e92d7c8ef
545886d5cad38314a95b7463b7fffbf30250436d
F20101108_AABMRH carr_s_Page_131.tif
4783bb1bf4842b4dad3b77099c9d3999
fe99d995b1387a091a9167629884972f3f72c5f0
51618 F20101108_AABMQS carr_s_Page_028.pro
c6cadef20d23395f20c2cc6828c9fbe6
aec8857cabc18b074fb3ed485cb183c75f43f8fd
117924 F20101108_AABLNQ carr_s_Page_037.jp2
a1811f83a5c94aeab6727804adc7f76d
cbf8ed6bb7c95d4116a9fb67aada907747cf0f25
F20101108_AABLOF carr_s_Page_051.txt
e3405b072fe8c039eb19577f632fcdb6
314b257cf29922cd21cc644146562e77314a7384
114027 F20101108_AABMRI carr_s_Page_126.jp2
17e03b8525ad7a239f5c9de9099b4166
ff3d21be4c6cd87a7c4996d79739904c7bc7b78d
926 F20101108_AABMQT carr_s_Page_142.txt
106c619d5b344c760eb2dcd94dc3d5ae
8c1eb951b7898aa3be8941b9a12b44492b2fa6e0
383 F20101108_AABLNR carr_s_Page_018.txt
fa7288eb8112a3d320c5b4459f5ec86d
8cfd9f40dbf78bdbc3faa39b50871145c5b5ddc4
F20101108_AABLOG carr_s_Page_105.tif
69ee991fec2ca040c0cc29a3cb6da7b5
12d88c679b94bffcf34ca0142e2069797178494a
34749 F20101108_AABMRJ carr_s_Page_167.pro
f846cac26d4ca0fba50fca2c388c12df
dcbbc1d7063af335a4b28532098bc7924f26f97c
F20101108_AABMQU carr_s_Page_088.tif
1f384af51ba5349458cd8ab1ba159985
4040cb417c11a65aa93e838e8c0025df3847a444
113464 F20101108_AABLNS carr_s_Page_111.jp2
4d4073f449e61468ff7e4875877d3155
75cca1d402a2dd95eb44b3cb70c96b6446953390
52586 F20101108_AABLOH carr_s_Page_035.pro
ae2127a41081ec50be62e5935b44e8ad
2ccadb9611dbd9058df940c266d017db6c1c9128
78058 F20101108_AABMRK carr_s_Page_042.jpg
541cfd46b6bb2892d1466da2697f49c5
b5d1e3f2bd56ed579d1f06cbc19a570584b0ef69
76040 F20101108_AABMQV carr_s_Page_116.jpg
b8aedbbbfc8d9a30de4a5b3859299e08
ba635fb4f4c2c73d17e441f2e8213d4df374b180
94646 F20101108_AABLNT carr_s_Page_174.jpg
0a9ca76989d848b17ff2f121c667d311
347941368a22458ff6353c51c6c17ef778d785f2
2667 F20101108_AABLOI carr_s_Page_068.txt
0ccdee015f019125a9535ad35856445b
edea6e4178cea11c9d6f2c2162db5488a0e5b56e
6321 F20101108_AABMRL carr_s_Page_046thm.jpg
8806a997d76bbe9bec5e526c8e9ce120
a097138d1520801db955ad842795e41f1fc68b10
F20101108_AABMQW carr_s_Page_009.tif
ddf1774367284d7941ac6db62dbf297d
b6a817f015da1c822c3b87f32f212bb6c63ae340
75024 F20101108_AABLNU carr_s_Page_126.jpg
744d5302b1a3dab944c23e4cb5b564c3
f47542681fa9c791283ee481d869d8ef9ecb886b
52592 F20101108_AABLOJ carr_s_Page_019.pro
ea29cd6fa1f5a219a0bd6965b1312034
ddcd110c917d915e8d4db72f65c8cd60a4b11398
127195 F20101108_AABMSA carr_s_Page_178.jp2
88ada9c72a7f8b8dee836ac8c1bc3b87
4cbee3b0e9539c08934745c5c22b7832fd059287
31952 F20101108_AABMRM carr_s_Page_134.jpg
ed8ed7afb0e249586ed64762bdc540cb
7085c65d6b83b030088ad2989e34c75c33acdf12
76588 F20101108_AABMQX carr_s_Page_168.jp2
4c195330c5568bca4f5feea57f49be0d
11c452683be757be606b4f4d57aed30c35538da3
810061 F20101108_AABLNV carr_s_Page_103.jp2
2aaf49d41ce2f797bbc16a9fe0abd2e0
3cef0aa3784140ec358d763eb6041d2a9799fe24
24515 F20101108_AABLOK carr_s_Page_013.QC.jpg
e34bebddc6d14ad6c39d8ac718c7266e
6e1facc6064c36184894e68995f0712216731dda
26005 F20101108_AABMSB carr_s_Page_112.QC.jpg
9267eaabc161219e9567cbe2d12caeb0
df083ea394053d9320ee29a82025e6c3e93ae303
F20101108_AABMRN carr_s_Page_058.tif
42dc76fc8bfef7e6c0535915f22024d9
8417c10d3b2c4152ca5e3abd0f34f482c9fba5e2
2023 F20101108_AABMQY carr_s_Page_034.txt
16cce023250303b747e3a85ee9efd83e
00209c2150816fecef3d4a8e9d71a4ae99490ded
52648 F20101108_AABLOL carr_s_Page_131.pro
42120cc8fd73f0f75d66339db96d308a
3eb5fe36b930d95e8f60c99f2507aeebbad00ada
9273 F20101108_AABLNW carr_s_Page_067.QC.jpg
31042a30203f57fc913f75f5b83228c9
4392f36df75143b652c4ddfca0f71570c19fa1bd
53601 F20101108_AABMSC carr_s_Page_116.pro
b11b063d391b22c7d042c71dcdc38ea1
5b5410cb3c272336b897bc4e0630c7533b469a23
2321 F20101108_AABMRO carr_s_Page_018thm.jpg
fdada7ab70184574c1c3720ffa21d958
eb6d54250ab0064284a2adee5b88bb403129a5fc
2571 F20101108_AABMQZ carr_s_Page_065.txt
2822960f8ef8c12fc705161cde8165cf
2d802912b0ad731f225f0ee3f36336d7d68e2bb3
32562 F20101108_AABLPA carr_s_Page_063.pro
8cce53d95dc3abab6f8139a8afeb324a
b776ff4fea92f9bd47f7e3f02b25d12190f6d111
F20101108_AABLOM carr_s_Page_110thm.jpg
f3bec5874c607a20a22ae49dcc0333f1
fa3831514db2b0617647105621d864017e9dfb7d
1999 F20101108_AABLNX carr_s_Page_036.txt
17f0ab86d96590e4d246320703f9b1c6
8885964290e116016cf03dcb9fa3d506dd508076
198 F20101108_AABMSD carr_s_Page_135.txt
040c4aa60093d606e75dde39e055b701
25acf48e87f98b6f4b05aa02341ae1e1e63be349
5055 F20101108_AABMRP carr_s_Page_009thm.jpg
48a2cfea495d81979cf9fa582892628d
47e71bff498ff86dee174c242c672760d6ad7a9a
72076 F20101108_AABLPB carr_s_Page_120.jpg
095c6df5bb9bd28f98c529d9990437a7
9e0986ed989620e45509b03e4302cf2590b5640b
31459 F20101108_AABLON carr_s_Page_169.jpg
b65f04c87ed88362445ee80c8ed0c289
71bc463ad8f7e9a8e16b960b5109d6d9664039a2
15136 F20101108_AABLNY carr_s_Page_071.QC.jpg
481fa8e6506b63782c9ea72722bb62dd
621d173b3f44faab2ad595cd5bbcb9b5ecefc2cd
12139 F20101108_AABMSE carr_s_Page_007.QC.jpg
3af6f6b1bcd5c095b348cdb7474feba6
d88712e5f2b5256db8558fc313f45502920d8d61
7102 F20101108_AABMRQ carr_s_Page_178thm.jpg
ce2e4c357c99511f285e29b745a5db50
8450a4db7e1fb1f2d70a8d2e745cc1ef9045f7aa
115399 F20101108_AABLPC carr_s_Page_032.jp2
f2d409cd6a5c5db1b48816be26135daa
d2b3359e2fec874925358f610bbbf82716dd892c
74706 F20101108_AABLOO carr_s_Page_017.jpg
65ca238f6f880d68792ed61d79c76d02
bba911fdf5f63c036827672d9df85050c15fbc87
54002 F20101108_AABLNZ carr_s_Page_099.jp2
f06d38b54562f19e1f5f6261e8baeafd
eeefc7b8f7c297000dcf44c816250b5f4670c25d
45136 F20101108_AABMSF carr_s_Page_156.pro
07766604962b4ebb29f9c4ff2dbf320d
9e90df126ae1f76eda75923213ba9ebd59cab7ea
2081 F20101108_AABMRR carr_s_Page_121.txt
5c8d276c20f0b4bf978d13666db788d1
0143ca04afd0ed0b4817ff13bf21e4b87cc3df71
113309 F20101108_AABLPD carr_s_Page_044.jp2
4cf0ef2959b54cf171e3563e5cbc298f
db203c7adc0181597e0cb69b83458559d822d7d6
14967 F20101108_AABLOP carr_s_Page_161.QC.jpg
a786ca7ab495f9444305982351e73834
9aaa4b513ab3067f67a2a58e94a0a73f03e4bbf1
118055 F20101108_AABMSG carr_s_Page_026.jp2
001cedcfd733525de2d077f31ac21e13
eab726253c51d0c8c5d92046d6e683a64ca6528a
F20101108_AABLPE carr_s_Page_034.tif
169be4cdf600c4cf418b04a0791248b7
2614e8c1a793edcff5327be014a685e555a97c73
3355 F20101108_AABMSH carr_s_Page_152.txt
81e7255eaa5d661be18ce70b27f62c47
260b492176c3a92cee0af7d474232a0eaa64b45f
52167 F20101108_AABMRS carr_s_Page_013.pro
ba5c0a71b28a171d2a746805546a97f4
3bd11711093d9c4b1190e73e8ba5427a7745e734
F20101108_AABLPF carr_s_Page_180.tif
ca31b42d730278f37ce1c9e2c788c9f3
b26c91072eee8eeb04f7a177ae982570a789afbe
F20101108_AABLOQ carr_s_Page_073.tif
96560d336a86d3528b7cfbf5ca99f66a
0768510f2ba7797f50647db65d7241bc42f6f4ae
24828 F20101108_AABMSI carr_s_Page_001.jpg
89cfa5764ba082d6ae1e07ded4ea3fde
594b8446b506bf734c0953bd06ce51a5e199010f
24469 F20101108_AABMRT carr_s_Page_018.jp2
92fff9ecb8f7f254b8e6a0d4a46afb72
3531f5c5f50178b14458a447aa9234ee749725ad
2204 F20101108_AABLPG carr_s_Page_077.txt
143dabffd3e352f04cdd6b3de91a24d2
1f5cc1d9ffb6b2adeaaa06d3470c9ee6b7090797
66165 F20101108_AABLOR carr_s_Page_069.jpg
4fd57b6fbba89fcbc9b295f7c25cab05
65552062ef8e358732341ee47e129fb494ace946
6611 F20101108_AABMSJ carr_s_Page_004thm.jpg
e97c21560b88a93786a32fd31faced66
783fa5c88755dc862cd840889fecec7dac60ae37
6662 F20101108_AABMRU carr_s_Page_032thm.jpg
4decd984e9c340a9a7f52e868ee7a128
3ec67ddab293aba3efbfe159303a517fae5ef36f
6724 F20101108_AABLPH carr_s_Page_143thm.jpg
33d2188e0c55332d0c0e80697c7d067f
d5176535ff32454fbd40e06d14982f1a7308670e
27755 F20101108_AABLOS carr_s_Page_173.QC.jpg
bacef142731221e025e1538e16428994
19b07d90ca530049d6eb38e0053aa4b67b9809fb
F20101108_AABMSK carr_s_Page_003.tif
55fb3552ad7962206b171bbeb54f8391
8b6b84335edc891b725553f4bc39febe36e5d899
7256 F20101108_AABMRV carr_s_Page_152thm.jpg
076c0a7d79147018ac42550f36db04d6
524d5c830ec7310326bb0140933abbf207afde20
2474 F20101108_AABLPI carr_s_Page_178.txt
4e49eeb90a0ddc4c9a22b38a154fdb91
32a0a310e9b1e29e959bb7f4e1b5229de0c9d527
52046 F20101108_AABLOT carr_s_Page_051.pro
b34b0db43f610909a9a7b40266884c65
4f06e1181ecba2e3d79e18efe045ef303c10ff9a
44798 F20101108_AABMSL carr_s_Page_005.pro
ddd601113914d18f02001ce60743dca2
4e4585546a90a39470ad4c754ecd45f748d5d919
1916 F20101108_AABMRW carr_s_Page_031.txt
1fb6f4fe806c048c1ff9f9d8bfa57ba5
fcd301a484a78fe9520bd9baa13fa25651165c7e
2009 F20101108_AABLPJ carr_s_Page_044.txt
01dc7d4c768fe2220d5966e43ce18f41
70da13b404d3b01cacdda2606082cbbcb56577e9
1948 F20101108_AABLOU carr_s_Page_080.txt
cd6fd26bbd819e51228c307652d6bd05
b4552922f57dccba4947578f0b4500734356a452
14586 F20101108_AABMTA carr_s_Page_159.QC.jpg
465710a4ffc3302d22886bb7b63c6e55
c6d76b66b9c5bf528bd572fdb084d6cb07add38f
724 F20101108_AABMSM carr_s_Page_002.pro
8a482eb0c34c78a993d7bea9342c8dc7
ce209eed5d101f9ea9f9125a18186682d69bfb74
1348 F20101108_AABMRX carr_s_Page_002thm.jpg
b82f516d1f5290eabfa9cb8625f21c90
a6cda963b8a5e39f195c5d84ae21271d00d5cb8d
F20101108_AABLPK carr_s_Page_165.tif
5549a7e485d4875fc0352712ae970f51
e6e7b8db5ceea56a552d6248e85268ac1af4bb64
108885 F20101108_AABLOV carr_s_Page_079.jp2
7cd68c07ffaad7cd85f116758bb1f42d
f8c2dd55331a80ad03026cb400e49c08bdc7f8c4
2087 F20101108_AABMTB carr_s_Page_039.txt
562c119d383a1b2ae0bb69f591368bc0
1f335c611df713902ba2ec9038dfa48633da36dd
5084 F20101108_AABMSN carr_s_Page_074thm.jpg
bcd061397399c1a0dd032be5a3f5e985
288f94c7f711041fb3da6514fceab43c0ea93944
14464 F20101108_AABMRY carr_s_Page_157.QC.jpg
17b6a359817aa401e8ef4231890aef84
783616537fee528116602f854c0c776d53c641a2
79705 F20101108_AABLPL carr_s_Page_115.jpg
eab18ab1495b58ffe342efd2af7fe8f7
59c9dab32ede98a191624461841730310862b031
76911 F20101108_AABLOW carr_s_Page_164.jp2
1512b655a1dfedceb47ed4e061add848
14294d7aeba76db24d06cc6a779c39262b748c55
72663 F20101108_AABMTC carr_s_Page_090.jpg
8030a085dc0ebd1c88a5b8fe323e679e
ee2e0283491b8714a35f7717a0853deb1fba02eb
83654 F20101108_AABMSO carr_s_Page_146.pro
74a4069442af63a7ce3940c4aff874c5
60d403a7c01cf5fa2f47e2a98ac732b6cfdd1229
4419 F20101108_AABMRZ carr_s_Page_156thm.jpg
30a8ea2e413b923dadb1c06701cc804d
ea2f5cb25b4891b2a18d71bc7d90c544cfb5b6f0
2722 F20101108_AABLPM carr_s_Page_067thm.jpg
25be93bf9f7f116013850539f4e8c239
fee15bcfaec04fa5fb286cc7a7e3618e01818f0f
115096 F20101108_AABLOX carr_s_Page_030.jp2
6da95397660a3e9413ca99c6e2a38378
6d4627b0d67ad804ae4c359e3a4c2a354f6e24a4
25075 F20101108_AABLQA carr_s_Page_143.QC.jpg
ce3eb8e9c95abdc95f85f9974f361b33
da3dbe153b91b542d54aaad8422d2ff0708e3be0
F20101108_AABMTD carr_s_Page_037.txt
e0648ab68fd83127d68ae7f07097846d
8dc62ac29218125bcc16be0066d6b7ef76c59eee
61285 F20101108_AABMSP carr_s_Page_177.pro
c6cace1f51355ee32012e4582c5830e5
03c85ef2449316e9444b5570a1b203c07ad1cf2c
45568 F20101108_AABLPN carr_s_Page_184.jp2
dc2d83b9765fda544959d2c967d6cc5a
1350738b29bbb77a34292ee92d8a298af98cc81d
46485 F20101108_AABLOY carr_s_Page_010.pro
e9b2e30223151e3ab1f640ec015f1436
d9b57f639552cfdf8bb09bac9a297fc9d298057d
23922 F20101108_AABLQB carr_s_Page_052.QC.jpg
8d41fcbd9791d54a2670bc7044935a1a
cdf2fc16c9eaa65d396259fb6693ae732c647f4b
116937 F20101108_AABMTE carr_s_Page_143.jp2
235fd9cf2e50ee8f6fdd2a50eefde711
b6789ed13725f8e14b97231b7d5de1d6f1da886d
F20101108_AABMSQ carr_s_Page_018.tif
d52b1ba444ae01b4ae403907282f679f
2c8684aaa499716c64353dcd8379e77d88663274
115012 F20101108_AABLPO carr_s_Page_013.jp2
85cd3c11f6115984bd219d18d639eff3
835b661362a954bfd72891231bba794910d830f0
14059 F20101108_AABLOZ carr_s_Page_164.QC.jpg
b5ddd5777eb0a16613b2ef61db3a3860
645b00b169a6e756e84ebecd3f937d25633cff6b
6240 F20101108_AABLQC carr_s_Page_114thm.jpg
dfb45c13305ee04d17f0dc74f69ef36e
fc2fa80ea7160c3222fbecd46c665338adc05c0b
24490 F20101108_AABMTF carr_s_Page_032.QC.jpg
b72122facc43f0883bb03278bf35ce57
d91e155a89be8e984aa092a5046e2d10ec783148
24774 F20101108_AABMSR carr_s_Page_087.QC.jpg
5a433a67b93d9bf4b36270f6d8f880cc
2eb293c55a31261d4f52866b75f799543d2f6cf3
F20101108_AABLPP carr_s_Page_154.tif
fe99c570a122a6aaf3703f57247b3f5e
69ba6f3f00e30350a5f8d223008e22f40e552af5
F20101108_AABLQD carr_s_Page_141.tif
3ac304b3b04796bab8c1ceed03941854
24449f44a52401a548328f7cf10a52f19623a881
6392 F20101108_AABMTG carr_s_Page_125thm.jpg
4f66e11894fe53d12364b95e55600e16
0087986f5935df6fb054cc73a36ae80e4d1d1bbf
1994 F20101108_AABMSS carr_s_Page_108.txt
f631491835fea7dd84e7b949fac8c9ef
0c694f96c48712bc7c4801cc5bfc274a228d43ec
6843 F20101108_AABLPQ carr_s_Page_055thm.jpg
e15a70339c6daf3a5a651b3a1bbdc906
a0e99c1a521f5c654ad696d9192e8cce917709d0
8362 F20101108_AABLQE carr_s_Page_073.pro
de4a813e7e1ca665e12fea6ff4108130
4bfff8d5df0c0ea48d6a409009ea18fe8ef0ec2c
F20101108_AABMTH carr_s_Page_087.tif
2286fbe521f20d7b1cdb73436650e9ce
375127ebf29214dc966335fd870a671a07979d2a
110610 F20101108_AABLQF carr_s_Page_052.jp2
c8746939ed3ae1b07d42c70c24cb2b05
1cc4e2739827e7a390eaa1f0d8a6dba4666bdbe6
13914 F20101108_AABMTI carr_s_Page_155.QC.jpg
04d10a8768205645443591167d631861
01679dc4cda2436831f3522a2be6054b3efea85e
6793 F20101108_AABMST carr_s_Page_037thm.jpg
6c54b8bc2561ca4bcf171da8fe2e01d6
50929b4614333b21e9b5223b372012d04d1c7ba9
157651 F20101108_AABLPR carr_s_Page_148.jp2
1f6fdff736781eef335095577832b753
6d100c74b2ff68ed66bb87265a080c5461fa8cc5
93082 F20101108_AABLQG carr_s_Page_133.jp2
2093acd65f260035fe3959ff1799f703
fc06b1bc42bd89e6cef7d3d6dbd566911d8a23c3
8423998 F20101108_AABMTJ carr_s_Page_136.tif
79ca22943bdd72293b6fc2741c1ae6ae
79c38b554a1d132826dd9e7a613bcfd6aa801b5c
840487 F20101108_AABMSU carr_s_Page_104.jp2
629c8e19ccabc305d3684951b18ef0b8
774b028425726dbcf6ef9607cd8c62cf2e61cdb8
110137 F20101108_AABLPS carr_s_Page_046.jp2
7616a8991046cd3dda3d4e099c2158ee
2b04ce9230bcf09def6cafc901d39d3c6054ad93
50519 F20101108_AABLQH carr_s_Page_025.pro
a08c104dba7261471481a04a32f32601
c0d36d89d07408f9cd341acd4a647b1961111da3
1855 F20101108_AABMTK carr_s_Page_040.txt
764b34c4f3f332b01081867f7f863f55
a47a9a7c10a54a05706f7cce30197bb8c9563715
F20101108_AABMSV carr_s_Page_088thm.jpg
5ca4c7918674e6cd956783e3c420ba3f
868521b6c231c7b61ae98f04356c809e8d95687d
112198 F20101108_AABLPT carr_s_Page_050.jp2
fa493697813b86c965fd507c513bebaa
b1547df132dbee32707fc94674a5562136c336eb
6803 F20101108_AABLQI carr_s_Page_129thm.jpg
fbabb3c23526dcbdccb04c6762f50933
1a19a10d2187f63e385aa70d33580215870058ac
77640 F20101108_AABMTL carr_s_Page_144.jpg
ba1c79f36ae8719a93257066e6ab9553
5394b8e024a54868c89ebec37b6832ce3b8316f3
25604 F20101108_AABMSW carr_s_Page_037.QC.jpg
2846b21e137ea915e266cb5c7537fa1b
50f636017094e55fa3ebe570bde26b96f2eb8805
F20101108_AABLPU carr_s_Page_021.tif
cf7b3cc7b22e2e63442eff46f98a6941
5bab47d0ac609577646e301547fd1e080ed0d9b4
82613 F20101108_AABLQJ carr_s_Page_151.pro
626bce0f97906a930b55df124077d013
ecb8b7b74cf58c13470d0b45db54148a11a4574e
F20101108_AABMUA carr_s_Page_020.txt
6fb9ee4f27afc40256fd5a475362fe38
b65575168631ca5d23140028655790e20a608c55
16133 F20101108_AABMTM carr_s_Page_138.pro
4dfebbdac714b8617ba560170b5297fd
1b50e32775a356fd48d8c2d8726387cef286e003
761 F20101108_AABMSX carr_s_Page_071.txt
4493f4e80498639f37f83fe2bcae53d7
b3ef821e82d15ae6211965a444513ae7788eda56
6819 F20101108_AABLPV carr_s_Page_038thm.jpg
8138fbd40bea4c0787700454eb27fb46
b73cc31a7328ef375da22da6470ba7745ab88ef0
2008 F20101108_AABLQK carr_s_Page_111.txt
de12af4392b0710a9573af27cc099702
21f2c749e95075bf62be160a9fc705edc0da9120
58950 F20101108_AABMUB carr_s_Page_175.pro
409577efe6fcc5a409cc2d242c17dfea
5b109d8c5273cb3c17cd15f6ca57df6712bc211d
25310 F20101108_AABMTN carr_s_Page_083.QC.jpg
9e35b8b2b8d28fd9e59bc6a7b130c0ad
7bbdd0ada83c9a3021a4bab80ea71a2198008961
1930 F20101108_AABMSY carr_s_Page_045.txt
6e7e4b76fcc951f05a6d325f1809927a
5b5b53cdba0c1ccf7e3de9aef096aad4142e63f3
5590 F20101108_AABLPW carr_s_Page_133thm.jpg
98849702396b2b3031c24b5430ea7a91
f1a1c06b3ff84fc7b30a0136af81611964167085
F20101108_AABLQL carr_s_Page_028.tif
1dc6dddd8b128ce8574e0a21d92456e6
c39d22d1966f24018c6d21908858dbb05692d4e9
1725 F20101108_AABMUC carr_s_Page_155.txt
e6021a334fc18935423f0580b6ef83de
144566eb519311c099c7a9b9ce3a3c8df0cd4434
60258 F20101108_AABMTO carr_s_Page_180.pro
25973fb6c0753c4a1c692bced7248287
9208f63f07d0c835c8c4f3272bad39f0920072b8
1030 F20101108_AABMSZ carr_s_Page_008.txt
a9639623e85987257510b2fcee5b28ee
113d7a84ba591fd71193a862e81e55e26240e4c7
48729 F20101108_AABLPX carr_s_Page_096.pro
e0e93ac493ff9d8a54951e397097db4e
79f1d18be286a37be83447d8e89b453fff0446da
16998 F20101108_AABLRA carr_s_Page_104.pro
db26bafe1b2d1d6cd215111291bb30a5
55e6d5e18cfec2c16ae204f088ab5b1f1d284b6f
F20101108_AABLQM carr_s_Page_012.tif
596905612d86efa96a2ed2d90b5f67eb
50e61d931fd596984bb706c18cb4e65ec60adc11
51944 F20101108_AABMUD carr_s_Page_083.pro
86f11e2e49690d282642a454f9cca0ed
cc77a7974649dbc4fb7eafc99f05fec82c8e5443
10982 F20101108_AABMTP carr_s_Page_060.QC.jpg
179d38163dac9316a65847a05d10e66f
b2e9d2b1f4fcc7020de737e859cb86d687457cda
54417 F20101108_AABLPY carr_s_Page_106.pro
2b174bedf6095222d6c00f5c234b5416
b723f587b7dea86514bc69ddbc194a091fa46cf6
75544 F20101108_AABLRB carr_s_Page_083.jpg
a566d96a4f70e43760176e32b8188b03
eac19306bd43b025197bd383d8963ea3c4de3774
51948 F20101108_AABLQN carr_s_Page_110.pro
832965986b6500c239bcfdb43b0db825
5143cfbd1e7700604ace944dd61c4e313281fae2
6306 F20101108_AABMUE carr_s_Page_119thm.jpg
7988499c7e24726eac4d162f21256c91
2bc6c3102b16abb0c97dc6c3df2e473c74aeaea4
114695 F20101108_AABMTQ carr_s_Page_084.jp2
fd610e354c20ecce215e89c374d3f960
e85da31284c2afd16ee93da2a01bc3149bbcfc4a
79724 F20101108_AABLPZ carr_s_Page_025.jpg
611daf05dd923d96acaa5243eda19953
73ca6b1c846ca7ad80c90b95fbfb415530f4ff4e
F20101108_AABLRC carr_s_Page_048.tif
a46a498bf8f6e8e8c649d34bdaf861cb
ca37641732da93025f9a85ab6f4ebc38936d3126
22898 F20101108_AABLQO carr_s_Page_114.QC.jpg
a46a5cccd192029b02f94b3a0124c9cd
0b23d763657129c0c11b6a226cbfefd3c2990775
75593 F20101108_AABMUF carr_s_Page_015.jpg
91f675f4498c211b125d4be9aecef435
32796c33c60f947c4afd7107d838629b4717e684
814 F20101108_AABMTR carr_s_Page_141.txt
269388be92d485ac569faef9be277b30
7473697360c9e06a7eab6e0512208c1800583e41
52425 F20101108_AABLRD carr_s_Page_122.pro
f1187ef0f5b2bcebbd098dc5cf50f7cc
13ae7280baabf2738477d5d130b17c167f2c3ace
50463 F20101108_AABLQP carr_s_Page_108.pro
9d1645ac6df484dd2268c6624782cfe0
775fcd587030b0f7f2050a8025de04be9a15f09e
114435 F20101108_AABNAA carr_s_Page_087.jp2
d5fe3806355ba0009e9c791d2cee1d1f
94dd24d738aac3326b37d5b3f034b1d8496d97a4
F20101108_AABMUG carr_s_Page_169.tif
7d948f6a95f6d2e1018e99248fded838
741b9dfba087e9eee97f94a30cfd4c0210f86fb7
23211 F20101108_AABMTS carr_s_Page_080.QC.jpg
6dbb167304e98c302cb1c1ff73556fed
d8dcd93528afce6596ec16c9d1872ab4e29834f8
25066 F20101108_AABLRE carr_s_Page_095.QC.jpg
1ee45dd7d52df55119be6e7db6923f98
65a42e080986e03291b512bf5d009dfc50745b84
154422 F20101108_AABLQQ carr_s_Page_147.jp2
37deb03b9fdde56660c745e2b00a9c8b
5b92829d02da84a2b60f6c3063b4943e76dd61b0
110710 F20101108_AABNAB carr_s_Page_090.jp2
37b4738279d070286cc14ab3c44bf017
4862b192555ce47ed12e7126d18340f8d73f61bb
31800 F20101108_AABMUH carr_s_Page_098.jp2
b4e719ecfe9b0ce63949515e60cb6aad
71763a7b7611a726b913cac53f470cb9878b44ed
118077 F20101108_AABMTT carr_s_Page_021.jp2
66d186a73c76809af6e637bd27351a8e
bfbd93ac45849d523cc045ca18f54583a4b87c33
F20101108_AABLRF carr_s_Page_085.tif
d440bd67340b2ba267d3532cccc460a1
8d362b33b7c02f552472861074b9b5379b708114
25420 F20101108_AABLQR carr_s_Page_105.QC.jpg
b54e6238f1a5b4afe6d1103227c797a3
8726517fd5eac289977f84f5eed8630a44724300
124371 F20101108_AABNAC carr_s_Page_093.jp2
de1bba21ff453033cbb66d3994497d0d
39f831db4667495428f997291a9436eab5897ee0
6576 F20101108_AABMUI carr_s_Page_122thm.jpg
93b7e863858424f07cc857b96011af3a
0bd96ae288f582d4714569030ab32ac5987c6b22
406 F20101108_AABLRG carr_s_Page_073.txt
56ad20f9dc99caa32b84b3f9a4600f65
01aa012640972f1de6591e0287a03aedaa294c1a
114330 F20101108_AABNAD carr_s_Page_105.jp2
f330b1542d1a8e6c064b25b2d9bc7bfd
6a4b39f97a6dbde2b1fc54028b88cba8ea2348a1
2129 F20101108_AABMUJ carr_s_Page_013.txt
a4dfc55212087f787960166c7b57ab7f
0620a344fce4b7348685e1e26f36453b80020ed4
27072 F20101108_AABMTU carr_s_Page_179.QC.jpg
3659d9edbae456c50798ff143581c3f9
3cb8508f0c1364eebb44c6eba55004bcf696a67a
25149 F20101108_AABLRH carr_s_Page_026.QC.jpg
37b6ee3345245770a6f520edf1fe688a
f43d82c4a4a113b037eff5f9e7d0022bdfea52be
44087 F20101108_AABLQS carr_s_Page_153.jpg
6cdece37e1254c335b29b2136427cdcc
cd8cf6f965ef3356143e0ca6021ffac03f88412f
104106 F20101108_AABNAE carr_s_Page_114.jp2
7ca30f487cd07e705e16878eda79a9da
dd4f04c2d25849c9ab1825a6da03d86dc6551b1b
57323 F20101108_AABMUK carr_s_Page_093.pro
13ba6af4c30987fea1e2c2ce24a0fdd8
0a265655f2c04e91855a3761bb40c091fef792aa
4799 F20101108_AABMTV carr_s_Page_163thm.jpg
cfd2c3c218b2067e69c7c95ca9735ecc
be010428494b5bb3358fb07f71c68f690250e413
F20101108_AABLRI carr_s_Page_086.tif
a4cf4e2bb66afdea8628707119d5b3d6
46fe02706cde2a6b70ba3581d56a9b8c02a4f4ba
5040 F20101108_AABLQT carr_s_Page_002.jp2
135fb78ccb7acaaef2a962ff7eb22d0d
1bb75083ed7dea6cafe8974d046031b3df0e6e2e
107805 F20101108_AABNAF carr_s_Page_119.jp2
581bd92ec8c82fe55ad7e6d0ee71029a
1f97034649deec8d1830f4df2c2aaf427bd9cbff
2077 F20101108_AABMUL carr_s_Page_131.txt
059e93b2ec4081a522f3ba662f71ebbc
40ffe663ae13a7d0e0a75604f3c7ba6286e33348
27704 F20101108_AABMTW carr_s_Page_162.pro
9f9e236c6f8fc7f1818d2c822f05398e
dc48e945162891e88d6344b2b462bdf382334c7d
121384 F20101108_AABLRJ carr_s_Page_065.jp2
f3cca46e1afd9903a2622cf1d7d9b24b
b80a27f855f787c8b68a117c22dcf95aa7d68216
F20101108_AABLQU carr_s_Page_053.tif
ec01fc6c23d28990151151cb118308af
407871871c8974f9d3ce3a2a488196b9640339a7
115194 F20101108_AABNAG carr_s_Page_121.jp2
fa71c627ca43c7ed60965da55dc893e6
b3d5d78417476b80763ae281b43a7f195080b77e
3467 F20101108_AABMVA carr_s_Page_146.txt
e3333dd1360b37c47edc2adc0ead9d96
cdb5fb55f554eb13137eb66e9f001edab0b5ac9f
77865 F20101108_AABMUM carr_s_Page_106.jpg
a42e559970b4aa7291ce29c1d51b47ee
953d45a4a2d1d87c3164907b8e7b94cc9fc00672
24273 F20101108_AABMTX carr_s_Page_129.QC.jpg
828390d72810f491f560350e91cd9b70
27a3747a134b8fea58a5a9ab6a1e6ab6d937aca4
7056 F20101108_AABLRK carr_s_Page_172thm.jpg
006ce4669ce7cef33d4fb145afd6da50
fc7df817157c24ffa2616f4fc9cb95905e9873b0
6768 F20101108_AABLQV carr_s_Page_039thm.jpg
6aa9dbd061a1429b1c21e5603f47d60f
e6bd20258d3951af1407fe1acbb48436cc08ac29
115387 F20101108_AABNAH carr_s_Page_122.jp2
5b0d0cac5c155e1d62963c9458c97e51
b7cc5d8678dcd584c826e7d4386709ec1016b184
47611 F20101108_AABMVB carr_s_Page_048.pro
e9f892cb92e912c03d3baaa8790930e7
020a7c511cc3248a83b8a10b0ab492ab84db6efb
762031 F20101108_AABMUN carr_s_Page_102.jp2
1d6a3b4b12943708b2d93735f15453f4
e15885623b27a72fcc86d09188de0cd0575fab9d
78064 F20101108_AABMTY carr_s_Page_118.jpg
a8e024e6f9b4f53fc3346429504f310a
49b3191f492b73c7db071e4f5264081c66c99a44
F20101108_AABLRL carr_s_Page_033.tif
be4bd93ca96067f315355931c6f2b480
c6d1a624b0f05118f6547ee9df715937139ac552
3717 F20101108_AABLQW carr_s_Page_138thm.jpg
aec006583e0194ebeea32f20cc043c6b
d2609a1a71423e6dfea763797ef63c9ae69d6a82
110726 F20101108_AABNAI carr_s_Page_124.jp2
e30e3a9fa55dd5f13a8418a2908ae291
45af6fa582f042b1ecb27983f4f547f92f7fbc2e
2700 F20101108_AABMVC carr_s_Page_171.txt
9452234be921e98f9e2082a989512710
2ebda9bb590e3707e089be49a8677f117af7bd09
F20101108_AABMUO carr_s_Page_077.tif
73ffeb6235f22030d40d63018d6d5132
1fc85a0b90028b137087cd52b79e0d1dd0fd049a
2110 F20101108_AABMTZ carr_s_Page_053.txt
8cd83fa91e71e2bd3bb1b9ea2bb0c112
8de7a5becc46a6e780db1d8e96dff0389f2e44c0
F20101108_AABLSA carr_s_Page_019.txt
4fe9dc69d5bbfa7ac2eafb74c8268181
e30b3115ddbb312ae949d0ef829bcc629a2ec12f
569 F20101108_AABLRM carr_s_Page_137.txt
e9a190d0932ed2113541d54ee541735e
7dd7fbb7e34ba1fb845ca89444c71552f41a336e
54296 F20101108_AABLQX carr_s_Page_055.pro
1a36fabffe2d8d677aa0b37a904db7e3
2092de1a1f5ef9011f9499f387b313a9b78fab8e
289410 F20101108_AABNAJ carr_s_Page_136.jp2
0e1c3eae8c68c0a14e22c8709d873e18
51775368f6c9ba1df27d4c1e01e8cf5db85368eb
3191 F20101108_AABMVD carr_s_Page_147.txt
4f48d6425239f1afbf91ae17ceeadb5f
e06e5a9741d7b417182291a50b175fc1dc422230
4269 F20101108_AABMUP carr_s_Page_160thm.jpg
d93a4604d44990eb3ecc4324f4ce29cc
50d94f185a7c4e71aaa9132c94950c98ab22e072
F20101108_AABLSB carr_s_Page_157.tif
bd463e36466b1dcb8fc22ec5ca590436
6aa824c4b8b11bf2bf8cf2fb760bb7fafb56f4ed
50755 F20101108_AABLRN carr_s_Page_166.jpg
0772223186a1a9e24a6aa85526f0f663
87a89f5e1a11d68484f31b9d683ed98b3a32b1aa
109241 F20101108_AABLQY carr_s_Page_120.jp2
6f8191336d42e191844b9c29e8766ad7
b18e44b8b1d8c2f23695328a4544c030d03b9707
367810 F20101108_AABNAK carr_s_Page_137.jp2
bcd2eda97b7eb5062a76690be5debb38
709757d0ace023bc495ec28a8acb175d473f6b9f
74680 F20101108_AABMVE carr_s_Page_016.jpg
9b88a519e8944ac5f0a18a115bf85b2a
299e24a8b0f0bfca77db1f02e3416b42431d3df6
49662 F20101108_AABMUQ carr_s_Page_079.pro
c20dc5af7ea18c5f3372e06e624036df
77679bd938c2e8fee12cfbb554e5186af385e9ee
111838 F20101108_AABLSC carr_s_Page_034.jp2
68806d8b9046e313e803fcda9a7ffa11
4214d66601f804d3b4634773f0202f34aac06353
7270 F20101108_AABLRO carr_s_Page_180thm.jpg
2eeb533c0e13a6ab93306aa7ecc5283c
ed963c525210a1db0e19b486630dd048e4e7a06c
112683 F20101108_AABLQZ carr_s_Page_113.jp2
ae4c2ef7079ba24a5ecf30b4a8340b33
01e5b30940e7dc2e49c8a27f65e5f6733d743675
F20101108_AABNBA carr_s_Page_010.tif
f83b591525e01ef641ae1293a524783b
0ae88163bc0b477ed34f51c15b97ff9e57feaab1
151791 F20101108_AABNAL carr_s_Page_151.jp2
93cebbcc7028eedd98124b617ebd41be
dc519f0c32543d7f514c946485bb057a91b19f6a
54805 F20101108_AABMVF carr_s_Page_026.pro
44332cfb996ab67b2987f69e488ad05d
374fbb167eff50d2157f47073551aadc1626f31d
16040 F20101108_AABMUR carr_s_Page_070.QC.jpg
3d455bcdd06119e194c9993f2bc65aa9
9becd8dee4d7a45c688c1c02cc6e3db133d4740a
F20101108_AABLSD carr_s_Page_181.QC.jpg
0ebec42440b476069f54e26f4b400fb7
0177834b0746d348f08bb15faee6751d5589a71f
53643 F20101108_AABLRP carr_s_Page_020.pro
3bcf69c4325515b210204876af1ef892
1c41cb1939de34979655f0f7e5dee13e0d5f4cc8
87665 F20101108_AABNAM carr_s_Page_156.jp2
c7d29465c75193171b4d34de8396e46c
c6759ce62ff36bf4cf648a9242e6354d9354051b
19906 F20101108_AABMVG carr_s_Page_145.QC.jpg
794cfaf3ba25cdd971602b4f8e49ea94
4e2e61d142e3ee84319586cbcec2f40766657e61
28327 F20101108_AABMUS carr_s_Page_151.QC.jpg
7fcff8036d680418385e76d5941ff7a1
0bb9c6423b64cb35011711fbd4e35383780b11ab
117506 F20101108_AABLSE carr_s_Page_051.jp2
aec4a351f1be2cfb75938661ae756c2b
605c8c125b5e6cea2d43fa6fd187d4791760bd5f
118220 F20101108_AABLRQ carr_s_Page_088.jp2
95eb921bafda2cc590cd4638f9ddea08
7b5792d8440efefd5948eb36f141dcad92b44436
F20101108_AABNBB carr_s_Page_011.tif
931838973d5f584c009c7aeaf6d23e87
408e796b1a427c1b514455f941b83a7acf11e216
85333 F20101108_AABNAN carr_s_Page_157.jp2
4d39dac5be013bdc7ef738fdbb513ec4
75415f78353d33e12e8ade1b5ce78774d654c93d
F20101108_AABMVH carr_s_Page_031.tif
6893088470e8e6ee686483616116b8f0
fbfcd2255cd55cd11a5d604843988e6439809f37
76279 F20101108_AABMUT carr_s_Page_167.jp2
0545a622da24849c0bbb9094c00f0e58
e685ef01f9ede1f27e17a2e84381502bd58f019b
54301 F20101108_AABLSF carr_s_Page_130.pro
d9dfe9c60742627d2c22129df50033bc
9d30e9e8946f8a4c498d43ec407687add833ecc3
76641 F20101108_AABLRR carr_s_Page_082.jpg
bfef3d82a72eb557ba41457b0eb7bd0d
72952cf12cafd4eb724272c204e220512762fbe3
F20101108_AABNBC carr_s_Page_013.tif
ee7fb64a7641b851a9e7c4b50620731f
ef3ffc7f71a03cb72c3b7b14b5cc3622540216ad
83074 F20101108_AABNAO carr_s_Page_158.jp2
e31cebb61b6ec5bbcc16c5a3ac289f6e
6ccde6eb3cfe7952de26a5a891f0021ed8a6d8a1
24661 F20101108_AABMVI carr_s_Page_128.QC.jpg
8b21cbb0edbc015fdd4fea6eaeb24be7
aaa3768e16c185ce5210405d29603ac939a9a45e
49454 F20101108_AABMUU carr_s_Page_004.pro
2b3ac8aa91496ce8f2bce73276b9625f
4060f200ad13ca959c06e62b2a5c72fab3dc3627
23153 F20101108_AABLSG carr_s_Page_046.QC.jpg
99015d84a45b3f8854c32a9c575b92ac
ea0593ce247c6647a3b50d8ebe981662e0ca896d
71573 F20101108_AABLRS carr_s_Page_165.jp2
a0f2c34d316f0041b883b54b2fddd71f
d659729a8131caa025bfcff33725acfc5be8623e
F20101108_AABNBD carr_s_Page_014.tif
966e4e596b2918950d054643c4ae33e1
fb888f7f748dc0d60f41e0d7dc0af28710247dc6
83626 F20101108_AABNAP carr_s_Page_160.jp2
515f1aadf6f04b7e49285f34a267abaa
8370d8b6fb23ff70db6e817f21255f5b8b25d6c4
51050 F20101108_AABMVJ carr_s_Page_063.jpg
79697cfbfdc2e13e2eac0d61b593aa25
99e570984ed3094148922c4ec4e835801ba8f3f2
24079 F20101108_AABLSH carr_s_Page_017.QC.jpg
25e24105b3c88aae426d662bd53d4921
240c07b0919fa7938f23b2867f328a1c4a9497c1
F20101108_AABNBE carr_s_Page_015.tif
766c8b7cb76d3dde0ac01f3a05874624
97a25c4fd9d70acc11fffe3d0fb2d6d285dc1a41
120498 F20101108_AABNAQ carr_s_Page_170.jp2
757e05b62dcffbd3d4074c1cbc84807b
70597a99fd8083ff847bd0a041feeb93e9eff074
72822 F20101108_AABMVK carr_s_Page_125.jpg
e580abf7f1e631820d93a76d53865d89
ea8aa1e35775a8e90e83f11f889b876502720c88
108656 F20101108_AABMUV carr_s_Page_004.jp2
02a5f66c6f87dd54dea8c7a5aff51956
d8b446772a8ec2f66a8adeecd2a761cf30f58602
52607 F20101108_AABLSI carr_s_Page_009.pro
86588ec98b4ee9bf78bc3fa950286963
e54ae448f23f7ab78c5cd800e16a755f90a7a993
9945 F20101108_AABLRT carr_s_Page_001.pro
9d991dd5fc01160e04ac2755d0490569
5c42b5c212c20337a3af8c4246f5f2eaeb40b88b
F20101108_AABNBF carr_s_Page_019.tif
10afe6afd4e9d4fcfceed1d698e3cff0
dc26d7c5b39c54cf7b2909e2db841a1f0a66b1dd
134828 F20101108_AABNAR carr_s_Page_171.jp2
5dffb964b20fcbcf6e402e8ae7833f0b
026fab30c5e709429c6f39c5825e72a9e3abc505
25993 F20101108_AABMVL carr_s_Page_085.QC.jpg
db1549eab6f2e2e7023add103f7052a5
1751bc8077956ee21df98881369179e50e7373b5
6689 F20101108_AABMUW carr_s_Page_084thm.jpg
99f2b808cb9018493e5fb02a1f3ee987
97d056f4f7561b0919d19f2a625f7bfd03e929d3
6879 F20101108_AABLSJ carr_s_Page_042thm.jpg
9aa6d175e69e055a63e01727264905ba
421c9171e5f51018b37d3e5cef96a94f998d8aac
F20101108_AABLRU carr_s_Page_134.tif
526b93213fae74b6808e41e06a3b35d8
e26fd6722b540ca957696272932617c6219bfa4e
F20101108_AABNBG carr_s_Page_023.tif
1afa99f4452df62c59399022c4215b0e
58009e0332bcb8e8c1d1e90c5275eb2357c10ff6
142667 F20101108_AABNAS carr_s_Page_174.jp2
1524a909e30cea9274121723eec352ac
7c72b3f9420feae9aa971b3517312bb619ce5494
F20101108_AABMWA carr_s_Page_066.tif
10d2e15a11b5097f19b0e49f466221ae
5b0fed4ac2516354ef14b262c801a52d6024849c
15287 F20101108_AABMVM carr_s_Page_142.pro
67c86381695c56e957ac1809f16c8943
ec1d2c6aa4f82c0673e353234c2e7973627f72ba
77808 F20101108_AABMUX carr_s_Page_107.jpg
a8759e96aef8d0d26e982d80a666ee57
7323e6fc7685e8619dab71b6c5bb4714db64271c
79353 F20101108_AABLSK carr_s_Page_089.jpg
7311a642b691f45f86e6075ddfdf58a5
715bca196229570a22200c5f090490beaa041032
39001 F20101108_AABLRV carr_s_Page_007.pro
281a6368b55277f8fffc1af3e622279b
1fe74f23fd2c49af1f18051c6731f58574385e8e
F20101108_AABNBH carr_s_Page_026.tif
a3976ca02b1a916d5ab9865046377e11
9c15957279eeca64283ee0ac10c9c0cb2a785df7
123663 F20101108_AABNAT carr_s_Page_176.jp2
2466c8983bbbc599012a458ade6c9790
27b90525887ce3634ee910ebf71121a02a638958
62057 F20101108_AABMWB carr_s_Page_009.jpg
5e1f7f3c2a2d1adfac9ac3836101d428
7e36d02da2b5b943e38d33c7a08036a912cac47e
56839 F20101108_AABMVN carr_s_Page_091.pro
14f017feb096bcf03879710de41b290c
466f303bd822ba032ae1fa65b16bf98b0260ed6d
563 F20101108_AABMUY carr_s_Page_001.txt
550120aa33fe1b8228504500a3527735
8b9af88e397c92b73eaa919463a325bb15209dde
6651 F20101108_AABLSL carr_s_Page_028thm.jpg
0fcb32398f98e87a381018e3e90e2fe8
d8f868e422a0283863dbe0f156ae0931b1e6bccc
6801 F20101108_AABLRW carr_s_Page_022thm.jpg
ab84dd0e1c86dff7785fb0b2575622d1
a6dde06a51f3193057cd80f66ee9d8a696378470
F20101108_AABNBI carr_s_Page_027.tif
7cc5e492805ba6c3d1a42b4cb79fbb1a
1af0dd72bdacf8c65f0db05317f07e929a7c9a39
128394 F20101108_AABNAU carr_s_Page_177.jp2
30252c9b04bbdb1ae2508e2948341be5
5930599b63dae5efecbc58d92d6288430302d5d2
68234 F20101108_AABMWC carr_s_Page_182.pro
8db5d27c80752e4645f3fb46124c2a29
dbea62247297cb8fc770f5f0dbed3becaf5e6e14
36088 F20101108_AABMVO carr_s_Page_100.jpg
1508ef3878db196534a19e624f716870
bc522d78e2b68e163462de83127f5b8edbfe5464
3135 F20101108_AABMUZ carr_s_Page_169thm.jpg
809a939e3c8cd3c9add8951c54a04e82
3afb44762a298cd38fc04d56f08736b3d72ead1c
4422 F20101108_AABLSM carr_s_Page_063thm.jpg
5ec3fd3dcd9b7ecdb57fe6cfeab60f75
f15b3aa911764f8d5c46dbd9b2a38369737cb3c1
121993 F20101108_AABLRX carr_s_Page_127.jp2
6f6371298a58967e032d60222f1c4ecd
e8ee0886151863c384312972d1a5a0c77164e9c7
F20101108_AABLTA carr_s_Page_004.tif
738de9224c5913a3ebcce0842ec0f1e7
f509d59291aa91191575d1bd4d98eefb165c1144
F20101108_AABNBJ carr_s_Page_032.tif
6ca46dba641c8604b1717bf5701d9f2a
f4f22c924e6db46159c790771b6c4848f9ba0c31
128330 F20101108_AABNAV carr_s_Page_181.jp2
556f4bbeab756928299071c9176cb0b2
4406f6826e400c519a90cdeffc6cb4b233a726dc
F20101108_AABMWD carr_s_Page_182.tif
66e8bb165f6d075b6139e419da5d80c4
a77d84c5388f7c54bb2df9aff752bac032131b9a
47256 F20101108_AABMVP carr_s_Page_139.jpg
bccfd51d0b31c40103e2bf450e6a1688
c5e6e27a27bca12eaef6e7784affc9b19b2f20d1
51801 F20101108_AABLSN carr_s_Page_086.pro
1db9922895385f5e026712921e5c4593
134086aac133e02ea70583f67364b08fef2d3063
F20101108_AABLRY carr_s_Page_001.tif
5d68ccb3855f9a49689ef0442696c1dd
40cd947c758178fda40517945838501a9123ee24
2194 F20101108_AABLTB carr_s_Page_054.txt
c19a3d5948c00dc273bf3ebf3bcc38d6
81d06685181feb8afe9f2d268ca047c0dbaba5bb
F20101108_AABNBK carr_s_Page_038.tif
102d81cb471725bded4e1d1e5640071a
a04ee16dcacd3b1489fc1e5a970344171ca616f7
F20101108_AABNAW carr_s_Page_002.tif
9fb45d222cff176f0f044efffac47957
9d384b1c269b750585ad111c1e6b66eee1ca4920
51833 F20101108_AABMWE carr_s_Page_084.pro
56fa5facf558ea8bd3dd88337637180a
cdb5f4aee8eaa406456255d574fcbe2e876902ee
85099 F20101108_AABMVQ carr_s_Page_178.jpg
69add001beb41da710a50de0fc3a5918
4312496e554ca8fe8f35c08379757e8b1e82a476
77644 F20101108_AABLSO carr_s_Page_105.jpg
6ae6e3ef842c7715c105a4404ab5c3d2
95b4fc6993744963e03c0a45a280f59c5f38e839
2068 F20101108_AABLRZ carr_s_Page_113.txt
789ff825053165608377d5d20ab80761
f338fb483c7aae47c472090334b3ccdbf7cba334
11296 F20101108_AABLTC carr_s_Page_137.QC.jpg
44859c2647dda8cdff78d2f6dbebba86
ad54ba65819b50ab0c0401380e6f8fd588885aa1
F20101108_AABNCA carr_s_Page_102.tif
ae6e1cd049b4ab738d3c531b714f8f89
b20c01f5e279cc8ba724ccb0378be2e98dc421d4
F20101108_AABNBL carr_s_Page_040.tif
35e58ff25b30f58068bf88753708ff58
a12ece12cdc6dff20c6d20eef652ab3c89ca915b
F20101108_AABNAX carr_s_Page_005.tif
a1c1aa4b1fa5e26c820060be9281cf50
22310a2ca30f8779ada1cad2118760726777e8b9
F20101108_AABMWF carr_s_Page_132.tif
ac5738fe813c213169f0a8b7043155d0
083fa9766b3d6a7aa3d99d8047fd995f8e654943
2347 F20101108_AABMVR carr_s_Page_170.txt
ae2f412e829215b9f89e43d6fa02879e
73600490582a57b70166eb7d6da272134d99b46e
24501 F20101108_AABLSP carr_s_Page_086.QC.jpg
bb2b09f6cbf37fe8c38537da8b5edf6c
cd91303eab28ea42cbb240015950b3131f3128c4
111476 F20101108_AABLTD carr_s_Page_043.jp2
c51295da2c7e95a064c148888b8bf851
19ab30202c9a299aba365eceb33fb1b2fcea193b
F20101108_AABNCB carr_s_Page_103.tif
c475305add44afb94539952839c7a207
3b10110a5040917dd05d652a61e222bcc27e5614
F20101108_AABNBM carr_s_Page_041.tif
d0acca820129dfb212dded08099f71d8
08b7b0c2f6e7604d83d79a8a0483b8eef3fcfc2e
F20101108_AABNAY carr_s_Page_007.tif
c8b4d1e058435e2a578222691dff8e2b
cf6d3985ee5f9a360ef34ff2dc498cb9c8aa7b79
116520 F20101108_AABMWG carr_s_Page_014.jp2
194e86cfb30f325d2d0e7f11747a6319
c5daf821f58f8cabfbd48f8a0652f4c4c74e49d3
44483 F20101108_AABMVS carr_s_Page_070.pro
286458ebf3a708b08339975df07ccefa
232559ea33229d0fab9daaaf0080dca70c000a17
50824 F20101108_AABLSQ carr_s_Page_044.pro
6c99b88aec2f045e42027c1dff23345c
63c87ed026e3b64556332064da4dda716ab646ef
51372 F20101108_AABLTE carr_s_Page_029.pro
6e3308e46028182e898a52a6507f073a
fbdb0e3d703c8d59bc9917f55f4a33effffcf05d
F20101108_AABNBN carr_s_Page_043.tif
44f222dc004f3d39e62b288faace2108
0566a2c6f722015a07a7a5f0dd4253b39d5b9b48
F20101108_AABNAZ carr_s_Page_008.tif
dbc1fe8a8770c3177572884ff95481dd
1ddea51c5cc1ad303a1210ad21e009276ab895de
4846 F20101108_AABMWH carr_s_Page_062thm.jpg
9a31e7cc7aa91646d25df24986a4dee6
b4fb3cd64ccc530c8714348b54bcb59b12c63d72
16696 F20101108_AABMVT carr_s_Page_064.QC.jpg
94676627be0286173b5722e5709c78c4
37971f487694323c9b56bbe93a5ff8849a41f2ce
1454 F20101108_AABLSR carr_s_Page_165.txt
b33dfe69f2fc1748821d0ca386c7b356
324b6050e17495b9c30f8900506db773b45e16be
F20101108_AABLTF carr_s_Page_067.tif
75ad9674602cf773883fd2142c3b2a26
c01da70b0a234e2625efcfe6a7987bffa7c64912
F20101108_AABNCC carr_s_Page_108.tif
9e690e9731b851223f058f7f9c1a712a
b752fffb34590fc1610ae0967425da2b55d75817
F20101108_AABNBO carr_s_Page_046.tif
1e86ba0c1b9f9d7bb5f7acd0f54aded2
81640992d7cda18bf88d324638f13b116bc0994d
2159 F20101108_AABMWI carr_s_Page_088.txt
6a81025703e81c447d7a5757c74f7f09
fa707b1e606e09dbe77cf0783d460c96817d9766
F20101108_AABMVU carr_s_Page_052.txt
061bca46e311001e8d70507800f592b4
7e91eac0efd817322095a0fdbfd66b53a6f9375c
40094 F20101108_AABLSS carr_s_Page_158.pro
761687b19a01442784d2820cead05b52
173997be95aac23fadb24f26dc80fbf735529027
6783 F20101108_AABLTG carr_s_Page_082thm.jpg
442c5afbc6c3b05f18e724279d24e899
e09ce035b2e4106066c054b69ced1d1f634b2295
F20101108_AABNCD carr_s_Page_111.tif
a64a89b3ebd53c7a38a7b2aba44ad3a7
e4d2c88709288866122d5de483994d0665c9be55
F20101108_AABNBP carr_s_Page_050.tif
fc08432999a485ffb44bb6b99f53f39a
443c226386146b6b2008e7a82919ebf214eec717
F20101108_AABMWJ carr_s_Page_044thm.jpg
a78ec4ed8d5e326ec7a40f763dd2f98d
e0708feaec30e58fc4a31701bd07afb8bb24ab09
4664 F20101108_AABMVV carr_s_Page_102thm.jpg
dab5e54eb9a22fb30e530c836a8db418
75bb4151960b3eb3d49ae97b6f70ae8d2ef37c36
115474 F20101108_AABLST carr_s_Page_117.jp2
cbab9ac4b8642faaafe0025d6dacf3a4
bb3c28cd8e500484ec130f44188d1fc6b7bb9035
43153 F20101108_AABLTH carr_s_Page_169.jp2
6ef915e9da289c0abd6b560732116d11
ffe1a3a5c7d939ee5d11c2a75fcdb200a00aad41
F20101108_AABNCE carr_s_Page_115.tif
c29f2a2398e1649e2d037e2bda8b0d7a
249fb49b66c0384bfa787687ddc497f988345294
F20101108_AABNBQ carr_s_Page_051.tif
f67553f2d6ca12938ac69e3182b865b5
e4181343d951a582a50f2ca1cff75f6c03132889
25329 F20101108_AABMWK carr_s_Page_055.QC.jpg
564399e01d42a216b1fb721bd01f4590
4c842fb147bbf3b81cb19ee5df960409fd21b847
4446 F20101108_AABLTI carr_s_Page_164thm.jpg
8e93c7af1d86485e8743e14503a11c71
b34a423c7794999f48c890338996ac7f5dae73c7
F20101108_AABNCF carr_s_Page_128.tif
624c5edfb7a1da68cb50e07a6ed06a5a
8031748f58daf1a212978cfe958f52437f13563f
F20101108_AABNBR carr_s_Page_057.tif
c0ee845c16855a3a162e7227b30f9f00
ac913e8db78d6d82fa307b52201c9c16f6e3df1b
6840 F20101108_AABMWL carr_s_Page_056thm.jpg
cee0847fe3600c044e18bfb830dd0d62
2b87e10ff4c0b2494a7cdb2436d771d18043d6e1
429501 F20101108_AABMVW carr_s_Page_060.jp2
1f49db543691d35afb7d541899cf5083
70f3773f55fea0421ea3307bcd77a107cc9c3d34
6626 F20101108_AABLTJ carr_s_Page_117thm.jpg
7510de9d4002f1c8b6bc6f643bd60148
f9209d6f81621c9ab845ab3ab4d93ecab140fcdd
52836 F20101108_AABLSU carr_s_Page_027.pro
d983ea20b24680a7131790c964528987
bef87203357d105e995731d40db9a57e31fd10c6
F20101108_AABNCG carr_s_Page_142.tif
83861328d633f89558de6622b969acb4
bcb7467818239cd169e6606e572935904fde4341
F20101108_AABNBS carr_s_Page_060.tif
cc5d84e3d648e190d15a3189fa6163d2
7d63e6961360776388c8a99589886248f056a9c5
75278 F20101108_AABMXA carr_s_Page_032.jpg
f7caee69c1d0edff375063db7a816312
c8790fbc09a6f530a5b47e3e67385886f90186d7
80839 F20101108_AABMWM carr_s_Page_166.jp2
0b2567d3737e643dde4a7f40cda2637c
cca16c7a1b3fc06b2e7500f65e9762167eee23e8
75343 F20101108_AABMVX carr_s_Page_183.jpg
8f0e7b0d7f68b20ad377892f45caeafa
517575c0ad399a2231dcf89dd5eece53bd44d982
14534 F20101108_AABLTK carr_s_Page_160.QC.jpg
860dd498e8531075bc87c4ca363660a4
982522118868e23d4ed8d0c0eb025a4cb0b085e0
418 F20101108_AABLSV carr_s_Page_012.txt
eb1e3f835fb130bb4cc26cc8fd70badb
78c174d3f9cf8e5596f57ef6ef96a0ec94bc3726
F20101108_AABNCH carr_s_Page_150.tif
c673c2b4d900965cfbae8de7ba3e8dc7
d6039f6d9c728c6ee701a5a772dfbbb8e7e39d18
F20101108_AABNBT carr_s_Page_062.tif
e2d53db16d5a1e9b4526bd079172ebb7
b23033db5d27a9852642d21e548c5f1bd1fae1f2
39422 F20101108_AABMXB carr_s_Page_163.pro
5d3930eb85a28b503ed7b031810ba25b
b0ed8ae9731a20c1b3547d7e96e94c992d6722cd
16116 F20101108_AABMWN carr_s_Page_154.QC.jpg
3104ef0a0e87aca5ff2ace71e8b7d30f
2437d41284566ed69e46416b64b67bc14b4f1f0f
121728 F20101108_AABMVY carr_s_Page_091.jp2
5e972ffd9487ce912325ad60cf9dab2d
355ace4dd44f6ce8aed514c25fdfcb716d9620e2
F20101108_AABLTL carr_s_Page_146.tif
bb51a735e77fce245f9e772a1075af45
2ff9c15553690038fe07c576e80a16d3d3f6e932
F20101108_AABLSW carr_s_Page_125.tif
abac4e342aa63ad6d3867888c8adfa57
a8b55c6249d7ec5cada93eef6939e4e011752e4e
F20101108_AABNCI carr_s_Page_156.tif
ee8a01f0214a9a92cea6130f27ab7e12
4160789426bedc4c0245bce175ab92f98cfe3c94
F20101108_AABNBU carr_s_Page_068.tif
6ead614dafea6d0dc6fa5bd54f5a5281
c3bf67ff5fba85082fa319f42e42c9e9f7f3eff5
53666 F20101108_AABMXC carr_s_Page_075.pro
91506460b8f9b3720c9750d4ce5bb679
fee760511e0d8d5dd822c1b9cf2708f3061346f9
F20101108_AABMWO carr_s_Page_011thm.jpg
d99884e414c42036d5edd46ff8d481cb
d6a625a4f0f9e8609ad7edad3b6997b9b5b5a9ab
1378 F20101108_AABMVZ carr_s_Page_064.txt
5a1a788c6f6d529570f3aa3ce750663e
b5a3c65ea5b7443b7a0f5406a49f7b76e2043411
F20101108_AABLUA carr_s_Page_016.txt
0dd726661768378d779e67af105a865d
8c9d2593bd8cbc78739e0db529bf58f876404f4d
27864 F20101108_AABLTM carr_s_Page_136.jpg
8633a9308df8f81a19d1b75865105022
527a3c3afba197f6b6327546ea78f0020c3768f3
22880 F20101108_AABLSX carr_s_Page_041.QC.jpg
69f4a158bde715d680416588f6a93e03
85127e4be8484e31e9f709c60ffb1129a934d0a2
F20101108_AABNCJ carr_s_Page_159.tif
3cf4bd6354be556a65774c929be08511
a0916ef163e87eba913307b9d14669c4765deab4
F20101108_AABNBV carr_s_Page_071.tif
ef6be56cb273215fd2c9d3e6759da8b9
5b3c14df63e614e234de5e69a865882ad22e5b69
30537 F20101108_AABMXD carr_s_Page_001.jp2
75df74506b186dbe9648d5e5352a042f
59caba11dc71551b62fc93d51b9088fe7575862e
113585 F20101108_AABMWP carr_s_Page_016.jp2
fa2addd4b023ad3cf084a36b08b0610d
9ee9d32b6bcccb784ff115869062be1d377e2271
6618 F20101108_AABLUB carr_s_Page_087thm.jpg
e4a52d96487c0c14d92be58bbaeb7b41
0373869734fa668924ee0be3f4b1cee535fab179
F20101108_AABLTN carr_s_Page_181.tif
47fc63efcab55a2ccbed86e18b281f35
3cc905737beb2d211fe67b1f73ee4c93421f0305
2019 F20101108_AABLSY carr_s_Page_050.txt
d200f896ca8c781774c2cba5b3bd2c4c
2183258a8935b9d2143d8763bc78f3414aed8f10
F20101108_AABNCK carr_s_Page_160.tif
b9813666766de79a733628a2b10b94f2
b504d5f7cbd12d32da3add1f15edf2a4f648db21
F20101108_AABNBW carr_s_Page_082.tif
01a34168ddabd1320b1d0b66067c0024
234fc5122473a06efde47212f3cbba0346dd586f
61793 F20101108_AABMXE carr_s_Page_153.jp2
057dc194b9f0704a0e2af9c2a6ddfc2c
505aeb2d9df3a30316974bda8d76a3c50d72aa47
19391 F20101108_AABMWQ carr_s_Page_069.QC.jpg
bbbae04788041261b76049492c9c613a
9ffe5208cd29b402b0510cb1ba9a0d2ee2d5f19a
48620 F20101108_AABLUC carr_s_Page_167.jpg
b6d4f30c489696a05eda494e3b659a62
2fd1284b8e6f7ee36f8fdbebe8fe512ac1c3dcdf
F20101108_AABLTO carr_s_Page_119.tif
dc1dc3d0de838b0ecb6af734b5e30225
acc4c2cb576859eaa018e297301037d14cbaa138
1624 F20101108_AABLSZ carr_s_Page_161.txt
5096b578adab77a2528a42a9ff3bfcf1
c0ecf7c31cdc9af1fba40111587c0c8dd3260263
53556 F20101108_AABNDA carr_s_Page_069.pro
289f58b7199303925f32477346b0fee0
7cffc738e6d8a797d602e53a57361b9a80f96ffd
F20101108_AABNCL carr_s_Page_161.tif
99ec1c3d0bf1d8204673087ad3c11490
ca2b84cc989d4cac3ed2c666dfde618b678118bc
F20101108_AABNBX carr_s_Page_083.tif
e0aeb94ea3599874f8b6c939e6dfe9d6
fd360d795826c456572f8a3177cc5e7c7d66ceec
2173 F20101108_AABMXF carr_s_Page_085.txt
65a4b59c32305af49ede5f142f30902e
9fe69f3206a0e4f2e2ebf9e7f143cc824d9a8666
54981 F20101108_AABMWR carr_s_Page_095.pro
8c9660e0536e35391cd5c963fe275ee8
dcad98e760819daf799ed6d3b576038928a69945
25750 F20101108_AABLUD carr_s_Page_089.QC.jpg
fe6e632202ed8db82c54041acee67ed3
bcf29ba1b8b4b379e776bb32bb81d1ec2228d536
F20101108_AABLTP carr_s_Page_157.txt
e0a4b8564f395264cd16dca2d4898763
24deb222ee0c92f0186f2de9e95e6a6d7d7eab44
48220 F20101108_AABNDB carr_s_Page_081.pro
2e548952a7106035a6600de6043ab1a4
526ec53067c5947f1bc82de5e231f0c2b891b70c
F20101108_AABNCM carr_s_Page_170.tif
0b3facadec65345c5c73d302fb705ace
f0656d9d239bf12f61e9f8659d088099ef05a9fd
F20101108_AABNBY carr_s_Page_095.tif
b7f24926995e835817dd50262a589b06
8e4b6d4dfebdacc70c9136b6866e19f408df8e08
25617 F20101108_AABMXG carr_s_Page_075.QC.jpg
9df7174cee7829dc52a28d27d43cffcb
b017f48366b03e78a085334c7a398dff5c52e2f7
F20101108_AABMWS carr_s_Page_025.tif
3abe7aee95c0badfe54dafedac8dbb43
8ca4022b0b3ac30fe3b6c419509f3e72274c2149
5268 F20101108_AABLUE carr_s_Page_104thm.jpg
e94441396d4aa3f9f457631ccf9e98d2
767d2b04d9791a3943234a2eebae1767d3b4c82a
6324 F20101108_AABLTQ carr_s_Page_023thm.jpg
94726b78dc131b1a50db6a59cf624c37
70b813cd22b9b86a0eba87da418259274fe25c86
19967 F20101108_AABNDC carr_s_Page_101.pro
8aa3ad4c128005d2801dda217bd3cce0
bb2ed0b869633a9a26569d5a7803e894010bc839
10403 F20101108_AABNCN carr_s_Page_012.pro
00188582b763308ad117b59384671270
2819f851f7540773446cc4bbbee1baf290189cd6
F20101108_AABNBZ carr_s_Page_097.tif
2ef89241024ee6718c7f9a33cd6b5fa7
47c21defd3076fa3cb3fbbfd68452bef603f6b59
58975 F20101108_AABMXH carr_s_Page_103.jpg
1fe0e1f2512283d2069e10902a2e1653
abdfc3357a6255a2aaa39d26022fb1ad7a433d4c
119528 F20101108_AABMWT carr_s_Page_076.jp2
4293e99db2e7d69d4333dc8a3350f60e
652de3e7c4d1cbad9ab2cc72a2a97f0999fcac7d
1892 F20101108_AABLUF carr_s_Page_048.txt
f8a44778f9f8983a2af77ab053731fb4
07b4b9478efa156fa257dd2ad739e06a0dd1bd2d
72947 F20101108_AABLTR carr_s_Page_109.jpg
43c68f22d054fbf8fc70fa319f9d7b41
56e90448ee86757a5616155fa697b1224a5ca7b7
14585 F20101108_AABMAA carr_s_Page_102.pro
ac6e615c8e5fe59bba890fbd8fab41ca
b4546d9de30f0d55ae559da88d080defee45fb48
9609 F20101108_AABNCO carr_s_Page_018.pro
7ebe448566f5951ddf632788ea40cfbb
8724a598545f1dba9c51b8cf19fd56f4d2bfdcb5
F20101108_AABMXI carr_s_Page_030.tif
496d56596becaf807677c10315098715
120829af67086d950c1d1832cfee6397820a2818
9879 F20101108_AABMWU carr_s_Page_100.pro
dd6e3816c13aad35a59ce41fa4a0857e
64d187e35861668c8a2aaf0c7467abf760ed12f3
F20101108_AABLUG carr_s_Page_016.tif
26353a8399fdb6379608d96dcbf5d4b6
5f65c0342441d36169a491e8134a3f8f5b653fce
2151 F20101108_AABLTS carr_s_Page_076.txt
dc159ea7aca79abaee03143ac451287d
ff5cda1fcf24785bbb076e45093f6ce73b2aac38
55646 F20101108_AABNDD carr_s_Page_107.pro
b004b961241d05192295222f044820c3
c1fc33b926afc52d9d61f1e6571f1c6c402e647e
52546 F20101108_AABNCP carr_s_Page_030.pro
36fd14e01c6d4695f3c82ef305a678d1
d77db50097f4e388d8485dc159e7f77d207e960f
61233 F20101108_AABMXJ carr_s_Page_059.jpg
5a613efe4597ab4019ecad1cb1628162
c2f051b31147ddb88b633c981b0bebaf9f661dbf
5523 F20101108_AABMWV carr_s_Page_145thm.jpg
2ea211c153656bbe3411d9e4899bd489
8c2dcf905dad5b7bebf1685f49ad2e0d3e1ec9a9
24082 F20101108_AABLUH carr_s_Page_099.pro
b2de901b6e13ec106480331415dbee59
a97972c041a608479f91fe300591d4ed099bb81f
F20101108_AABLTT carr_s_Page_006.tif
04bbec3fcc69849d5e01d23b0a9c7d6e
bd2a3695888dcbb45d8bb72c191782e1dbfe4039
2163 F20101108_AABMAB carr_s_Page_095.txt
c54579644d1d8888525c178fbdcc1827
242b26858e9eb4e601c8c096d121eb7e3696e162
50297 F20101108_AABNDE carr_s_Page_109.pro
c47ae5a9fb4ec49a3606debec1799a48
4f6801bc9cb05be44da21de4e4daa64f86f51e7e
52403 F20101108_AABNCQ carr_s_Page_032.pro
5c495bd99428dcc08dffcbbdfc2a7cbb
9bdc23bf7594adb5266997a089ea6e9a58c2d7f1
20746 F20101108_AABMXK carr_s_Page_072.pro
207957b432a384b18fec4c5397707740
7b837ca7401eac435b8f84a7c5eb604a1179fad7
1934 F20101108_AABMWW carr_s_Page_023.txt
daebba282fd1173c98e4dd559ec03eaf
6dae1036200ebce976d5cb45eb298889b0eb3156
12227 F20101108_AABLUI carr_s_Page_135.QC.jpg
4ae214aa162892b7462ac50eecc14e90
f0da3eb7c2a3f15ec4e48554aac8dbea54981c53
4676 F20101108_AABLTU carr_s_Page_135.pro
7eee29b42ef8af784bb00b9264a69a87
dd724d02bf7e1f6e7d5eeab368184ea8b902c343
2190 F20101108_AABMAC carr_s_Page_143.txt
db295f15745ea1422610ab18cf0d61e8
8a14387c75e899cafc0d3d3dfa955ce27a8119b9
52080 F20101108_AABNDF carr_s_Page_113.pro
2a0b7d2ad13a69cf7ef5195728e3123e
3d2ad52acd1154ac0beab89ebc633c445bdac509
49299 F20101108_AABNCR carr_s_Page_038.pro
135c126ffa2c9a40810a08857075abdc
43d92525d912ab6fcb67fb5789b6ec4b51596973
22950 F20101108_AABMXL carr_s_Page_040.QC.jpg
fcee60d47ec3ec1683c9c2cf4b7b038e
fede788cfa9b3ec0840ed99495a47648f5752ccd
F20101108_AABLUJ carr_s_Page_035thm.jpg
8861eccf49c4a1075e92c95989b359eb
9f310715af1873f248776cdd6c5fa5c1f96aec9b
24294 F20101108_AABMAD carr_s_Page_033.QC.jpg
79459515a13776a67fdf889b00347951
16192b5278b712161e88eec907b4e0d527281388
55939 F20101108_AABNDG carr_s_Page_127.pro
a9d3316ee8133dbcc687b0f9dc86cda8
87cabaa9df8134ca88afeff8705e64b6c14569e8
47994 F20101108_AABNCS carr_s_Page_046.pro
e726f1b92ddb58f94900933a2a50e148
95ea36ce92661ab2c29b0f89ceef0d762be34c12
73738 F20101108_AABMYA carr_s_Page_029.jpg
a7eda2349d5774f6e3a123a0b5af1120
0d4542c3c10a16c8435b9ec6646a88eafe055482
27267 F20101108_AABMXM carr_s_Page_171.QC.jpg
1b99f49e021c71d513fa62654ef63d5e
02260704858a12614a4ef5bc8493bc751021175b
37463 F20101108_AABMWX carr_s_Page_162.jpg
bec57d1c303b1a0cdb36a1a04dde9577
3fb0f7a006e56755d409a33b1ed66b0ad1c0ee8d
1066 F20101108_AABLUK carr_s_Page_099.txt
3dddcc0cf03cf60fea9ce7a8e05689de
47a8afaf5cba5deff0ff77d49e5bdf2bd750123c
117893 F20101108_AABLTV carr_s_Page_095.jp2
e13b1e88d37bc8081876c347dfa17f24
09419ee5faaf65ea7b5d8044becc5d239450aeed
3220 F20101108_AABMAE carr_s_Page_003.QC.jpg
af51a4c7aad510a71b1fc15298d2b0ee
883102b32973f5fefdfa062a3baf9b8286ccb70a
52349 F20101108_AABNDH carr_s_Page_129.pro
75c06bd6a455dd513a176e9eb5c660d6
8869478994217763776574bd3f84ae659c508fa2
50840 F20101108_AABNCT carr_s_Page_050.pro
ded567ebcd025165d0a29ff51bf13b18
bcbe4c3896942bce614d6283b5ed8368f209bccd
69346 F20101108_AABMYB carr_s_Page_031.jpg
d43a2f6fe2e84890ad876f8f20fd7079
a8420389a7895f5e4dc86f1a607f6322108036d3
F20101108_AABMXN carr_s_Page_104.tif
1434d9c5cbb7333a4790b941ad6fc6b2
b04502cd7174bf18e439c80f20277bd5a5d7bfba
1754 F20101108_AABMWY carr_s_Page_156.txt
bb8b3de6390a164126a3e17d7e0ea9c4
0ce504237dd6b0f773359f84759b11c213590476
F20101108_AABLUL carr_s_Page_116.jp2
b5de1867604ae4d83986911b721e8f0d
3ba138c31898e1e3ec51e712a02bc619999c878e
17407 F20101108_AABLTW carr_s_Page_103.QC.jpg
7dfdbc252464006d3a9697bd9cd1ecc0
01e4a3c4fb230e76d6a662b4a7d5ae38c9d0ee40
65179 F20101108_AABMAF carr_s_Page_133.jpg
d1aca287e0967968b32fd6a61f72b454
36c278bf2432f753b8b63638e197e9bfc17fefa5
11561 F20101108_AABNDI carr_s_Page_137.pro
a4c888b2bfb1dd172da78cc372d91e2f
a56af90019462f3760d1becf6e8756f311a8ec6c
52279 F20101108_AABNCU carr_s_Page_053.pro
8962052df5a1e04427fb30d855e92c87
28ebca84ec020596714a9780e5759fd52930a57b
77146 F20101108_AABMYC carr_s_Page_049.jpg
c9f8f821d864991a9d7aefa4f9feb6f4
8430a305516dc9cb901b6f0903bce55368c56a5f
3903 F20101108_AABMXO carr_s_Page_073thm.jpg
737a41c78a983e627a8ebd25fd67fa95
cf5f3cd5bb7fb52daf6f443624be528ca9f0b890
28760 F20101108_AABMWZ carr_s_Page_147.QC.jpg
e4e4177869cee6063fbdb9ff2a55548f
7e8975f07cd50b8e2d3f0d2ba9f210d1a8e5c512
135821 F20101108_AABLUM carr_s_Page_179.jp2
9c65e47f6e4a394bd3ac8bcf12ded0f1
cbb971f5e03ce4d37f362c041a2937f033beaaac
76405 F20101108_AABLTX carr_s_Page_143.jpg
da58ca3f5b8bbe55458697f7992f1813
97a6c822f0b8f7d724ea7773b941089a0b6ce6a6
F20101108_AABMAG carr_s_Page_184.tif
6ec68d07d441f2e57a99b493e9be3e6f
bff275462ea4a10fceb1e8bebe2463ed59b2f8db
2420 F20101108_AABLVA carr_s_Page_176.txt
847562862fdb4fb0401c70f95578116a
43b03b683460d349e9f7b673c596d6622aab03fa
20775 F20101108_AABNDJ carr_s_Page_140.pro
9d962c05ae409952790dbf4f98f80d73
6a9eb02fa6904587a0bb088492987fcbe3e80aa8
56012 F20101108_AABNCV carr_s_Page_054.pro
bc7c095dcd2c5436cb08809240289044
5385719b91df59666e8db78fa3cd014c84b3638c
78991 F20101108_AABMYD carr_s_Page_058.jpg
15cb5055c70d15f652cf21ac61e64fd7
3830d5079e683c152ac2bba73f729da77bfc0ed6
1561 F20101108_AABMXP carr_s_Page_158.txt
1a95639310c71b109920c6fdaab250ae
981e5f82102ffa6ed69eab5276bfb33256f09a2e
80942 F20101108_AABLUN carr_s_Page_064.jp2
b68a73f007e4efa2d05c0ab1e63c2f2e
a301936313d2083e8a8d7af3512f1df4cad2c055
118728 F20101108_AABLTY carr_s_Page_049.jp2
a883b85fecc8c8a94b0ea32788af36fd
972085f050367d8a17727b29519c09fe287a0bd4
7113 F20101108_AABMAH carr_s_Page_094thm.jpg
30c0e4d9c38054f99f06b9b1d10d7232
fb31b98a7e0db5ecb763444ef7939ff24c1f42c4
9380 F20101108_AABLVB carr_s_Page_136.pro
aaac191437b209a00b9092977e6c636b
7f15b8d66ec6b2274eb8832f122b27a79366bd0d