PAGE 1 1 LONGLEAF PINE ( Pinus palustris Mill) ECOSYSTEM RESTORATION ON COASTAL WET PINE FLATS: DEVELOPING A MONITORING PROGRAM USING VEGETATION AND SOIL CHARACTERISTICS By GEORGE L. McCASKILL 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 2008 PAGE 2 2 2008 George L. McCaskill PAGE 3 3 To my beloved wife Susana, my wonderful son Pedro, my mother, brothers and sisters PAGE 4 4 ACKNOWLEDGMENTS I would like to thank m y supervisory committ ee members, Drs. Shibu Jose (Chair), Andy Ogram, Alan Long, Wendell Cropper, Eric Jokela and Jack Putz for their advice, time and support. In particular, I want to thank Dr. Shi bu Jose for providing me the opportunity to attend graduate school and providing me with funds to co mplete the project. I would also like to thank Dr. Andy Ogram for his extensiv e advice and laboratory assistan ce concerning the soil microbial analysis procedures. His knowledge and help on soil microbial ecology helped make this study possible. Wendell Cropper with his speci alized knowledge on ecosystem modeling and ecosystems in general provided invaluable improve ments to my research. Thanks to Drs. Alan Long, Jack Putz and Eric Jokela for providing me with the scientific / philosophical continuity necessary for evaluating pine planta tions and natural forests. I woul d also like to thank Dr. Craig Ramsey and Dr. Ashvini Chauhan for their assistance from field work to la boratorial analysis of the project. The efforts of Tim Baxley in data collection work and Chris Dervinis with making the HLPC operational are truly appreciated. I would also like to thank my fellow graduate students, Drs. Susan Bambo, Pedram Daneshgar, Diomides Zamora, and Dawn Henderson for their friendship and support during my studies. In addition, I would also like to thank the Florida Division of Forestry for funding the restoration site, and the personnel of Topsail Hill State Park, St Marks National Wildlife Refuge, and the Chassahowitzka Wildlife Management Area of Floridas Fish & Wildlife Commission for providing permission and assistance in establishing the reference sites. I could not give a full acknowledgement without giving special thanks to my wife Susana and my son Pedro for their suppo rt and patience throughout our time at the UF. I would also like to thank my mother, brothers and sisters fo r always supporting me throughout my extended period of education. PAGE 5 5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................12 CHAP TER 1 MONITORING LONGLEAF PINE RESTOR ATION IN C OASTAL WET PINE FLAT COMMUNITIES......................................................................................................... 14 Longleaf Pine Ecosystems......................................................................................................14 Monitoring Restoration Success............................................................................................. 15 Monitoring Soil Characteristics.............................................................................................. 16 Developing a Monitoring Program......................................................................................... 17 2 FOREST STRUCTURE AND PLANT SPECIES DIVERSITY IN WET LONGLEAF PINE FLATS ACROSS A CHRONOSEQUENCE ............................................................... 21 Introduction................................................................................................................... ..........21 Materials and Methods...........................................................................................................25 Study Sites.......................................................................................................................25 Forest Age Classes.......................................................................................................... 27 Field Measurements......................................................................................................... 28 Data Analysis...................................................................................................................29 Results.....................................................................................................................................30 Overstory Stand Structure...............................................................................................30 Understory..................................................................................................................... ..30 Discussion...............................................................................................................................31 Overstory.........................................................................................................................31 Plant Species Diversity.................................................................................................... 32 Conclusions.............................................................................................................................34 3 PATTERNS OF SOIL CHEMICAL AND MICROBIAL PROPERTIES ALONG A CHRONOSEQUENCE IN W ET LONGLEA F PINE FLATS OF FLORIDA...................... 45 Introduction................................................................................................................... ..........45 Materials and Methods...........................................................................................................50 Study Areas.....................................................................................................................50 Soil Sampling and Preparation........................................................................................ 50 Soil Chemical Analysis................................................................................................... 51 Net Nitrogen Mineralization............................................................................................ 51 PAGE 6 6 Microbial Biomass........................................................................................................... 52 Data Analysis...................................................................................................................54 Results.....................................................................................................................................55 Soil Types, Soil Organic Matter, and Soil pH................................................................. 55 Net Nitrogen Mineralization............................................................................................ 55 Microbial Properties........................................................................................................ 56 Discussion...............................................................................................................................57 Conclusions.............................................................................................................................59 4 RELATIONSHIP BETWEEN VEGETATION AND SOIL CHARACTERISTICS IN WET L ONGLEAF PINE FLATS AL ONG FLORIDAS GULF COAST............................ 67 Introduction................................................................................................................... ..........67 Materials and Methods...........................................................................................................69 Study Areas.....................................................................................................................69 Field Measurements......................................................................................................... 69 Soil Sampling and Preparation........................................................................................ 70 Soil Chemical Analysis................................................................................................... 70 Mineral Nitrogen Fluxes..................................................................................................71 Bacterial Abundance and Microbial Dyna mics............................................................... 71 Experimental Design and Analysis................................................................................. 73 Results.....................................................................................................................................74 Nitrifying Bacteria and Nitrogen Mineralization............................................................ 74 Overstory.........................................................................................................................75 Understory..................................................................................................................... ..75 Discussion...............................................................................................................................76 Conclusions.............................................................................................................................78 5 MONITORING RESTORATION SUCCESS USING VE GETATION AND SOIL AS KEY INDICATORS: CASE STUDY OF A WET LONGLEAF PINE FLATS RESTORATION PROJECT................................................................................................... 85 Introduction................................................................................................................... ..........85 Materials and Methods...........................................................................................................88 Pt. Washington Restoration Site...................................................................................... 88 Pine Survival and Growth............................................................................................... 90 Vegetation Sampling....................................................................................................... 90 Reference Sites................................................................................................................ 91 Soil Sampling and Preparation........................................................................................ 91 Data Analysis...................................................................................................................92 Pine survival and growth.......................................................................................... 92 Understory................................................................................................................ 92 Biogeochemical indicators....................................................................................... 93 Results.....................................................................................................................................95 Ecological Classification................................................................................................. 95 Pine Growth and Vegetation Control.............................................................................. 96 Treatment Effects-Bioge ochem ical Indicators................................................................ 96 PAGE 7 7 Discussion...............................................................................................................................97 Conclusions...........................................................................................................................100 6 SUMMARY AND CONCLUSIONS...................................................................................112 Research Implications in Coastal We t Longleaf Pine Flats Restoration .............................. 117 APPENDIX SPECIES CODE LIST............................................................................................ 120 LIST OF REFERENCES.............................................................................................................122 BIOGRAPHICAL SKETCH.......................................................................................................138 PAGE 8 8 LIST OF TABLES Table page 2-1 Qualitative classification system for downed and standing deadwood............................. 36 2-2 Forest structural and plant spec ies diversity m eans among age classes............................ 36 2-3 Indicator values for plant species in three forest age classes. ............................................ 37 3-1 Soil and stand properties between reference sites..............................................................61 3-2 Soil chemical and microbial bi om ass means between age classes.................................... 62 3-3 Soil nitrogen mineralization means (Nmi n) for dry season 2002 and wet season 2005. ... 62 3-4 Differences in soil biogeochemical relationships based upon Spearm an rank correlations r as stratified by forest age class (n = 48)...................................................... 63 4-1 MPN enumerations of nitrifying bacteria in young and old longleaf pine forest soils. ..... 80 4-2 Ammonification and nitrification in y oung and old longleaf pine forest soils. .................80 4-3 Soil biogeochemical relationships with stand attributes based upon Spearm an Rank correlations r as stratified by forest age class (n = 48)...................................................... 81 5-1 Correlations and biplot scores for the biogeochem ical variables by pine.......................102 5-2 Plant Indicator Values (IndVal) (percent of perfect indicati on) with associated biogeochem ical variable by pine flat type....................................................................... 102 5-3 Correlations and biplot scores for the biogeochem ical variables by forest age class...... 103 5-4 Plant Indicator Values (IndVal) (percent of perfect indicati on) with associated biogeochem ical variable by forest age class.................................................................... 103 5-5 The means for soil biogeochemical variable s between reference si te locations and the Pt. W ashington restoration site........................................................................................ 104 5-6 Pt. Washington actual vs. predicted indicator values. ..................................................... 104 A-1 Species list.............................................................................................................. ........120 PAGE 9 9 LIST OF FIGURES Figure page 1-1 Floridas Gulf Coast Flatwoods zone wh ere the w et pine flat sites are located................ 20 2-1 Locations of the Pt. Washington Longleaf Pine Restoration site ( ) and the reference sites within Gulf Coast Flatw oods subecoregion of Florida. .............................................38 2-2 Nested plot sampling desi gn applied at three different s ites (age classes) for each reference location............................................................................................................. ..39 2-3 Mean stand density (trees per he ctare) along a 110-y ear longleaf pine chronosequence as m easured from 26 differently aged stands.......................................... 39 2-4 Mean stand DBH, height, BA, and volum e along a 110-year longleaf pine chronosequence as measured from 26 differently aged stands.......................................... 40 2-5 Downed woody debris and standing deadwood (snag) accum ulations along a 110year longleaf pine chronosequence as m easured from 26 differently aged stands............41 2-6 Decomposition levels by forest age cla ss as m easured from 26 differently aged stands..................................................................................................................................42 2-7 Composition of understory vegetation by forest age class. ............................................... 42 2-8 Shannon-Wiener Diversity and Coleman Rarefaction indices along a 110-year longleaf pine chronosequence as m easured from 26 differently aged stands.................... 43 2-9 Mean stand density versus the Shann on-W iener Diversity index and mean stand height versus the Coleman Rarefaction index as measured from the young, mid-age and mature age longleaf pine stands.................................................................................. 44 3-1 Soil organic matter content versus soil m oisture as measured from 26 differently aged stands.................................................................................................................... .....63 3-2 Soil pH versus soil organic matter content (percent) as m easured from 26 differently aged stands.................................................................................................................... .....64 3-3 Total net nitrogen mineralization, amm onification and nitrification rates (m g -1 nitrogen / kg -1 soil / month -1 ) along a 110-year chronosequence as measured from 26 differently aged stands..................................................................................................64 3-4 Trends for microbial biomass carbon (Cmb) and net nitrogen mineralization rates (Nmin) along a 110-year longleaf pine chronosequence as measured from 26 differently aged stands....................................................................................................... 65 PAGE 10 10 3-5 Microbial biomass carbon ve rsus net nitrogen m ineraliza tion rates as measured from 26 differently aged stands..................................................................................................65 3-6 Fungal biomass carbon ( C ) along a 110-y ear longleaf pine chronosequence as m easured from 26 differently aged stands......................................................................... 66 3-7 The fungal-to-microbial biomass ratio and fungal biom ass carbon levels (means) during the earlier and later portions of chronosequence respectively, as measured from 26 differently aged stands along th e 110-year longleaf pine chronosequence.......... 66 4-1 Net nitrogen mineralizati on versus stand volum e as measured from 26 differently aged stands.................................................................................................................... .....82 4-2 The fungal biomass (FB)-to-microbial biom ass (MB) ratio versus stand height as m easured from 26 differently aged stands......................................................................... 82 4-3 Fungal biomass carbon (Cfb) versus stand basal area (BA) as measured from stands grouped within the mid-aged and mature age classes only................................................ 83 4-4 Coarse woody debris accumulation versus fungal biomass carbon (Cfb) as measured from 26 differently aged stands......................................................................................... 83 4-5 Coleman Rarefaction index versus the fungal biom ass (FB)-to-microbial biomass (MB) ratio as measured from 26 differently aged stands.................................................. 84 4-6 Shannon-Wiener diversity H index vers us the fungal biom ass (FB)-to-microbial biomass (MB) ratio as measured fr om 26 differently aged stands.................................... 84 5-1 Pine flat type determined by a threedim ensional ordination biplot derived from Canonical Correspondence Analysis (CCA) of 192 plots usin g understory plant species abundance and soil biogeochemi cal data including the Pt. Washington restoration site..................................................................................................................105 5-2 A threer-dimensional ordination biplot derived from Canonical Correspondence Analysis (CCA) of 192 plots using unde rstory plant species abundance and soil biogeochemical data collected within the young, mid-aged, mature age class, and the Pt. Washington restoration site........................................................................................ 106 5-3 Monthly variation of total nitrogen mine ra lization, ammonification and nitrification rates (mg-1 kg-1 month-1) obtained from field incubation of soils (untreated) during 14 months before and after the 2002 treatments................................................................... 107 5-4 Net nitrogen mineralization means mg (NH4 + + NO3 -) / kg-1 soil / month for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometuron methylhexazinone mix and Arsenal: imazapyr........................................................................... 107 PAGE 11 11 5-5 Net ammonification mean monthly rates (mg-1 NH4 + / kg-1 soil / month) for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometuron methylhexazinone mix and Arsenal: imazapyr........................................................................... 108 5-6 Net nitrification mg -1 N03 / kg -1 soil / month; for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometur on methylhexazinone mix, and Arsenal: imazapyr; applied in different growing s easons, frequencies, and time of year.............. 108 5-7 Microbial biomass carbon (Cmb) mg -1 C / kg -1 soil; for the control, Oust: sulfometuron methyl, Velpar: hexazinone sulfometuron methylhexazinone mix, and Arsenal: imazapyr.....................................................................................................109 5-8 Microbial biomass carbon (Cmb) mg-1 carbon / kg-1 soil from soils treated only one year and two consecutive years of applications............................................................... 109 5-9 Microbial biomass carbon (Cmb) mg -1 carbon / kg -1 soil levels measured at the reference sites and the Pt. Wa shington restoration site.................................................... 110 5-10 Fungal biomass carbon mg -1 carbon / kg -1 soil; for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometur on methylhexazinone mix, and Arsenal: imazapyr....................................................................................................................... ....110 5-11 Pools and fluxes of nitrogen in the RE SDYN restoration m ode l. MP, metabolic pool; grass&forbs, holocellulose pool; shru bs, lignocellulosic pool; and CWD, woody pool..................................................................................................................................111 PAGE 12 12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy LONGLEAF PINE ( Pinus palustris Mill) ECOSYSTEM RESTORATION ON COASTAL WET PINE FLATS: DEVELOPING A MONITORING PROGRAM USING VEGETATION AND SOIL CHARACTERISTICS By George L. McCaskill August 2008 Chair: Shibu Jose Major: Forest Resources and Conservation Longleaf pine ecosystem rest oration should include more than reforestation or the application of prescribed fire. It must include the restoration of all the major functions and processes within the forest eco system along with restoring overs tory and understory species composition. Despite many longleaf pine restorati on projects on coastal pine flats, there is no monitoring protocol in place to evaluate the succe ss of an all-inclusive restoration effort. The goal of this study was to estab lish an ecological trajectory us ing selected indicators for wet longleaf pine flats as a monitoring framework for restoration projects. The first specific objective was to quantify the vegetational attributes of longleaf pine flat ecosystems along a chronosequence (2 -years after stand re placement to 110-years-old) of stands from within the Gulf Coast Flatwoods zone in Florida. Overstory structure and understory plant species diversity were quantified along the chro nosequence. Mean diamet er at breast height (dbh), height, and basal area in creased until 60-70 years, and then declined. Stand volume continued to increase. Stand density decrea sed before reaching a steady state. Coleman rarefaction and Shannon-Wiener diversity indices for understory plants ex hibited opposite trends during early stand development, but reached equilibrium during the mature (> 90 years) phase. PAGE 13 13 The second objective was to ex amine soil chemical and micr obiological properties along the same chronosequence. Ne t nitrogen mineralization (Nmin), soil microbial biomass carbon (Cmb), and fungal biomass carbon (Cfb) increased from the young to th e mid-aged age stands and declined from the mid-aged through the matu re age stands. Ammonium production dominated nitrogen cycling and ammonium enrichment occu rred on these wet sites by reduction of nitrate (the DNRA pathway). The biogeoche mical attributes showed that Floridas Gulf coastal pine flats reach a self-organizing threshold after 85-90 years. The third objective was to examine the in terrelationships between the structural (vegetative) and functional (so il biogeochemical) attributes. Nmin, Cmb and Cfb increased with increases in dbh, height, basal ar ea, and volume. Plant species di versity decreased as the FB-toMB ratio increased. Nitrate levels and nitrifyi ng bacteria numbers were higher in young forest soils than old forest soils. Based upon the indicat ors, coastal longleaf pi ne flats reach a steady state threshold with a lower and less variab le (tighter) nitrogen cycle at 90 years. The final objective was to determine if observe d structural and functi onal attributes were useful for evaluating restoration projects. An ongoing restoration project at the Pt. Washington State forest was evaluated for its ecological traj ectory following various restoration treatments involving herbicides. The site was determ ined to be a wet flatwoods based upon environmental ordination and plant species indi cator analysis. Herbicide use in creased soil microbial biomass carbon and net nitrogen mineralization rates. Imazapyr was the most effective herbicide treatment for this wet pine flat s site based upon the level of sh rub control, minimum impacts on herbaceous species diversity, and desired st ructural attributes of the overstory. Key words: Longleaf pine, reference communities, monito ring, ecological indicators, herbicides. PAGE 14 14 CHAPTER 1 MONITORING LONGLEAF PINE RESTORAT ION IN C OASTAL WET PINE FLAT COMMUNITIES Longleaf Pine Ecosystems The longleaf pine (Pinus palustris Mill) ecosystems that historically dominated the lower Coastal Plain from Virginia to Texas currently oc cupies less than 3 % of its original area (37 million ha) (Frost, 2006). This reduction in area has resulted in a great loss of habitat necessary for many plant and animal species (Wade et al 2000; Van Lear et al. 2005). Longleaf pine ecosystems are naturally maintained by frequent fires that reduce vegetative competition during pine seedling and sapling devel opment (Boyer, 1990). Fires, natura l or prescribed, have become severely restricted, especially by urban expans ion because of liability and property damage concerns (Achtemeier et al. 1998; Haines et al. 2001). For the last thirty years, forest industries in the South preferred to replace longleaf pine stands with slash pine ( Pinus elliottii Engelm.) on wet sites and with loblolly pine ( Pinus taeda L.) on upland areas (Croker & Lande rs, 1987). Slash and loblolly pines are considered easier to regenerate and managers have little need to address the long leaf pines unpred ictable period of establishment (grass stage). Furt hermore, they also reach comme rcial size faster than longleaf pine, which shortens the econom ic rotation (Outcalt, 2000). In recent years, there has been a great deal of attention given to the restoration of the extensive and species-ric h longleaf pine ecosystem. There ha ve been attempts to restore 400,000 ha of longleaf pine in the Sout heast during the past decade (W MI, 2006). This effort creates a need for monitoring protocols to be in place fo r evaluating the success of these restoration efforts. While established monitoring guideline s and programs are active for many of the other forest ecosystems in other parts of the U.S. (ERI 2003), the lack of such established directives PAGE 15 15 can hinder longleaf pine restorat ion projects in establishing functional and self-sustaining ecosystems across its originally exte nsive range (Devries et al. 2003). Monitoring Restoration Success Three established strategies for assessing a restoration effort are direct comparison, attribute analysis, and trajectory analysis (SER, 2004). Perm anent-plot studies have been used to directly identify changes over ti me after stand replacing harvests or other impacts. Attribute analysis uses the measurement of ecological indi cators to evaluate ecosystem conditions without directly considering patterns over time. Trajectory an alysis uses data collect ed over periodic time intervals to identify if restora tion trends are toward a refere nce condition (SER, 2004). We used a combination of attribute analysis and trajectory analysis to monitor our restoration project. Since expensive permanent-plot studies are genera lly limited to less than 30 years, conducting a trajectory analysis employing a space-for-time subs titution or chronosequence can be an efficient alternative for describing general trends over time or to test hypot heses based upon forest succession (Pickett, 1989). Chronosequencial studi es make use of a group of sites that have similar biotic, climatic, soil biogeochemical, and hi storical characteristics, but differ in age since a harvest or other stand replacing disturbance (Pickett, 1989). By comparing the different aged sites, one can identify changes in composition or function between decades, centuries, or even millenniums (Williamson et al. 2005). There is a critical requi rement for the different aged sites to be subjected to the same historical conditions and have the same speci es available over the ch ronosequence to give validity for using the space-for-time substitutions. One must also deal with separating spatial variability from the variability associated w ith time (Veldkamp et al. 1999). A recent chronosequence study examined the relationships between stand development and understory vegetation on 15 stands ranging from 7 to 427 years in a mixed conifer forest along the PAGE 16 16 California-Oregon border. Regression an alysis showed canopy openness was positively correlated with total understory cover, species richness, diversity, and composition. Surprisingly, no correlations were observed between any of the measured stand attributes. Shrub and graminoid species were negatively correlated, an d forbs were positively correlated, with stand age (Jules et al. 2008). Another study used detailed forest inventory and climatic data from 43 stands along a 250-year chronose quence to assess the effects of disturbance and climate on biomass accumulation patterns across Russia. Regr ession analysis indicated as expected the highest biomass increments in the warmest region s and the lowest in the coldest regions. Spruce ( Picea spp.) and birch (Betula spp.) forests had the highest biomass increments while larch ( Larix spp.) and aspen ( Populus spp.) forests had the lowest biomass accumulation. The faster growing spruce and birch forests had declines in biomass accumulation rates after 150 years whereas the slower growing la rch and aspen never showed declines duri ng the 250-year chronosequence (Krankina et al. 2005). Monitoring Soil Characteristics In addition to vegetative characteristics, m ineral pools, and the mineralization of key elements have been identified as important attrib utes for evaluating restor ation success in recent years (Mller et al. 2000; Mller and Lenz, 2006). Du ring the last decade there has been a major effort at assessing the effects of different forest management practices on the long-term soil productivity of southern pine forests (Burger a nd Kelting, 1999), including coastal wet pine flats (Lockaby and Walbridge, 1998; Lister, 1999; Burg er and Xu, 2001; Burdt, 2003). These studies have assessed treatment effects utilizing a set of so il indicators (Kelting et al. 1999) including soil pH, soil organic matter content, soil moistu re content, and the mi neralization levels of nitrogen, and phosphorus (Reynolds et al. 2000; Redding et al. 2004). For example, a recent chronosequence study examined the relationshi p between biomass accumulation and nitrogen PAGE 17 17 availability over 87 years in Populus grandidentata forests. Overstory biomass increment increased with stand age while understory bioma ss levels decreased. Net nitrogen mineralization rates were found to decrease during the first 18 years after harvest than increase over the next 70 years (White et al. 2004). In an earlier investig ation, forest floor microbial biomass was studied in a chronosequence of northern hardwood forest stands ranging from 3 years after clearcut to 120 years. Microbial biomass increased during the early successional stage, decreased during the mid-aged stage, and then increased during the late successional stage. Soil organic matter followed a pattern similar to microbial biomass. There was no trend in the fungal-to-bacterial ratio along the chronosequence. Soil moisture was strongly and positively correlated with fungal biomass. Soil pH was negatively correlated w ith fungal biomass. Finally, ammonium (NH4 +) production increased from the early to mid-aged stages and then decreased from the mid-aged to late successional stages (Taylor et al. 1999). Developing a Monitoring Program A good m onitoring program should be well focused on just a few key indicators to provide for statistically sound information (Lindenmay er, 1999). The standards for restoration are obtained from measuring key environmental indi cators at the restoration site and comparing them to established reference communities (SER, 2004). In ecological re storation, the pathway from the degraded condition to the restored, self-s ustaining condition is called the ecological trajectory (Stanturf et al 2001). Predicting the ecological traject ory of a longleaf pine forest is difficult because of the great variet y of disturbance regimes associat ed with southern pine forest ecosystems along the Gulf Coast (Palik et al. 2002). To define when a given ecological trajectory has reached a self-s ustaining state it is important to establish some specific goals for the restoration project (Hobbs & Harris, 2001). Two notable standards are to rest ore viable populations of key nativ e species in natural patterns PAGE 18 18 of abundance and distribution, and to sustai n key geomorphologic, hydr ological, ecological, biological, and evolutionary processes within the normal range s of variation (ecological integrity; Mller et al. 2000). Forest structur e and plant species com position are two of the indicators being monitored in this study to capture the successional and developmental forces. Soil chemical properties, net ni trogen mineralization, and soil mi crobial dynamics were also included as indicators to insure that key biogeochemical, ecologi cal, and biological processes are also being evaluated (Harri s, 2003; Mller and Lenz, 2006). How does one determine if these goals are being achieved along the successional pattern? The normal range of variation along a spatial s cale can be determined by using a series of reference communities that are evenly distributed along the distinct ecologically identified range, to compare with the restoration site (Harris, 1999). To evaluate changes in restoration along the chronosequence, each reference community had to contain stands representing distinct ages distributed evenly as possible al ong the 110-year scale (Mller, 1998). In summary, the following steps have been recommended to insure that a monitoring plan functions properly: a) Set monitoring goals, b) identify the resources to monitor, c) establish threshold levels, d) develop a sampling design, e) co llect and analyze data, a nd f) evaluate results (Block et al. 2001). The overall goa l of this study was to establis h an ecological trajectory using selected indicators for wet longleaf pine flats so that it can be used as a monitoring framework for restoration projects. The next four chapters will address the following four specific objectives of this study. 1. Quantify the vegetational attributes of longleaf pine flat ecosystems, along a chronosequence (2-years after a stand replaci ng disturbance to 110-ye ars) of stands from within the Gulf Coast Flatw oods zone of Florida. PAGE 19 19 2. Examine soil chemical (soil organic matter c ontent, pH, plant-available phosphorus, net nitrogen mineralization) and microbiologica l (microbial biomass carbon, fungal biomass carbon) properties along the same chronosequence. 3. Examine the interrelationships between the structural (vegetative) and functional (soil biogeochemical) attributes. 4. Determine if the observed structural and functio nal attributes could be used to evaluate restoration projects. PAGE 20 20 Figure 1-1. Floridas Gulf Coast Flatwoods zone wh ere the wet pine flat s ites are located (Florida DEP, 2002). PAGE 21 21 CHAPTER 2 FOREST STRUCTURE AND PLANT SPECIES DIVERSITY IN WET LONGLEAF PINE FLATS ACROSS A CHRONOSEQUENCE Introduction In recent years, there has been a great deal of interest in restorati on of the longleaf pine ecosystem one of the most threatened ecosystems in the United States with less than 3% of its original extent remaining. Ther e have been attempts to restor e 400,000 hectares of longleaf pine in the Southeast during the past decade alone (WMI, 2006). This situatio n creates a need for developing monitoring protocols to evaluate the success of these restoration efforts. While established monitoring programs ar e in place for many forest ecosystems in other parts of the U.S. (ERI, 2003), the lack of such establishe d guidelines can hinder th e restoration of the longleaf pine forest as a functiona l, self-sustaining ecosystem acro ss its former range (Devries et al. 2003). Community structure and species composition are two key attr ibutes often evaluated in restoration projects (Brockway et al. 2005). However, reliable information on the ecological trajectory of longleaf pine ecosystems have hamp ered monitoring of restoration projects in Florida and elsewhere in the Sout heast. Although past research has examined the structure and species composition of upland longleaf pine ecosyst ems, little information exists on the temporal patterns of forest structure and plant species diversity in wet longleaf pine flat communities located along the coastal lowlands of Floridas Gulf Coast (Michener, 199 9). Wet pine flats are pine-dominated, poorly drained, broad plain wetlands (Stout and Mari on, 1993; Harms et al. 1998). PAGE 22 22 In Florida, plant species ri chness has been found to increas e with soil moisture until an ecotone between wet pine flats and cypress swamps is reached (Huck, 1986; Walker, 1993; Kirkman et al. 2001; Walker and Silletti, 2006). This ecotone is the zone where one finds wet flatwoods and wet savanna subtypes of the coas tal wet pine flat (Mes sina and Conner, 1998). Their overstories are dominated with varying mixtur es of longleaf and slash pines, but also might contain a component of Choctawhatchee sand (Pinus clausa var. immuginata) and/or pond (Pinus serotina ) pine (Parker and Hamrick, 1996). The environment for Floridas wet pine flat s is the 1,240 km-long Gulf Coast, containing sounds, bays, and offshore islands. This coasta l landscape is continuous ly shaped by active fluvial deposition and shore z one processes which promote and maintain the formation of beaches, swamps and wet mineral flats. The lo cal relief ranges from 0 to 20 m in elevation. Annual precipitation ranges fr om 1300 mm and average annual temperatures vary between 19-21 C. Growing seasons are long, lasting 270-290 days (McNab and Avers, 1994). Soil parent material consists of marine deposits containing limestone, marl, sand, and clay. The dominant soils are Aquults, Aquepts, Aquods, and Aquents. These highly acidic soils have thermic and hyperthermic temperature regimes and an aquic moisture regime. The major forest type of this region is the longleaf-sla sh pine flatwoods, while water oak ( Quercus nigra), swamp tupelo ( Nyssa sylvatica var. biflora ), sweetbay (Magnolia virginiana ), and cypress ( Taxodium sp.) are found along the major river drainages and is olated depressions (McNab and Avers, 1994). Floridas subecoregional Gulf Coast Flatwoods (Figure 2-1) covers the majority of this geographical area where both pine savannas and coastal flatwoods occur in close association with cypress ponds (Myers and Ewel, 1990; Gri ffith et al. 1994). Because of the growing PAGE 23 23 conditions, wet pine flats are hi ghly productive ecosystems, and represent more than one million ha in the Southeast (Burger and Xu, 2001). There are almost 200 rare vascular plant ta xa found in the great variety of habitats classified as longleaf pine ecosystems. In additio n to the majority of them being found in Florida (Collins et al. 2001), the richest sites are found in these wet pine flats and their associated wetlands (Walker, 1993). The uniqueness of Florid as wet pine flat communities make them crucial for plant species diversity and the rarity of plants to be evaluated in any monitoring plan for restoration (Walker, 1993; Collins et al. 2001; LaSalle, 2002). One of the ways by which restoration progr ess can be monitored is by examining the ecological trajectory of the rest ored site and comparing it with the ecological trajectory of reference sites. Post-stand replacement seconda ry forest succession has been well studied in other ecosystems and is considered to follow four stages of development (Peet and Christensen, 1980; Oliver, 1981). Stand Initiation commences after a stand -replacing disturbance has occurred. Plants regenerate from sprouts, seed banks or newly dispersed seeds. Any advanced regeneration (saplings and seedlings) is releas ed by the disturbance and commences accelerated growth. The Stem Exclusion stage is when the individual tree s in the stand come under fierce competition for light, water, and nutrients. Canopy closure results in a great reduction of stand density as the residual stocking c ontains fewer, larger trees. Understory Reinitiation occurs when some of the dominant overstory trees be gin to die forming gaps for new regeneration. Finally, the Old Growth or Steady-State stage is reached when gap dynamics dominates the landscape and the forest is now all-aged. Snag s and downed logs are also found throughout the landscape (Perry, 1994). Gap dynamics is major sour ce of regeneration in na tural longleaf pine forests (Brockway and Outcalt, 1998). Steady-state is tied to the ability of a system to be self- PAGE 24 24 organizing and resilient (Mlle r et al. 2000; Mller and Lenz, 2006). Self-organizing forces become especially apparent during the understory reinitiation stage of forest succession when the steady-state mosaic begins to form. When identifying patterns of succession along a chronosequence, stand changes caused by disturbance must also be cons idered (Frelich, 2002; Pickett and Cadenasso, 1995). The longleaf pine forest is a pyro-climax ecosystem which relie s on short fire return intervals to maintain the steady-state stage over other woody plant species (Wade et al. 2000). In coastal wet pine flats, wind and precipitation are also major shapers of longleaf pine communities. Hurricanes directly affect the canopy stru cture of longleaf pine stands through gale-forced winds, opening them up to sunlight and changing the compositi on of the flora and fauna that occupy them. Hurricanes also affect longleaf pine stands by the extensive flooding that accompanies the wind. Extended flooding can cause changes in both the above and below ground productivity (Johnston and Crossley, 2002; Palik et al. 2002). Anthropogenic effects caused by human activities in forests can also change the forest structure. Timber harvesting, grazing, and prescrib ed fires can cause changes in the structural complexity of forests having negative effects by exposing surface soils and reducing biodiversity (Redding et al. 2004; Van Lear et al. 2005). In addition, climatic changes such as increased atmospheric CO2 levels may affect the soil biotic co mmunity, adding to th e potential negative feedbacks toward the productivity of abovegr ound plant communities (Peacock, 2001; Frelich, 2002). Measuring forest structural data along a chronosequence will make it possible to evaluate change along the life cycle of coastal wet longleaf pine flats. The objective of this study was to examine stand structural a ttributes and understory plant species diversity along a 110-year chronosequence. We hypothesized that stand DBH, height, PAGE 25 25 basal area (BA), and volume would increase while stand density and plant species richness and diversity would decrease through the mid-age. We expected th ese parameters to reach a threshold or steady-state during the mature phase when the unde rstory reinitia tion stage of succession has begun. Quantification of this eco logical trajectory w ould help establish monitoring thresholds in terms of stand struct ure and plant species composition for restoration projects. Materials and Methods Study Sites Three study sites were establishe d along a wet pine flats located within three kilom eters of Floridas Gulf Coast. They were Topsail Hill State Park, St. Marks National Wildlife Refuge, and Chassahowitzka Wildlife Management Ar ea. They were found between the cities of Pensacola and Tampa Bay (Figure 2-1), a narrow z one that makes up the majority of the Natural Resource Conservation Services Eastern Gulf Coast Flatwoods ecoregion (MLRA 152A) and the National Oceanic and Atmospheric Associations Panhandle Coast unit of the Louisianan reserve (National Estuary and Ri ver Reserve System). Both of these federal designations make this zone unique from an ecological as well as hy drological perspective. Within the Eastern Gulf Coast Flatwoods zone is the Gulf Coast Flatw oods (75I) subecoregion of Florida (Figure 2-1; Griffith et al. 1994). Any environmental varia tions between these sites were minimized by establishing very specifically defined spatial scales. This was accomplished by stratifying the important segments of the whole system (e.g. Floridas Gulf coastal flatwoods) down to the smallest distinct scale as possible (e.g. wet pine flats with in 3 kilometers of the coast) in order to take meaningful measurements (Chertov et al. 1999; Frelich, 2002; Mller et al. 2000). The first 120 years of longleaf pine succession has been incl uded with wet pine flats situated within 3 km of the Florida Gulf coast as the temporal and spatial scales in this study. These two scales were PAGE 26 26 determined from an in-depth preliminary survey of stand conditions found within the Gulf Coast Flatwoods zone of Florida a nd at the restoration site. The herbaceous ground cover of wet longleaf pine flats is very diverse due to the warm temperatures and high rainfall. Broomsedge ( Andropogon virginicus) wiregrass ( Aristida stricta var. beyrichiana) witch grass (Dichanthelium spp .), goldenrod ( Solidago odora), meadow beauty ( Rhexia alifanus) fetterbush ( Lyonia lucida) and aster ( Aster adnatus) are found on both subtypes (Brewer, 1998). Pine savannas are distinguished from wet flatwoods by a greater abundance of beak sedge (Cyperus) nut rush ( Scleria cilliata ), bloodroot ( Lachnanthes caroliniana) pitcher plants ( Sarracenia) and orchids ( Calopogon) or ( Platanthera) Coastal flatwoods have a greater presence of titi ( Cliftonia monophylla) swamp tupelo, gallberry ( Ilex glabra) saw palmetto ( Serenoa repens ), and sweetbay. Where fire is restricted, catbrier ( Smilax pumila) can be a prevalent vine species (LaSalle, 2002). All three of the selected sites have a moistu re gradient as represented by cypress swamps, wet pine savannas, and wet pine flatwoods. All th ree sites have active restoration management programs where fire has been used for more th an 20 years on approximately a three-year-return interval. All of the sites are primarily managed to enhance habitat for threatened species associated with longleaf pine ecosystems, and are managed by a state or federal agency. The southern site on the spatial gradient is the Chassahowitzka Wildlife Management Area in Hernando County, FL. It is approxima tely 12,140 ha, and the soils are dominated by Basinger fine sands (sandy, sili ceous, hyperthermic spodic Psammaquents) and Myakka fine sands (sandy, siliceous, hypert hermic aeric alaquods) (Hyde et al. 1977; Spencer, 2004). The St. Marks National Wildlife Refuge in Wakulla and Jefferson Counties, FL consists of 25,900 ha and the majority of the soils are mapped as the Scranton series (sandy, siliceous, PAGE 27 27 thermic humaqueptic Psammaquents) and the Leon series (sandy, siliceous, thermic aeric Alaquods) (Reinman, 1985; Allen, 1991). Topsail Hill State Park in Sant a Rosa County, FL, contains 610 ha of some of the oldest longleaf pine stands in Florid a. The soils are Pickney sand se ries (sandy, siliceous, thermic cumulic humaquepts) and the Leon series (Overing and Watts, 1989; White, 2001). Forest Age Classes The 110-year chronoseq uence starts from the point of stand replacement to the oldest stands measured in our reference sites. The fo llowing age classes were used to stratify and analyze changes in forest structure and plant sp ecies composition within the different stands at each of the reference sites. There are 12 repli cates per age class for the stand data and 48 replicates per age class for plan t species data. The age classes pr ovide a means to identify the structure of stands within speci fic time periods along the chronos equence and to detect changes from one time period to the next (Mller, 1998; Aravena et al. 2002). In this paper, a tree is defined as a woody plant with a diameter at breast height (DBH) of greater than 10 cm. A sapling is a woody plant with a DBH of less than 10 cm but gr eater than 2.5 cm. Finally, a seedling is a woody plant that is less than 91.5 cm in height (Wenger, 1984). The young age class: A young age stand exists when the majority of the stocking (> 70%) can be found as seedlings and saplings. The average stand age should be < 20 years since replacement. The mid-aged class: The mid-age stand should have stocking (>70%) dominated by a mixture of poles and small sawlog size trees (1 0-30 cm DBH). The average stand age should be between 20 and 55 years old. PAGE 28 28 The mature age class: A stand is considered wi thin the mature age class when the majority of stocking (>70%) can be found as dominant sawlog trees (30-45 cm DBH). The stand age should be > 55 years old. Field Measurements Each referen ce location had a cluster of th ree one-hectare blocks, containing stands representing each of the three previously defined age classes. Each one-hectare block was subdivided into four ra ndomly placed 400 m2 measurement plots. Tree height and DBH were measured on all trees > 10 cm DBH. At least tw o of the dominant trees were cored at breast height to determine stand age. St and density (trees/ha), basal area (m2/ha) and standing volume (m3/ha) were calculated from thes e data. In addition, the volume (m3/ha) of all snags and downed woody debris (CWD) were also calculated. The e quation used for tree and snag estimates was: V = (0.000078539816*(DBH2))*tree height. The volumes of downed logs were estimat ed with Smalians metric equation: V = [((D2) + (d2))*0.00003927]* log length (m), where D = diameter large end (cm) and d = diameter small end (Wenger, 1984). We adapted the system of five decomposition st ates for snags and downed woody debris used by Spetich et. al. (1999). The decomposition descriptions translated to five levels of decomposition deductions by percent (15, 30, 45, 60, and 75%; see Table 2-1). Each 400 m2 plot contained four smaller 1 m2 subplots randomly placed within the larger plot for understory sampling (Figure 2-3). Perc ent cover of each species was assessed using a modified Daubenmire method incorp orating eight different levels (Daubenmire, 1959). Coleman rarefaction and the Shannon-Weiner diversity indi ces were calculated for each stand (Koellner and Hersperger, 2004; Colwell, 2006). PAGE 29 29 Data Analysis A three stage balanced nested design was used to integrate the indicators measured at different scales, and among sites (Figure 2-2). Hy pothesis testing for differences between means was accomplished by using two-sample t-test with an alpha of 0.05 and a two-tailed confidence interval. The sampling of nine distinct refe rence locations produced a dataset where the assumptions for analysis of variance (ANOVA) wa s not ensured; therefor e, non-parametric tests were used to detect any significant differences among the reference sites and among the distinct forest age class segments (SAS, 2002). Trends over time and between variables were obtained from linear regression using the general linear model (PROC GL M) (Yang et al. 2006; SAS, 2002). Plant species indicator analysis (IndVal) was used to measure the level of relationship between a given plant species to categorical units such as pine flat subtypes or forest age classes. It calculates the indicator value d of species as the product of the relative frequency and re lative average abundance in each categorical cluster. Indicator species analysis is used to attribute species to particular environmental conditions based on the abundance and occurrence of that species within the selected group. A species that is a perfect indicator is consistent to a pa rticular group without fail. Indicator values range from 0 to 100, with 100 being a perfect indicator score. Because indicator species analysis is a statistical inference, a test of significance is applied to determine if species are significant indicators of the groups with which they are associated (Dufrene and Legendre, 1997). This is achieved by the M onte Carlo permutation test procedure (1000 iterations), where the significance of a P-value is determined by the number of random runs greater than or equal to the inferred value ( =0.10). Accuracy is defined from the binomial 95% confidence interval (Strauss, 1982). PAGE 30 30 Results Overstory Stand Structure The m ean stand DBH, height, BA, and volume varied significantly among the age classes (Table 2-2). For example, the mean DBH fo r the young stands was 6.0 cm, 23.4cm for the midage stands, and 30.0cm for the mature age sta nds. Height, BA, and volume exhibited similar results. Snags and downed woody debris had similar values for the young and mid-age stands, but was significantly higher for the mature stan ds (Table 2-2). Regression analysis revealed trends over the chronosequence for stand structural variables. Except for stand density, all of the stand variables increased with forest age class. Stand density was highl y variable over time and did not exhibit any specific patterns (Figure 2-3) As expected, stand DBH increased with age until 85-90 years then began to reach a steady-state (Figure 2-4). Stand height also increased over time, but reached an asymptote at 85-90 yrs. Stand basal area and volume followed similar regression curves as with DBH and height (F igure 2-4). Downed wood y debris accumulation levels were highly variable, but in general increased over the 110-year chronosequence (Table 22; Figure 2-5). The volume of standing deadwood (snag) increased through the mid-age stands and then decreased thereafter (F igure 2-6). The level of CWD decomposition remained the same for the young and mid-age stands, but was lower for the mature age stands (Figure 2-7). Understory The abundance of grasses and forbs decreased, and the abundance of shrubs increased over the three forest age classes (p < 0.05; Figure 2-8). The Shannon-W iener diversity index ranged from 1.28 2.40 for the dataset. In general, Shan non-Wiener diversity index decreased over time (Figure 2-9). Shannon-Wiener dive rsity index also decreased w ith stand density during the young age class, but increased with stand density during the mature age cl ass (Figure 2-10). The Coleman rarefaction index ranged from 7.2-22.0 for the dataset (Table 2-9). The Coleman PAGE 31 31 rarefaction index increased with stand height during the young age class, but decreased with stand height during the midaged class (Figure 2-10). Bluestem grass, blueberry ( Vaccinium spp.), and witchgrass were the dominant plant species indicators for the young ag e stands (p < 0.022). Meadow b eauty, wiregrass, and Carolina redroot were the dominant plant species indicators for the mid-aged stands (p < 0.067). Gallberry, running oak ( Quercus pumila ), and dangleberry were the best plant species indicators for the mature age stands (p < 0.1; Table 2-3). Discussion Overstory The overstory variables of m ean stand DBH, stand height, stand BA, and volume exhibited strong positive relationships with stand age as expected. Even downed logs and snags, heterogeneous variables among the sites and with in age classes, produced a weakly positive trend with stand age (p < 0.042). Stand de nsity showed no clear pattern along the chronosequence, owing to the high variability f ound within density across the age gradient. The data showed most of the gr owth variables reaching an as ymptote around 85-90 years. When the chronosequence stand data were co mpared to growth and yield of natural longleaf pine stands, our stands were found to have lower basal area (14 m2 vs. 25 m2) at age 30, but comparable stand volumes (150 m3 vs. 130 m3) at age 60 (Farrar, 1985). The steady-state phase for these forests is reached around 85-90 year s, may be shorter than the steady state of 110 years for longleaf pine ecosystems in Texas, reported by Chapman (1909). All of the stands over 85 years measured at our reference sites had large gaps containing saplings and some poles. This structure would in dicate that the understory reinitiation stage of secondary succession was well along and the ec osystems self-organization capacity was apparent. Of the aboveground indicators, both stand density and CWD had the greatest PAGE 32 32 heterogeneous datasets based upon statistical analysis. This great variability in stand density and CWD reflects the differences that must be evaluated when comparing natural versus intensively managed forest stands. Stand growth is high during the early phase, but is slowed during the mid-aged class when the stem exclusion state is reached. Interspecifi c competition results in mortality of individuals and the snag accumulation rates increase. This was evident along the chronosequence when snag accumulation peaked between 60 and 80 years. During the mature phase, stand growth is slowed as the forest reaches a steady state. This indicat es that the mid-aged class was the major period for CWD accumulation brought on by both competi tion and disturbance (Spetich et al. 1999). Downed logs and snags continued to accumulate during the mature phase, but with a decreasing rate of decomposition. Plant Species Diversity The species diversity exhibited interesting patterns along the chronosequence. Stand density had a strong negative relationship with the Shannon-W iener diversity index within the young age class, but a positive relationship dur ing the mature age class (Figure 2-10). The Coleman rarefaction index increased with stand height within the young age class then decreased during the mid-aged class (Figure 2-10). The ch ange in response between stand height and Coleman rarefaction, and stand density w ith the Shannon-Wiener diversity index is due to how these two distinct indices calculate species diversity. The Coleman rarefaction function gives more weight to the rarity than commonness of species. The function looks at the diversity where the turnov er of species between two distinct species pools is measure d. There was an expected number of species E(s) where Monte Carlo iterations were performed to predict which species would be more likely appear (Hurlbert, 1971). In this case, species diversity became related to habitat diversity or habitat heterogeneity PAGE 33 33 (Hersperger and Koellner, 2004). As habitats beco me more complex (layer ed) with tree growth during early stand development, ra refaction index increases with species turnover rates, and as the number of rare species increase. This approa ch can better assess disturbance changes to the site than the (information theory measure) Sha nnon-Wiener diversity inde x (Gotelli and Colwell, 2001). The Shannon Wiener diversity index responde d positively to a stand with higher tree density during the mature age class. These mature higher density stands may have more habitat homogeneity than a stand with greater openness. Habitat homogeneity influences the ShannonWiener indexs evenness function J. Species evenness may be easier to attain where habitat homogeneity is greater. This is why the Shannon-Wiener indexs H value decreased as habitat heterogeneity increased during ea rly stand development, or may increase in older stands with higher stand density (homogeneity). Since the Shannon-Wiener indexs evenness function gives equal weight to rare and common species, it doe s not measure local patterns of assemblage where disturbance impacts could be assesse d (Pianka, 1966). The Shannon-Wiener diversity index still should be included w ith a rarefaction index during assessments because it is an abundance-based function where the total number of species (richness S) that are found within an area are measured. In addition, a measuremen t of the relative abund ance (N) and degree of equality among species (evenness J) ar e also calculated (Poole, 1974). The young, mid-aged, and mature age classes va ried in the abundance of grasses, forbs, and shrubs. Even with the goal of applying prescribed fire every thre e years at the reference sites, shrub species increased and graminoid species de clined over the age classes. The mature age class changed with running oak for mesic sites and gallberry for the we tter sites. The young age class had blueberry for mesic conditions and bluest em grass for the wetter sites. In this case, PAGE 34 34 more forest structure brought on by stand maturation could re present a drying effect on soil conditions as represented by a change in plant species. Conclusions Stand DBH, height, and basal area increased until 85-90 years wh en they began to reach a steady state. Coarse woody debris accum ulation levels were highly variable, but tended to increase with age. Standing deadwood also incr eased with age up to 60-80 years and began to decline thereafter. The decomposition levels of CWD were c onstant through the mid-aged class, but declined from the mid-age to the mature stage. The levels of shrub sp ecies were significantly higher in the mature sites than eith er the young or the mid-aged classes. Tree growth during early stand development tran slates to habitat heterogeneity as partial shading brings in new groups of plant species. At this point, stand height had a strong positive relationship with Coleman rarefaction index and stand density had a stro ng negative relationship with Shannon-Wiener diversity index. The plant species turnover rates as indicated by the Coleman rarefaction index were high and the ev enness of plant species as indicated by the Shannon-Wiener were very low. The evenness of plant species was not attained until the mature age class when competition was reduced, and the nu mber of plant species entering the ecosystem was equal to the number of plant species leav ing it. During this time period, Shannon-Wiener diversity index had a strong positive relationship w ith stand density and the Coleman rarefaction index had a negative relations hip with stand height. The results have shown interesting trends along the chronosequence for wet longleaf pine flat communities along the Gulf coast of Florida. The results indica te that Floridas Gulf Coastal longleaf pine flats reach the unders tory reinitiation condition at a pproximately 85-90 years. This would mean the forest is neari ng a steady, self-organizing state, perhaps a threshold point for attaining restoration success in terms of structural attributes. PAGE 35 35 Three areas of this research warrant further attention. Firs t, investigations concerning coarse woody debris in southern pine forests is lacking, probabl y due to a perception that any accumulation would be limited by prescribed fire. In this research, we found heavy accumulations at sites in each of the forest age classes. Secondly, experiments in plant community assemblage should be conducted to take a closer look at the relationships between the commonly used Shannon-Wiener diversity index and the Coleman rarefaction index. Coleman rarefaction was a stronger index duri ng early stand development, but showed no advantage over the Shannon-Wiener index during la ter stages of stand de velopment. Finally, our research found that shrub species dominated the mature aged stands even with aggressive fire management programs. Many of the plant species that were classified as woody do not have pioneer patterns similar to gallberry, saw-palme tto, or runner oak. They never dominated the site. There should be studies that focus on the less kn own woody species and their benefits to longleaf pine forest ecosystems. PAGE 36 36 Table 2-1. Qualitative classification sy stem for downed and standing deadwood. Characteristic Decay Class I II III IV V Leaves Present Absent Absent Absent Absent Twigs Present Present Absent Absent Absent Bark Present Present Often Pr esent Often Absent Absent Bole Shape Round Round Round Round to Oval Oval to Flat Wood Consistency Solid Solid Semi-Solid Partly Soft Soft Wood Strength Firm Firm Firm Breakable Fragmented Decomposition Deduction 15% 30% 45% 60% 75% Spetich et al. 1999; Adapted from Spies et al (1988). Table 2-2. Forest structural and plant species diversity means among age classes. Stand Age (years) Age Class Stand Diameter (cm) Stand Height (m) Stand Basal Area (m2/ha) Stand density (Trees/ha) Stand Volume (m3/ha.) CWD (m3/ha) Coleman Rarefaction ES(63) ShannonWierner Diversity 6Young0.30.20.03000.017.510.962.04 8Young0.72.80.12000.013.312.932.24 9Young8.34.90.3501.31.315.781.91 10Young8.56.00.3751.82.615.781.91 17Young16.18.95.625060.811.116.972.25 *Mean6.0a3.3a1.3a175a12.9a9.5a14.51a2.07a24Mid-Aged19.510.72.335026.02.516.692.2927Mid-Aged25.711.75.412562.00.716.692.29 29Mid-Aged13.68.27.340084.04.320.111.91 31Mid-Aged19.810.17.125078.01.517.902.32 34Mid-Aged23.311.510.1225128.076.013.281.72 36Mid-Aged17.211.111.1425125.025.320.111.91 40Mid-Aged29.919.41.87534.03.512.992.14 42Mid-Aged21.910.71.02510.110.012.992.14 46Mid-Aged25.818.020.6425395.880.79.991.28 50Mid-Aged35.020.94.950101.25.89.991.28 52Mid-Aged26.812.311.5275174.51.87.991.78 *Mean23.4b13.0b7.7b216a112.2b9.3a14.41a1.97a60Mature27.719.16.1250115.018.115.551.9461Mature18.412.214.1450192.313.717.902.3262Mature41.717.824.2225433.158.310.911.4168Mature24.912.57.3125122.3105.812.801.6271Mature35.115.510.2225156.161.57.991.78 86Mature26.016.610.7175206.39.612.801.62 95Mature37.518.211.1175202.512.717.891.56 101Mature30.213.213.7175234.111.47.201.16 105Mature33.417.511.0125194.31.017.891.56110Mature34.017.415.9250278.712.317.891.56*Mean30.0c15.5c11.8c217a201.7c30.0b14.25a1.86a Means followed by the same lower case letters are not signif icantly different (alpha=0.05). CWD represents snags and downed logs. The sample size for stand data by age class was n=12 and the understory vegetation data by age class was n=48. PAGE 37 37 Table 2-3. Indicator values for plant sp ecies in three forest age classes. Age Class Plant Species Age Class Young Mid-Aged Mature SD P-Value Veg Type Young Andropogan virginicus 34 3 1 3.09 0.001 Grass Dichanthelium ovale 33 10 2 3.49 0.002 Grass Vaccinium sp. 25 2 9 3.27 0.019 Shrub Pteridium aquilinum 12 1 0 2.28 0.021 Forb Mid-Aged Rhexia alifanus 0 28 0 3.03 0.001 Forb Cyperus sp. 0 25 0 2.20 0.001 Grass Lachnanthes caroliana 8 25 1 2.98 0.005 Forb Arisitida var. beyrichiana 1 23 0 2.54 0.001 Grass Solidago odora 0 9 1 2.09 0.066 Forb Mature Ilex glabra 20 11 37 3.04 0.007 Shrub Quercus pumila 17 3 29 3.31 0.014 Shrub Gaylussacia frondosa 2 3 13 2.81 0.099 Shrub Licania michauxii 0 0 10 2.17 0.040 Shrub Kalmia hirsuta 0 0 7 1.96 0.094 Shrub INDICATOR VALUES (% of perfect indication based on co mbining the values for relative abundance and relative frequency). The sample size for the understory vegetation data by age class was n=48. PAGE 38 38 Selected Reference Communities 1. Chassahowitzka Wildlife Management Area 2. St. Marks National Wildlife Refuge 3. Topsail Hill State Park Figure 2-1. Locations of the Pt. Washington Longleaf Pine Restora tion site ( ) and the reference sites within Gulf Coast Flat woods subecoregion of Florida (Griffith, 1994). 3 1 2 PAGE 39 39 Figure 2-2. Nested plot sa mpling design applied at three differe nt sites (age classes) for each reference location. 0 50 100 150 200 250 300 350 400 450 500 020406080100120 Stand Age (Years)Stand Density (trees / ha) Figure 2-3. Mean stand density (trees per hectare) along a 110-year longleaf pine chronosequence as measured from 26 differently aged stands. 1 ha Age Class Site 400 m2Forest p lot 1m2Ve g / soil q uadrat PAGE 40 40 Figure 2-4. Mean stand DBH, height, BA, and volume al ong a 110-year longleaf pine chronosequence as measured from 26 differently aged stands. y = -0.0043x2 + 0.7571x + 0.4671 R2 = 0.75 p < 0.0013 0 5 10 15 20 25 30 35 40 45 50 0255075100125Mean Stand DBH (cm) y = -0.0026x2 + 0.4109x + 0.9501 R2 = 0.72 p < 0.0004 0 2 4 6 8 10 12 14 16 18 20 22 0255075100125Mean Stand Height (m ) y = -0.0018x2 + 0.3297x 1.9355 R2 = 0.46 p < 0.0025 0 5 10 15 20 25 0255075100125Mean Stand Age (Years)Mean Stand Basal Area (m2 / ha) y = -0.0115x2 + 3.6533x 24.902 R2 = 0.76 p <0 .0004 0 50 100 150 200 250 300 350 0255075100125Mean Stand Age (Years)Mean Stand Volume(m3 / ha) PAGE 41 41 y = 0.3717x + 0.5181 R2 = 0.16 p < 0.0414 0 20 40 60 80 100 020406080100120Downed Woody Debris (m3 / ha) y = -0.0094x2 + 1.1249x 13.057 R2 = 0.25 p < 0.086 0 20 40 60 80 100 020406080100120 Stand Age (Years)Standing Deadwood (m3 / ha) Figure 2-5. Downed woody debris and standing deadwood (snag) accumulations along a 110year longleaf pine chronosequence as m easured from 26 differently aged stands. PAGE 42 42 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Forest Stage ClassDecomposition Level (%) Young Mid-Aged Mature Figure 2-6. Decomposition levels by forest age class as measured from 26 differently aged stands. Percent levels repr esent the amount of decay. 0 1 2 3 4 5 6 GrassesForbsShrubsVines Vegetative TypePercent Cover (%) Young Mid-Aged Mature b a a a a b a a b a ab a Figure 2-7. Composition of understory vegetation by forest age class. a a b PAGE 43 43 Figure 2-8. Shannon-Wiener Diversity and Co leman Rarefaction indices along a 110-year longleaf pine chronosequence as measur ed from 26 differently aged stands. y = 7E-05x3 0.01x2 + 0.3696x + 12.054 R2 = 0.23 p < 0.0373 0.00 5.00 10.00 15.00 20.00 020406080100120 Stand Age (Years)Coleman Rarefaction ES (63) y = 5E-07x2 0.0071x + 2.19 R2 = 0.40 p < 0.0006 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 020406080100120Shannon-Wiener Diversity H' PAGE 44 44 y = -1E-05x2 + 0.004x + 1.7514 R2 = 0.62 p < 0.0001 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 0100200300400500600Shannon-Wiener Diversity H' Young Age Class y = -0.0609x3 + 0.8965x2 2.5441x + 12.187 R2 = 0.70 p < 0.0001 5 8 11 14 17 20 03691 2Coleman Rarefaction ES (63) y = 0.149x2 5.0778x + 52.855 R2 = 0.61 p < 0.0001 5 8 11 14 17 20 23 691215182124Coleman Rarefaction ES (63) y = 0.4086x2 11.76x + 94.552 R2 = 0.37 p< 0.0024 5 8 11 14 17 20 23 691215182124Stand Height (Meters)Coleman Rarefaction ES (63) y = -6E-06x2 + 0.0023x + 1.7807 R2 = 0.08 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 0100200300400500Shannon-Wiener Diversity H' Mid-Aged Class y = 9E-06x2 0.0023x + 1.6464 R2 = 0.56 p < 0.0032 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 0100200300400500Stand Density (Trees / ha)Shannon-Wiener Diversity H' Mature Age Class Figure 2-9. Mean stand density versus the Shan non-Wiener Diversity index and mean stand height versus the Coleman Rarefaction index as measured from the young, mid-age and mature age longleaf pine stands. PAGE 45 45 CHAPTER 3 PATTERNS OF SOIL CHEMICAL AND MICROBIAL PROPERTIES ALONG A CHR ONOSEQUENCE IN WET LONGLEAF PINE FLATS OF FLORIDA Introduction Soil nutrient dynamics and their relationshi p to forest stand development have been under investigation for some time (Odum, 1969; V itousek and Reiners, 1975). Studies describing soil nutrient status, in particul ar nitrogen, and its influen ce on stand productivity and canopy nutrient dynamics have focused mostly on plantations (Morris and Boerner, 1998; Kirkman et al. 2001; Allen and Schlesinger, 2003), but research conducted in natu ral stands are also found in the literature (Zak et al. 1990; Vance and Entry, 2000; Arav ena et al. 2002; Chapman et al. 2003; White et al. 2004). Similar studies are rare in the longl eaf pine ecosystem, one of the most threatened ecosystems in the United States (W ilson et al. 2002). For example, patterns of nitrogen mineralization, the relationship between nitrogen levels and soil mi crobes, and how this relationship changes over time, have not been given much attention (Johnston and Crossley, 2003). Such information could aid the efforts of re storation professionals who are interested in not only restoring the structural attributes of the longleaf pine ecosy stem, but its functional attributes as well. One way to study the relationship between forest stand development and soil microbial dynamics is to use a chronosequence of simila r stands having differe nt ages since stand replacement (Pickett, 1989; Williamson et al. 2005) In an earlier investigation, Taylor et al. (1999) studied forest floor microbial biomass of northern hardwood forest stands ranging from 3 years after clearcut to 120 years. These authors reported an increasing tre nd in microbial biomass with age during the early successional stage. Ho wever, microbial biomass decreased with age during the mid-aged stage, but increased again during the late successional stage. Soil organic matter followed a pattern similar to microbial bioma ss. They further reported that fungal biomass PAGE 46 46 was positively correlated soil moisture and ne gatively correlated with soil pH. Finally, ammonium (NH4 +) production increased from the early to mid-aged stages and decreased from the mid-aged to the late successional stage (Tay lor et al. 1999). In another study, investigators wanted to detect the effects of plant diversity on the levels of fungal microbes by measuring the populations of fungal-feeding nematodes. As the plant community succeeded toward late successional conditions, there was little effect on the numbers of fungal-feeding nematodes (Kardol et al. 2005). The relationship between fungal growth and nematode populations was more complex than the investigators surmised. Recently, increased emphasis has been placed on examining soil microbial communities during soil assessments, especially when monito ring restoration projects (Harris, 2003; Johnston and Crossley, 2003). Some of the measured soil biotic variables have included microbial biomass carbon and nitrogen (Vance and Entry, 2000; Wilson et al. 2002), most probable numbers (MPN) of microbial functional groups (Schmidt and Belser, 1982), fungal biomass estimates (Montgomery et al. 2000), and complete co mmunity profiling (Bailey et al. 2002). Studies from other parts of the U.S. and the world have also contributed to our understanding of the soil community relationships For example, a growing number of studies have indicated that soil microbial communities with distinct f unctional groups inhabit different forest types (Pennanen et al. 1999). A black pine ( Pinus nigra) forest in Austria was found to have higher relative amounts of f ungi and actinomycetes in the soil microbial biomass than were found in a neighboring oak-beech (Quercus petrea Fagus sylvatica ) hardwood forest (Hackl et al. 2005). Researchers conducting a study in England found that soil moisture, pH, and microbial biomass levels decreased along a successional grad ient from moorland to grassland to mature pine forest (Chapman et al. 2003). Researchers in Finland found great variability within soil PAGE 47 47 microbial populations when measured beyond a 3-4 m radius (Pennanen et al. 1999). This result could imply soil microbial measurements farther than 4 meters apart may need to be considered as independent samples and not replications. A recent study in a longleaf pine ecosystem found that nitrogen mineraliza tion rates declined sign ificantly along a moisture gradient from xeric sandhills through the pine scrub to wet pine flat woods, but the levels of soil carbon, nitrogen, microbial biomass carbon, and aboveground net prim ary productivity (ANPP) had reverse trends (Wilson et al. 2002). These studies illustrate the strong interac tions that exis t between soil biogeochemical properties and vegetative ch anges in the aboveground cover type. Fires are important agents of natural disturba nce in many forest types, and especially in maintaining southern pine forests (Outcalt, 2000). Fires have variable e ffects on soil microbial biomass in forest soils where fire intensity, se ason, and weather play a role (Wilson et al. 2002). Hurricanes, another prominent di sturbance along the Gulf coast, can have major impacts on forest structure by strong winds (Palik et al. 2002), but the associated flooding may have bigger impacts on soil productivity, biogeochemistry and soil microbial popu lations (Lockaby and Walbridge, 1998). Anthropogenic disturbance effects from human activities in fore sts can also have effects on the functioning of forest soils. In a military installation study focused on the effects of different levels of vehicle traffic on the soil microbi al community within a longleaf pine forest, investigators found that increasing levels of tr affic produced a decrease in the fungal biomass (Peacock et al. 2001). Fertilization can have negative feedbacks. In a grassland restoration study, fertilizer additions caused increases in the num ber of bacteria, but a decrease in the fungal population. Sites with no fertil izer had larger fungal biomass levels, a greater number of legumes, and higher plant species richness than th e fertilized areas (Smith et al. 2003). Changes PAGE 48 48 in soil microbial functional groups caused by natural or human-induced disturbances can have negative impacts on long term soil nutrient cycles. Ecosystem health has been described in terms of nutrient retention or the ability of an ecosystem to prevent nutrient loss (Odum, 1969). Ecosystems have been identified as leaky where nitrate (NO3 -) was found in higher concentrations, or as nitrogen-limited or having a tight nutrient cycle where th e less mobile ammonium (NH4 +) ion was in higher concentrations (Davidson, 2000). The steady-state development st age of succession (Oliver, 1981) has been described as the time period of forest succession when the ecosystems nutrients are held tightly within (Odum, 1969). Vitousek and Reiners (1975) concluded the tightest period of nutrient retention was during the mid-aged period when nutrients are brought into short supply by heavy competition. They further concluded that nutrient retention in an ecosystem actually reflects biomass accumulation patterns. They suggested th at differences between net nitrogen input and output were proportional to the ra te nitrogen was incorporated into net biomass increment. Biogeochemical equilibriu m would be signified when the di fferences between nitrogen inputs and outputs would be equal to ze ro, or during the period of late succession when net biomass accumulation is close to zero (Zak et al. 1990). Given the relationship between biomass accumulation and nutrient retention, the biogeoche mical thresholds should be found when the ecosystem is self-organizing or during the understory reinitiation stage of succession (Oliver, 1981). Recent investigations have f ound an internal mechanism by which excessive nitrate is conserved in wet forested ecosystems. In upla nd forested environments, examined within both the temperate and tropical zones, investigator s have discovered that dissimilatory nitrate reduction to ammonium (DNRA) was a major pa thway for transforming nitrate to ammonium, PAGE 49 49 and preventing losses from leaching or by the de nitrification pathway (Silver et al. 2001; Huygens et al. 2007). Through 15N tracing, investigators discovered the majority of any surplus nitrate was reduced by DNRA, rather than reduce d by denitrification or immobilized by soils. The common conditions found at the research si tes were wet soils, hi gh organic carbon, and normally nitrogen-limited environments. We conducted our studies within longleaf pi ne ecosystems located along the Gulf Coast Flatwoods zone. This coastal region is found between the panha ndle community of Pensacola and Tampa Bay, Florida. Various ecological studies have investigated the changes in plant community composition along soil moisture gradient s within the Gulf Coast Flatwoods zone, but none have examined the soil chemical and microbial properties al ong a chronosequence. Previous research has concluded that plant species richness increases along a soil moisture gradient until an ecotone betw een mesic flatwoods and cypress swamps is reached (Huck, 1986; Walker, 1993; Kirkman et al. 2001). This ecotone is the interface where one would find the wet flatwoods and wet pine savanna su btypes of the coastal longleaf-slash pine flat (Messina and Conner, 1998). There are almost 200 rare vascular plant taxa found in the great variety of habitats classified as longleaf pine ecosystems. In addition to the majority of them being found in Florida (Collins et al. 2001), the ri chest sites are found in these wet pine flats and their associated wetlands (Walker, 1993). Wet pine flats repres ent more than 1 million ha in the Southeast (Burger and Xu, 2001). Plant species richness of wet longleaf pine communities has been positively correlated with soil productivity (Kirkman et al. 2001), and specific soil properties (Wilson et al. 2002). Soil characteristics need to be included with plant species richness in any restoration assessment of coasta l wet longleaf pine flat ecosyst ems to couple functional with structural attributes (Joh nston and Crossley, 2002). PAGE 50 50 Our main objective was to measure soil pH, moisture content, organic matter content (SOM), plant-available phosphorus, soil nitrogen mineralization rates (Nmin), soil microbial biomass carbon (Cmb) and fungal biomass (Cfb) along a 110-year chrono sequence to determine the ecological trajectory in terms of soil chemical and microbial characteristics for longleaf pine in coastal wet pine flat communities. We specifically tested our hypothesis that this group of soil biogeochemical indicators measured along a 110year chronosequence would follow a pattern similar to the biomass accumulation curve of fore st succession (Vitousek and Reiners, 1975). In response to rapid increase in grow th during the early years of sta nd establishment, we predicted a similar increase in net nitrogen mineralizati on rates, microbial biomass and fungal biomass levels. We hypothesized that thes e variables would decrease at some point during the late midaged phase and reach a threshold some time during the mature phase. Materials and Methods Study Areas Three representative locations along Floridas Gulf Coast Flatwoods zone (720 km ) were selected for this study. The three locations we re Topsail Hill State Park, St. Marks National Wildlife Refuge, and the Chassahowitzka Wildlif e Management Area of the Florida Fish and Wildlife Conservation Commission. At each location, four 400 m2 plots representing each of early, mid, and mature age classes of longleaf pine stands were laid out for vegetation (reported in Chapter 1) and soil sampling. The different successional ages (age classes) represented a chronosequence of 110-years. Soil Sampling and Preparation Soils sam ples (> 500 g) were taken from four (1m2) quadrats taken in each of the 400 m2 plots during September of 2005 and 2006 for general analysis. The samples were taken from the upper 10 cm of the A horizon, not including the organic layers. An additional sampling was PAGE 51 51 conducted in August, 2005, at which time a paired-s oil sample was buried in place for incubation (Eno, 1960; described later). The incubated samp les were taken out during the September, 2005 sampling period. All samples were immediately stor ed at 4C until analysis. A sub-sample (20g) from all of the samples was used for determinin g moisture content after oven drying at 105C for 72 hours. Soil Chemical Analysis The soil s amples were analyzed at the Univer sity of Florida soil testing laboratory (UF Analytical Research Laboratories), Gainesville, Florida. Soil water pH was determined from prepared slurries using a soil-to-water rati o of 1-to-2 (EPA method 150.1). Plant-available phosphorus was determined with the use of Mehlich-1 extractant (H2SO4 & HCL) and measured on an inductively coupled plasma (ICP) spect rophotometer (EPA method 200.7; Nelson et al. 1953). Soil organic matter content (SOM %) wa s determined by the Walkley-Black method (Walkley, 1947). Net Nitrogen Mineralization Net nitrogen m ineralization was determined by the buried bag technique (Eno, 1960). Forty eight samples were collected from each reference location and the re storation site for a total of 144. In general, one bag was buried in situ for incubation during August 2005 (Eno, 1960) and the other bag taken to the soil lab for an alysis. The incubated bags were collected and analyzed after 30 days. Mineral nitrogen was extracted from 20 g of both soil samples with 60 ml 2N KCL and placed in a shaker for one hour. They were then filtered through # 42 Whatcom filter papers into 20 ml scillination bottles. The samples were analyzed by the University of Floridas Analytical Research Laboratory fo r ammonium (EPA method 350.1) and nitrate (EPA method 353.2) with a continuous auto-flow analy zer. Net mineralization was calculated as the PAGE 52 52 difference between incubated-N and initial-N (c orrected for soil moisture) (Keeney & Nelson, 1982). Microbial Biomass Soil m icrobial biomass C was determined by ch loroform fumigation-extraction (Vance et. al., 1987), with the following modifications. Twelve grams of soil were sieved from soil samples stored at 4 C and then placed in 50 ml centrifuge test tubes. Matching 12 g soil samples were set aside in additional 50 ml centr ifuge tubes as the control. The soil samples were fumigated in a desiccator with 40 ml of alc ohol-free chloroform placed into a center beaker, an additional 0.5 ml of chloroform was placed into each centrifuge tube. The top of the desiccator was pressure sealed and vacated until the chloroform began to boil. The tubes were then incubated for 24 hours at 25C. The dessicator was then opened, re sealed, and after the chloroform was reboiled, incubated for an additional 24 hours. The control and fumigated samples were extracted with 36 ml of 0.05 M K2SO4, shaken (360 rpm) on an orbital shaker for 1 hour, and centrifuged @ 6000 rpm for 15 minutes. The supernatant was then filtered through # 42 What com filter papers into 20 ml scintillation vials and frozen until analysis. Levels of total organic carbon (TOC) were determined on a Shimadzu TOC-VCSH analyzer (Vance & Entry, 2000). Microbial biomass carbon was calculated as: [(fumTC ConTC) / 0.51] / (Soil Wt.) = mg C kg-1 dry wt. soil (Joergensen, 1996). The value of 0.51 is the conversion factor equal to the extractable porti on of microbial biomass in a forest soil. Fumigated and non-fumigated blanks were measured to correct for the chloroform and potassium persulfate. Soil fungal biomass levels were determined by a physical disruption method for extraction of ergosterol from soil samples (Gong et al. 2001) with the following modifications. Six grams of soil were mixed with 9 ml of 0C methanol and 1.9 g of glass beads into 20 ml scintillation PAGE 53 53 vials. The vials were vortexed for 30 seconds, sh aken (360 rpm) on an orbital shaker for 1 hour, and refrigerated over night. An a liquot of 1.8 ml was plac ed into 2 ml micro-centrifuge tubes and centrifuged @ 11,000 rpm for 20 minutes. After extr action, the samples needed to be filtered before running through a High Performance Li quid Chromatography (HPLC) computerized machine. A syringe was used to remove 1.5 ml of the supernatant from the micro-centrifuge and filtered through a 0.20 m filter into amber colored 2 ml glass HPLC vials. The HLPC vials were covered with aluminum foil and stored in the dark at 0C until ready to inject into the HPLC. Each sample was quantified on a Beckman C oulter HPLC equipped with an UV detector, a pump, an auto-sampler, and through a C-18 re verse-phased analytic column (4.6 x 250 mm). The UV detector was set at 282 nm and pure metha nol was used as the mobile phase at a flow rate of 1 ml per minute. Extracts (100 l) were injected while the column pressure was maintained at 1000 psi. Pure ergos terol (Sigma) was recrystalized in pure methanol at different concentrations to establish a set of standards. The standard curve was constructed from a linear regression relationship between peak area and ergosterol concentration Ergosterol recoveries were calculated from the difference between spiked and non-spiked paired samples divided by the amount of ergosterol added Under such conditions, an isolated peak was identified from field samples at approximately 13 minutes, based upon the peaks obtained from the ergosterol standards. Establishe d from results of previo us investigations, an average conversion factor for 3.65 g ergosterol pe r mg of soil is converted to fungal biomass (mg-1 /g -1 soil) when multiplied by 220 (Montgomery et al. 2000). Fungal-to-microbial biomass ratios were determined using a ratio of the calculated soil fungal biomass carbon and the soil microbial biomass carbon for each sample. PAGE 54 54 Coastal wet longleaf pine flats experience l ong periods of standing water (Harms et al. 1998). This flooding causes changes in the biogeochemical cycling of nutrients. These forested wetlands also contain highly acidic soils that require modifications to standard soil biochemical analysis techniques normally used in moderate pH (6.0) wetlands. The following modifications were necessary in order to produce good laboratory results. The microbial biomass carbon was extracted from soils using a lower 0.05 M K2SO4 extractant instead of the standard molarity (0.5 M) for improved efficiency in these low pH soils (Haney et al. 2001). The samples were centrifuged before filtering to re duce the high amount of woody material found present in the soil samples. A relatively new ergosterol extraction method by physical disruption was utilized to simplify the process for an alyzing fungi in a large number of soil samples (Gong et al. 2001). A lower conver sion factor for fungal biomass was used to account for the flooded conditions on soil fungal growth (Montgomery et al. 2000). Data Analysis A three stage balanced nested design was used to integrate the indicators measured at different scales and among sites. Hypothesis testing for differences between means was accomplished by using two-sample t-test with an alpha of .05 and a two-tailed confidence interval. Since the monitoring of the restoration site with nine distinct reference locations produced a dataset where the assumptions for an alysis of variance (ANOVA) were not ensured, non-parametric tests were used to detect any significant differences among the reference sites and among the distinct age class segments (SAS, 2002). Correlations between soil moisture, soil chemical and microbial abundances were determined using Spearman's rank ( r ) correlations (Dumortier et al. 2002; SAS, 2002; Spyreas and Mathews, 2006). Trends between variables we re obtained from linear regression using the general linear model (PROC GLM) (Yang et al. 2006; SAS, 2002). The chronosequencial trends PAGE 55 55 were enhanced by incorporating moving average smoothing (MA model) as a data filter to reduce cyclical and seasonal variations found in the datasets fo r a number of the indicators affected by climate ( Platt and Denman 1975; Kumar et al. 2001; Ittig, 2004). The trend analysis was followed by log10 data transformations where necessary to stabilize variances prior to analysis. Partial Canonical Correspondence An alysis with multivariate regression (proc CANCORR) was used to determine the relative c ontributions of the different variables to the relationship (SAS, 2002; Fortin and Dale, 2005). Results Soil Types, Soil Organic Matter, and Soil pH All three ref erence sites contained taxonomically equivalent soil types. All of the soils had similar soil properties (sandy, acidic, thermi c, aquic). The soils were also found to be functionally equivalent even when compared by dr ainage class (Table 3-1). Soil organic matter content (SOM) was found to increase from 1% to 4.5% as gravimetric soil moisture increased from 20% to 60% of soil weight (Figure 3-1). Soil pH decreased from a pH of 5.0 to 4.0 as SOM increased from 1% to 4.5% (Figure 3-2). The plant-availabl e phosphorus tests produced too many non-detectable samples for any m eaningful results (Table 3-2). Net Nitrogen Mineralization Net nitrogen mineralization rates (Nmin) increased during the y oung age class, peaked during the mid-age class, and then decreas ed after 60 years (Figure 3-3). Mean Nmin rates were 12 mg N / kg soil / month for the young age stan ds, 14 mg N / kg soil / month during the midaged class, and 8 mg N / kg soil / month during the mature age class for the reference sites (table 3-2). The pattern for Nmin rates followed microbial biomass levels (Cmb) over the 110-year chronosequence (Figure 3-4). Nmin rates increased from 5 mg N / kg soil / month to 20 mg N / kg soil / month as Cmb increased from 100 mg-1 C / kg soil to 1000 mg-1 C / kg soil (Figure 3-5). PAGE 56 56 Nitrate production was 13.8 mg-1 NO3 / kg soil during April 2002 and 135.7 mg-1 NO3 / kg soil during August 2002. In comparison, Nitrate production was 4.7 mg-1 NO3 / kg soil during April 2005 and 1.9 mg-1 NO3 / kg soil during August 2005 (Table 3-3; Figure 5-3, Chapter 5). Ammonium production was 13.4 mg-1 NH4 / kg soil during April 2002 and 104.7 mg-1 NH4 / kg soil during August 2002. During 2005, ammonium production was 8.9 mg-1 NH4 / kg soil during April and 9.6 mg-1 NH4 / kg soil during August (Table 3-3). Nmin was positively related to ammonification (NH4 +) (r > 0.810; p < 0.0001) during all three age classes, but not co rrelated with nitrification. Nmin became positively correlated with soil moisture and SOM ( r > 0.460 (p < 0.01) during the mid-aged class and remained so through the mature age class (Table 3-4). Ammonium production was negatively related to nitrate production (NO3 -) during the mid-age and mature ( r = 0.470; p < 0.001) age classes (Table 3-3). Microbial Properties Mean soil microbial biomass carbon (Cmb) levels were 275 (mg C / kg soil) for the young age stands, 416 (mg C / kg soil) during the mid-ag ed class, and 339 (mg C / kg soil) during the mature age class for the reference sites (T able 3-2). Mean soil fungal biomass carbon (Cfb) levels were 102 (mg C / kg soil) for the young age class, 163 (mg C / kg soil) for the mid-aged stands, and 125 (mg C / kg soil) during the mature age class at the reference sites (Table 3-2). Fungal biomass carbon increased during the first 60 years (~200 mg C / kg soil), then decreased down to 110 years (~100 mg C / kg soil; Figure 3-6). The fungal-to-microbial biomass ratio (FB-to-MB) decreased from a mean value of 0.4 to 0.2 during th e first fifteen years after establishment, and then increased to 0.8 at 50 years (Figure 3-7). Microbial biomass (Cmb) had a negative relationship (r > 0.400 (p < 0.01) with soil pH during the mid-aged and mature age classes (Table 3-3). PAGE 57 57 Discussion The nitrogen m ineralization (Nmin) process in high soil moisture conditions was dominated by ammonium production (NH4 +), with low concentrations of nitrate being measured. The net nitrification rates represente d 50% of the production during 20 02 and less than 25% during 2005. The net nitrogen mineralization rates were 10 magnitudes greater during 2002 compared to 2005 (Table 3-3). Similar results between NO3 and NH4 + were measured in a study comparing xeric, mesic, and wet longleaf pine sites in s outhern Georgia (Wilson et al. 2002). When Nmin became positively correlated with soil moisture and SOM during the mid-age and mature age classes, nitrate levels (NO3 -) became negatively correlated to ammonium (NH4 +) production. The dynamics indi cates a portion of the NO3 was converting to NH4 + during saturated conditions. This condition might be i ndicating the dissimilatory nitrate-reduction-toammonium (DNRA) process is taking place during flooded conditions. Little dinitrogen (N2) gas is lost to the atmosphere or NO3 by leaching when the DNRA pathway is dominant. Flooding causes a lower redox potential (Eh < 0.6), and with a sufficient supply of NO3 and labile carbon, DNRA became the preferred pa thway over denitrification, resulting in the enriched pool of NH4 + (Stevens et al. 1998). Investigators examined the changes in nitrogen and phosphorus availability in longleaf pine sites from wetlands through an ecotone to upland sites, and they measured higher levels of nitrate a nd phosphorus taken from soils in the middle of wetland sites than found in the ecotone or uplan d sites. However, the upland sites had higher amounts of labile nitrate than the we tter sites (Craft and Chiang, 2002). The anaerobic conditions and a high supply of nonlabile nitrate in we t longleaf pine sites are conducive to DNRA. During anaerobic conditions, the DNRA pathway provides NH4 + to plants and microbes, requiring le ss energy to assimilate than NO3 assimilation (Silver et al. 2001). The characteristics favoring DNRA over denitr ification are high rainfa ll, a high C:N ratio, PAGE 58 58 and a forested ecosystem that is naturally N-limited during dry conditions. The DNRA pathway has been determined to be less sensitive to higher soil temperatures and a lower pH than denitrification. The DNRA pathway is now cons idered a major nitrogen conservation mechanism in humid forest ecosystems (Silver et al. 2001; Huygens et al. 2007). One of the reasons why the nitrogen mineraliz ation patterns closely followed microbial biomass changes over the chronosequence (Figure 3-4) was because the nitrogen mineralization data did not contain significan tly high levels of nitrate during the 2005 growing season. Heterotrophic bacteria and fungi that dominate the soil microbial biomass produce ammonium from organic nitrogen. Only a fe w chemoautotrophic bacteria produc e the majority of nitrite and nitrate (Richards, 1987). Soil nitrogen mineralization ra tes increased during stand esta blishment, but eventually decreased after canopy closure as longleaf pine st ands entered the stem exclusion phase of stand development (Oliver, 1981). Other studies in hard wood forests have found increases in nitrogen mineralization rates after stand re placing harvests up to 60 years, then declining to a constant range (Zak et al. 1990). Investigators evaluating th e affects of ponderosa pine restoration treatments on mycorrhizal fungi, determined treatments prom oting graminoid and herbaceous ground cover had a positive relationship to levels of arbuscu lar mycorrhizal (AM) f ungi (Korb et al. 2003). These authors also discovered a positive re lationship between stand BA and levels of ectomycorrhizal (EM) fungi (Korb et al. 2003). In an earlier experi ment in a slash pine forest, Sylvia and Jarstfer (1997) reported a strong competition that ex ists between AM weeds and EM pine roots (Sylvia and Jarstfer, 1997). The implicat ion would be that AM fungal levels were high at harvest, declined during th e first 15 years as growing EM trees crowded out the AM PAGE 59 59 groundcover. Eventually the AM fungi were replaced with EM fungi, and the overall fungal biomass levels increased after 15 years. This pattern is simila r to our results (Figure 3-7). Phosphorus availability was very limited in these sites as indicated by the poor results. Similar results have been reported in loblolly pine plantations throughout the South (Martin and Jokela, 2004). Fertilization can dramatically improve biomass accumulation, but unless it is maintained, nutrient-deficient soils can result fr om the fast pine growth (Adegbidi et al. 2005). Phosphorus levels were found to be higher and P-mi neralization rates lower in wet southern pine forests (Grierson et al. 1999; Craft and Chia ng, 2002). In our study, soil or ganic matter content (SOM) was found to increase with soil moisture, and increased levels of SOM caused decreases in soil pH. As soil organic matter increases, it forms complexes with Mg2+ and Ca2+cations in solution, releasing H+ ions into soil solution from organic acids (Brady and Weil, 2002). This relationship was confirmed by a negative relationship between SOM and soil pH. A lower soil pH usually leads to lower nitrogen mineralizat ion rates (Morris and Boerner, 1998). Active bacterial respiration and microbial biomass levels substantially decline below a soil pH threshold of 5.0, resulting in lower rate s of nitrogen mineralization (Baath and Anderson, 2003). Lower mineralization rates results in higher organic matter accumulation. Conclusions Nitrogen cycling was dom inated by ammoni um production during the wet 2005 growing season when compared to a drier 2002. There was ammonium enrichment at the cost of nitrate levels. This probably indicates that the dissimilatory-nitrate reduction-to-ammonium (DNRA) pathway was prominent during the flooded 2004-2005 growing seasons. The net nitrogen mineralization rates, microbial biomass carbon, and fungal biomass carbon increased between the young and mid-aged classes, th en decreased between the mid-ag ed and mature age classes. The FB-to-MB ratios increas ed dramatically up to 60 years, th en decreased to 110 years. Finally, PAGE 60 60 soil organic matter content (SOM), increased wi th soil moisture. Based upon the results, this group of soil biogeochemical indicators follows biomass accumulation patterns and will attain biogeochemical equilibrium after a stand age of approximately 60-70 years. The threshold would be during the mature age class after the un derstory reinitiation phas e of forest succession has started. Soil biogeochemical studies require a great amount of resources and equipment to conduct an ecosystem-level analysis. The research could have been improved if a series of soil samples were analyzed over a two-year period, at 3-month intervals instead of annual sampling. However, the cost of running net nitrogen mi neralization, microbial biomass and ergosterol determinations would be quite high. Our resear ch has shown some interesting results, but additional research is required to explore the biogeochemistry of wet longleaf pine flats. This would include exploring the soil organic matte r accumulation vs. flooding cycles in facultative wetland pine sites, the relations hip between tree root mass a nd fungal biomass during longleaf succession, and the effects of competition betw een mycorrhizal and saprophytic fungi during longleaf pine development. PAGE 61 61 Table 3-1. Soil and stand properties between reference sites. LOCATION SOIL GREAT GROUPS SOIL TEXTURE (Top 10 cm) MOISTURE REGIME TEMPERATURE REGIME DRAINAGE CLASS Chassahowitzka Wildlife Management Area Psammaquent Sandy Aquic Hyperthermic Very poorly drained Alaquod Sandy Aquic Hyperthermic Poorly drained St. Marks National Wildlife Refuge Psammaquent Sandy Aquic Thermic Very poorly drained Alaquod Sandy Aquic Thermic Poorly drained Topsail Hill State Preserve Humaquept Sandy Aquic Thermic Very poorly drained Alaquod Sandy Aquic Thermic Poorly drained STAND BASAL AREA AND SOIL BIOCHEMICAL PROPERTIES (Mean Values*) DRAINAGE CLASS STAND BASAL AREA (m2 / ha) pH [ H+] NET NITROGEN MINERALIZATION RATES (mg N / kg soil / month) MICROBIAL BIOMASS CARBON (mg C / kg soil) FUNGAL BIOMASS CARBON (mg C / kg soil) Very poorly drained 6.5a 4.4a 11.6a 374.3a 133.8a Poorly drained 8.3a 4.5a 9.9a 356.1a 135.3a Means followed by the same lower case letter s are not significantly different. (alpha=0.05) PAGE 62 62 Table 3-2. Soil chemical and microbial biomass means between age classes. Stand Age (years) Age Class Net Nmin (mg-1/kg-1 Soil / year)Cmb (mg-1 / kg-1 / soil) SOM Content (%) Soil pH [ H+] Plant Avail P (mg/kg soil)Cfb mg/kg soil FB to MB Ratio 6Young2.492.22.754.100.4029.07.08 8Young8.333.71.564.70ND31.86.76 9Young23.6253.01.894.500.0457.312.73 10Young22.3448.61.964.60ND72.815.82 17Young5.3674.51.694.30ND102.223.77 *Mean12.0a275.4a1.97a4.44a*101.5a14.48a24Mid-Aged21.81169.50.834.50ND39.80.0327Mid-Aged9.6703.61.294.70ND24.10.03 29Mid-Aged40.8402.13.554.400.0234.00.08 31Mid-Aged9.2762.62.094.30ND83.20.11 34Mid-Aged2.2285.13.024.60ND191.10.67 36Mid-Aged17.0403.82.494.30ND83.50.21 40Mid-Aged-0.730.31.034.60ND98.51.00 42Mid-Aged3.598.61.034.80ND41.90.43 46Mid-Aged16.4321.01.164.001.28117.10.36 50Mid-Aged11.4131.81.295.30ND95.70.73 52Mid-Aged19.1275.83.954.100.14154.40.56 *Mean14.0a416.3b1.98a4.51a*162.9b14.45a60Mature0.833.41.364.900.838.41.0061Mature10.7519.60.704.50ND92.90.1862Mature32.6753.63.624.10ND59.90.0868Mature10.0511.53.554.00ND189.30.3771Mature-0.4132.54.544.00ND36.60.28 86Mature6.4241.80.904.40ND53.90.22 95Mature7.71.01.634.300.040.31.00 101Mature5.4570.21.494.30ND88.00.15 105Mature6.083.01.494.50ND28.50.34110Mature3.782.71.634.700.585.91.00*Mean8.4b338.7ab2.02a4.44a*125.3ab14.25a Means followed by the same lower case letters are not significantly different (alpha=0.05). The sample size for the soil data by age class was n=48. Table 3-3. Soil nitrogen mineralization means (Nmin) for dry season 2002 and wet season 2005. Date Nmin (mg N / kg soil) Nmin (mg NO3 / kg soil) Nmin (mg NH4 / kg soil) Apr-02 27.2b 13.8b 13.4b Aug-02 240.4a 135.7a 104.7a Sep-02 189.1a 128.4a 60.7a Apr-05 7.6b 4.7b 8.9b Aug-05 11.5b 1.9b 9.6b *Means followed by the same lower case letters are not significantly different (alpha=0.05). The sample size for the 2002 soil data was n=180 and for the 2005 soil data was n=144. PAGE 63 63 Table 3-4. Differences in soil biogeochemical relationships based upon Spearman rank correlations r as stratified by forest age class (n = 48). Net NminNH4 + MinNO3 Min MoistureSoil pHSOM CmbNet Nmin 0.885**** 0. 333* NH4 + Min 0.342* Net NminNH4 + MinNO3 Min MoistureSoil pHSOM CmbNet Nmin 0.805**** 0. 367* 0.468** NH4 + Min -0.310*0.513*** NO3 Min Soil pH -0.411** Net NminNH4 + MinNO3 Min MoistureSoil pHSOM CmbNet Nmin 0.860**** 0. 471***-0.413** 0.470** NH4 + Min -0.528***0.449**-0.329* Soil pH-0.412** Prob > |r| under H0: Rho=0 Young Age Class (6-20) Mid-Aged Class (25-55) Mature Age Class(60-110) Significance of the Spearman rank correlation test: blank: non-significant, *0.05 < p 0.01, **0.01 < p 0.001, ***0.001 < p 0.0001, **** p < 0.0001. y = 6.5506x 0.2389 R2 = 0.49 p < 0.0001 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.00.10.20.30.40.50.60.70.8 Soil Moisture Content (g H2O / g Soil dwt) Soil Organic Matter content (%) Figure 3-1. Soil organic matter content versus soil moisture as measured from 26 differently aged stands. PAGE 64 64 y = -0.1615x + 4.7683 R2 = 0.31 P< 0.0029 3.8 4.0 4.2 4.4 4.6 4.8 5.0 0.001.002.003.004.005.00 Soil Organic matter content (%)Soil pH [H+] Figure 3-2. Soil pH versus soil organic matter cont ent (percent) as measured from 26 differently aged stands. y = -0.0016x2 + 0.1273x + 5.7195 R2 = 0.46 p < 0.0258 0 5 10 15 20 020406080100120 Stand Age (Years)Net Nitrogen Mineralization Rates (mg N / kg soil / month) Figure 3-3. Total net nitrogen mineralization, ammonification and nitr ification rates (mg -1 nitrogen / kg -1 soil / month -1 ) along a 110-year chronosequence as measured from 26 differently aged stands. The data was filtered with moving average smoothing to remove seasonal and cyclic effects. PAGE 65 65 1 10 100 1000 020406080100120 Stand Age (Years)Cmb and Nmin Patterns (mg-1 /kg-1 soil) Microbial Biomass Carbon Net Nitrogen Mineralization Figure 3-4. Trends for microbial biomass carbon (Cmb) and net nitrogen mineralization rates (Nmin) along a 110-year longleaf pine chronosequence as measured from 26 differently aged stands. The data was filtered with moving average smoothing to remove seasonal and cyclic effects. y = 0.0157x + 6.2013 R2 = 0.31 p < 0.004 0 5 10 15 20 25 30 02004006008001,000 Microbial Biomass Carbon (mg C / kg soil)Net N mineralization (mg N / kg soil / month) Figure 3-5. Microbial biomass carbo n versus net nitrogen minerali zation rates as measured from 26 differently aged stands. The data was filtered with moving average smoothing to remove seasonal and cyclic effects. PAGE 66 66 y = -0.0328x2 + 3.8711x + 57.154 R2 = 0.31 p < 0.0054 0 30 60 90 120 150 180 210 240 270 300 020406080100120 Stand Age (Years)Fungal Biomass (mg-1 C / kg-1 soil) Figure 3-6. Fungal biomass car bon ( C ) along a 110-year longl eaf pine chronosequence as measured from 26 differently aged stands. The data was filtered with moving average smoothing to remove seasonal and cyclic effects. Figure 3-7. The fungal-to-microbial biomass ra tio and fungal biomass carbon levels (means) during the earlier and later portions of chronosequence respectively, as measured from 26 differently aged stands along th e 110-year longleaf pine chronosequence. y = 0.0252x2 5.7258x + 424.05 R2 = 0.40 p < 0.03210 50 100 150 200 250 300 350 485460667278849096102108114Stand Age (years)Fungal Biomass Carbon (mg/kg soil ) y = 0.0011x2 0.0483x + 0.7088 R2 = 0.80 p < 0.00030 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0612182430364248Stand Age (years)Fungal to Microbial Biomass rati o PAGE 67 67 CHAPTER 4 RELATIONSHIP BETWEEN VEGETATION AND SOIL CHARACTERISTICS IN WET LONGLEAF PINE FLATS ALON G FLORIDAS GULF COAST Introduction There have been m any efforts on assessing th e inter-relationships between community structure, plant species composition and soil bioc hemical attributes of forested ecosystems around the globe (Goebel et al. 2001; Peacock et al. 2001; Wilson et al. 2002; Allen and Schlesinger, 2003). For example, researchers in northwest Spain found that specific herbaceous species assemblages were indicato rs for soil pH, soil organic matte r levels, C: N ratios, and high or low levels of soil nitrogen, phosphorus, potas sium, calcium, and magnesium (Zas and Alonso, 2002). South American researchers co mpared mature natural alerce ( Fitzroya cupressoides ) forests with mature mixed beech ( Nothofagus-Podocarpus ) forests in the Chilean Andes and found that the mixed beech-conifer forests that contained greater tree and plant species biodiversity, had significantly hi gher soil nitrogen minera lization rates (Perez et al. 1998). These and other studies have greatly contributed to ou r understanding of the so il-vegetation community relationships for various ecosystems in the U.S. and other parts of the world (Vance and Entry, 2000; Reynolds et al. 2000; Chapma n et al. 2003; Korb et al. 2003; Hackl et al. 2005). However, the reasons why soil nutrient and microbial dynamics influence community structure and composition of the longleaf pine eco system of the southeastern U. S. still remains unexplored. Longleaf pine ecosystem is one of the most threatened ecosystems in the U.S. Knowledge about the interrelationships among soil chemical, micr obial and vegetational ch aracteristics of the longleaf pine ecosystem may aid in restoring it to a healthy, functiona l ecosystem across its range. The concept of using soil chemical and microbial properties in combination with vegetation attributes for monitori ng restoration projects has gained momentum in the recent past. PAGE 68 68 For example, researchers monitoring the restoration of ponderosa pine ( Pinus ponderosa ) forests in Arizona explored the relati onship between mycorrhizal and pl ant functional groups (Korb et al. 2003). They discovered that arbuscular mycorrhizal (AM) fungi were highly positively correlated with increases in grasses and forbs, and negatively correlated w ith tree cover and pine litter. Ectomycorrhizal (EM) fungi had no response to the restoration treatments, but had a high positive correlation to stand basal area (Korb et al. 2003). A companion study found that as plant species richness increased primarily due to an increase in legumes and stress tolerant plants, there was a corresponding in crease in soil fungi and an abundance of fungi relative to bacteria (Smith et al. 2003). A growing number of studies have indicated that soil microbial communities with distinct functional groups inhabit different forest types (Pennanen et al. 1999). A black pine forest in northeastern Austria was found to have higher relative amounts of fungi and actinomycetes in the soil microbial biomass than were found in a nei ghboring oak-beech hardwood forest (Hackl et al. 2005). Chapman et al. (2003), investigating nativ e woodland expansion in England, found that soil moisture, pH, and microbial biomass levels decreased along a successional gradient from moorland to grassland to mature pine forest, but the fungal component increased. In beech ( Fagus sp.) forests of Denmark, researchers found that different fractions of coarse woody debris supported distinct fungal species. Larger trees parts contained more f ungal species, smaller pieces had higher densities of a few species, and snags were species-poor. They concluded that coarse woody debris should be left as whole trees compared to smaller or larger pieces to insure high species richness in the fungal community of the forest floor (Heilmann-Clausen and Christensen, 2004). These studies illustrate th e strong interactions that exist between soil biogeochemical properties and aboveground cover type. PAGE 69 69 The objective of this study was to examine th e relationships between key soil chemical and microbial properties and the oversto ry and understory plant characte ristics of a wet longleaf pine flat community in the Gulf Coastal Plain of Florida. We hypothesized stand volume will show a positive relationship with soil nitrogen mineraliza tion, which, in turn, will be driven by the microbial community dynamics in the soil. We also hypothesized that the fungal biomass will increase as coarse woody debris accumulated on the forest floor and the standing stock increased over time Materials and Methods Study Areas Three reference site locations along a spatial gradient from within the Coastal Flatwoods subecoregion of Florida (Chapt er 1), sub-divided into three one-hectare blocks, representing young, mid-aged, and mature age classes, were used in this study. The different successional age classes represented a chronosequence of 110-year s across a moisture gradient containing mesic flatwoods, wet flatwoods, wet savannas, and bo rdered by cypress ponds. The three hectares established at each reference location was scaled to match the three hectares established at the restoration site (Chapter 5). The three locations are Topsail Hill State park, St. Marks National Wildlife Refuge, and the Chassahowitzka Wildlif e Management Area of the Florida Fish and Wildlife Commission (Chapter 2). Field Measurements Each site had a clus ter of three one-hectare bl ocks, containing stands representing each of the three previously defined age classes. Each one-hectare block was sub-divided into four randomly placed 400 m2 measurement plots. To assess the fo rest structure, tree height and diameter-at-breast height (DBH), were measured on all trees 10cm DBH. At least two of the dominant trees were cored to determine sta nd age. Stand density (trees/ha), basal area (m2/ha) PAGE 70 70 and volume (m3/ha) were calculated from this data. In addition, the volume (m3/ha) of all snags and coarse woody debris (CWD) were also measured (Spetich et al. 1999). Each 400 m2 plot contained four randomly placed 1 m2 subplots for understory sampling (Chapter 2; Figure 2-2). Percent cover of each species was assessed using the Daubenmire method modified to estimate eight levels of percent cover (Daubenmire, 1959). Coleman rarefaction and the Shannon-Weiner diversity indices were calc ulated for each stand based upon four sub-samples (Colwell, 2006; Koellner and Hersperger, 2004). Soil Sampling and Preparation Soils were sam pled (> 500 grams) from within the vegetation survey (1m2) quadrats taken in the top 10cm of the A horizon. The samp ling took place during August and September of 2005, and September of 2006, at each of the reference sites and the restorati on test site. They were immediately stored at 4C until analysis. A sieved and oven dried (105C) sub-sample (20g) was used for determining moisture content. Soil Chemical Analysis The soil s amples were analyzed at the University of Florida Soil Testing Laboratory (UF ARL), Gainesville, Florida. Soil water pH was determined from prepared slurries using a soil-towater ratio of 1-to-2 (EPA method 150.1). Plantavailable phosphorus was determined with the use of Mehlich-1 extractant (H2SO4 & HCL) and measured on an inductively coupled plasma (ICP) spectrophotometer (EPA method 200.7). So il organic matter content (SOM %) was determined by the Walkley-Black method. The grav imetric soil water content was determined in 2005 and 2006 for all of the samples analyzed. PAGE 71 71 Mineral Nitrogen Fluxes Net nitrogen m ineralization was determined by comparing collected paired soil samples contained in plastic bags. Forty eight samples we re collected from each reference location for a total of 144. One bag was buried in situ for incubation during August, 2005 (Eno, 1960) and the other bag taken to the soil lab fo r analysis. The incubated bags were collected and analyzed after 30 days. Mineral nitrogen was extracted from 20 g of both soil samples with 60 ml 2N KCL and placed in shaker for one hour. They were then filtered through # 42 Whatcom filter papers and analyzed by the UF ARL for ammonium (EPA method 350.1) and nitrate (EPA method 353.2) with a continuous auto-flow analyzer. Net Mineralization was calculated as the difference between incubated-N and initial-N (corrected for soil moisture) (K eeney & Nelson, 1982). Bacterial Abundance and Microbial Dynamics Enum eration of nitrifying bacteria was de termined by the most probable numbers (MPN) method for densities of ammonium and nitrite oxidizing bacteria using a five tube dilution (Schmidt and Belser, 1982). The ammonium oxidizi ng bacteria were incubated in a medium of di-ammonium sulfate, and the nitrite oxidizing bact eria were incubated in potassium nitrite. The tubes were incubated for 8 weeks for the first read ings and 16 weeks for the final readings. A pH indicator of bromothymol blue was used to determine pH changes caused by increased respiration of the ammonium oxidizing bacteria. Positive readings for the nitrite oxidizing bacteria were determined from a nitrate test r eagent of diphenylamine in sulfuric acid solution (Schmidt and Belser, 1982). Soil microbial biomass C was determined by ch loroform fumigation-extraction (Vance et. al., 1987), with the following modifications. Siev ed 12 grams of soil were taken from soil samples stored at 4 C and then placed in 50 ml centrifuge test tubes. Matching 12 gram soil samples were set aside in additional 50 ml centrif uge tubes as the control. The soil samples were PAGE 72 72 fumigated in 24 tubes per desiccator with 40 ml of alcohol-free chloroform placed into a center beaker and an additional 0.5 ml of chloroform was placed into each centrifuge tube. The top of the desiccator was pressure sealed and vacated until the chloroform began to boil. The tubes were then incubated for 24 hours at 25C. The de ssicator was then opened, resealed, and after the chloroform was reboiled, incubated for an additional 24 hours. The control and fumigated samples were extracted with 36 ml of 0.05 M K2S04, shaken (360 rpm) on an orbital shaker for 1 hour, and centrifuged @ 6000 rpm for 15 minutes. The supernatant was then filtered through # 42 Whatcom filter papers into 20 ml scintillation vials and frozen until analysis. Levels of total organic carbon (TOC) were determined on a Shimadzu TOC-VCSH analyzer (Vance & Entry, 2000). Microbial biomass carbon was equal to [(fumTC ConTC) / 0.51] / (Soil Wt.) = mg C kg dry wt. Soil -1 (Joergensen, 1996). The value of 0.51 is the conversion factor equal to the extractable portion of microbial biomass in a forest soil. Fumi gated and non-fumigated blanks were measured to correct for the ch loroform and potassium persulfate. Soil fungal biomass levels were determined by a physical disruption method for extraction of ergosterol from soil samples (Gong et al. 2001 ); with the following modifications. Weighed 6 grams of soil were mixed with 9 ml of 0C me thanol and 1.9 grams of glass beads into 20 ml scintillation vials. The vials were vortexed for 30 seconds, shaken (360 rpm) on an orbital shaker for 1 hour, and refrigerated over night. An a liquot of 1.8 ml was plac ed into 2 ml microcentrifuge tubes and centrifuge d @ 11,000 rpm for 20 minutes. A syringe was used to remove 1.5 ml of the supernatant from the micro-cent rifuge tubes and filtered through a 0.20 m filter into amber colored 2 ml glass HPLC vials. Th e HLPC vials were covered with aluminum foil and stored in the dark at 0 degrees C until rea dy to inject into the HPLC. Each sample was quantified on a Beckman Coulter HPLC equi pped with an UV detector, a pump, an auto- PAGE 73 73 sampler, and through a C-18 reverse-phased anal ytic column (4.6 x 250 mm). The UV detector was set at 282 nm and pure methanol was used as the mobile phase at a flow rate of 1 ml per minute. Extracts (100 l) were injected while the column pressure was maintained at 1000 psi. Pure ergosterol (Sigma) was recrystalized in pur e methanol at different concentrations to establish a set of standards. The standard curve was construc ted from on a linear regression relationship between peak area and ergosterol concentration Ergosterol recoveries were calculated from the difference between spiked and non-spiked paired samples divided by the amount of ergosterol added Under such conditions, an isolat ed peak was identified from field samples at approximately the 13 minutes, base d upon the peaks obtained from the ergosterol standards. An averaged conversion factor for 3.65 g ergosterol per mg of soil translates to a fungal biomass (mg /g -1 soil) when multiplied by (220) (Montgomery et al. 2000). Fungal: microbial biomass ratios were re presented by a ratio of the calcul ated soil fungal biomass, and the soil microbial C biomass levels for each sample. Experimental Design and Analysis A three stage balanced nested design was used to integrate the indicators measured at different scales, and between sites. Since the monito ring of the restoration s ite with nine distinct reference locations produced a dataset where the assumptions for analysis of variance (ANOVA) were not ensured, non-parametric tests were used to detect any signifi cant differences between the reference sites and between the dis tinct forest age classes (SAS, 2002). Inter-relationships between forest structural variables, understory sp ecies diversity indices, and the soil biogeochemical variables we re determined by Spearman's rank ( r ) correlations using SAS 8.2 (Dumortier et al. 2002; SAS, 2002; Spyr eas and Mathews, 2006). Trends between variables were obtained from linear regressi on using the general lin ear model (PROC GLM) PAGE 74 74 (Yang et al. 2006; SAS, 2002). These trends we re enhanced by incorporating moving average smoothing (MA model) as a data filter to reduce seasonal variations found in the datasets for a number of the indicators affected by climate ( Platt and Denman 1975; Kumar et al. 2001; Ittig, 2004). The trend analysis was followed by log10 data transformations where necessary. Results Nitrifying Bacteria and Nitrogen Mineralization Young forest soils at one of the reference si tes, S t. Marks, had numbers of ammonium oxidizing bacteria (AOB) that were 34 times gr eater (14,690 / g soil) than that found in soils from the mature sites (427 / g soil) (Table 41). The higher AOB number s in the young forest soils corresponded to lower a mmonium production (0.14 mg NH4 + / kg soil/month) and higher nitrate production (Table 4-2). Topsail Hill State Preserve also had numbers of ammonium oxidizing bacteria that were 60 times greater in the young forested soils (240 / g soil) than found in soils from the mature sites (4 / g soil) (Table 4-1). However, the young wet pine savanna had very high ammonium production compared to nitrif ication (Table 4-2). The mesic mature forest soils at Topsail had lower amm onium levels than the wet young forest soils (Table 4-2). The numbers of AOB at St. Marks (14,690) were significantly larger compared to the numbers measured at Topsail Hill (240). The higher AOB numbers in the soil under the young forest at St. Marks resulted in lower ammonification 0.14 mg NH4 + / kg soil/month compared to the soil from the young forest at Topsail Hill 17.9 mg NH4 + / kg soil/month. St. Marks had larger numbers of AOB in the old forest soils (427 vs. 4.0), but the level of ammonium production was smaller 2.98 vs. 4.98 mg NH4 + / kg soil/month when compared with the soils from the Topsail old forest site (Table 4-2). The numbers of nitrite oxidizing bacteria (NOB) showed differences between the age groups (427 / g -1 soil, in young soil vs. 4 / g-1 soil, in old soil), but not between the sites PAGE 75 75 (Table 4-1). The nitrate production levels in young (2.39 vs. 1.74 mg NO3 / kg soil/month) and old (1.57 vs. 0.9 mg NO3 / kg soil/month) soils were simila r between the sites (Table 4-2). Overstory Stand volume increased with net nitrogen m ineralization (Nmin) until the volume reached 200 m3 / ha, when Nmin decreased substantially (Figure 4-1). All of the overstory stand variables were positively correlated with microbial biomass carbon (Cmb) during the young age class, but mean stand DBH and height were negatively correlated with Cmb during the mid-aged and mature age classes (Table 4-3). Similar to Cmb, all of the forest structural variables were positively correlated with fungal biomass carbon (Cfb) during the young age class, but remained positively correlated with Cfb during the mid-aged class (Table 4-3). FB-to-MB ratios increased with stand height during the mid-aged and mature age classes, when the mean stand height was greater than 7.5 m (Figure 4-2). Cfb increased by more than 130% as stand BA approached 10 m2 / ha. However, Cfb declined by 30 % as the stand BA grew from 10 m2 / ha to 20 m2 / ha (Figure 4-3). Coarse woody debris CWD was positively correlated with Cfb during the mid-aged and mature age class (Table 4-3). Cfb increased by more than 45 % as CWD increased from 1 to 55 m3 / ha (Figure 4-4). Stand density had a positiv e relationship with soil organic matter content (SOM) during the young and mid-aged class, bu t not during the mature age (Table 4-3). Understory Coleman rarefaction was positively correlated with Cmb during the young and mid-aged class, and negatively correlated to Cmb during the mature age class (Table 4-3). The Coleman rarefaction index decreased by 50% and ShannonWierner diversity inde x by 25% as the FB-toMB ratio approached 1.0 (Fi gure 4-5; Figure 4-6) The Coleman rarefaction index and the Shannon-Wiener diversity index were also negatively correlated with soil organic matter content (SOM), during the young age and mature age classes (Table 4-3). PAGE 76 76 Discussion Net nitrogen m ineralization declin ed at a stand volume of 200 m3 / ha which corresponds to a stand age of 90 years (Chapter 2; Figure 2-9) This could be a stand volume threshold where fungi and actinomycetes have become the major decomposers in the microbial community due to lignin concentrations (Richards, 1987). Even in high soil moisture condit ions, the forest soils from young longleaf pine stands ha d significantly higher levels of nitrifying bacteria than soils from mature pine sites. The nitrifying bacteria data confirmed that nitrif ication rates were higher during the young age class than m easured in the mature aged stands. The AOB numbers were highly variable between sites, bu t the NOB numbers were similar. Nitrate levels were lower and ammonium levels were higher in the soils from the mature forest sites compared to the soils from the young forests. The higher levels of ammonium a nd lower levels of nitrate in mature forest soils could be an indication of a nitrogen c onserving (tighter) ecosystem (Davidson, 2000). There was an exception with the wet young Topsail pine savanna soil that had higher ammonification levels than the mesic mature Topsail soil. Higher ammonium levels and lowe r nitrification levels have been measured in wet longl eaf pine sites when compared to more xeric sites (Wilson et al. 2002). Ammonium production was hi gher and nitrate production was lower in the soils from the unburned Topsail Hill sites compared to St. Marks. The larger numbers of nitrifying bacteria measured at St. Marks NWR compared to Tops ail Hill State Park were probably due to the higher frequency of prescribed fire implemented at St. Marks. Higher ni trification rates after prescribed fire have been meas ured in a number of studies (Cooks on et al. 2007; Hart et al. 2005; Wilson et al. 2002). In addition, researchers studying disturbance in a Norway spruce ( Picea abies ) forest measured large enumerations of ammonium oxidizing bacteria (AOB) in sites recently harvested, but only detected very small numbers (< 10 / gm) in mature undisturbed sites (Paavolainen and Smolander, 1998). PAGE 77 77 The effect of forest growth on the environmen t represents more than creating a preference for shade tolerant plant species or the creation of a multi-layered architecture. It also represents the evolution of soil organic matter (SOM) input s from an easily decomposed substrate to a SOM complex having a higher portion of recalcitrant material. As the inputs to the soil change, there is a corresponding change in the soil microbial commun ity as ectomycorrhizal and saprophytic fungi play greater ro les. This relationship between the aboveground component and the belowground biological community is important in shaping the ecolo gical trajectory of ecosystems (Hackl et al. 2005). The positive relationship between stand de nsity and soil organic matter (SOM) through the mid-aged class illustrates the effect of site quality on stand productivity. Stand BA and volume had strong positive relationships with Cmb up to the mature age class (60 years+). Correlations also showed strong positive relationships between most of the forest growth variables (DBH, height, BA) and Cfb levels, again up to the mature age class (60 years+). These two relationships reinforce how the rate of stand volume growth is interdep endent on the rate of organic matter decomposition and nutrien t cycling (Vitousek and Reiners, 1975). Regression analysis produ ced a trend showing that Cfb increased dramatically as stand basal area decreased. A ponderosa pine restoration study produced similar results showing positive relationships between increases in forest basal area and higher levels of ectomycorrhizal (EM) fungi (Korb et al. 2003). The relationship be tween the fine root biomass of trees and EM fungal levels has also been well established (Hendricks et al. 2006; Wallander et al. 2001; Sylvia and Jarstfer, 1997). Cfb was also found to increase with higher accumulations of CWD. Researchers in Demark determined that a comb ination of larger DBH logs and the greater surface area of smaller diameter CWD, promoted the highest le vel of fungal species richness PAGE 78 78 (Heilmann-Clausen and Christensen, 2004). Whether the increase in Cfb during longleaf pine succession was due to the size of a trees root sy stem (mycorrhizal fungi) or in part due to increases in coarse woody debris accumulation (saprophytic fungi), fungal biomass (Cfb) increased as the average stand height incr eased. These results indicate that both Cmb and Cfb are important soil variables for longleaf pine flat development, but the Cfb portion of the biomass becomes more important over time as the eco system requires the decomposition of larger amounts of CWD and the improved cycling of nutr ients (Hackl et al. 2005; Leckie et al. 2004; Pennamen et al. 1999). The relationship between the FB-to-MB ratio and the Coleman rarefaction index was similar to Shannon-Wiener diversity index, species diversity decreased as the fungal component increased. Both Coleman Rarefaction and Shannon-Wiener diversity H indices were also negatively related to SOM during the young and ma ture age classes. Through a restoration study in England, researchers also f ound a negative relationship between native plant species richness and soil fertility. Never the less, in contrast to our results, they found a positive relationship with plant species richness and FB-MB ratios. The inves tigators attributed this positive relationship to a greater presence of legumes in the lo wer fertile soils (Smith et al. 2003). Conclusions The m ajority of the soil biogeochemical indica tors influenced longleaf pine stand growth, and as stands developed, changes in aboveground vegetation influenced the soil biogeochemical indicators. Net nitrogen mineralization increase d with stand volume until a threshold of 200 m3 / ha (stand age = 90 years). Nitrat e was found to be in higher con centrations in the young forest soils than the mature forest soils. Populations of nitrifying bacteria (AOB + NOB) were also found to be higher in the young forest soils. At T opsail Hill, ammonium levels were found to be higher in the wet young pine savanna soils than the mesic mature soil. Higher soil moisture PAGE 79 79 translates to lower nitrificati on levels. The rela tionships between fungi and increases in stand height or coarse woody debris accumulation indicate a strong c ontinual relationship between the soil biogeochemical indicators and longleaf pine stand development. The dynamics of this relationship might be better unde rstood if the measured fungal biomass could have been identified as arbuscular mycorrhizal (AM) f ungi, ectomycorrhizal (EM) fungi, or saprophytic fungi along the chronosequence. The dominance of fungi negatively affected the Coleman Rarefaction and Shannon-Wiener di versity indices. This may have indicated a decrease in species richness, but the functi onal redundancy component of ecosy stem resilience has probably been strengthened. The strong relationships between forest biomass accumulation and soil biogeochemistry should be assessed in any m onitoring event. Nitroge n cycling appears to become tighter in mature forests at a thres hold of 90 years. This condition is dependent on mycorrhizal and saprophytic fungi do minating the soil microbial biomass. PAGE 80 80 Table 4-1. MPN enumerations of nitrifying bacter ia in young and old longleaf pine forest soils. Enumerations ( MPN g -1 ) All values are expressed in units of MPN per gram (wet weight) of 0 to 10-cm soil and are averages of three replicates. Lower and upper limits in pare ntheses reflect 95% confidence intervals. St. Marks Mature site (100 yrs.) Topsail Hill Sapling site (19 yrs. Topsail Hill Mature site (100 yr s 0.0040 X 103 (0.005, 0.123) 0.4273 X 103 (0.103, 1.385) 0.0036 X 103 (0.005, 0.123) 0.0427 X 104 (0.103, 1.385) 0.0240 X 104 (0.047, 0.965) 0.0004 X 104 (0.005, 0.123) Ammonium Oxidizers Nitrite Oxidizers St. Marks Seedling site (6 yrs.) Site Locations 0.4273 X 103 (0.103, 1.385) 1.4690 X 104 (0.278, 6.318) Table 4-2. Ammonification and nitrification in y oung and old longleaf pine forest soils. Site Locations Ammonification (mg NH4 / kg soil / month ) Nitrification (mg NO3/ kg soil / month ) St. Marks Seedling site (6 yrs.) 0.14 2.39 St. Marks Mature site (100 yrs.) 2.98 1.57 Topsail Hill Sapling site (19 yrs.) 17.9 1.74 Topsail Hill Mature site (100 yrs.) 5.98 0.9 Values expressed as me an monthly rates and based on the dry weight of soil. PAGE 81 81 Table 4-3. Soil biogeochemical relationships with stand attributes based upon Spearman Rank correlations r as stratified by forest age class (n = 48). CmbSOMCfbFB-to-MB ratio Stand Height0.524**** 0.543**** Stand Density 0.394** Stand DBH0.513**** 0.510**** Stand BA0.542**** 0.593**** Stand Volume0.540**** 0.564**** CWD Shannon Diversity -0.584**** Coleman Rarefaction 0.464***-0.363**CmbSOMCfbFB-to-MB ratio Stand Height-0.345* 0.296*0.476*** Stand Density0.509***0.581**** -0.358* Stand DBH-0.401**-0.365**0.319*0.539**** Stand BA0.465***0.585****0.348* Stand Volume0.360*0.502***0.457** CWD 0.326*0.339* Shannon Diversity -0.396**-0.290* Coleman Rarefaction 0.351*0.302*-0.322*-0.517***CmbSOMCfbFB-to-MB ratio Stand Height-0.646**** -0.289*0.446** Stand Density Stand DBH-0.429**0.422** Stand BA Stand Volume CWD0.326*0.648****0.293* Shannon Diversit y -0.439** Coleman Rarefaction -0.348*-0.565****Mid-Aged Time Interval ( 2555 ) Mature Time Interval ( 60 110 ) Prob > |r| under H0: Rho=0 Regeneration Time Interval ( 6 20 ) Significance of the Spearman rank corre lation test: blank: non-significant, *0.05 < p 0.01, **0.01 < p 0.001, ***0.001 < p 0.0001, **** p < 0.0001. PAGE 82 82 y = 1.8739x2 35.731x + 233.45 R2 = 0.44 p < 0.0038 0 4 8 12 16 20 050100150200250300 Stand Volume (m3 / ha)Net Nitrogen Mineralization (mg N / kg soil /month) Figure 4-1. Net nitrogen minerali zation versus stand volume as measured from 26 differently aged stands. y = 0.0041x2 0.0621x + 0.4083 R2 = 0.41 p < 0.0017 0 0.2 0.4 0.6 0.8 1 1.2 51 01 52 02 5 Stand Height (meters)FB-to-MB ratio Figure 4-2. The fungal biomass (FB)-to-microbial biomass (MB) ratio vers us stand height as measured from 26 differently aged stands. PAGE 83 83 y = 9.5282x + 93.367 R2 = 0.35 p < 0.0154 0 50 100 150 200 250 300 350 400 0510152025 Stand BA (m2 / ha)Cfb (mg C / kg soil) Figure 4-3.Fungal biomass carbon (Cfb) versus stand basal area (BA) as measured from stands grouped within the mid-aged and mature age classes only. y = 1.5077x + 105.9 R2 = 0.33 p < 0.0029 0 50 100 150 200 250 300 -551525354555 Coarse Woody Debris (m3 / ha)Cfb (mg C / kg soil) Figure 4-4. Coarse woody debris accumul ation versus fungal biomass carbon (Cfb) as measured from 26 differently aged stands. The data was filtered with moving average smoothing to remove seasonal and cyclic effects. PAGE 84 84 y = -10.243x + 18.279 R2 = 0.37 p < 0.0025 5 10 15 20 00.20.40.60.81 FB-to-MB ratioColeman Rarefaction ES(63) Figure 4-5.Coleman Rarefaction index versus the fungal biomass (FB) -to-microbial biomass (MB) ratio as measured from 26 differently aged stands. The data was filtered with moving average smoothing to remove seasonal and cyclic effects. y = -0.9518x + 2.2266 R2 = 0.37 p < 0.0017 1.00 1.50 2.00 2.50 00.20.40.60.81 FB-to-MB ratioShannon-Wiener Diversity H' Figure 4-6. Shannon-Wiener divers ity H index versus the fungal biomass (FB)-to-microbial biomass (MB) ratio as measured from 26 differently aged stands. The data was filtered with moving average smoothing to remove seasonal effects. PAGE 85 85 CHAPTER 5 MONITORING RESTORATION SUCCESS US ING VE GETATION AND SOIL AS KEY INDICATORS: CASE STUDY OF A WET LONGLEAF PINE FLATS RESTORATION PROJECT Introduction Ecosystem s ecology requires the integration of structural and functional characteristics for developing a holistic understandi ng of ecological change caused by natural or anthropogenic disturbance. However, these two characteristic s of ecosystems have generally been studied separately along vegetative and geochemical gradients (Muller, 1998). Soil chemical and biotic properties need to be included as indicators with forest structural and vegetative compositional measurements for the integration to take pl ace (Johnston and Crossley, 2002). Soil microbial community analysis also provide s a means to measure how respons ive soils are to disturbance and restoration treatments (Harris, 2003). Additionally, the inter-relationships between vegetation and soil characteristics have also been identified and used to assess site quality. In pine plantation research, specifi c soil properties have been found to be associated with the growth of specific tree and plant species. Similarly, certain groups of plant species may indicate specific soil conditions (Burger and Kelting, 1999; Zas and Alonzo, 2002). This combination of above and below ground data can also be used to ecologically verify if a restoration site falls within the spatial gradie nt of the reference sites (Goebel et al. 2001). The research reported in Chapter 3 determin ed that net nitrogen mineralization rates increased until 90 years. It was also determined that the soil fungal-to-microbial biomass ratio increased with stand growth and total woody de bris accumulation. Fina lly, soil fungal biomass increased with mean stand height (Chapter 4). These results show strong relationships exist between stand development and soil biochemical dynamics. This paper examines a case study of a restoration project, hereafter re ferred to as the Pt. Washington restoration project in Florida. PAGE 86 86 The Pt. Washington restoration project was in itiated in 2001 to convert a slash pine plantation to a longleaf pine ecosystem. The e ffects of low-level herb icide applications on longleaf pine development and understory species richness were evaluated. The central goal of this experimental application of herbicides was to determine which herbicide, as a substitute for fire, would produce the best results for longleaf pine seedling su rvival and growth, understory plant species richness and composition, and soil nitrogen mineralization. Herbicides are currently being used in restoration projects thr oughout the United States for promoting the establishment of native grasslands assisting in the control of exotic invasive species as part of integrated pest management programs, for weed control during early forest stand development, and to combat eutrophicat ion from unwanted plant growth in aquatic ecosystems (Sigg, 1999). Yet, many environmentally-sensitive managers and scientists are hesitant to support the use of herb icides making it imperative that th e correct herbicide is used in the proper environment, with th e lowest feasible application rates (Murphy, 1999; Sigg, 1999). What primary factors make the restoration of coastal longleaf pine flats unique compared to other pine ecosystems? First, longleaf pine regeneration is de pendent on a grass stage when the pine seedlings are able to survive light surface fires and during fierce vegetative competition (Boyer and Peterson, 1983; Boyer, 1990). Longleaf pine seedlings have been known to stay in the grass stage from 5-20 years. This protective state can make th e growth rates of longleaf pines unpredictable (Haywood, 2000). Secondly, altho ugh many longleaf pine ecosystems are found on a variety of upland sites (Peet and Allard, 1993), coastal wet pine flats are unique because they are located on low, rain-fed coastal terr aces where weather patterns maintain high soil moisture conditions for extended periods duri ng the growing season (M essina and Conner, 1998). PAGE 87 87 The longleaf pine grass stage becomes the critic al factor in the amount of time the pine remains in the seedling stage when prescribed fire is restricted. On e study found that the application of a hexazinone herb icide helped to release longleaf pine seedlings from fierce vegetative competition, enhancing conditions fo r leaving the grass stage (Haywood, 2000). A recent study of herbicides in an old field rest oration project found a hi gher first-year seedling survival rate and a higher percentage of seedlings out of the grass stage (2nd year) than with no herbicide treatment (Ramsey et al. 2003). There have also been studies to control fuel loads by different fuel reduction technique s, including herbicide applicatio ns. An earlier study found that fire and mechanical removal of fuels had imme diate but short-term impacts on reducing fireline intensity levels, but herbicide applications had longer term affects on reducing fuel levels, starting in the second year afte r treatment and lasting up to si x years (Brockway et al. 1998). Herbicides can be the short-term substitute for fire when applied in the correct manner, and within the correct environment. In 2004, the Pt. Washington restoration projec t was expanded to include the development of a monitoring program, the addi tion of bio-indicator s for evaluation incl uding soil microbial dynamics, and the use of reference sites for estab lishing a chronosequence for the restoration site evaluation. We predict the overstory, understory, and soil biogeochemical indicators will be useful for ecologically classifyi ng the Pt. Washington restoration site as a mesic flatwoods, wet flatwoods, or wet savanna. We will use them fo r trying to detect differences among the four herbicide treatment effects applied on the restorati on site. Finally, we will use them to predict the development or ecological trajecto ry in wet longleaf pine flat restoration. The predicted values will be presented with pine growth results on th e effects of herbicide treatments applied in the second year after planting compared to first year only, consecutive herbicide treatments (1st & PAGE 88 88 2nd Year), and whether an early or late spring application changes the effects (McCaskill data, 2006). Materials and Methods Pt. Washington Restoration Site The longleaf pine restoration project is lo cated on the Point Washington State Forest (30020.04 N, 860 4.22 W) in southern Walton County, Florida. This coastal wet pine flats site was approximately a 4-ha, 26 year-old slash pine plantation having a basal area of 1.85 m2 / ha and an average dbh of 19.1 cm as measured in 1991. It contained scattered residual longleaf pine saplings and poles as part of the stands stocking. The adjacent area makes up approximately 15 ha of mixed slash and l ongleaf pine surrounding a cypress dome, and contained within the greater 6800 ha Pt. Wash ington State Forest. The understory plant community was dominated by broomsedge, a smaller component of wiregrass, and a group of shrub species highlighted by gallberry, saw palmetto, running oak, and dangleberry ( Gaylussacia frondosa). The annual precipitation av erages 1500 mm with most of it occurring during the late summer. The soil belongs to the Leon series and classified as sandy, siliceous, thermic aeric Alaquods. This soil series signi fies that they are very poorly drained soils (Jokela and Long, 1999). Since this pine flats forest is found very close to the coast (within 3 kilometers), its soils were formed on sandy quaternary parent material derived from marine deposits (Stout and Marion 1993). These soils are described as high ly weathered, acidic, infertile substrates (LaSalle, 2002). The surrounding area consists of wet pine sa vannas and wet flatwoods sites that are found within Floridas Gulf Coast Flatwoods zone (Chapter 1; Griffith, 1994). Floridas Gulf coastline is continuously shaped by active fluvial deposition and shoreline processes which promote and maintain the formation of beaches, swamps and mineral flats. The local relief is less than 20 m in PAGE 89 89 elevation. The annual precipitation ranges from 1300,600 mm, and the average annual temperatures vary between 19 to 21 C. The growing season is long, lasting 270-290 days. The parent material consists of marine deposits containing limestone, marl, sand, and clay. The dominant soils are Aquults, Aquepts, Aquods, and Aquents. These acidic soils have thermic and hyperthermic temperature regimes and an aquic mo isture regime. The soils are poorly drained, deep, and moderately textured. Th e dominant vegetative cover c onsists of longleaf-slash pine forests with a smaller co mponent of Choctawhatchee sand and/or pond pine (McNab and Avers, 1994; Parker and Hamrick, 1996). As the first step towards re storing the Pt. Wash ington site back to longleaf pine, the overstory of slash pine was clearcut during August 2001. The site was roller chopped once and prescribed burned in October 2001. There wa s no existing bedding or any other hydrologicalmodifying practice applied. A randomized comple te-block design (RCB) with six blocks was used to measure the effects of four vegetationcontrol chemical mixtures on the dynamics of the understory plant species and pine growth and survival. Five plot s were randomly located within each of the six blocks. All treatment plots we re 26.6m x 24.4 m, and included at least a 3-m buffer strip between plots. The six blocks with buffers make up approximately 3.5 ha within the 4 ha clearcut. In December 2001, one-year-old containerized longleaf pine seedlings were hand-planted at 3.1 x 1.8 m spacing. Seedlings were planted in rows to facilitate the app lication of herbicides. In March 2002, four herbicide treatments Sulf ometuron methyl (methyl 2-[[[[(4,6-dimethyl-2 pyrimidinyl)amino]carbonyl] amino]sulfonyl]benzoate) at 0.26 ai kg ha-1, Hexazinone (3cyclohexyl-6-(dimethylamino)-1-methyl-1,3,5-triazine-2,4(1H,3H)-dione) at 0.56 ai kg ha-1, Sulfometuron (0.26 ai kg ha-1) + Hexazinone (0.56 ai kg ha-1) mix, and Imazapyr (4,5-dihydro- PAGE 90 90 4methyl-4(1-methylethyl)-5-oxo-1-H-imidazol2-y l-3 pyridinecarboxylic acid) at 0.21 ai kg ha-1, were applied in a 1.2 m band over th e top of seedlings using a knaps ack sprayer. In each block, one treatment plot was kept herbicide-free as a control plot (Ranasinghe, 2003). Pine Survival and Growth Pine survival and growth (root co llar diameter and height) were monitored at the end of the growing season every year, through 2006. Seedling hei ght was measured using a ruler, from the soil surface to the top of the bud. Root collar diameter (RCD) was measured using a digital caliper. Stem volume index (SV I) was calculated with the measured RCD and height data. Vegetation Sampling A prelim inary vegetation survey was conducted (June 2001) prior to overstory harvest and site preparation to assess the initial per cent cover of understory species. After study establishment and herbicide ap plication, four vegetation surveys were conducted. Two randomly selected 1m2 quadrats were sampled within each tr eatment plot and the same location was revisited for subsequent surveys. In every su rvey, all plants found within the quadrat were identified to species and assigne d to shrub, graminoid, forb, or fe rn vegetation classes. Percent cover was ocularly estimated by species using the modified Daubenmire scale (Daubenmire, 1959). In addition to percent cover, the number of stems and average stem height were collected for the woody understory species. These plant surveys were c onducted concurrently at the reference sites (described below) during the 2004 growing season. Coleman rarefaction and the Shannon-Weiner diversity indices were calculat ed for each stand (Colwell, 2006; Koellner and Hersperger, 2004). The assemblage pathway for th e plant community was determined from these measurements over time using Canonical Corresp ondence Analysis (CCA) ordination (Palmer, 1993). PAGE 91 91 Reference Sites Three representative locations along a spatial gradient from Pensacola to Tampa Bay (720 km) sub-divided into three one-hectare blocks representing young, mid-aged, and mature age class; were used in this study. Th e different stages (age classes) represente d a chronosequence of 110-years. The three locations are Topsail Hill St ate Park, St. Marks National Wildlife Refuge, and the Chassahowitzka Wildlife Management Area of the Florida Fish and Wildlife Commission. A three stage balanced nested design was used to integrate the indicators measured at different scales, and between site s. Each reference site had a cl uster of three one-hectare blocks containing stands that repr esent young, mid-aged, and 100+ year-old age class. Each one-hectare block was sub-divided into four randomly placed 400 m2 measuring plots where forest structure and coarse woody debris (CWD) were determined. Within each 400 m2 subplot, vegetation was inventoried on four randomly placed 1 m2 quadrat using the same modified Daubenmire scale method utilized at the restoration site. Soil Sampling and Preparation Soils were s ampled from within the vegetation survey quadrats taken from the top 10cm, at each of the reference sites and the restoration test site during August of 2005, and September of 2005 and 2006. They were stored at 4 C until analysis. Sub-samples were sent to the University of Florida soil testing lab for analysis of soil pH by prepared slurries using a soil-to-water ratio of 1-to-2 (EPA method 150.1), organic matter content (%) by the Walkley-Black method, and plant-available phosphorus by the use of Mehlich-1 extractant (H 2SO4 & HCL) and measured on an inductively coupled plasma (ICP) spect rophotometer (EPA method 200.7). Soil microbial biomass was determined by chloroform fumiga tion-extraction extracti on (Vance et. al., 1987). Net nitrogen mineralization rates were estimated from in-situ incubation of soil samples (Eno, PAGE 92 92 1960). Fungal biomass levels were determined by soil ergosterol analysis (Gong et al. 2001). Fungal-to-microbial biomass ratios were calcula ted along the gradient (Montgomery et al. 2000). A sieved and dried (105C) sub-sample wa s used to determine moisture content. Data Analysis Pine survival and growth Pine survival, RCD, and height data collected during five growing seasons were analyzed using analysis of variance (ANOVA) within the fram ework of a randomized complete block design (RCBD) using JMP IN version 5 (SAS In stitute, Inc.). Height and RCD comparisons were made separately for seedlings in the gras s stage (GS) and out of the grass stage (OOGS) using a threshold height of 12 cm (Haywood, 2000) The study addressed only the main effects of herbicide treatment, and tests of these e ffects were not dependent on the assumption of no treatment x block interaction. Block effects were therefore tr eated as random effects in a univariate ANOVA model with two independent variables: treat ment with Block&Random as a covariate. Data were log-transformed where necessary to meet the assumptions of ANOVA. Significant differences between treatments were se parated with the Tukey-Kramer HSD test. Following a prescribed fire in February of 2007, post-fire seedling mortality was assessed in June 2006. Post-fire survival was analyzed with ANCOVA, using pre-fire su rvival as a covariate (Freeman, 2008). Understory The effects of treatm ents on stem counts and he ights of the major shr ub species were also analyzed using ANOVA for a randomized comple te block design. The study addressed only the main effects of herbicide treatment, and test s of these effects were not dependent on the assumption of no treatment x block interaction. Block effects were therefore treated as random effects in a univariate ANOVA model with tw o independent variable s: treatment with PAGE 93 93 Block&Random as a covariate. Data were log-transformed where necessary to meet the assumptions of ANOVA. Significant differences between treatments were separated with Tukeys HSD or Hsus MCB. Post-fire treatme nt differences were analyzed with ANCOVA, using pre-fire distributions as a c ovariate (Freeman, 2008; Ranasinghe, 2003). Biogeochemical indicators A three stage balanced nested design was used to integrate the indicators measured at different scales, and between site s. Significant treatment effects on the biogeochemical indicators ( =0.05) were also compared with the contro l using Dunnetts t-test for multiple means comparison. Hypothesis testing for differences between means was accomplished by using twosample t-test with an alpha of .05 and a two-tailed confidence interval. Since the monitoring of the restoration site with nine distinct refe rence locations produced a dataset where the assumptions for analysis of variance (ANOVA) was not ensured, non-parametric multiple and linear regression, and multivariate Canonical Corre spondence Analysis (CCA) tests (ter Braak, 1994) were used to analyze for similarities a nd differences between the reference sites and between the distinct age class segments using SAS version 8.2 (SAS, 2002). For identifying which variables contribute the most to a gi ven relationship, partial Canonical Correspondence Analysis using univariate multiple regression (PROC CANCORR) was used to determine the relative contributions of each indicat or (Fortin and Dale, 2005; SAS, 2002). Trend analysis was enhanced by incorporating moving average smoothing (MA model) as a data filter to reduce cyclical and seasonal variations found in the datasets for a number of the indicators affected by climate (Platt and Denm an 1975; Kumar et al. 2001; Ittig, 2004). The trend analysis was followed by log10 data transformations where necessary. PAGE 94 94 CCA multivariate analysis functions by relating a primary matrix of plant species abundance data with a secondary matrix of environmental or soil data PC-ORD, a PC-based program (McCune and Meffrod, 1999) containing an algorithm for Canonical Correspondence Analysis (CCA), was used to examine the overall spatial structure of the individual reference sites, the restoration site with the understory plant species along vectors (gradients) for soil chemical, net nitrogen mineralization, and soil microbial values found among the study sites (Heady and Lucas, 2004) (Palmer, 1993). Linear co mbinations of environmental variables are used to maximum the separation of plant species along four dimensional axes. Site scores are derived from the weighted averages of the associ ated species scores. The sites are located in the biplot where the center of the associated spec ies cluster exists. Community structure is illustrated by the influence of different envir onmental variables on its ordination (ter Braak, 1994). Plant species indicator analysis (IndVal) wa s used to measure the level of relationship between a given plant species to ca tegorical units such as pine fl at subtypes or age class. It calculates the indicator value d of species as the product of the relative frequency and relative average abundance in each categorical cluster. In dicator species analysis was used to attribute species to particular environmental conditions based on the abundance and occurrence of that species within the selected group. A species that was a perfect indicator was consistent to a particular group without fail. Indicator values range from 0 to 100 with 100 being a perfect indicator score. Because indicator species analysis is a statistical inference, a test of significance was applied to determine if species are significa nt indicators of the groups to which they are associated (Dufrene and Legendre, 1997). This was achieved by the Monte Carlo permutation test procedure (1000 runs) where the significance of a P-value was determined by the number of PAGE 95 95 random runs greater than or equal to the inferred value ( =0.1). Accuracy was defined from the binomial 95% confidence interval: p +/accuracy (Strauss, 1982). Growth predictions were determined from linear regression using the general linear model (PROC GLM) (Yang et al. 2006). The multiple regression model selection procedures Rsquared, Backward Elimination, and Mallow Cp we re used to determine the combination of indicators for prediction of each variable. The results from regression analysis were based on best model selection criteria of minimizing Ma llow Cp and maximizing R2 and included only those indicators having a biological si gnificance level of p < 0.05 (SAS, 2002). Results Ecological Classification The Pt. W ashington restoration site can be cla ssified ecologically as a wet pine flatwoods subtype of the coastal pine flat based upon th e results from Canonical Correspondence Analysis (CCA) ordination, indicator species analysis (In dVal), and pre-harvest stand data. Canonical Correspondence Analysis ordination indicated th e majority of the plots measured at Pt. Washington fall in between the environmental pa tterns (moisture05, pH, SOM) (Table 5-1) for mesic flatwoods and wet flatwoods measured at th e reference sites (Figure 5-1). Indicator species analysis produced results showi ng that gallberry was the indicator for both wet flatwoods and the Pt. Washington restoration site (Table 5-2). When the data were analyzed by age class, the ordination produced the same vectors of mois ture05, soil pH, and soil organic matter content (SOM), but with stronger results .(Table 5-3). CCA ordination di d not show any clear separation along age class (Figure 5-2). Indicator species analysis did produce results showing that the restoration site had similar plant species as the young age class of the reference sites (Table 5-4). Witch grass and blue stem grass were species indicators for the young age class, while witch grass and wiregrass were found to be the species indicators of the Pt. Washington restoration site PAGE 96 96 (Table 5-4). The means for each of the soil biogeoc hemical variables measured at the restoration site were found to be within the range of mean values measured at the reference sites, except for the significantly higher soil microbial biomass levels (Cmb) (Table 5-5). The seasonal trend for nitrogen mineralization fluxes ove r a 14 month period was for the rates to increase from winter through spring, to peak during the middle of August, and to decline throu gh the fall and winter (Figure 5-3). Pine Growth and Vegetation Control A concurrent study produced results from five years of pine growth and four years of vegetation surveys showing imazapyr and sulf ometuron-hexazinone herbicide treatments significantly reduced longleaf pine seedling survival after four growing seasons. Imazapyr, followed by hexazinone treatments produced significantly higher numbers of pine seedlings in the out-of-the-grass stage when compared to the other treatments. Im azapyr also produced significantly taller pines in th e out-of-the-grass stage compared to the other trees. Imazapyr treatments also resulted in the best control of the overall cover (%) and stem counts of the major shrub species, while producing the highest level of herbaceous richness (Freeman, 2008). From the same four years of pine data (2002-2006) both imazapyr and hexazinone produced better pine growth when applied during the second growing season compared to the first (Table 3-6). They both had higher survival rates as indicated by higher stand densities. Imazapyr produced the best pine growth of all the treatments when app lied during April instead of March. Hexazinone produced better pine growth when applied in March (Table 3-6; McCaskill data, 2006). Treatment Effects-Biogeochemical Indicators Im azapyr produced the highest monthly m ean nitrogen mineralization rates while sulfometuron methyl treatments produced the lo west monthly rates. Most of the herbicides increased the nitrogen mineralization rates, but imazapyr was the only treatment to produce PAGE 97 97 statistically significant higher leve ls of net nitrogen mineralizati on when compared to the control (Figure 5-4). This difference was more pronounced for the ammonification data (Figure 5-5). The sulfometuron methyl mixed with hexazinone treatment produced a higher mean than imazapyr for the nitrification data (Figure 5-6). Only the sulfometuron methyl treatment produced significantly lower microbial biomass leve ls when compared to the control (Figure 57). Two years of herbicide applications resulted in a significant increase in the soil microbial biomass carbon when compared to a single year application (Figure 5-8). The mean microbial biomass carbon levels were higher at the Pt. Washington restoration site than any of the reference sites (Table 5-2; Figure 5-9). Sulfomet uron methyl also produced the lowest levels of fungal biomass, although not significantly different (Figure 5-10). Fungal biomass carbon (Cfb) levels failed to detect significant differences among any of the treatments (Figure 5-10). The predicted values for mean stand DBH, st and density, and stand ba sal area, were close to the actual values (Table 5-7). Predicted valu es involving stand height were different than the actual restoration site. Discussion The vegetative and soil biogeochem ical variab les collected from the reference sites were effective for ecologically classifying the restorati on site at Pt. Washington They were able to determine the pine subtype and the age class. The environmental gradients as evaluated by the soil biogeochemical indicators were stronger dete rminants of ecosystem conditions than was age (Figure 5-1; Figure 5-2). The power of the soil indicators can be realized by the results of the CCA ordination of the sites along the environmental axes and the plant species indicator analysis (Figure 5-1; Table 5-2; Table 5-4). An analysis of all of the treatment effects indicated th at Imazapyr produced the best improvements in pine seedling development and vegetative control while having the smallest PAGE 98 98 impact on the natural patterns of understory herbaceous richness. These gains are offset by Imazapyr producing one of the lowest pine seedling survival rates. The survival rate can be improved if the herbicide is ap plied at the beginning of the second growing season during April instead of March (Table 5-6). Most herbicides are readily broken down by soil microbes causing an increase in their numbers and activity (Haney et al. 2002). Some researchers have found certain herbicides cause a reduction in microbial biomass accompanied by an increase in nitrogen mineralization rates (Busse et al. 2006). The decrease in microbial biomass was attributed to a corresponding decrease in organic matter inpu ts from the vegetative control, and not from direct microbial mortality. In any case, the genera l response following the application of herbicides has been an increase in the soil nitrogen mine ralization rates (Li et al. 2003). If Imazapyr, a leucine, and isoleucine protei n inhibitor, was the only treatment to produce significantly higher ne t nitrogen mineralization rates when compared to the control, then some factor must have partially interf ered with the affects of the othe r chemical treatments on nitrogen cycling. The factor of interference may be tie d to Imazapyr being the only chemical amongst this group of herbicides to be currently register ed for use in aquatic systems (Langeland et al. 2006). The Leon soil series found on this restorati on site has a moderate soil leaching rating and a high soil runoff rating for pesticide selec tion (Obreza and Hurt, 2006). The concerns for hexazinone, an photosystem II quinone inhibitor, are m obility in soils and persistence in water. It was also found to inhibit ammoni fication and promote denitrifica tion, dominant transformations during flooding events (V ienneau et al. 2004). Sulfometuron methyl, an acetolactate synthase inhibitor, has been found to quickly move off-site when applied to sites in contact with wetlands (Michael et al. 2006). The chemical has PAGE 99 99 also been found to be toxic in low concentra tions to many strains of pseudomonas, a major heterotrophic bacteria commonly fo und in forest soils (Boldt and Jacobsen, 1998). This mortality was attributed to the acetolactate synthase i nhibition (ALS) property of sulfometuron methyl (Whitcomb, 1999). This finding might explain the reduction in microbial biomass found in our experiment from applying sulfometuron methyl. The chemical properties of hexazinone a nd sulfometuron methyl limited the treatment effects on this coastal wet pine flat when flooding and the associated high water tables were present. Nitrification was impacted more than ammonification by excessive water from flooding. This condition might explain why the sulfometuron methyl-hexazinone treatment had a significant difference with the control in the nitrification data, but not the net nitrogen mineralization or ammonification data. The effect of soil moisture content on herbicides was observed when comparing the ammonification treatment data with the nitrification results (Figure 5-3; Figure 5-5; Figure 5-6). The results show microbi al biomass measurements are able to detect differences between sites where herbic ides have and have not been applied. These Cmb measurements were also sensitive to the number of herbicide applications used on a given site. Fungal biomass carbon did not detect any differe nces among the treatments. Previous studies have failed to detect any herbicide effect s on fungal communities (Busse et al. 2004). A primary reason for the fungal biomass meas urements failing to de tect statistically significant differences between the treatments may be attributed to the time since the treatments were applied. Most of the individual effects of he rbicides on microbial biom ass levels are greatly reduced beyond two years after application (Li et al. 2003). W ith the exception of the Pt. Washington nitrogen studies, the soils were co llected and analyzed for fungal and microbial biomass 40 months after the second year treatmen ts. The predictions were generally good except PAGE 100 100 for height and volume estimates (Table 5-6). Mean stand height values were skewed due to a group of the 400 m2 forest structure plots measured within the young age class containing naturally regenerated stands. These naturally re generated stands are dominated with larger saplings, poles and some small sawlog-size trees causing the predicted va lues for height and volume in a 6-year old stand to be exaggerated. Why was Imazapyr more effective than the other herbicide treatments in reducing vegetation competition without sign ificantly impacting natural pa tterns of understory succession within this wet flatwoods site? The answer to this question goes back to achieving the central goal of this experiment. The Point Washington State Forest restorati on site suffers from extensive seasonal flooding and dr ought, which adds to pine seed ling mortality and complicates the selection of the proper herbicide for vegetation control. Imazapyr is a broader spectrum herbicide (less selective) and more effective at controlling perennial woody species. These properties are critical in mimick ing fire effects. Secondly, Imazapyr is more persistent in wet sandy soils than the other herbicide treatments. This is also a critical factor in wet longleaf pine flat sites where the water table is constantly near the surface and the effects of herbicide treatments can be reduced by flooding. Conclusions The Pt. W ashington restoration si te contains elements of me sic flatwoods, wet flatwoods, and wet savannas. However, based upon CCA envir onmental ordination, plant species indicator analysis, and pre-harvest stand data it is a wet flatwoods site. These multivariate techniques were also useful in determining similarities between the Pt. Washington rest oration site and the young age class data of the reference s ites. Imazapyr was the best herbicide treatment for this site based on its ability to control shrubs and remain effective during flooding events. In general, herbicide use increased nitrogen mineralization rates, but imazapyr was the only treatment to produce PAGE 101 101 statistically significant higher le vels of net nitrogen mineralization when compared to the control. Both imazapyr and the sulfometuron methyl-hexazinone treatments had a significant difference with the control in the nitrification data. The herbicide-treated restoration site had higher soil microbial biomass carbo n levels than the reference sites. Two years of herbicide applications increased soil microbial bioma ss carbon over a single application. There was an indication that sulfometuron met hyl treatments caused soil microbi al mortality. Higher nitrogen mineralization rates at Pt. Washington were nega tively correlated with both of the species diversity indices. The net nitrogen mineralization data proved eff ective at detecting differences between the herbicide treatments. Soil microbial biomass carbon was sensitive to the amount of herbicide applied. The predictions were generally good except for height and volume estimates. Mean stand height values were skewed due to a group of the 400 m2 forest structure plots measured within the young age class containing naturally regenerated all-aged stands. PAGE 102 102 Table 5-1. Correlations and biplot scores for the biogeochemical variables by pine flat type. Variable Axis 1Axis 2Axis 3Axis 1Axis 2Axis 3 Moisture 05-0.7720.245-0.194 -0.351 0.082-0.058 Soil pH 0.358-0.8750.2720.163 -0.292 0.081 SOM -0.8350.136-0.477 -0.380 0.045-0.142 NetNmin -0.087-0.297-0.349-0.039-0.099-0.104 MBc -0.0480.018-0.026-0.0220.006-0.008 FBc -0.2910.1240.539-0.1320.042 0.160 FBc:MBc 0.4450.144-0.3020.2020.048-0.090 *The Pearson correlations are "intraset correlations of ter Braak (1986). Correlations Biplot Scores Table 5-2. Plant Indicator Values (IndVal) (percent of perfect indica tion) with associated biogeochemical variable by pine flat t ype. P-values repres ent the proportion of randomized runs (1000) equal to or less than observed values ( =0.1). Pine Subtype Plant Species Pine Subtype Mesic Wet Flatwoods Wet Savanna SD PValue Veg Type Mesic Smilax pumila 25 1 5 4.69 0.038 Vine Hypericum hypericoides 17 1 0 3.08 0.024 Forb Gaylussacia frondosa 16 0 4 3.30 0.057 Shrub Pteridium aquilinum 12 0 1 3.00 0.066 Fern Wet Flatwoods Lachnanthes caroliana 0 52 4 3.57 0.001 Forb Arisitida beyrichiana 0 36 0 3.51 0.001 Grass Dichanthelium ovale 6 36 7 4.41 0.007 Grass Cyperus 1 11 1 2.67 0.088 Grass Wet Savanna Ilex glabra 19 13 38 3.55 0.009 Shrub Scleria 17 3 29 3.31 0.014 Grass Pt. Washington Blocks 1&2 Blocks 3&4 Blocks 5&6 SD PValue Veg Type Blocks 1 & 2 Arisitida beyrichiana 34 10 7 5.97 0.039 Grass Tragia urens 13 0 0 2.11 0.016 Forb Blocks 3 & 4 Smilax pumila 0 25 0 5.58 0.001 Vine Pteridium aquilinum 2 18 0 3.42 0.013 Fern Blocks 5 & 6 Scleria 0 3 25 6.16 0.078 Grass Lachnanthes caroliana 0 0 25 4.82 0.024 Forb INDICATOR VALUES (% of perfect indication based on combining the values for relative abundance and relative frequency) n=48 PAGE 103 103 Table 5-3. Correlations and biplot scores for the biogeochemical variables by forest age class. VariableAxis 1Axis 2Axis 3Axis 1Axis 2Axis 3 Moisture 05-0.7720.245-0.194 -0.521 0.142-0.106 Soil pH0.358-0.8750.2720.242 -0.505 0.148 SOM-0.8350.136-0.477 -0.563 0.079-0.260 NetNmin-0.087-0.297-0.349-0.058-0.171-0.190 MBc-0.0480.018-0.026-0.0320.011-0.014 FBc-0.2910.1240.539-0.1960.0720.294 FBc:MBc0.4450.144-0.3020.3000.083-0.164 CORRELATIONS AND BIPLOT SCORES (7 Biogeochemical Variables) Correlations Biplot Scores *The Pearson correlations are "intraset correlations of ter Braak (1986). Table 5-4. Plant Indicator Values (IndVal) (percent of perfect indica tion) with associated biogeochemical variable by forest age class. P-values represent the proportion of randomized runs (1000) equal to or less than observed values ( =0.1). Species codes are found in Appendix A. SpeciesAgeGroupIndValp-value Dich-AnviRegen 32.90.0010 Rhal MidAged27.80.0010 Ilgl Mature 28.40.0160 Dich-ArbePt Wash36.40.0010 PAGE 104 104 Table 5-5. The means for soil biogeochemical variab les between reference site locations and the Pt. Washington restoration site. Site Time Interval (years) Stand Age (years) Soil Moisture 2005 Soil Moisture 2006 N et Nmin (mg-1/kg-1 Soil / year)Cmb (mg/kg/ soil) SOM Content (%) Soil pH [H+] Plant Avail-P (mg-1/kg-1 soil)Cfb (mg/kg /soil) FB to MB Ratio St. Marks Seedling60.340.2221152.84.30.24510.44 St. Marks Mid-Aged360.260.28155891.44.6-0.24750.13 St. Marks Mature1100.250.115491.54.60.31871.66 ChassahowSapling90.440.07111862.94.3-0.091050.57 ChassahowMid-Aged450.230.0881451.14.7-0.041611.11 ChassahowMature710.570.22173694.64.10.231560.42 Topsail HillSapling190.450.10205242.94.3-0.361710.33 Topsail HillMid-Aged490.310.1055591.94.5-0.501790.32 Topsail HillMature1010.320.0974902.04.2-0.401900.39 Pt WashSeedling 60.270.072 1198 1.44.6-0.271260.16 Table 5-6. Pt. Washington actual vs predicted indicator values. Predicted Values Age-6 Reference SitesControl Velpar 1st Year Only Velpar 2nd Year Only Velpar March Application Arsenal 1st Year Only Arsenal 2nd Year Only Arsenal April Application DBH (cm) 3.73 2.873.033.143.313.263.233.62 R2 = 0.81 p < 0.0001 Height (m) 2.090.170.180.250.190.260.290.34 R2 = 0.82 p < 0.0001 Density (trees/ha) 265.57 259268298302224244218 R2 = 0.1 p < 0.0083 BA (m2/ha) 0.09 0.170.190.230.260.190.200.22 R2 = 0.56 p < 0.0001 Volume (m3/ha) 17.50 0.030.030.060.050.050.060.07 R2 = 0.52 p < 0.0001Predicted DBH = [(-0.00510*Age 2 ) + (0.82861* Age) 1.06278], Predicted Height = [(-0.00288*Age2) + (0.45277*Age) 0.526 32], Predicted Density = [(-0.83301*Age) + 270.56934], Predicted Basal Area = [(-0.00190*Age2) + (0.3323*Age) 1.95528], Predicted Volume = [(2.46264*Age) + 2.72759]Pt. Washington Actual Values (2006) PAGE 105 105 Figure 5-1. Pine flat type determined by a thre e-dimensional ordination biplot derived from Canonical Correspondence Analysis (CCA) of 192 plots usin g understory plant species abundance and soil biogeochemi cal data including the Pt. Washington restoration site. PAGE 106 106 Figure 5-2. A three-dimensiona l ordination biplot derived from Canonical Correspondence Analysis (CCA) of 192 plots using unde rstory plant species abundance and soil biogeochemical data collected within the young, mid-aged, mature age class, and the Pt. Washington restoration site. PAGE 107 107 -200 -150 -100 -50 0 50 100 150 200 250 300Jan-02 Feb-02 Mar-02 Apr-02 May-02 Jun-02 Jul-02 Aug-02 Sep-02 Oct-02 Nov-02 Dec-02 Jan-03 Feb-03 Mar-03N Mineralization Rate (mg N / kg soil / month) Ammonification Nitrification Figure 5-3. Monthly variation of total nitrogen mineralization, a mmonification and nitrification rates (mg-1 kg-1 month-1) obtained from field incubation of soils (untreated) during 14 months before and after the 2002 treatments. 0 20 40 60 80 100 120 140 160 TreatmentsNitrogen Mineralization (mg (NH4+NO3) / kg soil) Control Oust Velpar Oust-Velpar Arsenal Figure 5-4. Net nitrogen mine ralization means mg (NH4 + + NO3 -) / kg-1 soil / month for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometuron methylhexazinone mix and Arsenal: imazapyr. Resu lts are from soil samples collected during 14 months before and after the 2002 treatments. a a a ab b PAGE 108 108 0 10 20 30 40 50 60 70 TreatmentsNet Ammonification (mg NH4 / kg soil / month) Control Oust Velpar Oust-Velpar Arsenal Figure 5-5. Net ammonification mean monthly rates (mg-1 NH4 + / kg-1 soil / month) for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometuron methylhexazinone mix and Arsenal: imazapyr. Resu lts are from soil samples collected during 14 months before and after the 2002 treatments. 0 10 20 30 40 50 60 70 80 90 TreatmentsNitrification (mg N03 / kg soil / month) Control Oust Velpar Oust-Velpar Arsenal Figure 5-6. Net nitrification mg -1 N03 / kg -1 soil / month; for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometur on methylhexazinone mix, and Arsenal: imazapyr; applied in different growing s easons, frequencies, and time of year. Results are from soil samples collected during 14 months before and after the 2002 treatments. a a a b b a a a a b PAGE 109 109 0 200 400 600 800 1000 1200 1400 1600 1800 TreatmentsCmb (mg C / kg soil) Control Oust Velpar Oust-Velpar Arsenal Figure 5-7. Microbial biomass carbon (Cmb) mg -1 C / kg -1 soil; for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometuron methylhexazinone mix, and Arsenal: imazapyr. Results are 40 mont hs after second treatment (2006). 0 200 400 600 800 1000 1200 1400 1600 1800 Number of ApplicationsCmb (mg C / kg soil) One Year Two Years Figure 5-8. Effects of one year and two consecutive years of herb icide applications on microbial biomass carbon (Cmb) mg-1 carbon / kg-1 soil from soils. Results are forty months after second treatment (2006). b b b b a a b PAGE 110 110 St. Marks ChassTopsail Pt. Wash0 250 500 750 1000 1250 1500Cmb (mg Carbon / kg soil) Figure 5-9. Soil levels of Microbial biomass carbon (Cmb) mg -1 carbon / kg -1 measured at the reference sites and the Pt. Washington restor ation site. Results are forty months after second treatment (2006). 0 20 40 60 80 100 120 140 160 180 TreatmentsFungal Biomass (mg carbon / kg soil) Control Oust Velpar Oust-Velpar Arsenal Figure 5-10. Fungal biomass carbon mg -1 carbon / kg -1 soil; for the control, Oust: sulfometuron methyl, Velpar: hexazinone, sulfometur on methylhexazinone mix, and Arsenal: imazapyr. Results are four year s after second treatment (2007). a a b c PAGE 111 111 Figure 5-11. Pools and fluxes of nitrogen in the RESDYN restoration model. MP, metabolic pool; grass&forbs, holocellulose pool; sh rubs, lignocellulosic pool; and CWD, woody pool. There are distinctive stabilization coefficients for microbial biomass, young soil organic matter (Y-SOM), and old so il organic matter (SOM) (adapted from Corbeels et al. 2005). PAGE 112 112 CHAPTER 6 SUMMARY AND CONCLUSIONS Ecosystem restoration requires a good monitoring system that allows for the tracking of success by measuring key ecological indicators at th e restoration site and comparing those results with reference communities. The measured ecol ogical indicators must include monitoring changes belowground as well as in the abovegro und vegetation for the coup ling of functional and structural attributes. The overa ll objective of this study was to examine stand structure, understory species composition, and soil ch emical and microbial properties along a chronosequence in longleaf pine wet flats along Floridas Gulf Coast in an attempt to develop an ecological trajectory for this co mmunity. Such an ecological traj ectory would serve as the basis for developing a monitoring framework for restora tion projects in the southern Gulf Coastal Plain. We selected three reference sites within the Gulf Coas t Flatwoods subecoregion to accomplish our objective. Within each reference site we sampled a total of 12 plots, 4 plots each in the early, mid and mature age classes. This experimental design resulted in 29 different age groups representing a chronosequenc e of 2 to 110-year-old stands. The selected reference locations not only represented the highest quality sites that could be found in Florida, but were also located within the specific range for coas tal wet longleaf pine flats found along Floridas Gulf coast. Monitoring this very specific biog eographical area (Gulf Coast Flatwoods subecoregion of Florida) created a spatial gradient pertinent to the restoration site that we wanted to evaluate. The time scale was limited to the oldest available longleaf pine stands (110-year old) distributed along the specified spatial range. The major focus in Chapter 2 was to examine overstory stand structur e data and understory plant species composition along the 110year chronose quence. As expected, stand DBH, height, PAGE 113 113 and basal area increased with age, but reached a steady state plateau ar ound 80-90 years. when they began to decline. Coarse woody debris accumulation levels were highly variable, but tended to increase with age. The decomposition levels of CWD were constant through the mid-aged class, but declined from the mid-age to the mature age class. The level of shrub species was significantly higher in th e mature sites than found in either the young or the mid-aged classes. Stand growth during early development transl ates to habitat heterogeneity as partial shading brings in new groups of plant species. At this point, stand height had a strong positive relationship with the Coleman rarefaction i ndex and stand density has a strong negative relationship with the Shannon-Wiener diversit y index. The plant spec ies turnover rates as indicated by Coleman rarefaction values were high and the evenness of plant species as indicated by Shannon-Wiener was very low. The evenness of plant species was not attained until the mature stage when the number of plant species entering the ecosystem was equal to the number of plant species leaving it. At this point, Sh annon-Wiener diversity values had a strong positive relationship with stand density and the Colema n rarefaction index had a negative relationship with stand height. The equilibrium between Co leman rarefaction and Shannon-Wiener diversity indices at this stage indicate s a steady state in the overstor y. Based upon the chronosequencial trends, Floridas Gulf Coastal longleaf pine fl ats reach the understory reinitiation stage at approximately 90 years. This would mean the fo rest is self-organizing, a threshold point for restoration. In Chapter 3, Our main objective was to measure soil pH, moisture content, organic matter content (SOM), plant-ava ilable phosphorus, soil nitrog en mineralization rates (Nmin), soil microbial biomass carbon (Cmb) and fungal biomass (Cfb) along the same 110-year chronosequence for determining the ecological traj ectory in terms of soil chemical and microbial PAGE 114 114 characteristics of longleaf pine in coastal wet pi ne flat communities. We specifically tested our hypothesis that this group of so il biogeochemical indicators measured along the chronosequence would follow a pattern similar to the biomass accumulation curve for forest succession (Vitousek and Reiners, 1975). In response to rapid increas e in growth during the early years of stand establishment, we predicted a similar increase in net nitrogen mineralization rates, microbial biomass and fungal biomass levels. We hypothesized that these variables would decrease at some point during the mid-aged stage and reach a th reshold steady-state some time during the early mature stage when the understory reinitia tion process of forest succession has begun. Nitrogen cycling was dominated by ammoni um production during the wet 2005 growing season when compared to a drie r 2002. Nitrification represented 50% of the production during 2002 and less than 25% during 2005. There was ammonium enrichment by nitrate reduction. This probably indicates that the dissimilatory-n itrate reduction-to-ammonium (DNRA) pathway was prominent during the flooded 2004-2005 growing seasons. The net nitrogen mineralization rates, microbial biomass carbon, and fungal biomass carbon increased between the young and mid-aged classes, then decreased between the mi d-aged and mature age classes. The FB-to-MB ratios increased dramatically up to 60 years, then decreased to 110 year s. Finally, soil organic matter content (SOM), increased with soil moisture. Based upon the results this group of soil indicators follows biomass accumulation patterns and will attain biogeochemical equilibrium after a stand age of approximately 60-70 years. The threshold would be during the mature age class after the understory reinitiation phase of fore st succession has started. The objective of Chapter 4 was to examine the relationships between key soil chemical and microbial properties and the oversto ry and understory characteristic s of a wet longleaf pine flat community in the Gulf Coastal Plain of Flor ida. We hypothesized stand volume will show a PAGE 115 115 positive relationship with soil nitrogen mineraliza tion, which, in turn, will be driven by the microbial community dynamics in the soil. We also hypothesized that the fungal biomass will increase as coarse woody debris accumulated on the forest floor and the standing stock increased over time The majority of the soil biogeochemical indica tors influenced longleaf pine stand growth, and as stands developed, changes in aboveground vegetation influenced the soil biogeochemical indicators. Net nitrogen mineralization increase d with stand volume until a threshold of 200 m3 / ha (stand age = 90 years). Nitrat e was found to be in higher con centrations in the young forest soils than the mature forest soils. Populations of nitrifying bacteria (AOB + NOB) were also found to be higher in the young forest soils. At T opsail Hill, ammonium levels were found to be higher in the wet young pine savanna soils than the mesic mature soil. Higher soil moisture translates to lower nitrificati on levels. The rela tionships between fungi and increases in stand height or coarse woody debris accumulation indicate a strong c ontinual relationship between the soil biogeochemical indicators and longleaf pine stand development. The dynamics of this relationship might be better unde rstood if the measured fungal biomass could have been identified as arbuscular mycorrhizal (AM) f ungi, ectomycorrhizal (EM) fungi, or saprophytic fungi along the chronosequence. The dominance of fungi negatively affected the Coleman Rarefaction and Shannon-Wiener diversity indices. This may indicate a decrease in species richness, but the functional redundancy component of ecosystem resilience is probably being strengthened. The strong relationships be tween forest biomass accumulation and soil biogeochemistry should always be studied in any monitoring event. Nitrogen cycling appears to become tighter in mature forests at a thres hold of 90 years. This condition is dependent on mycorrhizal and saprophytic fungi do minating the soil microbial biomass. PAGE 116 116 The objectives of Chapter 5 were to use the i ndicator data to ecological classify the Pt. Washington restoration site as a mesic flatwoods, wet flatwoods or wet savanna. Secondly, to use the soil biogeochemical indicato rs for trying to detect differe nces among the four herbicide treatment effects applied on the restoration site. Finally, we will use both the vegetative and soil biogeochemical data to predict th e development or ecological trajec tory in wet longleaf pine flat restoration. The predicted values will be presented with pine gr owth results on the effects of herbicide treatments applied in the second year after planting compared to first year only, consecutive herbicide treatments (1st & 2nd Year), and whether an earl y or late spring application changes the effects (McCaskill data, 2006). The Pt. Washington restoration si te contains elements of me sic flatwoods, wet flatwoods, and wet savannas. However, based upon CCA envir onmental ordination, plant species indicator analysis, and pre-harvest stand data it is a wet flatwoods site. These multivariate techniques were also useful in determining similarities between the Pt. Washington rest oration site and the young age class data of the reference s ites. Imazapyr was the best herbicide treatment for this site based on its ability to control shrubs and remain effective during flooding events. In general, herbicide use increased nitrogen mineralization rates, but imazapyr was the only treatment to produce statistically significant higher le vels of net nitrogen mineralization when compared to the control. Both imazapyr and the sulfometuron methyl-hexazinone treatments had a significant difference with the control in the nitrification data. The herbicide-treated restoration site had higher soil microbial biomass carbo n levels than the reference sites. Two years of herbicide applications increased soil microbial bioma ss carbon over a single application. There was an indication that sulfometuron met hyl treatments caused soil microbi al mortality. Higher nitrogen mineralization rates at Pt. Washington were nega tively correlated with both of the species PAGE 117 117 diversity indices. The net nitrogen mineralization data proved eff ective at detecting differences between the herbicide treatments. Soil microbial biomass carbon was sensitive to the amount of herbicide applied. The predictions were generally good except for height and volume estimates. Mean stand height values were skewed due to a group of the 400 m2 forest structure plots measured within the young age class containing naturally regenerated all-aged stands. Research Implications in Coastal We t Longleaf Pine Flats Restoration The monitoring study proved effective at evaluating our restoration site with a set of indicators that integrat ed the structural and functional at tributes of the wet longleaf pine ecosystem. The aboveground vegeta tive variables and the soil biogeochemical measurements produced similar threshold periods. The selectio n process for the reference sites also proved fruitful based upon the sites having similar sta nd, soil properties, and common understory plant species among the locations. It was critical to restrict the location of the reference sites to within the 3 kilometers of the Gulf coast. Our set of reference sites were selected to evaluate southern coastal pine communities that are directly affected by tr opical storms. The restoration of Gulf coastal wet longleaf pine flats is distinct from other longleaf pine communities. Flooding caused by active hurricane seasons can leave these sites inundated for more th an two years. This condition causes two major results in the biogeochemistry of these pinelands. First, extended floodi ng causes the nitrogen cycle to be dominated by ammonium production. When ammonium becomes scarce, nitrate is converted to ammonium thr ough the DNRA pathway conserving nitrogen losses. Secondly, long term flooding results in the accumulation of so il organic matter, causing the pH of the soil medium to drop. This condition favors fungi and anaerobic bacteria over the aerobes. When the conditions become dry, there is a great flush of growth in both the overstory and understory vegetation. The effects of this flooding cycle are greater on younger forests than mature forests PAGE 118 118 where nitrate is in greater demand because of stand growth requirements. This demand was expressed by the nitrate levels and numbers of nitrifying bacter ia being significantly higher in soils from the young stands compared to the mature stands. When prescribing fire in these sites, it is critical not to burn them during a flooding cycle before the flush of growth is completed. Based upon the conditions at our four sites, that can take 12-14 months af ter the drying process has started. The understory vegetation was also distinct in these wet pine flat s. There are higher densities of facultative wetland grasses and forbs and fewer ha rdwoods, especially the oaks. Very few oaks were measured on any of our site s other than the creepers (running oak). Some of these sites have not been burned in over 5 years. The implication here is th e fire return-intervals can be extended well beyond 2-3 years if flooding conditions exist. The mesic mature sites had a higher composition of shrub species than the young mesic stands, even under fire return-intervals of 3 years. Soil moisture in the terms of extende d flooding can enrich wet l ongleaf pine flat soils, conserve their nitrogen supply, and prevent invasion by shrub species. The flooding cycle can provide as many benefits to coastal wet pine ecosystems as fire does. In summary, monitoring needs to include indicators that meas ure the functions as well as structural attributes of a given ecosystem. This proved to be extremely important in Gulf coastal pine communities where soil cond itions are distinct from in land ecosystems. It was also important to restrict the sites to within the Gulf Coast Flatwoods subecoregion of Florida and to within 3 kilometers of the coast for insuring th e same climatic effects that occurred at the restoration site occurred at each of the reference sites. One result of these stratifications was that all of the sites had 63 understory plant species in common. This may not have been attainable had the spatial scale been broader. This set of indicators and the time scale for the PAGE 119 119 chronosequence can be utilized at other envir onments where longleaf pine ecosystems are found. The chronosequence approach is strengthened by having as many replications as economically feasible at each of the differently aged sites. A difficult and important aspect to monitoring is selecting the spatial scale for the reference sites. If one is looking to monitor longleaf pine in mountain terrain found in the norther n limit of its range, it would be more effective to restrict the reference sites to within that environment in order to capture the eco logical differences found within the local c limatic and soil conditions. A key directi on for future research is to conduct investigations for improving our understanding of the biogeochemical dynamics that take place in facultative pine wetlands (i.e., wet pine-dominated mineral flats). This research would need to include molecular analysis for the identification species that change in soil microbial community between wet and dry conditions. PAGE 120 120 APPENDIX SPECIES CODE LIST Table A-1. Species list. Scientific name Code Common name Shrubs Asimina incana Asin Wooly paw paw Cyrilla racemiflora Cyra Titi Gaylussacia dumosa Gadu Drawf huckleberry Gaylussacia frondosa Gafr Dangleberry Ilex coriacea Ilca Large gallberry Ilex glabra Ilgl Gallberry Ilex vomitoria Ilvo Yaupon Kalmia hirsuta Kahi Hairy wicky Licania michauxii Limi Gopher apple Lyonia lucida Lylu Fetterbush Magnolia virginiana Mavi sweet bay Myrica cerifera Myce Wax myrtle Photinia pyrifolia Phpy Red choke berry Quercus pumila Qupu Running oak Serenoa repens Sere Saw palmetto Stillangia sylvatica Stsy Queens delight Vaccinium spp Vacc Blueberry spp Grasses Andropogan virginicus Anvi Bluestem grasses Aristida stricta var. beyrichiana Arbe Wiregrass Calamovilfa curtissii Cacu Curtis sandgrass Ctenium aromaticum Ctar Toothache grass Cyperus Cype Sedge spp Eragrostis spectabilis Ersp Purple lovegrass Dichanthelium ovale Dich Eggleaf witch grass Panicum Dichanthelium Pani Panicum spp Dichanthelium erectifolium Paer Erect leaf witchgrass Panicum laxiflorum Pala Velvet Witchgrass Scleria Scle Nutrush spp Xyris caroliniana Xyca Yellow eyed grass Forbs Asclepias viridula Asvi Southern milkweed Aster adnatus Asad Scaleleaf aster Aster eryngiifolius Aser Thistleleaf aster Aster reticulatus Asre White top aster Aster tortifolius Asto Dixie aster PAGE 121 121 Table A-1. Continued Carphephorous pseudoliatris Caps Bristleleaf chaffhead Carphephorus odoratissimus Caod Deer tongue Chrysopsis Chry Silkgrass spp Conyza canadensis Coca Canadian horseweed Coreopsis linifolia Coli Texas tickseed Desmodium rotundifolium Dero Tricklyfoil Drosera capillaris Drca Pink sundew Elephantopus tomentosus Elto Devils grandmother Eupatorium capillifolium Euca Dog fennel Eupatorium compositifolium Euco Yankee weed Eupatorium mohrii Eumo Mohrs thoroughwort Eupatorium pilosum Eupi Rough Boneset Euthamia graminifolia Eugr Flat top goldenrod Gelsemium sempervirens Gese Yellow jessamine Gratiola hispida Grhi Rough Hedgehyssop Hypericum hypericoides Hyhy St. Andrews cross Hypoxis sessilis Hyse Glossyseed yellow stargrass Hypoxis spp Hypo Stargrass spp Lachnanthes caroliniana Laca Carolina redroot Lechea Lech Pineweed spp Lechea pulchella Lepu Leggetts pineweed Liatris gracilis Ligr Slender gayfeather Liatris tenuifolia Lite Shortleaf gayfeather Mimosa quadrivalvis Miqu Sensitive brier Oenothera fruticosa Oefr Evening primrose Opuntia humifusa Ophu Prickly pear Pityopsis graminifolia Pigr Silkgrass Pterocaulon pycnostachyum Ptpy Blackroot Rhexia alifanus Rhal Meadow beauty Rhexia petiolata Rhpe Fringed meadow beauty Sabatia brevifolia Sabr Shortleaf Rosegentian Seymeria cassioides Seca Yaupon Blacksenna Smilax laurifolia Smla Laurel green brier Smilax pumila Smpu Green brier Solidago odora Sood goldenrod Stylisma patens Stpa Coastal plain dawn flower Tragia urens Trur Wavyleaf noseburn Verbena brasiliensis Vebr Brazilian vervain Viola septemloba Vise Blue violet Vitis rotundifolia Viro Muscadine PAGE 122 122 LIST OF REFERENCES Achtem eier, G., Jackson, W., Hawkins, B., Wade, D.D. and C. McMahon. 1998. The Smoke Dilemma: A Head-On Collision. Transcript In: The 63rd North American Wildland and Nature Resource Conference 415-421. Adegbidi H.G., Jokela, E. J. and N.B. Co merford. 2005. Factors influencing production efficiency in intensively managed loblolly pine plantations in a 1-to-4-year old chronosequence. Forest Ec ology and Management 218:245-258. Allen A.S. and W.H. Schlesinger. 2003. Nutr ient limitations to soil microbial biomass and activity in loblolly pine forests. Soil Biology and Biochemistry 36:581-589. Allen M.F., Allen, E.B., Zink, T.A., Harney, S., Yoshida, L.C., and R. MacAller. 1999. Soil Microorganisms, In L. Walker (ed.) Ecosys tems of disturbed ground. 16. Ecosystems of the World. Elsevier Press, New York 22:521-544. Allen, W.J., 1991, Soil survey of Wakulla County, Fl orida: U.S. Department of Agriculture Soil Conservation Service 163 p. Allison, V.J., Condron, E.M., Peltzer D.A., Richar dson, S.J. and B.L. Turner. 2007. Changes in enzyme activities and soil microbial comm unity composition along carbon and nutrient gradients at the Franz Josef ch ronosequence, New Zealand. Soil Biology and Biochemistry 39:1770-1781. Aravena, J. C., Carmona, M.R., Perez, C.A., and J.J. Armesto. 2002. Changes in tree species richness, stand structure and soil properties in a succession al chronosequence in northern Chiloe Island, Chile. Revista chil ena de historia natural 52:1-30. Bth, E., Frostegrd,, A., Pennanen, T., and H. Fritze. 1995. Microb ial community structure and pH response in relation to soil organic ma tter quality in wood-ash fertilized, clear-cut or burned coniferous forest soils. Soil Biology and Biochemistry 27:229-240. Bth, E., and T.H. Anderson. 2003. Compar ison of soil fungal/bacterial ratios in a pH gradient using physiological and PLFA-based techniques. Soil Biology and Biochemistry 35:955-963. Bailey, V.L., Smith, J.L. and H. Bolton Jr. 2002. Fungal-to-bacteria ratio s in soils investigated for enhanced C sequestration. Soil Biology and Biochemistry 34:997-1008. Bailey, V.L., A. D. Peacock, J.L. Smith, and H. Bolton 2002. Relationships between soil microbial biomass determined by chloroform fumigation-extracti on, substrate-induced respiration, and phospholipid fatty acid analys is. Soil Biology and Biochemistry 34:13851389. Barnett, J.P. 1999. Longleaf Pine Ecosystem Restoration: The Role of Fire. Journal of Sustainable Forestry 9:89-96. PAGE 123 123 Belser, L.W. 1979. Population ecology of nitrifying bacteria. Annual Reviews of Microbiology 33:309-333. Block, W.M., Franklin, A.B., Ward, J.P., Ga ney, J.L., and C. White 2001. Design and Implementation of Monitoring Studies to Eval uate the Success of Ecological Restoration on Wildlife. Restoration Ecology 9:293-303. Boldt, T. S. and C.S. Jacobsen. 1998. Diffe rent toxic effects of the sulfonylurea herbicides metsulfuron methyl, chlorsulfuron and thifensulfuron methyl. On fluorescent pseudomonads isolated from an agricultura l soil. FEMS Microbi ology Letters 161:29-35. Bormann, F. H. and G.E. Likens. 1979. Pattern and process in a forested ecosystem. SpringerVerlag, New York 253 p. Boyer, W. D. 1990. Pinus Palustris, Longleaf pine. In: Burns, R. M. and B. H. Honkala. Silvics of North America, Vol. 1. Conife rs. USDA Forest Serv. Washington, DC. 405412. Boyer, W. D., and D.W. Peterson 1983. L ongleaf pine. In: Burns, R. M., Complier. Silvicultural Systems for the Major Forest Types of the United States. Agricultural Handbook # 445 USDA Forest Service, Washington, DC. 153-156. Brady, N. C., and R.R. Weil 2002. The Nature and Properties of Soils 13th ed. Prentice Hall Publishing N.J. 960 p. Brewer, J. S. 1998. Patterns of plan t species richness in a wet slash-pine ( Pinus elliottii ) savanna. Journal of the Torre y Botanical Society 125:216-224. Brockway, D. G. and K.W. Outcalt. 1998. Ga p-phase regeneration in longleaf pine-wiregrass ecosystems. Forest Ecology and Management 106:125-139. Brockway, D. G., Outcalt, K. W., Tomczak, D. J. and E. E. Johnson. 2005. Restoration of longleaf pine ecosystems. General Technical Report SRS-83. U.S. Forest Service, Southern Research Station, Ashville, N.C. 34 p. Brockway, D. G., K. W. Outcalt, and R. N. Wilkins. 1998. Restoring lo ngleaf pine wiregrass ecosystems: Plant cover, dive rsity and biomass following low rate hexazinone application on Florida sandhills. Forest Ecology and Management 103:159-175. Brose, P. and D. Wade. 2002. Potential Fire Behavior in Flatwoods Forests Following Three Different Fuel Reduction Techniques Forest Ecology and Management 163:71-84. Burdt, A.C. 2003. Hydric Soil Properties as In fluenced by Land Use in Southeast Virginia Wet Flats. Master Thesis. Virginia Polytechnic Institute and State University. Blacksburg. 191 p. PAGE 124 124 Burger, J. A. and Y. Xu 2001. Sustaining the Productivity and Function of Intensively Managed Forests. Final Project Report # DE-FC07-97ID 13551 U.S. Department of Energy. Virginia Polytechnic Institute and State University 38 p. Burger, J. A. and D.L. Kelting 1999. Using soil quality indicators to assess forest stand management. Forest Ecology and Management 122:155-166. Busse, M. D., Beattie, S. E., Powers, R. F., Sanchez, F. G., and A.E. Tiarks. 2006. Microbial community responses in forest mineral soil to compaction, organic matter removal, and vegetation control. Canadian Journal of Forest Research 36:577-588 Busse, M. D., Fiddler, G. O. and A.W. Ratcliff. 2004. Ectomycorrhizal formation in herbicide treated soils of differing clay and organic matter content. Water, Air, and Soil Pollution 152:23-34. Chapman, S.J., Campbell, C.D., and G. Puri 2003. Native Woodland Expansion: Soil Chemical and Microbiological Indicators of Change. Soil Biology and Biochemistry 35:753. Chauhan, A., Ogram, A.V., and K.R. Reddy. 2004. Syntrophic-Methan ogenic Associations along a Nutrient Gradient in the Florida Ever glades. Applied Envi ronmental Microbiology 70: 3475. Chertov, O.G. Komarov, A.S., Karev, G.P. 199 9. Modern Approaches in Forest Ecosystem Modelling. European Forest Institute Resear ch Report 8 Brill, Leiden Boston Kln. 30 p. Collins, B. White, P.S., and D.W. Imm. 2001. Introduction to Ecology and Management of Rare Plants of the Southeast. Natural Areas Journal 21:4-11. Corbeels, M., McMurtrie, R.E., Pepper, D.A. and A.M. OConnell (2005a) A process-based model of nitrogen cycling in forest plantations Part I. Structure, calib ration and analysis of the decomposition model. Ec ological Modell ing 187:426-448. Corbeels, M., McMurtrie, R.E., Pepper, D.A. and A.M. OConnell (2005b) A process-based model of nitrogen cycling in forest plantations Part II. Simulating growth and nitrogen mineralisation of Eucalyptus globulus plantations in south-western Australia. Ecological Modelling 187:449-474. Corbeels M, McMurtrie RE, Pepper DA, Mendha m DS, Grove TS and O'Connell AM (2005c) Long-term changes in productivity of eucalypt plantations under different harvest residue and nitrogen management practices: A m odelling analysis. Forest Ecology and Management 217:1-18. Colwell, R.K. 2006. EstimateS : Statistical estimation of species richness and shared species from samples. Version 8. Persis tent URL . PAGE 125 125 Cookson W.R., Osman, M., Marschner, P., Abaye, D.A., Clark, I., and C.A. Watson. 2007. Controls on soil nitrog en cycling and microbial comm unity composition across land use and incubation temperature. Soil Biology and Biochemistry 39:744-756. Covington W.W., Fule P. Z., Hart S.C., and R.P. Weaver. 2001. Modeling Ecological Restoration Effects on Ponderosa Pine Forest Structure. Restoration Ecology 9:421. Craft, C.B. and C. Chiang. 2002. Forms and Amounts of Phosphorus Across a Longleaf PineDepressional Wetland Landscape. Soil Scie nce Society American Journal. 66:17131721. Daubenmire, R. F. 1959. Canopy coverage method of vegetation analysis. Northwest Science 33:43-64. Davidson, E.A., Keller, M., Erickson, H.E., Verc hot L.V. and E. Veldkamp. 2000. Testing a conceptual model of soil emissions of nitr ous and nitric oxides, BioScience 50:667. Devries et. al., 2003. Intensive monitoring of fore st ecosystems in Europe. Forest Ecology and Management 174:77-95. Drake, J. A. 1991. COMMUNITY-ASSE MBLY MECHANICS AND STRUCTURE OF AN EXPERIMENTAL SPECIES ENSEMBLE. American Naturalist 137:1-26. Dufrene, M. and P. Legendre. 1997. Species a ssemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Monographs 67:345-366. Dumortier, M., Butaye, J., Jacquemyn, H., Van Camp, N., Lust, N., and M. Hermy. 2002. Predicting vascular plant species richness of fragmented forest s in agricultural landscapes in central Belgium. Forest Ecology and Management 158:85-102. Ecological Restoration Institut e (ERI). 2003. Multiparty Monitoring Handbook Series. Collaborative Forest Restoration Program. USDA Forest Service, Inventory and Monitoring Institute. Ft. Collins, CO. 290 p. Eno, C. 1960. Nitrate production in the field by incubating the so il in polyethylene bags. Soil Science Society American Journal 24:277-279. Farrar, R. M. 1985. Volume and growth predictions for thinned even-aged natural longleaf pine stands in the East Gulf area. USDA Forest Service, Research Paper SO-220. Southern Forest Experiment Station, New Orleans, LA. 171 p. Fierer, N., and R.B. Jackson 2006. The diversity and biogeography of soil bacterial communities. Proceedings Natural Academy of Science 103:626-631. Fortin, M.J., and M.R.T. Dale. 2005. Spatial Analysis: A Guide for Ecologists. Cambridge University Press Cambridge, UK. 261 p. PAGE 126 126 Fraterrigo, J.M., Balser, T.C., and M.G. Turner. 2006. Microbial community variation and its relationship with nitrogen minera lization in historically altered forests. Ecology 87:570579. Frelich, L.E. 2002. Forest Dynamics and Disturbance Regimes. Cambridge University Press Cambridge, UK. 261 p. Frey, S.D., Elliot, E.T., and K. Paustian. 1999. Bacterial and fungal abundance and biomass in conventional and no-tillage agroecosytems along two climatic gradients. Soil Biology and Biochemistry 31:573-585. Frost, C.C. 1993. Four Centuries of Ch anging Landscape Patterns in the Longleaf Pine Ecosystem. In: Proceedings of the 18th Tall Timbers Fire Ecology Conference. The longleaf pine ecosystem: ecol ogy, restoration, and management 1991 Tallahassee, Fl Tall Timbers Research Station 18:17-42. Frost, C.C. 2006. History and Future of the L ongleaf Pine Ecosystem. Pages 9-47 In: Jose, S., Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and Restoration. Springer Science + Business Media, LLC., New York. Gallo, M., Amonette, R., Lauber, C., Sinsabaugh, R.L., and D.R. Zak. 2004. Microbial Community Structure and Oxidative Enzy me Activity in Nitrogen-Amended North Temperate Forest Soils. Microbial Ecology 48:218. Giardina, C.P., Ryan, M.G., Hubbard, R.M., a nd D. Binkley. 2001. Tree Species and Soil Texture Controls on Carbon and Nitrogen Mine ralization Rates. Soil Science Society American Journal 65:1272. Glitzenstein, J. S., and D. Wade. 2003. Fire Frequency Effects on Longleaf Pine Vegetation in South Carolina and Northeast Florida. Natural Areas Journal 23:22-37. Goebel, C. P., Palik, B.J., Kirkman, L.K., Drew M.B., West, L., and D. C. Pedersen. 2001. Forest Ecosystems of a Lower Coastal Plai n Landscape: Multifactor Classification and Analysis. Journal of the Torrey Botanical Society 128:47-75. Gong, P., Guan, X., and E. Witter. 2001. A ra pid method to extract ergosterol from soil by physical disruption. Applied Soil Ecology 17:285-289. Gotelli, N. J., and R.K.. Colwell. 2001. Qu antifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness. Ecology Letters 4:379-391. Grierson, P.F., Comerford, N.B. and E.J. J okela. 1999. Phosphorus mineralization and microbial biomass in a Florida Spodosol: eff ects of water potential, temperature, and fertilizer application. Biology and Fertility of Soils. 28:244-252. Griffith, G. E., Omemik, J.M., Rohm, C.M., and S.M. Pierson. 1994. Florida Regionalization Project. U. S. Environmental Protection Agency, National Health and Environmental Effects Research Laboratory, Corvallis, Oregon. EPA/600/Q-95/002 83 p. PAGE 127 127 Hackl, E., Bachmann G., and S. Zechmeister-Bolte rnstern. 2004. Microb ial nitrogen turnover in soils under different types of natural forest. Forest Ecology and Management. 188:101 112. Hackl, E., Pfeffer M., Donat, C., Bachmann G ., and S. Zechmeister-Bolternstern. 2005. Composition of the microbial communities in the mineral soil under different types of natural forest. Soil Biology and Biochemistry 37:661-671. Haines, T.K., Busby, R.L. and D.A. Cleaves. 2001. Prescribed Burning in the South: Trends, Purpose, and Barriers. Southern Journal of Applied Forestry 25:149-153. Haney, R. L., Franzluebbers, A. J., Hons, F. M ., Hossner L. R., and D. A. Zuberer. 2001. Molar concentration of K2SO4 and soil pH affect estima tion of extractable C with chloroform fumigationextraction. So il Biology and Biochemistry. 33:1501-1507. Haney, R. L., Senseman, S. A., and F. M. Hons 2002. Effect of Roundup Ultra on Microbial Activity and Biomass from Selected Soils. Journal of Environmental Quality 31:730-735. Harris J.A. 2003. Measurements of the soil microbial community for estimating the success of restoration. European Journal of Soil Science 54:801-808. Harm s, W.R., Aust W.M., and J. A. Burger. 1998. Wet flatwoods. In : M.G. Messina and WH Conner (ed.) Southern Forested Wetlands: Ecology and Management. Lewis Publ. Boca Raton, FL., NY. 422-444. Harris, R.R. 1999. Defining Reference C onditions for Restoration of Riparian Plant Communities: Examples from California, USA. Environmental Management 24:55-63. Hart, S.C., DeLuca H.T., Newman G.S., MacKenzi e M.D. and S. I. Boyle. 2003. Post-fire vegetative dynamics as driver s of microbial community stru cture and function in forest soils. Forest Ecology and Management 220:166-184. Haywood, J. et al. 2001. Vegetative Responses to 37 years of Seasonal Burning on a Louisiana Longleaf Pine Site. Southern Jour nal of Applied Forestry 25:122-130. Haywood, J. D. 2000. Mulch and hexazinone herbicid e shorten the tim e longleaf pine seedlings are in the grass stage and increases height growth. New Forests 19:279-290. Heady, R.B. and J.L. Lucas. 2004. PERMAP 11.3 Operation Manual for Multidimensional Scaling Computer Program. University of Louisiana at Lafayette. 75 p. Heilmann-Clausen, J., and M. Christensen 2004 Does size matter? On the importance of various dead wood fractions for fungal diversity in Danish be ech forests. Forest Ecology and Management 201:105-117. Hedman, C.W., Grace, S.L., and S.E. King 2000. Vegetation composition and structure of southern coastal plain pine forests: an ecological comp arison. Forest Ecology and Management 137:126-238. PAGE 128 128 Hendricks J.J., Mitchell, R.J. Kuehn, K.A. Pecot, S.D. and, S.E. Sims 2006. Measuring external mycelia production of ectom ycorrhizal fungi in the field: the soil matrix matters. New Phytologist 171:179-186. Hersperger, A.M., and T. Koellner 2004. Rarefaction method for assessing plant species diversity on a regional scal e. Ecography 27:532-544. Hobbs, R. J., and J.A. Harris. 2001. Restora tion Ecology: Repairing the Earths Ecosystems in the New Millennium. Restoration Ecology. 9:239-246 Horwath, W.R., and E. A. Paul. 1994. Microbia l Biomass. In: Methods in Soil Analysis Part 2 Microbiological and Biochemical Prope rties (2nd Edition). Soil Science Society American. 649 p. Huck, R.B. 1987. Plant communities along ed aphic continuum in a Central Florida watershed. Florida Science 50:111-128. Hurlbert, S. H. 1971. The nonconcept of sp ecies diversity: a cri tique and alternative parameters. Ecology 52:577-586. Huygens, D., Rutting, T., Boeckx, P., Van Cleemput, O., Godoy, R. and C. Muller. 2007. Soil nitrogen conservation mechanisms in a pr istine south Chilean Nothofagus forest ecosystem. Soil Biology and Biochemistry 39:2448-2458. Hyde, A.G., Law, L. Jr., Weatherspoon, R.L., Cheyney M.D., and J.J. Eckenrode. 1977. Soil Survey of Hernando County, Florida. Natural Resource Conservation Service. U. S. Department of Agriculture. 152 p. Ittig, P.T. 2004. Comparison of Efficient Seasonal Indexes. Journal of Applied Mathematical Decisional Science 8:87-105. Izak, J. and L. Papp. 2000. A link between ecological diversity indi ces and measures of biodiversity. Ecologica l Modelling 130:151-156. Johnston, J.M., and D.A. Crossley. 2002. Forest ecosystem recovery in the southeast US: soil ecology as an essential component of ecosy stem management. Forest Ecology and Management 155:187-203. Jokela, E.J. and AJ. Long 1999. Using Soils to Guide Fertilizer Recommendations for Southern Pines. Publ. FRO-53. IFAS Coope rative Extension Service University of Florida, Gainesville. 10 p. Jorgensen, S.E. and G. Bendoricchio 2001. Fundamentals of Ecological Modelling (3rd Edition) Elsevier Science Ltd. Oxford, UK 530 p. Jorgensen, S.E., 1999. State-of-the-art of ecological modeling with emphasis on development of structural dynamic models. Ecological Modelling 120:75-96. PAGE 129 129 Jules, M.J., Sawyer, J.O. and E.S. Jules. 2008. Assessing the relationships between stand development and understory vegetation using a 420-year chronosequence. Forest Ecology and Management (Articles in Press). Keeney, D. R., and D.W. Nelson. 1982. Nitr ogen-Inorganic Forms. In: Methods in Soil Analysis, Part 2. Chemical a nd Microbiological Properties (2nd Edition) 649 p. Kelting, D.L., Burger, J. A., Patterson S.C., Au st, W.M., Miwa, M., and C.C. Trettin. 1999. Soil quality assessment in domestic forests a southern example. Forest Ecology and Management 122:167-185. Kirkman L. K., Coffey K. L., Mitchell R. J ., and E. B. Drew. 2004. Ground cover recovery patterns and life-history traits: implications for restoration ob stacles and opportunities in a species-rich savanna. Jour nal of Ecology 92:409-421. Kirkman L. K., Mitchell R., Helton R.C., a nd M. Drew. 2001. Productivity and Species Richness Across An Environmental Gradient In A Fire-Dependent Ecosystem. American Journal of Botany 88:2119. Kirkman L. K. 1995. Impacts of Fire and Hydrological Regimes on Vegetation in Depression Wetlands of Southeastern USA. Tall Timbers Fire Ecology-Fire in Wetlands: A Management Perspective Conference Proceedings 19:10-20. Tall Timbers Research Station, Tallahassee, FL. Koellner, T. and A.M. Hersperger. 2004. Ra refaction method for assessing plant species diversity on a regional sc ale. Ecography 27:532. Korb, J. E., Johnson, N.C., and W.W. Covingt on. 2003. Arbuscular mycorrhizal propagule densities respond rapidly to ponderosa pine re storation treatments. Journal of Applied Ecology 40:101. Kovach, W. L., 1999. MVSPA mul tivariate statistical package for windows, ver. 3.1. Kovach Computing Services, Pentraeth, Wales, UK. Krankina, O.N., Houghton, R.A., Harmon, M. E., Hog, E.H., Butman, D., and G. Spycher. Effects of climate, disturbance, and species on forest biomass across Russia. Canadian Journal of Forest Research 35:2281-2293. Kuai, L., Nair, A.A., and M. F. Polz. 2001. Rapid and Simple Method for the Most-ProbableNumber Estimation of ArsenicReducing Bacteria. Applied Environmental Microbiology 67:3168. Kumar, V., Steinbach, M., Tan, P., Klooster, S., Potter, C., and A. Torregrosa. Mining scientific data: Discovery of patterns in the global climate system In Proceedings of the Joint Statistical Meetings (Athens, GA, Aug. 5 --9). American Statistical Association, Alexandria, VA, 2001. 8 p. PAGE 130 130 LaSalle, M.W. 2002. Recogni zing Wetlands in the Gulf of Mexico Region: Regulatory Definition of Wetlands. Publication # 2157 Extension Service of Mississippi State University 25 p. Langeland, K., Netherland, M., Haller, W., and T. Koschnick. 2006. Efficacy of Herbicide Active Ingredients Against Aquatic Weeds. Publ. SS-AGR-44. IFAS Cooperative Extension Service University of Florida, Gainesville. 4 p. Lindenmayer, D.B. 1999. Future directions fo r biodiversity conservatio n in managed forests: indicator species, impact studies and mon itoring programs. Forest Ecology and Management 115:277-287. Lister, T.W. 2003. Forest Harvesting Distur bance and Site Prepar ation Effects on Soil Processes and Vegetation in a Young Pine Plan tation. Master Thesis. Virginia Polytechnic Institute and State Univer sity. Blacksburg. 104 p. Lockaby, B.G., and M. R. Walbridge. 1998. Biogeochemistry: In : M.G. Messina and WH Conner (ed.) Southern Forested Wetlands: Ecology and Management. Lewis Publ. Boca Raton, FL., NY. 149-171. Lugo A. E. 1995. Fire and Wetland Manage ment. Tall Timbers Fire Ecology-Fire in Wetlands: A Management Perspective Conference Proceedings 19: 1-9. Tall Timbers Research Station, Tallahassee, FL. Martin, T.A. and E.J. Jokela. 2004. Stand de velopment and production dynamics of loblolly pine under a range of cultural treatments in nor th-central Florida USA. Forest Ecology and Management 192:39-58. McCune, B., and M. J. Meffrod. 1999. PC-ORD Multivariate Analysis of Ecological Data. Version 4.0. MjM software, Gleneden Beach, OR. McNab, W. H. and P. E. Avers. 1994. Ecologi cal Subregions of the United States, section descriptions: Florida Coastal Lowlands (Western). U. S. D. A. Forest Service. Administrative Publication WO-WSA-5 267 p. Messina, M.G., and W.H. Conner. 1998. Southern Forested Wetlands: Ecology and Management. Lewis Publishers, New York. 616 p. Michael, J.L., Batzer, D.P., Fisher, J.B., and H.L. Gibbs. 2006. Fate of herbicide sulfometuron methyl (Oust) and effects on invertebrates in drainages of an intensively managed plantation. Canadian Journal of Forest Research 36:2497-2504. Michener, W.K. 1999. Wetland Restoration: Still More Chance Than Science? (book review). Ecology 80:2132-2133. Montgomery, H. J., Monreal, C.M., Young, J.C., and K.A. Seifert. 2000. Determination of soil fungal biomass from soil ergosterol analysis Soil Biology and Biochemistry 32:1207 1217. PAGE 131 131 Morris, S.J., and R.E.J. Boerner, 1998. Intera ctive influences of silv icultural management and soil chemistry upon soil microbial abundance and n itrogen mineralization. Forest Ecology and Management 103:129-139. Mller, F., and R. Lenz, 2006. Ecological in dicators: Theoretical funda mentals of consistent applications in environmental manage ment. Ecological Indicators 6:1-5. Mller, F., Hoffmann-Kroll, R. and H. Wiggering 2000. Indicating ecosystem integrity theoretical concepts and environmental requ irements. Ecologi cal Modelling 130:13-23. Mller, F. 1998. Gradients in ecological systems. Ecologi cal Modelling 108:3-21 Murphy, S. 1999. Eight Questions for Ecological Restorationists. Alternative Journal 25:19-20. Myers, R.L. and J.J. Ewel 1990. Ecosystems of Fl orida. University of Central Florida Press. Orlando. 765 p. Myster, R. W. 2001. What is Ecosystem Stru cture ? Caribbean Journal of Science 37:131134. NCSSF. 2005. Science, Biodiversity, and Sustainable Forestry: A Findings Report of the National Commission on Science for Sustainabl e Forestry (NCSSF) Washington D.C. 22 p. Nelson, W., Mehlich, A. and E. Winters 1953. The development, evaluation, and use of soil tests for phosphorus availability. Agronomy 4:153-188. Obreza, T.A. and G. W. Hurt. 2006. 1998. Soil Ra tings For Selecting Pesticides For Water Quality Goals. Circular 959, Soil and Water Science Department, Cooperative Extension Service University of Florida, Gainesville, Florida. 5 p. Odum, E.P. 1969. The strategy of ecosystem development. Science 164:262. Ohtonen, R. and H. Vare. 1998. Vegetation Com position Determines Microbial Activities in a Boreal Forest Soil. Microbial Ecology 36:328. Oliver, C.D. 1981. Forest development in North America following major disturbances. Forest Ecology and Management 3:153-168. Outcalt, K. W., 2000. The longleaf pine ecosyst em of the south. Native Plants Journal 1:43-51. Overing, J.D. and F.C. Watts. 1989. So il Survey of Walton County, Florida Natural Resource Conservation Service. U. S. De partment of Agriculture. 234p. Paavolainen, L. and Smolander, A., 1998. Nitrifica tion and denitrification in soil from a clear-cut norway spruce (Picea abies) stand. S oil Biology & Biochemistry 30:775. PAGE 132 132 Palik, B. J., Mitchell, R.J., and J.K. Hiers. 2002. Modeling silv iculture after natural disturbance to sustain biodivers ity in the longleaf pine ( Pinus palustris ) ecosystem: balancing complexity and implementation. Forest Ecology and Management 155:347356. Palmer, M. W. 1993. Putting things in even be tter order: The adva ntages of canonical correspondence analysis. Ecology 74:2215-2230. Parker, A. J. and J. A. Hamrick. 1996. Genetic Variation in sand pine ( Pinus clausa ). Canadian Journal of Forest Research 26:244-254. Peacock, A.D., MacNaughton, S.J. Cantu, J.M., Dale, V.H. and D.C. White. 2001. Soil microbial biomass and community compos ition along an anthropogenic disturbance gradient within a long leaf pine habita t. Ecological Indicators 1:113-121. Peet, R.K. and D.J. Allard 1993. Longleaf Pine Vegetation of the Southern Alantic and Eastern Gulf Coast Regions: A Preliminary Classification. In: Proceedings of the 18th Tall Timbers Fire Ecology Conference. The longl eaf pine ecosystem: ecology, restoration, and management. 1991 Tallahassee, Fl Tall Timbers Research Station 18:45-81. Pennanen, T., Liski, J., Baath, E., Kitunen, V., Uotil a, J., Westman, C.J., and H. Fritze. 1999. Structure of the Microbial Communities in Coni ferous Forest Soils in Relation to Site Fertility and Stand Developmen t. Microbial Ecology 38:168-179. Perez, C.A., Hedin L.O., and J.J. Armesto. 1998. Nitrogen Mineraliza tion in Two Unpolluted Old-Growth Forests of Cont rasting Biodiversity and Dyna mics. Ecosystems 1:361. Perry, D. A. 1994. Forest ecosystems. Baltim ore. Johns Hopkins University Press, Baltimore. 649 p. Peterson, C.H. 1976. Measurement of community pattern by indices of local segregation and species diversity. Jour nal of Ecology 64:157. Platt, T. and K.L. Denman. 1975. Spectral An alysis in Ecology. Annual Review of Ecological Systems 6:189-210. Pianka, E. R. 1966. Latitudinal gradients in sp ecies diversity: A review of concepts. The American Naturalist 100:33-46. Pickle, J. 1999. Microbial biomass and ni trate immobilization in a multi-species riparian buffer. Masters Thesis. University of Iowa, Ames, Iowa. Pickett, S. T. A. 1989. Space-for-time substituti on as an alternative to long-term studies, in: Longterm studies in ecology, edited by: Likens G. E., New York, Sp ringer Verlag New York. 110. Pickett, S.T.A. and M.L. Cadenasso. 1995. Landscape Ecology: Spatial Heterogeneity in Ecological Systems. Science 269:331-334. PAGE 133 133 Pickett, S.T.A., S.L. Collins and J.J. Armesto. 1987. A hierarchical consideration of causes and mechanisms of succession. Plant Ecology 69:109-114. Pickett, S.T.A. and White, P.S., 1985. The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, FL. Poole, R.W. 1974. An Introduction to Quantitati ve Ecology. McGraw-Hill New York N.Y. 532 p. Pywell, R.F., Bullock, J.M., Roy, D.B. Warman, L. Walker, K.J., and P. Rothery. 2003. Plant traits as predictors of performance in ecologi cal restoration. Jour nal of Applied Ecology 40:65-77. Qingchao, Li., Allen, H.L., and C.A. Wilson. 2003. Nitrogen Mineralization dynamics following the establishment of a loblolly pine plantation. Canadian Journal of Forest Research 33:364-374. Ramsey, C.L., Jose, S., Brecke, B.J., and S. Merritt. 2003. Growth response of longleaf pine (Pinus palustris Mill.) seedlings to fertilization and herbaceous weed control in an old field in southern USA. Forest Ecology and Management 172:281-289. Ranasinghe, S. 2003. Role of Herbicides in Longleaf Pine Flatwoods Restoration: Pine Growth, Understory Vegetation Response and th e Fate of Applied Herbicides. Master Thesis. University of Florid a. Gainesville. 64 p. Redding, T., Hope G.D., Schmidt, M.G., and M.J. Fortin. 2004. Analytical methods for defining stand-clearcut edge effects demonstrated for nitrogen mineralization. Canadian Journal of Forest Research 34:1018-1024. Reinman, J.P. 1985. A Survey of the Understo ry Vegetation Communities of the St. Marks National Wildlife Refuge. Technical Repor t # 41640-79-1. U. S. Fish and Wildlife Service. 89 p. Reynolds, P.E., Thevathasan, N.V., Simpson, J.A., Gordon A.M., Lautenschlager, R.A., Bell, W.F., Gresh, D.A., and D.A. Buckley. 2000. Alternative conifer release treatments affect microclimate and soil nitrogen mineralizat ion. Forest Ecology and Management 133:115125. Richards, B.N. 1987. THE MICROBIO LOGY OF TERRESTRIAL ECOSYSTEMS. University of New England Austrailia John Wiley & Sons NY. 399p. Richardson, D. R. 1985. Allelopathy and fire in the Florida scr ub. American Journal of Botany 72:864-865. Sandberg, D.V., R.D. Ottmar, and G.H. Cu shon. 2001. Characterizing fuels in the 21st century. International Journal of Wildland Fire 10:381-387. PAGE 134 134 SAS (Statistical Analysis System) Institute Inc. 2002. Version 8.2. SA S Institute Inc., Cary, NC. Schmidt, E.L., and L.W. Belser. 1982. N itrifying Bacteria: Enumer ation by Most Probable Number. In: Methods in Soil Analysis, Pa rt2. Chemical and Microbiological Properties (2nd Edition) 1027. Seiter, S., Ingham, E.R., and R.D. William 1999. Dynami cs of soil fungal and bacterial biomass in a temperate climate alley cr opping system. Applied Soil Ecology 12:139-147. SER. 2004 The SER International Primer on Ecological Restoration. Science & Policy Working Group, Society for Ec ological Restoration. www.ser.org/ Sigg, J. 1999. California Exotic Pl ant Pest Council News, Summ er/Fall 10-13. Silver, W.L., Herman, D.J. and M.J. Firestone 2001. Dissimilatory Nitrate Reduction to Ammonium in Upland Tropical Fore st Soils. Ecology 82:2410-2416. Simpson, T.B. 2005. Ecological Restorati on and Re-Understanding Ecological Time. Ecological Restoration 23:46-51. Slesak, R.A. 2007. Control of Competing Vegetation Following Harvesting: Effects on Nitrogen Leaching, Mineralization, and Douglas-fir Foliar Status In: Proceedings of the ASA-CSSA-SSSA International Annual Meetings A Century of Integrating Crops, Soils & Environment. New Orleans, La. Smith, O.L. 1982. Soil Microbiology: A model of Decomposition and Nutrient Cycling. CRC Press 273 p. Smith, R.S., Shiel, R.S., Bardgett, R.D., Milkwa rd, D., Corkhill, P., Rolph, G., Hobbs, P.J., and S. Peacock. 2003. Soil microbial comm unity, fertility, vegeta tion and diversity as targets in restoration management of a meadow grassland. Journal of Applied Ecology 40:51-64. Spencer, S. 2004. Plant Communities of Chassa howitzka Wildlife Management Area. Division of Wildlife and Florida Natural Areas Invent ory, Florida Fish and Wildlife Conservation Commission. 31 p. Spetich, M.A., Shifley, S.R, and G.R. Parker. 1988. Regional Distribution and Dynamics of Coarse Woody Debris in Midwestern Old-Gr owth Forests. Forest Science 45:303-313. Spies, T.A. and S.P. Cline. 1999. Coarse w oody debris in forests and plantations of coastal Oregon. P. 5-24 In : From the forest to the sea: A story of fallen trees. Maser, C.M. et al. (eds.). U.S. Department of Agriculture, Fo rest Service General Technical Report PNW229. PAGE 135 135 Spyreas, G., and J.W. Mathews. 2006. Floristic conservation value, ne sted understory floras, and the development of second growth fore sts. Ecological Applications 16:1351-1366. Stanturf, J.A., Schweitzer, C.J., Schoenholtz, S. H., and J.P. Shepard. 2001. Achieving restoration success: myths in bottomland hard wood forests. Restor ation Ecology 9:189200. Steiner, M., Linkov, L. and S. Yosida 2001. The role of fungi in the tr ansfer and cycling of radionuclides in forest ecosystems. Journa l of Environmental Ra dioactivity 58:217-241. Stevens, R. J., Laughlin, R.J., and J. P. Malone. 1998. Soil pH affects the processes reducing nitrate to nitrous oxide and di-nitroge n. Soil Biology and Biochemistry 30:1119-1126. Stout, I. J., and W. R. Marion. 1993. Pine flatwo ods and xeric pine forests of the southern (lower) coastal plain. In: Mart in, W. H., S. G. Boyce and A. C. Echternacht. Ed. Biodiversity of the Southeastern United States, Lowland Terrestrial Communities. John Wiley and Sons, Inc. New York. 373-446. Strauss, R.E. 1982. Statistical significance of sp ecies clusters in associ ation analysis. Ecology 63:634-639. Subba Rao, N.S. 1995. Soil Microorganisms and Plant Growth (3rd Ed.). Science Publishers 335 p. Sylvia, D.M. and A.G. Jarstfer. 1997. Distribution of mycorrhiz a on competing pines and weeds in a southern pine plan tation. Soil Science Society Amer ican Journal. 61:139-144. Sylvia, D.M., Hartel, P., Fuhrmann, J., and D. Zube rer. 1998. Principles and Applications of Soil Microbiology. Prentice Hall 500 p. Taylor, L.A., Arthur, M.A., and R. D. Yanai. 1999. Forest floor microbial biomass across a northern hardwood successional sequence. Soil Biology and Biochemistry 31:431-439. Ter Braak, C. J. F. 1994. Canonical community ordination. Part I: Basic theory and linear methods. Ecoscience 1:127-140. Vance, N.C., and J. A. Entry 2000. Soil propertie s important to the restoration of Shasta red fir barrens in the Siskiyou Mount ains. Forest Ecology a nd Management 138:427-434. Vance, E.D., Brookes, P.C., and D. S. Jenkinson 1987. An extraction method for measuring soil microbial biomass carbon. Soil Bi ology and Biochemistry 19:703-707. Van Lear, D.H., Carroll, W.D., Kapeluch, P.R., a nd R. Johnson 2005. History and restoration of longleaf pine-grassland ecosyst em: Implications for species at risk. Forest Ecology and Management 211:150-165. PAGE 136 136 Veldkamp, E., Davidson, E., Erickson, H., Keller, M., and A. Weitz. 1999. Soil nitrogen cycling and nitrogen oxide emissions along a pasture chronosequence in the humid tropics of Costa Rica. Soil Biology and Biochemistry 31:387-394. Vienneau, D.M., Sullivan, C.A., House, S.K., a nd G. W. Stratton. 2004. Effects of the herbicide hexazinone on nutrient cycling in a low-pH bluebe rry soil. Environmental Toxicology 19:115-122. Vitousek, P.M. and W. A. Reiners. 1975. Ecosystem Succession and Nutrient Retention: A Hypothesis. Bi oscience 25:376-381. Wade, D. et al. 2000. Chapter 4: Fire in Ea stern Ecosystems. U.S.D.A. Forest Service, General Technical Report RMRS-GTR-42. 2:53-96. Walker, J. 1993. Rare Vascular Plant Taxa A ssociated with the Longleaf Pine Ecosystems: Patterns in Taxomony and Ecology. In: Herma nn, S. M. (ed), Proc. Tall Timbers Fire Ecol. Conf. Vol. 18. Tall Timbers Research Station, Tallahassee, FL 105-130. Walker, J. and A.M. Silletti. 2006. Restoring the ground layer of longleaf pine ecosystems. Pages 297-333 in Jose, S., Jokela, E., and Miller, D.L., editors. The Longleaf Pine Ecosystem: Ecology, Silviculture, and Restor ation. Springer Science + Business Media, LLC., New York. Walkley A. 1947. A critical examination of a ra pid method for determining organic carbon in soils. Effects of variations in digestion conditions and inorganic soil constituents. Soil Science 63:251-263. Wallander, H., Nilsson, L.O., Hagerberg, D., and E. Baath 2001. Estimation of the biomass and seasonal growth of external mycelium of ectomycorrhizal fungi in the field. New Phytologist 151:753. Wenger, K. F. 1984. Forestry Handbook (second ed ition). Society of American Foresters John Wiley and Sons New York 1335 p. Whitcomb, C. E. 1999. An introduction to ALS-inhibiting herbic ides. Toxicology and Industrial Health 15:232-240. White, B. A. et al. 2000. Topsail Hill Preserve State Park Unit Management Plan. Division of Recreation and Parks, Florida Department of Environmental Protection. 104 p. White, L.L., Zak, D.R. and B.V. Barnes. 2004. Biomass accumulation and soil nitrogen availability in an 87-year-old Populus grandidentata chronosequence. Forest Ecology and Management 191:121-127. Wildlife Management Institute (WMI). 2003. Regional Wildlife Habitat Needs Assessment for the 2007 Farm Bill: A Summary of Successes an d Needs of the Farm Bill Conservation Programs. WMI, Washington, D.C. PAGE 137 137 Williamson, W.M., Wardle, D.A., and G.W. Yeat es. 2005. Changes in soil microbial and nematode communities during ecosystem declin e across a long-term chronosequence. Soil Biology and Biochemistry 37:1289-1301. Wilson, C.A., Mitchell, R.J., Boring, L.R., and J.J. Hendricks. 2002. Soil nitrogen dynamics in a fire-maintained forest ecosystem: results ov er a 3-year burn in terval. Soil Biology and Biochemistry 34:679-689. Yang. Y., Watanabe, M., Li, F., Zh ang, J., Zhang, W., and J. Zhai. 2006. Factors affecting forest growth and possible effects of climate change in the Taihang Mountains in northern China. Forestry (Oxf ord Journals) 71:135-147. Yin. B., Crowley, D., Sparovek, G., Jose De Me llo, W., and J. Bornem an. 2000. Bacterial Functional Redundancy along a Soil Reclamati on Gradient. Applied Environmental Microbiology 66:4361-4365. Zak, D.R., Grigal, D.F. Gleeson, S. and D. Tilman 1990. Carbon and nitrogen cycling during old field succession: Constraints on plant a nd microbial biomass. Biogeochemistry 11:111-129. Zas, R. and M. Alonso 2002. Understory vegetation as indicators of soil characteristics in northwest Spain. Forest Ecol ogy and Management 171:101-111. PAGE 138 138 BIOGRAPHICAL SKETCH George McCaskill s tarted his doctoral study at the School of Forest Resources and Conservation, University of Florida in January, 2003. Before joining the University of Florida, he was Associate Faculty teaching multiple course s at the College of the Redwoods in Eureka, California. Prior to that he worked as a Bili ngual Forestry instructor at Mt. Hood Community College. For three years, he was a State Lands Timber Sales Forester for the Washington Department of Natural Resources. He spent 3.5 years in the U.S. Peace Corps serving as an Environmental Program Specialis t evaluating Chilean forest pr actices as applied to their Monterey pine plantations and their native Nothof agus forests. While working with the Chilean Forestry Corporation, he served as interprete r/translator/editor duri ng the Sixth Congress on Criteria and Indicators for th e Conservation and Sustainable Management of Temperate and Boreal Forests. Also known as the Montral Prot ocol, he helped to finalize the treaty where all the Pacific Rim countries signed the document. In 1990, Mr. McCaskill completed his Masters program at California Polytechnic in San Lu is Obispo, California. He is a Registered Professional Forester in California.