<%BANNER%>

Interactions between Plants and Soil Microbes in Florida Communities: Implications for Invasion and Ecosystem Ecology


PAGE 1

INTERACTIONS BETWEEN PLANTS AND SOIL MICROBES IN FLORIDA COMMUNITIES: IMPLICATIONS FO R INVASION AND ECOSYSTEM ECOLOGY By SARAH RENEE BRAY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Sarah Renee Bray

PAGE 3

iii ACKNOWLEDGMENTS Many people and organizations have supporte d my research and this dissertation would not have completed without their help. Kaoru Kitajima, my advisor, has allowed me to pursue my own research interests ev en as they diverged from her own. My committee members, Alison Fox, Doria Gordon, Michelle Mack, and David Sylvia, and the Plant Ecology Group have been wonderful sounding boards, colleagues, and editors. Michelle Burch, Robert Bilba o, Jason Alexander, Debbie Rene lus, Jennie DeMarco, Abid Al-Agely, Alex Reinstein, Thai Van, Ellen Di ckstein and Grace Crummer have helped in the field and lab. The Botany Department st aff made sure my bills were paid and deadlines were met. The National Science Foundation, the Departme nt of Environmental Protection, the Florida Exotic Pest Plant C ouncil, and the College of Liberal Arts and Sciences have all provided me funding to comp lete my research. The real reason that I have been able to finish my doctorate was the unwavering support of my friends and family. I thank Sara Smith, Melissa “Fig” Bonfig, Amy Mill er Jenkins, Silvia Alvarez, Erika Gubrium, and Aline Gubrium for the conv ersations that encouraged me to continue when I was frustrated. I thank my parents, Steve and Susan Bray, for always encouraging me to achieve my goals and picking me up wh en I fell short. I thank my brother and sister-in-law, Seth and Maureen Bray, for remindi ng me of the fun things in life. Most of all, I thank my husband, Anthony Allen. He spent many hours working in the lab and field with me, supported us financially, and ne ver lost faith in me even when I doubted my abilities. Without him, I could not have finished this dissertation.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ix CHAPTER 1 INTRODUCTION........................................................................................................1 2 MYCORRHIZAE DIFFERENTIALLY ALTER GROWTH, PHYSIOLOGY AND COMPETITIVE ABILITY OF AN INVASIVE SHRUB................................10 Introduction.................................................................................................................10 Materials and Methods...............................................................................................12 Species.................................................................................................................12 Experiment 1: Effects of Soil-P, Li ght and Inoculum Source on Growth and Allocation.........................................................................................................13 Experiment 2: Effects of Mycorrhizae on Seedling Competition.......................16 Statistical Analyses..............................................................................................18 Results........................................................................................................................ .19 Experiment 1: Effect of Light Soil-P and Inoculum Type.................................19 Experiment 2: Effects of Mycorrhizae on Seedling Competition.......................21 Discussion...................................................................................................................22 Effect of Inoculum Source on Ardisia .................................................................22 Competitive Interactions.....................................................................................24 Implications of Effects of Myco rrhizae on Exotic Species Invasion..................25 3 SOIL MICROBIAL COMMUNITY STRUCTURE AND FUNCTION IN FLORIDA PLANT COMMUNITIES PR ONE TO NON-NATIVE PLANT INVASION.................................................................................................................32 Introduction.................................................................................................................32 Methods......................................................................................................................35 Species, Sites, and Sampling...............................................................................35 Microbial Community Compositi on and Nutrient Analysis...............................37

PAGE 5

v Statistical Analyses..............................................................................................39 Results........................................................................................................................ .40 Discussion...................................................................................................................43 Habitat Controls of Microbial Community Composition....................................43 Alteration of Microbial Communities by Invasion.............................................45 Conclusions.........................................................................................................46 4 LINKS BETWEEN LITTER QUAL ITY, DECOMPOSITION AND MICROBIAL COMMUNITY COMP OSITION ON NATIVE AND NONNATIVE PLANT LITTER.........................................................................................60 Introduction.................................................................................................................60 Methods......................................................................................................................63 Litter Collection and Experimental Design.........................................................63 Phospholipid Fatty Acid Analysis.......................................................................65 Statistical Analysis..............................................................................................66 Results........................................................................................................................ .69 Decomposition of Litter......................................................................................69 Microbial Commun ity Com position....................................................................71 Linking Litter Quality, Decompositi on and Microbial Communities.................74 Discussion...................................................................................................................74 Factors Controlling Microbial Community Composition...................................75 Factors Controlling Decomposition Rate............................................................78 Implications for Invasion.....................................................................................79 Conclusions.........................................................................................................80 5 CONCLUSIONS........................................................................................................98 LIST OF REFERENCES.................................................................................................103 BIOGRAPHICAL SKETCH...........................................................................................115

PAGE 6

vi LIST OF TABLES Table page 2-1 ANOVA summarizing the effects of light and soil on LAR and R/S in experiment 1.............................................................................................................28 2-2 Means of leaf area ratio (LAR) and root:shoot ratios (R /S) from all soil-P levels in experiment 1.........................................................................................................28 2-3 Comparison of light saturated net photosynthesis rate (Amax) and dark respiration under moderate vs. low light treatments ...............................................28 2-4 ANOVA summary of the effects of sp ecies, competition, and mycorrhizae on RGR, LAR, R/S, and colonization rates from experiment 1....................................29 3-1 Location, mean annual temperature, mean annual rainfall, soil type and dominant vegetation of the thre e sites in each habitat type......................................48 3-2 Soil characteristics of the five habitats examined....................................................49 3-3 Relationship of total PLFA and fungal:b acterial ratio with % moisture, %C, %N and C:N using quadratic fit for total PLFA and simple linear regression for fungal : biomass ratio...............................................................................................49 3-4 ANOVA results for the effects of habitat, invasion-status and their interaction on the relative representation of Gram – and Gram + PLFAs......................................49 3-5 Correlation between %N, %C, and %moist ure and the relative representation of Gram – and Gram + PLFAs ....................................................................................49 3-6 Results of MANOVA for the effects of habitat, invasion and their interaction on principle component axes 1 and 2 scor es from the ordination of 23 PLFAs...........50 3-7 Loadings for the first two axes in a principle components analysis of 23 common PLFAs extracted from soil samples..........................................................50 3-8 Correlations (r) of soil characteristics with the first two axes from the ordination of 23 PLFAs.............................................................................................................51

PAGE 7

vii 3-9 The results of MANOVA fo r the effects of habitat, i nvasion and their interaction on principle component axes 1 and 2 sc ores derived from an ordination of metabolic activity of soil microbes on 95 substrates................................................51 4-1 Characteristics of plant species whos e litter was used in this experiment...............82 4-2 Mean and SE for decomposition ra tes of native and non-native litters....................84 4-3 Result of simple linear regression fo r log-transformed decomposition constants against individual variables of litter chemical composition.....................................84 4-4 Factor loadings for individual variable s that contribute to litter quality on the leaf chemistry PCA axis 1........................................................................................85 4-5 Factor loadings of individual PLFAs upon the first two main axes of PCA (PC1 and PC2) and CCA (CC1 and CC2).........................................................................85 4-6 MANOVA summarizing the effects of time, litter species and litter species*time on PCA and CCA axes 1 and 2 scores from ordination of all samples....................86 4-7 Loadings for the first two PCA axes for PCA run separately for individual sampling dates (t = 1, 2, and 5)................................................................................86 4-8 Correlation between each lit ter quality variable and th e amount of PLFA (nmol) indicating monounsaturated Gram – bacteria at times 1, 2, and 5...........................87 4-9 Correlation between each lit ter quality variable and th e amount of PLFA (nmol) indicating branched Gram+ fatty acids at times 1, 2, and 5.....................................87 4-10 Correlation between each lit ter quality variable and th e amount of PLFA (nmol) indicating fungal fatty acid (18:2 6c) at times 1, 2, and 5......................................88 4-11 Correlation between each litter quality variable a nd ratio fungal : bacterial PLFAs at times 1, 2, and 5.......................................................................................88 4-12 Correlation between each lit ter quality variable and the total PLFA (nmol) at times 1, 2, and 5.......................................................................................................89 4-13 Correlation between each lit ter quality variable and th e amount of PLFA (nmol) indicating cyclopropyl Grambacteria at time 5......................................................89 4-14 Coefficient of determination (r2) of abundance of PLFAs for microbial functional groups with leaf chemistry scores at times 1, 2, and 5............................90 4-15 Coefficient of determination (r2) of correlations of mi crobial community PCA axes vs.decomposition rate (k), and leaf chemistry axis (from ordination of initial litter quality ) vs. microbial community PCA axes.........................................90

PAGE 8

viii LIST OF FIGURES Figure page 1-1 Conceptual diagram of interactio ns between soil microbes and plants......................9 2-1 Response of Ardisia to light and inoculum type at 5 mg kg-1 P...............................30 2-2 Response of Ardisia and Prunus to heterospecific or conspecific competition and mycorrhizal status..............................................................................................31 3-1 Mean soil microbial community total PLFA (nmol g-1 + S.D.) for five habitats prone to invasion by five non-native species...........................................................52 3-2 Fungal:bacterial ratios (+ S.D.) for fi ve habitats prone to invasion by five nonnative species............................................................................................................53 3-3 Relative representation (% of total nmol es extracted) of PL FAs across habitats and invasion-status...................................................................................................54 3-4 Mean (+ S.D.) principle components sc ores by habitat and invasion-status from ordination of 23 most common microbi al PLFAs found in soil samples.................55 3-5 Mean (+ S.D.) relative abundance of 5 fatty acids with high loadings on the principle component axes.........................................................................................56 3-6 Mean (+ S.D.) relative abundance of PL FAs across habitats and invasion-status...57 3-7 Mean (+ SD) principle components sc ores by habitat and invasion-status from the ordination of metabolism of 95 carbon sources.................................................59 4-1 Mean + SE of the percent mass remaining of 20 species.........................................91 4-2 Mean + SE for nitrogen in litter of 20 plant species at six collection dates.............92 4-3 Ordinations of the 17 most common microbial PLFA found on 11 litter species from three sampling dates........................................................................................93 4-4 PCA and CCA analyses for 17 most co mmon microbial PLFA and litter quality variables run separately for individual sampling dates............................................95

PAGE 9

ix 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 INTERACTIONS BETWEEN PLANTS AND SOIL MICROBES IN FLORIDA COMMUNITIES: IMPLICATIONS FO R INVASION AND ECOSYSTEM ECOLOGY By Sarah Renee Bray August 2005 Chair: Kaoru Kitajima Major Department: Botany Among ecologists there is an increasing awareness and inte rest in the role of soil microbes in the distribution of plants and f unctioning of ecosystems. This dissertation relates soil microbial community compositi on to plant growth, habitat type, and decomposition with particular emphasis on invasive plants. I examined growth, physiology and compe titive ability of an invasive shrub, Ardisia crenata in two greenhouse experiments. When grown singly, relative growth rates (RGR) and leaf area ratio (LAR) were higher for seedlings inoculated with mycorrhizal fungi isolated from Ardisia roots than those inocul ated with single-spore isolates and nonmycorrhizal contro ls. In a second experiment, Ardisia was grown with a conspecific or heterospecific ( Prunus caroliniana ) competitor. While neither identity of competitor nor mycorrhizal status had a great effect on Ardisia growth, Prunus growth was significantly depressed in competition with Ardisia in the absence of mycorrhizal fungi.

PAGE 10

x In chapter 2, I examined soil microbial communities from five different habitats prone to invasion by an invasi ve plant using phospholipid fatty acids (PLFA) and Biolog substrate utilization. Habitat type had the largest effect on microbial community composition. Moisture content of soils a nd, to a lesser extent, carbon and nitrogen contents appeared to be dr iving differences in biomarker PLFAs. Although the largest differences in soil microbial community composition were found among habitats, invasion altered microbial commun ity composition within habitats. In chapter 3, I placed litters of 20 nati ve and non-native plant species of varying decomposability in a common site and quan tified their decompos ition. The composition of microbial communities on 11 of the litters was examined by PLFA at 28, 56 and 238 days. Microbial communities at the early (low moisture) sampling dates were more similar to one another than to the late (h igh moisture) sampling date. In addition to moisture effects, litter quality had a significant effect on microbial community composition. Both decomposition and microbial community composition were correlated with leaf chemistry. The best single pred ictor of decomposition rate was microbial community composition. These results suggest that plant-microbial interactions are important in plant invasion, and explicit examination of a poten tial positive feedback on invasion through the microbial community s hould be further explored.

PAGE 11

1 CHAPTER 1 INTRODUCTION Historically, most ecologists have studied aboveground macroorganisms even though they have long recognized that proces ses and organisms aboveground can modify processes and species distributions be lowground and vice versa (Jenny 1941, Brown 1958, Garrett 1963, Hudson 1968, Janos 1980). Much of this aboveground bias has been due to the difficulty of examining organisms a nd tissues in the soil matrix. Traditionally, culture-based techniques have been used to identify and isolate soil microbes. However, only a small proportion of soil microbes (<1%) ar e believed to be cu lturable (Atlas and Bartha 1998), and many researchers have treated the soil as a “black box.” The advent of new techniques to study belowground microor ganisms using molecular markers has led to a “renaissance” of sorts in the study of interactions of soil microorganisms with aboveground macrophytes, allowing ecologist s to proceed “through a ped darkly” (Coleman 1985). Soil microbes have direct and indirect in teractions with plan ts (Figure 1-1). Symbiotic microbes directly affect the host plan t’s fitness, resulting in alteration of plant species abundance and distribution. Pat hogenic microbes have negative effects on individual plant fitness. Due to pathogens’ high host-spec ificity, these microbes may be responsible for density-dependant plant dist ributions predicted from the Janzen-Connell hypothesis (e.g., Packer and Clay 2000). Mutu alistic bacteria and fungi (e.g., nitrogenfixing bacteria and mycorrhizal fungi) also di rectly interact with their hosts receiving carbon substrates directly from their host and supplying the plant with nutrients. Free-

PAGE 12

2 living saprophytic fungi and bacteria in th e soil interact indire ctly with plants by decomposing their senescent ti ssue and replenishing inorganic soil nutrients. Therefore, whereas mutualistic soil microbes may be a proximal source of nutrients, free-living microbes supply the ultimate source of nutrients. Availability of nutrients in the soil is one of the primary controls of productivity in ecosystems, and competition for those limiting resources is a major control of community composition. Plant species that can draw down the most limiting resource below the level at which other plants can survive (i.e., the pl ant with the lowest R*) should be the best competitor within a given system and will displace other species (Tilman 1982). To become a superior competitor for this limiting resource, the plant must allocate a greater proportion of its biomass to the acquisition, c onservation, and/or efficient use of that resource. Such biomass investment limits th e ability of the plant to compete for other limiting resources. The resultant trade-off m eans that species with different allocation patterns or suites of functional traits rela ted to resource acquisition will be superior competitors based upon resource availability in the habitat (Chapin 1980, Tilman 1988, Chapin and Aerts 2000). Competition for resources by plants may be modified by interactions with freeliving and symbiotic microbes. Mycorrhizal fungi increase the volume of soil a plant can access, increasing their ability to acquire phos phorus. Increasing the mycorrhizal fungal richness leads to a greater use of soil phosphorus by the plant community, greater productivity and plant species richness (van der Heijden et al. 1998). Not all species of mycorrhizal fungi benefit all plants equally, however, as carbon cost of some fungal partners exceeds the benefit in increased phosphorus nutrition (Bethlenfalvay et al. 1982,

PAGE 13

3 Smith and Smith 1996, Johnson et al. 1997, Graham and Eissenstat 1998). Such differential response of plants to the species of mycorrhizal fungi pr esent may modify the competitive hierarchy (Moora and Zobel 1996, Smith et al. 1999). If dominant, superior competitors harbor fungi para sitic to themselves, their competitive ability should be decreased. Work by Bever et al. (1996) has, in fact, shown that there are host-specific sporulation rates of mycorrhiz al fungi. They found that these differential rates of sporulation maintained diversity through nega tive feedback on domi nant plants as the fungus that preferentially sporul ated with the dominant plant was also least beneficial to it (Bever 2002). Conversely, dominance ma y be achieved through the presence of mutualistic microbes or the absence or resist ance to pathogenic microbes (Bever et al. 1997). While studies of mycorrhizal fungi have demonstrated that in teractions between fungi and plants influence the distributi on and abundance of both groups, our knowledge and understanding of the interact ions between plants and free-living microbes are more limited. Because of the difficulty in examin ing non-culturable soil microbes, ecologists have generally left this community as a “black box,” examining the effect of soil communities on plant growth without identifying the agents (pathogenic, mutualistic or free-living) responsible. In these types of studies, plants are grown in a soil for an extended period of time to allow for the soil community differentiation over time due to plant inputs. Plants are then grown in thei r own or another species’ soil. Some studies have shown that plant growth is higher in their own soil (a “home-soil” advantage) while others have shown the reverse (Bever 1994, Ca llaway et al. 2004). This would suggest

PAGE 14

4 the possibility for negative or positive regula tion of plant species density through the soil community although those agents respons ible for the regulation are unknown. There has been an increased interest in moving away from the “black box” and examining what free-living microbes are actua lly responding to different plant species and what effects changes in free-living micr obes might be. The advent of molecular techniques for examining soil microbes, in pa rticular the use of phospholipid fatty acids (PLFA) and nucleic acids, has resulted in a large number of studies of soil microbial community composition (Tiedje et al. 1999). Techniques based upon DNA and RNA have greater power in distinguishing micr obial species than do PLFA techniques. However, due to the specificity of primers and probes used, nucleic acid based studies tend to examine a phylogentic subset of soil mi crobes. Additionally, as all nucleic acid techniques are dependant on PCR, only pr esence/absence, rather than quantitative analysis, is possible. While PLFAs are lim ited in their ability to identify species, and thus can never truly answer questions about diversity, they can be used to identify different functional groups of microbe s and do allow for a broad, quantitative examination of the entire soil microbial co mmunity. Thus the use of both types of techniques will help to advance understan ding of soil microbial community composition. The field of soil ecology is still very much in an exploratory phase; however, some trends are beginning to emerge Different environmental factors such as temperature, moisture, and soil nutrients, and different plant species support different microbial communities (Bossio and Scow 1998, Bossio et al. 1998, Staddon et al. 1998, Pennanen et al. 1999, Priha and Smolander 1999, Myers et al. 2001). Less research has been directed at the effects of free-living microbes on the plant community. Because microbes

PAGE 15

5 decompose organic matter, releaing inorganic nutrients, microbial influence on nutrient availability may result in plant community changes. As decomposition rates and microbial communities are known to diffe r among plant communities, microbial community composition may be ultimately re sponsible for differences in decomposition rate. These various plant-microbial interactions have for the most part been documented in systems with well-established associat ions of plant species. Plant community composition and diversity, however, are changi ng at global and local scales. Plant invasions, in effect, offer a “natural” experi ment in which to examine the effect of changing species composition on plant-microbial interactions. Plant invasions also offer a chance to examine the interact ion of native microbes and nonnative plants that likely do not share an evolutionary history. The goal of this dissertation is to use the natural experiment of non-native plant invasion to further our unde rstanding of plant controls on symbiotic and free-living microbial community composition and vice versa. Such studie s will also shed light on the functional reason for dominance of invasive plants. My disserta tion is divided into three sections examining the interaction of myco rrhizal fungi with an invasive shrub, differences between soil microbial communities from different habitats with and without plant invasion, and the links between nativ e and non-native plant litter, microbial community composition and decomposition. In Chapter 2, I study the effect of native mycorrhizal fungi on the growth, physiology and competitive ability of an invasive shrub Ardisia crenata in two experiments. I hypothesized that Ardisia ’s successful invasion may be in part due to its

PAGE 16

6 ability to form beneficial partnerships with native mycorrhizal f ungi it encounters in the areas it invades. This was tested in a greenhouse experiment in which Ardisia was grown in sterile soil, with one of two native single fungal species, or with a mix of mycorrhizal fungi isolated from Ardisia roots. I also hypothesized that Ardisia performs better in heterospecific competition than conspecific co mpetition especially when mycorrhizae are present to modify resource competition. This was tested in a second greenhouse experiment in which Ardisia was grown with a single conspecific or a native heterospecific competitor, Prunus carolinana with or without mycorrhizae. As the results from Ch apter 2 indicated that Ardisia ’s growth was improved by the presence of mycorrhizal fungi found natura lly occurring in its roots, I wanted to determine if invasion could result in the alte ration of the soil micr obial community. If invaders alter the composition of the soil microbial community to their advantage, there might be potential for positive feedback on th e invasion. In a natural survey of five habitats prone to invasion by non-native plan ts, I sought to determine if plant invasion predictably alters soil microbial community composition and function as examined with phospholipid fatty acids (PLFA) and Biolog substrate utilization. Three alternative hypotheses could explain soil microbial commun ity composition at the landscape level. One, habitat characteristics such as soil type, soil nutrients and moisture contents, temperature, and native plan t community composition could have such a strong control on soil microbial community composition that even dense invasions may not result in a change in soil microbial community compos ition. Alternatively, dense invasions may alter the microenvironment experienced by so il microbes enough to alter the composition of their communities. There may also be an intermediate hypothesis where some, but not

PAGE 17

7 all invaders alter soil microbial commun ity composition as a result of functional characteristics of the invader. The results of Chapter 3 suggested that wh ile there were larger differences between habitats than within habitats due to in vasion status, invasion does alter microbial community composition. The final chapter (Cha pter 4), I sought to determine if the chemical composition of litter determines the microbial community present and how microbial community compositi on is related to decompositi on. I hypothesized that more chemically recalcitrant litter (i .e., litter high in lignin, C:N, and low N:lignin) would have a higher proportion of fungi and would decompos e more slowly. Litter with highly labile chemistry (i.e., litter high in non-fiber carbon fractions and nitrogen and low in lignin) would decompose quickly as a result of hi gher total microbial biomass and a higher proportion of bacteria. To test these hypotheses, I created a “common garden” experiment in which litter of varying litter quality from native and non-native plant species were allowed to decompose for the period of one year. The composition of microbes on the plant litter was examined for 11 of the species at 28, 57, and 238 days. Composition of microbial communities varied with time; however, microbial communities on leaf litter of similar quality were more similar to one another than to microbial communities on litter of very diffe rent quality. I also found the microbial community composition was the best predicto r of decomposition rate of all factors examined. Together these studies begin to paint a picture of how plant and microbial communities may feedback on one another’s composition. This work additionally shows that plant-microbial interactions may be an important pathway for the success of invasive

PAGE 18

8 plants in new ranges. Further studies s hould explicitly examine the potential positive feedbacks on invasion through soil microbial communities.

PAGE 19

9 Figure 1-1: Conceptual diagram of intera ctions between soil microbes and plants. N-fixing bacteria Mycorrhizal fungi Organic Matter Free-living microbes Inorganic nutrients Plant litter decomposition

PAGE 20

10 CHAPTER 2 MYCORRHIZAE DIFFERENTIALLY ALTER GROWTH, PHYSIOLOGY AND COMPETITIVE ABILITY OF AN INVASIVE SHRUB Introduction Plants must overcome obstacles of disper sal, abiotic conditio ns, and competition with existing species in order to colonize a nd establish in a new geographical locality.1 Arbuscular mycorrhizal fungi can aid or hinde r the establishment of a new species by ameliorating or intensifying th e abiotic stresses en countered in the new range. Arbuscular mycorrhizae (AM) may improve phosphorus (P) availability and enhance leaf photosynthetic rates and growth rates of the hosts (Sique ira et al. 1998, Sharma and Adholeya 2000). Due to improved P nutriti on, mycorrhizal plants may allocate proportionally less to roots while increasing leaf area ratio and specific leaf area (Berta et al. 1995, Gavito et al. 2000, Lovelock et al. 1996, Son and Smith 1988). Change in allocation patterns of the host, in turn, may affect its intera ction with neighboring plants for light and soil nutrients. Because the res ponse of plants to AM is dependant on both soil-P levels and light availa bility (Gavito et al. 2000, Graham et al. 1997, Peng et al. 1993), effects of AM fungi on the plant inva sion process must depend on these abiotic factors. Although AM can infect a wide range of hos ts from various geographical localities, the responses of host plants to mycorrhizae vary greatly depending on the combination of 1 The information in this chapter was published in: Bray, S.R., K. Kitajima, and D.M. Sylvia. 2003. Mycorrhizae differentially alter gr owth, physiology, and competitive ability of an invasive shrub. Ecological Applications 13 : 565-574 and is used here with the permission of the Ecological Society of America.

PAGE 21

11 plant and fungus genotypes (Johnson et al.1997, Smith and Smith 1996). Different fungal genotypes can have positive, negative or little effect on the growth of the same host species (Boerner 1990, Monzon and Azcon 1996, van der Heijden and Kuyper 2001), because AM may differ in their ability to infect a given host, efficiency of P transferred to the host, carbon demand, soil ad aptation, and host compatibility (Graham et al. 1996, Johnson 1993, Johnson et al. 1997, Monzon and Azcon 1996). Thus, assessment of the effects of AM on plant inva sion must consider the genotype and source of fungal isolates. Mycorrhizal fungi may alter competitive interactions between invading and local plants. Although AM have been largely i gnored as a mediator of plant invasion (Richardson et al. 2000), they have been show n to increase the growth of an invasive plant species over natives and accelerate the pr ocess of invasion in a grassland ecosystem (Marler et al. 1999). More generally, diffe rences in competitive ability under the influence of mycorrhizae can alter community composition by favoring mycorrhizalresponsive, inferior competitors (Hartnett et al. 1993, Moora and Zobel 1996, Smith et al. 1999) or causing competitive exclusion of non-responsive dominants (Gange et al. 1999, Marler et al. 1999). Whether mycorrhizal fungi promote or inhibit the process of plant invasion must be determined by examining res ponses of exotic plants to multiple fungal genotypes with and without competition w ith native species that occupy similar ecological niches. Here we report the results of two experi ments that examined the effects of fungal isolates and abiotic environment on growth and competitive interactions of an exotic invasive shrub. Specifically, we examined the effects of various isolates of AM fungi,

PAGE 22

12 soil-P, light and competition type on the gr owth, physiology and biomass allocation of Ardisia crenata Sims (Myrsinaceae, hereafter Ardisia ). In the first greenhouse experiment, we examined the effects of light, soil-P content, and AM fungal isolates on growth, photosynthetic rates and bi omass allocation patterns of Ardisia seedlings grown singly in pots. We hypothesized that mycorrhizal plants w ould exhibit higher growth rates, invest more bioma ss aboveground, and maximize leaf area ratio compared to nonmycorrhizal plants, and plant response to phosphorus would vary among mycorrhizal isolates. In the second gr eenhouse experiment, we examined the effects of AM on intervs. intraspecific competition between seedlings of Ardisia and a native shade-tolerant subcanopy tree, Prunus caroliniana (Mill) Aiton (R osaceae, hereafter Prunus ). We hypothesized that Ardisia would be less affected by c onspecific than heterospecific competition, particularly when mycorrhizal. Materials and Methods Species Ardisia is a woody evergreen shrub that was introduced as an ornamental to the southeastern United States from east As ia ca. 100 years ago (Dozier 1999). Ardisia is actively invading mesic forest understory in Lo uisiana, Texas, Hawa ii, and north-central Florida (Singhurst et al. 1997). Ardisia forms dense monodominant patches in the understory and suppresses local species richne ss and diversity of na tive understory plant species (A. Fox and K. Kitajima, unpublished data). The architecture of the plants creates strong selfand neighbor-shading, even when plants are not in dense clumps (K. Kitajima and M. Dooley, unpublished data). Growth of Ardisia seedlings in the field has been shown to be positively correlated with soil-P cont ent (Dozier 1999) and Ardisia roots are highly colonized by AM in the field (S. Bray, unpublished data). Prunus was

PAGE 23

13 chosen as a heterospecific competitor in the experiment because its juveniles are abundant in forest understories where Ardisia typically invades (A. Fox and K. Kitajima unpublished data). These species have similar se ed size, and the juveniles of both species have evergreen leaves that are common under the partially deciduous canopy of mesic hardwoods forests in north -central Florida. Experiment 1: Effects of Soil-P, Light and Inoculum Source on Growth and Allocation Three inocula and a sterile control were us ed in this experiment. The inocula were Glomus etunicatum (S3029), G. fasciculatum (S3060), and host-associated fungi. S3029 has been maintained in pot culture for 15 years; S3060 was isolated in 1997 from a tomato field in north-central Florida (Sylvi a et al. 2001). These isolates have caused positive growth responses in both agronomic and native woody plants (Sylvia 1990, Sylvia et al. 1993) and will be collectively referred to as standard inocula. The hostassociated inoculum (HA) was composed of a corn trap culture initiated with washed Ardisia roots gathered from a north-central Florid a hardwood forest (29’40” N, 82’9” W). Inocula were composed soil, roots and spor es produced in the corn trap cultures. Infection potential of the inocula was determined by growing corn ( Zea maize ) with 5 g of inocula for 4 wk (Sylvia 1994). Infecti on potential rather than most probable number was used because we were interested in comparing inocula rather than determining absolute numbers of propogules. S3060 had the highest infection potential (51.7% + 10.2%) followed by HA (46.7% + 7.21%) and S3029 (33.0% + 7.00%). Ardisia seeds were gathered from four popul ations in Gainesville, FL, mixed, cleaned of pulp and stored in moist sand at 4oC for 24 wk. Seeds were germinated in petri dishes lined with moist filter paper at 26oC. Before leaf development, seedlings

PAGE 24

14 (approx. 15 d after radicle emergence) were tran sferred to bleach-ster ilized Deepots (6.2 cm top diameter x 24.5 cm, J.M. McConke y & Co., Sumner, WA) containing a 1:1:1 steam-pasteurized mixture of soil:sand:peat moss. The soil was collected from Austin Carey Forest (29’43” N, 82’ 13” W). The soil had a pH of 5.7 (2:1, H2O:soil), 0.1% organic matter, and 1.6 mg Mehlich-I-extractable P kg-1. The sand was acid washed to remove excess P, and rinsed until neutral pH was achieved. Phosphorus was added in the form of KHPO4 at 0, 5, 30 or 60 mg P kg-1 soil. Pots were filled full with the growth medium; 5 g of inoculum were added to inoc ulated treatments and thoroughly mixed with the growth medium. After adding seedlings of equal mass and remaining soil, pots were randomly assigned to shading treatments with one (moderate light) or two (low light) layers of shade cloth supported by PVC frames. These treatments created mean photosynthetic photon flux densit ies (PPFD) of 412 mol m-2 s-1 (moderate) and 212 mol m-2 s-1 (low) at mid-day. The shade treatment s were randomly assigned to locations within each of three blocks along a greenhouse bench. Each block contained 3 to 4 plants per treatment group. For the control, S3060 a nd S3029 inoculum types, there were four P levels by two light levels by 10 replicates for a total of 240 plants. The HA inoculum was used only at the 5 mg kg-1 P level, but had 10 replicates in each light treatment for 20 HA inoculated plants. Plants were wate red when necessary and biweekly given a modified Hoagland’s solution of 0.1x concentrat ion of all nutrients, except P at 0.01x concentration (Sylvia et al. 2001). The photosynthetic rates of the most recen tly fully expanded leaf were measured with a Li-6400 gas-exchange system (Li-Cor, Lincoln, NE) for three seedlings per light and inoculum treatment combination at the 5 mg kg-1 P level during the fourth month

PAGE 25

15 after planting. All measurements were taken between 0800 and 1200. A CO2 mixer unit maintained the CO2 concentration of incoming refere nce air at 380 ppm. Temperature of the thermister block was maintained at 26oC. Flow rate was 250 mL min-1. Light was supplied with a blue-red LED (LI620002B). Leaves were exposed to 500 mol m-2 s-1 PPFD for 15 min for photosynthetic inducti on, after which quasi steady-state gas exchange rates were recorded at lig ht levels of 800, 500, 300, 100, 60, 40, 20, 10, and 0 mol m-2 s-1. Plants were harvested after 258 d and leaf area was measured immediately with a portable area meter (Li-3000, Li-Cor, Lincoln, NE). Roots of five randomly selected plants within each treatment combination we re weighed for fresh mass and set aside for analysis of AM colonization. Roots of the re maining plants along with stems and leaves of all plants were dried at 60o C until constant mass was r eached. The dry mass of roots used for assessment of colonization was estimat ed from fresh:dry mass ratios of roots. To examine biomass allocation pattern, root:s hoot ratio (R/S), specific leaf area (SLA, leaf area divided by leaf mass), net assim ilation ratio (NAR, net carbon assimilation on leaf area basis) and leaf area ratio (LAR, leaf area divided by total mass) were calculated. Relative growth rate (RGR) was determined using the following equation: Tissue phosphorous contents were determin ed after grinding the dried stems and leaves with a Wiley Mill with a 20-mesh sc reen (Thomas Scientific, Swedesboro, NJ). Due to the small size of seedlings, two plants per block per treatment group at the 5 mg kg-1 P level were combined. Samp les were ashed overnight at 500oC and digested with (days) study of duration mass) seedling (initial harvest) at mass (seedling ) day g (mg RGR ln ln1 1

PAGE 26

16 12N HCl, followed by colorimetry methods of Murphy and Riley (1962) to determine tissue-P concentration a nd content per plant. To quantify mycorrhizal colonization, Ardisia roots were cleared in 10% KOH at 80o C for 45 min, while corn roots used for determination of inoc ulum potential were cleared for 15 min. A longer clearing time was necessary for Ardisia roots due to their high tannin content. Ardisia roots were then rinsed and soaked in H2O2 at 50oC for 10 min for additional clearing. Roots were again rinsed and acidified in concentrated HCl (5 ml HCl per 200 ml-1 H2O) for five minutes. The roots were then soaked overnight at room temperature in trypan blue stain, whic h had been used successfully to stain fieldcollected Ardisia roots (Bray, unpublished data). To estimate mycorrhizal colonization, twenty one-cm root fragments were mounted on microscope slides and examined at 100x. Roots were scored as mycorrhizal if they c ontained coils or arbuscules, spores, vesicles or typical AM hyphae (aseptate, larg e diameter, angular branching). Experiment 2: Effects of Mycorrhizae on Seedling Competition As a competitor of Ardisia we chose Prunus a shade-tolerant subcanopy tree found in north-central Flor ida hardwood forests. Prunus seedlings with four to eight leaves and 7-12 cm height gr own in soil-free medium were acquired from a commercial nursery (Urban Forestry Services, Micanopy, FL). Examination of cleared and stained roots of 10 seedlings revealed no mycorrhizal colonization. Although Prunus spp. Have been shown to be ectomycorrhizal in some cases (Smith and Read 1997), we found no evidence of ectomycorrhizae in Prunus caroliniana Ardisia seedlings were collected from the same invaded forest as the inoculum in experiment 1. Ardisia seedlings had four to eight leaves and heights of 5-10 cm. Cleared and stained Ardisia roots revealed a total mycorrhizal colonization level of 52% + 2%.

PAGE 27

17 Prunus and Ardisia were grown in heteroand conspecific competition with (AM) or without (NM) mycorrhizal inoculum. Two plants were potted in each 3.8 L pot containing the same medium as in experiment 1 with no additional P. Plants were paired according to height and number of leaves. Twenty pots contained two Prunus seedlings ( Prunus conspecific competition), twenty contained two Ardisia seedlings ( Ardisia conspecific competition), and forty contained one Ardisia and one Prunus seedling ( Prunus heterospecific competition and Ardisia heterospecific competition). Half of the pots were randomly assigned to the NM treat ment and were drenched with 75 mg of benomyl (Benlate) dissolved in one L of deionized water. Although benomyl may have phytotoxic effects in some species, no phytotoxic effects have been reported in either Prunus or Ardisia and benomyl is commonly used to control fungal pathogens in horticultural nurseries growing Prunus persica Prunus dulcis and Prunus serotina (Fontanet et al.1998, Stanosz 1992). Plants in the AM treatm ent received 5 g of G. fasciculatum (S3060) inoculum to supplement indigenous AM fungi and one L of deionized water. In th e AM treatments containing Ardisia each pot contained S3060 and fungi already inhabiting the Ardisia seedling; treatments without Ardisia contained only S3060. Pots were then randomly placed under a shade frame (approx. 20% open-sky light) in the greenhouse. Pots were wate red as needed and received the modified Hoagland’s solution used in experiment 1 biweekly. The minimum and maximum temperatures averaged 18oC and 34oC, respectively. The average maximum PPFD in the greenhouse was 910 mol m-2 s-1. At 160 d after transplanting, growth an alysis was conducted with one randomly selected plant per pot to ensure statistical independence. Root:shoot, LAR, SLA, and

PAGE 28

18 RGR were determined as in experiment 1. A subsample of each root system of six randomly selected individuals per treatment was used to estimate percent colonization after determining fresh mass. Fresh:dry mass ratios of remaining roots were used to estimate dry mass of the subsamples used to determine colonization. Initial mass of seedlings was estimated through an allometr ic relationship of leaf number (mean + S.E.: Ardisia 5.1 + 0.24; Prunus 5.7 + 0.41) and height ( Ardisia 7.2 + 0.23 cm; Prunus 9.3 + 0.35 cm) with the total dry mass for each species ( Ardisia 1.31 + 0.05 g; Prunus 0.46 + 0.04 g). Percent mycorrhizal colonization was determined as in experiment 1. Tissue-P content was quantified only for leaves with the same method as in experiment 1. Statistical Analyses Allocation and growth data from both expe riments were analyzed with factorial model fitting (JMP 4.0, SAS Institute). For the analysis of the first experiment, the effects of soil-P, light and inoculum type (control vs standard inocula), a nd their interactions on R/S, LAR, SLA, NAR, and RG R were analyzed. Because th e host-associated inoculum was given only at 5 mg kg-1 P level, the effects of inocul um type (control, S3029, S3069, HA), light and their interactions at the 5 mg kg-1 P level were then examined. The results of experiment 2 were analyzed with a m odel that included the effects of species ( Ardisia or Prunus ), competitor (conspecific or heterospecif ic), mycorrhizal status (AM or NM), and their interactions. When treatment a nd interaction terms were not significant ( P > 0.1), they were dropped from the model and th e results of the analysis with the reduced model were reported. Tukey HSD at alpha = 0.05 was used to compare differences between means. Effects of LAR and R/S on RGR were examined with multiple regression analysis. Differences in survi vorship among treatment groups were evaluated

PAGE 29

19 with logistic regression models in both expe riments. Percent colonization levels were converted by square root of the arcsine to achieve normality and analyzed by ANOVA. Results Experiment 1: Effect of Light, Soil-P and Inoculum Type In the first analysis, effects of th e standard inocula (S3029 and S060) and nonmycorrhizal control were examined at fact orial combinations of two light levels and four soil-P levels. Overall, there was no diffe rence in RGR or biomass allocation patterns among three mycorrhizal treatments (S3029, S 3060 and control) in any combination of light or soil-P. Thus, inoculum type was dropped from the ANOVA model. Neither soilP nor light affected RGR, SLA or NAR. Leaf area ratio and R/S, however, were affected by light and soil-P (Table 2-1, 2-2). Leaf area ratio was pos itively correlated with RGR and explained 37% of the variance in RGR ( P < 0.0001) while R/S was negatively correlated with RGR and explained 15% of the variance ( P < 0.0001). Leaf area ratio and R/S together explained 39% of the variance in RGR ( P < 0.0001) in a multiple regression. A total of twen ty-seven seedlings of 260 died over the course of the experiment, but survivorship was not affected by treatments. In the second analysis, the effects of HA inoculum on seedling growth and biomass allocation were compared to the standard inocula and control, at the 5 mg kg-1 P level (Figure 2-1a-d). Relative growth rates of plants with HA inoc ulum were twice that of the standard inocula and control ( P < 0.0001), while LAR was 2.5-3x that of the other treatment groups ( P < 0.0001). For pooled data acro ss treatments, LAR was positively correlated with RGR ( P < 0.0001) and explained 35% of the variance in RGR. Conversely, R/S of plants i noculated with HA were half that of the other inoculum treatments ( P = 0.0006). Root:shoot ratio was negatively correlated with RGR ( P <

PAGE 30

20 0.0001) and explained 23% of the variance in RGR. Together LAR and R/S explained 39% of the variance in RGR ( P < 0.0001). Leaf area ratio was the only morphological measure affected by light treatment, with plan ts in low light having higher values than those in moderate light ( Plight = 0.003). Inoculum effects on SLA were nearly significant ( P = 0.06); plants inoculated wi th HA had greater SLA (193.1 + 17.49 g/cm2 vs. 136.9 + 14.47 g/cm2) than controls and other inoculum sour ces. NAR was not affected by light or inoculum. Both shoot-P concentration and content differed among treatments ( P = 0.005 and P = 0.04, respectively). Plants inoculated with S3060 had the highest P concentration and content, while HA inoculum had the lo west P concentrati on and second-highest content (Figure 2-1 e, f). Leaf gas exchange data showed similar tre nds to RGR in relation to inoculum type (Table 2-3). The maximum photosynthetic ra te of plants inoculated with HA were approximately twice that of plants with S3060 and controls under moderate light. Due to the small size of plants and leaves inoc ulated with S3029 under moderate light, no gas exchange rates are available for this treatme nt group. Plants inoculated with HA had rates approximately twice that of controls in both light treatments. Dark respiration rates of all inocula were similar to the dark respiration rates of control plants. Mycorrhizal colonization of plants mirrored the growth and biomass allocation differences among inocula ( P <0.0001); plants with HA i noculum were highly colonized (mean + S.E.: 67 + 7%) whereas standard isolates (S3029 & S3060) had low colonization rates (17 + 7% and 7 + 7%, respectively). Controls were virtually non-colonized (0.8 + 0.8%). Septate hyphae, presumably saprophyt ic, were observed externally with many

PAGE 31

21 root samples, primarily concentrated in plants inoculated with S3060 and S3029. Colonization levels were not related to orig inal inoculum potential as determined by the corn assay. Experiment 2: Effects of Myco rrhizae on Seedling Competition Mycorrhizal colonization was significantly lower in plants treated with benomyl (NM treatment) than AM plants of both species ( P < 0.0001). Benomyl was more effective in reducing colonization in Prunus (AM = 73 + 7%, NM= 9 + 10%) than Ardisia (AM = 59 + 1%, NM = 38 + 6%; Table 2-4). Competition type had no effect on colonization levels. Nine of 80 seedlings died over the cour se of the experiment. Mortality was significantly higher for NM Prunus seedlings (7 of 9 dead seedlings, log-likelihood 2 = 6.57, P = 0.038, Pmyc = 0.037) than AM Prunus seedlings. Relative growth rates differed significantly between treatments ( P < 0.0001) with Prunus having greater RGR than Ardisia ( P < 0.0001, Figure 2-2a). Prunus and Ardisia also responded differently to the competition treatment (Table 2-4, significant species x competition interaction). While Ardisia had the highest RGR in heterospecifc competition, the RGR of Prunus decreased approximately by half when grown in heterospecific competition without mycorrhizae (Figure 2-2a). Leaf area ratio also differed among treatment groups as Prunus and Ardisia responded differently to mycorrhizal status ( Psp*myc = 0.015, Table 2-4). Unlike in experiment 1, Ardisia tended to have higher LAR when nonmycorrhizal in heterospecific competition, whereas Prunus had higher LAR with mycorrhizae in both competition types (Figure 2-2b). Ardisia had a higher R/S than Prunus ( P < 0.0001, Figure 2-2c); however, R/S of Ardisia decreased under conspecific competition when mycorrhizal

PAGE 32

22 ( Pcomp*myc = 0.04). Specific leaf area did not vary significantly among treatments and species ( P = 0.18, Figure 2-2d). Leaf-P concentration and content differed among species and treatments (Table 5; Fig. 2-2e, f). Ardisia had lower P concentration than Prunus and was not affected by competition or mycorrhizal treatment (Figure 2-2e). Prunus had its highest P concentration in the NM, heterospecific co mpetition treatment, while the other three treatments did not differ significantly from one another (Figure 2-2e). Ardisia in the NM, conspecific treatment had a significantly lo wer total P content than the other three treatment groups (Figure 2-2f). Prunus had lower P content in NM than AM treatments with no effect from competition (Figure 2-2f). Discussion Effect of Inoculum Source on Ardisia Although all three types of inocula colonized Ardisia roots, they had strikingly different colonization levels and effects on biom ass allocation and growth of the host. In experiment 1, only the HA inoculum isolated from field-collected Ardisia roots, but not the standard inocula, improved seedling RGR over the nonmycorrhizal control. As Ardisia had no response to soil-P nor was highe r RGR accompanied by an increase in P content or concentration, it appears that the benefit of the HA mycorrhizae was mediated through changes in allocation and physiology ra ther than improved P nutrition. Plants inoculated with HA inoculum ha d less relative investment in roots, greater leaf area and higher Amax than the control. HA inoculum phenotypically altered LAR of Ardisia seedlings from a low value typical for shade tolerant species to a higher value. LAR is generally thought to be the primary determin ant of RGR both across and within species

PAGE 33

23 (Poorter and Remkes 1990) and in this study accounted for 35-37% of the variation in RGR. In a study by Lovelock et al. (1996), shade-tolerant seedlings of Beilschmiedia pendula also increased their RGR through an in crease in LAR when mycorrhizal and its morphology became more similar to that of more light-demanding plants. The lack of positive effects on morphology and growth by the standard inocula despite their positive effects on tissue phos phorus concentration was surprising, especially given that they have been shown to increase the biomass of several species, including woody, native plants (Sylvia 1990, Sylvia et al. 1993, Sylvia et al. 2001). Possibly the costs of the two Glomus species were greater than their benefits at the light levels used in this experiment. Mycorrhizal fungi can demand up to 20% of the total C budget of a plant in extreme cases (Peng et al. 1993), and carbon costs can vary widely among fungal genotypes (Graha m et al. 1996). The differences between inoculum types may also be mediated by fungal diversity. Isolates S3029 and S3060 represent singlespore cultures, while the HA inoculum was likely composed of mu ltiple fungal species and strains. At least one of these may have been more effective than the two Glomus species. Greater numbers of fungal species ha ve been shown to increase the productivity of grass macrocosms In contrast to the strong effect of HA i noculum in experiment 1, suppression of mycorrhizal fungi in experime nt 2 had no effect on RGR of Ardisia seedlings, even though colonization rates were reduced from 59% to 38%. These seedlings were collected from a dense Ardisia population in the field and were presumably colonized with mycorrhizae similar to HA. The magnitude of reduction of mycorrhizal colonization (33% reduction) in this study was comparable to the reductions in many

PAGE 34

24 other studies, (Kahiluoto et al. 2000, M oora and Zobel 1996, Smith et al. 1999). However, it has been suggested that the re lationship between mycorrhizal colonization and plant benefit is curvilinea r with benefit to the plant ev entually reaching a plateau at some colonization level (Gange and Ayers 1999). Ardisia may have reached its maximal benefit at or before a colonization rate of 38%, as found in the benomyl treatments. That Ardisia has a differential response to di fferent inoculum types may have important implications for its invasive ability. In heavily invaded areas, Ardisia is already associating with e ffective mycorrhizal fungi that alter its morphology and physiology to that of faster-growing plants The main mode of resource competition by Ardisia is through casting dense shade to its ne ighbors. Increased LAR, enabled by the reduction in R/S due to my corrhizae, must enhance Ardisia ’s competitiveness for light in the forest understory. Competitive Interactions Ardisia seedlings grew better in heterospecific competition with Prunus seedlings than in conspecific competition. Conversely, Prunus seedlings had lower survival and growth with Ardisia seedlings than with conspeci fic seedlings, especially in nonmycorrhizal treatment (Fig. 2-2a). The architecture of Ardisia results in a higher amount of selfand neighbor-shading than that of Prunus (K. Kitajima, unpublished data). Hence, each Ardisia seedling is more shaded by a conspecific neighbor than a heterospecific neighbor. In conspecific competition, Ardisia responded with greater phenotypic plasticity of increasing LAR th an in competition with less shade-casting Prunus Prunus seedling growth and survival wa s reduced to a greater extent by heterospecific competition in the nonmycorrhizal than in the mycorrhizal treatment. The

PAGE 35

25 presence of HA and S3060 inocula reduced the negative effect of interspecific competition on Prunus as is often observed in mo re mycorrhiza-dependant species (Hartnett et al. 1993, Moora and Zobel 1996, Smith et al. 1999). More mycorrhizadependant species often have a low total investment in root s (Jakobsen 1991). Prunus had overall higher RGR than Ardisia and this difference was associated with inherently higher LAR and lower R/S of Prunus (Fig. 2-2 a-c). Mycorrhizae apparently allowed Prunus seedlings to invest less in roots and more to leaf area, and enabled them to compete more effectively with Ardisia seedlings for light. This finding of an apparent greater AM dependency by a faster growing species in a compet itive regime is interesting because often the opposite trend has been found in the absence of heterospecific competitors (e.g. Janos 1980, Zangaro et al. 2000). Different plant species in the same co mmunity can support different mycorrhizal communities in their rhizosphere (Bever 1994) and cause differential rates of sporulation (Bever et al. 1996). A high percentage of Ardisia roots are colonized by AM fungi in the field (Bray, unpublished data), likely dominated by preferre d mycorrhizal fungi. The field-collected Ardisia seedlings in Experiment 2 ha d been colonized by mycorrhizal fungi, some of which remained after the benomyl treatment. These fungi colonized Prunus seedlings at a low level (9%), but they did not benefit growth of Prunus seedlings in the heterospecific competition treatment. The community composition of mycorrhizal fungi may be highly modified in the dense clump of Ardisia in the invaded forest. If Ardisia alters the composition of AM, the competitiveness of Ardisia may be increased. Implications of Effects of Mycorrh izae on Exotic Species Invasion As a new colonist in Florida, Ardisia is apparently not limited by the lack of potential mutualists and, in fact, benefits fr om the local mycorrhizal fungi. Unlike typical

PAGE 36

26 invaders, Ardisia is highly shade-tolerant and has low RGR. Many other exotic species that have a higher RGR than Ardisia have a negative or neut ral response to mycorrhizae when they are grown alone (Marler et al. 1999, Philip et al. 2001, Richardson et al. 2000). This appears to be the first study to docu ment a positive response of a slow-growing exotic to native mycorrhizae. It is likely, however, that further study will show that there is no link between life history and mycorrhizal dependency in exotic species as has been found in the mycorrhizal literature as a whole (Allsopp and Stock 1992, Janos 1980, Smith and Smith 1996, Zangaro et al. 2000). Until a greater predictive framework for mycorrhizal response is developed, invasi ve plant response must be examined on a species by species basis. The results of our study suggest that it is difficult to predict how competitive interactions between exotic and native plants are modified by mycorrhizae. The exotic plant’s response in isolation does not necessar ily predict its response to mycorrhizae in a competitive environment. Another study of co mpetition between an exotic forb and native grass found that neither species’ bi omass was altered by mycorrhizae when grown in isolation, but when grown in mixture, my corrhizae increased the growth of the exotic plant to the detriment of the native (Marler et al. 1999). Our results also suggest that the type of mycorrhizal in oculum must be considered when evaluating mycorrhizal effects on compe titive interaction between native and exotic species. The results vary depending on whethe r the fungal inoculum is the one preferred by the native or the exotic. Thus, studies should incorporate eval uation of both species and the mycorrhizae of the ecosystem invaded by exotic plant species.

PAGE 37

27 The role of AM in mediating plant invasi ons and competitive interactions needs to be examined carefully. The response of exo tic plants to mycorrhizae is highly variable depending on genotype interactions both in is olation and in competitive environments. Understanding how native and exotic plants respond to the local microbial community will be important for understanding the mechan isms and impacts of community invasion. Similarly, it is also imperative that we dete rmine how exotic species potentially alter the microbial community and its ecosystem functions.

PAGE 38

28 Table 2-1: ANOVA summarizi ng the effects of light a nd soil on LAR and R/S in experiment 1 (model P <0.05). Fungal isolate effects (control, S3029, and S3060) were not significant and were pooled. LAR R/S Source F P F P Model 4.51 0.0001 3.24 0.003 Light 9.24 0.002 0.20 0.66 P-level 3.11 0.03 3.94 0.009 Light P-level 3.37 0.02 3.24 0.02 Notes : Fungal isolate effects (control, S3029, and S3060) were not significant and were pooled. Abbreviations are: LA R, leaf area ratio; R/S, root to shoot ratio; P, phosphorus. Table 2-2: Means of leaf ar ea ratio (LAR) and root:shoot ra tios (R/S) from all soil-P levels in experiment 1. Different letters in the same column signify significant difference in Tukey HSD values (alpha = 0.05). [P] LAR (cm2 g-1) R/S RGR (mg g-1 day-1) Moderate Light 0 20.5 a 2.38 a 5.05 5 17.5 a 2.24 a 5.08 30 22.5 a 1.83 a 6.05 60 28.9 ab 1.52 b 6.62 Low Light 0 29.1 ab 1.91 a 5.77 5 24.9 a 2.12 a 4.75 30 38.0 b 1.66 b 7.02 60 24.2 ab 2.08 a 5.18 Notes : Abbreviations: RGR, relative growth rate ; other abbreviations as in Table 2-1. Table 2-3: Comparison of light saturated net photosynthesis rate (Amax) and dark respiration under moderate vs. low light treatments (means + 1 SE) from gas exchange measurements of three individuals from experiment 1. Amax Dark Respiration Inoculum Moderate Low Moderate Low Control 2.02 + 0.55 2.02 + 0.53 -0.279 + 0.062 -0.310 + 0.045 S3029 NA 3.43 + 1.19 NA -0.307 + 0.132 S3060 2.65 + 0.38 2.19 + 0.35 -0.382 + 0.078 -0.253 + 0.015 HA 4.04 + 0.28 4.42 + 0.78 -0.356 + 0.046 -0.313 + 0.065 Notes : No data were collected for the moderate light treatment of isolate S3029 due to small size of leaves in this group. Abbrevia tions: NA, not available; HA, host-associated.

PAGE 39

29Table 2-4: ANOVA summary of the effects of species, competition, and mycorrhizae on RGR, LAR, R/S, and colonization rates from experiment 1 (model P < 0.05). A dash (-) indicates that th e specified effect was not significant ( P > 0.1) and was dropped from the model. RGR LAR R/S Colonization Source F P F P F P F P Species 29.1 <0.0001 0.129 0.73 62.5 <0.0001 2.03 0.16 Myc 2.94 0.09 1.16 0.28 2.93 0.09 38.9 <0.0001 Comp 3.19 0.08 2.20 0.14 0.059 0.81 Sp*Myc 1.10 0.30 6.31 0.01 12.0 0.001 Myc*Comp 0.481 0.49 4.54 0.04 Sp*Comp 11.9 0.001 3.36 0.07 Sp*Myc*Comp 3.78 0.06 Notes : Degrees of freedom were: RGR, 1,63; LAR, 1,65; R/S, 1, 66; colonization, 1,67. NS indicates that the specified effect was no t significant (P > 0.1), and was dropped from the model. Table 2-5: ANOVA summary of the effect of species, competition, and mycorrhizae for P concentration and total P content from th e competition study. P concentration (mg P g-1 tissue) Total P content (mg P) Source F P F P Species 43.6 <0.0001 19.1 <0.0001 Myc 0.086 0.70 24.9 <0.0001 Comp 9.6 0.003 2.57 0.11 Sp*Myc 3.81 0.06 Myc*Comp 4.15 0.05 4.51 0.04 Sp*Comp 7.06 0.01 5.27 0.02 Notes : Degrees of freedom were, P concentration, 1,64; total P conten t, 1,65. The three-way interaction was dropped due to lack of significance in both tests as was the species mycorrhizae interaction in the P content test.

PAGE 40

30 Figure 2-1: Response of Ardisia to light and inoculum type at 5 mg kg-1 P. Light treatment is distinguished only wh en it had a significant effect ( P < 0.05). a) Relative growth rate in response to inoculum ( P < 0.0001). b) Leaf area ratio in response to light ( P = 0.003) and inoculum ( P < 0.0001). Open bars = moderate light; hatched bars = low light c) Root:shoot ratio in response to inoculum ( P = 0.0006). d) Specific leaf ar ea in response to inoculum ( P = 0.06). e) Shoot-P concentration (mg P g-1 tissue) in response to inoculum ( P = 0.005). f) Shoot-P content (mg P) in response to inoculum ( P = 0.04). Error bars represent + 1 SE. Letters signify difference between Tukey HSD values at alpha = 0.05. RGR (g g -1 day -1 ) 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016 LAR (cm 2 g-1 total biomass) 10 20 30 40 50 60 70 80 Inoculum Source ControlS3029S3060HA Root : shoot ratio 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 a a a b a a a b a a a c SLA (cm 2 g-1 leaf biomass) 0 50 100 150 200 ab ab ab bc P concentration (mg P g -1 tissue) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Inoculum Source ControlS3029S3060HA P content (mg P) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 ab ab a b b ab a aba b c d e f

PAGE 41

31 Figure 2-2: Response of Ardisia and Prunus to heterospecific or conspecific competition and mycorrhizal status. H = heterosp ecific competition; C = conspecific competition. a) Relative growth rate ( P < 0.0001). b) Leaf area ratio ( P = 0.04). c) Root to shoot ratio ( P < 0.0001). d) Specific leaf area ( P = 0.18). e) Leaf-P concentration (mg P g-1 tissue) ( P < 0.0001). f) Leaf-P content (mg P) in response to inoculum source ( P < 0.0001). Error bars represent + 1 SE. Letters signify difference between T ukey HSD values at alpha = 0.05. RGR (g g -1 day -1 ) 0.002 0.004 0.006 0.008 0.010 LAR (cm 2 g-1 total biomass) 20 40 60 80 100 Root:shoot 0.0 0.2 0.4 0.6 0.8 1.0 1.2 SLA (cm 2 g-1 leaf biomass) 50 100 150 200 250 ab ab ab a bcd cd ab d a ab a a b b b b a a a P concentration (mg P g-1 tissue) 2 4 6 8 10 12 P content (mg) 0 2 4 6 8 10 a ab a a cb ab c ab a a a bc abc ab bc cAM NM AM NM H C H C H C H C Ardisia Prunus AM NM AM NM H C H C H C H C Ardisia Prunusa b c d e f

PAGE 42

32 CHAPTER 3 SOIL MICROBIAL COMMUNITY STRUCT URE AND FUNCTION IN FLORIDA PLANT COMMUNITIES PRONE TO NON-NATIVE PLANT INVASION Introduction Microbes are a key component of ecosystems that function as pathogens, mutualists and decomposers. Recent studies have shown that the composition of microbial communities differs among plant communities, as influenced by biotic and abiotic factors (Waldrop et al. 2000, Myers et al. 2001, Gallo et al. 2004, Leckie et al. 2004, Waldrop and Firestone 2004). Microbial community me tabolic potential and structure change across climatic gradients (Staddon et al. 1998) nitrogen levels (P ennanen et al. 1999, Gallo et al. 2004, Leckie et al. 2004), and soil moisture contents (Bossio and Scow 1998). Microbial community compositi on also varies under differe nt plant species within a given community (Grayston et al. 1998, Ba rdgett et al. 1999, Saetre and Baath 2000, Priha et al. 2001). These differe nces are likely due to the di fferences in root exudates and turnover and the quantity and quality of aboveground litter in puts (Grayston et al. 1996). Plant species composition is changing in many plant communities due to the introduction of non-native species. These inva sions result not only in shifts in plant community composition, but can also have ecosystem-level effects such as altered nutrient levels, hydrology, a nd soil accumulation (Gordon 1998). Invasions, then, offer a “natural experiment” in which to examine th e role of species composition in ecosystem function. It is likely that altered substr ate composition due to pl ant composition shifts accompanied by altered abiotic soil environment will result in altered microbial

PAGE 43

33 community composition. In a temperate forest, two non-native understory species supported microbial communities significantly different from a native understory species within the same forest stand (Kourtev et al. 2002). These di fferences extended below the root zone of the vegetation but larger scale effects were not examined. This leaves interesting questions on the potential impact s of exotic species on microbial community composition: 1) Can invasive plants alte r microbial community composition beyond their crownand root-zone of influence? 2) Do i nvaders have a consiste nt effect on microbial community composition across a landscape scale? and, 3) Does the identity of the invader and the community type it invades influence the potential effects of that invasion on microbial community composition? The state of Florida presents many oppor tunities to examine a number of nonnative invaders across a variety of plant co mmunities. The Florida Exotic Pest Plant Council lists over 120 non-native species as invasive (FLEPPC 2003). These species range from grasses to trees and invade communities from freshwater marshes to upland pine savannas in freeze and freeze-free climate zones. We examined the impact of five invasive plants, ranging from trees to unde rstory herbs, on microbial communities of plant communities in saturated south Florid a everglades marsh soils to well-drained north-central Florida forest soils. We us ed phospholipid fatty acids (PLFA) to examine the structure of microbial communities and Bi olog substrate utilization to examine the function of those communities. Biolog and PLFA profiles have become common methods to examine microbial community composition as they are rapid and of low enough cost to allow for the analysis of the number of samples required for ecological studies. Biolog plates assay the

PAGE 44

34 functional traits of microbial communities by testing their cumulative ability to metabolize 95 different carbon s ubstrates. Because Biolog de pends upon the growth of organisms within the 95 wells, it has many of the same limitations as culture-based techniques and organisms adapted to high reso urce availability are likely to be overrepresented (Smalla et al. 1998). However, Biolog has been successfully used to differentiate microbial communities in different soil types, land management treatments and rhizospheres of different plant speci es (Gorlenko and Kozhevin 1994, Bossio and Scow 1995, Garland 1996). PLFA profiles ar e a useful technique for examining the structure of the microbial community at many levels. A common use is to examine overall “fingerprints” of microbial comm unities by subjecting the abundance of the various fatty acids within a sample to multiv ariate analysis (Frostegard et al. 1993). Major taxonomic groups of microbes, such as fungi, microeukaryotes, and Gramnegative and Gram-positive bacteria can be specifically examined by the use of biomarkers (Vestal and White 1989). In a ddition, functional groups such as aerobes, anaerobes, methanotrophs, and sulfate reducers can be examined by the use of fatty acid biomarkers (Harwood and Russell 1984, Dow ling et al. 1986, Parkes 1987, Hill et al. 2000). We used these two measures of microbial community structure and function to examine microbial communities within invade d and non-invaded areas within the habitat prone to invasion by each of the five invasive species. As both biotic and abiotic factors are known to influence microbial community composition, we propose two alternative hypotheses for the composition of microbial comm unities. Inherent characteristics (e.g. soil moisture, nutrient levels, physiognomy of vegetation) of habitats may drive

PAGE 45

35 composition of microbial communities regardless of plant species. Alternatively, the invasion and dominance of non-native species may override the habitat effect, resulting in changes in the microbial community because of changes in vegtation. Methods Species, Sites, and Sampling The invaders we examined included trees, Sapium sebiferum Schinus terebinthifolius and Melaleuca quinquenervia an understory shurb, Ardisia crenata and an herbaceous perennial, Ruellia brittoniana In addition to being dominants in the habitats they invade, Sapium Schinus and Melaleuca have been shown to have ecosystem-level impacts on native communitie s such as increased soil elevation, increased litter accumulation, or altere d mineralization, disturbance or hydrology (Woodall 1981, Cameron and Spencer 1989, Greenway 1994, Laroche 1994, Gordon 1998). Secondary defensive chemicals in Schinus are known to have a llelopathic effects on other plants (Morton 1978, Mahendra K. J. et al. 1995) and the conspecific Melaleuca alternifolia has antibiotics that depress decom position rates (Boon and Johnstone 1997, Bailey et al. 2003). Thus, we felt that th ese three species would have the greatest potential for altering microbial community composition. We chose the additional understory plants, Ardisia crenata and Ruellia brittoniana as cases in which invaders do not contribute a large percentage of th e total biomass of a plant community. The habitats that these species invade vary in their lo cations, physiognomy of native vegetation, moisture regimes and sus ceptibility to invasion by specific invaders (Table 3-1). For ease of discussion we will id entify these habitats by the species to which they are most prone to invasi on, i.e., Sapium-prone habita t (SA), Schinus-prone habitat (SC), Melaleuca-prone (ME), Ruellia-prone (RU), and Ardisiaprone (AR). For

PAGE 46

36 consistency of presentation thes e habitats will appear in order of hydrological position from wettest to driest in th e results figures and tables. We selected three sites for each habitat. Most habitats included only one native plant community type (Myers and Ewel 1990) except for RU habitats in which two sites were in bottomland swamp forests and one in a cabbage palm hammock and SA habitats sampled in 2 wet prairies in Texas on clay -dominated soil and one site in Florida on sandy muck soil. The two southern-most habitats, SC and ME, were sampled in Cladium -dominated everglades and tropical ha mmock rocklands, respectively. The AR habitats were sampled exclusively in mixed hardwood forests. We chose sites that had areas of high invader density (invaded) and areas free of the invade r (non-invaded). In order to minimize the chance that differen ces between invaded and non-invaded areas were based on site conditions prior to invasi on, we chose only sites where other workers knew the history of the site. The abse nce of invasion in non-invaded areas was maintained either by human intervention or a clear “invasion front” was apparent. Except for the presence of the invasive species, the invaded areas appeared similar to noninvaded areas used for the study in terms of topography and soil type. Within each site, six 5 m x 5 m plots were established—three in invaded areas and three in non-invaded areas. In most cases, ea ch invaded plot was sp atially separated from other distinct areas of invasion. When this was not possible, invaded plots were separated by at least 50 m in a large patch of the invade r. Corresponding non-invaded plots were located at least 30 m from the invasion front. Within each 5 m x 5 m plot, 20 soil cores (diameter 2 cm) were extracted systematically, approximately 80-100 cm apart, from the top 10 cm of soil, including the

PAGE 47

37 litter layer. Cores within a plot were comb ined to create a composite sample. Samples were kept cool (~ 4 C) during collection in th e field and frozen at the end of each day to prevent alteration of the microbial community during the remainder of field collection. The southern most habitats, ME and SC, we re collected within 3 days of each other in the July of 2002 and Florida SA and AR in August 2002. The Texas SA sites were collected in September 2002. In the summer of 2003, RU samples were collected. All samples were collected within the wet season. Microbial Community Composit ion and Nutrient Analysis Microbial community structure was ex amined using phospholipid fatty acid (PLFA) biomarkers. All glassware was heated at 500 C for five hours and PTFE lined caps rinsed with hexane to remove any orga nic matter. A subsample weighing a total of 5 g (dry mass) from each plot was extracted in glass centrifuge t ubes using single-phase phosphate-buffered methane chloroform solvent (White et al. 1979). After two hours, an additional 5 mL each of chloroform and me thanol were added to break the phase. Samples were briefly shaken, vented, and allo wed to separate overnight. Samples were then centrifuged and the or ganic layer passed through a Wh atman #2 paper into a test tube where the solvent was driven off under N2 gas. The organic phase was resuspended and separated on a silcic acid column. The phospholipid fraction was transesterified to fatty acid methyl esters (FAME) by mild alkaline methanolysis (Findlay and Dobbs 1993). For identification, samples were suspen ded in hexane with 19:0 fatty acid as an internal standard and analyzed with an Agilent Technologies 6890 gas chromatograph (Palo Alto, CA) with a 25m U ltra 2 phenyl methyl silicone column. The temperature program increased from 170 C to 270C at 5 C per minute. Peaks were identified by

PAGE 48

38 MIDI peak identification software (MIDI, Inc., Newark, DE) and co-elution with standards. Fatty acids were defined in terms of th e ratio of total num ber of carbon atoms : number of double bonds. The position of the double bond from the methyl end of the molecule is signified by the symbol followed by the carbon position; cis and trans geometry are referred to by “c” and “t.” Th e prefixes “i” and “a” signify and iso and anteiso branching; “cy” signi fies a cyclopropyl fatty acid; 10Me indicates a methyl group on the tenth carbon atom from the carboxyl e nd of the fatty acid. The position of hydroxy groups is indicated by xOH. F ungal:bacterial ratio was calculated as (18:2 6c)/(i15:0 + a15:0 + 15:0 + i16:0 + 16:1 5c + i17:0 + a17:0 + 17:0 + 18:1 7c + cy19:0) (Frostegard and Baath 1996). Gram – bacteria were represented by fatty acids 16:1 7c, 16:1 w7t, 17:0cy, and 18:1 7c; Gram + bacteria by i15:0, a15:0, i16:0, a17:0, and 17:0 10 ME (O'Leary and Wilkinson 1988, Wilkinson 1988). Total PLFA (nmol g-1) was used as a proxy for total biomass. Microbial community function was examined as the ability to metabolize the 95 substrates in Biolog Gm plates (Biolog, Inc., Haywood, CA). We inoculated one plate per plot with a 10-3 dilution of one-gram subsamples. Absorbance was measured at 12, 24, 48, 72, and 96 hours with a microplat e reader (Bio-Rad, Hercules, CA). A subsample of soil from each plot was dr ied at 60 C until constant weight was achieved to determine gravimetric moisture c ontent. Subsamples were then ground with a mortar and pestle to homogenize. Th e carbon and nitrogen contents of soil were determined by combustion on an ECS4010 el emental combustion system (Costech Analytical Technologies, Inc, Valencia, CA).

PAGE 49

39 Statistical Analyses In statistical analyses, site was treated as the unit of replication (i.e., means of the three plots per site), yielding an n of 30. Di fferences in %C, %N, C:N, %moisture, total PLFA (nmol g-1), and fungal:bacterial ratio among ha bitats and invasion-status were tested with ANOVA. Variables were log-tr ansformed for normality and homogeneity of variances when necessary. When interaction terms of the two-way ANOVA were significant, Bonferroni-correct ed t-tests were used to look for differences between invaded and non-invaded soils within each community. Correlations between soil characteristics and fungal:bacterial ratio a nd total PLFA were ex amined with simple linear and quadratic regressions. We used principle components analysis (PCA) to summarize the multivariate data sets and create microbial community fingerpri nts for both PLFAs and substrate utilization from Biolog. The twenty-three most comm on, positively identified PLFAs comprising > 1% of the total amount of fatty acids extr acted were included in the ordination as percentage of the total fatty acids per sample after arcsine-square root transformation. Absorbance values at 48 hours for 96 carbon s ources for four of the five habitats examined (excluding RU) were ordinated to examine substrate utilization patterns. Average scores per site from the first two PC A axes of each ordination were subjected to MANOVA to examine the effects of habitat, in vasion and habitat by invasion interaction. Finally, differences among habitats and inva sion-status in the relative abundance of PLFAs that were important in structuring the PCA were determined with oneand twoway ANOVAs. All univariate statistics were performed in JMP 4.0 (SAS Institute) and ordinations were performed in PCOrd 4.2 (MJM So ftware Design, Gleneden Beach, OR).

PAGE 50

40 Results Soil characteristics varied among habitats but invasion had no significant effect. ME habitats were most different from the other habitats examined with the highest moisture, carbon and nitrogen contents in the soil and lowest C:N ratio (Table 3-2). Percent C, N and moisture were highly correlated with one another (C-N r2 = 0.88, Cmoisture r2 = .61, N-moisture r2 = 0.74). Total phospholipids and funga l:bacterial ratios differed significantly among habitat types. Only habitat type had a significant effect on total PLFA (F4,20 = 22.29, p < 0.0001), with RU, SA, and SC habitats havi ng the highest total PLFA (Figure 3-1). Fungal:bacterial ratio was highest in RU and AR habitats and lowest in ME habitats (F4, 20 = 69.5, p <0.00001, Figure 3-2). Although ther e was no main effect of invasion, there was a significant habitat by invasion interaction (F4, 20 = 3.06, p = 0.04). AR and ME habitats showed a trend of in creased fungal : bacter ial ratio with invasion, while SC, SA and RU habitats showed a trend towards decr eased fungal : bacterial ratio with invasion (Figure 3-2). Both total PLFA and fungal : bacterial ratio were correlated with the soil characteristics examined. The relationship of total PLFA with soil characteristics was unimodal (i.e., maximum biomass at intermediate values of each soil characteristic) and quadratic equations produced a better fit than linear equations. Total PLFA was best correlated with percent carbon and nitroge n (Table 3-3). Relationships between fungal:bacterial ratio and so il characteristics were monotonic and simple linear regressions produced the best fits. Mois ture, nitrogen, and carbon contents were negatively correlated with fungal:bacterial ratio explaining 70%, 48%, and 26% of the

PAGE 51

41 variance, respectively. Conversely, C:N was positively correlated w ith fungal:bacterial ratio (Table 3-3). We examined differences in relative repr esentation of Gram – and Gram + bacterial groups as related to habitats a nd invasion-status. Gram + bact eria were highest in SA and AR habitats while the inverse was true for Gram – bacteria (Figure 33). All effects were significant in a two-way ANOVA of Gram – PLFAs (Table 3-4). Gram – PLFAs increased with invasion in SC and RU hab itats, decreased in AR habitats and had no change in ME and SA habitats. Nitrogen a nd carbon contents of th e soil were positively correlated with Gram – bacteria and nega tively correlated with Gram + bacteria. Moisture content was positively correlated only with Gram – bacteria and C:N had no significant relationship with ei ther Gram negative or positive bacteria (Table 3-5). To examine overall differences in microbi al community composition in relation to habitat and invasion status, we ordinate d the PLFA data by principle components analysis. The first two co mponents explained 22.3% and 17.2% of the variance, respectively. Microbial communities from different habitats were primarily separated by the first axis, although th e two wettest habitats were also separated from the remaining communities by the second axis (Figure 3-4). Within habitats, differences between invaded and non-invaded samples were primar ily on the first axis. In MANOVA, both habitat and invasion effects were significant (Table 3-6). The PLFA with high loadings on th e first two PCA axes (loadings > 0.3) included i11:0 3OH, 16:0, i16:0, 16:1 7c, 16:1 5c, 16:1 2OH, cy17:0, and 18:2 6c (Table 3-7). PCA 1 was negatively correlated with 16:0 (ubiquitous fa tty acid) and 16:1 7c (Gram – PLFA), and positively correlated with i 16:0 (Gram + PLFA) and 16:1 2 OH (Gram

PAGE 52

42 PLFA). PCA 2 was negatively correlated wi th i11:0 3OH, 16:1w5c (mycorrhizal PLFA, also often classified as Gram -), cy17:0 (G ram – PLFA, possibly an anaerobic marker), and positively correlated with 18:2 6c (fungal PLFA). Soil characteristics were not strongly correlated with PCA 1, but were correlated with PCA 2 (Table 3-8). Moisture content, followed by %N and %C, was strongl y negatively correlated with PCA 2. C:N was positively correlated with PCA 2. We then explored the effect of habi tat and invasion-status on the relative representation of these eight PLFAs with high loadings. Only habitat effect was significant for five PLFAs (i11:0 3OH, i16:0, 16:1 5c, 16:0, and cy17:0) whereas both habitat and invasion had significan t effects on the abundance of 16:1 7c, 16:1 2OH, and 18:2 6c. Two somewhat surprising patterns di d arise. A purported mycorrhizal PLFA, 16:1 5c, was higher in wetter communities and lowe r in drier communities (Figure 3-5). Similarly, a purported anaerobic marker, cy17:0, was higher in drier communities than in the wet ME and SC habitats. It is unknown what group i11: 0 3OH may represent, but its concentration appeared to decrease across the moisture gradient of habitats. The remaining fatty acids, i16:0 and 16:0 were variable across the moisture gradient. Three fatty acids, 16:1 2OH, 16:1 7c (both Gram – PLFAs), and 18:2 6c (a fungal PLFA) exhibited both a signi ficant main effect of habitat and also a significant interaction effect. There was a general decr ease in 16:1 2OH across the habitat moisture gradient (Figure 3-6a). At the extremes of the gradient, 16:1 2 OH tended to decrease with invasion, while the middle of the grad ient showed no effect of invasion on its relative abundance. In contrast, for 16:1 7c, there was no clear pattern across the habitat gradient as SA and AR habitats had th e lowest relative concentration of 16:1 7c (Figure

PAGE 53

43 3-6b). The effect of invasion ranged from a decrease in 16:1 7c with invasion in AR and ME habitats, an increase in RU habitats, and no discernable effect in SC and SA habitats. Finally, the fungal PLFA, 18:2 6c, generally increased across the habitat moisture gradient (Figure 3-6c), increasing wi th invasion in AR and ME habitats. The catabolic potential of microbial communities from ME, SC, SA, and AR habitats was examined as the ability of those microbes to metabolize 95 carbon substrates. Although the majority of subs trates were metabolized and there was no significant difference in the number of subs trates metabolized between habitats or invasion-status, rates of utilization differed. Substrate utilizati on patterns show lesser degree of discrimination among microbial co mmunities than PLFAs (Figure 3-8). A MANOVA indicated only a significant effect of habitat type (Table 3-9). Discussion Habitat Controls of Microbi al Community Composition Our results clearly indicate that habitat was the primary control of microbial community composition across the sites we examined. Microbial communities from different habitats differed in total PLFA, f ungal:bacterial ratio, rela tive representation of Gram – and Gram + bacteria, and several other individual PLFAs. Because plant community composition is not independent of th e soil characteristics in this study, it is difficult to separate the effects of plant species composition from soil environmental variables. The major environmental gradient across the habitats examined, however, was water content of the soil and was apparently the major control of microbial community composition. Total PLFA was highest at interm ediate moisture lowest at the wettest and driest habitats (ME and AR, Fi gure 3-1). At high water contents microbial activity is limited by oxygen availability, while at low matr ic potential microbial activity is limited

PAGE 54

44 by water availability (Griffin 1985). Fungi are better able to w ithstand low matric potentials and therefore tend to dominate in drie r soils and are at low levels or absent in water-logged soils (Bossio and Scow 1995, 1998, Nakamura et al. 2003). This pattern was corroborated by our result s for the fungal PLFA 18:2 6c that decreased from AR (driest) to ME (wettest) habitats (Figure 6c ). Such differential responses of fungi and bacteria resulted in a decrease in fungal:bacte rial ratio across the gr adient from low to high water contents, with moisture explaini ng 51% of the variance in fungal:bacterial ratio (Figure 3-2). Water content of the soil was also asso ciated with higher organic matter as indicated by the high carbon and nitrogen cont ents in waterlogged ME habitats. Thus although carbon and nitrogen availability was highest in ME habita ts, total PLFA was highest in communities with intermediate water, carbon and nitrogen contents. Heterotrophic soil microbes are generally t hought to be carbon(Alden et al. 2001, Ekbald and Nordgren 2002) or nitrogen-l imited (Hart and Stark 1997). Our results indicate oxygen availability becomes a grea ter limitation to microbial activity and biomass than substrate availability under saturated conditions, however, when not under saturated conditions, microbial biomass (as measured by total PLFA) is positively correlated with C and N. Fungi and bacteria however, responded in a predicted, linear fashion to increasing carbon and nitrogen av ailability. Fungi (as measured by 18:2 6c) and fungal:bacterial ra tio decreased with increasing C and N concentrations, and increased with C:N. Conversely, Gram – bact eria were positively correlated with carbon and nitrogen concentrations. Bacteria have higher nitrogen requirements and metabolic and growth rates than fungi (Griffin 1985). Th is leads to the pattern of fungi dominating

PAGE 55

45 in soils with low nitrogen with bacteria dom inating in soils with high nitrogen (Bardgett and McAlister 1999, Priha et al. 2001, Leckie et al. 2004). The increase in bacterial biomass with soil carbon and nitrogen concen trations in this study was primarily due to an increase in Gram – biomass. Gram – bact eria are able to quick ly respond to nutrient enrichment (Griffiths et al. 1999) and have b een shown to be higher in communities with higher soluble organic contents (Leckie et al. 2004). Gram + bacteria may be being outcompeted at high resource availability a nd others studies have shown this group to decrease with increasing carbon avai lability (Bossio and Scow 1998). These individual responses of particular functional groups led to an overall differentiation of microbial communities amon g habitats (Figures 3-4 and 3-7). There was greater separation of habita ts by PLFA than by substrat e utilization method. This trend has been seen in other studies (Buyer and Drinkwater 1997); however, we feel that the reduced separation discrimi nation of habitats by substrate utilization data could reflect an artifact of freezing soil prior to analysis. In effect, only the catabolic potential of microbes able to withstand the freezi ng and thawing process was examined. In examining the effect of storage on the catab olic potential of pl ant growth promoting bacteria, (Shishido and Chanway 1998) found that frozen soil samples were more similar to one another than to its fresh sample c ounterpart. Nevertheless, different habitats supported microbial communities distinct in both function and structure. Alteration of Microbial Communities by Invasion Overall, invasion-status had smaller, but significant, effects on soil microbial structure, as measured by PLFAs, than those of habitat (Figure 3-4, Table 3-6). Four of the five invaders (all but Schinus ) appeared to alter microbial community structure. We predicted that invaders that dominate the co mmunity to a greater extent in terms of

PAGE 56

46 biomass would have larger impacts on micr obial community composition. We did not find that trend. Instead, it was only the two ha bitats at the extreme ends of the gradient, AR and ME, that invasion had a significan t effect on PCA scores within habitats. Melaleuca converts wetlands dominated by Cladium (sawgrass, a sedge) to dense forest while Ardisia adds a monodominant shrub-layer wit hout altering the overstory. Chapin and D’Antonio (Chapin et al. 1996, D'Antonio et al. 1999) pred icted that invaders that change the structure of a community would ha ve larger impacts on that system. This would seem to explain the great Melaleuca effects, but does not explain the effects of Ardisia Kourtev et al. (2002) did find that on a local scale shrubs could alter microbial community composition. We believe that Ardisia may be altering microbial community composition due to its high density of carbohydr ate-rich roots (S. Bray unpublished data) that may be providing a different or larger source of exudate s in the top 10 cm of soil. Our data indicate that invasion alters mi crobial community co mposition not only in the zone of influence of a single plant as pr eviously shown (Kourtev et al. 2002) but also at a landscape level. There was a large variati on in the abiotic characteristics, such as soil moisture, that we examined across sites and likely large variation in other variables not measured in this study. This variation across sites was likely responsib le for the scatter in microbial community composition and the relatively low total variation (39.5%) explained by ordination of PLFA profiles. Conclusions Our data contribute to a gr owing literature base that demonstrates both soils beneath different plants support different mi crobial communities (Grayston et al. 1996, Westover et al. 1997, Grayston et al. 1998, Katajisto et al 1999, Priha et al. 1999, Kourtev et al. 2002, 2003, Bardgett and Walk er 2004) and microbial communities differ

PAGE 57

47 among habitats with contrasting types of ve getation (Waldrop et al. 2000, Myers et al. 2001, Leckie et al. 2004). Habitat is th e main control of microbial community composition, but invasion significantly modified microbial communities within a given habitat. The direction of these changes, however, was not predictable across habitat types. Changes in microbial community structure with invasion will likely be influenced by abiotic conditions in the community and the identity of the invader. As microbes are primary responsible for decomposition in most ecosystems, any alteration in microbial community structure and function with invasi on may result in altere d nutrient cycling and availability. The link between microbial co mmunity composition and process rates needs to be explicitly examined across plant communities.

PAGE 58

48Table 3-1: Location, mean annual temperatur e, mean annual rainfall, soil type and domi nant vegetation of the three sites in eac h habitat type. Site Location Mean Annual Temperature (C)a Mean Annual Rainfall (mm)a Soil Order Dominant Native Vegetation ME Habitats 1 2624’ N 8014’ W 26.04 1560 Histosol Sedges 2 2555’ N, 8026’ W 24.44 1487 Histosol Sedges 3 263’ N, 8033’ W 26.04 1560 Histosol Sedges SC Habitats 1 268’ N, 813’ W 23.83 1376 Entisol Tropical Hardwoods 2 262’ N, 8117’ W 23.83 1376 Entisol Tropical Hardwoods 3 2556’ N, 8118’ W 23.83 1376 Entisol Tropical Hardwoods RU Habitats 1 2937’ N, 8219’ W 20.33 1228 Alfisol Deciduous Softand Hardwoods 2 287’ N, 829’ W 22.83 1137 Mollisol Deciduous and Evergreen Hardwoods 3 2846’ N, 8112’ W 22.67 1228 Alfisol Evergreen Hardwoods and Palms AR Habitats 1 2937’ N, 8217’ W 20.33 1228 Ultisol Deciduous and Evergreen Hardwoods 2 2933’ N, 8221’ W 20.33 1228 Alfisol Deciduous and Evergreen Hardwoods 3 2940’ N, 829 W 20.33 1228 Alfisol Deciduous and Evergreen Hardwoods, Softwoods SA Habitats 1 2936’ N, 8219’ W 20.33 1228 Alfisol Grasses, forbs, 2 2923’N, 951’ W 21.78 1113 Vertisol Grasses, forbs 3 2922’N, 952’ W 21.78 1113 Vertisol Deciduous Hardwoods a Mean annual temperature and rainfall data are 30-year (1971-2000) means from the n earest weather monito ring station of the National Oceanic and Atmo spheric Administration.

PAGE 59

49 Table 3-2: Soil characteristics of th e five habitats examined. Means + S.D, different letters within the same column signify significant differences by Tukey HSD. Df = 1, 25. Habitat % Moisture %C %N C:N ME 79.29 + 18.66 a 32.82 + 8.81 a 2.05 + 0.71 a 16.54 + 1.98 c SC 40.74 + 8.82 b 8.81 + 9.49 bc 0.391 + 0.18 b 20.26 + 10.11 bc SA 43.28 + 5.49 bc 9.37 + 3.68 c 0.432 + 0.15 b 21.91 + 6.69 bc RU 28.34 + 7.68 c 12.08 + 2.02 b 0.377 + 0.045 b 32.61 + 7.39 a AR 22.20 + 9.03 c 3.80 + 1.42 d 0.158 + 0.076 c 27.56 + 10.86 ab Table 3-3: Relationship of total PLFA and fungal:bacterial ratio with % moisture, %C, %N and C:N using quadratic fit for to tal PLFA and simple linear regression for fungal = biomass ratio (n = 30). All significant models had p-value < 0.005 except %C vs. fungal : bacterial ra tio which was significant at p = 0.03. Total PLFA Fungal : Bacterial Ratio R2 Direction r2 Direction % Moisture 0.32 Convex 0.70 Negative % C 0.45 Convex 0.26 Negative % N 0.48 Convex 0.48 Negative C:N NS NS 0.52 Positive Table 3-4: ANOVA results for the effects of ha bitat, invasion-status and their in teraction on the relative representation of Gram – and Gram + biomarkers. Gram Gram + Effect Df F P F P Model 9, 20 26.07 <0.0001 17.96 <0.0001 Habitat 4 45.69 <0.0001 38.13 <0.0001 Invasion 1 4.23 0.0529 2.27 0.15 Hab x Inv 4 11.91 <0.00001 1.72 0.19 Table 3-5: Correlation between %N, %C, and %moisture and the relative representation of Gram – and Gram + biomarkers (n= 30). All comparisons were significant at p <0.01. Gram Gram + r2 Direction r2 Direction %C 0.37 Positive 0.32 Negative %N 0.31 Positive 0.19 Negative %moisture 0.21 Positive NS NS

PAGE 60

50 Table 3-6: Results of MANOVA for the effects of habitat, invasion and their interaction on principle component axes 1 and 2 scor es from the ordination of 23 PLFAs. Effect DF F P Model 9, 20 28.4 <0.0001 Habitat 4 61.6 <0.0001 Invasion 1 5.74 0.03 Hab Inv 4 0.92 0.47 Table 3-7: Loadings for the first two axes in a principle components analysis of 23 common PLFAs extracted from soil sample s. Bold face indicates loadings > 0.3. Fatty Acid PCA 1 Loading PCA 2 Loading Gram + i15:0 0.1311 0.1373 a15:0 0.1980 -0.0118 i16:0 0.2952 0.1035 i17:0 0.1958 -0.1475 a17:0 0.2146 -0.1363 16:0 10ME 0.2456 -0.2619 Gram 15:0 -0.1103 0.2806 16:1w7c -0.2922 -0.2198 16:1w5c 0.0164 -0.3532 cy17:0 -0.2140 -0.3141 16:1 2 OH 0.3478 0.0410 17:0 10ME 0.2610 0.1541 18:1w7c -0.2419 -0.1602 cy19:0 0.1892 0.2399 Fungi 18:2w6c -0.0881 0.3527 18:1w9c -0.1377 0.2555 Microeukaryote 20:4 0.0797 0.1116 Actinomycete 18:0 10ME 0.2080 0.0591 No classification i11:0 3OH 0.0480 -0.3176 15:0 3OH 0.1190 0.0696 14:0 -0.1978 -0.0207 16:0 -0.3237 0.1680 18:0 -0.2192 0.2497

PAGE 61

51 Table 3-8: Correlations (r) of soil character istics with the first two axes from the ordination of 23 PLFAs. Soil characteristic PCA 1 PCA 2 % N -0.135 -0.634 % C -0.250 -0.538 % Moisture -0.025 -0.737 C:N -0.331 0.388 Table 3-9: The results of MANOVA for the effects of habitat, invasion and their interaction on principle component ax es 1 and 2 scores derived from an ordination of metabolic activity of soil microbes on 95 substrates. Effect DF F P Model 7, 16 2.91 0.037 Habitat 3 6.67 0.004 Invasion 1 0.32 0.57 Hab x In 3 0.007 0.99

PAGE 62

52 0 20 40 60 80 100 120 140 160 MESCSARUAR Habitatbiomass (nmol g-1) Figure 3-1: Mean soil microbia l community total PLFA (nmol g-1 + S.D.) for five habitats prone to invasion by five non-native speci es. Black bars = non-invaded; white bars = invaded. Habitat types marked by different letter s are significantly different from one another by Tukey’s HSD (alpha = 0.05). B B A A A Wet Dry

PAGE 63

53 Figure 3-2: Fungal:bacterial ratios (+ S.D.) for five habitats prone to invasion by five non-native species. Different letters over bars indicat e significant differences between habitats as determined by Tukey’ s HSD (alpha = 0.05). Black bars = non-invaded; white bars = invaded. 0 0.5 1 1.5 2 2.5 MESCSARUAR HabitatFungal : Bacterial RatioC C B A A Wet Dry

PAGE 64

54 Figure 3-3: Relative represen tation (% of total nmoles extracted) of PLFAs across habitats and invasion-status. a) Gram + and, b) Gram – PLFA biomarkers (mean + S.D.) Black bars = non-invaded; white bars = invaded. Different letters over bars indicate significan t differences between habitats as determined by Tukey’s HSD (alpha = 0.05) after pooling invaded and noninvaded areas within a habitat. Si gnificant differences between invasionstatus within a community by Bonferroni -corrected t-tests are indicated by asterisks (* = p < 0.05; ** p < 0.01). 0 5 10 15 20 25 30 MESCSARUAR HabitatRelative abundance Gram PLFA (%mol) 0 10 20 30 40 50 MESCSARUAR HabitatRelative abundance Gram + PLFA (%mol)B B B A A Wet Dry Wet Dry a A A A B B * b

PAGE 65

55 PCA 1 (22.3%) -6-4-2024 PCA 2 (17.2%) -4 -3 -2 -1 0 1 2 3 4 Schinus invaded Schinus non-invaded Melaleuca invaded Melaleuca non-invaded Ardisia invaded Ardisia non-invaded Sapium invaded Sapium non-invaded Ruellia Invaded Ruellia non-invaded Figure 3-4: Mean (+ S.D.) principle components scores by habitat and invasion-status from ordination of 23 most common microbial PLFAs found in soil samples. Downward triangles = ME, circles = SC diamonds = SP, upward triangles = RU, squares = AR habitats. Open sy mbols = non-invaded, closed symbols = invaded.

PAGE 66

56 Figure 3-5: Mean (+ S.D.) relative abundance of five PLFAs that had high loadings on the principle component axes that show ed significant habita t effects, but no invasion effect. Solid bars = ME, op en bars = SC, diagonal stripes = SA, stipled bars = RU, horizontal stripes = AR. Different letters over bars indicate significant differences between habitats as determined by Tukeys HSD (alpha = 0.05). 0 5 10 15 20 25 30 35 40 i11:0 3OH16:0i16:016:1w5ccy17:0 Fatty AcidsRelative abundance (%mol) Ubiquitous Gram + Gram AB A A A A A A A A BC BC BC BC C C C C B B B B B B B B

PAGE 67

57 Figure 3-6: Mean (+ S.D.) relative abundance of PLFAs with high loadings (>0.3) in the PCA analysis showing significant habitat and invasion effects. a) 16:1 2OH, Gram – biomarker, b) 16:1 7c, Gram – biomarker, and c) 18:2 6c, fungal biomarker. Black bars = non-invaded, ope n bars = invaded. Different letters over bars indicate significant differences between habitats as determined by Tukey’s HSD (alpha = 0.05) after po oling invaded and non-invaded areas within a habitat. Significant differences between invasion-status within a community by Bonferroni-corrected t-test s are indicated by asterisks (* = p < 0.05; ** p < 0.01). 0 1 2 3 4 5 6 7 8 MESCSARUAR HabitatRelative abundance 16:1 2OH (%mol)A AB BC C CWet Dry a

PAGE 68

58 Figure 3-6. Continued 0 2 4 6 8 10 MESCSARUAR HabitatRelative abuncdance 18:2w6c (%mol) 0 1 2 3 4 5 6 7 8 MESCSARUAR HabitatRelative abundance 16:1w7c (%mol)A AB* *BC C Wet Dry Wet Dry b c C B B A A

PAGE 69

59 PCA 1 (26%) -10-8-6-4-20246810 PCA 2 (11%) -6 -4 -2 0 2 4 6 Figure 3-7: Mean (+ SD) pr inciple components scores by habitat and invasion-status from the ordination of metabolism of 95 carbon sources by soil microbial communities. Symbols are as described in Figure 3-4.

PAGE 70

60 CHAPTER 4 LINKS BETWEEN LITTER QUALITY, DECOMPOSITION AND MICROBIAL COMMUNITY COMPOSITION ON NATIV E AND NON-NATIVE PLANT LITTER Introduction The decomposition of plant litter is de pendant on climate, substrate quality and decomposers. Within a given climate, leaf chemical composition is an excellent predictor or decomposition. The optimal measure of li tter quality as a predictor of decomposition depends upon the system examined and the le ngth of the study. Nitrogen content and carbon to nitrogen ratio (C:N) are often pos itively and negatively correlated with decomposition rates, respectively, in studies of short duration and in species or soils with a high percentage of more labile carbon fr actions (Flanagan and van Cleve 1983, Pastor et al. 1987, Taylor et al. 1989). Other studies have shown that carbon quality, as measured by proportional representation of labile and recalcitrant s ubstrates, to be the major factor limiting decomposition (Meente meyer 1978, Berg and Taum 1991, Hobbie 1996). Lignin is one such recalcitrant co mpound degraded by only a limited number of organisms, predominately brownand white-ro t Basidiomycete fungi, can degrade it. In many cases, lignin:N is the best integrator of litter quality for the duration of its decomposition (Melillo et al. 1982). Suites of litter quality characteristics te nd to be correlated with one another as a function of habitat quality. In low-resour ce habitats, species tend to have low nutrient contents, high carbon-based st ructural defenses, and longlived leaves (Chapin 1980). Conversely, species from high-resource habita ts have higher nutrient contents, lower

PAGE 71

61 levels of defense, and shorter-lived leaves (Chapin 1980, Grime et al 1996, Wright et al. 2001). Moisture levels in soil can indirectly affect resource availabi lity. At high soil moisture contents, oxygen becomes limiti ng and decreases decomposition (Haynes 1986). These lowered oxygen levels lower nut rient availability and reinforce slow decomposition rates by leading to suites of leaf traits that result in slower decomposition rates. While the role of substrate quality in deco mposition is well established, the role of the microbial community has been less studied and, in general, is treated as a ‘black box.’ It seems likely, however, that different comm unities of microbial decomposers would be found on litter of different quality and that microbes are probably responding to the same litter quality factors that control decompositi on rates. As bacteria have higher nitrogen requirements and faster grow th and reproduction rates than fungi, bacteria might be expected to dominate on litte r with high nitrogen and low C:N. Fungi, conversely, due to their slower growth, lower nutrient requireme nts and ability to decompose lignin, might be expected to dominate litter with high C:N a nd lignin:N. Changes in litter quality over the duration of decomposition will also likel y control the succession of microbes on the litter. Early studies of litter decomposers have shown that th e ratio of fungi to bacteria increased with increasing C:N (Witkamp 1963, 1966). Previous studies of decomposer microbes have primarily relied upon culture-based techniques. As only an estimated <1% of soil microbes are culturable (Torsvik et al. 1996, Atlas and Bartha 1998), culture-based techniques are limited in their ability to examine the microbial community. Phospholipid fatty acid (PLFA) analysis offers a nonculture based technique that provides a measure of the living microbial community.

PAGE 72

62 PLFA analysis has the advantage of pr oviding information on specific groups of microbes through the use of signature fatty acids as well as providing an overall “fingerprint.” Thus, not onl y can functional or taxonomic gr oups be examined, but also the similarities (or dissimilarities) of different microbial communities can be compared. We applied PLFA analysis as a new appr oach to examine decomposer microbes on different plant litters over time. Differences between plant species, by wa y of their chemical composition affecting litter quality, are known to affect decom position rate and we hypothesize it will also determine the composition of the decomposer community. Plant community composition is changing globally with the spread of non-na tive species and such changes can lead to changes in ecosystem function (Vitousek a nd Walker 1989, Mack and D'Antonio 2003, Allison and Vitousek 2004). While a consen sus predictive framework for determining which invaders will be more likely to have ecosystem-level impacts has not been developed, invaders that are qualitatively different from nativ e species are believed to be more likely to alter ecosystem processes (C hapin et al. 1996, D'An tonio et al. 1999). Those species that have traits that overla p with native species are expected to have limited or slower impacts on ecosystem tra its (Mack and D'Antoni o 2003). As litter quality is a continuum in whic h both native and exotic species are distributed, we would expect that the potential impacts of e xotic species invasion on decomposition and microbial community composition would be de termined by its relative position on this continuum. The evolution of increased co mpetitive ability (EICA) hypothesis predicts that due to the lack of herbivore pressure in their new range, invasi ve plants evolve to

PAGE 73

63 invest less in defense (Blossey and Notzold 1995) potentially resulting in higher quality litter in invaders. The primary goal of this study was to examine the links between decomposition and microbial community composition through litter quality. To that end, we examined the decomposition of litter of twenty species of plants, both native and exotic, from a variety of habitats varying in litter quality, in a common site. By including a variety of plant species from a variety of habitats, we attempted to sample a broad range of litter quality. We hypothesized that species identity of the litter (hereaf ter “litter species”) significantly affects decomposition rates in rela tion to the litter species’ typical habitats, leaf longevity, and other leaf functional tr aits. We hypothesized that the same litter quality factors that control decomposition in th is site should also control the composition of the microbial community. Litters of higher quality should support a higher total biomass with proportionally more bacteria, especially Gram-negative bacteria, while fungi should dominate litters of lower quali ty. The difference in microbial community composition between plant litters should decrease as decomposition proceeds and remaining litter is dominated by more recalcitrant substrates. Finally, because we hypothesize that the same litte r quality factors control both decomposition rate and microbial community composition, the corr elation between them should be high. Methods Litter Collection and Experimental Design We chose 20 plant species with the goals of representing 1) a broad range of litter quality as measured by leaf ha bit and lifespan, nitrogen c ontent and carbon fractions and 2) non-native and native species from a variety of habitats (Tab le 4-1). Leaf litter of each species was collected from a minimum of five sites and a minimum of five individuals

PAGE 74

64 per site between the months of September a nd December. Only leaves that fell freely from shaken plants or had a clear abscissi on zone were collected; obviously green leaves and leaves with heavy herbivore damage we re excluded. Litter was pooled by species and dried at room temperature for a mini mum of four weeks prior to litter bag construction. Litter bags were constructed of 1-mm fiberg lass window screen and filled with 5 g of air-dried litter. Subsamples of ai r-dried litter were we ighed, dried at 60oC and reweighed to determine initial dry mass of litter bags. A total of 30 litter bags per species were made to allow for 5 replicates at each of 6 collection dates. We strung a total of 20 litter bags (1 per species) onto a nylon line for each harvest date. Litter bags were placed in a common hardwood-dominated forest (2940’N, 829’W; mean annual rainfall 1200 mm; mean annual temperature 20.3C) in February 2004. Six lines of litter bags, representing the six sampling dates, were placed at 5 randomly determined locations within the study site. Lines radiated out from a central flag on a litter layer dominated by Quercus nigra Quercus hemispherica and Pinus taeda One line from each replicate was collected at 28, 57, 112, 180, 238, and 319 days. Litter bags were transported to the lab where th e exterior of the bags were brushed free of adhering soil. Roots, soil, invertebrates a nd frass were removed and the remaining litter weighed. A subsample of litter was immediat ely frozen for later phospholipid analysis; the remaining litter was weighed, dried at 60 C and reweighed after 5 days of drying. Dried samples were ground on a Wiley Mill (Thomas Scientific) through a 40 mesh. Carbon and nitrogen contents of initial l itter and litter bags was determined by combustion on an elemental analyzer (Costech, Inc., Valencia, CA). Multiplying

PAGE 75

65 nitrogen concentration by litter mass and dividing by the initial mass of nitrogen determined percent initial nitrogen remaini ng. Carbon fractions were extracted with increasingly acidic solutions (van Soest 1963) usi ng an Ankom 220 Fiber Analyzer (Ankom, Macadon, NY). Neutral detergent rem oved non-polar extracts (fats, oils waxes) and soluble cell contents ( carbohydrates, starch, non-bound pr oteins) and comprises the non-polar fraction (NPE). Dilute acid de tergent removed hemi-cellulose and bound proteins comprising the water-s oluble fraction (WS). Cellulo se, the acid-soluble fraction (AS), was separated from lignin with 72% H2SO4. The lignin fraction was corrected for ash content by ashing samples at 500 C after the sulfuric acid step. Carbon fractions were expressed as a percentage of total mass. Phospholipid Fatty Acid Analysis Prior to analysis, frozen samples were ground with a Wiley Mill to pass through a 40-mesh. All glassware was heated at 500C for 5 hours and PTFE caps were rinsed with hexane to remove organics. Phospholipids were extracted from litter using the methods of Wilkinson et al (2002). Lipids were extracted from 250 mg lit ter samples in two 30minute baths in a 37C water bath in a single-phase phosphate buffered methanechloroform solvent. After each extraction, the supernatant liquid was transferred to a second test tube. The phase of the second test tube was broken by the addition of 4 ml of chloroform and 4 ml of buffer. After vorte xing, the phases were allowed to separate overnight. The organic phase was fractio nated on silicic acid columns and the phospholipid fraction collected a nd transesterified to fatty ac id methyl esters (FAMEs) by mild alkaline methanolysis (Findlay and Dobbs 1993). Samples were resuspended in hexane with 19:0 fatty acid as an internal standard and analyzed with an Agilent

PAGE 76

66 Technologies 6890 gas chromatograph (Palo Alto CA) with a 25m Ultra 2 phenyl methyl silicone column. The temperature program increased from 170C to 270C at 5C per minute. Peaks were identified with MIDI peak identification software (MIDI, Inc., Newark, DE) and by co-elution with standards. Fatty acids were defined in terms of total number of carbon atoms : number of double bonds. The position of the double bond from the methyl end of the mol ecule is signified by the symbol followed by the carbon position; cis and trans geometry are referred to by “c” and “t.” The prefixes “i” and “a” signify and iso and anteiso branching; “cy” signifies a cyclopropyl fatty acid. Statistical Analysis In order to compare litter decay rates among species in relation to litter quality, decomposition rate constants (k) for each litter type were determined using a negative exponential model: ln(Xt) = ln (X0) – kt where Xt equals the amount of mass left at time = t, X0 is the initial mass of litter, and t = time in years (Olson 1963). In order to improve normality, k’s were log transformed and subjected to ANOVA examin e differences among species. Tukey’s HSD was used to examine differences between sp ecies. To examine difference between leaf lifespan categories and habitats, species aver ages were subjected to a Kruskal-Wallace rank test, as there were uneven sample si zes among groups and unequal variances. The relationship between decomposition rate and lit ter quality was examined in two ways. The individual effects of mean initial %N, %NPE, %WS, %AS, %lignin, C:N and lignin:N on mean decomposition co nstants were examined in simple regressions. As are auto-correlated and cannot be consider as inde pendent of each other, we sought to create an integrative measure of overall litter qua lity. Therefore, we ordinated the average

PAGE 77

67 initial values of %N, C:N, %NPE, %WS, %AS, %lignin, and lignin:N for each species in a principle components analysis The first axis from that PCA explained 56% of the variance in litter quality data and we use this axis as a proxy for a litter quality ranking and refer to it as the “leaf chemistry axis ” herein. Average species k’s were then regressed against this leaf chemistry axis. Similarity of microbial communities was examined using both constrained (canonical correspondence ordination, CCA) and unconstrained (principle components analysis, PCA) ordination. We performe d these two types of ordinations because comparison of their result is informative. Constrained ordination is best used in ordinations when a set of independent e nvironmental variables is believed to be structuring the community. As we hypothe size that microbial communities should be structured by litter quality characteristics, c onstrained ordination allo ws us to directly examine how litter quality is structuring micr obial communities as the ordination axes are constrained to these independe nt variables. Unconstrai ned ordination, conversely, extracts the major variation in community data irrespective of any environmental variables. Therefore, if a PCA and CCA give similar results and explain a similar percentage of variance in the community da ta, it is assumed that the environmental variables examined are primarily responsible for the structuring of the community. Additionally, as CCA is subject to the same limitations as multiple regression and CCA can be sensitive to noise in the envi ronmental matrix (M cCune 1997), use of unconstrained ordination along with constr ained ordination is often suggested (McGarigal et al. 2000).

PAGE 78

68 We ran PCA and CCA ordinations on all samples simultaneously and on samples from each time point individually. We performe d both types of analyses so that we could examine overall changes and controls of microbial community composition over decomposition and the differences and primary controls on microbial community composition at a given sampling date. Only th e 17 PLFAs that compri sed at least 1% of the total PLFAs extracted were included in the multivariate analyses. The PCA analysis was performed only on the main matrix of th e 17 PLFA values. The CCA analysis was performed on the main matrix and a second matrix containing the 7 measures of litter quality, initial mass remaining (%IMR), and percent moisture. Individual PLFAs were log-transformed to improve normality. MANOVA was performed on PCA and CCA scores to determine the eff ects of time, litter species and litter species*time. The relationship between litter quality and functional groups of microbes (terminally branched Gram-positive bacterial PLFAs: i14: 0 + i15:0 + a15:0 + i16:0 + a16:0 + i17:0 + a17:0; monounsaturated Gram-neg ative bacterial PLFAs: 16:1 7 + 17:1 7 + 18:1 7; cyclopropyl Gram-negative bacterial PL FAs: cy17:0 + cy19:0; and fungi: 18:2 6), fungal:bacterial ratio (18:2 6) / (i15:0 + a15:0 + 15:0 + i16:0 + 16:1 5c + i17:0 + a17:0 + cy17:0 + 17:0 + 18:1 7c + cy19:0), and total biomass was examined in simple regressions for each time point. When nece ssary, PLFAs were transformed to achieve normality. All regressions, analysis of variance, and non-parametric tests were performed in JMPIN 4.0 (SAS Institute 2000, Cary, NC) wh ile PCA and CCAs were performed in PCORD 4.20 (MJM Software, Gleneden Beach, OR 1999).

PAGE 79

69 Results Decomposition of Litter Most species showed two periods of rapid decomposition—during the initial 28 days and again between 112 and 238 days, co rresponding with the wet season (Figure 41). Decomposition rate constants (k) ranged from < 0.4 to nearly 2.0 yr-1 with significant differences among species (p < 0.0001, F19,80 = 23.47; Table 2). Leaf lifespan class of the litter species had no effect on decomposition rate ( 2 = 2.19, df = 2, p = 0.33). Habitat affiliation of the litter species, however, did have a significant effect on decomposition rate ( 2 = 9.72, df= 2, p = 0.0077) such that decomposition rate was lowest in species from dry habitats and higher for litter speci es from wetter habitats. However, some individual species from “wet” habitats (e.g. Juncus Taxodium and Acer ) had relatively slow decomposition rates (T able 4-1 and 4-2). There was also a significant difference in decomposition constants between native and non-native species (F1,98 = 47.09, p < 0.0001); this result, however, should be interpreted with caution as the result would likely change with diffe rent representative native and exotic species. When the non-na tive species are co mpared with native dominants in the plant communities they invade, Ruellia Causurina Schinus and Sapium have higher decomposition rates, Imperata had comparable rates with native dominants, and Ardisia which can co-occur with severa l of the deciduous species with moderate decomposition rates, was comparab le with those species and faster than Pinus with which it also co-occ urs (Tables 4-1 and 4-2). We also examined the relationship between initial litter quality and decomposition rates. Of the seven measures of litter qua lity examined, the concentration of non-polar extracts and lignin:N individua lly explained the greatest am ount variation. Lignin and

PAGE 80

70 nitrogen contents as well as the relative availability of nitrogen (C:N), were also significantly correlated with decomposition rate (Table 4-3). According to a backwards stepwise regression with %N, %NPE, %lignin, C:N and lignin:N, the variables %lignin and C:N were retained and produced the following model: Ln(k) = 0.650 – 0.0275(%lignin) – 0.00907(C:N) This model explained 60% of the variation (F2,17 = 12.76, p = 0.0003). Multiple regression models can be problematic because of collinearity of predictor variables and the order in which these variables are added or removed in stepwise regression. As litter characteristics are known to co-vary, they were analyzed by principle components analysis. The first PCA axis from this analysis explained 55.6% of the total variance (eigenvalue = 3.89), with high loadings of C:N, %NPE, %WS, and lignin:N. Thus PCA 1 can be interpreted as a litter quality axis in which low values indicate high litter quality i.e., low C:N, lignin:N, %WS, and high %NPE ) and high values indicate low litter quality (Table 4-4). Decomposition rate was negativ ely correlated with the leaf chemistry axis (F1,18 = 17.57, p = 0.0005, r2 = 0.49). Nitrogen concentration of litter increased ove r time (Figure 4-2a). Approximately half of the species showed no change or net in crease of nitrogen relative to the initial total with time, indicating net immobilization of n itrogen. The three species with the fastest decomposition rates ( Ruellia Sapium and Taxodium Table 4-2) show ed the largest nitrogen mobilization (= net nitrogen loss, Fig. 4-2b). Immobiliza tion varied over the course of the study, with several species s howing increased mobilization at time points with higher decomposition (Figure 4-1 and 4-2b).

PAGE 81

71 Microbial Community Composition We examined the structure of microbial communities of eleven species at 28, 57, and 238 days by multivariate analysis of PLFAs. We performed both PCA and CCA analysis on samples from all sample dates simultaneously and then each sample date individually. The PCA analysis of all sample dates explained a grea ter proportion of the total variance (52% in the first 2 PCA axes) th an the CCA analysis (32.3% of variance in first 2 CCA axes, Figure 4-3, a-b). This i ndicates that while the leaf litter quality variables we measured were important, other, unmeasured variables were also important to the structuring of microbial communities. In the PCA analysis, there is a large cluste r of samples with poor litter quality from t = 1 and 2 samples dates and a smaller clus ter composed of highe r litter quality from primarily t = 2 samples and two t = 1 samples ( Ruellia and Causurina ). By t = 5, microbial communities were beginning to conve rge as indicated by the ellipse indicating 9 of the 11 litter species at t = 5. The CCA analysis allows differentiation of PLFA composition (i.e., microbial community composition) on the basis of variation in litter quality. Like the PCA, CCA axis 1 differentiates microbial communities of the dry periods (t = 1 and 2) with lo w CCA-1 scores, from t = 5 mi crobial communities from wet periods with high scores. CCA axis 2, in contrast, differentiates microbial communities in relation to litter quality, such th at species with high scores on the leaf chemistry axis are associated with low CCA axis 2 scores (Fig 4-3b), especially among samples from t = 1 and t = 2. In MANOVA of PCA and CCA axes 1 and 2 scores, time, species and time species effects were all significant (Table 4-6). When ordination analyses were done for each sampling period separately, the first two axes of the PCA analyses of individual sampling dates again de scribed more of the

PAGE 82

72 total variation than the first two axes of th e CCA analyses (time 1: 50.5% vs. 36.6%, time 2: 51% vs. 35.9%, time 5: 64% vs. 38.2%) Constraining the axes by litter quality variables in CCA altered the distribution of litter species in species-score space (Figure 44, a-f). In the PCA of t = 1, Ruellia was most different from other litter species, but there was no obvious pattern of microbial community composition in relation to litter quality (Figure 4-4a). The first axis was composed of high loadings for two Gram PLFAs and the fungal PLFA, while the second axis wa s composed of negative Gram + PLFA loadings (Table 4-7). The simultaneous analysis of litter quality in the CCA for t = 1 microbial communities isolated the species with high litter quality, Ruellia Sapium and Schinus, from the remaining samples on the basi s of their higher non-polar extract and nitrogen contents which were negativel y loaded on CCA axis 1(Figure 4-4b). Pinus and Juncus were very different from other litter sp ecies due to their high amount of initial mass remaining, and high lignin:N and wate r-soluble fiber content, respectively. In the PCA of t = 2 microbial communiti es, Gram + PLFAs were again negatively loaded on the second axis while saturated fatty acids, fungi and tw o Gram – PLFAs were positively loaded on the first axis (Table 4-7). Inclusion of litter quality in the CCA for t = 2 samples, did not drastically change the relative position of litter species to one another (Figure 4-4, c-d). The primary diffe rence between the t = 2 PCA and CCA was a greater separation of Acer and Ardisia microbial communities from one another and a better separation of poor litter quality species, Aristida Imperata and Juncus, from the remaining litter by CCA. The separation of microbial communities in the CCA was

PAGE 83

73 driven by the positive correlation of NP E on CCA axes 1 and 2, and the positive correlation of WS, AS, and initial mass remaining on CCA axis 1 (Table 4-7). At t = 5, there was generally less separation of microbial communities on high and moderate litter quality (Figur e 4-4, e, f). Fungi and some saturated PLFAs were negatively loaded on PCA 1 and two Gram – PLFAs and one Gram + PLFA were negatively loaded on PCA 2 (Table 4-7). Inclusion of litter quality in the CCA better separated low quality litters Pinus Aristida and Juncus from one another and the remaining litter species on the basis of in itial mass remaining and C:N (Figure 4-4f). We explored the effect of litter quality on microbial groups in simple regressions at each collection date. In gene ral, the strength of relati onship between litter quality variables and microbial groups increased across time (Tables 4-8 through 4-13). The fungal PLFA and total PLFA, howev er, had strongest correlations with litter quality at t = 2. Three litter quality factors, %N, C:N and %IMR, explained the greatest proportion of variance across all microbial groups. C:N a nd %IMR were negatively correlated with monosaturated and cyclopropyl Gram – bacter ia, Gram + bacteria, fungi, and total PLFA (nmol g-1) and positively correlated with fungal : b acterial ratio. The opposite trends were true for %N. Some surprising trends relating to carbon fract ions were identified. Lignin content was not correlated with fungal PLFAs, but was positively correlated with Gram – and Gram + bacteria and total PLFA at t = 5, a lthough these correlations were weak. On the other hand, the relationship between lignin:N and microbial groups was as expected, being negatively correlated with bacter ial groups and positivel y correlated with fungal:bactieral ratio. The relationship be tween water and acid soluble fractions and

PAGE 84

74 different microbial groups was not consis tent and changed depending upon collection date examined. For example, water and acid soluble fractions were positively correlated with Gram + bacteria at t = 2, but negatively correlated at t = 5. Similar to individual measures of litter quality, the strength of correlation betw een the leaf chemistry axis (Table 4-4) and microbial groups increased with tim e (Table 4-14). Linking Litter Quality, Decompos ition and Microbial Communities Just as the leaf chemistry axis wa s correlated with decomposition rate (r2 = 0.49) and microbial groups (r2 up to 0.77), it was highly corre lated with microbial community principle components axes. Th e strength of the correlation between leaf chemistry and microbial community composition increased w ith time (Table 4-15). Leaf chemistry explained 88% of the variation in the first mi crobial axis from the fifth collection date and 65% of the variation in the first mi crobial axis from all samples ordinated simultaneously. Microbial community compositi on also explained the greatest amount of variation in decomposition rate consta nts (k) of any measures explored (r2 = 0.76 for t = 5 microbial communities). As with the rela tionship between microbial community and litter chemistry, the strength of the co rrelation between microbial community and decomposition increased with time (Table 4-15). Discussion In this study we endeavored to examine the role of plant litter quality in the structuring of microbial communities and how those communities changed over time. In our experimental design of us ing field common gardens, en vironmental variables other than litter quality were standardized acro ss litter species, although the environment did not stay constant. In addition to change in litter quality as decomposition progressed, the background moisture availability changed as th e rainy season started in the middle of the

PAGE 85

75 experiment. Indeed, some of the tempor al differences in microbial community composition as detected by PLFA biomarke rs appear to be better explained by differences in litter moisture than by chemical differences of the litter. Hence, sampling date appeared to have the greatest effect on microbial community composition (Figure 43a), but there were also significant differe nces among litter species due, in part, to differences in litter characteristics (Figur e 4-3b). While non-pola r extracts, lignin and lignin:N (Table 4-3) were the best individual predictors of decomposition rate, non-polar, acidand water-soluble carbon fractions and C:N (Figure 4-3b) were most important in structuring microbial communities. What was particularly novel about this study was the discovery that the best correlate of decomposition rate was not a measure of litter quality, but the composition of the microbial community (r2 = 0.76 for correlatio n of PCA1 of the microbial community at t = 5, Table 4-15). Factors Controlling Microbial Community Composition Time, litter species and litter species tim e all had a significant effect on microbial community structure. The most obvious effect is that of time with samples collected at t = 5 being most different from other sampling dates (Figure 4-3). The t = 5 samples were collected on October 5, 2005 w ithin four weeks of hurricanes Frances and Jeanne which together produced 325 mm of rain in th e area and these samples had much higher moisture contents than samples collected at t = 1 and 2. The separation of the t = 5 group is due to an increase in Gram-neg ative bacterial PLFAs cy17:0, cy19:0, and 18:1w7c and Gram-positive bacterial PLFAs i15:0 and i16:0 in those samples. Other studies have shown increases in Gram-positive, terminally branched, saturated fatty acids such as i15:0, a15:0, i17:0, and a17:0 in soil and litter microbial communities under flooded conditions (Bossio and Scow 1998, Na kamura et al. 2003). Cyclopropyl fatty

PAGE 86

76 acids cy17:0 and cy19:0 have additionally been suggested as biomarkers for anaerobic bacteria (Guckert et al. 1985, Vestal and White 1989), but see (Parkes and Taylor 1983, Bossio and Scow 1998). In the only anal ysis of microbial communities of early decomposition of plant litter in upland conditi ons of which we are aware, Wilkinson et al (2002) found that a15:0 and cyclopropyl fatty acids were generally higher in regularly watered samples, although the effect was depe ndant on litter species examined. In this study, bacterial dispersal and activity were lik ely limited in the first two collections by low moisture, while the saturating conditions of the fifth collection resulted in increased bacterial biomarkers overall and espe cially putative anaerobic biomarkers. While time had the largest effect on microbial communities in PC and CC analyses, litter species and litter species*time interaction effects were also significant (Table 4-6). Microbial communities on poor qual ity litter changed less from sample date to sample date than did microbial communities on litter with high initial litter quality (Figure 4-3a). Two alternative hypotheses may explain this patt ern. Resource availability may be high enough in high quality litter that decom position and microbial succession can proceed even under low moisture levels. Alternatively (but not nece ssarily exclusively), greater changes in resource availability in high quality litter may be responsible for shifts in microbial community composition. Ordination of each sampling date indivi dually shows litter species and quality effects on microbial community structure. In early decomposition under low moisture levels, high concentration of non-polar ex tracts and to a le sser extent nitrogen concentration separated high quality litter mi crobial communities from moderate to low quality litters with higher wate r and acid soluble fiber conten ts that retained a greater

PAGE 87

77 proportion of the initial mass expressed as %IMR (Figure 4-4 b, d). The only litter quality characteristic that was important to structuring microbial communities at all samples dates was %IMR. Under the higher moistu re levels of t = 5, moisture, C:N, and %N had become important to the structuring of the microbial communities (Figure 4-4f). Other studies of soil microbial communities have shown differen tiation of microbial communities from different plant communities differing in litter or soil C:N or nitrogen availability (Eiland et al. 2001, Gallo et al. 2004, Leckie et al. 2004, Myers et al. 2001, Waldrop and Firestone 2004a, b). It would appear from this study that moisture levels and availability of labile car bon control early colonization of litter and as decomposition proceeds, nitrogen availability becomes more important. Examination of functional groups of micr obes also shows that moisture levels controlled colonization by microbes as litter quality had low discrimination of microbial community composition at t = 1 and 2 (Tables 4-8 through 13). However, at t = 5 when water was not limiting, %N, C:N and %IMR we re generally good predictors for bacterial functional groups, fungal : bacterial ratio, and to tal biomass as inferred by total microbial PFLA. This is consistent with other st udies of soil microbial communities with both Gram-negative and Gram-positive bacteria generally increasing with nitrogen availability (either through nitrogen fertil ization or decrease in C:N ra tio)(Eiland et al. 2001, Leckie et al. 2004, Waldrop et al. 2004). The response of fungi, however, appears to be equivocal with fungi or fungal:bacterial biom ass having a negative response (Eiland et al. 2001, Leckie et al. 2004), no response (Waldrop et al. 2004) or, as seen in this study, a positive response (Gallo et al. 2004) to increa sed nitrogen availability. Many of these studies show that there is not always a consistent re sponse of individual PFLA

PAGE 88

78 biomarkers within a functional group to litter quality treatments (see especially Gallo et al. 2004, Waldrop et al. 2004). We found this to be true in this study and the low amount of variance explained for a given functional group by different carbon fractions is a result of differential response of indicator fatty acids within a functional group. Factors Controlling Decomposition Rate Early decomposition is dominated by th e leaching and decomposition of soluble and low molecular weight compounds (Swift et al. 1979, Berg et al. 1982). In our study, only 182 mm of rain fell in the first 112 days of the experiment, with 56% of that rain falling before the first sampling date. Between samples 3 and 5 (129 days), encompassing hurricanes Frances and Jeanne nearly 900 mm of rain fell. This distribution in rainfall is likel y responsible for two major peri ods of leaching resulting in greater relative mass losses at those points (Figure 4-1). Alth ough a single exponential model may not be the best model to explain ma ss loss, we chose this model to enable us to have a single variable to examine th e relationship between decomposition and litter quality. All litter quality factors examined except for acid-soluble fiber were correlated with decomposition (Table 4-3). As with the microbial communities, it is difficult to determine which measure of initial litter qua lity is most responsible for controlling decomposition. However, across the three methods we used to examine relationship between litter quality and decomposition (si ngle regression, multiple regression and the ordination of a leaf chemistry axis), ligni n:N, C:N and non-polar carbon fractions were important factors. This is consistent with a large body of lite rature that shows decomposition rate is related to relative av ailability of carbon a nd nitrogen (Coulson and Butterfield 1978, Melillo et al. 1982, Taylor et al. 1989). Recalcitrant fractions such as

PAGE 89

79 lignin are not degraded in a significant way unt il later in decomposition when most of the more labile fractions have been redu ced significantly—more than 2 years into decomposition in a study of Scots pine litte r (Berg et al. 1982). As this study was conducted for less than a year, it is not surpri sing that the non-polar fractions, which are readily available for fast-growing microor ganisms, were highly correlated with decomposition rate. The novel information from this study is the linking of litter quality, microbial community composition and decomposition rate Non-polar fracti ons and C:N, which were strongly correlated with decomposition rate, were also the litter quality measures most strongly correlated with CCA axes or dinating the microbial communities (Figure 43b). Additionally, the leaf chem istry axis explained a large proportion of the variation in decomposition rate, functional groups, and overall microbial community composition (Tables 4-14 and 4-15). The best predicto r of decomposition rate, however, was not any measure of litter quality. Instead, microbial community composition explained the greatest amount of variation in decomposition ra te (76%). These results indicate that treating microbes as a black box may limit our understanding of controls on decomposition. Implications for Invasion We believe that our study indicates that invaders will be more likely have a large impact on microbial community composition or nutrient release if the invader’s litter quality, especially C:N, lignin: N and non-polar extracts is significantly different from the native community as these were the best predictors of decom position and microbial community composition. Invasive species may have a greater potential to alter microbial communities and ecosystem process if they inve st less in defense. Upon introduction to a

PAGE 90

80 new range, several invasive species have been shown to evolve toward less allocation to defense (EICA hypothesis) as a result of decrea sed herbivore pressure in their introduced range (Blossey and Notzold 1995, Siemann a nd Rogers 2003). Such reductions in defense allocation would result in high litte r quality of invaders higher decomposition rates and a shift from oligotrophic (i.e., able to live on low nutrient levels) to copiotrophic (i.e. requiring high levels of high quality nutrients, sensu Oh ta and Hattori 1983) microbial communities. These shifts could lead to positive feedbacks on invasion by increasing nutrient availabi lity (Ehrenfeld 2003). Invasive species need to contribute a large proportion of the total litter inputs in order to affect nutrient cy cling on an ecosystem scale. In our study, although Ardisia and particularly Ruellia have litter characteristics different from those in the plant communities they invade, they are understory shrubs without a large leaf biomass and thus will be less likely to affect nutrien t cycling and microbial communities at the ecosystem scale. Sapium Causurina Schinus and Imperata conversely, can form monodominant stands and produce large amounts of litter that would enable them to affect ecosystem-level change. Conclusions We investigated the differences in microbial community composition and decomposition of litter among different species of litter in a common site. Even though the plant litter was subjected to the same envi ronment, with the same land-use history and the same potential colonizing microbial community, composition of the microbial community differed among litter species. Th ese differences in microbial community composition were a function of initial litter quality and litter quality at the time of collection. Changes in litter quality due to both decomposition and changing

PAGE 91

81 environmental factors, predominately moisture content, resulted in a significant effect of time on microbial community composition. Microbial community composition was more strongly correlated with decomposition rates th an any measure of litter quality. This study suggests that explicitly examining the microbial community rather than treating it as a black box may improve our understan ding of the controls of early-phase decomposition.

PAGE 92

82Table 4-1: Characteristics of plant species whose litter was used in this experiment including leaf lifespan class (1 = <1 year 2 1 year, 3 = > 1 year), leaf type (broadleaf needle leaf, monocot), position on moisture gradient (dry, mesic, wet), and native status, along with the chemical compositi on of freshly collected litter (senescen t leaves that were easily abscised by shaking). Species Lifespan Type Habitat Native? C:N NPE(%) WS(%) AS(%) Lignin(%) Acer rubrum 1 Broad Wet Yes 55.9 60.1 10.6 12.3 14.3 Ardisia crenata 3 Broad Mesic No 44.5 52.0 19.7 20.5 7.43 Aristida stricta 2 MonocotDry Yes 128.5 12.3 40.8 22.9 23.6 Carya glabra 1 Broad Mesic Yes 38.8 55.5 12.2 18.5 13.6 Casuarina glauca 1 ** Mesic No 25.1 32.7 17.2 25.6 24.2 Imperata cylindrica 2 MonocotDry No 97.2 19.2 31.0 38.4 10.9 Juncus roemerianus 2 MonocotWet Yes 69.2 20.1 34.9 36.0 34.9 Liquidambar styricaflua 1 Broad Mesic Yes 51.9 57.4 9.21 16.2 16.9 Magnolia virginiana 3 Broad Mesic Yes 57.5 51.8 14.1 20.3 13.1 Pinus palustris 3 Needle Dry Yes 58.2 34.6 14.5 27.7 22.5 Quercus chapmannii 2 Broad Dry Yes 59.7 50.9 11.7 19.3 17.9

PAGE 93

83 Table 4-1. Continued Species Lifespan Type Habitat Native? C:N NPE(%) WS(%) AS(%) Lignin(%) Quercus geminata 2 Broad Dry Yes 74.9 42.4 12.9 25.1 19.1 Quercus laevis 1 Broad Dry Yes 36.0 32.0 11.9 26.6 29.2 Quercus nigra 1 Broad Mesic Yes 47.7 44.2 13.6 20.7 21.3 Ruellia brittoniana 1 Broad* Wet No 32.4 72.2 9.29 10.2 7.92 Sabal palmetto 3 MonocotMesic Yes 30.3 25.27 20.2 33.9 19.1 Sapium sebiferum 1 Broad Wet No 32.4 73.7 9.85 10.5 5.02 Schinus terebinthifolius 2 Broad Mesic No 37.8 61.1 5.05 11.9 21.6 Taxodium distichum 1 Needle Wet Yes 28.8 45.9 12.9 15.1 25.9 Typha latifolia 2 MonocotWet Yes 54.5 21.4 23.4 17.0 37.5 Ruellia brittoniana was the only non-woody broad leaf ex amined. **Photosynthetic tissue of Casuarina gluaca treated as leaves in this study, are modified stems, needle-like in appearance.

PAGE 94

84 Table 4-2: Mean and SE for decomposition rate s of native and non-native litters (n=5). Different letters indicate significant differences (alpha = 0.05) between species by Tukey HSD. The leaf chemistry axis is the score from the ordination of initial plant li tter chemistry. High scores indicate poor quality (high lignin:N, C:N), and low scores indicate high quality (high non-polar fraction, nitrogen). DF = 19, 80. Species Native? k (year-1) Leaf Chemistry Axis Score Mean SE Sapium No 1.91a 0.0852 -2.70 Ruellia No 1.79a 0.150 -2.51 Causurina No 1.06b 0.0546 -1.26 Ardisia No 0.990bc 0.159 -1.14 Carya Yes 0.849bc 0.0311 -1.45 Liquidambar Yes 0.789bcd 0.0451 -0.80 Schinus No 0.724bcd 0.0354 -0.38 Q. nigra Yes 0.731bcd 0.0673 -1.48 Magnolia Yes 0.723bcd 0.0853 -0.64 Typha Yes 0.710bcd 0.0708 2.37 Sabal Yes 0.668bcd 0.0449 -0.49 Q. geminata Yes 0.655cde 0.0333 0.56 Acer Yes 0.647cde 0.0357 -0.81 Taxodium Yes 0.586cdef 0.0684 -1.16 Imperata No 0.531def 0.0506 2.26 Q. chapmannii Yes 0.404ef 0.0167 -0.36 Pinus Yes 0.401fg 0.0165 1.51 Juncus Yes 0.394fg 0.0170 3.49 Q. laevis Yes 0.388fg 0.0242 -0.30 Aristida No 0.387fg 0.159 5.28 Table 4-3: Result of simple linear re gression for log-transformed decomposition constants against individual variable s of litter chemical composition. DF = 1,18. %NPE = non-polar fraction, %W S = water-soluble fraction, %AS = acid-soluble fraction. Litter variable Slope r2 P %N 0.583 0.24 0.028 C:N -0.0101 0.32 0.0088 %NPE 0.0173 0.45 0.0011 %WS -0.0214 0.19 0.052 %AS NS NS NS %Lignin -0.0305 0.34 0.0066 Lignin:N -0.0193 0.47 0.0009

PAGE 95

85 Table 4-4: Factor loadings fo r individual variables that cont ribute to litter quality on the leaf chemistry PCA axis 1. In general, high leaf chemistry score indicates low quality litter with low %N, high C:N, low%NPE, high %WS, high %lignin, and high lignin:N ratio. Litter Quality Measure Loading %N -0.353 C:N 0.431 %NPE -0.425 %WS 0.442 %AS -0.0490 %Lignin 0.282 Lignin:N 0.481 Table 4-5: Factor loadings of individual PLFAs upon the fi rst two main axes of PCA (PC1 and PC2) and CCA (CC1 and CC2). Loadings > 0.3 in bold Fatty Acid PC 1 PC 2 CC 1 CC 2 Unclassified 14:0 0.245 -0.100 -0.011 0.086 15:0 0.331 0.059 0.314 -0.118 16:0 0.114 -0.417 -0.153 -0.033 17:0 0.154 -0.318 -0.123 0.058 18:0 0.179 -0.316 -0.133 -0.060 i19:0 -0.033 -0.068 -0.182 0.512 20:1 9c 0.006 -0.262 -0.267 -0.161 Gram + i15:0 0.348 0.191 0.536 -0.077 a15:0 0.280 -0.189 0.036 -0.124 i16:0 0.351 0.136 0.371 -0.023 Gram 16:1 7c 0.300 -0.188 0.039 0.059 17:1 8c 0.084 -0.337 -0.143 0.100 cy17:0 0.343 0.165 0.476 -0.145 18:1 7c 0.306 0.211 0.520 0.080 cy19:0 0.332 0.224 0.590 -0.015 Fungi 18:2 6c 0.041 -0.422 -0.188 0.068 18:1 9c 0.076 -0.025 0.050 0.033

PAGE 96

86 Table 4-6: MANOVA summarizing the effects of time, litter species and litter species*time on PCA and CCA axes 1 a nd 2 scores from ordination of all samples simultaneously. PCA CCA F P F P Model 9.01 <0.0001 27.14 <0.0001 Time 25.89 <0.0001 194.55 <0.0001 Species 14.50 <0.0001 34.13 <0.0001 Sp*Time 4.73 <0.0001 3.97 <0.0001 Note : Model df = 29, 112; Species df = 9, 112; Time df = 2, 112; Species*time df = 18, 112. Table 4-7: Loadings for the first two PCA axes for PCA run separately for individual sampling dates (t = 1, 2, and 5). T = 1 T = 2 T = 5 Fatty Acid PCA 1 PCA 2 PCA 1 PCA 2 PCA 1 PCA 2 Unclassified 14:0 0.1307 0.3481 0.1745 0.2407 -0.1945 -0.0666 15:0 0.3287 0.0477 0.3158 -0.0603 -0.1978 -0.3765 16:0 0.3605 0.1458 0.3440 0.2246 -0.2842 -0.2803 17:0 0.2457 -0.2206 0.3597 -0.0408 -0.2059 -0.3638 18:0 0.2188 0.0340 0.3091 0.1865 -0.2990 -0.1028 i19:0 0.1531 0.3016 NA NA NA NA 20:1 9c 0.1751 -0.0434 0.1597 0.3788 -0.1396 -.0899 Gram + i15:0 0.1022 -0.4039 0.1825 -0.4010 -0.3037 0.1943 a15:0 0.2567 -0.3499 0.1795 -0.4548 -0.2692 0.2642 i16:0 0.1533 -0.4933 0.1843 -0.4157 -0.2742 0.2461 Gram 16:1 7c 0.3741 -0.0441 0.3399 -0.1433 -0.2952 0.2216 17:1 8c 0.3617 -0.0086 0.3184 0.0161 -0.1413 -.1089 cy17:0 0.0299 -0.0451 0.1439 0.1822 -0.2756 0.2736 18:1 7c 0.1860 0.1305 0.1439 0.1822 -0.3068 0.1148 cy19:0 -0.0550 0.1330 -0.0326 -0.1557 -0.3001 0.1010 Fungi 18:2 6c 0.3870 0.1301 0.3694 -0.0044 -0.0847 -.4984 18:1 9c 0.1536 0.3576 -0.0623 0.0443 -0.2768 -.2205

PAGE 97

87 Table 4-8: Correlation between each litter quality variable and the amount of PLFA (nmol) indicating monounsaturated Gram – b acteria at times 1, 2, and 5. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = non-polar fraction, %WS = watersoluble fraction, %AS = acid-solubl e fraction, %IMR = initial mass remaining, %moisture on a dry mass basis. T = 1 T = 2 T = 5 Direction r2 Direction r2 Direction r2 % N + 0.18** NS – + 0.41*** C:N 0.23** NS – 0.40*** %NPE NS – NS – + 0.10* %WS NS – NS – 0.09* %AS NS – NS – 0.26*** %lignin NS – NS – + 0.19** Lignin:N 0.29*** NS – 0.12** %IMR NS – NS – 0.45*** %moisture NS – NS – + 0.09* Table 4-9: Correlation between each litter quality variable and the amount of PLFA (nmol) indicating branched Gram+ fatty acids at times 1, 2, and 5. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = non-polar fracti on, %WS = watersoluble fraction, %AS = acid-solubl e fraction, %IMR = initial mass remaining, %moisture on a dry mass basis. T = 1 T = 2 T = 5 Direction r2 Direction r2 Direction r2 % N + 0.16** NS – + 0.40*** C:N NS – NS – 0.41*** %NPE NS – NS – NS – %WS + 0.12** + 0.17** 0.08* %AS NS – + 0.05* 0.18** %lignin NS – 0.13** + 0.19** Lignin:N 0.10* NS – 0.06 %IMR 0.16** NS – 0.33*** %moisture NS – NS – + 0.11**

PAGE 98

88 Table 4-10: Correlation between each litter quality variable and the amount of PLFA (nmol) indicating fungal fatty acid (18:2 6c) at times 1, 2, and 5. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = non-polar fracti on, %WS = watersoluble fraction, %AS = acid-solubl e fraction, %IMR = initial mass remaining, %moisture on a dry mass basis. T = 1 T = 2 T = 5 Direction r2 Direction r2 Direction r2 % N + 0.09* + 0.47*** NS C:N 0.10* 0.45*** NS %NPE NS – + 0.31*** NS %WS NS 0.20** NS %AS NS 0.32*** NS %lignin NS NS NS Lignin:N 0.12** 0.25*** NS %IMR 0.43*** NS %moisture NS 0.06* NS 0.29*** Table 4-11: Correlation between each litter quality variable and ratio fungal : bacterial PLFAs at times 1, 2, and 5. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = non-polar fraction, %WS = water-sol uble fraction, %AS = acid-soluble fraction, %IMR = initial mass remain ing, %moisture on a dry mass basis T = 1 T = 2 T = 5 Direction r2 Direction r2 Direction r2 % N NS NS 0.41*** C:N + 0.11* NS + 0.32*** %NPE NS + 0.15** NS %WS NS 0.21** NS %AS NS 0.11* + 0.10* %lignin NS + 0.05 NS Lignin:N + 0.11* + 0.05 + 0.19** %IMR + 0.07* NS + 0.37*** %moisture NS NS 0.27***

PAGE 99

89 Table 4-12: Correlation between each litter quality variable and the total PLFA (nmol) at times 1, 2, and 5. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = non-polar fraction, %WS = water-sol uble fraction, %AS = acidsoluble fraction, %IMR = initial mass remaining, %moisture on a dry mass basis. T = 1 T = 2 T = 5 Direction r2 Direction r2 Direction r2 % N + 0.15** + 0.41*** + 0.15** C:N 0.08* 0.39*** 0.14** %NPE NS + 0.25*** + 0.21** %WS NS 0.26*** 0.20** %AS NS 0.29*** 0.17** %lignin NS NS + 0.08* Lignin:N NS 0.11** NS %IMR NS 0.27*** 0.17** %moisture NS NS NS Table 4-13: Correlation between each litter quality variable and the amount of PLFA (nmol) indicating cyclopropyl Grambact eria at time 5. Only t = 5 is presented as the large number of zeros at other time points did not allow for parametric statistics. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = nonpolar fraction, %WS = water-soluble fr action, %AS = acid-soluble fraction, %IMR = initial mass remaining, %moisture on a dry mass basis Direction r2 % N + 0.26*** C:N 0.18** %NPE + 0.05 %WS NS %AS 0.23** %lignin + 0.07* Lignin:N NS %IMR 0.28*** %moisture + 0.8*

PAGE 100

90 Table 4-14: Coefficient of determination (r2) of abundance of PLFAs for microbial functional groups with leaf chemistry scores at times 1, 2, and 5. Df time 2 and 5 = 1,8; df time 1 = 1, 7. p < 0.05, ** p <0.01, *** p < 0.0001 Time = 1 Time = 2 Time = 5 Total PLFA NS 0.54** 0.62** Fungal:Bacterial Ratio NS NS 0.49** Fungi NS 0.53** NS Gram+ NS NS 0.77** Branched GramNS NS 0.70** Cyclopropyl GramNS NS 0.60** Note : All significant relationships were negati ve correlations except for fungal:bacterial ratio which was positive. Table 4-15: Coefficient of determination (r2) of correlations of microbial community PCA axes (independent) vs.decomposition rate (k) (dependant), and leaf chemistry axis (from ordination of initi al litter quality) (independent) vs. microbial community PCA axes (dependent). All relationships were significant at p <0.05. T=1 PCA1 T=2 PCA1 T=5 PCA1 All PCA1 All PCA2 K 0.53 0.62 0.76 0.76 0.58 Litter chemistry 0.44 0.52 0.88 0.65 0.53 Note : Microbial PCA vs. k df = 1,9 (except T=1 PCA, df = 1,8); litter quality vs. PCA df = 1,90 (except T=1 PCA, df = 1,8).

PAGE 101

91 Percent mass remaining 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Ruellia Sapium A. ruburm Taxodium Ardisia Liquidamber Causurina Schinus Q.nigra C. glabra Magnolia Aristida Q. laevis Q. champanii Sabal Q. geminata Imperata Typha Juncus Pinus Days 050100150200250300350 Rainfall from previous sample date (mm) 0 100 200 300 400 500 600 700 Figure 4-1: Mean + SE of the percent mass remaini ng of 20 species at six collection dates. Species are shown in order of decreasing mass loss at the first collection date. Total rainfalls between sample collections are also displayed as bars.

PAGE 102

92 Figure 4-2: Mean + SE for nitrogen in litter of 20 plant species at six collection dates. (a) nitrogen concentration, a nd (b) percent initial nitrogen remaining Species are shown in order of decreasing mass lo ss at the first collection date. Time (days) 050100150200250300350 % N 0 1 2 3 Ruellia Sapium Acer Taxodium Ardisia Liquidambar Causurina Schinus Q. nigra Carya Magnolia Aristida Q. laevis Q.chapmannii Sabal Q. geminata Imperata Typha Juncus Pinus Time (days) 050100150200250300350 Percent inital N remaining 0.25 0.50 0.75 1.00 1.25 1.50 Ruellia Sapium A. rubrum Taxodium Ardisia Liquidambar Causurina Schinus Q. nigra Carya Magnolia Aristida Q. laevis Q.chapmannii Sabal Q. geminata Imperata Typha Juncus Pinus a b

PAGE 103

93 Figure 4-3 Ordinations of the 17 most common microbial PLFA found on 11 litter species from three sampling dates, showi ng mean values for five replicates per litter species per sampling date.: a) Pr incipal Component Analysis (PCA). b) Canonical Correspondence Analysis includi ng nine litter quality measures as a secondary matrix. Biplots for litter quality measures with an r2 > 0.3 with at least one of the two canonical axes displayed. Open symbols signify nonnative species, solid symbols signify native species. Litter species are presented in increasing leaf chemistry ax is scores, with the 4 lowest scores representing high litter quality in green, middle litter quality in yellow, and the 4 highest scores representing low litter quality in red. Ellipses are drawn by hand to indicate groups of points from the same collection date. The large cloud of symbols not encompassed in an ellipse in (a) are predominantly litter species with low litter quality and tend to move to the right on PCA 1 over time as indicated by the arrows. PCA 1(35%) -4-20246 PCA 2 (27%) -6 -4 -2 0 2 4 Sapium Ruellia Taxodium Schinus Ardisia Causurina Acer Aristida Imperata Juncus Pinus t=2 t=5t=1 a

PAGE 104

94 Figure 4-3. Continued. CCA 1 (24.9%) -2-10123 CCA 2 (7.4%) -2 0 2 4 t=5 IMR NPE C:N %N ASb

PAGE 105

95 Figure 4-4: PCA and CCA analyses for 17 most common microbial PLFA and litter quality variables run separately for i ndividual sampling dates. A) 28 day PCA, b) 28 day CCA, c) 57 day PCA, d) 57 day CCA, e) 238 day PCA, f) 238 CCA. Only litter quality measures with r2 > 0.3 with one of the two canonical axes are displayed. See Figure 3 for symbol descriptions. CCA 1 (25.4%) -3-2-1012 CCA 2 (12.2%) -3 -2 -1 0 1 2 3 %NPE Lignin:N C:N, %IMR AS, WS PCA 1 (35.0%) -4-20246 PCA 2 (15.5%) -4 -3 -2 -1 0 1 2 3 Sapium Ruellia Taxodium Schinus Ardisia Causurina Aristida Imperata Juncus Pinus a b

PAGE 106

96 Figure 4-4. Continued. PCA 1 (38%) -4-202468 PCA 2 (13%) -4 -3 -2 -1 0 1 2 3 4 Sapium Ruellia Taxodium Schinus Ardisia Causurina Acer Aristida Imperata Juncus Pinus CCA 1 (21.4%) -3-2-10123 CCA 2 (14.5%) -3 -2 -1 0 1 2 3 4 NPE, N WS, AS, C:N Lignin:N, IMR c d

PAGE 107

97 Figure 4-4. Continued. e f PCA 1 (48%) -6-4-20246 PCA 2 (16%) -4 -3 -2 -1 0 1 2 3 Sapium Ruellia Taxodium Schinus Ardisia Causurina Acer Arisitida Imperata Juncus Pinus CCA (31.2%) -2-10123 CCA 2 (7.0%)-NS -3 -2 -1 0 1 2 3 4 %N Moisture %IMR C:N

PAGE 108

98 CHAPTER 5 CONCLUSIONS The goal of this dissertation was to evaluate the role of plant species composition in the interaction between plants and soil micr obes using invasive plants as a natural experiment. I examined three primary inter actions: native mycorrhizal fungi with a nonnative plant, variation in fr ee-living soil microbes in five different habitats with and without non-native plant i nvaders, and the links betw een plant litter chemical composition, microbial community compos ition and decomposition of plant litter. In Chapter 2, I examined the effect of native mycorrhizae on the growth, physiology and competitive ability of an invasive shrub, Ardisia crenta I hypothesized that Ardisia may be a successful invader because it has been able to find a beneficial fungal partner that improves its competi tive ability for phosphorus. I found that Ardisia had the highest growth rate and photosynthe tic rate when grown with the fungi with which it associates in an invaded habitat. Ardisia ’s growth appears to be increased due to higher photosynthetic rates and a greater allo cation to aboveground biomass, particularly in leaf area ratio (LAR). In a competitive e nvironment, however, neither identity of the competitor nor mycorrhizal status had an effect on the growth rate of Ardisia The absence of mycorrhizae in heterospecific competition did decrease the growth of Prunus While Ardisia benefits from local mycorrhizal fungi, removal or disturbance of mycorrhizal fungi may be of greater detriment to natives than to Ardisia Further work with the Ardisia system could clarify the ro le of mycorrhizal fungi in mediating plant invasion. It would be usef ul to determine which species of fungi are

PAGE 109

99 present in the host-associated inoculum a nd how their composition and density compare to non-invaded areas. As my results showed that the growth response of Ardisia is dependent upon the identity of the fungus, fu rther growth experiments with different plant and fungal species may also be needed. For example, the results of the Ardisia growth and competition experiments may have b een very different if fungi isolated from Prunus ’ roots were used in these experiments. My second objective was to determine if non-native plants could alter the composition of free-living microbes (Chapter 3). The largest differences seen in microbial community structure, as measur ed by PLFA, and function, as measured by Biolog substrate utilization, were between habitat types. The two southern most communities, those prone to Melaleuca and Schinus invasion, were most similar to one another in comparison to the other habitats examined. Total PLFA and fungal : bacterial ratio were highly correlated with moisture content of the soil indicating that moisture content and oxygen availability may be a major control of microbial community composition. While I found that variati on among microbial comm unities was greater between habitats than within them, there wa s a significant overall effect of invasion on microbial community structure. There was no effect of invasion on substrate utilization, however, suggesting that although invasion may result in a change in microbes present, due to redundancy of metabolic ability, micr obial community function may not be altered by invasion. Conclusions made from this study, howeve r, are limited as it is only a single snapshot in time. It may be possible that at different times of year (e.g., during the dry season) similarities in soil moisture contents may cause a convergence in composition of

PAGE 110

100 microbial communities across habitats. Conve rsely, the larger differences in soil temperature during the dry season may lead to greater differentiation of microbial communities according to lat itude. Additionally, although I attribute much of the differences between microbial communities to differences in soil moisture, several environmental characteristics were not measur ed and may be responsible for differences in microbial community composition. A mo re thorough examination of environmental variables in situ or experime ntal alteration of these variables would lead to a better understanding of controls on microbial community composition Finally, in Chapter 4, I explicitly examined the role of litter quality in microbial community composition. The composition of microbial communities on eleven species of native and invasive plant l itter of varying quality was followed over decomposition. I also examined the relationship between the composition of the microbial communities and decomposition. I hypothesized that plan t litters with similar litter quality would support similar microbial communities. This proved to be true. Whether the plant litter was native or invasive did not seem to affect the composition of the microbial community. Microbial community composition differed most among sample dates, with microbial communities of 28 and 57 days clus tering together in ordination of PLFAs separate from microbial communities at 238 da ys. These large differences in microbial communities at early and late sample dates c ould be a result of mi crobial succession, but may also be due to moisture limitation at early sample dates. The most novel result of this study, however, was the linking of microbial community composition to decomposition rate I examined the correlation between seven measures of leaf chemistry, an integr ated measure of leaf chemistry, microbial

PAGE 111

101 community composition (the first axis from principle components ordination of PLFA data), and decomposition rate constants. Of all these relationships examined, microbial community composition was the best predictor of decomposition rate with a correlation of 76%. This result is very compelling evid ence that opening the “black box” of the soil is not only of interest to microbial ecol ogists, but has broader implications These results suggest many more studies. Although one of the general goals for this dissertation was to examine interactions between invasive species and microbes, this study least explicitly examined this interaction. Better pa iring of native and invasive plants with respect to dominance and functi onal groups would place the results in better context to evaluate the impact of invasion. Additionally, many of the species used in this study are not found in hammock communities. A more realistic study would be to examine the decomposition and microbial succession of native and invasive litter pairs in invaded and non-invaded areas in which they would normally be found. Finally, the relationship found between litte r quality, microbial compos ition and decomposition rate in this study is correlative not causal. Microcosms initiated with different starting microbial communities subjected to litter of low, moderate, and high quality might help to determine if microbes merely respond to the substrate avai lable or are act ually drivers of decomposition. Taken together, these studies have shown that invasive plants can benefit from specific native mycorrhizal fungi, can alte r microbial communities on a landscape level and may result in different decomposer mi crobes and decomposition rates if the litter quality of the invader is much different than native litter quality. These results suggest that plant-microbial could be an important aspect in the success of non-native plant.

PAGE 112

102 Further studies should explore the potent ial for positive feedbacks on plant invasion through alterations of the soil mi crobial community by invaders.

PAGE 113

103 LIST OF REFERENCES Alden, L., F. Demoling, and E. Baath. 2001. Rapid method of determining factors limiting microbial growth in soil. App lied and Environmental Microbiology 67:1830-1838. Allison, S. D., and P. M. Vitousek. 2004. Rapid nutrient cycling in leaf litter from invasive plants in Hawai'i. Oecologia 141:612-519. Allsopp, N., and W. D. Stock. 1992. Mycorrhizas seed size and seedli ng establishment in low nutrient environment. Pages 59-64 in D. J. Read, D. H. Lewis, A. Fitter, and I.J. Alexander, editors. Mycorrhizas in Ecosystems. CAB International, Wallingford, UK. Atlas, R. M., and R. Bartha. 1998. Microbi al Ecology: Fundamentals and Applications. Benjamin/Cummings, Redwood City, CA. Bailey, P. C. E., S. C. Watkins, K. L. Morris, and P. I. Boon. 2003. Do Melaleuca ericifolia Sm. leaves suppress organic matte r decay in freshwater wetlands? Archiv Fur Hydrobiologie 156:225-240. Bardgett, R. D., and E. McAlister. 1999. The measurement of soil fungal:bacterial biomass ratios as an indicator of ecosy stem self-regulation in temperate meadow grasslands. Biology and Fertility of Soils 29:282-290. Bardgett, R. D., and L. R. Walker. 2004. Impact of coloniser plant species on the development of decomposer microbial communities following deglaciation. Soil Biology & Biochemistry 36:555-559. Bardgett, R. D., J. L. Mawdsley, S. Edwards, P. J. Hobbs, J. S. Rodwell, and W. J. Davies. 1999. Plant species and nitrogen e ffects of soil biologi cal properties of temperate upland grasslands. Functional Ecology 13:650-660. Berg, B., and C. O. Taum. 1991. Decomposition and nutrient dynamics of litter in longterm optimum nutrition experiments. Sca ndinavian Journal of Forest Research 6:305-321. Berg, B., K. Hannus, T. Popoff, and O. Th eander. 1982. Changes in organic-chemical components during decomposition. Long-term decomposition in a Scots pine forest I. Canadian Journal of Botany 60:1310-1319.

PAGE 114

104 Berta, G., A. Trotta, A. Fusconi, J. E. Hooke r, M. Munro, D. Atkinson, M. Giovannetti, S. Morini, P. Fortuna, B. Tisserant, V. Gianinazzipearson, and S. Gianinazzi. 1995. Arbuscular mycorrhizal induced change s to plant growth and root-system morphology in Prunus cerasifera Tree Physiology 15:281-293. Bethlenfalvay, G., M. Brown, and R. Pascovsky. 1982. Parasitic and mutualistic associations betwen a mycorhizal fungus and soybean: development of the host plant. Phytopathology 72:889-893. Bever, J. D. 1994. Feedback between plants and their soil communities in an old field community. Ecology 75:1965-1977. Bever, J. D. 2002. Negative feedback within a mutualism: host-specific growth of mycorrhizal fungi reduces plant benefit. Proceedings of th e Royal Society of London Series B-Biological Sciences 269:2595-2601. Bever, J. D., J. B. Morton, J. Antonovics and P. A. Schultz. 1996. Host-dependant sporulation and species di versity of arbuscular mycorrhizal fungi in a mown grassland. Journal of Ecology 84:71-82. Bever, J. D., K. M. Westover, and J. An tonovics. 1997. Incorporating the soil community into plant population dynamics: the utility of the feedback approach. Journal of Ecology 85:561-573. Blossey, B., and R. Notzold. 1995. Evolution of increased competitive ability in invasive nonindigenous plants–a hypothe sis. Journal of Ecology 83:887-889. Boerner, R. E. J. 1990. Role of mycorrhizal fungus origin in grow th and nutrient-uptake by Geranium robertianum American Journal of Botany 77:483-489. Boon, P. I., and L. Johnstone. 1997. Organi c matter decay in coastal wetlands: An inhibitory role for essential oil from Melaleuca alternifolia leaves? Archiv Fur Hydrobiologie 138:433-449. Bossio, D. A., and K. M. Scow. 1995. Impact of carbon and flooding on the metabolic diversity of microbial communities in soils. Applied Environmental Microbiology 61:4043-4050. Bossio, D. A., and K. M. Scow. 1998. Impacts of carbon and flooding on soil microbial communities: phospholipid fatty acid profile s and substrate utilization patterns. Microbial Ecology 35:265-278. Bossio, D. A., K. M. Scow, N. Gunapala, and K. J. Graham. 1998. Determinants of soil microbial communities: effects of agricultural management, season, and soil type on phospholipid fatty acid pr ofiles. Microbial Ecology 36:1-12. Brown, J. C. 1958. Soil fungi of some Brit ish sand dunes in relation to soil type and succession. Journal of Ecology 46:641-664.

PAGE 115

105 Buyer, J. S., and L. E. Drinkwater. 1997. Comparison of substrate utilization assay and fatty acid analysis of soil microbial co mmunities. Journal of Microbiological Methods 30:3-11. Callaway, R. M., G. C. Thelen, A. Rodri guez, and W. E. Holben. 2004. Soil biota and exotic plant invasion. Nature 427:731-733. Cameron, G., and S. Spencer. 1989. Rapid leaf decay and nutrient release in a Chinese Tallow forest. Oecologia 80:222-228. Chapin, F. S. 1980. The mineral nutrition of plants. Annual Review of Ecology and Systematics 11:233-260. Chapin, F. S. I., and R. Aerts. 2000. The mine ral nutrition of wild pl ants revisited: a reevaluation of processes and patterns. Advances in Ecological Research 30:1-67. Chapin, F. S. I., H. Reynolds, C. M. D'Antonio, and V. Eckhart. 1996. The functional role of species in terrestrial ecosys tems. Pages 403-430 in B. Walker and W. Steffan, editors. Global Change in Terre strial Ecosystems. Cambridge University Press, Cambridge. Coleman, D. C. 1985. Through a ped darkly : an ecological assessment of root-soilmicrobial-faunal inte ractions. Pages 1-21 in A. Fitter, D. Atkinson, D. J. Read, and M. B. Usher, editors. Ecological Interactions in Soil: Plants, Microbes and Animals. Blackwell, Oxford. Coulson, J. C., and J. Butterfield. 1978. An i nvestigation of the bio tic factors determining the rates of plant decomposition on blanket bog. Journal of Ecology 66:631-650. D'Antonio, C. M., T. L. Dudley, and M. C. Mack. 1999. Disturbance and biological invasions: direct effects a nd feedbacks. Pages 429-468 in L. R. Walker, editor. Ecosystems of Disturbed Gr ound. Elsevier, Amsterdam. Dowling, N. J. E., F. Widdle, and D. C. White. 1986. Phospholipid ester-linked fatty acids biomarkers of acetate-oxidizing su lphate-reducers and other sulphide-forming bacteria. Journal of General Microbiology 132:1815-1825. Dozier, H. 1999. Plant introductions and invasion: history, pub lic awareness, and the case of Ardisia crenata Ph.D. University of Florida, Gainesville. Ehrenfeld, J. G. 2003. Effects of exotic plan t invasions on soil nutrient cycling processes. Ecosystems 6:503-523. Eiland, F., M. Klamer, A. M. Lind, M. Leth, and E. Baath. 2001. Influence of initial C/N ratio on chemical and microbial com position during long-term composting of straw. Microbial Ecology 41:272-280.

PAGE 116

106 Ekbald, A., and A. Nordgren. 2002. Is growth of soil microorganism in boreal forest limited by carbon or nitrogen av ailability? Pl ant and Soil 242:115-122. Findlay, R. H., and F. C. Dobbs. 1993. Quantita tive description of microbial communities using lipid analysis. Pages 777 in P. F. Kemp, editor. Handbook of Methods in Aquatic Microbial Ecology. Le wis Publishers, Boca Raton. Flanagan, P. W., and K. van Cleve. 1983. Nu trient cycling in relation to decomposition and organic matter quality in taiga ecos ystems. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 13:795-817. Florida Exotic Pest Plant Council. 2005. List of Florida's Invasive Sp ecies. Florida Exotic Pest Plant Council. Internet: h ttp://www.fleppc.org/list05web.pdf. June, 2005. Fontanet, X.,V. Estaun,A. Camprubi and C. Calvet. 1998. Fungicides added to potting substrate affect mycorrhizal symbiosi s between a peach-almond rootstock and Glomus sp. Hortscience 33: 1217-1219. Frostegard, A., and E. Baath. 1996. The us e of phospholipid fatty acid analysis to estimate bacterial and fungal biomass in soil. Biology and Fertility of Soils 22:5965. Frostegard, A., E. Baath, and A. Tunlid. 1993. Shifts in the structur e of soil microbial communiteis in limed forests as revealed by phospholipid fatty acid analysis. Soil Biology & Biochemistry 25:723-730. Gallo, M., R. Amonette, C. Lauber, R. L. Sinsabaugh, and D. R. Zak. 2004. Microbial community structure and oxidative enzy me activity in nitrogen-amended north temperate forest soils. Microbial Ecology 48:218-229. Gange, A. C. and R. L. Ayres. 1999. On the relation between arbuscular mycorrhizal colonization and plant benefit. Oikos 87: 615-621. Gange, A., D. Lindsay, and L. Ellis. 1999. Can arbuscular mycorrhizal fungi be used to control the undesirable grass Poa annua on golf courses? Journal of Applied Ecology 36:909-919. Garland, J. L. 1996. Patterns of potential C source utilization by rhizophere communities. Soil Biology & Biochemistry 28:223-230. Garrett, S. D. 1963. Soil Fungi and So il Fertility. Pergamon Press, Oxford. Gavito, M. E., P. S. Curtis, T. N. Mikke lsen, and I. Jakobsen. 2000. Atmospheric CO2 and mycorrhiza effects on biomass allocation and nutrient uptake of nodulated pea ( Pisum sativum L.) plants. Journal of Experimental Botany 51:1931-1938. Gordon, D. R. 1998. Effects of invasive non-indigenous plant species on ecosystem processes: lessons from Flor ida. Ecological Applications 8:975-989.

PAGE 117

107 Gorlenko, M. V., and P. A. Kozhevin. 1994. Differentiation of soil microbial communites by multivariate testing. Microbiology 63:158-161. Graham, J., and D. Eissenstat. 1998. Field evidence for the car bon cost of citrus mycorrhizas. New Phytologist 140:103-110. Graham, J., D. Drouillard, and N. Hodge. 1996. Carbon economy of sour orange in response to different Glomus spp. Tree Physiology 16:1023-1029. Graham, J., L. Duncan, and D. Eissensta t. 1997. Carbohydrate a llocation patterns in citrus genotypes as affected by phosphorus nutrition, mycorrhizal colonization and mycorrhizal dependenc y. New Phytologist 135:767-776. Grayston, S. J., D. Vaughan, and D. Jones. 1996. Rhizosphere carbon flow in trees, in comparision with annual plants: the importa nce of root exudation and its impact on microbial activity and nutrient availability. Applied Soil Ecology 5:29-56. Grayston, S. J., S. Wang, C. D. Campbell, and A. C. Edwards. 1998. Selective influence of plant species on microbial diversit y in the rhizosphere. Soil Biology and Biochemistry 30:369-378. Greenway, M. 1994. Litter accession a nd accumulation in Melaleuca quinquenervia (Cav.) S.T. Blake wetland in Southeaste rn Queensland. Australian Journal of Marine and Freshwater Research 45. Griffin, D. M. 1985. A comparison of the roles of bacteria and fungi. Pages 221-255 in E. R. Leadbetter and J. S. Poindexter, editors. Bacteria in Nature: Bacterial Activities in Perspective. Plenum Press, New York. Griffiths, B. S., K. Ritz, N. Ebblewhite, and G. Dobson. 1999. Soil microbial community structure: effects of substrate load ing rates. Soil Biology & Biochemistry 31:145153. Grime, J. P., J. H. C. Cornelissen, K. Thompson, and J. G. Hodgson. 1996. Evidence of a causal connection between anti-herbivore defense and the decomposition rate of leaves. Oikos 77:489-494. Guckert, J. B., C. P. Antworth, P. D. Nic hols, and D. C. White. 1985. Phospholipid, esterlinked fatty acid profiles as reproducible assays fo r changes in prokaryotic community structure of estuarine se diments. FEMS Microbiology Ecology 31:147158. Hart, S. C., and J. M. Stark. 1997. Nitrogen limitation of the microbi al biomass in an oldgrowth forest. Ecoscience 4:91-98. Hartnett, D. C., B. A. D. He trick, G. W. T. Wilson, and D. J. Gibson. 1993. Mycorrhizal influence on intraspecific and inters pecific neighbor interactions among cooccurring prairie gras ses. Journal of Ecology 81:787-795.

PAGE 118

108 Harwood, D. L., and N. J. Russell. 1984. Lipids in plants and microbe s. George Allen and Unwin Ltd., London. Haynes, R. J. 1986. The decomposition proces s: mineralization, immobilization, humus formation, and degredation. Pages 52-126 in R. J. Haynes, editor. Mineral Nitrogen in the Plant-Soil System. Academic Press, Orlando. Hill, G. T., N. A. Mitkowski, L. Aldrich-Wolfe, L. R. Emele, D. D. Jurkonie, A. Ficke, S. Maldonado-Ramirez, S. T. Lynch, and E. B. Nelson. 2000. Methods for assessing the composition and diversity of soil mi crobial communities. Applied Soil Ecology 15:25-36. Hobbie, S. E. 1996. Temperature and plant spcies control over litter decomposition in Alaskan tundra. Ecological Monographs 66:503-522. Hudson, H. J. 1968. The ecology of fungi on plant remains above the soil. New Phytologist 67:837-874. Jakobsen, I. 1991. Carbon metabolism in mycorrhiza. Methods in Microbiology 23:149180. Janos, D. P. 1980. Vesicular-arbuscular mycorrh izae affect lowland tropical rain forest plant growth. Ecology 61:151-162. Jenny, H. 1941. Factors of Soil Fo rmation. McGraw-Hill, New York. Johnson, N. C. 1993. Can fertilization of soil select less mutu alistic mycorrhizae? Ecological Applications 3: 749-757. Johnson, N. C., J. Graham, and F. Smith. 1997. Functioning of mycorrhizal associations along the mutualism-parasitism conti nuum. New Phytologi st 135: 575-586. Kahiluoto, H., E. Ketoja, and M. Vestberg 2000. Creation of a nonmycorrhizal control for a bioassay of AM effectiveness. 1. Comparison of methods. Mycorrhiza 9:241258. Katajisto, J., V. Huhta, and J. Laakso 1999. Plant effects on the soil community: A microcosm experiment. Europ ean Journal of Soil Biology 35:17-21. Kourtev, P. S., J. G. Ehrenfeld, and M. Haggblom. 2002. Exotic plant species alter the microbial community structure a nd function in the soil. Ecology 83:3152-3166. Kourtev, P. S., J. G. Ehrenfeld, and M. Haggblom. 2003. Experimental analysis of the effect of exotic and native plant speci es on the structure and function of soil microbial communities. Soil Biology & Biochemistry 35:895-905. Laroche, F. B. 1994. Melaleuca Management Plan for the State of Florida. Florida Exotic Pest Plant Council, West Palm Beach.

PAGE 119

109 Leckie, S. E., C. E. Prescott, S. J. Gray ston, J. D. Neufeld, and W. W. Mohn. 2004. Characterization of humus microbial communities in adjacent forest types that differ in nitrogen availa bility. Microbial Ecology 48:29-40. Lovelock, C. E., D. Kyllo, and K. Wint er. 1996. Growth responses to vesiculararbuscular mycorrhizae and elevated CO 2 in seedlings of a tropical tree, Beilschmiedia pendula Functional Ecology 10:662-667. Mack, M. C., and C. M. D'Antonio. 2003. Exotic grasses alter controls over soil nitrogen dynamics in a Hawaiian woodland. Ecological Applications 13:154-166. Mahendra K.J., Y. B., Rogers J.M., Smith A. E., Boger E.T.A., Ostrander R.L., Rheingold A.L. 1995. Specific competitive inhibitor of secreted phospholipase A2 from berries of Schinus terebinthifolius. Phytochemistry 39:537–547. Marler, M. J., C. A. Zabinski, and R. M. Callaway. 1999. Mycorrhizae indirectly enhance competitive effects of an invasive forb on a native bunchgrass. Ecology 80:11801186. McCune, B. 1997. Influence of noisy envi ronmental data on canonical correspondence analysis. Ecology 78:2617-2623. McGarigal, K., S. Cushman, and S. Stafford. 2000. Multivari ate statistics fo r wildlife and ecology research. Springer-Verlag, New York. Meentemeyer, V. 1978. Macroclimate and ligni n control of litter decomposition rates. Ecology 59:465-472. Melillo, J. M., J. D. Aber, and J. F. Mu ratore. 1982. Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology 63:621-626. Monzon, A., and R. Azcon. 1996. Relevance of mycorrhizal fungal origin and host plant genotype to inducing growth and nutrient uptake in Medicago species. Agriculture Ecosystems & Environment 60:9-15. Moora, M., and M. Zobel. 1996. Effect of arbuscular mycorrhiza on interand intraspecific competition of two grassland species. Oecologia 108:79-84. Morton, J. F. 1978. Brazilian pepper, its impacts on people, animals, and the environment. Economic Botany 32:353-359. Murphy, J. and J. P. Riley. 1962. A m odified single solution method for the determination of phosphate in natura l waters. Analytica Chemica Acta 27: 31-36. Myers, R. L., and J. J. Ewel, editors. 1990. Ecosystems of Florida. University of Central Florida Press, Gainesville.

PAGE 120

110 Myers, R. T., D. R. Zak, D. C. White, a nd A. Peacock. 2001. Landscape-level patterns of microbial community composition and substr ate use in upland forest ecosystems. Soil Science Society of America Journal 65:359-367. Nakamura, A., C. C. Tun, S. Asakawa, and M. Kimura. 2003. Microbial community responsible for the decomposition of rice straw in a paddy field: estimation by phospholipid fatty acid analysis. Biology and Fertility of Soils 38:288-295. Ohta, H., and T. Hattori. 1983. Agromonas oligotrophica gen. nov., sp. nov, a ntirogenfixing oligotrophic bacterium. Antonio van Leeuwenhoek 49:429-446. O'Leary, W. M., and D. G. Wilkinson. 1988. Gram-positive bacteria. Pages 117-202 in C. Ratledge and D. G. Wilkinson, editors. Mi crobial Lipids. Academic Press, London. Packer, A., and K. Clay. 2000. Soil pathogens an d spatial patterns of seedling mortality in a temperate tree. Nature 404:278-281. Parkes, R. J. 1987. Analysis of microbi al communities within sediments using biomarkers. Pages 147-178 in M. Fletcher, T. P. G. Gray, and J. G. Jones, editors. Ecology of Microbial Communities. Cambridge University Press, London. Parkes, R. J., and J. Taylor. 1983. The relatio nship between fatty acid distributions and bacterial respiratory types in contempor ary marine sediments. Estuaries and Coastal Shelf Science 16:173-189. Pastor, J., A. Stillwell, and D. Tilman. 1987. Little bluestem litter dynamics in Minnesota old fields. Oecologia 72:327-330. Peng, S. B., D. M. Eissenstat, J. H. Gr aham, K. Williams, and N. C. Hodge. 1993. Growth depression in mycorrhizal citrus at high-phosphorus supply–analysis of carbon costs. Plant Physiology 101:1063-1071. Pennanen, T., J. Liski, E. Baath, V. Kitunen, J. Uotilia, C. J. Westman, and J. Fritze. 1999. Structure of the microbial communities in coniferous forest soils in relation to site fertility and stand deve lopment stage. Microbial Ecology 38:168-179. Philip, L. J., U. Posluszny, and J. N. Klironomos. 2001. The influence of mycorrhizal colonization on the vegetative growth and sexual reproductive potential of Lythrum salicaria L. Canadian Journal of Botany 79:381-388. Poorter, H., and C. Remkes. 1990. Leaf-area ratio and net assimilation rate of 24 wildspecies differing in relative growth-rate. Oecologia 83:553-559. Priha, O., and A. Smolander. 1999. N itrogen transformations in soil under Pinus sylvestris, Picea abies and Betula pendula at two forest sites. Soil Biology & Biochemistry 31:965-977.

PAGE 121

111 Priha, O., S. J. Grayston, R. Hiukka, T. Pe nnanen, and A. Smolander. 2001. Microbial community structure and characteristics of the organic matter in soils under Pinus sylvestris, Picea abies and Betula pendula at two forest sites. Biology and Fertility of Soils 3:17-24. Priha, O., S. J. Grayston, T. Pennanen, a nd A. Smolander. 1999. Microbial activities related to C and N cycling and microbial community structure in the rhizospheres of Pinus sylvestris, Picea albies and Betu la pendula seedlings in an organic and mineral soil. FEMS Microbiology Ecology 30:187-199. Richardson, D. M., N. Allsopp, C. M. D'Ant onio, S. J. Milton, and M. Rejmanek. 2000. Plant invasions–the role of mutualisms Biological Reviews of the Cambridge Philosophical Society 75:65-93. Saetre, P., and E. Baath. 2000. Spatial variat ion and patterns of soil microbial community structure in a mixed spruce-birch stand. Soil Biology & Biochemistry 32:909-917. Sharma, M. P., and A. Adholeya. 2000. Response of Eucalyptus tereticornis to inoculation with indigenous AM fungi in a semiarid alfisol achieved with different concentrations of available soil-P. Microbiological Research 154:349-354. Shishido, M., and C. P. Chanway. 1998. St orage effects on indigenous soil microbial communities and PGPR efficacy. Soil Biology & Biochemistry 30:939-947. Siemann, E., and W. E. Rogers. 2003. Reduced resistance of invasive varieties of the alien tree Sapium sebiferum to a generalist he rbivore. Oecologia 135:451-457. Singhurst, J. R., W. J. Ledbetter, and W. C. Holmes. 1997. Ardisia crenata (Myrsinaceae): New to Texas. Southwestern Naturalist 42:503-504. Siqueira, J. O., M. A. C. Carneiro, N. Curi S. C. d. Silva Rosado, and A. C. Davide. 1998. Mycorrhizal colonization and mycotr opic growth of native woody species as related to successional groups in Sout heastern Brazil. Forest Ecology and Management 107:241-252. Smalla, K., U. Wachtendorf, H. Heuer, W. T. Liu, and L. Forney. 1998. Analysis of Biolog GN substrate utilization patterns by microbial communities. Applied and Environmental Microbiology 64:1220-1225. Smith, F. A., and S. E. Smith. 1996. Mutualis m and parasitism: Diversity in function and structure in the ''arbuscular'' (VA) mycorrh izal symbiosis. Advances in Botanical Research 22:1-43. Smith, M. D., D. C. Hartnett, and G. W. T. Wilson. 1999. Interacting influence of mycorrhizal symbiosis and competition on plant diversity in tallgrass prairie. Oecologia 121:574-582. Smith, S. E. and D. Read. 1997. Mycorrhizal Symbiosis. Academic Press, London.

PAGE 122

112 Son, C. L., and S. E. Smith. 1988. Mycorrhi zal growth-responses interactions between photon irradiance and phosphorus-n utrition. New Phytologist 108:305-314. Staddon, W. J., J. T. Trevors, L. C. Duchesne, and C. A. Colombo. 1998. Soil microbial diversity and community structure across a climatic gradient in western Canada. Biodiversity and Conservation 7:1081-1092. Stanosz, G. R. 1992. Effect of cherry leaf -spot on nursery black-cherry seedlings and potential benefits from control. Plant Disease 76: 602-604. Swift, M. J., O. W. Heal, and J. M. A nderson. 1979. Decomposition in Terrestrial Ecosystems. Blackwell Science Publications, Oxford, England. Sylvia, D. M. 1990. Inoculation of native woody-plants with vesicular arbuscular mycorrhizal fungi for phosphate mine land reclamation. Agriculture Ecosystems & Environment 31:253-261. Sylvia, D. M. 1994. Vesicular-arbuscular mycorrhizal fungi. pp. 351-378 eds. Methods of Soil Analysis, Part 2. Microbiological and Biochemical Properties, vol. 5. Soil Science Society of America, Madison, WI. Sylvia, D. M., A. Alagely, D. O. Chellemi, and L. W. Demchenko. 2001. Arbuscular mycorrhizal fungi influence tomatoes co mpetition with bahia grass. Biology and Fertility of Soils 34:448-452. Sylvia, D. M., D. O. Wilson, J. H. Graham, J. J. Maddox, P. Millner, J. B. Morton, H. D. Skipper, S. F. Wright, and A. G. Jarstf er. 1993. Evaluation of vesicular-arbuscular mycorrhizal fungi in diverse plants and soils. Soil Biology & Biochemistry 25:705713. Taylor, B. R., D. Parkinson, and W. F. J. Parsons. 1989. Nitrogen and lignin content as predictors of litter decay rates: a microcosm test. Ecology 70:97-104. Tiedje, J. M., S. Asuming-Brempong, K. Nussl ein, T. L. Marsh, and S. J. Flynn. 1999. Opening the black box of soil microbi al diversity. Applied Soil Ecology 13:109122. Tilman, D. 1982. Resource Competition and Community Structure. Princeton University Press, Princeton. Tilman, D. 1988. Plant Stategies and the D ynamics and Structure of Plant Communities. Princeton University Press, Princeton. Torsvik, V., R. Sorheim, and J. Goksoyr. 1996. Total bacterial diversity in soil and sediment communities–a review. A pplied Environmental Microbiology 56:776781.

PAGE 123

113 van der Heijden, E. W., and T. W. Kuyper. 2001. Does origin of mycorrhizal fungus or mycorrhizal plant influence effectiveness of the mycorrhizal symbiosis? Plant and Soil 230:161-174. van der Heijden, M. G. A., J. N. Klironomos M. Ursic, P. Moutoglis, R. StreitwolfEngel, T. Boller, A. Wiemken, and I. R. Sanders. 1998. Mycorrhizal fungal diversity determines plant biodiversity, ecosystem variability and productivity. Nature 396:69-72. Vestal, J., and D. C. White. 1989. Lipid anal ysis in microbial eco logy: Quantitative approaches to the study of micr obial communities. Bioscience 39:535-541. Vitousek, P. M., and L. R. Wa lker. 1989. Biological invasion by Myrica faya in Hawaii plant demography, nitrogen-fixation, eco system effects. Ecological Monographs 59:247-265. Waldrop, M. P., and M. K. Firestone. 2004a. Altered utilization patterns of young and old soil C by microorganisms caused by temp erature shifts and N additions. Biogeochemistry 67:235-248. Waldrop, M. P., and M. K. Firestone. 2004b. Microbial community utilization of recalcitrant and simple carbon comp ounds: impact of oak-woodland plant communities. Oecologia 138:275-284. Waldrop, M. P., D. R. Zak, and R. L. Sinsabaugh. 2004. Microbial community response to nitrogen deposition in northern forest ecosystems. Soil Biology & Biochemistry 36:1443-1451. Waldrop, M. P., T. C. Balser, and M. K. Firestone. 2000. Linking microbial community composition to function in a tropical soil. Soil Biology and Biochemistry 32:18371846. Westover, K. M., A. C. Kennedy, and S. E. Kelley. 1997. Patterns of rhizosphere microbial community structure associated with co-occurring plant species. Journal of Ecology 85:863-873. White, D. C., W. M. Davis, J. S. Nickels, J. D. King, and R. J. Bobbie. 1979. Determination of the sedimentary microb ial biomass by extractible lipid phosphate. Oecologia 40:51-62. Wilkinson, D. G. 1988. Gram-negative bacteria. in C. Ratledge and D. G. Wilkinson, editors. Microbial Lipids. Academic Press, London. Wilkinson, S. C., J. M. Anderson, S. P. Scardelis, M. Tisiafouli, A. Taylor, and V. Wolters. 2002. PLFA profiles of microbial communities in decomposing conifer litters subject to moisture st ress. Soil Biology & Biochemistry 34:189-200

PAGE 124

114 Witkamp, M. 1963. Microbial populations of leaf litter in relati on to environmental conditions and decomposition. Ecology 44:370-377. Witkamp, M. 1966. Decomposition of leaf litte r in relation to environment, microflora, and microbial respiration. Ecology 47:194-201. Woodall, S. L. 1981. Evapotranspiration and Melaleuca Pages 117-123 in R. K. Geiger, editor. Proceedings of Melaleuca Symposium. Florida Division of Forestry, Tallahassee, FL. Wright, I. J., P. B. Reich, and M. Westoby. 2001. Strategy shifts in leaf physiology, structure and nutrient conten t between species of highand low-rainfall and highand low-nutrient habitats. Functional Ecology 15:423-434. Zangaro, W., V.L.R. Bononir and S.B. Tr uffen. 2000. Mycorrhizal dependency, inoculum potential and habitat preference of native woody species in South Brazil. Journal of Tropical Ecology 16: 603-622.

PAGE 125

115 BIOGRAPHICAL SKETCH Sarah Renee Bray was born in Des Moines, Iowa, on February 23, 1976. Her family soon moved to Nebraska and Sarah was raised in Lincoln and Seward, Nebraska. Her love of nature sprang from family v acations spent camping and hiking in Nebraska, the Rockies and the Southwest. In junior and senior high she was inspired to study biology by Terry Loontjer and James Landon. In 1994, Sarah graduated from Seward High School and started her B.A. in biology a nd environmental science at Coe College in Cedar Rapids, Iowa. There she participated in a summer research program with Paula Sanchini when she discovered her interests in plant ecology. While at Coe she also participated in research programs at Kenne dy Space Center and in Costa Rica. Sarah graduated magna cum laude with college honor s and started her graduate career under the supervision of Dr. Kaoru Kitajima in the De partment of Botany at the University of Florida. While at Florida, Sarah pursued her dual interests in teaching and research. After completing her doctorate, Sarah and her husband will move to Nebraska where she has accepted a position as assistant biology pr ofessor and curator of the arboretum at Midland Lutheran College in Fremont, Nebraska.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101130_AAAAAG INGEST_TIME 2010-11-30T06:13:22Z PACKAGE UFE0010945_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 107270 DFID F20101130_AAAFAR ORIGIN DEPOSITOR PATH bray_s_Page_043.jp2 GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
e092d0886e8ffa2d0c1e3239af5d138a
SHA-1
3871759e31592e6dae00c92a3e5e88238321d8d9
1333 F20101130_AAAFZB bray_s_Page_107.txt
acdf9d4462ce7a229379d86fa554fb2b
257bb2458f28d15854286ef2fc183af8a6f14952
1903 F20101130_AAAFYN bray_s_Page_082.txt
b72037d8de70a318ed7b9fdf30caa09a
a68000c0436d60a240e2ea4e05a6c7b1e5ef6c26
1968 F20101130_AAAFXY bray_s_Page_047.txt
33de54e6ccb3d6d76a5327adc6f6a7bc
fe4a64183eb611dd9755230a4f0f965e40e6b4e6
1053954 F20101130_AAAFBF bray_s_Page_024.tif
f451b860ecaeddc2befb5eb4d7a75be1
0afa7f10552d236cd1154b4d997cfb4b4501b66d
2714 F20101130_AAAEWA bray_s_Page_069thm.jpg
cc97a77f136d77a3d2cb3d1cc1d80949
cdf64669a2076eecf47adcea1d2db3964d9e0f06
23861 F20101130_AAAFAS bray_s_Page_023.QC.jpg
c81592733e6220bbc958fa99a2b0bc9e
0be262194c0743d1348b9cc9b055a2d5d6bc5e1d
214 F20101130_AAAFZC bray_s_Page_112.txt
04cef3fc2e5242b6c1fd48cd8822ec43
0d2366ed287e1c4b276c7c14ddd226beddd0c353
2028 F20101130_AAAFYO bray_s_Page_086.txt
ba3be0c1edc0ea9d24eca01b0f65c639
81dd3278df23991ee0daede6a0a106fa3ed0e8d8
1935 F20101130_AAAFXZ bray_s_Page_051.txt
f13881998c3ede7a5698c53071eee7ce
d42fd5d7ad6008e7a492ab1bbd7807059fb56ed6
20050 F20101130_AAAFBG bray_s_Page_096.QC.jpg
061a1b70703973726e3fa9bafc3b0ac5
4d49ad27e7be511e60a5f95847c0bc80ba3477f6
16779 F20101130_AAAEWB bray_s_Page_097.QC.jpg
835434c3eb4c298004e250e9c1d568db
3fd26bb4cc1477813c7a9d28b18b145af25fb321
64800 F20101130_AAAFAT bray_s_Page_011.jpg
0523028d7d25bbd8c22ab0c65be9aec4
b9ba0bfa0095872694a45687039383073e94e3dd
2519 F20101130_AAAFZD bray_s_Page_115.txt
121618bf784f1ef2104c262d4dd7a7c7
1f36ea2195b4531c287fbcdd1a5dccd554913e1c
F20101130_AAAFYP bray_s_Page_090.txt
4e91b8aa3a0f6c00801b20a8d92e4fde
0cf42c9ef02ff38ffc9c5ae4f1901234620bba0e
19266 F20101130_AAAFBH bray_s_Page_010.QC.jpg
d81bec8a5b4698e929e64253d2c01ef4
a31e7b2c345acd6f54188c1d694bb01f30f2264a
93167 F20101130_AAAEWC bray_s_Page_116.jpg
e33c6c2d031f8c9474ee3f961db1aa69
1d815a3ac2219983371670c2e632afd48fd16f26
1848 F20101130_AAAFAU bray_s_Page_048.txt
7956712899f31a7a519f5e1c21e4a41e
5c23a73289abd46d70dacf7a6b8625fdd01f7ca2
2671 F20101130_AAAFZE bray_s_Page_120.txt
6a12853f3fb18524a0b05e3e3dd7bb12
ee9326ef699599ea596736aae51111401ab6a78f
513 F20101130_AAAFYQ bray_s_Page_091.txt
09f544ab3a3a249220dcea6b40b78407
7831f4e3afd72763f7829fe6d93fd6aba8e2c8e9
122026 F20101130_AAAFBI bray_s_Page_118.jp2
c42f37fb9cd6c18c5d084df26af7debc
b0d9d36cbc0e43445113a88049725e109c4fe797
23514 F20101130_AAAEWD bray_s_Page_033.QC.jpg
4d5a346245578cc06a2874e8fc65568a
7ed141ae8c58a83e171252837c38864b1325ff8b
3465 F20101130_AAAFAV bray_s_Page_107thm.jpg
6e5f07ded487053121651236d4deb208
81361fa55cf44e771fc1fa9a03b63089634a7f58
2748 F20101130_AAAFZF bray_s_Page_121.txt
4111b0f20b42b40456326d70d096fab0
b79b5b7304ec71379429ea41968f0fbaa8df408c
1934 F20101130_AAAFYR bray_s_Page_092.txt
bd96ee671a0478ac59bdc208983db813
45a6f08ea753281cb1f0a31a76f0460fb8e44c3b
110190 F20101130_AAAFBJ bray_s_Page_088.jp2
620453d365ffa896a363783a04733fc9
5255522b968f43fed080bdc2b556bba2c8fd639b
39844 F20101130_AAAEWE bray_s_Page_095.pro
764a2e05efa87c8c40f1798e47fbf229
0f3b66f7da74abce1ea576ac894af98fdb7622f6
26363 F20101130_AAAFAW bray_s_Page_120.QC.jpg
94b36d23f608609aaab072e24f773d8f
9bf694e7cf62dfea62c4c12609f3d5d5a3b97ecb
2405 F20101130_AAAFZG bray_s_Page_122.txt
9674b9f9e67357cf15dc79a162cdd1b7
6f3e3ae8908f70aab9db34d79fa290d6f89706a4
1173 F20101130_AAAFYS bray_s_Page_093.txt
be66e97eae8559ffce82060f1e8af2d6
5cdc8749cf809fc0e665f0475baae0b356892695
25367 F20101130_AAAFBK bray_s_Page_119.QC.jpg
5415ac416ec4cfb4ea9f9c9bb9670c88
4aff10088cb881a941875f1030e2595952b5b120
70829 F20101130_AAAEWF bray_s_Page_044.jpg
8cce563c7cd2933fcef65f0525b545dc
0cd421a903d220098a9dd3ed716f894d70d32a94
1888 F20101130_AAAFAX bray_s_Page_108.txt
081e1a4d4fb48c08c27bd90015824e4d
ccdb4759ab96d1ddfa095792e3582e595c315c5f
2436 F20101130_AAAFZH bray_s_Page_123.txt
ba0ea2ef5079009fb9d82695d6821d7e
ce258a1dacccd4ed636c8c7c662de77899c9d8d1
2886 F20101130_AAAFYT bray_s_Page_094.txt
d28b0fc77e8b4ab96d8dd64dbc373289
b994369be10d685a7a096f2469572272f520c5cc
49177 F20101130_AAAFBL bray_s_Page_076.pro
61c77c94da5bb7c7fb8e1b0eca537d53
f740ab50518b526d4fd4b1e8317fafc8120ae670
2395 F20101130_AAAEWG bray_s_Page_118.txt
c89c7f00a21e3555eac5a0d75d360ba6
e0b62e75c18c127d2e94a68c348e488b0b0dfc66
21926 F20101130_AAAFAY bray_s_Page_015.QC.jpg
371b7097755b8267eb5305df402353d8
5666f1df4e53abbba766cd8d0a765b05bded1b22
956 F20101130_AAAFZI bray_s_Page_124.txt
d2b53586e71f8d4ebe82aaa6ddc621c0
64cc090279b8bb31105394d7669d6d78020e6c70
29517 F20101130_AAAFCA bray_s_Page_061.jpg
192a2eebb6a804610378eab7479d7965
adf2213f365e1274f43611a6317d57197ec01d26
3294 F20101130_AAAFBM bray_s_Page_106thm.jpg
5c4756e89926f363bcd8eed3740228cd
ca25c1806e1872aef1e8c8bc07a03438ef44bf60
2022 F20101130_AAAFAZ bray_s_Page_073.txt
db00f5dca0777911ea617c815917dad3
3fce92778b7d5277a44a70cd8f446877489c8f61
1347 F20101130_AAAFZJ bray_s_Page_125.txt
c5c0131899e68b9edbffe28c001cf532
6801d59099396de8dc98ea8974e5221b8736ba8a
2087 F20101130_AAAFYU bray_s_Page_095.txt
590b56d57916101d16387e9dfc4a1bc0
53bcab28b7189d457be7e7c3ceb76fb35e85cc09
6356 F20101130_AAAFCB bray_s_Page_110thm.jpg
2ab13cceaa439883859e7fb7fb5d5241
235f69d49b24d43fc25b7d37b197a5ed10838998
23963 F20101130_AAAFBN bray_s_Page_087.QC.jpg
809e16265eea342e70990fe32b3fb97e
d6ad6322e64225ff21a80e0e97a937aa31efc162
F20101130_AAAEWH bray_s_Page_057.tif
f8fd21ea39bac78ea59d7c2b6746c291
6f8b9f7953918d38a86545cd47beb1e33ef172c5
568382 F20101130_AAAFZK bray_s.pdf
4ae71e81f617aa9475d9479db9b9dade
51a4fc1cac0d5c9802952c52303e8a874b722515
2307 F20101130_AAAFYV bray_s_Page_098.txt
df180f5c20ee71c1d7de5329df4c614c
ab6cf29a39f0b44631668e75f0fe469f0fcc841b
269722 F20101130_AAAFCC bray_s_Page_104.jp2
5974b155c2beb4514604edc602db6a69
6d519cf584a207c5a2b35c32c62300726fcd5970
53634 F20101130_AAAFBO bray_s_Page_098.jpg
f48c1b879a849168958c1c2e8566629e
c0d9555d1ff9c968993a128edacc4317ebae4d56
2683 F20101130_AAAEWI bray_s_Page_116.txt
80a91e57cd120639d56025a13da24d66
382e6f7b94cd9bccaefe87cc39e07c26ecbcf1c6
2600 F20101130_AAAFZL bray_s_Page_001thm.jpg
5969378c1eeda0590d4405a947acee19
8575cbd7379c95605acd6f6595b042246d55a0e4
F20101130_AAAEVU bray_s_Page_020.tif
90da6e8c37d4b719a7920c671cd603cb
415dc55e6b07e7598698024c0cf9b50c475012e3
1852 F20101130_AAAFYW bray_s_Page_101.txt
74f676c53e4238cf65a0532bcd9ece1a
be47566b0c904edbcb91fefa5539cd82aebf6151
90618 F20101130_AAAFCD bray_s_Page_119.jpg
501c6a01a2cf566fa4eba5a534b4183d
3dea57749e28e5a6ccc5c17041d3b5d6f43a97f1
1341 F20101130_AAAFBP bray_s_Page_064.txt
610f680e6deb7596ffd07f1923d2d18f
88a87406a654ae281bcf3baa3ea072124bd9b084
2043 F20101130_AAAEWJ bray_s_Page_084.txt
a22c963deed9f6e41bc221cf8ebee075
8429c6b72d2a3dd92e626c8e66d224fb005b8148
3295 F20101130_AAAFZM bray_s_Page_002.QC.jpg
7ee775db79ff82638c28e9337f7d57dc
8c33d322852f8b66c3ffe4ccd3e41f341cb3b228
21281 F20101130_AAAEVV bray_s_Page_011.QC.jpg
ce288b587dcf432764f2390c60461970
05d50429bad876b4b12a2ec824d1d5286727868d
3892 F20101130_AAAFYX bray_s_Page_102.txt
f96c48a8501f385d07edbc9ccc0ba815
0b0a1ef5cece16a668d004f1c283d838b657c47a
57093 F20101130_AAAFCE bray_s_Page_040.jpg
1005df049cb3a22d6bce5ed9000b274e
53f8cbed0c4178386781234a834c18d9ecf067ba
F20101130_AAAFBQ bray_s_Page_001.tif
4387025eccc8bcf5341e332755aa2cad
bc48931fae521fdf8284d065d6c423e0a322e3e9
50396 F20101130_AAAEWK bray_s_Page_023.pro
8deadfac595b40df430a1e5c6b361b3e
32f7784fb27b8d88394a2fffac189e9811e0e69f
5177 F20101130_AAAFZN bray_s_Page_004thm.jpg
ea8b3f02ea4176cb135d6b0a58f261f2
89b2bdc7861283942a1479b8ef6a375ca54e0492
2056 F20101130_AAAEVW bray_s_Page_056.txt
23030f0fcfe907856320c98676f3929f
c9158961d956c325312b57151776ce3d98bd39a3
2411 F20101130_AAAFYY bray_s_Page_103.txt
7805cbe1d95ab45bb5f77fa406e98a87
5bed7d5f79d60399dc57a548aa0320aa66d45f39
68315 F20101130_AAAFCF bray_s_Page_017.jpg
c3091bf1b7f8649e7e904573c9d0382b
82b5f4cfe99e7ee57f7c106c1db0c147a08c2d7c
9219 F20101130_AAAFBR bray_s_Page_061.QC.jpg
d2104eea9f60cef34a6483460e578d35
d6cf37f886a5692d3ef9344055c79bd718da8fbf
69802 F20101130_AAAEWL bray_s_Page_079.jpg
705d682a5763c75f24a24dc99cf29674
b1d04114c05bcd0027034b1b54427e73cb79a644
16277 F20101130_AAAFZO bray_s_Page_005.QC.jpg
7ccba3f388c66ac11f4c4dc91317b691
8a175bb4cc33e20ee1f824ceb77c6c9d9bd5e1d0
2911 F20101130_AAAEVX bray_s_Page_062thm.jpg
2580359d93f10d7cad1278b8be11532a
547d7c440a42103b919874c9e46c44c187c2980a
1700 F20101130_AAAFYZ bray_s_Page_105.txt
3578d3d0b5e38afa051fd9069346c634
48b83478c82f758668c85971f9f133a608aaf88a
F20101130_AAAFCG bray_s_Page_083.tif
b8e1c8d7ddf315acd46cf9251bd34d42
b2bd5aeb440a3aee2f8c6d00d1914cdb7e5fe4fe
39463 F20101130_AAAEXA bray_s_Page_065.jp2
18eccdee7c5a4f7c10bdac3b85bf5824
ec5123db6812d15e6360087fbbf5253243005feb
13333 F20101130_AAAFBS bray_s_Page_066.QC.jpg
fd94e9038e7da132baf7d135c1c656f2
d11aa5aa49f695c56c0cb81322a36036679ef9a6
F20101130_AAAEWM bray_s_Page_032.tif
0c45ba8667adda28423bb344e028a27c
8277edebd80265b9e785fc8af920320b8c4f8172
4246 F20101130_AAAFZP bray_s_Page_005thm.jpg
62762a88a7a2dd00e4b19f88047b295a
0650a85797329388723e9573ac8eb313d6439c55
23115 F20101130_AAAEVY bray_s_Page_084.QC.jpg
35734eb08ee6bd40a4768740389a8f0c
52ded095ad6816e61f633f88173d86b4c45232a0
6428 F20101130_AAAFCH bray_s_Page_025thm.jpg
bd6b7e2fb6c81d8d8ef4ffc07bca6fc0
143bc0d523c235791a2b9477d94bf29ff0ad8137
3626 F20101130_AAAEXB bray_s_Page_068thm.jpg
177e3157c5e84438816b56b537f179c6
bd827f399c093ec7794b6b95c01e25aebeb67e59
6318 F20101130_AAAFBT bray_s_Page_032thm.jpg
70f6504e72d43aa1a033f4833227ad6b
beb9285556bf7be23fc2e02e4de6aed9d0fb312a
2052 F20101130_AAAEWN bray_s_Page_016.txt
838f0a3d4e3642eaf6efe3ca1d0ea5ca
06490da2dd2e819a0022e22fc0b6a3c2c102fee6
7104 F20101130_AAAFZQ bray_s_Page_007thm.jpg
dbd7b375503ba59282562afa4bfa4790
193475abf8cfc9985bf16296acfe491ec3466e8d
10456 F20101130_AAAEVZ bray_s_Page_068.QC.jpg
4c6e55b7343870871dd82e69cc200006
9a16328c3f96bc6f9d2ef60caff2a3d309e4fa17
11834 F20101130_AAAFCI bray_s_Page_091.pro
8088019311666e87f935f07ce6ab8408
d1386f055cba7fb2633757299fc3090a918b6284
1279 F20101130_AAAEXC bray_s_Page_067.txt
9ef6dea8d482d22364f913fa510a16da
38d613ed129c43a8e5de8f4a918ca583273a58d1
65403 F20101130_AAAFBU bray_s_Page_108.jpg
9b6597289028103ae9d3fc38f28a4f99
9d91b287224428552b9fca9f337bf0985b7581c1
111368 F20101130_AAAEWO bray_s_Page_014.jp2
7aac908cfbf2d428ce02fab61cde7bf1
03dc4502fefd8a5f46be67cdf67eba8a6b9241b6
5967 F20101130_AAAFZR bray_s_Page_008thm.jpg
7a4fd2dfe8cc7d2bb123ce6b1fe982b3
5048f9c65d96b21d2f8fb338823db099847ef863
F20101130_AAAFCJ bray_s_Page_022.tif
a13af88a5537d5cc81a66a90155f421d
0d199a930f53efcc832c8ad201be54c61fe67821
4921 F20101130_AAAEXD bray_s_Page_125thm.jpg
8de24ab4f03a84b54ef767e3afa24302
fa82a68ff6b82d7f69c3cdfb39a8305f5d3dfb56
F20101130_AAAFBV bray_s_Page_109.tif
c6ad15e2e9b9214cad265c79c80d677b
a314a238ebdb22f3fa2041e416aa5b8bde581699
2055 F20101130_AAAEWP bray_s_Page_097.txt
fba2dc62a644251ace626760851add85
af40f4a4d27b4f96d2210b308b58bb9cd5c8e3a4
2002 F20101130_AAAFCK bray_s_Page_052.txt
c69c254997c9ad92b9649cec9c98a847
e17747da49f50777dbd1d9fa3d9395ff16f45827
5776 F20101130_AAAEXE bray_s_Page_002.jp2
c3d33b06c2006006234d78f523eb1ce2
b9c2e6fecb04ce58f2f3257510b02ff92e3798a8
52237 F20101130_AAAFBW bray_s_Page_056.pro
58e235a94e70e29a13bab9247a9c450c
17d12e1ea24a1ccb44dd78faac9c0abd4c6278cf
4408 F20101130_AAAEWQ bray_s_Page_104.pro
24f685c8b31f886764611eb39e3044fe
05369c27b37dde2ae768d744331060f912f08942
4939 F20101130_AAAFZS bray_s_Page_009thm.jpg
6ed1449be863253c3eef1e9be834ae21
7be055dc80bdbca0a8437cdd2c6c44041343f41d
51714 F20101130_AAAFCL bray_s_Page_101.jpg
676feea6fe2803fd4ba8e143bdde948e
c40d29965f990215f2e0b848ff8983e452eaf383
1054428 F20101130_AAAEXF bray_s_Page_093.tif
53bc10bd10068c75b26c6bda4ee9a73e
f9472481a95f09e2aa7d6f8bfd8edb89af997443
649 F20101130_AAAFBX bray_s_Page_037.txt
a8d0ab1e8ddeb9017fc2ffd662bc8d85
c9d1de0e88a2deeef6056a84a687efece78e0735
70526 F20101130_AAAEWR bray_s_Page_034.jpg
de8ed99d30465ad19b72665bdb2628a2
d18d9616bb8d01d7f57a0ba35b25fc6a6e88364b
6001 F20101130_AAAFZT bray_s_Page_011thm.jpg
ea539f7b79127e2e29475aa59b5027da
f0d1b2469ee5df235a31210abf7d489ae709e3de
58699 F20101130_AAAFDA bray_s_Page_010.jpg
ff0d8ca2dd8badbd8c9465bb62861764
8ffe925532c883b57c578cf643cd524a8418a761
3327 F20101130_AAAFCM bray_s_Page_124thm.jpg
c17e97087d445046991befbc4f73432f
8c85b22ac01069d9d0693c8fabe739978cc9c455
1926 F20101130_AAAEXG bray_s_Page_109.txt
0aa948c16310c5b4c6396cce8d0d2064
7262e5e4b86685b41968853de815b7ca45024ba9
104681 F20101130_AAAFBY bray_s_Page_013.jp2
31a980604fb61edd28ef54727ea6ce7a
cb2cf3b93485530cec4929b490af822249370fdb
6581 F20101130_AAAEWS bray_s_Page_026thm.jpg
f10c9ddad287d659a30d8d5b63c8804a
be89b3fdab7383d8a94e3b06016dede03d3d8cee
6498 F20101130_AAAFZU bray_s_Page_012thm.jpg
e026eacd8d9be39380f6dd4ec4c7dbd6
d82948f0167fa8ce3decf1d51c686b18b7367101
F20101130_AAAFDB bray_s_Page_065.tif
453f56f03123c3e840140d5053f4d851
db47f727326a6303e2b57d697720fad414a7bd70
23351 F20101130_AAAFCN bray_s_Page_012.QC.jpg
a71dba03f33c5fff527fc8c4b0c41237
32e286bbf8fd9ce38aeab16b700875497f7c9e5c
F20101130_AAAEXH bray_s_Page_099.tif
2dba5ebfc095bcb9c28f3ce6d90e710e
75cc43a95851c13ae9eda11389e0a140b325a15e
67456 F20101130_AAAFBZ bray_s_Page_050.jpg
57700a1ef8ff06d6efdd3856b2b6d072
7aed7181159a3f55390f7d9de3549297ddf6550a
107076 F20101130_AAAFDC bray_s_Page_049.jp2
6113a993ac7c41e0901240ac87f9b9b2
a973d5b8e01a1fdfd64d2fa35e0b358ec420c7ca
F20101130_AAAFCO bray_s_Page_050.tif
aa5bb8e85502c5ab4d231794cc5f46db
3b03425149404f18633fddf72a1b2b1fb832f8ea
11786 F20101130_AAAEXI bray_s_Page_058.QC.jpg
e97abdcb2aa8c391730be5471c5b1e54
5979a585c221f25b5e2ed0f32b1d35870abe8cb1
112788 F20101130_AAAEWT bray_s_Page_085.jp2
faa5511982d6ab28c4e92a9784dfe9a3
9880149cafc67c9c20b81a594f3c09404cb43319
22039 F20101130_AAAFZV bray_s_Page_013.QC.jpg
15c84c3c56df5f78f0cf9579007b61ef
6d88981cfa140fef787996bdd766464082571778
23456 F20101130_AAAFDD bray_s_Page_083.QC.jpg
0f01a2219dada797652a363e6dc1709d
f2174d0b1e240646494dc25f3ba58d88f7dcd0aa
8423998 F20101130_AAAFCP bray_s_Page_063.tif
63e4c48dc0b78b27757eb64256609bb9
ddcee4ac83d68325ce386adda470d91b6faf9228
46088 F20101130_AAAEXJ bray_s_Page_040.pro
34d80a620db2526fdffbe216f0d0288d
b0397f0427369984e508d0091112a5822352ca4b
4310 F20101130_AAAEWU bray_s_Page_112.pro
5424fd5730c3438e20987228b5c851a8
97f549edb9d78f60986a094cc96433bbb12007fd
6148 F20101130_AAAFZW bray_s_Page_015thm.jpg
b76ff62f1d6eb25a66492dfe90b73705
2f201fdbe5f56de53efde34e4d6ff17cb428ff91
23592 F20101130_AAAFCQ bray_s_Page_028.QC.jpg
39d3a6fb00c47e82e5981f999bee0574
5e2c0a53f862ae24ebdb287ac7d7e83e66884b7e
101935 F20101130_AAAEXK bray_s_Page_080.jp2
71b1e4328dc13d2a8b2d0e8f955a4447
368747a5dd08b48649f2f05ba30a502992c4e813
5784 F20101130_AAAEWV bray_s_Page_096thm.jpg
811b247fefb7cb872076cf857e7204b6
f66f4a35c75a181d3afaa5635425ab57b3017ffa
4552 F20101130_AAAFZX bray_s_Page_018.QC.jpg
616cf3c458e3f9c26093ccd7a803ee05
1b67691ce62ba96da8c75b579a39b8dc7e6db637
25271604 F20101130_AAAFDE bray_s_Page_007.tif
2ff9431a1d78b7f7d4255c7b56ce6a92
98135b8d0749845e4641f74b3021ff02e20245a1
2648 F20101130_AAAFCR bray_s_Page_117.txt
346968dfb6bb35b631412788bd91eced
5bc1b90c8743aeab0bd3a955324f3078e62dba0c
108338 F20101130_AAAEXL bray_s_Page_071.jp2
3a9bf2a63d5dae626ee41f7ca403f463
422be0a4efddf6546018ec9be83ff8c1396ef511
48195 F20101130_AAAEWW bray_s_Page_013.pro
3f70d2104c03d2035b52e0bc889645e3
f18e7c03999bd084f58c9d1dd38a100c1da57de1
1679 F20101130_AAAFZY bray_s_Page_018thm.jpg
61bcb700b5d12c6a7c737c9a0de51a75
bde234f2819419a548007cc125c7bd56f7567f56
103 F20101130_AAAFDF bray_s_Page_002.txt
4753068e89fe40230139f30a295cd94c
2eefa89b8047b6bf95e14153bea4edd27aab7e37
F20101130_AAAEYA bray_s_Page_015.tif
3b9716e1629e587fd04d78fbee4130b8
3d4a77835e03b229fdc5a9cb2d09bd8ee8d44415
35745 F20101130_AAAFCS bray_s_Page_061.jp2
8136215781d5de8a0843fc6ea695cb03
f819263b54fe65c32c71e8d6e2b2950ab34c0b5e
2418 F20101130_AAAEXM bray_s_Page_119.txt
246bd9d55e44381e5e8df5dbd94d9e2e
3230f6d24f78ac40e0fcf7a7565094e272af87f6
71904 F20101130_AAAEWX bray_s_Page_059.jpg
d4b0c40f9c235ae2953113aa243b7768
3565554bd1c2dd9b45d8c844a49581a7608071f7
11854 F20101130_AAAFZZ bray_s_Page_019.QC.jpg
e7887688e5fffb05a7c16ebdb802e46a
5c3dccb6004cdf54d1c41232e177e2dafd0d49b7
73236 F20101130_AAAFDG bray_s_Page_113.jpg
98d26b08df8a3e286ce6a6b56a43dd1d
95361568249967a8f884d810cc82be0fbe2d9603
1958 F20101130_AAAEYB bray_s_Page_027.txt
f53d0862b0591fb142d480638fd2d322
f36af9e0a43a54fcdc6e4397d8d1c505e2e71165
9289 F20101130_AAAFCT bray_s_Page_106.QC.jpg
15a1b0270d1dfed3c3e7b5d4a4f23e8f
354f79db0a36f2758093357dd933b95d65f20b1c
98496 F20101130_AAAEXN bray_s_Page_003.jp2
f6ebb635d1e7801a0fce55287ab472f1
9c43342f93b00a41718809acedb7a2ebe4c42e5e
777 F20101130_AAAEWY bray_s_Page_062.txt
2317baafab950738b2f94dd12e9f8dd3
511e5088effac8add7eff9b748282f0b2e283f58
48548 F20101130_AAAFDH bray_s_Page_060.jpg
1f22040c80eed9fa55d8eb76cbbf8f75
20717ebbf232bca39d9b0ee00117806bb0faee80
110948 F20101130_AAAEYC bray_s_Page_086.jp2
0397aba7a38fd0d8d9db596c0cef692a
abbc92bba6fd5cccff2c772e187d78167e3e1845
F20101130_AAAFCU bray_s_Page_087.txt
bc2184658b9d72f3af061c4180f77aee
c8ecda4b1cefa60d5c20c563f65a4bc093cedab8
7156 F20101130_AAAEXO bray_s_Page_116thm.jpg
d8e7e173bbfc763f1f4871f40ab3fd1a
012eefce733ed09564b898c7934e7a9329e31d33
F20101130_AAAEWZ bray_s_Page_016.tif
5b66e2eec11dede1da0ee777959d41c8
a40312702b207e4d5d49a9bd751c64e56b4efcd7
F20101130_AAAFDI bray_s_Page_123.tif
85e310fd46582732c938ef484e76c161
6fd0197df5963bb0912339db61d434f6e490c38d
50727 F20101130_AAAEYD bray_s_Page_099.jpg
16b9d73f9333465d530bd7fdde76146f
3f18a1f5cdcf7cb3b0109f5b2378d6de304501c7
22168 F20101130_AAAFCV bray_s_Page_017.QC.jpg
58dc828de077342deb060de7dfeb8ed0
1ff3a98d766bc75f011bbca70a262cd768ed7f00
6421 F20101130_AAAEXP bray_s_Page_036thm.jpg
1543cf4f3e5f9981a4eca0707755a8f4
b05db55bc880c26e9912856c21304c54b476a18a
1673 F20101130_AAAFDJ bray_s_Page_058.txt
df23d9a9baa7055f322467b3192c9b22
807f8181b9502849edb91205c594c2ba3250afab
6077 F20101130_AAAEYE bray_s_Page_077thm.jpg
3aac0cb710bedf6f7dc9b0fa901c6e85
ed9f0be545d20adae9f0a6537641e2392ff2cbcc
15904 F20101130_AAAFCW bray_s_Page_095.QC.jpg
d391c1e2b334f4eaa4285113ad4b25bf
3bcf4edd6d175a33c3be400378a119837a69ea96
18898 F20101130_AAAEXQ bray_s_Page_094.QC.jpg
9b9c5362bc0c830e81463bd047c98a0f
0536eb482e1d712b894f486a61dbb1da3ee2e6a0
109856 F20101130_AAAFDK bray_s_Page_028.jp2
b7902864e94d6966f24c02af1a57ce92
4fab2341ba0fb48d866c1fe23825a3f3b8cc4394
6455 F20101130_AAAEYF bray_s_Page_086thm.jpg
ce7e25088101a1d9f4e8badf65801d8d
6139f5880f133b7e03313e1d040504273dcbe3d9
15513 F20101130_AAAFCX bray_s_Page_065.pro
30d21af5a4429267ef1d22573117619f
2b632b782c9407528959cbec7c07c2fe03a6895d
4562 F20101130_AAAEXR bray_s_Page_060thm.jpg
bfbc1aeedc8b29fd58fe9e6ede547ec3
017b460827c1e38cc0349bf47f6e6fba99267493
1051982 F20101130_AAAFDL bray_s_Page_008.jp2
e2bf172862bde21fd7aa5edae9bf31a7
261f0e42cb5972ac3bb0683d941efa88057a4dbf
2639 F20101130_AAAEYG bray_s_Page_114.txt
0378ac27ae3ca13853f7669a49ba25d9
7fe162599aecad49ba7a76921b46840022e5fcef
1567 F20101130_AAAFCY bray_s_Page_100.txt
4d8e2b1040aa6f9ea33cbc3da31356ed
90f93cf177ef400a9592bc4b150db98aaecfae09
49573 F20101130_AAAEXS bray_s_Page_089.pro
4e4eba91a0b6dbf1c499081193b06e23
383021629fb24f113dc5df9b90f175dd5f413e96
107968 F20101130_AAAFEA bray_s_Page_044.jp2
8617787357d5d09d21f933fd04dbe736
0f30a0b12f890da2fb9f1f7db12fbf68341e525d
22457 F20101130_AAAFDM bray_s_Page_008.QC.jpg
279370d0078fa3239dabbcd8f6f1f2bc
31ff6d3dcbab88568537b150ced303256f77a1e5
115483 F20101130_AAAEYH bray_s_Page_035.jp2
197a2260701b14d2261b16719daea17c
bd1f1ba28188a3082edef8c110072f6b9ea5d90b
17221 F20101130_AAAFCZ bray_s_Page_125.QC.jpg
f5c8d24d26467f6c7adc145fb0149dd7
dd899b60af455bb9d61996e4ea603529dfebf67d
4261 F20101130_AAAEXT bray_s_Page_019thm.jpg
12716d019e177ee566c686404e55d25e
ee31b951cc3047564fb93f4cfc0e34ec38b56456
49682 F20101130_AAAFEB bray_s_Page_083.pro
9498d5063f6a8718b9ac2b8a9239ee9f
5a2f1ebfd7b42bbcac5554ec20483903137459cd
64731 F20101130_AAAFDN bray_s_Page_100.jp2
39e72235ef19f6cf423f841a1dbc55be
07e9a1fc4b1c6613628e93a1f57b357084db58ef
4358 F20101130_AAAEYI bray_s_Page_101thm.jpg
9017bc69b25d14ced1e4c8b42d9bb6a8
9f774e5160ed633a9fe61fa04a165dc7ddf6ea5e
64884 F20101130_AAAFEC bray_s_Page_117.pro
f616994d4d5dce56cb7e8c0d9c7cbe91
04a70be5e8449370fbdb746f1c35776e6e9e3772
2035 F20101130_AAAFDO bray_s_Page_088.txt
3eba42bf6289ae849175c590b108e274
1fe3484efb0679152980749c051368fad6ee665f
73558 F20101130_AAAEYJ bray_s_Page_026.jpg
3ea7b9d895b344da82ec7be608c71764
6575cb6d7f82caf4e3000c7089b27e95264fb39f
5789 F20101130_AAAEXU bray_s_Page_038thm.jpg
1b71d548296bbbc9a55cc7493a72fccd
13d878ed7bc964469a72339100436496668cd908
F20101130_AAAFED bray_s_Page_107.tif
54e89b0878e17903f320a835b906d4cc
7d4a453530e579eb74640e92c70988ac13716000
49156 F20101130_AAAFDP bray_s_Page_043.pro
88fed5a50472a4e43f1112e0d7952de0
1887cf57d4e0fc90e55c6e2180564f4ff61e1cf6
46647 F20101130_AAAEYK bray_s_Page_048.pro
01d2e6f89df03fd0287423a5d5ce434b
6c3f87047cab5145efa379191e5b2ab2497bb90f
F20101130_AAAEXV bray_s_Page_030.tif
618a64f14ba28d743e363bee43b7aca8
72dcae9942f0c832f5ef28199090e1f39f40b739
6934 F20101130_AAAFEE bray_s_Page_114thm.jpg
d23beb9acefdb326f2c6a1b1668b0e1b
1febe854b3bc7fb081660298db144999e173856a
6505 F20101130_AAAFDQ bray_s_Page_016thm.jpg
69c29484299b9f8f2f93eaf44c70b375
1c4d098fd62a10f0e4b27e541057a2f51550927e
44768 F20101130_AAAEYL bray_s_Page_042.pro
322f8d1d9f85b11176236fdfc02372e6
0042e945ca415622921673cc5fb4318f3fa5e155
88644 F20101130_AAAEXW bray_s_Page_114.jpg
74869542ef6f6f5cfa6497c1e0dba02d
38da641940b798965c4e819810fcc23310f361bf
1963 F20101130_AAAFEF bray_s_Page_083.txt
ca0ad68bf1795a46defb50cf440bd64c
9609d667a25e04e5cfc417a7864641c79e17c08a
F20101130_AAAFDR bray_s_Page_077.tif
1465edad35c6a226c83ed9e6d0588355
b692c5c8cde90aeb05663007dce7e15e8ee698ad
1609 F20101130_AAAEYM bray_s_Page_009.txt
55a34842942411688dbf0b6c9eebe5c8
16d2d5fc6e34f245764d0c41119382b48404d652
F20101130_AAAEXX bray_s_Page_010.tif
afc914dea669571e1f1b77b1145ab8d5
e9fcca5bf07f3624cf374acd26419c28c3763973
25822 F20101130_AAAFEG bray_s_Page_117.QC.jpg
e47b0fd6e56b4fa50b59c6e497ea9cfc
36550285bf10525e76d228efb25685d64d8bee45
139704 F20101130_AAAEZA bray_s_Page_121.jp2
6e3b7671dea30c5192281b9a60e9c405
603d82bab1336a1ab8673cb7e750f004c6a7afe4
97401 F20101130_AAAFDS bray_s_Page_020.jp2
2d7531d11a38fe00504dd31214941f35
f0ced77a7c6fafeea03f91c653dc46c17213a6ef
F20101130_AAAEYN bray_s_Page_013.tif
b3e845315b9197839e12122172aa0f78
15e0ad478a1f3b5e229312ce592a1be274216209
442 F20101130_AAAEXY bray_s_Page_069.txt
115057e6abd32c48e1dbf4288624e26d
d2148b360c84366da9eaf4f8a00b162de5b4f1e5
21328 F20101130_AAAFEH bray_s_Page_108.QC.jpg
ddd2e573a970a94d5bfb693fc9b27107
e9cac1e37cf3db2acc22ddc81f4c3ab2e315beb8
98440 F20101130_AAAEZB bray_s_Page_011.jp2
8a91d7499b5085892b1aac48b06d1894
747f6e61bb7e1d472ce1fcab403799ebd7782f8c
48176 F20101130_AAAFDT bray_s_Page_082.pro
be3d26b9202e62a83975d8073da47049
3f684db7a6dae5f0b2d7dad320267c519bd03394
6511 F20101130_AAAEYO bray_s_Page_083thm.jpg
1982a95f31e0bfeeed0c55707aaf6ea5
7a191834562cd06bb45ce0d7563e3dd95e26f840
1810 F20101130_AAAEXZ bray_s_Page_077.txt
cb1bbeaf5eacb40e281ca4281396b59a
a73f862d1f2fce7abc4eadaffea21f10f1994450
66339 F20101130_AAAFEI bray_s_Page_060.jp2
892f4d4b2c5074ffd5e7a8a726aab1f3
3e3b76c788f2d3d62f66b800fefb801137e74128
70189 F20101130_AAAEZC bray_s_Page_025.jpg
3282ad77c0d248b2810ade58eaec7c4e
5cb2f4c3ef3acffd8495b29e2eea5c301e874e4b
46638 F20101130_AAAFDU bray_s_Page_032.pro
8ff313d7e0a01d328851a787650bc4e6
73143d20471802ffedf0b5d3c249595b4925f241
108808 F20101130_AAAEYP bray_s_Page_074.jp2
76357a0aa037d9049e5d29476b556a85
9d21a78b3df060549479f73a3ca0da65fce13c92
F20101130_AAAFEJ bray_s_Page_061.tif
22afd0c706ed910fc04c29f2ef64cf16
606a1a516f6e227477c3adcb241729923a7f213c
24010 F20101130_AAAEZD bray_s_Page_052.QC.jpg
4f60f4409c1369b5c96cbef2d579b55b
9fcae66e2b2dac8dd91e752944c17bf9c0f8d78b
51356 F20101130_AAAFDV bray_s_Page_026.pro
c593144d61aa965374e5f4370629877d
70698632b4535ba080a8434f411776c94e06cf2a
2036 F20101130_AAAEYQ bray_s_Page_053.txt
d1febc8e799f02a9900111abfd3f6fba
b5cea308acbd6cafa85eb4cfba3e8caaa0f7a43c
6558 F20101130_AAAFEK bray_s_Page_075thm.jpg
19596e22f29aa46e4fdabd19b6482f70
3fcd35f987984882b46aba56603e5ba644378954
69348 F20101130_AAAEZE bray_s_Page_089.jpg
2a475e940dcf474f3d5564b2e4764f44
3f69a4aea2b01814329033c17917e83d385ce38f
12230 F20101130_AAAFDW bray_s_Page_105.QC.jpg
a704834f709152fa28f1e560461a6e0a
1505d4f4863858da045b4e01efa7b94cdc739138
68690 F20101130_AAAEYR bray_s_Page_036.jpg
11aabdf6fee7ff79de24ba3a2f88256e
3157ee5fe8b0f75e52145a138acdedb4b9bcb2f3
109768 F20101130_AAAFFA bray_s_Page_033.jp2
ffd2c020a2ee9218fb87cb4b66af020b
144d5656ab5328d1e07d13e2e1ae2005ba1ceade
6458 F20101130_AAAFEL bray_s_Page_013thm.jpg
2eade8ce5980884b28a70c2bdada1529
d8a8a643442986a521a5e54c5b7bd1b440f64b46
5700 F20101130_AAAEZF bray_s_Page_006thm.jpg
1723b7fca1f667d72a6ac25623564fbf
308bec384226591c1e21e860239f65b87fd9ed63
113156 F20101130_AAAFDX bray_s_Page_055.jp2
c21370bbe2f403bd14a877219aa6afb1
5a7212ebb3b2458370fdf5def341065b04dead81
6182 F20101130_AAAEYS bray_s_Page_003thm.jpg
0189990143db0fbc24a9b825c8c8f31e
1d5e5b04ce1858d0fbaa11e14068ff6d55e8dc0c
77096 F20101130_AAAFEM bray_s_Page_008.jpg
58f89877e184468add7b6fdfdb192845
0c12624f2dc34aa35df9a152f45244ac53f7b581
F20101130_AAAEZG bray_s_Page_073.tif
fbdd4b0260fd4c698365a83759b2a863
6843864f1178656fbcea684ad7bb747aad1956a9
5420 F20101130_AAAFDY bray_s_Page_010thm.jpg
5258ae2ebf6abe880dbf9adb28b07771
743266721ff2b443038849fab37cc28c562b3315
72192 F20101130_AAAEYT bray_s_Page_075.jpg
79daa733ff42b43766aea78b7db98fce
72e84a5ae86d3d96587a0c963e47ec16a8db8f84
F20101130_AAAFFB bray_s_Page_067.tif
5c5d79b42263599524643d3d4fc128d1
d671563757506f3c15137dee6f163a566ca75263
101485 F20101130_AAAFEN bray_s_Page_050.jp2
f3e86d44c032f1c6bc0b40cb803d9124
acd5a9070d59df302d32601f492489299f7559b9
45551 F20101130_AAAEZH bray_s_Page_078.pro
dada7811089e26f7e9064faef19d1f56
2370684678be1b65561548ba7eab81a74120b198
101603 F20101130_AAAFDZ bray_s_Page_015.jp2
23260dd8c4db9f94a329192f4c6c3649
886cf3cf6cbde16693c847648e4d3c5a3941eee5
4049 F20101130_AAAEYU bray_s_Page_019.pro
b58d5159474224da1e34e58f43618bf2
0665282775243196fd6b00594c10b405cc3ae104
112110 F20101130_AAAFFC bray_s_Page_026.jp2
93f136568d30c91a7cb18f1b747d28bf
1c59122eeff5c43231ead91858ce91a81883b33d
92793 F20101130_AAAFEO bray_s_Page_121.jpg
6f19e328c5a5855bdb7973bd27304e3f
c7973bfbbf6f2f60108ca31f4c3527eb4fe3d48a
2432 F20101130_AAAEZI bray_s_Page_091thm.jpg
23014adc13ff57cc5a32a1cd4306185a
4da0fcaa79d19164e762880f2fb5056c50bb5e73
27142 F20101130_AAAFFD bray_s_Page_063.jpg
69ac359fa016c632b76fea1a374f59bc
3a61efa3701a401fd97d7fba5888117fc03984b8
19151 F20101130_AAAFEP bray_s_Page_102.QC.jpg
38e71a3ed1f3fa9db50b9815ce798eb1
0e34a4525b9c68909ffe16c094371633789930cf
69655 F20101130_AAAEZJ bray_s_Page_072.jpg
4bf85576ae7f828d36919212f9517b25
4bf8e176cb93b1bb45455c105e35f979622e679d
F20101130_AAAEYV bray_s_Page_085.tif
1dfdaa4760f7994bb0e965d1b045d5cf
4c220d8e1b31f322d041adca8e9cc54922426431
101357 F20101130_AAAFFE bray_s_Page_048.jp2
5a007037dcc087a24706162cf2a79688
1ca3b393b96270f4821b43684f2b7a206b06790e
100869 F20101130_AAAFEQ bray_s_Page_032.jp2
2ef27e221136cb6c9757b325a339bd5e
376379ef563b0217dc2ceeb5bef29cecc57d486f
72664 F20101130_AAAEZK bray_s_Page_095.jp2
3b18a2d768b40ba2cbcfeb2c19e2e460
dc72d398f894b4a50da6eab8660ea841ee271dff
1874 F20101130_AAAEYW bray_s_Page_031.txt
113d8783eca925b3894f1271bd6493d1
47a6b2cb093bc0215b2e928cee0513748fde4a10
106790 F20101130_AAAFFF bray_s_Page_083.jp2
222c9ca64280fc17048809f45d27732d
fca4ba5b8e2bacd2749d6643e5602985e93455ef
22147 F20101130_AAAFER bray_s_Page_003.QC.jpg
3ef4a93afd3f4e0b59b603b7b8360ab7
9d1dc518a806db4413400b2aa176e1380773aa52
48156 F20101130_AAAEZL bray_s_Page_072.pro
cba4c16949031d66342fe0845572523a
13fc8f7fc912dca4b0036b2710bb01826fe72aca
1928 F20101130_AAAEYX bray_s_Page_049.txt
54d1439e593a8606cfd07c18eebe99ed
5aa0c2c52386f272ea3a2ff436a8c5b159c3708f
F20101130_AAAFFG bray_s_Page_005.tif
ccc93cacde6bbde9672a13184dbc535d
c1b100b8448c0edfc66bad21a3e2f1620c62481b
40666 F20101130_AAAFES bray_s_Page_039.jpg
b2fe644904f6e88db874991a851c4de3
89c476861df996859ee80af869be7353523ae919
70836 F20101130_AAAEZM bray_s_Page_046.jpg
3060d06fdac865b74f229c12fa8c0376
cee9e44205cdd2b6241f0c3216eac3c614cc7581
1580 F20101130_AAAEYY bray_s_Page_099.txt
7a2069e38b5b0ee87d0ad5d725f53a54
8e6b99da614be643ceda9eff164584f874aa8a26
F20101130_AAAFFH bray_s_Page_106.tif
89592665981646cbdfca8cf3517357fb
84cb6f829d8e0d0e8033a53b687364b13bfdca20
52402 F20101130_AAAFET bray_s_Page_087.pro
1921e6470e333aa181ddc96d3ba0f04a
726c2a824ddf219631fdf99c581d6092256045fb
1893 F20101130_AAAEZN bray_s_Page_050.txt
4edd5cdb12e42d8e286c7837ced59ae2
26a59fad332662a422445536dcde91706021a677
22106 F20101130_AAAEYZ bray_s_Page_036.QC.jpg
70d55389ee57675fb950d31d0ecd3a8a
0116734352da35d922c7057b2d8f7f38ec452520
9082 F20101130_AAAFFI bray_s_Page_001.pro
ee0af7b0b144fbfb7437c4ea51504042
87194d9a073eb2b18666cf994906c7f6a120aa67
24045 F20101130_AAAFEU bray_s_Page_026.QC.jpg
ecb1362fe35c0dd4b1ddaf5899ca50a8
41f7b4bc91c92757cf3cad55105fe014ee22953f
F20101130_AAAEZO bray_s_Page_029.tif
c0b429f8ca6d547b3ec5809c71878429
17eeb81ed6a68ca624bc9a4a8315259625d487a3
72994 F20101130_AAAFFJ bray_s_Page_086.jpg
3c65fd4113065afb011cd694a9e94a88
9e5d8698aa28a0d120928c9bfdee5f51f752acda
50598 F20101130_AAAFEV bray_s_Page_124.jp2
91c8be6bbd7d1f4b83248dd5f7ad8eac
30f42a6c555b94c7d621cbc3cceaa2b4f44603e1
105644 F20101130_AAAEZP bray_s_Page_030.jp2
84995af1de0d3199b2cc4a05b9eeedbf
98f04a4c3ec5b59facc01233809d9d61213f5a6e
82175 F20101130_AAAFFK bray_s_Page_122.jpg
6fe0ba32a3ac9b0877b4f995e2a5ed4e
67a84bfad327491c4f71b990ed206e33ba6b4fb4
4712 F20101130_AAAFEW bray_s_Page_097thm.jpg
e5f8cf9465b9337cde5cb65063394119
52650f6321721417511d996f27e66dcbd989cf38
F20101130_AAAEZQ bray_s_Page_044thm.jpg
91b59a3f68c4edb93ad0a6d1684525fc
fe63bbc89a0e77111c4c47490c4fa704e73369e3
103520 F20101130_AAAFFL bray_s_Page_082.jp2
f05f59b50334bc00c225e917ae53730e
ed414934e99f8300c511e7c0f2508df8f2aa908d
67575 F20101130_AAAFEX bray_s_Page_032.jpg
5b6cfdd10f14eec3536d9272e9390d97
a36ef34ec8656d5d9e0cc954c502127935c437f6
F20101130_AAAEZR bray_s_Page_113.tif
00b2b3fed3101dcb91dda2bddb3ba075
fa8f53fbe0278534b62e77d126a34bbfb6583ed6
22999 F20101130_AAAFGA bray_s_Page_091.jpg
33dd835935d082641a09b285bfb22485
953f955a9545125645af76cb5ce7d41724744d6a
29796 F20101130_AAAFFM bray_s_Page_062.jpg
7018c4a3261175e628302953dfbb3c2e
350e69d0bb9c2cf52a2ba0ac9b9ccaad5e2308c3
134155 F20101130_AAAFEY bray_s_Page_120.jp2
5cc5e452cdb27c92fab10be5c4760934
2712be93e464ca9b6e0004fd68ce746eaf6049a1
F20101130_AAAEZS bray_s_Page_101.tif
efeab171698a7c609840548207578073
fc5dda5cddf53cf171ba0487bf22310ea537d345
6087 F20101130_AAAFGB bray_s_Page_070thm.jpg
64d96b3c8b630cb2a7049bc00d2b1bfd
e8ec7157fb3cd687f2c4973334460592ba67bccc
12236 F20101130_AAAFFN bray_s_Page_092.QC.jpg
4d28365cb5ba3c338d9666b6640f2104
bf5fe028590685778d85ee5fe0f923c846edd751
23769 F20101130_AAAFEZ bray_s_Page_016.QC.jpg
4e02338cd3a663569f3737f6ec78453a
107032d337f6d944fa68878f6f2272d494d33197
50274 F20101130_AAAEZT bray_s_Page_029.pro
c7140e376e4fae247e5a989693d0a0e6
87d09529a98101648d095cddb7e2583b875908a8
70931 F20101130_AAAFFO bray_s_Page_045.jpg
bdff02b3a6c982af9718ff3c1297ba0e
5f8ccf950dfd4c77a2079320d6cbb03ee29c3593
F20101130_AAAEZU bray_s_Page_068.tif
736501fa3f53ec2d875128a97234dc58
a5f4d2e69fddc35025a27595d3e97a35215a9c67
16228 F20101130_AAAFGC bray_s_Page_098.QC.jpg
46676e59ffe92dde54d1597ee6d87d0c
fda48a46aeaa5d41b0687bcc275596b96ac95b50
2902 F20101130_AAAFFP bray_s_Page_005.txt
39603efa3a66dc346a1eae9f4ea08f21
9b4f892e58de40af70d3dd6e71a19fb3711ec581
1834 F20101130_AAAEZV bray_s_Page_011.txt
177fe3643c1ed4fc901ff5ca9e5db247
9a808f465356ddaf6138301dd760f748f1b9e073
F20101130_AAAFGD bray_s_Page_048.tif
6d98e0da884ff70f88917ff34831bcd9
0a8abf2f45b5d96985e1fdb21a8ed06f015e259a
1447 F20101130_AAAFFQ bray_s_Page_060.txt
d6d5c4e324f546f268b1546f3eba15bd
eb122546b130847f4352ccf4b3023ca06e69d9fd
17659 F20101130_AAAFGE bray_s_Page_103.QC.jpg
9a46c9d54347916b0d86a60b1d4bad33
1573af1d386254309656e3b4b3ad5ca53d92ad3b
50960 F20101130_AAAFFR bray_s_Page_012.pro
2f8e303f3e145647f13c8477a8287912
d7ec0317edec2ce2218c7bdc87ee6736661a0a13
1403 F20101130_AAAEZW bray_s_Page_002thm.jpg
13b779c833a4435984fc49461d04ae2d
7a6ea014fffc63a4f2b151f7d73ca70d735d084c
33795 F20101130_AAAFGF bray_s_Page_065.jpg
0d860deda3824343033117afa36f33f9
de2a738d3447fd7f661483d2d5a6fd357cf35d55
69573 F20101130_AAAFFS bray_s_Page_041.jp2
ee0b24a8c3f7e6af804e50d618fd0313
6ea88bacf3ed3fd779129bb4737372123084d5ff
F20101130_AAAEZX bray_s_Page_012.tif
8e5e63839b23c7710a5a19bc29e28c77
78e7e94d577b679525c336ef2564e6e4029df697
67731 F20101130_AAAFGG bray_s_Page_121.pro
efdcdc4536504f88bb1159b9dad3a77e
1a7fe6996d1338f7382e193ad435ef88d2bbbaad
1921 F20101130_AAAFFT bray_s_Page_034.txt
1e6960a8a8dd4601261a8e717e105986
974ee3d3a74be923f647e162816dd3aecdf652d2
21641 F20101130_AAAEZY bray_s_Page_006.QC.jpg
a73d8de81b0ccd9d13b2ec5294240b19
7a86a1f7408db785db2a814e6ede4887049e25f9
17622 F20101130_AAAFGH bray_s_Page_040.QC.jpg
14566be7d423fc2a68dca7604caa77d8
612d30eaf9738077b389b28d12df0db521ce8f24
15389 F20101130_AAAFFU bray_s_Page_064.QC.jpg
9c85d6a4a9110e08a5aca3353f1c83d0
dde6110dca04efa6f2364417fec9d6def083eae2
24757 F20101130_AAAEZZ bray_s_Page_055.QC.jpg
240efde83a6a1fe1aa2eeaad1de58cf7
52903801069c806bb34678cce411d897a4492521
2079 F20101130_AAAFGI bray_s_Page_085.txt
b5d396866df256f93767c0f2c4a31625
8af5098e343373456b1e944fc8e78548948a37ce
24066 F20101130_AAAFFV bray_s_Page_024.QC.jpg
3796bfc4802feec7d18598c9330f794c
e625ee7575d1f001c4178e59b2b44f99999c2d37
5840 F20101130_AAAFGJ bray_s_Page_020thm.jpg
167c8b828e6321a24c8606b6d48230c2
1197949c1a4d0be0873f73f11cdf762c2e4ef97b
F20101130_AAAFFW bray_s_Page_096.tif
f8e60c89741ddb37c630fcf5c699e3f8
138892624d10eec088d5a53e87e6a5046d4811fa
72885 F20101130_AAAFGK bray_s_Page_047.jpg
5e3074ec86d7d4a7f038b6ffec8b3428
32d2b59b56496aef49dd1e38e92fc6c629ca4fa4
51692 F20101130_AAAFFX bray_s_Page_014.pro
73c6a48cf0605d8744919fb2c152834e
f3770ecf0a2961cbb69dcc4165cd8b530f11a4b6
F20101130_AAAFHA bray_s_Page_102.tif
a09841e907f5013886beeab4858a4c9d
744e485d180a565d17d85be7546c9ce70f2ada67
6541 F20101130_AAAFGL bray_s_Page_052thm.jpg
7e399010cb465b2081e0bfe6faaf1ed9
e661ed98882d0e669d83da4e05afa2aa4209986e
5397 F20101130_AAAFFY bray_s_Page_094thm.jpg
3084c659aee6f692f9ab523865ffea12
a82e85fd224a46d2a5ba9495ebafcd726c3c591a
6759 F20101130_AAAFHB bray_s_Page_055thm.jpg
2682582089008a8556b780a46fb08a73
d1533c871d818e350d109647bf8b707d0c1aacac
6768 F20101130_AAAFGM bray_s_Page_029thm.jpg
1f39362ef27b915271e7dfba1852587d
753b20b8508256422067fa5fbdaa34ca44f2b831
23146 F20101130_AAAFFZ bray_s_Page_067.pro
dc39713a3e95949adfecc6497efe3d57
e8eabeef7b041e55e376dc62687fff4c7f036967
24207 F20101130_AAAFHC bray_s_Page_047.QC.jpg
9e09305d4764169cec5915b94e843c35
5f5a276c1c1bc3d1b2ceb7983b6fa049548396fc
72647 F20101130_AAAFGN bray_s_Page_028.jpg
252927ff0394ae9ca9f363ff91817392
80539932804051e049cae8674171c9f415b8bf3c
5804 F20101130_AAAFGO bray_s_Page_059thm.jpg
d8dc7168d5c28e46505b5de5c8705cf2
ecb0fb4332c6f7ee7c717816f38d4ab11c1e4f4d
39822 F20101130_AAAFHD bray_s_Page_058.pro
094d9f419b79de8ab1915eaf0bcf61fa
1169869e140387603b89860fea329a02db58feb0
72094 F20101130_AAAFGP bray_s_Page_027.jpg
2db5df27a6c53425a12f71b4159e4672
085583fef5b5a0d012ce90c56ba179eb8cab1d13
1299 F20101130_AAAFHE bray_s_Page_041.txt
b698fd28672b2cc6b7ec899a38fe5ab3
7848564300a4535dd9b8cca1a1b2797ad281b6b1
22777 F20101130_AAAFGQ bray_s_Page_089.QC.jpg
54961388a27ef6c38712e74b6bdd0b81
fa4360b923db4b02f1a1b9ab417f2009141cf974
50854 F20101130_AAAFHF bray_s_Page_053.pro
6c91e4bba8efd753ad97ded5fa485005
2bff48ee1b2f3764a38eb94ad23a2353104d0a1a
105702 F20101130_AAAFGR bray_s_Page_109.jp2
763af4f0ac146e9b76acb52aebd72893
b028ff62ca64ee7420193631bb5a899c06b2b1f0
6128 F20101130_AAAFHG bray_s_Page_021thm.jpg
898ebd2556df11ad20ca4d3b3bae1f8b
3db26c6af2847ac3785b67ba78f908157f4f1431
1906 F20101130_AAAFGS bray_s_Page_017.txt
a2dd44ded4818e81aba906f15b8d3c54
b9ae20b66a4def84b3a11786c36b289bf470e9a9
472 F20101130_AAAFHH bray_s_Page_063.txt
d239c57b4780aff57977f3551fb045f2
5ce31d708384c50364bdf72cbd49f3f63ea84e9d
65166 F20101130_AAAFGT bray_s_Page_114.pro
028af2685ab17d6904671fe44f06ebc0
8106bb5e9e5432b82a7141f430b77bcc10304259
23336 F20101130_AAAFHI bray_s_Page_076.QC.jpg
bd6cae31eae8d7180cb063d2364f77df
057f89bafbdb3cdee6b184676a9b79c3ac7d7592
73131 F20101130_AAAFGU bray_s_Page_052.jpg
4eb8bcb86d049450bfe70a3427065c30
72efa321701db59437927c5150170c044b60b036
110297 F20101130_AAAFHJ bray_s_Page_045.jp2
b0c21987e59eca124bbf53c20664fc62
53f4b300304d89bcc349b5df79b41a0a24ec850b
107312 F20101130_AAAFGV bray_s_Page_113.jp2
44649dd097c1a12f67e6ddd2b5ff13df
13e4120315895a86f0ee7216b0cb3e3e934fe97f
6610 F20101130_AAAFHK bray_s_Page_028thm.jpg
05ddf400d722e370218b25a05459c76c
82a89068eb68d3a98c0aac07057830b0f8c6c2cd
2366 F20101130_AAAFGW bray_s_Page_008.txt
f77dff4940ad4ecee1e47a2f404dd610
dd6f502cf3b9f2b5ebcd1936362434b448e68398
1564 F20101130_AAAFIA bray_s_Page_010.txt
d91d5033f4d80ebb37575ba8d84e9e8a
7d19091f46ccd5706d54897e5565ade261c5e557
52443 F20101130_AAAFHL bray_s_Page_095.jpg
867db30b90ba3c6e5dd3de1312ff9cb0
bbda9ff464b466bd0a501e19c05ca93daf60ae71
F20101130_AAAFGX bray_s_Page_041thm.jpg
b8fbc44609b06192ea18d196245d5c03
b8746a04c7b46b383af6ffb645a3feb295cea23e
106396 F20101130_AAAFIB bray_s_Page_034.jp2
93c40ec30d00dd629f3b7777aa813e82
6ed8e20fb217fc1ae4431608eef8189f22e5a52b
7059 F20101130_AAAFHM bray_s_Page_117thm.jpg
614e43b78da97cb32222038cddfb0e2a
fbbd2c7bda01a65bb6c38ac5bc4901aa3058d29f
3604 F20101130_AAAFGY bray_s_Page_039thm.jpg
bce2a7a4bf68359bfd94f285e058718f
3f4b311bd68f18522084a598fe3a2a35e725e888
22213 F20101130_AAAFHN bray_s_Page_072.QC.jpg
ebb9ab5c59aea1893d754566677b8e7b
9143cd77fef633b67f3f5058d9d52bd0698a8322
3138 F20101130_AAAFGZ bray_s_Page_065thm.jpg
d48f9af8ef8a9b64ab5133864cb581c7
9b46479ec90f43524c9b6b60bf2314aa08d94df6
17809 F20101130_AAAFIC bray_s_Page_041.QC.jpg
70afa3a5fcf7eb13423b0fe8e5c339f4
f36d81c3a5a5fd380ff7ac3db74576207e490b20
14312 F20101130_AAAFHO bray_s_Page_100.QC.jpg
48a137c4c3c03db3bbbb1f3c4d2bbc8b
eb75882daeb74ac113fb17049d53d4b97bd4c544
2094 F20101130_AAAFID bray_s_Page_055.txt
ab96c450d253e6cfb79cd3792da6a91c
a126b0aaf1af092502325e2d484afe55fba58edb
6578 F20101130_AAAFHP bray_s_Page_053thm.jpg
ad2e24ba6dd49f92d92d273c1eac2d7f
0ef8aa76d5a079055518fc736d20aab4746114c4
46537 F20101130_AAAFHQ bray_s_Page_064.jpg
e2fc2125e1dd9616ad647bf1ecd8b1cb
1650221fd95809314999e1cd3690b5df334b3a00
F20101130_AAAFIE bray_s_Page_045.tif
ab01fbd12977f242d47ecc8f2afea4b3
28f55ca2baee7e07a712893e022d170fb575e4b0
73293 F20101130_AAAFHR bray_s_Page_014.jpg
4db2f763a9855f33e38dfd0769379042
0164f30cb4b53e28d20fd91aa6571f42e5320e58
69594 F20101130_AAAFIF bray_s_Page_090.jpg
cc0402679082a8fb3ca44dbfa9ba6078
51bb99a6ef915ef8e91a9fa51fa0c93cc3a6696b
F20101130_AAAFHS bray_s_Page_115.tif
ae90ca6e8ca35e3dae8936790c4b78c7
31aa4d918a38625ecdc8df0b126b1952788b2838
97927 F20101130_AAAFIG bray_s_Page_042.jp2
b13ffb7f7ff854004819cb0319a4bf41
eea8f5fcc322be388b5b3aabef5b69a882dc2fa6
3076 F20101130_AAAFHT bray_s_Page_093thm.jpg
8491550801b2e09c0daa59864e869f73
c7dda7c6a6b856e070c22ea440c7eb4ecf2f1316
2005 F20101130_AAAFIH bray_s_Page_012.txt
8e43ea015f160d45ecbd91ce0c9832c9
1a2fa58ce7a73ecd558e1a71336b5a959a95952b
933280 F20101130_AAAFHU bray_s_Page_103.jp2
0c6b8bad2dafd2dff3d57e77316283bd
079a14d9d92828318240fc5286d5342d71ca4a00
71537 F20101130_AAAFII bray_s_Page_022.jpg
2fd96bfe13850a7ee487d9568083c249
dcec3d05d3f66179df7d81769c75746239a84386
F20101130_AAAFHV bray_s_Page_011.tif
24df682396449751d957861ad0adde9c
89a939c078256f2b33f8dc570f66a79239fc0585
2034 F20101130_AAAFIJ bray_s_Page_014.txt
b23e28db7401e2ef989fc58490d53917
f5a0cee9bf4c823fe57c74a71d5a60be376699d2
21777 F20101130_AAAFHW bray_s_Page_080.QC.jpg
939b104915727c0e3c0a2181dc8c34fd
34c7f29dd0eebfa933ba04b14166dccebfbbc568
2090 F20101130_AAAFIK bray_s_Page_113.txt
550b132fc81b31731ed244bd6827cb68
f19b3b5c72e1e6eaf3204132789d525d8dcf0bb7
89540 F20101130_AAAFHX bray_s_Page_117.jpg
291644bc738903111a276a4c5d1bb4fa
40bc96e84af052919395246ace1ddd0b58de3123
F20101130_AAAFJA bray_s_Page_111.tif
4fc067e1b5ee9b99f7ac1db13b033457
da110403683f4a98f20546810b47901f834984b2
24855 F20101130_AAAFIL bray_s_Page_118.QC.jpg
4ded7f567f32a5bfbe4b14a3ae7a050d
d151ffed42bec8fb70018690937d927673e4b793
6658 F20101130_AAAFHY bray_s_Page_023thm.jpg
786da3e5c8be46fffa726438d724219b
00e03ea25c93b4639dc17f70a1560ee411f71a4d
F20101130_AAAFJB bray_s_Page_105.tif
2ea3426aee44648074b1a56a4ef1fbe6
fe5c84fecefb2965b5eb9c0f04d64a8f6a319afa
6023 F20101130_AAAFIM bray_s_Page_048thm.jpg
c7e9042a2fd38b93e09f6ddaa0e697e7
2c443d3064a646a6ba9b6068171f078cb67c3013
1978 F20101130_AAAFHZ bray_s_Page_110.txt
1df87ab2f1574ce26737aaa293354db5
37190926854ae708ed5837834effad7e812ed7db
F20101130_AAAFJC bray_s_Page_120.tif
97d0d6810dc66aef1e4888916bf9e927
bf625bf41440adf94f77bceba3c9a2b3b431db54
90215 F20101130_AAAFIN bray_s_Page_058.jp2
ec1c494f479ae6cff45370a633211b84
73e82f1daa565f0b034c3ba16e857b873d1ea3dc
F20101130_AAAFJD bray_s_Page_082.tif
114350db4abd3d5ac2587bf0a9e675a6
e4737e5129fe8d9b5abbefe17e91e825550d9cca
84860 F20101130_AAAFIO bray_s_Page_123.jpg
df854e014909fb709369a22de84d5710
05b75d3f8f6f7a032c98c1a2ce93462cc16f8524
F20101130_AAAFJE bray_s_Page_028.tif
e8ddec6a8a53023f32ee40521b940521
d70947f3f91490b3290cad7847f9737ec5b1665a
38116 F20101130_AAAFIP bray_s_Page_097.pro
b1b82859654c8ac517ccf81ea02e1ae1
d00afde484adf1bbd3db692932ac724bcd9d8b4e
51042 F20101130_AAAFIQ bray_s_Page_084.pro
7597cef83664ec911b9a305027fc5322
9c5c26fb6fb87522041248a316dbc95ab104976d
38636 F20101130_AAAFJF bray_s_Page_124.jpg
f7211027254064a26a0d098b046cdbc2
76aa6d58a49f616f99f038733971756226eb7123
5963 F20101130_AAAFIR bray_s_Page_113thm.jpg
8054227e5799871e53c6914ed4f6f978
78a0167c2b0b40a0bdd69067a9994b0f9ed8066a
72420 F20101130_AAAFJG bray_s_Page_038.jpg
1121be3c77dca6ed9eb394e8638d0db0
162e98505f6264e797a43084845b04aa216e2b66
69206 F20101130_AAAFIS bray_s_Page_109.jpg
a24cdbedfde3992139417729808b983d
c816a29b08f670d559f99b59ed05c33378801515
104624 F20101130_AAAFJH bray_s_Page_007.jpg
a952460947751fd4a1bec07b85527e6d
9838841239a54ff42c9ad48fe26aa0e6f2b2e4a3
1809 F20101130_AAAFIT bray_s_Page_070.txt
9c48164e41505ee688f87080d869fe62
4f359a2e35a40fd43b9bc0be3d3bc6c5e22b918b
50438 F20101130_AAAFJI bray_s_Page_073.pro
65a149b321979ad1347c37eaeb6ee9bb
328c8c5320e58450eb1a90f1e20f824a5372bc6d
49786 F20101130_AAAFIU bray_s_Page_074.pro
5f825461d146dc5dd0074f0ee7d24082
8e43ae08c9255bb218fb228ee6b4f58b2e8c9172
F20101130_AAAFJJ bray_s_Page_019.tif
0848b6653668894171afdece72de7486
bd9584cdcb5e58fffba9dc7c50785eeb4562b66e
F20101130_AAAFIV bray_s_Page_023.tif
b03d3272503d7a63e85b1b38c264fcf3
8fd759685205c6623a55ccef1ad4cd61ca009982
F20101130_AAAFJK bray_s_Page_118.tif
a1aa0a39ef42c3d6164a929d84bffff5
16d341acf4582e0a3fd14738dbd17e023c84c831
6527 F20101130_AAAFIW bray_s_Page_027thm.jpg
e58d376266265bf74bc1fd9eb9a65acb
e6a586fb3f936ae57ac088ec01e5b6b48e366f8e
21768 F20101130_AAAFKA bray_s_Page_032.QC.jpg
18f04740458ae4267a3535178d101daf
a20ccc6adc295cd99ca5333eafc2abd33182a952
57974 F20101130_AAAFJL bray_s_Page_008.pro
290e3d5b282834a86de8fbfc21148127
16ba8fe175400e418957f24ebd8a00379cb5de42
6154 F20101130_AAAFIX bray_s_Page_017thm.jpg
060380a1bb477a83375831333629ef81
cea14de9f713b0b7b1287dd85c5a1f0d377dd8bb
1051959 F20101130_AAAFKB bray_s_Page_007.jp2
fbaa73bcf89d9e9dd5586309c6208bda
c7f609687cb78481d7109e374f4df8b25c5a0cf8
15351 F20101130_AAAFJM bray_s_Page_060.QC.jpg
103af2addfcfec0cdcf3fa212d38c6c5
b6d1b7ed04205ff5de8c856e6f86854cfc46dad0
2245 F20101130_AAAFIY bray_s_Page_096.txt
2f1c77c474f26b5d097b4f898f9f429f
ec166a53c908f5e51aeccd7e179223de89a49bc7
4895 F20101130_AAAFKC bray_s_Page_103thm.jpg
bcd74347d7d1d2d12892fd81a5dd6f1a
f38c5de0d76a5c362e862b16dce615c118576823
470969 F20101130_AAAFJN bray_s_Page_067.jp2
ad3d20cdac33b61ef06fe0d27209f14b
43fbefd2d6ffda32618c2974727cce7b3d2b9a4b
1987 F20101130_AAAFIZ bray_s_Page_045.txt
d1263fb0c2d24d076b45fa2831e82d48
9ae7016e3a8cc84184a2fd1fe84a685fb5e8875a
22710 F20101130_AAAFKD bray_s_Page_049.QC.jpg
bda7a9808cd4e8b8735df86fdf3af5d1
bbb4b02bcb24c0a4b9aac4b69eef9bc49fad3bb3
22542 F20101130_AAAFJO bray_s_Page_082.QC.jpg
d16a252f723a7b5a1606738a5a19d493
cfc6c40ffde29f66141279795f63822fb94ece6f
69116 F20101130_AAAFKE bray_s_Page_082.jpg
043fbae7b5113f49ad5fa4d61bb5e1ca
f5b725c8569d1f45fa684706e5e26f4ae85d52ca
F20101130_AAAFJP bray_s_Page_070.tif
f271b47dba5f8044b95b0fa9caab7fff
de80992194a6064359f75e63bf5b551ffef076fb
2568 F20101130_AAAFKF bray_s_Page_059.txt
a14b09e3815d9fcedae35cdea7257600
1b54dece03be971ccf725cb9c8d8151e51c7cca6
11679 F20101130_AAAFJQ bray_s_Page_057.QC.jpg
e9b3a8009a4682f6043cc2f01e140066
af9adf32a0be4803ddc548bbf27048e7e68c1370
2512 F20101130_AAAFJR bray_s_Page_039.txt
eddc2f05286ab0766da979e857898490
d6f49b7721ec599aed962c94d508095c38c1fec7
98114 F20101130_AAAFKG bray_s_Page_108.jp2
fcec36d7c832a940c24552cdc8fd8b38
6950543fb4cff55b5a02420ff075e751d949e34a
26193 F20101130_AAAFJS bray_s_Page_116.QC.jpg
2006a878ad336f3ea0f903777ba8855e
6da732cb87933a9304df2400111d9265c90c9ff6
16384 F20101130_AAAFKH bray_s_Page_106.pro
674eadc051402b06a32b1fcfd8d22a37
f5c319c86a4a22703bf4fd6cee73015c8fcab005
6684 F20101130_AAAFJT bray_s_Page_085thm.jpg
b0c8a95b980121f0b3947747c59970b3
5b756ce2108f1c973e86d4f895a5b52c81c23063
109520 F20101130_AAAFKI bray_s_Page_053.jp2
43e7ae89381abcae5fc69da4ab22dbed
68bd08871207d1ac4b0a8e39c39582181b47d0af
F20101130_AAAFJU bray_s_Page_103.tif
83ca43a2736575ee385bdadedf6e417d
27ae10ed52f761a8574ab1caba695d588b35939e
67622 F20101130_AAAFKJ bray_s_Page_015.jpg
471b798697fa928cb6f29fa72dbfa562
e3ac4201f96376dc9bb9ca4d5e0c76a53f3ad4f1
74350 F20101130_AAAFJV bray_s_Page_097.jp2
0c5515157cd5c675ff8a38235f77f1dd
9020493c178b23e67d26b391758fa3f9264019e3
6358 F20101130_AAAFKK bray_s_Page_034thm.jpg
7df2fcc1b2d4dff7a3794e95d9eca6b7
665f371d39ddf71c15bdcccdbc22e0ff94579c9f
109601 F20101130_AAAFJW bray_s_Page_029.jp2
42a69c42e3b35cf835d7b9a57eeba431
4896a4bab3fb7cc042243491248e93439a03033e
F20101130_AAAFKL bray_s_Page_038.tif
97ea0f1d6408f8e7feac6069aec83294
d0224d655bb72fb615b7c3cb568c88c7ca53187f
72612 F20101130_AAAFJX bray_s_Page_053.jpg
a0e2276161a3e21576db5e904ab85ab2
8c05ddb113ac851100406708522776df69c628da
4151 F20101130_AAAFLA bray_s_Page_066thm.jpg
75742eaba90e077663d03ea4a8912660
b57bf3c62cca6b364ffb5e36a696744529551f40
22818 F20101130_AAAFKM bray_s_Page_021.QC.jpg
085fc462a94cd165f7c9d644ff4d1d84
ab3691ff346b53e8f5b08c654f6493647f553c97
49749 F20101130_AAAFJY bray_s_Page_046.pro
45a8b80d5c3992fa1eaa3b1e01e1b104
502a2dc58eccfa884afa7b10322cdb314a356a4c
6809 F20101130_AAAFLB bray_s_Page_088thm.jpg
16b952ccd85504744e61c0c9248e866e
b763c69a6d32b4a3156f77ccdab756b54f39c993
28803 F20101130_AAAFKN bray_s_Page_007.QC.jpg
af511fb6119c35b9bbb6de0bf06475e9
7af5512e9f236ac6e7446758edfc33cf207efab2
2881 F20101130_AAAFJZ bray_s_Page_061thm.jpg
5d95ee87e914a5c4e26441da4740b118
e5d08aa7dd00ce4f1d4049eec01034d740788b18
2583 F20101130_AAAFLC bray_s_Page_104thm.jpg
fa3f0f813f70bf727901f39ca0be8713
1fa8dce54ed34f1b5462670a64e56b654b299358
69553 F20101130_AAAFKO bray_s_Page_013.jpg
276bbc9a2e39281872d134dfeb12cb7c
778f93c1326b560259f622c654277c5670356b16
53289 F20101130_AAAFLD bray_s_Page_055.pro
3ecbe9b229c8e4fe47c078370e21f2c1
0db759c0acb39e7aed308c449687042ce1c6549c
3633 F20101130_AAAFKP bray_s_Page_092thm.jpg
62b3ede4414a54854187fb1ca6d25752
8f1538813684c34be26164ce962553cb295ae4c0
107865 F20101130_AAAFLE bray_s_Page_110.jp2
ba8b86ba699abffcdc0d13c1aebc9ab1
4d1a7f2a9027b186ceb4711ffc3bd520f311c531
F20101130_AAAFKQ bray_s_Page_097.tif
d42b29d6e3ad45026c5e07e6564f724f
9a5d60b6585c778c08f32b37e3b272f328b84cee
51009 F20101130_AAAFLF bray_s_Page_111.pro
b28baa4965595787e7453d04d02106fa
9cd33217fd0758f8dad3fde21c6030422260dc9b
F20101130_AAAFKR bray_s_Page_056.tif
ed92b93829da9078f477d30a4e31d934
71ea93fa483f199f34f082dd029ffad299392519
3593 F20101130_AAAFLG bray_s_Page_058thm.jpg
086c58093647038c55ca1efb3144ae13
b2bca93a9394158708fd266a490293e75a553f5c
55579 F20101130_AAAFKS bray_s_Page_097.jpg
b3940567520fd4fb4bc87d2737d2b412
d7a5dd0beea4460ba49d021b62259a9d99196c03
23913 F20101130_AAAFKT bray_s_Page_056.QC.jpg
7910685be59eaaba2a0c1d322c1bc19a
02f4c2b47f89c72fb88adff6725e7bb2e5b07b98
10134 F20101130_AAAFLH bray_s_Page_107.QC.jpg
9661189e493c318da14576efd61a2257
0e3a5805ca077e8b2ac4fe85d2fc8c7253c98f18
2007 F20101130_AAAFKU bray_s_Page_111.txt
9e595bedf1a0ae28ed07ff6bdbbf6b0c
8bedd00dbe48df7b3287c4e874195c81623fade3
6591 F20101130_AAAFLI bray_s_Page_024thm.jpg
7f43ef54e0fc32b8e55258faab87c708
e3f30a9b665df1692d579f8f839685b86f9c6aa3
6407 F20101130_AAAFKV bray_s_Page_111thm.jpg
abf8e44eb91d5094cc9acf34735a12d6
70f03efa06e93e80b3ad311c4fc3759b884c66dc
F20101130_AAAFLJ bray_s_Page_086.tif
e63c74e2ae9ed82eaa3d80068c18b78e
a321b9e0a8b969e2336e5dee897484df464ec7a2
23112 F20101130_AAAFKW bray_s_Page_124.pro
9d7b321fbb2f460da6235ad341109b41
b5cb79b08b9b40bbb4da5337dd9f4cb6cb531d12
39803 F20101130_AAAFLK bray_s_Page_103.pro
72d0b2507e81933ff64c79e68d9aa22c
bb8578c86ca7125d71d89d56ff6311a7b4b78618
357258 F20101130_AAAFKX bray_s_Page_106.jp2
42dd8c2527b8f4295b54f218b6e316fc
a004d41b891dd63586c52462c21ff1a82023b9a6
6557 F20101130_AAAFMA bray_s_Page_014thm.jpg
e0950ece37b37c3a8a7e10b91960c6d6
4bac431b7f5f7351ff376014b5ab52f50a73af77
2949 F20101130_AAAFLL bray_s_Page_063thm.jpg
19fe9cb28d1df57775a92e838b501812
1428f36bbd5dec04e63b55e0e7f5eb602c39fd12
70599 F20101130_AAAFKY bray_s_Page_083.jpg
813ebe9ceda9fa9661a57c6a91b1ab41
1012bca5525e540a3a0b76d03dd53782e45bb4ab
1946 F20101130_AAAFMB bray_s_Page_033.txt
b4bda5510c2e2e98a7b1151d2e912f0a
267fa2254c68fb9d73ee1fc6159dd8de25237616
5603 F20101130_AAAFLM bray_s_Page_040thm.jpg
79099d021361a23f6e392cec822ef9ab
99b375ecd4e9bef6dfaa1c91dd547885966300fb
27146 F20101130_AAAFKZ bray_s_Page_037.jpg
491146020a249d558ad2224c9b027e36
ef1bc4017be20c6426d7539d907e92ea8271d64a
102795 F20101130_AAAFMC bray_s_Page_017.jp2
9b7f1b32ad9b284e2fdc23ab3034c13f
394ebfc002bce5d3e242e5cc09ca93edb8fc7422
535512 F20101130_AAAFLN bray_s_Page_064.jp2
5364dd85756e1ac27bbbf73e0e3e6d92
1c079138509c2ca395ef57fcf15dd4ccea6d9436
374 F20101130_AAAFMD bray_s_Page_104.txt
c2b7d8f4bad1196ccad752f616a37dd8
6575220d9c208dd84ba60cf3faf4a224fe1572c1
19887 F20101130_AAAFLO bray_s_Page_004.QC.jpg
dac855320c59b729f96b6877c4db619b
b2e35b5d4a4d7a8064effa16e6c8f8e35e78ca95
2017 F20101130_AAAFME bray_s_Page_029.txt
db85e2bd25c804ffea23a9eb85de1880
c775fc840caf6cde2d729c73aff36059a59ace42
45765 F20101130_AAAFLP bray_s_Page_077.pro
8a30e7278389fc9a896834fddbbd0422
cd4ca78553d6c136b500b230edfd04f0bae85f27
92689 F20101130_AAAFMF bray_s_Page_096.jp2
2c82cc309e7a0b237d1c02a78aa62e14
2cd9608bd980339c1a7c10d0cca18963692b159d
43854 F20101130_AAAFLQ bray_s_Page_070.pro
e0d7d29ee6f34ea0d7b502520c84121f
d62f234a82452a3170a624ff2a43bbe61c8c8f0a
7027 F20101130_AAAFMG bray_s_Page_121thm.jpg
d2e78f8715564c96fdf458b1103da916
3834cdb8a112dae8edc08725c8132c34ad419f31
7945 F20101130_AAAFLR bray_s_Page_001.QC.jpg
2cf8daa1b3c49dd08529d657545bd796
23a34c9d5ad3707ef54461c8cb40b16bef735655
F20101130_AAAFMH bray_s_Page_046.tif
dfcae92136a41bdfa57ceec28aa4d276
684a21f331050ce5eb1def4dce2c12969369e643
F20101130_AAAFLS bray_s_Page_051.tif
500d4e926fdd021be44177e191769535
cd70daf4a90e783b39a3861c9b415d5e5eb777dd
F20101130_AAAFLT bray_s_Page_027.tif
76d10799428d2af897b48c2de0b43080
13f7d300c1a3fdef83b2f1bc5817d91df223b01c
109800 F20101130_AAAFMI bray_s_Page_052.jp2
92d2c37bfaa70a37264f6fd06f960f89
e469d56039f51a53a16549ad651be3ca795edcbf
F20101130_AAAFLU bray_s_Page_058.tif
a2d54e4d1ec89fc3863c25479d5dc4fa
2f1283c0df546ec25bef45034c55f472da4529f0
F20101130_AAAFMJ bray_s_Page_114.tif
16f58268c85d0ba9bc4f2c4b8cfc57d0
3033aab2315f6e6e267cbb0d3f28a498c2000d3b
23226 F20101130_AAAFLV bray_s_Page_111.QC.jpg
153f00a503414c9ba0bf82f638ced98d
27dfe6f0fae95cb965c84c7867b3e0bc0989a7f3
F20101130_AAAFMK bray_s_Page_043thm.jpg
6030d35da53c34b33c930904aef94ffe
d522033aedc81194c0542c1a01a4c49579125d3d
6841 F20101130_AAAFLW bray_s_Page_118thm.jpg
6c36bfef2e9babc0929489dfc9e04bb4
df6be04c4b4b9d05571d07dc2c1a685ae64c8c6c
F20101130_AAAFML bray_s_Page_036.tif
ce75217c1c08c94459dab76a0bb5c742
70f3620d574d85f2e60d607fa7e3100e51469eb2
F20101130_AAAFLX bray_s_Page_078.tif
fda1088164e6b917c5b2d917057e0437
ece89c0d6816d0f9ed8677edbb0477303deb13e8
25301 F20101130_AAAFNA bray_s_Page_001.jpg
c06bd4b9b8231a8fafd618fd5b181d1c
d4d9a15607942ec39587f661297238dbf137701a
22977 F20101130_AAAFMM bray_s_Page_071.QC.jpg
d430b09d465481a207319dbaab3b3c79
41fa0f10719614c7737b634e0ea15e5b96ebff8c
28781 F20101130_AAAFLY bray_s_Page_091.jp2
df8c39727fb31e330db765974fdd9b28
90f2a771275fcf74d1c016a64bf6825e29ac3a10
10168 F20101130_AAAFNB bray_s_Page_002.jpg
a414bb756df4bc4dce9f826289b5b5e2
4371cffd16d05b9dc5759ea415ba785621691419
23289 F20101130_AAAFMN bray_s_Page_110.QC.jpg
e8de2516899cfbf503089271b26c4f71
b7f245c48451cc2257f7d76be55ba5fcb5d0a771
1051912 F20101130_AAAFLZ bray_s_Page_006.jp2
727371b02b1c4a538895c2b5e4c2a987
a57fdae3bb257bb9ee843b2eaed1efaeb7b1fe3b
67476 F20101130_AAAFNC bray_s_Page_003.jpg
1b278d7855d5c0c900ac41eada390bee
25373c829609ea8da8a053fdd2b25bd01a05b165
46447 F20101130_AAAFMO bray_s_Page_080.pro
722ecb449eec92038226b11ff89e8ebf
aa1d40c1a722304b8dcfe89d237bba147a416334
73841 F20101130_AAAFND bray_s_Page_004.jpg
e86a9b2aba0fd8d41c0d377ecedec8b7
70ddbb9dc0d26d3bda4b2988c046136c31349b92
50168 F20101130_AAAFMP bray_s_Page_110.pro
d1d86759f76593ec18494ed114a9c858
84027fdaa07dfeb914e5eab0a1eaa288ee641759
78547 F20101130_AAAFNE bray_s_Page_006.jpg
0e3f0e19f1155ee4e8dedfd87d26a5e4
4effa5b74a71e2ad3b1d9b3e35c06d1d53ea66d9
111359 F20101130_AAAFMQ bray_s_Page_087.jp2
e517989659acc907a92b0185fe687d4e
f55993cd8815a3f862f64ac59023c14a9ad2631d
57150 F20101130_AAAFNF bray_s_Page_009.jpg
41bf34ef23d5c146238e604c3f241f7d
b258418086b7d5ac419840fec36543bfee4eb25b
1896 F20101130_AAAFMR bray_s_Page_072.txt
f4a4b3f1bba46662331699641334434b
1114baf72b88e5a14527b3a68aa8de8883ad1923
71452 F20101130_AAAFNG bray_s_Page_012.jpg
2ae1bdd6f7004c38e361b0fb294b6192
f4debffebd370f11d4e3dbff7fe34c8e71fcc830
6496 F20101130_AAAFMS bray_s_Page_084thm.jpg
6a717dc99c21c3629bf3db846fb96623
3bc049eb17d5d65d42e54749ad9651c08ca42e31
73751 F20101130_AAAFNH bray_s_Page_016.jpg
10b4f0396dbc1b98a2a14a504b8bedf6
effedc9af3acba5904899ef4f30d23ca3ef230f9
F20101130_AAAFMT bray_s_Page_122.tif
bfa8c706173fc34112cdfd7d6f32752c
98f785790f2d5ae6cbdce217ee9dd93b88fc1531
13542 F20101130_AAAFNI bray_s_Page_018.jpg
c40be0c0d2423c27f40eb809f54640f8
5ac41131b5af0608060cb139f73f5885f0f58d27
58100 F20101130_AAAFMU bray_s_Page_041.jpg
d2cca6068a84094d6efa3df52bd7f08e
eb5cc997b8ca0cf7dcaddb0b74b9ff81a64b673d
22896 F20101130_AAAFMV bray_s_Page_031.QC.jpg
a28d51a23b65dfe0b3cd1da6d03635f4
058924b46423e2dfc83cfd896cae020824ec428a
34385 F20101130_AAAFNJ bray_s_Page_019.jpg
1609e5313675ab86fb9e83f3a5825a12
3da57807d46e588bba19d459ce139b264eaec877
6239 F20101130_AAAFMW bray_s_Page_072thm.jpg
401b95eb6af74daca87aa8ebf0b6d482
14e97b601c65664ed46a6aca7397e556deccc0d9
66078 F20101130_AAAFNK bray_s_Page_020.jpg
3363a1b0ff07198f80edcbf1a1b9e8e1
fd5cf0e46d4741fa4b008832738ffebea43cbea9
185756 F20101130_AAAFMX UFE0010945_00001.xml FULL
2a12d5525a786822a02cb83dd771c6c4
8278aef486af273f3e9eaad842d93d10e49712bf
39681 F20101130_AAAFOA bray_s_Page_058.jpg
2c424639677a8631f7bba1be9dc3d51a
f39d04cedd65d85d04877eb2ff3ee0b8b0480070
69003 F20101130_AAAFNL bray_s_Page_021.jpg
2f34f0bab5fd5bdfcce0ec042e9bcf8b
362996acc87e73bae8056b4e3228a39f7eead102
41293 F20101130_AAAFOB bray_s_Page_066.jpg
24d89927b297fd117f722ba0e468ac58
9aa31b8f600621079245be0366c106df3bc0a531
72924 F20101130_AAAFNM bray_s_Page_023.jpg
55e71ba34ea6a61cbcd9d3bc00b1b469
6afc3f9604540c216918dc629451d4babf2bd32e
40614 F20101130_AAAFOC bray_s_Page_067.jpg
dc0cad21af7e79ef55beafbfff9ffadb
53a887f3edf4aa98e0c27555b668b62bbefaedc4
73214 F20101130_AAAFNN bray_s_Page_024.jpg
a7633044ce2d46a4852d2cc2a9f34939
c49ec82e70a81f146a3b14d6c0bf896b7a4a0886
24537 F20101130_AAAFOD bray_s_Page_069.jpg
4e498823e37fdc81c2ac6f7bb6d8ec7c
f6cbd2a47651157de5d9f264b761f2b4b50fb6ef
73405 F20101130_AAAFNO bray_s_Page_029.jpg
93596f0a730d8b148e2cf48a0cf58cdf
28cf644c73615237bad7b88ce66e76f8c7bb646a
65327 F20101130_AAAFOE bray_s_Page_070.jpg
e9f6a7ec90ecba85ecdda147aa823e9e
44d71867bee00fafb942730bd293cf86fdd9eca1
70610 F20101130_AAAFNP bray_s_Page_030.jpg
201f8e3ec65ac2582f906ad49953725e
f06bd8459c95398d0a7badf749e63384bb491298
71113 F20101130_AAAFOF bray_s_Page_071.jpg
5f609a6905df44f0be236af2469b313d
fc20d915235f23aaa37fdcdd437f3a0f8252fa5f
70302 F20101130_AAAFNQ bray_s_Page_031.jpg
a67d898fb79badf552dc7026f121234e
a4ba3b7e09bb502a6952f2c5921298955c6b3d29
70536 F20101130_AAAFOG bray_s_Page_073.jpg
ce53233fc7cf6e93ecf3cecbc5ab9230
83b7c6651f2f4e0e75533e7d29bb12a52514b158
71495 F20101130_AAAFNR bray_s_Page_033.jpg
4ee4a38b948749972dd707b072c7cb78
efd7ea25fccfab9af51dc2c808e03dc4898632e0
70351 F20101130_AAAFOH bray_s_Page_076.jpg
36d7766a29609d8c2a7e1039b7f47cce
561975827aa9aef3ac95b40da4721b236c537950
76531 F20101130_AAAFNS bray_s_Page_035.jpg
a05c647c0046a7b38d54a7c0f12242ba
7a8e9107a3d226e03433c1898f54dcf820261bff
66914 F20101130_AAAFOI bray_s_Page_077.jpg
cb9f0a081d3eab78fc6edfa6aa8c5621
266e7ac4769bd0bab670a9cda75997458df72744
67203 F20101130_AAAFNT bray_s_Page_042.jpg
dc20ad98db81459b9ad4badb9e1e39f3
d86e73e8221e3ae5cde6ee31f890f1ade41b3305
67513 F20101130_AAAFOJ bray_s_Page_078.jpg
d287991731d9352cbd12cea505bb9b99
91c6a32c1e1cb894b6c02fed77252e2461750373
70262 F20101130_AAAFNU bray_s_Page_043.jpg
c9fd8efca315a8c4a7b1b9ef0334012f
05c4375d56ff57feb0f17144c0566cf6ee912537
70957 F20101130_AAAFNV bray_s_Page_049.jpg
e822dd8c364e45062f0240e61272f896
f310cee5dc2d459ce811018aef131404d79afa70
67527 F20101130_AAAFOK bray_s_Page_080.jpg
3a7e82bfcc06dd9461db28929bd83cbe
f83c1ae519502f6f4c9d84c8e75e832fd8fefd13
70746 F20101130_AAAFNW bray_s_Page_051.jpg
3c35be72f5a82737f550fdae4729e01f
59eddbc191e99a398f5c4515c0b8eef685f40e0c
71096 F20101130_AAAFPA bray_s_Page_110.jpg
2d8625e91a5e49832be13558277fc446
1595becc9bcd1909c54e673bca73a735e9c3b369
74286 F20101130_AAAFOL bray_s_Page_081.jpg
f9fc4b2dea4c6073b62c0031c5eb4a94
64419bb0ac81839d94699f68431cfb3d564ec449
75270 F20101130_AAAFNX bray_s_Page_055.jpg
4121b663ca03894ea21e6c9bf00ce146
d74e0dbbe6d0c913239d6e98cbec48d205ef4b89
71816 F20101130_AAAFPB bray_s_Page_111.jpg
d5b840712c6b44b6e3d64dad32d7c481
ffa67e6ec141d145e791c98b8e4864955de137a4
72095 F20101130_AAAFOM bray_s_Page_084.jpg
aa9af4d53716b2803fb357c2d9b1587d
935dfe133653a610d6ad26d869803b4bc216e75e
73835 F20101130_AAAFNY bray_s_Page_056.jpg
4746b521be73c13c69f4f517c4ace447
9042c9bd0d4465d50804f7b1d7b78c7e70b2ae8b
13873 F20101130_AAAFPC bray_s_Page_112.jpg
62d82fb332f150dd551808894ca43867
612ec032da73f205465cd32452c6663570912984
75542 F20101130_AAAFON bray_s_Page_085.jpg
48479c505fb4590c18549a414adcd01a
694b0dc08c8de8313f8173af7b2e4f61ef9d201e
36387 F20101130_AAAFNZ bray_s_Page_057.jpg
af47f38507342f968d7b62f0f9e457c8
8ef75d8f40166a1193a4662166f56efcf512e253
84201 F20101130_AAAFPD bray_s_Page_115.jpg
ecd0b2f992ea164ff4f49605dff54911
8f5adfea341f5286f204f5a1891662259debe1f7
73961 F20101130_AAAFOO bray_s_Page_087.jpg
9718403b95b2cfa47b12bd523ba1573f
2f8f09aca477feb1df2266cf1156f4c16b1651a4
81426 F20101130_AAAFPE bray_s_Page_118.jpg
e2260a336546e384bcab0362e13a03dc
0b7eea694f8da95923556032d59d78a7fc216cd8
72445 F20101130_AAAFOP bray_s_Page_088.jpg
312edca997e4835e498df1130d031212
ef83a5b479614ab018c738fc682e47735d8d18df
93886 F20101130_AAAFPF bray_s_Page_120.jpg
7d02ac47e1db9860f25b99d2234b5ed1
37a90b62a315694ed5ca51cd8af8db2ab6dea227
33140 F20101130_AAAFOQ bray_s_Page_092.jpg
06cb5afa7d3fb5cdcc26218115fe9810
0d488a804c17b39c50f0da568b43125b9fbd8be5
52088 F20101130_AAAFPG bray_s_Page_125.jpg
57881ff795be8276849eb81cc53bfb5e
5b4707badcd1061ad6fc5ee6334aff27217e91fc
65641 F20101130_AAAFOR bray_s_Page_094.jpg
bd9df96c7279ec5673599207c5cc4c41
85b93a51bcf51c07f556257e19a62d3593d5dbb4
26775 F20101130_AAAFPH bray_s_Page_001.jp2
cfbfef3f76e828332d382e4e713d6c0d
9694358baa775788cd49a8f56ff8ffe0188165f8
65585 F20101130_AAAFOS bray_s_Page_096.jpg
a2e3f793d626c0596093917f9420bdd5
2fbe9c4bdd4db7839d36278e9887f04885bc1f97
F20101130_AAAFPI bray_s_Page_004.jp2
77bbeb735229845e71ea590cc3b84a42
6e7b963de7f71c63a20d797e0f0894140b723fe3
46440 F20101130_AAAFOT bray_s_Page_100.jpg
c59c5bacfb887703b1745733ddfc4e2e
259167e261e836e00f938fe9039c77e52f78b147
1051975 F20101130_AAAFPJ bray_s_Page_005.jp2
c4aa2edce80158d3fbcef9a7094f8a1c
a7b95cb0024ae607cad1fe98846b514dcef17450
67043 F20101130_AAAFOU bray_s_Page_102.jpg
822f40fcfcf6fc645cbe88b3e6c39697
8a4ec097880a4f290be9c8e1ba3fc3ab49fd0148
83150 F20101130_AAAFPK bray_s_Page_009.jp2
f71aafcab46d2977ae1462a47ccad7a9
fe3d2a3eb746cc715b8014c5a014944f5b5ce387
64160 F20101130_AAAFOV bray_s_Page_103.jpg
489ac887843217bce367b36f32942de6
06093b080cb6d293369a747ca14363c06d97d95c
21658 F20101130_AAAFOW bray_s_Page_104.jpg
38580fc3f9267bbe85f3c7852653db23
36823693e66faa41946d37588ea991a38f96790d
86314 F20101130_AAAFPL bray_s_Page_010.jp2
8dcd57b57c25261f0ea535f76110704c
e0214be53efbd1f16af8bdd0fed910cfbbc3f0df
38690 F20101130_AAAFOX bray_s_Page_105.jpg
f87ee8d0e479a1b3922d17b8025157fb
58788f70dce2d82d2602cf7bb79cb2e243a199b6
86321 F20101130_AAAFQA bray_s_Page_039.jp2
9edc79f40b7f8158b86f2a81d326f2ba
2f1bf246977a16c9291b8682615bb07c0f98b5f2
107936 F20101130_AAAFPM bray_s_Page_012.jp2
2690a0d9e45e2d69c7492628450fb56d
6872d997871786b10e609d44fab7a85db00963b2
27655 F20101130_AAAFOY bray_s_Page_106.jpg
00972ff791597c7e85319c47efc64104
56c2b17da6d1827bcd0ebe581b37ac6d6dc11a16
68055 F20101130_AAAFQB bray_s_Page_040.jp2
e82c41211bbe174d1eba0d4c0cf2db55
737a4563ce633502dc1823cab6920ffb9e2718e6
111695 F20101130_AAAFPN bray_s_Page_016.jp2
93dca923a6d385173efd06dc502d43db
fe6bcdb5db4e33417bf2396f352d20a4f9df0bb7
29315 F20101130_AAAFOZ bray_s_Page_107.jpg
00255d4b5cd8ccf9c0bbf0b1820e3ee1
8ff0cb6f22cc999db065069acc7a83507cfb4a7c
108141 F20101130_AAAFQC bray_s_Page_046.jp2
6cba2e8511afadff26cfd593a2d92035
48a8de49deca814070953ab819148e3b50506c03
11931 F20101130_AAAFPO bray_s_Page_018.jp2
a57ef19cda46d3ca781075052b971547
12be685b5706edea323ce49823bc1255dc5e8934
110060 F20101130_AAAFQD bray_s_Page_047.jp2
bf25ed7738d3a95e5463961a7a4c2a9e
8bb777292aeaedb6353373dffb1d0aa36a062596
462109 F20101130_AAAFPP bray_s_Page_019.jp2
98c282698003e4e27ac714c3b01cb268
eb69dcf1ef68d3d12e548e32eb41de3d004dfb93
106567 F20101130_AAAFQE bray_s_Page_051.jp2
53544d705bf80245845a549c05b535ac
2b2bab82ce0d7f5225e3861e9aafe070eda7ddf1
105968 F20101130_AAAFPQ bray_s_Page_021.jp2
a89e7369371f70c66be3756d6cf7a445
9c24f876c4ee77fe9fee7b8793c264d825c44762
112155 F20101130_AAAFQF bray_s_Page_056.jp2
500805733deaf2d10a5d68920abd4281
93889f5eb26bc34d35167b9bdbc1768b33c7a09f
110424 F20101130_AAAFPR bray_s_Page_022.jp2
8f6a17fbad5c9789cf139c285540e96b
371903eff8aca37eb1022c529f72dc3e9df25d23
50002 F20101130_AAAFQG bray_s_Page_057.jp2
94179e6c64688b67ba9acf14cd807ae7
8b3e15bc47fa45f392ac31d74a1f2e618ca71685
110624 F20101130_AAAFPS bray_s_Page_023.jp2
cec364e4fc7f368caedfd68f952f7e5a
ace8bf640bba44ea23141f0cbbf96a7d6fc57115
99301 F20101130_AAAFQH bray_s_Page_059.jp2
bd365a8da0fcaae0648b0c085ae11de7
acae8528da4138fe7c5af6156ef8ccdc9db1001d
110321 F20101130_AAAFPT bray_s_Page_024.jp2
6fbfb5723ab7bef798812a152c114ea8
5065d58c38a8b8655502699f5289af250dee0e1f
291707 F20101130_AAAFQI bray_s_Page_062.jp2
a5272c24b89023fb9e6c2dfb40cdd9f1
4ccf828a7f5b8f8b1d66c89d14e35bcfde0cdcaf
107052 F20101130_AAAFPU bray_s_Page_025.jp2
95fcce18ce6644ae588fbfdf748507c9
134dd47e1c8a24e15395ecf476104e95bb9ad012
271457 F20101130_AAAFQJ bray_s_Page_063.jp2
aa40a22d3b4288272ddd38aa1f84160c
b56253c393a058e273bc2a230a04998d29b08cbb
110777 F20101130_AAAFPV bray_s_Page_027.jp2
4430ea33d5d47165b7490d475a483f1e
21b7715e491d355d8a6503f4401053d770b85657
481754 F20101130_AAAFQK bray_s_Page_066.jp2
31988b06257f1c975fc18c155aba37aa
eec329011b5f9ac91d00c45d1a9c020fc43f1b67
105404 F20101130_AAAFPW bray_s_Page_031.jp2
cc800f08337fc020442d84e9b4650bfd
7e2ccc99a620df286dedbb2aa101b04a593eeb5b
232022 F20101130_AAAFQL bray_s_Page_068.jp2
8fea762e50ec1af3dd784d82f8e7fa8c
3b18fddfafc466850e05c3ee5d147e2479080f89
97596 F20101130_AAAFRA bray_s_Page_094.jp2
13a76fb233f32ab470014e242623cf65
2cd90c7fc8f1bdcd834c1d08f90608dbcf01f168
103882 F20101130_AAAFPX bray_s_Page_036.jp2
9e0f09c83e8d40fb84d9ed0b5bf10db3
8c58c83a173e691595142f1d67aa1ddff7516c1e
71635 F20101130_AAAFRB bray_s_Page_098.jp2
0738c2ee25e5566de288ffd9a1da2d6d
c445502c533644e319a6aa57fad548fb24f53677
23171 F20101130_AAAFQM bray_s_Page_069.jp2
01e1d29d18ad255e940e41d5cb693b97
ba3f98066624c937cfd078125047a72fd8b79172
36109 F20101130_AAAFPY bray_s_Page_037.jp2
a62ba605b526ca54c71be64101e105ab
b32f63f5270f45dea47bb40ef4922d5d59243c08
69451 F20101130_AAAFRC bray_s_Page_099.jp2
f49a30a6ed93f51927cc9d5611b7a994
6cc8464772bff56fe892fea529379098c041d1ed
96010 F20101130_AAAFQN bray_s_Page_070.jp2
7c9f2b7e576cffe2a0a7fa620c0d63e8
b3adc0dd8fcf71fe81791e1d69aa51d70b843540
102706 F20101130_AAAFPZ bray_s_Page_038.jp2
ff4a4de81ac389cff8ca27dcc552fb2d
8fe5107e7978262224be3d56a86cfe8fcc05d9c7
867019 F20101130_AAAFRD bray_s_Page_101.jp2
7eedbc6e6288d5aed0d96a9b6725a19c
a8bab5bedc7b5a37bbed7ec665be579a355bf831
105836 F20101130_AAAFQO bray_s_Page_072.jp2
0cf7ba00476f825537bd2c221fad61a9
fc919efee7cad01e336fec9a403d1b3c71ff8556
1016269 F20101130_AAAFRE bray_s_Page_102.jp2
fac20189319c52940cab7b80c2f3b875
10edad296ea8986a17d503c3008bc940cdd6c6fe
106911 F20101130_AAAFQP bray_s_Page_073.jp2
fde04efcd9def5ab0688f52e352d1e68
97566031bf16fc7d3720d544d056a3c4ae823e6b
507552 F20101130_AAAFRF bray_s_Page_105.jp2
9267e306329781efb2a862518e8fcde5
b2b582e8ac2c0b547b4c3a7d122946f61770a3af
109021 F20101130_AAAFQQ bray_s_Page_075.jp2
c5634e58257de7ff61f3b70a84b18ebd
b230c26bb1a5d639320c4cc82c517e1e8b09e582
359630 F20101130_AAAFRG bray_s_Page_107.jp2
e1a7832b3494f3d7e7763e1cf0dd9cd3
ea3985037f31b899dd7c33e5f1207b9c7bff961f
107327 F20101130_AAAFQR bray_s_Page_076.jp2
40f3060d712f5b88f33e729e0f8029c6
3fbc06b26ac503d10ec4153067262f38079bb8a2
109040 F20101130_AAAFRH bray_s_Page_111.jp2
249e39100ec477e090baf409f2326466
e4c44cf3cfdcb80802f58aec3c94c2c06574f5e4
100406 F20101130_AAAFQS bray_s_Page_077.jp2
e9334da5d12490496d2c431f0a270034
d01c8c203c99ff2fff827c2dee1b5f04fb835953
12140 F20101130_AAAFRI bray_s_Page_112.jp2
113b9566520e10c35f0ab90740400ee1
89c7a8fe31099e7a234e19deb6199f8949d72525
99773 F20101130_AAAFQT bray_s_Page_078.jp2
fdebe07cba6318beef7657601a953ab7
3d82ae734361515eedb5f0d7a196c49ebc23f03c
134861 F20101130_AAAFRJ bray_s_Page_114.jp2
3053681cfb3f7bb0f97e9d879b05c99e
32ea91741b90d1f3330dd193cdad7815feaff232
106944 F20101130_AAAFQU bray_s_Page_079.jp2
c59dbc63e1ca99992bc52d0ee9097a37
2cb2049657ed3c1bc3bb89f6c0258913f12f6a04
126456 F20101130_AAAFRK bray_s_Page_115.jp2
b303a986aebb5719f703eb64cc849639
89d36c87fcdc4b6fa21dcf459b0e9b1233277590
110287 F20101130_AAAFQV bray_s_Page_081.jp2
a184afac10d3afef9561d0c61b964a7b
c2ab72fc4ce9ddedb73defc70075d8bdd360da0d
136298 F20101130_AAAFRL bray_s_Page_116.jp2
b53daba0d467468e4ccdeb77545eb68a
cea9282442877826ec448de31fd72d16b64b61f5
109473 F20101130_AAAFQW bray_s_Page_084.jp2
c38a3678bcff6c14df974a5e0cc78ced
9ab50537b5e44f26a8a20f5d8dae186976ade28b
132402 F20101130_AAAFRM bray_s_Page_117.jp2
2a43487c9b8ec273688767f5033f29b4
2734a6aac9ce4e8c2d218c9c976890f47c1df1f4
106105 F20101130_AAAFQX bray_s_Page_089.jp2
27b742a9502dca3ae7d5597ec36daf5f
56a3d34d16652a5409a6e2da5fee32c2fd545a05
F20101130_AAAFSA bray_s_Page_021.tif
e54b2c972ce746c043090a9a8f0265a6
98a86f02f2f4c4f38bb06a2a2c7e77a02bea9480
107425 F20101130_AAAFQY bray_s_Page_090.jp2
b01a06886d9fc676a162e6b808c19c13
7ff76db0a7ff52556b4ad070f4d3b0a2902abaf4
F20101130_AAAFSB bray_s_Page_025.tif
c6e599c643d63c921ee6d921ecd2cbc9
6e39ee6beace3a3276c732d83507181cb38592ff
124736 F20101130_AAAFRN bray_s_Page_119.jp2
b9362b31291fa75133ad1dadeff07911
dccde1cedc2515866be8d10031d3d5806686ee23
57128 F20101130_AAAFQZ bray_s_Page_093.jp2
4ac372185df2e21920f0f4b10982324d
d5eaf3d23c9e7a445048ceb96e8a25ad3b266cfb
F20101130_AAAFSC bray_s_Page_026.tif
eee5ea1ccdfb5b5eaad063d47eccd4cd
3e6dbde65c762967dc49afea9bc08a9745aa0c99
122070 F20101130_AAAFRO bray_s_Page_122.jp2
2907e0cc179bbacbd5ede9eee42147da
ec192aef8a9addfa50f48496f04fa501aeac26b0
F20101130_AAAFSD bray_s_Page_031.tif
d19e67a4478c14f0f96fcbb4d950ddfa
3b43f8667bd97b44ea8cf4ef62d45f17e1cf8e92
124884 F20101130_AAAFRP bray_s_Page_123.jp2
09e5e5f9f456bba2ba72555cb6104a47
c2ce1e43313e68507d0ff1966d891758dd6d106a
F20101130_AAAFSE bray_s_Page_033.tif
3ab122ff63b14b36f1e85dcc12d0fef3
6405b8b8b2838552ff53ba9867a780dcced6d6b1
74892 F20101130_AAAFRQ bray_s_Page_125.jp2
5da483735949ab315311c5fab014b9b3
c9aa47a832fafa788a2b97d43f19adc17add32f6
F20101130_AAAFSF bray_s_Page_034.tif
df3bb5232de3da18346f6181ecba6d57
8627b3ef3beb595b101b49cfd998009dcd97f89c
F20101130_AAAFRR bray_s_Page_002.tif
743c967d8e883a13bee25a39613ed7ff
7fa92e321a8c199b93f4ff08d8ccd3e0f076a2e9
F20101130_AAAFSG bray_s_Page_035.tif
fd7fc2bed889fbe98820100536261d12
e2fe2a66b0d5686bf9e5714c7b5b190ee73fb43c
F20101130_AAAFRS bray_s_Page_003.tif
7c75b2c008b63fe893954ad41efec9a3
6f1c638ca37e916f4a726fc85f5df466f08b32c1
F20101130_AAAFSH bray_s_Page_037.tif
d7cf9b64221b640c55dca824c68d5ad1
1325515599b995f5bd1ca455b0f5e067862d79ea
F20101130_AAAFRT bray_s_Page_004.tif
e582ce069df49baae0ef11ab4302fcb5
0130edce3a85c771068773563a3ea65390d0ad44
F20101130_AAAFSI bray_s_Page_039.tif
b350f078e666ae6dd74b1ab5cf0cefc0
4e3e9ae5ee64a5516c465df1a0a88224051e90df
F20101130_AAAFRU bray_s_Page_006.tif
8db97084d60cb75c8e88c25663f0ca5c
fa6068c0268aef367ba98b648eaf6767381ea8be
F20101130_AAAFSJ bray_s_Page_040.tif
b3bb123d7e77041842b0c17debb0ff69
e445ea1799157a97569cabb9555919b09172669d
F20101130_AAAFRV bray_s_Page_008.tif
6cfff333083141a39f2ef232d8327868
aa3f985991c6b829b889382e12108ec9f0274df1
F20101130_AAAFSK bray_s_Page_041.tif
d8108e1d5d1cf998c346ea9abddc2369
191fc395fbbc782f9ea88fbd2c4f15030f1d2692
F20101130_AAAFRW bray_s_Page_009.tif
2d53d2e1dc45345f11d5fd6b453f1d1e
b01c9652a023667898c9b19251006b925d5a08ba
F20101130_AAAFSL bray_s_Page_042.tif
abdeb4fd31527deb0d66a648aeb41ff1
90c4509e2dacddea2c9574921c35057fffab6217
F20101130_AAAFRX bray_s_Page_014.tif
6e2b047fc21e1d3db53b3c09e8e5aa1c
4f4ce9e05c3486dcef40956a8dda0f65c72dd75b
F20101130_AAAFTA bray_s_Page_071.tif
863ce10d71f56e0af6edb3451db41c65
c92902c70ec5e2913c25ee293793583344a5bf90
F20101130_AAAFSM bray_s_Page_043.tif
05e5868d1ff4e82280bc4cd371fdcc99
1d23f072700e2a80ee6a5aa131af04ced813b5d7
F20101130_AAAFRY bray_s_Page_017.tif
1b491c2b96984ff731e7eb8821e84c98
7b043795e47c42fb9541bc0e68e2bcb5ec9a71cc
F20101130_AAAFTB bray_s_Page_072.tif
67433cfee921a6648c8b2b90333b1bcd
a14dcc68d162b692714e7ce4ec81562d373a7e78
F20101130_AAAFSN bray_s_Page_044.tif
83af4feb8ebb5c534d3eb0f2a25d2d33
3a589a514c9759db08cdf38fd983473c5964d7b2
F20101130_AAAFRZ bray_s_Page_018.tif
60deb414c97574630641d19b3b1eddcd
ffaed2c8256b11c03c723160a8ab4a8c3cb48f4b
F20101130_AAAFTC bray_s_Page_074.tif
fb7dbc643426b0d815641ac8b67ed168
ac41037514778aaa4efbcd1c35f2cfd6f77aa278
F20101130_AAAFTD bray_s_Page_075.tif
b1a1ba96436c79efd346b930d59e0ac6
f9a5a9050d558e4a23b48c22bea3e48367555a05
F20101130_AAAFSO bray_s_Page_047.tif
121c2e7e72e4292eea9fbc7ceadddc3c
356268df762a880e214f3ace12f235a5b6ca3315
F20101130_AAAFTE bray_s_Page_076.tif
23fae966febcdedc19be7541318a5c5e
cc3d76ef8567f46c3c980632a3ce10b70715a2a1
F20101130_AAAFSP bray_s_Page_049.tif
080525d789d7e990f69bd76c78ecadc4
1d40dfe51443d2fc011ff6af7178bfa1b066edee
F20101130_AAAFTF bray_s_Page_079.tif
54bf467df846e647d303ed65f5d3ea1a
4f73bb3ca1db3749c9e492ce54771aa6f1f56947
F20101130_AAAFSQ bray_s_Page_052.tif
2d49ea85fb3f844d7a5f992ca08659a6
cf738eda1b2703feba18e20ee68c66dc35ea30f5
F20101130_AAAFTG bray_s_Page_080.tif
90d29a4a877571f9550a0cef888be432
1a4000ecba1fa2561a7aaa54276e58a0d4f0ede5
F20101130_AAAFSR bray_s_Page_053.tif
7cf0b9d6598cf2d1a49a70dd1a535ec6
0735bf2f5fa4911c82070dd806280af458781131
F20101130_AAAFTH bray_s_Page_081.tif
13c8986dce972e304421db2fbc8a09f9
97c325d78a5c4e378ca9bee52a796393a4f88c93
F20101130_AAAFSS bray_s_Page_054.tif
dc027cbc168575f3f665ec75b10fcf8c
99959ee07a518bb1d1df86ef8bbba0681044f976
F20101130_AAAFTI bray_s_Page_084.tif
ef7edaae6386ea71fd4dde40317ed9d1
58b3191ef5c7a0374db06cea8fbce0f9c9eab51c
F20101130_AAAFST bray_s_Page_055.tif
108c52ddd2ed309a00855ce4190195a6
a7b4e32ecbc79336ea6b76c27fad93cd0a6d9cbe
F20101130_AAAFTJ bray_s_Page_088.tif
6dea7e6a0ab688954376f5002f1d2202
aa2872eaeb21d36e75c4119c17199494c5be975a
F20101130_AAAFSU bray_s_Page_059.tif
da1835baf2bbcee0a28be79208a9fb48
151d02d88f3ea1915637c03cda246edb4477a8d0
F20101130_AAAFTK bray_s_Page_089.tif
507eef09b1af9ed734df05d8b0e19606
4e895162dca4eb4ac045e1d2c17072b191a03e86
F20101130_AAAFSV bray_s_Page_060.tif
ae852e26c95bb89ce4704d1ffb5dd11e
61a74e7256c4d4de2aeff13837b5ca04739166b0
F20101130_AAAFTL bray_s_Page_090.tif
c54d55f4f9295dc9f4e9270d4a193f04
7c5a411d9f7568e9e9dd40d77e353eadaf66d738
F20101130_AAAFSW bray_s_Page_062.tif
0d8983b09f936e9aba20852d3476a6d9
ecb9c56e9a6346a39f1455b8bd2d979739d683bc
F20101130_AAAFUA bray_s_Page_124.tif
c1e2eb4a35fc0576a288246f79afc4c3
01091457f6b61706b25172cd71ab9088ff12c3c0
F20101130_AAAFTM bray_s_Page_091.tif
c170871a9c5f960862be0bdc32aa4529
ed6dd418afe70a38465bbd973a29786359d7ee32
F20101130_AAAFSX bray_s_Page_064.tif
ac97619ff89334795bb88b77e8a2635b
6b8c524d975ff37177aa150665a036b9d95bc4c2
F20101130_AAAFUB bray_s_Page_125.tif
8917f4a2b6da45f8d22792c7e4c99ddb
b08ac8c1069fca5632da32a5ab7a3265e7a02326
F20101130_AAAFTN bray_s_Page_092.tif
883db54766be67f3eb504b6f824ebffc
0d3a970b091c9663eb44a5f8d34333b2f71f1144
F20101130_AAAFSY bray_s_Page_066.tif
76245993d0f7fb8b0a1ec101cbd70f20
dd47aa1c93b24af384341dbfce42dd6fc0c6da94
1199 F20101130_AAAFUC bray_s_Page_002.pro
8f6e8a861c5b04ad3db1b21c15dad4ce
821db2dd264c1d2a24b9890c9964851c5ebc256b
F20101130_AAAFTO bray_s_Page_094.tif
749b1a3f350f2aaadc933c2b459e798a
334f32cc1d9dc73ef7b4b076c9b0762f785ab956
F20101130_AAAFSZ bray_s_Page_069.tif
7ffccfff445473efae923eae2bdf1bd4
8aebae2a60760845660bb218cf8678efd11c75f9
45331 F20101130_AAAFUD bray_s_Page_003.pro
241c9c1bd407480ad9974832d2ac7503
a2050b1549765e3a7044e24e726e9adb9b42b9d5
F20101130_AAAFUE bray_s_Page_004.pro
9313fe0da8e9dfc21225d47632f61be8
885a644f5108896733f38a4e0380ff15384a2874
F20101130_AAAFTP bray_s_Page_095.tif
2d5014f5c6685558dcb7e59d5554d8e1
cf0b0bb19c4eb620e71da1da2644419846ae4772
68821 F20101130_AAAFUF bray_s_Page_005.pro
197317e296332210eade7f80a78060f0
c43a61bf0b685a4f6edaecf85291895b057f839d
F20101130_AAAFTQ bray_s_Page_098.tif
4b2a2367d034e2a62c70745400084992
611daed498d1aafa0d24abd6c76a4295c5e2fb98
21118 F20101130_AAAGAA bray_s_Page_020.QC.jpg
1f1d519cdf2b3b75229619da98f1faa1
b2f19d96fc0c0a9118780e602ec8c5b52ef1347a
59385 F20101130_AAAFUG bray_s_Page_006.pro
96faac36c362abedb9c9724d6dc10860
c0cac1fc40bdaac38351d1b334557c0ecad2f75d
F20101130_AAAFTR bray_s_Page_100.tif
a2d3dc86898049dae73b58fc30f4064e
5dc3455a9d139eced17df907059f4164e68d3cd8
23333 F20101130_AAAGAB bray_s_Page_022.QC.jpg
d847315f7a676f3a646664c040a40f15
b1324c1b0b4e59af6a2f8fd033865f950141db2a
77364 F20101130_AAAFUH bray_s_Page_007.pro
b737ec006689397950a8040875347c26
b4e820e07326105f11b10c813cf505030bc4ee03
F20101130_AAAFTS bray_s_Page_104.tif
e5a47f580ee7a5cdf99711429b0ca092
ca25397bac85b994b743206fc66535ecc049c823
23035 F20101130_AAAGAC bray_s_Page_025.QC.jpg
afeb8dffe28436e5f728f5e60076d848
43417be1ff9a8947b177462498630c867cf585a2
36383 F20101130_AAAFUI bray_s_Page_009.pro
39b38de370f217c867c4f98a1fa606ca
793abc740a3e7f17d244bc1b26c5dcbe961f4bdb
F20101130_AAAFTT bray_s_Page_108.tif
6efd31ab023c801c6cc7589c318b3db5
005f9497ca755da22d42c942ea2b2e99fed65166
23678 F20101130_AAAGAD bray_s_Page_027.QC.jpg
730bde8700afb9e0044c6db1042817fa
30c233e7b8d290cd5f900b1a0aebb125d9090bc5
38906 F20101130_AAAFUJ bray_s_Page_010.pro
09096de508450121509f634279976169
dacc9f8650fd7ba7abc4b5887d09a87d15073d63
F20101130_AAAFTU bray_s_Page_110.tif
393e0c599667009d1473eecd945e29b9
a1dd19dec5d584af6f644c046711ea120d15b92d
24230 F20101130_AAAGAE bray_s_Page_029.QC.jpg
6c0581be974941864fbd641da07aca1e
f3beb0971c7fd74b80bbc1282547ed21014a06e7
44720 F20101130_AAAFUK bray_s_Page_011.pro
ed4b240e95202c438b28b5ceda456637
80a0b89943f83605a93f4c0f54f6c103ea1fe195
F20101130_AAAFTV bray_s_Page_112.tif
46818e9b603fa86768d6414868bff5e3
8e3ad15ca36957d9f0629b967ba1345eb100cad7
22899 F20101130_AAAGAF bray_s_Page_030.QC.jpg
5ba6263f6ba3501e67acfa26d31f6b35
97d874b493f79add1a214f35a517e32abe504026
47043 F20101130_AAAFUL bray_s_Page_015.pro
d886d73bab8d8e9a3a9487cc6d929ef8
f09039d70836affb07fcebebdf793d4ec883a61a
F20101130_AAAFTW bray_s_Page_116.tif
1418b21f29b924a7bde1fee1dd5dca5f
24af59f5db5af14c31cd9aba423b4465012fdfa2
6328 F20101130_AAAGAG bray_s_Page_030thm.jpg
835d25a898048406bb6b7f9774e33776
985a9c618e18e07dea82de876d486a74bee381be
52326 F20101130_AAAFUM bray_s_Page_016.pro
bda0d873eeff45731250c78e3c615867
ce24745e0b7bd94dbf7ea30eecb03e8b0e99ab37
F20101130_AAAFTX bray_s_Page_117.tif
4b20e636985727f101a3b4ba34b16b0c
abb7a2a3be87c4041ff33372203af1ec44a8aaa0
47851 F20101130_AAAFVA bray_s_Page_036.pro
5648c697edbafb2a2ea17096019c7146
55f732eb7911f054780fea8e721c859145623ed2
6521 F20101130_AAAGAH bray_s_Page_031thm.jpg
49b6cf889bc4a1ee6ca213faf60de9b4
ba593f5526f60a1ea2c9897ef131492dd6e955ab
48219 F20101130_AAAFUN bray_s_Page_017.pro
0923d4e5bda1ac418fa3c8bb71ed8c1c
0d4dc7b09ab0d0dc8086741b9a47c24fdfd4de7f
F20101130_AAAFTY bray_s_Page_119.tif
af8bc4955a7798cc56cb13c41f4194d5
dd185afe15957aec878bb65fd3615a01ed872591
15151 F20101130_AAAFVB bray_s_Page_037.pro
67c542651cfb6b005594b8c117c294ce
e2695259f5d1efd32609a10cc84103c16a800755
6444 F20101130_AAAGAI bray_s_Page_033thm.jpg
893c5e72c6e4c26b9283bc8a2eeb0ed5
d51891fdd4583bed9e5b7a5d769b114edde85dca
4106 F20101130_AAAFUO bray_s_Page_018.pro
239116398037f85bda2e5dde6a86cf14
f168c266498008b1c6e7fb65364457ce577996c6
F20101130_AAAFTZ bray_s_Page_121.tif
c82b7c23db24f949c7e65e08fd5cf80d
27ab2d2f079cdd5c835b74925d223f11c1e00073
54469 F20101130_AAAFVC bray_s_Page_038.pro
23c2e3e469f7e00d35fca997df4217ec
448b82adb87f90911b27f1bf70f4d25ae996e55a
23050 F20101130_AAAGAJ bray_s_Page_034.QC.jpg
9797b204a1064876f1d41ff8c020fa9d
931b46cec30ce7527bb7499ffb0fb6fa0cf96afe
46524 F20101130_AAAFUP bray_s_Page_020.pro
139531828f453597afe2bf4e4f6a141d
c32b4bafbdcc1576d3836bd0c720609ecbbc67ea
47598 F20101130_AAAFVD bray_s_Page_039.pro
cc9c62744f68437ee158bce99b0b20b6
0b4e64e827bff659357dc03fed3e20b549b7db40
24521 F20101130_AAAGAK bray_s_Page_035.QC.jpg
b73c9c6a92a0849dce9b0a412f915ac9
dff166b4716086ec759447ec6b211d56f2164bd3
26313 F20101130_AAAFVE bray_s_Page_041.pro
96834c6a63e0e3552c85735edaf632e9
24ab9eb276baf00397aebb2edebe3d5fd50d411b
6805 F20101130_AAAGAL bray_s_Page_035thm.jpg
ad2f0a7ba69f1a4ee63b4ae0aa13ea8d
cd5a061c53e586359f7411537b8b36a825b7e08e
48217 F20101130_AAAFUQ bray_s_Page_021.pro
4dc38ad7bc0d413af8364a0ce3eb6188
8ad822a32a172cf00ef77ad8bb4c490a656779c3
49221 F20101130_AAAFVF bray_s_Page_044.pro
1e22a13a6378abb0c025db64718270f6
db41ecb4b083f55dd0ea8b141815dabfa8ccecb3
21871 F20101130_AAAGBA bray_s_Page_050.QC.jpg
503e0e664bc1069cc0c5c06c890594ab
e529f5661a90635d2042cad351f3c833c0dcb0b2
9140 F20101130_AAAGAM bray_s_Page_037.QC.jpg
e2892a7455fbe4605673de08cdf9f79c
ca6e328c917742abeec94487a80094dbd68aeb9e
49752 F20101130_AAAFUR bray_s_Page_022.pro
6f31084db0f65d686795de1d8eae321a
a6e3e954a09af3f8a19bf556d26cb231ef0c0015
49554 F20101130_AAAFVG bray_s_Page_045.pro
46226cb7cef175d104a1438e835dfb84
0252bdad94c9eadb7c61827a1abc1b81c368e53c
6051 F20101130_AAAGBB bray_s_Page_050thm.jpg
75ad12e08d3830d261f4da9486308597
0870b1d33538b4458f4d60a2aafaa56994cc085e
2795 F20101130_AAAGAN bray_s_Page_037thm.jpg
9c55fc86e65f1306f117f9d919f5d48f
bc754cbfed3bf4f1f2f02446d333afbfb14e146f
50912 F20101130_AAAFUS bray_s_Page_024.pro
401b11a99be0a56753e66a3f4b68dccc
f61d57fd68e01aeedb857c3d039bcc72d6c38a2f
49880 F20101130_AAAFVH bray_s_Page_047.pro
e8c9c559ef24fd6092a13b5b789c00d0
6d98f13c5b45082c067dd846bdff19203cd697e2
22733 F20101130_AAAGBC bray_s_Page_051.QC.jpg
65d602ee736da548c61e6ae1f2a61fc0
7b1d7221c469186f9f0211640c0e04ceaaa3480f
21516 F20101130_AAAGAO bray_s_Page_038.QC.jpg
2861354272874d96b1a2e8e8d94b7a4a
90fdf0f14f3a884296a643b2a290df5e579ce374
48980 F20101130_AAAFUT bray_s_Page_025.pro
88d86f4bec3add2a3e4e4158f1790d78
5c74cd41efe3a6e76cdd37ed9f315f93d15bfa70
48700 F20101130_AAAFVI bray_s_Page_049.pro
3df25dfe3da300119e83a0c35c26ee45
308a9de9af9d9eb52a5de3cc38e5d74d24a6b286
6299 F20101130_AAAGBD bray_s_Page_051thm.jpg
bdace6a64edcee48f2416cd220b4df89
a8a0a4e8227500799bc324d9f155e3d0facfd2d4
12229 F20101130_AAAGAP bray_s_Page_039.QC.jpg
a50d841878dcabe93bb17daad63cefc7
de1b6e06679f9e2be1de8f2f024bcb415130cdc5
46817 F20101130_AAAFVJ bray_s_Page_050.pro
b566b9c56d4ea0834436ebb6795ae791
1830c2951ed3e742807049d3af64af8cb6668501
49101 F20101130_AAAFUU bray_s_Page_027.pro
9d3b7631bee9e960c94a12a6d881026b
ebdccc9f4f01e23921e8893e6fc6ff5dd350a175
24025 F20101130_AAAGBE bray_s_Page_053.QC.jpg
d0e50d6c1fe46157d1c1fd271be1c87c
47e25a4f81e01300341fa22c1995370ed4b3ec1e
21498 F20101130_AAAGAQ bray_s_Page_042.QC.jpg
0d46e877aadc92f2ab1bf62294aee683
3c9a98bae036cfa9de31de620596efee462ad8a3
48935 F20101130_AAAFVK bray_s_Page_051.pro
6585fffca9c995376a636f3a66f3461e
1a4ad0795af8facd12c73f98f01a9b687bb69e4c
50593 F20101130_AAAFUV bray_s_Page_028.pro
5ff41f370964cb5c73e42f20bdc5e12d
6f972dec6b3022e6326e14447700eac1d16e7327
23808 F20101130_AAAGBF bray_s_Page_054.QC.jpg
1049c022496c8621a218174334a1ee17
2fd39d90226c5706c8aa3c41750d48841f6ff090
6149 F20101130_AAAGAR bray_s_Page_042thm.jpg
0645a1664447daedc8f01b6e237b3b49
70bd30360aa4471cbab0724ae840b8627aa575a6
50879 F20101130_AAAFVL bray_s_Page_052.pro
be194cea01c4145a930467fdbfe61111
c6851a8f5b615894161fb50ede780dfe2e6909e5
48274 F20101130_AAAFUW bray_s_Page_030.pro
d36a9602e423b043afb5cd972de9e420
d64f56dba1d43f644c6a94cde62354fde5ca54de
6405 F20101130_AAAGBG bray_s_Page_054thm.jpg
8d5e4bfe744690301185e4ec85122ad1
00d4c8d3a0c95c7759efb10c63df826b26e2bb2f
23207 F20101130_AAAGAS bray_s_Page_044.QC.jpg
f9c06dd4ec45f2a6cb5265c420f58503
2ce603a94e624f38f963fd6d6dfe04e35feb720a
53113 F20101130_AAAFWA bray_s_Page_085.pro
0aba36090691fa4025408df57e7b43e2
e13ba3889081e9f05e7553648e029cb98bf8fae1
50368 F20101130_AAAFVM bray_s_Page_054.pro
36b1af12dea2cd5d1a94248b1f403ccf
e70ed58d7188e6c15be0028a95d10c8402605376
47290 F20101130_AAAFUX bray_s_Page_031.pro
897efadb1a5b8b1aa164b43e2d71e70e
4f87685d90dcda9ce77e0e144369941decfd8c63
6506 F20101130_AAAGBH bray_s_Page_056thm.jpg
73ba7e82ff1f71a5897eb7ebbbe5b7f2
1bcb9257a55ac561144d1505e0b79b07a8df1f54
23433 F20101130_AAAGAT bray_s_Page_045.QC.jpg
52be1c1c2aa8fac96cad76d17284b863
01f9004e6d3d5aa7f941c3d4bbc3c054b4455130
51543 F20101130_AAAFWB bray_s_Page_086.pro
4e081c8a38283bc34c1d04f4c926038c
cc798f2d5e5006dc99f9a00ae35b816def732aae
21897 F20101130_AAAFVN bray_s_Page_057.pro
ebcf3a56da794593646c7fe3070de914
8990533fe78ad097921a58fc8f11ca702ed15faf
49362 F20101130_AAAFUY bray_s_Page_033.pro
bf0359fd1037c8ee99f6fa0eabb63b63
faa7978da5b9adf326f2198435017005d2277ab3
3617 F20101130_AAAGBI bray_s_Page_057thm.jpg
7d917a856c9008417b5353f5122b0e7f
0746267f31ecb51c466bef256af8c34b89338672
6516 F20101130_AAAGAU bray_s_Page_045thm.jpg
3e6b8399577ee9ffa2926a8a4ede84cb
977bd3a85db40555af35c89d9e8aa27115f68768
51846 F20101130_AAAFWC bray_s_Page_088.pro
f5f29b28e0128cb952da157d09a29deb
3aa106c342349ce31d3ef403a710ebb94c154d7a
52024 F20101130_AAAFVO bray_s_Page_059.pro
0bd45ab51d16c4412f0a9b7463304a6e
897c7c8f7fc700de1b5460bcf7efeb924ab4496a
52977 F20101130_AAAFUZ bray_s_Page_035.pro
7da684eb85000457bf93c5e8bd6b566b
9c1bd44b6889cea06734e362bbb1b4ea1bb57fa0
21396 F20101130_AAAGBJ bray_s_Page_059.QC.jpg
60219cc25171bf995b0e3b73c976a5e7
d9340e5c3ac1100241280435243502d14cb37c28
23101 F20101130_AAAGAV bray_s_Page_046.QC.jpg
2788cef2cf2943efa74f98ef4eda5f8a
f6f4defae24ef5d60a3bbdb5cc65c2278c9e34b3
50057 F20101130_AAAFWD bray_s_Page_090.pro
e20f5da99de4c5d865d04784efefc5f0
93783ee47924dbe3831dec166df349abe6cb05cc
30464 F20101130_AAAFVP bray_s_Page_060.pro
dc0a1023631bb07c09d0e74ef23c3e04
5e3c85cc0be7862f03161af1042c3674bbf9532f
9363 F20101130_AAAGBK bray_s_Page_062.QC.jpg
a4215fc3efc566d370b3956656d6ca4c
7e81535526c31ac74b4069c8e26a6b4cf21ed7ca
6468 F20101130_AAAGAW bray_s_Page_046thm.jpg
1d4218bae974e306216a65aa00a652e3
e357477377da73d3e17760f4c47bfdb7fa167875
37652 F20101130_AAAFWE bray_s_Page_092.pro
212771bcceaf5f1e96f03ce5261df326
5f9729f2764a8b74e45e5cc8db3541d03cfea972
17760 F20101130_AAAFVQ bray_s_Page_061.pro
ca224375dabbcca7fe97dfabb2ed5b36
ab77c5ae14a0b1c02953f66133a46da959c3b5a5
9320 F20101130_AAAGBL bray_s_Page_063.QC.jpg
d44697463d880214eda062c5007ed317
e9b21d270dcbb90832cca519bc34aa142c83aa86
6534 F20101130_AAAGAX bray_s_Page_047thm.jpg
909ea4f2c8f6c8e0590a2bcd66ca46c0
3541cfdd9895036ae30b7139d9fb4cacaf457a9a
24289 F20101130_AAAFWF bray_s_Page_093.pro
887c6b0a79365fa41a40eddda46f6593
50d72f87e4b123784caad3f764c278e7b7ce9cfd
6291 F20101130_AAAGCA bray_s_Page_078thm.jpg
33e0e9f6b3a7a55d21cbbe27b6570208
ef202f87a786b1e678da5a7509657f378d51b4fa
4523 F20101130_AAAGBM bray_s_Page_064thm.jpg
5376c72b260f3f2a8d94b8539ab6d40d
c54f9bf75f046e19e70eb41986c9303e919ab176
22059 F20101130_AAAGAY bray_s_Page_048.QC.jpg
770a15f7bc8a6b74d921863de73ebf21
4b772a522b51a9b3c6d53dea631cee189ff49edd
55131 F20101130_AAAFWG bray_s_Page_094.pro
8501f9e278096604b944ecea8ea607c3
73de402a7f8a7b020ade8b1c5171357679622838
12793 F20101130_AAAFVR bray_s_Page_062.pro
9c9a8ae9efb1c69dc28005878c8dd780
4f70ddf2824d46655b9c9a911f7260287964566c
22808 F20101130_AAAGCB bray_s_Page_079.QC.jpg
3ad8b8e215ca5fbaab8e215ad09c4815
57f1531af0d1984e0e3d5422a6432da164716166
10571 F20101130_AAAGBN bray_s_Page_065.QC.jpg
356d0bd72ff998f7318437d049203dff
b221b71028d538e568d91667a66971028486c69e
6537 F20101130_AAAGAZ bray_s_Page_049thm.jpg
c1623c701dbe1885a1ebb16b14d0facc
65d7e66211a7466be1ad08fa8f6cf1f9107ed4aa
45892 F20101130_AAAFWH bray_s_Page_096.pro
f3edca6f00e89e7703aa49cabd4c8ffd
fa0a613dee40550064ff63714c0457fa255ac5e8
9621 F20101130_AAAFVS bray_s_Page_063.pro
59c1443e4eb071f685b241e3deb39f80
c89da3de0dcb7beb410ad9eec3f8faa0a26231cb
6385 F20101130_AAAGCC bray_s_Page_079thm.jpg
3322907a3ac343b8adf8ab04dafbba54
38a3ca17f8841d0bc58a0d5a10f31030db7d86f4
12058 F20101130_AAAGBO bray_s_Page_067.QC.jpg
1a5f71a063a25b392445298ffa1fdf11
9672119c50ad2f3b9c3cb241888f2c4f8f8547b6
32970 F20101130_AAAFWI bray_s_Page_099.pro
f0282760e1ea4108b919379cb0b1fec4
dfb799c93f17672103b437b7bc722e8b97cd643a
19174 F20101130_AAAFVT bray_s_Page_066.pro
04d0c64ae9d000bb223902bd24953e9a
274a4399bb410d075a17130afad14a07a94e395b
6316 F20101130_AAAGCD bray_s_Page_080thm.jpg
64799199581637c9b44ba4668f58a074
68cabddb1ad28676a876788070327f552f80f95a
3500 F20101130_AAAGBP bray_s_Page_067thm.jpg
95d1473bfbc41f99fa3ce8aa3ce8b9cf
3ff77dbefce9c9dbd2f516a2a7624028fa1f62ea
33111 F20101130_AAAFWJ bray_s_Page_100.pro
09e4664d5c4cb0dcfd70c8f3bd4decc3
679418d653196588e0d9eaf7e269f91e54af0ac1
5099 F20101130_AAAFVU bray_s_Page_068.pro
7fd6533ccdd9bea2c5cd15c735dca0e9
72b5d988357a605c5e549438af3263ec5cc470b2
23918 F20101130_AAAGCE bray_s_Page_081.QC.jpg
9c003e8ecc8782e79f26c31eb5f02c53
d13ca6589d12f5b1cf65920c3935c3dd075e584c
8194 F20101130_AAAGBQ bray_s_Page_069.QC.jpg
838886e8f5d90732ef9c21adfa0f846e
35bdabb6ac8c6d6c2c042e5d1162027369c15cce
25724 F20101130_AAAFWK bray_s_Page_101.pro
a26a823d3a5c9b633d8d2902f524e819
445942b80e01fb1181e06d6e6d9fca2b1ae9931a
7627 F20101130_AAAFVV bray_s_Page_069.pro
dabda41846a85d5a23205705e63099d1
bc607a5e4b61afb7a5f8458a6f381b0f2ddad4e3
6765 F20101130_AAAGCF bray_s_Page_081thm.jpg
d0aefd6f5f8738d53316b2b131dcdf37
ab1bf54a451073b9e03afacd6f16ba865ae769a4
F20101130_AAAGBR bray_s_Page_071thm.jpg
5ed27886b8a27fa20f029fe0a40af897
b5c7f360652c5a2fb7c8908841bee40ce84cfa94
35647 F20101130_AAAFWL bray_s_Page_102.pro
be5ebb27914f0df8a2c640ae4b50ee1b
707b406e332633763e6aa6b393b4040bfbee0a7d
50442 F20101130_AAAFVW bray_s_Page_071.pro
d29b19d27a500244caabd763853a3e91
2af5b229393c619ccf8995bde64c479ec951eb20
6493 F20101130_AAAGCG bray_s_Page_082thm.jpg
55bb4a20642ed33933e78c8182ff57a9
73751a8f9f5ba525e94875dbe5dd01f27bf98694
23219 F20101130_AAAGBS bray_s_Page_073.QC.jpg
55659e5c0d810b61c20f2c14e20e2b78
6942a996a7b345eed7f230ce2a0a5cd6b4e97e12
27392 F20101130_AAAFWM bray_s_Page_105.pro
5b0f3dc6364c7fd0e3a4cf832ca9df56
37f208f56de6bce0da0c91ff673396394c71c187
49514 F20101130_AAAFVX bray_s_Page_075.pro
538b29f988ff08c76172a8c653882afc
ccd14de8d9ebf293bbc2c58a1da365742abd6eab
1814 F20101130_AAAFXA bray_s_Page_003.txt
1d589495147233f922034e64c1784de8
d0010072f3c0ec5e619df7003b43a7fe9ac36276
24641 F20101130_AAAGCH bray_s_Page_085.QC.jpg
253408ea4cd20bf4e326e0f74cd91ba2
d77fb6b30bb0ce4bb6ab0d66e0a66995b24ca732
6370 F20101130_AAAGBT bray_s_Page_073thm.jpg
48d5876b950a4f395e3154c3d917671d
3b97a04e1983adc6c3f254c091e02fc314d3256b
18077 F20101130_AAAFWN bray_s_Page_107.pro
4a2f5fb2bf531a8d52296ef5e34b20a8
ba3b37f3daa3fcf0e99ec549d72693ddc00d60b3
49208 F20101130_AAAFVY bray_s_Page_079.pro
9a8e44b6723d47f4b04dc794a063f2e1
d976db024974762ea55a788221e171fe21cd563f
2963 F20101130_AAAFXB bray_s_Page_004.txt
61834750a0be0941e3ed84b81b4d976b
99fa148edc3e69df277a46e2d8dd4e51d86ed3df
23600 F20101130_AAAGCI bray_s_Page_086.QC.jpg
bc362639e56cac140a6261ce29e24043
bcc945f68bac39f8ec41620ca0f91f874b8191e9
22965 F20101130_AAAGBU bray_s_Page_074.QC.jpg
b48bc2844f05f13aee4d9902f35bdcb3
e1c23091447d78bc45a8c10a72e63966c42ae422
45848 F20101130_AAAFWO bray_s_Page_108.pro
bd2490ab989e95bfec5cc48cec5394dd
39a86ee9a5920acf1077becd27a6bbf4678872a9
51399 F20101130_AAAFVZ bray_s_Page_081.pro
988920c8bd0a5443fb0aaf9d646a892a
196273a4f105bbf751a96528c66bdefafc352140
2421 F20101130_AAAFXC bray_s_Page_006.txt
b28ba58e73de47e9cbe5869ed081ba5a
2c3d9dcc7cb843e0d60e5132f073298725cd60e1
6727 F20101130_AAAGCJ bray_s_Page_087thm.jpg
2e8228fbdc030db8289f03e2133b36f9
b49039bcfdc6b93d8d4ceb63154a714d779b655b
6659 F20101130_AAAGBV bray_s_Page_074thm.jpg
e367bd56d77faddad4f77ad8053fb662
b99678b6714a84731f073bdad432f90655782449
48762 F20101130_AAAFWP bray_s_Page_109.pro
94ee64b788ac8f033b8e8bf3cb0f6917
f6e5339d79e82ece00f09a54be24ea6dcf359c5b
3104 F20101130_AAAFXD bray_s_Page_007.txt
cf2404d4cf6774a254238c1ac5a6fc15
f7bbd1ca7d12fee43d4be1e54dea68b3e394aa18
23752 F20101130_AAAGCK bray_s_Page_088.QC.jpg
d632baf0545238e4e06e58292df43e18
67a936373b0979f0e9d3f75f6595d0931a4b519d
22919 F20101130_AAAGBW bray_s_Page_075.QC.jpg
fbea33d78f0c5fc3164549fbc6970503
767c26bff36d94e9af7e38886424a373c6a94c24
51258 F20101130_AAAFWQ bray_s_Page_113.pro
eef7549de3b0616776c08efaf7cc4abe
beaa635d816d27e4a127a4bb73d44f5e0cbaab4a
1897 F20101130_AAAFXE bray_s_Page_013.txt
270389047dd48e769e56531caa88ccae
2e99752869e5f199ea34760e4cf519d80c63cf1a
6399 F20101130_AAAGCL bray_s_Page_089thm.jpg
70b764112c318c8a14b89a59d2faa1f4
9c0b37b8f87c1264380ff336506f3c0b8fed0808
6437 F20101130_AAAGBX bray_s_Page_076thm.jpg
271872e609f8533b92ae0e2caf9e3f82
b084151b7996fc987dd2349132722753bec2ae28
61734 F20101130_AAAFWR bray_s_Page_115.pro
a8e83dbf11ffd073ec19d14e7a15ea47
f5b8b13267e9508e2533721de18f347e608641e0
1865 F20101130_AAAFXF bray_s_Page_015.txt
0f8c3680c8a8cab64b816e2b107f8b06
4c2dc2526f35c7e87f260a2cb822122a04b93ea6
22496 F20101130_AAAGDA bray_s_Page_109.QC.jpg
eb3a0b87a38de05148bd73a4a82c7e8e
abfa7e70c62bb724b6ff2a7f5df6b53bb952ac35
23475 F20101130_AAAGCM bray_s_Page_090.QC.jpg
c9ea29e1206a55d64e2c32d9f30103c5
89c942778b5846cea1d4a4af032506705e00045c
21638 F20101130_AAAGBY bray_s_Page_077.QC.jpg
8ac2376ce3c9e3d6f4b37dfbc911b72d
d86d5f603b9be6d147c001aae312a5e216aebd90
207 F20101130_AAAFXG bray_s_Page_018.txt
03b6337ecce4324c07efcecbb5d2b294
9cfbbbe39e5e747e23ae87a0bfffda5398e6fc48
6140 F20101130_AAAGDB bray_s_Page_109thm.jpg
d634d169af9023b2239d67ead90f9a5c
9de539cd67995c00b77423d9b1c354bc14ebeb59
6499 F20101130_AAAGCN bray_s_Page_090thm.jpg
77df10974c14ba75fb71c8b539a25e49
cb63890eb250b1b3579754b27dd17ef23aeb28e3
21699 F20101130_AAAGBZ bray_s_Page_078.QC.jpg
941291d88ccf3806dbc031c4af5458e5
7d2c178fe896f638a975d3e177ab5e98e1e10d0f
66235 F20101130_AAAFWS bray_s_Page_116.pro
3caad15014be88cd77c5ec7ca98206bb
6ea4f51093893b8b60901187fac76b5c78686901
178 F20101130_AAAFXH bray_s_Page_019.txt
0e222d88f61b99e2bb89d8ff26ed6999
da556b4ab8b5553c77b48d3ed5ab03989d1c8fed
4733 F20101130_AAAGDC bray_s_Page_112.QC.jpg
717867a0851601f77d3033d6788ac7c0
1a1f11cd20b06e8a78e982ae1ea421c314af9b47
7573 F20101130_AAAGCO bray_s_Page_091.QC.jpg
c58d1a240e5da3cac21a34e83faa1995
52824fc1bfaee2b695dc2eea11315c30a26ed7ec
59400 F20101130_AAAFWT bray_s_Page_118.pro
49b5ea2d33509dc98ee887c17ff7cac0
176863e2828bfea4a274d9640c8dab8035469bd5
F20101130_AAAFAA bray_s_Page_021.txt
8aee237744d351b0acc7af62d4da2e0c
59101655c2b43de1653131fcf890221b24613894
1919 F20101130_AAAFXI bray_s_Page_020.txt
9406a9cd79519617c5583b11c9ab0746
c1000aa3f0eb3b8bb8a9973006bde2e345d0bbfa
1699 F20101130_AAAGDD bray_s_Page_112thm.jpg
6b585ef0f7da7a21b13f04a474adad54
4c763c20c27e4b61745f29c9ef146277c52e612e
10010 F20101130_AAAGCP bray_s_Page_093.QC.jpg
e18199d64ddc14459b2d7a25a0be2228
669145a52f73dc826c40cf29ade42f0ba6747424
59522 F20101130_AAAFWU bray_s_Page_119.pro
03439058a94536667b92c324e8482750
d661515d05e7217997bc89192cace0c6ae7c6de2
73347 F20101130_AAAFAB bray_s_Page_092.jp2
cbcb4a20f44b9013bfb1e2dbc74f6587
49a270728e87feef1b478ece985e99e3d5f906d7
1982 F20101130_AAAFXJ bray_s_Page_022.txt
6beab7b7eecdc256ccb30b89b5d62453
4ea0218425e528fb4c58be7f296d309f96a994dc
21476 F20101130_AAAGDE bray_s_Page_113.QC.jpg
36ea56d0bf9d9002fbd3ca0228458569
066b1556614c7643c7b4f98c6293ab2e57c1c931
4643 F20101130_AAAGCQ bray_s_Page_095thm.jpg
9ec3e481b100c5d97a7c8289495682c5
283f8bf13dfd2b08e9a95f8de30feafdd884fada
65752 F20101130_AAAFWV bray_s_Page_120.pro
3a425f782f69cf7380e8c5e458554aa7
9214fa616629e1a85f4256b01b1f2a24537437bf
28299 F20101130_AAAFAC bray_s_Page_068.jpg
5db853dbe5511a8f792fc210b2f7c7e5
bb5fb9f5ab24b0ccdb06730aab58bf4a71dccf4c
F20101130_AAAFXK bray_s_Page_023.txt
8fcff285c6913a6707239f628a81bc58
40cb24d92cf1b5cd3786f55385bf2386dea586ee
25799 F20101130_AAAGDF bray_s_Page_114.QC.jpg
efa627ced58faa1a97973dd9ac3277b2
810371543b3d8f0cca88a263bb4451e25b0c55f5
4826 F20101130_AAAGCR bray_s_Page_098thm.jpg
fee105206df46981f6ed1d42031e82a7
556eddb91b09ae9ffb9e2f029fd9dd202f594940
59158 F20101130_AAAFWW bray_s_Page_122.pro
ded3f11306c061dbbba1638f1ea0814d
84eaade07f16aa2c063639d471185efe18b5ed52
6866 F20101130_AAAFAD bray_s_Page_120thm.jpg
26a9d9da9f275cf356dbc298f54feebb
389d4f3979c217b991be7f9fef042ffd67eea20b
F20101130_AAAFXL bray_s_Page_024.txt
2ce61f8fe4b7d1bbd25c29405bbd358a
78863aed8691a5fd5d25a502fceacb3095c0337a
24484 F20101130_AAAGDG bray_s_Page_115.QC.jpg
010bc04112522364728460dfbda2e83d
8fe56bb4d00aac219f011ef21cfba07c19f5b1d5
14799 F20101130_AAAGCS bray_s_Page_099.QC.jpg
5aca0e50f27019e22b11daaec7e46b86
7710aa7b58a2b45096175898d0b40d8e2cc8609a
59803 F20101130_AAAFWX bray_s_Page_123.pro
51d1a252f9f051832928e5588d3067c3
f6a835ca06274ce2bf1c2f968709d6cd7a01524b
65576 F20101130_AAAFAE bray_s_Page_005.jpg
a88d694f62667792c9245afa16fd54fe
db8c0e40dc26971a6d1230f7f709214c29629a3f
1977 F20101130_AAAFYA bray_s_Page_054.txt
b2a6153e74cb7e29dab37148fba088ae
1ce9ab7d0dd96babd66469cc5b9a144cc57be976
1971 F20101130_AAAFXM bray_s_Page_025.txt
25efceadb65976e721fc4ee659db5d44
d468d37051a5cf456d793674beef18d217483828
6840 F20101130_AAAGDH bray_s_Page_115thm.jpg
cbecd5f8efede93c927c5f9258bc1fd9
5bada31a59ab86227f74738e3b9f420a19337072
4349 F20101130_AAAGCT bray_s_Page_099thm.jpg
762e899dee56d1670b477bce1ed5fde2
984be4ba7d90c0cf740b56f5a156440be2a9a1cc
32751 F20101130_AAAFWY bray_s_Page_125.pro
9f85e039a6a92b64382a8bddd08674c0
adfa2d4f31a324421f8c696f39170f1b63c830db
6345 F20101130_AAAFAF bray_s_Page_022thm.jpg
87cc6f4af129bc096a15d8ee581960dd
cd9c43d7026a8dff6e29b2c12c0732d67e619359
870 F20101130_AAAFYB bray_s_Page_057.txt
3c9a0223672d971e8cc5c08ae8ff2699
9a072d966856240f600055b7a4302603af136f8d
F20101130_AAAFXN bray_s_Page_026.txt
0352a64cbed903ece2f8a5cc36564377
44f3c7dd1c2cc3bc3dc7cc40af4ea0ecb5af2e0f
6630 F20101130_AAAGDI bray_s_Page_119thm.jpg
6a236e45388aa3739a320915adfd78f3
4d8b75285352767fc621b4c3bea2227e6865b8f5
3895 F20101130_AAAGCU bray_s_Page_100thm.jpg
526debe4be4c2f078e9e01d7e92e48b1
6e7bc635befb3c15ad82f7e7f043e0caf99368b3
489 F20101130_AAAFWZ bray_s_Page_001.txt
db59b70a4b28d13822a42bca84c83deb
1edfe5285d98bfdbd3e85ceb357ceb0251fb49c3
21316 F20101130_AAAFAG bray_s_Page_070.QC.jpg
a721e2c45d75cfbbb6f6929fb2c731fc
e63beea6fac22da75a917a791199be25c3f991d7
872 F20101130_AAAFYC bray_s_Page_061.txt
1f6560321ad6c289f4032b2e523e227c
ef874bf084e4f6835c900acb11d12da664fcf431
F20101130_AAAFXO bray_s_Page_028.txt
e97aad1bf6ea05fbf05f32342763eb2f
5e3587082970c601d415dd8216e3e7d4ad83d5b4
26946 F20101130_AAAGDJ bray_s_Page_121.QC.jpg
2dda1b427b3ae576769f88deeb50d804
ba4d6f822419a7b8925d7a4cac74812c09946340
15404 F20101130_AAAGCV bray_s_Page_101.QC.jpg
9704e2cb9125cb5c0cf4502e364eb8ce
975f8d4130e022d36fdcf37ef4920cadd03468e0
22536 F20101130_AAAFAH bray_s_Page_064.pro
34b51b74cf375a6d84ca3b36f0300b81
d42c61627de13823247b6fa87fccc5fd1cd173d9
699 F20101130_AAAFYD bray_s_Page_065.txt
091610318e2d4ed7f35e09a2431eedf2
584fbedc2ec9f1da46ed35db017d0b217d340798
1909 F20101130_AAAFXP bray_s_Page_030.txt
24b64cbe81c2765b9a6a28b187201cef
90ba67d3584b35d5c70bfaf2332c6503b2917fa9
24051 F20101130_AAAGDK bray_s_Page_122.QC.jpg
6de099730bcdad03ea94d876fc90b26c
ded6d7c8c4ded071c828d3ea64fc96b7c3416465
5487 F20101130_AAAGCW bray_s_Page_102thm.jpg
540c4342e0103b1f719c5ed9d625f9e7
9211a85a33391b847235a8fe43d7b83c14820696
72059 F20101130_AAAFAI bray_s_Page_074.jpg
9f9b22a0f000be45821c505b6daf761f
ed2ae046ce5f1272c405a4f574913f115baaf7e2
307 F20101130_AAAFYE bray_s_Page_068.txt
11960d8c75910d39b3a99099b9f7ae04
a81a28827d73f00754423dbfdd010f82045dbad9
1878 F20101130_AAAFXQ bray_s_Page_032.txt
61df9853c716041421a23e8a85abe33c
6490c64ce63ab1ab20f96c89861ea0eafaeebbc2
6827 F20101130_AAAGDL bray_s_Page_122thm.jpg
fd4ff1c73ff6ae09161f0a7a6dda9ccf
68dd1b95cb8ec464c3035faeb8c3cf5ab264ee98
7180 F20101130_AAAGCX bray_s_Page_104.QC.jpg
a8066712d4fba221e0b99768307de27d
7814f45c19e0576d73a1ad7c8b19c241c6b8d976
23915 F20101130_AAAFAJ bray_s_Page_014.QC.jpg
4061e8f129414546b1e318286bfc7a23
12015768921778e3a9ba897fec15e1fdacd37ecf
1986 F20101130_AAAFYF bray_s_Page_071.txt
b6ca3573d70f77046df27afce7b34db4
d94aa99f6527ca231b778ee1c102ef01e6ccee98
2081 F20101130_AAAFXR bray_s_Page_035.txt
e7fa4db932d764077664da1cf0c000de
8ddd213a254267dd049bebe9bc67af61fa2e2c42
24192 F20101130_AAAGDM bray_s_Page_123.QC.jpg
bc4979620a6b99b72df7d3b3ab803ff3
28a1adb8a41769b4370f107e499be2c765080f9c
3850 F20101130_AAAGCY bray_s_Page_105thm.jpg
6629fdd23eff1641a3d957a8475029aa
327ea126e2b248c314d28f39fbaa38fddf829dd6
109832 F20101130_AAAFAK bray_s_Page_054.jp2
419b79af801aead26e8f8e00c9889754
1a3b57f6c61bdd7e9e81d5d34d00a022cedad249
1967 F20101130_AAAFYG bray_s_Page_074.txt
e8048979419f45016bd86244381fbb55
35feaa7c6fe3e8986e3ecd37822b141d86c63af7
1891 F20101130_AAAFXS bray_s_Page_036.txt
817941ff9814c6ee445e0a0e0e358385
501a8e9e3f77d6ce10dfb6ab320bcd16c827baec
6683 F20101130_AAAGDN bray_s_Page_123thm.jpg
4fb2b000a2e0ed11733425c060c5bebe
4bef3ed39b7f546eb2e29ba8df0ab9a2e3979405
5996 F20101130_AAAGCZ bray_s_Page_108thm.jpg
2abc783ae2e04f0ff400bc5e1d0c75ee
2203ba02bff9a1d9530678e73c066e64c71661dc
22750 F20101130_AAAFAL bray_s_Page_043.QC.jpg
6e76b93479f3adb47df595db2a368466
198f2e8464864248539932c90046fe520e10b1fa
1947 F20101130_AAAFYH bray_s_Page_075.txt
295b7d20cc73127de17796e6c982f41e
ac2fc6c0c004c99d3a92142287417318fe483147
11296 F20101130_AAAGDO bray_s_Page_124.QC.jpg
c9c3f1812c584fe46271284b32b7a38f
dcdaca008205a3a80e2b499bbec4c9e48092195b
1948 F20101130_AAAFAM bray_s_Page_089.txt
b91df1218dd7128840a5b09955faaa46
f42097a75937695beb812b2cfb5b43d67cf13cd0
1942 F20101130_AAAFYI bray_s_Page_076.txt
ed8f9923d0de7b8ce39d0e7c6e37c093
345e45e6781b213b56d0ca0054308e6e80884a0e
2832 F20101130_AAAFXT bray_s_Page_038.txt
7fe563fdeab1c4d0e1392c870c03adb9
0ccf82ddca4d24acd6c23e2460a84259e1221758
1153 F20101130_AAAFBA bray_s_Page_066.txt
addd5ee5556e0f430b2ac47dd2c5c1c3
9c700a5fbee206b6acb626d724dd53619e2d443d
143442 F20101130_AAAGDP UFE0010945_00001.mets
57182766eadfb97b462800ecbcb334b9
cbae5f7e22dabef58e5fec6dbabb230f698f22cf
72117 F20101130_AAAFAN bray_s_Page_054.jpg
87b3f46ce6dbb827c6c0d360f22c2416
7ef69ea6c3ffca0666c8f233365061cf2aac3065
1803 F20101130_AAAFYJ bray_s_Page_078.txt
f5c949ee3d3f259ece9e26a78dc636ec
b112a86d3ded1ac6f402110b90c9371afb3c5115
2640 F20101130_AAAFXU bray_s_Page_040.txt
274398f7b083891b3141450022165c8b
6fc2100a31469a2ab1c41b9cca208585f1675237
F20101130_AAAFBB bray_s_Page_087.tif
18d0bf1ce2d7659dd2ff432e9c279ddc
70a9c66451a2f373e91c14501d867e78e799fcf8
43488 F20101130_AAAFAO bray_s_Page_098.pro
7430f6e66280454f03f2b0175fce4bfd
8e7f3840a5e23fdaf166851c2e5f85a9fecfea10
1983 F20101130_AAAFYK bray_s_Page_079.txt
19c6a1d9c34d6c9674ba6a5ed63df927
dcf2851e9634b4eadfaef2bf6538272bcd45d5ba
1941 F20101130_AAAFXV bray_s_Page_043.txt
50aba3397e7da6c89ff7f7f3d64c32bf
d0a55c1023fe9fc27c9ffa167bdb47261c5c009e
1844 F20101130_AAAFBC bray_s_Page_042.txt
09d91a569b83386c1f10618fcf5d9b3e
04e8ac584c2548520d5241cea0aceeb71b0607c0
48516 F20101130_AAAFAP bray_s_Page_034.pro
37bbd43cde95d1d87cef18b4a66eaf03
040fda12195506b8e6629514ff0e17008edc9665
1845 F20101130_AAAFYL bray_s_Page_080.txt
5b57cfb07ed297381d6b230ba8e5a277
4733fcaae8e23c10aecd414f0a814e30aec683ab
1933 F20101130_AAAFXW bray_s_Page_044.txt
5f015d95d5f941f706076793b0564788
e31f0c117e5b527e61576a9f0a96c4fbbbd43e70
27365 F20101130_AAAFBD bray_s_Page_093.jpg
e50d1778bf4b2d1f641ee1149f4c4fe3
fbe848d360a4e1f02db53b6ee117e554527c1a32
17578 F20101130_AAAFAQ bray_s_Page_009.QC.jpg
954aabcfc9da395a15b8867ed4344064
897fa94ee977d20b77f457629e9e8ab08bf77b4c
1074 F20101130_AAAFZA bray_s_Page_106.txt
bd3c2e4b625c9af6d2c69c4c0d7ab731
1a9632aea9d28a1fe49d02ab47aa987c5f12c85a
F20101130_AAAFYM bray_s_Page_081.txt
11fb0a53a8ec57f8f73fe2c0742d4cab
be98654b6ea5edbf3add9a23d065f96c6385e98e
F20101130_AAAFXX bray_s_Page_046.txt
2d83074c332cadf4d8f2cf4cbe0b2a72
8895f1d5825f3397570b651c213c3ad5fa227235
68413 F20101130_AAAFBE bray_s_Page_048.jpg
91ceec2003fa414ffc87b2f4fe51c628
3b85b7c52810707f6fcbe57fe61e1cceec54a91e


Permanent Link: http://ufdc.ufl.edu/UFE0010945/00001

Material Information

Title: Interactions between Plants and Soil Microbes in Florida Communities: Implications for Invasion and Ecosystem Ecology
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010945:00001

Permanent Link: http://ufdc.ufl.edu/UFE0010945/00001

Material Information

Title: Interactions between Plants and Soil Microbes in Florida Communities: Implications for Invasion and Ecosystem Ecology
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010945:00001


This item has the following downloads:


Full Text












INTERACTIONS BETWEEN PLANTS AND SOIL MICROBES IN FLORIDA
COMMUNITIES: IMPLICATIONS FOR INVASION AND ECOSYSTEM ECOLOGY















By

SARAH RENEE BRAY


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

UNIVERSITY OF FLORIDA


2005





























Copyright 2005

by

Sarah Renee Bray















ACKNOWLEDGMENTS

Many people and organizations have supported my research and this dissertation

would not have completed without their help. Kaoru Kitajima, my advisor, has allowed

me to pursue my own research interests even as they diverged from her own. My

committee members, Alison Fox, Doria Gordon, Michelle Mack, and David Sylvia, and

the Plant Ecology Group have been wonderful sounding boards, colleagues, and editors.

Michelle Burch, Robert Bilbao, Jason Alexander, Debbie Renelus, Jennie DeMarco, Abid

Al-Agely, Alex Reinstein, Thai Van, Ellen Dickstein and Grace Crummer have helped in

the field and lab. The Botany Department staff made sure my bills were paid and

deadlines were met. The National Science Foundation, the Department of Environmental

Protection, the Florida Exotic Pest Plant Council, and the College of Liberal Arts and

Sciences have all provided me funding to complete my research. The real reason that I

have been able to finish my doctorate was the unwavering support of my friends and

family. I thank Sara Smith, Melissa "Fig" Bonfig, Amy Miller Jenkins, Silvia Alvarez,

Erika Gubrium, and Aline Gubrium for the conversations that encouraged me to continue

when I was frustrated. I thank my parents, Steve and Susan Bray, for always encouraging

me to achieve my goals and picking me up when I fell short. I thank my brother and

sister-in-law, Seth and Maureen Bray, for reminding me of the fun things in life. Most of

all, I thank my husband, Anthony Allen. He spent many hours working in the lab and

field with me, supported us financially, and never lost faith in me even when I doubted

my abilities. Without him, I could not have finished this dissertation.
















TABLE OF CONTENTS

page

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

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

LIST OF FIGURES ............. ................... ............ .......... ................ viii

ABSTRACT ........ .............. ............. ...... ...................... ix

CHAPTER

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

2 MYCORRHIZAE DIFFERENTIALLY ALTER GROWTH, PHYSIOLOGY
AND COMPETITIVE ABILITY OF AN INVASIVE SHRUB ................................10

Intro du action ...................................... ................................................ 10
M materials and M methods ....................................................................... .................. 12
Species .................. ................ ......... .... .. .................................12
Experiment 1: Effects of Soil-P, Light and Inoculum Source on Growth and
Allocation...................... .... .. ......... ..... ........ 13
Experiment 2: Effects of Mycorrhizae on Seedling Competition .....................16
S statistical A n aly ses........... .......................................................... .. .... .. .. ... 18
Results......................... ....................... .............. ...... ........... 19
Experiment 1: Effect of Light, Soil-P and Inoculum Type ..............................19
Experiment 2: Effects of Mycorrhizae on Seedling Competition .....................21
D iscu ssion ............... .. .............. .................................. ............... ..... 2 2
Effect of Inoculum Source on Ardisia.......................... .............................. 22
C om petitive Interactions ........................ .. ...... ............... ....24
Implications of Effects of Mycorrhizae on Exotic Species Invasion ..................25

3 SOIL MICROBIAL COMMUNITY STRUCTURE AND FUNCTION IN
FLORIDA PLANT COMMUNITIES PRONE TO NON-NATIVE PLANT
IN V A SIO N ...................................... ................................................. 32

In tro du ctio n ...................................... ................................................ 3 2
M eth o d s ............... ................... ....................... ................ 3 5
Species, Sites, and Sam pling .......................................................... ..................35
Microbial Community Composition and Nutrient Analysis ............................37









Statistical A n aly ses........... ...... ........................................................ ..........3 9
R e su lts .........................................................................................................................4 0
D iscu ssion ............................... ......................... ......................... 43
Habitat Controls of Microbial Community Composition.............................. 43
Alteration of Microbial Communities by Invasion ............................. .......45
C o n clu sio n s ..................................................... ................ 4 6

4 LINKS BETWEEN LITTER QUALITY, DECOMPOSITION AND
MICROBIAL COMMUNITY COMPOSITION ON NATIVE AND NON-
N A T IV E PL A N T L IT TE R .............................................................. .....................60

Introdu action ............. ..................................................................................... 60
M methods .................................... ..... ...... ........................ ........ 63
Litter Collection and Experim ental Design......................................................63
Phospholipid Fatty A cid A nalysis.................................... ....................... 65
Statistical A analysis ........................ .............. ...........................66
R e su lts .................. .................................................................... ..................................6 9
D ecom position of L itter ............................................ ....................................69
Microbial Community Composition.................... .......................71
Linking Litter Quality, Decomposition and Microbial Communities ...............74
D discussion .................................. .. ... .. ... .. ..... ............ ...... ..... 74
Factors Controlling Microbial Community Composition ..................................75
Factors Controlling Decomposition Rate ................................. ................ 78
Im plications for Invasion.......................................................... ............... 79
C o n c lu sio n s ................................................................................................... 8 0

5 CON CLU SION S .................................. .. .......... .. .............98

L IST O F R E FE R E N C E S ....................................................................... .................... 103

BIOGRAPHICAL SKETCH ........................................................................115





















v
















LIST OF TABLES


Table p

2-1 ANOVA summarizing the effects of light and soil on LAR and R/S in
ex p erim e n t 1 ....................................................... ................ 2 8

2-2 Means of leaf area ratio (LAR) and root:shoot ratios (R/S) from all soil-P levels
in ex p erim en t 1 .................................................... ................ 2 8

2-3 Comparison of light saturated net photosynthesis rate (Amax) and dark
respiration under moderate vs. low light treatments ............................................ 28

2-4 ANOVA summary of the effects of species, competition, and mycorrhizae on
RGR, LAR, R/S, and colonization rates from experiment 1 ....................................29

3-1 Location, mean annual temperature, mean annual rainfall, soil type and
dominant vegetation of the three sites in each habitat type................ .......... 48

3-2 Soil characteristics of the five habitats examined ............................... ...............49

3-3 Relationship of total PLFA and fungal:bacterial ratio with % moisture, %C, %N
and C:N using quadratic fit for total PLFA and simple linear regression for
fungal : biom ass ratio ......................... ........ ........... ................49

3-4 ANOVA results for the effects of habitat, invasion-status and their interaction on
the relative representation of Gram and Gram + PLFAs. ....................................49

3-5 Correlation between oN, %C, and %moisture and the relative representation of
Gram and Gram + PLFA s .............................................................................. 49

3-6 Results of MANOVA for the effects of habitat, invasion and their interaction on
principle component axes 1 and 2 scores from the ordination of 23 PLFAs. .........50

3-7 Loadings for the first two axes in a principle components analysis of 23
common PLFAs extracted from soil samples .................................. ............... 50

3-8 Correlations (r) of soil characteristics with the first two axes from the ordination
of 2 3 P L F A s. ......... .. ............. ............................. ................... ..................5 1









3-9 The results of MANOVA for the effects of habitat, invasion and their interaction
on principle component axes 1 and 2 scores derived from an ordination of
metabolic activity of soil microbes on 95 substrates...............................................51

4-1 Characteristics of plant species whose litter was used in this experiment ..............82

4-2 Mean and SE for decomposition rates of native and non-native litters ..................84

4-3 Result of simple linear regression for log-transformed decomposition constants
against individual variables of litter chemical composition................................ 84

4-4 Factor loadings for individual variables that contribute to litter quality on the
leaf chem istry PC A axis 1 .............................................. .............................. 85

4-5 Factor loadings of individual PLFAs upon the first two main axes of PCA (PC1
and PC2) and CCA (CC1 and CC2) ...................................................................... 85

4-6 MANOVA summarizing the effects of time, litter species and litter species*time
on PCA and CCA axes 1 and 2 scores from ordination of all samples....................86

4-7 Loadings for the first two PCA axes for PCA run separately for individual
sam pling dates (t = 1, 2, and 5). ........................................ ......................... 86

4-8 Correlation between each litter quality variable and the amount of PLFA (nmol)
indicating monounsaturated Gram bacteria at times 1, 2, and 5 .........................87

4-9 Correlation between each litter quality variable and the amount of PLFA (nmol)
indicating branched Gram+ fatty acids at times 1, 2, and 5...................................87

4-10 Correlation between each litter quality variable and the amount of PLFA (nmol)
indicating fungal fatty acid (18:2co6c) at times 1, 2, and 5 ......................................88

4-11 Correlation between each litter quality variable and ratio fungal : bacterial
PL FA s at tim es 1, 2, and 5 .............................................. ............................. 88

4-12 Correlation between each litter quality variable and the total PLFA (nmol) at
tim e s 1, 2 an d 5 .................................................................... 8 9

4-13 Correlation between each litter quality variable and the amount of PLFA (nmol)
indicating cyclopropyl Gram- bacteria at time 5..................................... ........... 89

4-14 Coefficient of determination (r2) of abundance of PLFAs for microbial
functional groups with leaf chemistry scores at times 1, 2, and 5..........................90

4-15 Coefficient of determination (r2) of correlations of microbial community PCA
axes vs.decomposition rate (k), and leaf chemistry axis (from ordination of
initial litter quality) vs. microbial community PCA axes......................................90















LIST OF FIGURES


Figure p

1-1 Conceptual diagram of interactions between soil microbes and plants..................9

2-1 Response of Ardisia to light and inoculum type at 5 mg kg-1 P.............................30

2-2 Response ofArdisia and Prunus to heterospecific or conspecific competition
and mycorrhizal status.................... ............. ................. 1

3-1 Mean soil microbial community total PLFA (nmol g-1 + S.D.) for five habitats
prone to invasion by five non-native species. .................................. .................52

3-2 Fungal:bacterial ratios (+ S.D.) for five habitats prone to invasion by five non-
n ativ e sp ecies......................................................................................... .... 5 3

3-3 Relative representation (% of total nmoles extracted) of PLFAs across habitats
and invasion-status ......................... ......... ..... ........ .. .............. 54

3-4 Mean (+ S.D.) principle components scores by habitat and invasion-status from
ordination of 23 most common microbial PLFAs found in soil samples ...............55

3-5 Mean (+ S.D.) relative abundance of 5 fatty acids with high loadings on the
principle com ponent axes......... ............................ ........................ ............... 56

3-6 Mean (+ S.D.) relative abundance of PLFAs across habitats and invasion-status... 57

3-7 Mean (+ SD) principle components scores by habitat and invasion-status from
the ordination of metabolism of 95 carbon sources ............................ ............... 59

4-1 Mean + SE of the percent mass remaining of 20 species...................................91

4-2 Mean + SE for nitrogen in litter of 20 plant species at six collection dates............92

4-3 Ordinations of the 17 most common microbial PLFA found on 11 litter species
from three sam pling dates ............................................... ............................. 93

4-4 PCA and CCA analyses for 17 most common microbial PLFA and litter quality
variables run separately for individual sampling dates ........................................ 95















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

INTERACTIONS BETWEEN PLANTS AND SOIL MICROBES IN FLORIDA
COMMUNITIES: IMPLICATIONS FOR INVASION AND ECOSYSTEM ECOLOGY

By

Sarah Renee Bray

August 2005

Chair: Kaoru Kitajima
Major Department: Botany

Among ecologists there is an increasing awareness and interest in the role of soil

microbes in the distribution of plants and functioning of ecosystems. This dissertation

relates soil microbial community composition to plant growth, habitat type, and

decomposition with particular emphasis on invasive plants.

I examined growth, physiology and competitive ability of an invasive shrub,

Ardisia crenata, in two greenhouse experiments. When grown singly, relative growth

rates (RGR) and leaf area ratio (LAR) were higher for seedlings inoculated with

mycorrhizal fungi isolated from Ardisia roots than those inoculated with single-spore

isolates and nonmycorrhizal controls. In a second experiment, Ardisia was grown with a

conspecific or heterospecific (Prunus caroliniana) competitor. While neither identity of

competitor nor mycorrhizal status had a great effect on Ardisia growth, Prunus growth

was significantly depressed in competition with Ardisia in the absence of mycorrhizal

fungi.









In chapter 2, I examined soil microbial communities from five different habitats

prone to invasion by an invasive plant using phospholipid fatty acids (PLFA) and Biolog

substrate utilization. Habitat type had the largest effect on microbial community

composition. Moisture content of soils and, to a lesser extent, carbon and nitrogen

contents appeared to be driving differences in biomarker PLFAs. Although the largest

differences in soil microbial community composition were found among habitats,

invasion altered microbial community composition within habitats.

In chapter 3, I placed litters of 20 native and non-native plant species of varying

decomposability in a common site and quantified their decomposition. The composition

of microbial communities on 11 of the litters was examined by PLFA at 28, 56 and 238

days. Microbial communities at the early (low moisture) sampling dates were more

similar to one another than to the late (high moisture) sampling date. In addition to

moisture effects, litter quality had a significant effect on microbial community

composition. Both decomposition and microbial community composition were correlated

with leaf chemistry. The best single predictor of decomposition rate was microbial

community composition.

These results suggest that plant-microbial interactions are important in plant

invasion, and explicit examination of a potential positive feedback on invasion through

the microbial community should be further explored.














CHAPTER 1
INTRODUCTION

Historically, most ecologists have studied aboveground macroorganisms even

though they have long recognized that processes and organisms aboveground can modify

processes and species distributions belowground and vice versa (Jenny 1941, Brown

1958, Garrett 1963, Hudson 1968, Janos 1980). Much of this aboveground bias has been

due to the difficulty of examining organisms and tissues in the soil matrix. Traditionally,

culture-based techniques have been used to identify and isolate soil microbes. However,

only a small proportion of soil microbes (<1%) are believed to be culturable (Atlas and

Bartha 1998), and many researchers have treated the soil as a "black box." The advent of

new techniques to study belowground microorganisms using molecular markers has led

to a "renaissance" of sorts in the study of interactions of soil microorganisms with

aboveground macrophytes, allowing ecologists to proceed "through a ped darkly"

(Coleman 1985).

Soil microbes have direct and indirect interactions with plants (Figure 1-1).

Symbiotic microbes directly affect the host plant's fitness, resulting in alteration of plant

species abundance and distribution. Pathogenic microbes have negative effects on

individual plant fitness. Due to pathogens' high host-specificity, these microbes may be

responsible for density-dependant plant distributions predicted from the Janzen-Connell

hypothesis (e.g., Packer and Clay 2000). Mutualistic bacteria and fungi (e.g., nitrogen-

fixing bacteria and mycorrhizal fungi) also directly interact with their hosts receiving

carbon substrates directly from their host and supplying the plant with nutrients. Free-









living saprophytic fungi and bacteria in the soil interact indirectly with plants by

decomposing their senescent tissue and replenishing inorganic soil nutrients. Therefore,

whereas mutualistic soil microbes may be a proximal source of nutrients, free-living

microbes supply the ultimate source of nutrients.

Availability of nutrients in the soil is one of the primary controls of productivity in

ecosystems, and competition for those limiting resources is a major control of community

composition. Plant species that can draw down the most limiting resource below the

level at which other plants can survive (i.e., the plant with the lowest R*) should be the

best competitor within a given system and will displace other species (Tilman 1982). To

become a superior competitor for this limiting resource, the plant must allocate a greater

proportion of its biomass to the acquisition, conservation, and/or efficient use of that

resource. Such biomass investment limits the ability of the plant to compete for other

limiting resources. The resultant trade-off means that species with different allocation

patterns or suites of functional traits related to resource acquisition will be superior

competitors based upon resource availability in the habitat (Chapin 1980, Tilman 1988,

Chapin and Aerts 2000).

Competition for resources by plants may be modified by interactions with free-

living and symbiotic microbes. Mycorrhizal fungi increase the volume of soil a plant can

access, increasing their ability to acquire phosphorus. Increasing the mycorrhizal fungal

richness leads to a greater use of soil phosphorus by the plant community, greater

productivity and plant species richness (van der Heijden et al. 1998). Not all species of

mycorrhizal fungi benefit all plants equally, however, as carbon cost of some fungal

partners exceeds the benefit in increased phosphorus nutrition (Bethlenfalvay et al. 1982,









Smith and Smith 1996, Johnson et al. 1997, Graham and Eissenstat 1998). Such

differential response of plants to the species of mycorrhizal fungi present may modify the

competitive hierarchy (Moora and Zobel 1996, Smith et al. 1999). If dominant, superior

competitors harbor fungi parasitic to themselves, their competitive ability should be

decreased. Work by Bever et al. (1996) has, in fact, shown that there are host-specific

sporulation rates of mycorrhizal fungi. They found that these differential rates of

sporulation maintained diversity through negative feedback on dominant plants as the

fungus that preferentially sporulated with the dominant plant was also least beneficial to

it (Bever 2002). Conversely, dominance may be achieved through the presence of

mutualistic microbes or the absence or resistance to pathogenic microbes (Bever et al.

1997).

While studies of mycorrhizal fungi have demonstrated that interactions between

fungi and plants influence the distribution and abundance of both groups, our knowledge

and understanding of the interactions between plants and free-living microbes are more

limited. Because of the difficulty in examining non-culturable soil microbes, ecologists

have generally left this community as a "black box," examining the effect of soil

communities on plant growth without identifying the agents (pathogenic, mutualistic or

free-living) responsible. In these types of studies, plants are grown in a soil for an

extended period of time to allow for the soil community differentiation over time due to

plant inputs. Plants are then grown in their own or another species' soil. Some studies

have shown that plant growth is higher in their own soil (a "home-soil" advantage) while

others have shown the reverse (Bever 1994, Callaway et al. 2004). This would suggest









the possibility for negative or positive regulation of plant species density through the soil

community although those agents responsible for the regulation are unknown.

There has been an increased interest in moving away from the "black box" and

examining what free-living microbes are actually responding to different plant species

and what effects changes in free-living microbes might be. The advent of molecular

techniques for examining soil microbes, in particular the use of phospholipid fatty acids

(PLFA) and nucleic acids, has resulted in a large number of studies of soil microbial

community composition (Tiedje et al. 1999). Techniques based upon DNA and RNA

have greater power in distinguishing microbial species than do PLFA techniques.

However, due to the specificity of primers and probes used, nucleic acid based studies

tend to examine a phylogentic subset of soil microbes. Additionally, as all nucleic acid

techniques are dependant on PCR, only presence/absence, rather than quantitative

analysis, is possible. While PLFAs are limited in their ability to identify species, and

thus can never truly answer questions about diversity, they can be used to identify

different functional groups of microbes and do allow for a broad, quantitative

examination of the entire soil microbial community. Thus the use of both types of

techniques will help to advance understanding of soil microbial community composition.

The field of soil ecology is still very much in an exploratory phase; however, some

trends are beginning to emerge. Different environmental factors such as temperature,

moisture, and soil nutrients, and different plant species support different microbial

communities (Bossio and Scow 1998, Bossio et al. 1998, Staddon et al. 1998, Pennanen

et al. 1999, Priha and Smolander 1999, Myers et al. 2001). Less research has been

directed at the effects of free-living microbes on the plant community. Because microbes









decompose organic matter, releaing inorganic nutrients, microbial influence on nutrient

availability may result in plant community changes. As decomposition rates and

microbial communities are known to differ among plant communities, microbial

community composition may be ultimately responsible for differences in decomposition

rate.

These various plant-microbial interactions have for the most part been documented

in systems with well-established associations of plant species. Plant community

composition and diversity, however, are changing at global and local scales. Plant

invasions, in effect, offer a "natural" experiment in which to examine the effect of

changing species composition on plant-microbial interactions. Plant invasions also offer

a chance to examine the interaction of native microbes and non-native plants that likely

do not share an evolutionary history.

The goal of this dissertation is to use the natural experiment of non-native plant

invasion to further our understanding of plant controls on symbiotic and free-living

microbial community composition and vice versa. Such studies will also shed light on

the functional reason for dominance of invasive plants. My dissertation is divided into

three sections examining the interaction of mycorrhizal fungi with an invasive shrub,

differences between soil microbial communities from different habitats with and without

plant invasion, and the links between native and non-native plant litter, microbial

community composition and decomposition.

In Chapter 2, I study the effect of native mycorrhizal fungi on the growth,

physiology and competitive ability of an invasive shrub Ardisia crenata in two

experiments. I hypothesized that Ardisia's successful invasion may be in part due to its









ability to form beneficial partnerships with native mycorrhizal fungi it encounters in the

areas it invades. This was tested in a greenhouse experiment in which Ardisia was grown

in sterile soil, with one of two native single fungal species, or with a mix of mycorrhizal

fungi isolated from Ardisia roots. I also hypothesized that Ardisia performs better in

heterospecific competition than conspecific competition especially when mycorrhizae are

present to modify resource competition. This was tested in a second greenhouse

experiment in which Ardisia was grown with a single conspecific or a native

heterospecific competitor, Prunus carolinana, with or without mycorrhizae.

As the results from Chapter 2 indicated that Ardisia's growth was improved by the

presence of mycorrhizal fungi found naturally occurring in its roots, I wanted to

determine if invasion could result in the alteration of the soil microbial community. If

invaders alter the composition of the soil microbial community to their advantage, there

might be potential for positive feedback on the invasion. In a natural survey of five

habitats prone to invasion by non-native plants, I sought to determine if plant invasion

predictably alters soil microbial community composition and function as examined with

phospholipid fatty acids (PLFA) and Biolog substrate utilization. Three alternative

hypotheses could explain soil microbial community composition at the landscape level.

One, habitat characteristics such as soil type, soil nutrients and moisture contents,

temperature, and native plant community composition could have such a strong control

on soil microbial community composition that even dense invasions may not result in a

change in soil microbial community composition. Alternatively, dense invasions may

alter the microenvironment experienced by soil microbes enough to alter the composition

of their communities. There may also be an intermediate hypothesis where some, but not









all invaders alter soil microbial community composition as a result of functional

characteristics of the invader.

The results of Chapter 3 suggested that while there were larger differences between

habitats than within habitats due to invasion status, invasion does alter microbial

community composition. The final chapter (Chapter 4), I sought to determine if the

chemical composition of litter determines the microbial community present and how

microbial community composition is related to decomposition. I hypothesized that more

chemically recalcitrant litter (i.e., litter high in lignin, C:N, and low N:lignin) would have

a higher proportion of fungi and would decompose more slowly. Litter with highly labile

chemistry (i.e., litter high in non-fiber carbon fractions and nitrogen and low in lignin)

would decompose quickly as a result of higher total microbial biomass and a higher

proportion of bacteria. To test these hypotheses, I created a "common garden"

experiment in which litter of varying litter quality from native and non-native plant

species were allowed to decompose for the period of one year. The composition of

microbes on the plant litter was examined for 11 of the species at 28, 57, and 238 days.

Composition of microbial communities varied with time; however, microbial

communities on leaf litter of similar quality were more similar to one another than to

microbial communities on litter of very different quality. I also found the microbial

community composition was the best predictor of decomposition rate of all factors

examined.

Together these studies begin to paint a picture of how plant and microbial

communities may feedback on one another's composition. This work additionally shows

that plant-microbial interactions may be an important pathway for the success of invasive






8


plants in new ranges. Further studies should explicitly examine the potential positive

feedbacks on invasion through soil microbial communities.
















Plant litter


N-fixing bacteria


Free-living microbes


Figure 1-1: Conceptual diagram of interactions between soil microbes and plants.















CHAPTER 2
MYCORRHIZAE DIFFERENTIALLY ALTER GROWTH, PHYSIOLOGY AND
COMPETITIVE ABILITY OF AN INVASIVE SHRUB

Introduction

Plants must overcome obstacles of dispersal, abiotic conditions, and competition

with existing species in order to colonize and establish in a new geographical locality.1

Arbuscular mycorrhizal fungi can aid or hinder the establishment of a new species by

ameliorating or intensifying the abiotic stresses encountered in the new range. Arbuscular

mycorrhizae (AM) may improve phosphorus (P) availability and enhance leaf

photosynthetic rates and growth rates of the hosts (Siqueira et al. 1998, Sharma and

Adholeya 2000). Due to improved P nutrition, mycorrhizal plants may allocate

proportionally less to roots while increasing leaf area ratio and specific leaf area (Berta et

al. 1995, Gavito et al. 2000, Lovelock et al. 1996, Son and Smith 1988). Change in

allocation patterns of the host, in turn, may affect its interaction with neighboring plants

for light and soil nutrients. Because the response of plants to AM is dependant on both

soil-P levels and light availability (Gavito et al. 2000, Graham et al. 1997, Peng et al.

1993), effects of AM fungi on the plant invasion process must depend on these abiotic

factors.

Although AM can infect a wide range of hosts from various geographical localities,

the responses of host plants to mycorrhizae vary greatly depending on the combination of

1 The information in this chapter was published in: Bray, S.R., K. Kitajima, and D.M. Sylvia. 2003.
Mycorrhizae differentially alter growth, physiology, and competitive ability of an invasive shrub.
Ecological Applications 13: 565-574 and is used here with the permission of the Ecological Society of
America.









plant and fungus genotypes (Johnson et al. 1997, Smith and Smith 1996). Different

fungal genotypes can have positive, negative or little effect on the growth of the same

host species (Boerner 1990, Monzon and Azcon 1996, van der Heijden and Kuyper

2001), because AM may differ in their ability to infect a given host, efficiency of P

transferred to the host, carbon demand, soil adaptation, and host compatibility (Graham et

al. 1996, Johnson 1993, Johnson et al. 1997, Monzon and Azcon 1996). Thus,

assessment of the effects of AM on plant invasion must consider the genotype and source

of fungal isolates.

Mycorrhizal fungi may alter competitive interactions between invading and local

plants. Although AM have been largely ignored as a mediator of plant invasion

(Richardson et al. 2000), they have been shown to increase the growth of an invasive

plant species over natives and accelerate the process of invasion in a grassland ecosystem

(Marler et al. 1999). More generally, differences in competitive ability under the

influence of mycorrhizae can alter community composition by favoring mycorrhizal-

responsive, inferior competitors (Hartnett et al. 1993, Moora and Zobel 1996, Smith et al.

1999) or causing competitive exclusion of non-responsive dominants (Gange et al. 1999,

Marler et al. 1999). Whether mycorrhizal fungi promote or inhibit the process of plant

invasion must be determined by examining responses of exotic plants to multiple fungal

genotypes with and without competition with native species that occupy similar

ecological niches.

Here we report the results of two experiments that examined the effects of fungal

isolates and abiotic environment on growth and competitive interactions of an exotic

invasive shrub. Specifically, we examined the effects of various isolates of AM fungi,









soil-P, light and competition type on the growth, physiology and biomass allocation of

Ardisia crenata Sims (Myrsinaceae, hereafter Ardisia). In the first greenhouse

experiment, we examined the effects of light, soil-P content, and AM fungal isolates on

growth, photosynthetic rates and biomass allocation patterns of Ardisia seedlings grown

singly in pots. We hypothesized that mycorrhizal plants would exhibit higher growth

rates, invest more biomass aboveground, and maximize leaf area ratio compared to

nonmycorrhizal plants, and plant response to phosphorus would vary among mycorrhizal

isolates. In the second greenhouse experiment, we examined the effects of AM on inter-

vs. intraspecific competition between seedlings of Ardisia and a native shade-tolerant

subcanopy tree, Prunus caroliniana (Mill) Aiton (Rosaceae, hereafter Prunus). We

hypothesized that Ardisia would be less affected by conspecific than heterospecific

competition, particularly when mycorrhizal.

Materials and Methods

Species

Ardisia is a woody evergreen shrub that was introduced as an ornamental to the

southeastern United States from east Asia ca. 100 years ago (Dozier 1999). Ardisia is

actively invading mesic forest understory in Louisiana, Texas, Hawaii, and north-central

Florida (Singhurst et al. 1997). Ardisia forms dense monodominant patches in the

understory and suppresses local species richness and diversity of native understory plant

species (A. Fox and K. Kitajima, unpublished data). The architecture of the plants

creates strong self- and neighbor-shading, even when plants are not in dense clumps (K.

Kitajima and M. Dooley, unpublished data). Growth of Ardisia seedlings in the field has

been shown to be positively correlated with soil-P content (Dozier 1999) and Ardisia

roots are highly colonized by AM in the field (S. Bray, unpublished data). Prunus was









chosen as a heterospecific competitor in the experiment because its juveniles are

abundant in forest understories where Ardisia typically invades (A. Fox and K. Kitajima

unpublished data). These species have similar seed size, and the juveniles of both species

have evergreen leaves that are common under the partially deciduous canopy of mesic

hardwoods forests in north-central Florida.


Experiment 1: Effects of Soil-P, Light and Inoculum Source on Growth and
Allocation

Three inocula and a sterile control were used in this experiment. The inocula were

Glomus etunicatum (S3029), G. fasciculatum (S3060), and host-associated fungi. S3029

has been maintained in pot culture for 15 years; S3060 was isolated in 1997 from a

tomato field in north-central Florida (Sylvia et al. 2001). These isolates have caused

positive growth responses in both agronomic and native woody plants (Sylvia 1990,

Sylvia et al. 1993) and will be collectively referred to as standard inocula. The host-

associated inoculum (HA) was composed of a corn trap culture initiated with washed

Ardisia roots gathered from a north-central Florida hardwood forest (29'40" N, 82'9" W).

Inocula were composed soil, roots and spores produced in the corn trap cultures.

Infection potential of the inocula was determined by growing corn (Zea maize) with 5 g

of inocula for 4 wk (Sylvia 1994). Infection potential rather than most probable number

was used because we were interested in comparing inocula rather than determining

absolute numbers of propogules. S3060 had the highest infection potential (51.7% +

10.2%) followed by HA (46.7% + 7.21%) and S3029 (33.0% + 7.00%).

Ardisia seeds were gathered from four populations in Gainesville, FL, mixed,

cleaned of pulp and stored in moist sand at 40C for 24 wk. Seeds were germinated in

petri dishes lined with moist filter paper at 260C. Before leaf development, seedlings









(approx. 15 d after radicle emergence) were transferred to bleach-sterilized Deepots (6.2

cm top diameter x 24.5 cm, J.M. McConkey & Co., Sumner, WA) containing a 1:1:1

steam-pasteurized mixture of soil:sand:peat moss. The soil was collected from Austin

Carey Forest (29'43" N, 82' 13" W). The soil had a pH of 5.7 (2:1, H20:soil), 0.1%

organic matter, and 1.6 mg Mehlich-I-extractable P kg-1. The sand was acid washed to

remove excess P, and rinsed until neutral pH was achieved. Phosphorus was added in the

form of KHPO4 at 0, 5, 30 or 60 mg P kg-1 soil. Pots were filled 34 full with the growth

medium; 5 g of inoculum were added to inoculated treatments and thoroughly mixed with

the growth medium. After adding seedlings of equal mass and remaining soil, pots were

randomly assigned to shading treatments with one (moderate light) or two (low light)

layers of shade cloth supported by PVC frames. These treatments created mean

photosynthetic photon flux densities (PPFD) of 412 rmol m-2 s-1 (moderate) and 212

mrnol m-2 S-1 (low) at mid-day. The shade treatments were randomly assigned to locations

within each of three blocks along a greenhouse bench. Each block contained 3 to 4 plants

per treatment group. For the control, S3060 and S3029 inoculum types, there were four P

levels by two light levels by 10 replicates for a total of 240 plants. The HA inoculum

was used only at the 5 mg kg-1 P level, but had 10 replicates in each light treatment for 20

HA inoculated plants. Plants were watered when necessary and biweekly given a

modified Hoagland's solution of 0. Ix concentration of all nutrients, except P at 0.01x

concentration (Sylvia et al. 2001).

The photosynthetic rates of the most recently fully expanded leaf were measured

with a Li-6400 gas-exchange system (Li-Cor, Lincoln, NE) for three seedlings per light

and inoculum treatment combination at the 5 mg kg-1 P level during the fourth month









after planting. All measurements were taken between 0800 and 1200. A CO2 mixer unit

maintained the CO2 concentration of incoming reference air at 380 ppm. Temperature of

the thermister block was maintained at 26C. Flow rate was 250 mL min Light was

supplied with a blue-red LED (LI6200-02B). Leaves were exposed to 500 [amol m-2 s-1

PPFD for 15 min for photosynthetic induction, after which quasi steady-state gas

exchange rates were recorded at light levels of 800, 500, 300, 100, 60, 40, 20, 10, and 0

[mol m-2 S-1

Plants were harvested after 258 d and leaf area was measured immediately with a

portable area meter (Li-3000, Li-Cor, Lincoln, NE). Roots of five randomly selected

plants within each treatment combination were weighed for fresh mass and set aside for

analysis of AM colonization. Roots of the remaining plants along with stems and leaves

of all plants were dried at 600 C until constant mass was reached. The dry mass of roots

used for assessment of colonization was estimated from fresh:dry mass ratios of roots.

To examine biomass allocation pattern, root: shoot ratio (R/S), specific leaf area (SLA,

leaf area divided by leaf mass), net assimilation ratio (NAR, net carbon assimilation on

leaf area basis) and leaf area ratio (LAR, leaf area divided by total mass) were calculated.

Relative growth rate (RGR) was determined using the following equation:


RGR (g g -1day 1 In (seedling mass at harvest) In (initial seedling mass)
RGR (mg-g 'day ) -
duration of study (days)


Tissue phosphorous contents were determined after grinding the dried stems and

leaves with a Wiley Mill with a 20-mesh screen (Thomas Scientific, Swedesboro, NJ).

Due to the small size of seedlings, two plants per block per treatment group at the 5 mg

kg-1 P level were combined. Samples were ashed overnight at 500C and digested with









12N HC1, followed by colorimetry methods of Murphy and Riley (1962) to determine

tissue-P concentration and content per plant.

To quantify mycorrhizal colonization, Ardisia roots were cleared in 10% KOH at

800 C for 45 min, while corn roots used for determination of inoculum potential were

cleared for 15 min. A longer clearing time was necessary for Ardisia roots due to their

high tannin content. Ardisia roots were then rinsed and soaked in H202 at 50C for 10

min for additional clearing. Roots were again rinsed and acidified in concentrated HC1 (5

ml HC1 per 200 ml-1 H20) for five minutes. The roots were then soaked overnight at

room temperature in trypan blue stain, which had been used successfully to stain field-

collected Ardisia roots (Bray, unpublished data). To estimate mycorrhizal colonization,

twenty one-cm root fragments were mounted on microscope slides and examined at 100x.

Roots were scored as mycorrhizal if they contained coils or arbuscules, spores, vesicles

or typical AM hyphae (aseptate, large diameter, angular branching).

Experiment 2: Effects of Mycorrhizae on Seedling Competition

As a competitor of Ardisia, we chose Prunus, a shade-tolerant subcanopy tree

found in north-central Florida hardwood forests. Prunus seedlings with four to eight

leaves and 7-12 cm height grown in soil-free medium were acquired from a commercial

nursery (Urban Forestry Services, Micanopy, FL). Examination of cleared and stained

roots of 10 seedlings revealed no mycorrhizal colonization. Although Prunus spp. Have

been shown to be ectomycorrhizal in some cases (Smith and Read 1997), we found no

evidence of ectomycorrhizae in Prunus caroliniana. Ardisia seedlings were collected

from the same invaded forest as the inoculum in experiment 1. Ardisia seedlings had

four to eight leaves and heights of 5-10 cm. Cleared and stained Ardisia roots revealed a

total mycorrhizal colonization level of 52% + 2%.









Prunus and Ardisia were grown in hetero- and conspecific competition with (AM)

or without (NM) mycorrhizal inoculum. Two plants were potted in each 3.8 L pot

containing the same medium as in experiment 1 with no additional P. Plants were paired

according to height and number of leaves. Twenty pots contained two Prunus seedlings

(Prunus conspecific competition), twenty contained two Ardisia seedlings (Ardisia

conspecific competition), and forty contained one Ardisia and one Prunus seedling

(Prunus heterospecific competition and Ardisia heterospecific competition). Half of the

pots were randomly assigned to the NM treatment and were drenched with 75 mg of

benomyl (Benlate) dissolved in one L of deionized water. Although benomyl may have

phytotoxic effects in some species, no phytotoxic effects have been reported in either

Prunus or Ardisia, and benomyl is commonly used to control fungal pathogens in

horticultural nurseries growing Prunus persica, Prunus dulcis, and Prunus serotina

(Fontanet et al.1998, Stanosz 1992). Plants in the AM treatment received 5 g of G.

fasciculatum (S3060) inoculum to supplement indigenous AM fungi and one L of

deionized water. In the AM treatments containing Ardisia, each pot contained S3060 and

fungi already inhabiting the Ardisia seedling; treatments without Ardisia contained only

S3060. Pots were then randomly placed under a shade frame (approx. 20% open-sky

light) in the greenhouse. Pots were watered as needed and received the modified

Hoagland's solution used in experiment 1 biweekly. The minimum and maximum

temperatures averaged 18C and 34C, respectively. The average maximum PPFD in the

greenhouse was 910 utmol m-2 s-1

At 160 d after transplanting, growth analysis was conducted with one randomly

selected plant per pot to ensure statistical independence. Root:shoot, LAR, SLA, and









RGR were determined as in experiment 1. A subsample of each root system of six

randomly selected individuals per treatment was used to estimate percent colonization

after determining fresh mass. Fresh:dry mass ratios of remaining roots were used to

estimate dry mass of the subsamples used to determine colonization. Initial mass of

seedlings was estimated through an allometric relationship of leaf number (mean + S.E.:

Ardisia 5.1 + 0.24; Prunus 5.7 + 0.41) and height (Ardisia 7.2 + 0.23 cm; Prunus 9.3 +

0.35 cm) with the total dry mass for each species (Ardisia 1.31 + 0.05 g; Prunus 0.46 +

0.04 g). Percent mycorrhizal colonization was determined as in experiment 1. Tissue-P

content was quantified only for leaves with the same method as in experiment 1.

Statistical Analyses

Allocation and growth data from both experiments were analyzed with factorial

model fitting (JMP 4.0, SAS Institute). For the analysis of the first experiment, the effects

of soil-P, light and inoculum type (control vs. standard inocula), and their interactions on

R/S, LAR, SLA, NAR, and RGR were analyzed. Because the host-associated inoculum

was given only at 5 mg kg-1 P level, the effects of inoculum type (control, S3029, S3069,

HA), light and their interactions at the 5 mg kg-1 P level were then examined. The results

of experiment 2 were analyzed with a model that included the effects of species (Ardisia

or Prunus), competitor (conspecific or heterospecific), mycorrhizal status (AM or NM),

and their interactions. When treatment and interaction terms were not significant (P >

0.1), they were dropped from the model and the results of the analysis with the reduced

model were reported. Tukey HSD at alpha = 0.05 was used to compare differences

between means. Effects of LAR and R/S on RGR were examined with multiple

regression analysis. Differences in survivorship among treatment groups were evaluated









with logistic regression models in both experiments. Percent colonization levels were

converted by square root of the arcsine to achieve normality and analyzed by ANOVA.

Results

Experiment 1: Effect of Light, Soil-P and Inoculum Type

In the first analysis, effects of the standard inocula (S3029 and S060) and

nonmycorrhizal control were examined at factorial combinations of two light levels and

four soil-P levels. Overall, there was no difference in RGR or biomass allocation patterns

among three mycorrhizal treatments (S3029, S3060 and control) in any combination of

light or soil-P. Thus, inoculum type was dropped from the ANOVA model. Neither soil-

P nor light affected RGR, SLA or NAR. Leaf area ratio and R/S, however, were affected

by light and soil-P (Table 2-1, 2-2). Leaf area ratio was positively correlated with RGR

and explained 37% of the variance in RGR (P < 0.0001) while R/S was negatively

correlated with RGR and explained 15% of the variance (P < 0.0001). Leaf area ratio

and R/S together explained 39% of the variance in RGR (P < 0.0001) in a multiple

regression. A total of twenty-seven seedlings of 260 died over the course of the

experiment, but survivorship was not affected by treatments.

In the second analysis, the effects of HA inoculum on seedling growth and biomass

allocation were compared to the standard inocula and control, at the 5 mg kg-1 P level

(Figure 2-la-d). Relative growth rates of plants with HA inoculum were twice that of the

standard inocula and control (P < 0.0001), while LAR was 2.5-3x that of the other

treatment groups (P < 0.0001). For pooled data across treatments, LAR was positively

correlated with RGR (P < 0.0001) and explained 35% of the variance in RGR.

Conversely, R/S of plants inoculated with HA were half that of the other inoculum

treatments (P = 0.0006). Root:shoot ratio was negatively correlated with RGR (P <









0.0001) and explained 23% of the variance in RGR. Together LAR and R/S explained

39% of the variance in RGR (P < 0.0001). Leaf area ratio was the only morphological

measure affected by light treatment, with plants in low light having higher values than

those in moderate light (Plight = 0.003). Inoculum effects on SLA were nearly significant

(P = 0.06); plants inoculated with HA had greater SLA (193.1 + 17.49 g/cm2 vs. 136.9 +

14.47 g/cm2) than controls and other inoculum sources. NAR was not affected by light or

inoculum.

Both shoot-P concentration and content differed among treatments (P = 0.005 and

P = 0.04, respectively). Plants inoculated with S3060 had the highest P concentration

and content, while HA inoculum had the lowest P concentration and second-highest

content (Figure 2-1 e, f).

Leaf gas exchange data showed similar trends to RGR in relation to inoculum type

(Table 2-3). The maximum photosynthetic rate of plants inoculated with HA were

approximately twice that of plants with S3060 and controls under moderate light. Due to

the small size of plants and leaves inoculated with S3029 under moderate light, no gas

exchange rates are available for this treatment group. Plants inoculated with HA had

rates approximately twice that of controls in both light treatments. Dark respiration rates

of all inocula were similar to the dark respiration rates of control plants.

Mycorrhizal colonization of plants mirrored the growth and biomass allocation

differences among inocula (P <0.0001); plants with HA inoculum were highly colonized

(mean + S.E.: 67 + 7%) whereas standard isolates (S3029 & S3060) had low colonization

rates (17 + 7% and 7 + 7%, respectively). Controls were virtually non-colonized (0.8 +

0.8%). Septate hyphae, presumably saprophytic, were observed externally with many









root samples, primarily concentrated in plants inoculated with S3060 and S3029.

Colonization levels were not related to original inoculum potential as determined by the

corn assay.

Experiment 2: Effects of Mycorrhizae on Seedling Competition

Mycorrhizal colonization was significantly lower in plants treated with benomyl

(NM treatment) than AM plants of both species (P < 0.0001). Benomyl was more

effective in reducing colonization in Prunus (AM = 73 + 7%, NM= 9 + 10%) than

Ardisia (AM = 59 + 1%, NM = 38 + 6%; Table 2-4). Competition type had no effect on

colonization levels.

Nine of 80 seedlings died over the course of the experiment. Mortality was

significantly higher for NM Prunus seedlings (7 of 9 dead seedlings, log-likelihood 2 =

6.57, P = 0.038, Pmyc = 0.037) than AM Prunus seedlings.

Relative growth rates differed significantly between treatments (P < 0.0001) with

Prunus having greater RGR than Ardisia (P < 0.0001, Figure 2-2a). Prunus and Ardisia

also responded differently to the competition treatment (Table 2-4, significant species x

competition interaction). While Ardisia had the highest RGR in heterospecifc

competition, the RGR of Prunus decreased approximately by half when grown in

heterospecific competition without mycorrhizae (Figure 2-2a).

Leaf area ratio also differed among treatment groups as Prunus and Ardisia

responded differently to mycorrhizal status (Psp*myc = 0.015, Table 2-4). Unlike in

experiment 1, Ardisia tended to have higher LAR when nonmycorrhizal in heterospecific

competition, whereas Prunus had higher LAR with mycorrhizae in both competition

types (Figure 2-2b). Ardisia had a higher R/S than Prunus (P < 0.0001, Figure 2-2c);

however, R/S of Ardisia decreased under conspecific competition when mycorrhizal









(comp*myc = 0.04). Specific leaf area did not vary significantly among treatments and

species (P =0.18, Figure 2-2d).

Leaf-P concentration and content differed among species and treatments (Table 5;

Fig. 2-2e, f). Ardisia had lower P concentration than Prunus, and was not affected by

competition or mycorrhizal treatment (Figure 2-2e). Prunus had its highest P

concentration in the NM, heterospecific competition treatment, while the other three

treatments did not differ significantly from one another (Figure 2-2e). Ardisia in the NM,

conspecific treatment had a significantly lower total P content than the other three

treatment groups (Figure 2-2f). Prunus had lower P content in NM than AM treatments

with no effect from competition (Figure 2-2f).

Discussion

Effect of Inoculum Source on Ardisia


Although all three types of inocula colonized Ardisia roots, they had strikingly

different colonization levels and effects on biomass allocation and growth of the host. In

experiment 1, only the HA inoculum isolated from field-collected Ardisia roots, but not

the standard inocula, improved seedling RGR over the nonmycorrhizal control. As

Ardisia had no response to soil-P nor was higher RGR accompanied by an increase in P

content or concentration, it appears that the benefit of the HA mycorrhizae was mediated

through changes in allocation and physiology rather than improved P nutrition. Plants

inoculated with HA inoculum had less relative investment in roots, greater leaf area and

higher Amax than the control. HA inoculum phenotypically altered LAR of Ardisia

seedlings from a low value typical for shade tolerant species to a higher value. LAR is

generally thought to be the primary determinant of RGR both across and within species









(Poorter and Remkes 1990) and in this study accounted for 35-37% of the variation in

RGR. In a study by Lovelock et al. (1996), shade-tolerant seedlings of Beilschmiedia

pendula also increased their RGR through an increase in LAR when mycorrhizal and its

morphology became more similar to that of more light-demanding plants.

The lack of positive effects on morphology and growth by the standard inocula

despite their positive effects on tissue phosphorus concentration was surprising,

especially given that they have been shown to increase the biomass of several species,

including woody, native plants (Sylvia 1990, Sylvia et al. 1993, Sylvia et al. 2001).

Possibly the costs of the two Glomus species were greater than their benefits at the light

levels used in this experiment. Mycorrhizal fungi can demand up to 20% of the total C

budget of a plant in extreme cases (Peng et al. 1993), and carbon costs can vary widely

among fungal genotypes (Graham et al. 1996). The differences between inoculum types

may also be mediated by fungal diversity. Isolates S3029 and S3060 represent single-

spore cultures, while the HA inoculum was likely composed of multiple fungal species

and strains. At least one of these may have been more effective than the two Glomus

species. Greater numbers of fungal species have been shown to increase the productivity

of grass macrocosms

In contrast to the strong effect of HA inoculum in experiment 1, suppression of

mycorrhizal fungi in experiment 2 had no effect on RGR of Ardisia seedlings, even

though colonization rates were reduced from 59% to 38%. These seedlings were

collected from a dense Ardisia population in the field and were presumably colonized

with mycorrhizae similar to HA. The magnitude of reduction of mycorrhizal

colonization (33% reduction) in this study was comparable to the reductions in many









other studies, (Kahiluoto et al. 2000, Moora and Zobel 1996, Smith et al. 1999).

However, it has been suggested that the relationship between mycorrhizal colonization

and plant benefit is curvilinear with benefit to the plant eventually reaching a plateau at

some colonization level (Gange and Ayers 1999). Ardisia may have reached its maximal

benefit at or before a colonization rate of 38%, as found in the benomyl treatments.

That Ardisia has a differential response to different inoculum types may have

important implications for its invasive ability. In heavily invaded areas, Ardisia is

already associating with effective mycorrhizal fungi that alter its morphology and

physiology to that of faster-growing plants. The main mode of resource competition by

Ardisia is through casting dense shade to its neighbors. Increased LAR, enabled by the

reduction in R/S due to mycorrhizae, must enhance Ardisia's competitiveness for light in

the forest understory.

Competitive Interactions

Ardisia seedlings grew better in heterospecific competition with Prunus seedlings

than in conspecific competition. Conversely, Prunus seedlings had lower survival and

growth with Ardisia seedlings than with conspecific seedlings, especially in

nonmycorrhizal treatment (Fig. 2-2a). The architecture of Ardisia results in a higher

amount of self- and neighbor-shading than that of Prunus (K. Kitajima, unpublished

data). Hence, each Ardisia seedling is more shaded by a conspecific neighbor than a

heterospecific neighbor. In conspecific competition, Ardisia responded with greater

phenotypic plasticity of increasing LAR than in competition with less shade-casting

Prunus.

Prunus seedling growth and survival was reduced to a greater extent by

heterospecific competition in the nonmycorrhizal than in the mycorrhizal treatment. The









presence of HA and S3060 inocula reduced the negative effect of interspecific

competition on Prunus, as is often observed in more mycorrhiza-dependant species

(Hartnett et al. 1993, Moora and Zobel 1996, Smith et al. 1999). More mycorrhiza-

dependant species often have a low total investment in roots (Jakobsen 1991). Prunus

had overall higher RGR than Ardisia, and this difference was associated with inherently

higher LAR and lower R/S of Prunus (Fig. 2-2 a-c). Mycorrhizae apparently allowed

Prunus seedlings to invest less in roots and more to leaf area, and enabled them to

compete more effectively with Ardisia seedlings for light. This finding of an apparent

greater AM dependency by a faster growing species in a competitive regime is interesting

because often the opposite trend has been found in the absence of heterospecific

competitors (e.g. Janos 1980, Zangaro et al. 2000).

Different plant species in the same community can support different mycorrhizal

communities in their rhizosphere (Bever 1994) and cause differential rates of sporulation

(Bever et al. 1996). A high percentage of Ardisia roots are colonized by AM fungi in the

field (Bray, unpublished data), likely dominated by preferred mycorrhizal fungi. The

field-collected Ardisia seedlings in Experiment 2 had been colonized by mycorrhizal

fungi, some of which remained after the benomyl treatment. These fungi colonized

Prunus seedlings at a low level (9%), but they did not benefit growth of Prunus seedlings

in the heterospecific competition treatment. The community composition of mycorrhizal

fungi may be highly modified in the dense clump of Ardisia in the invaded forest. If

Ardisia alters the composition of AM, the competitiveness of Ardisia may be increased.

Implications of Effects of Mycorrhizae on Exotic Species Invasion

As a new colonist in Florida, Ardisia is apparently not limited by the lack of

potential mutualists and, in fact, benefits from the local mycorrhizal fungi. Unlike typical









invaders, Ardisia is highly shade-tolerant and has low RGR. Many other exotic species

that have a higher RGR than Ardisia have a negative or neutral response to mycorrhizae

when they are grown alone (Marler et al. 1999, Philip et al. 2001, Richardson et al. 2000).

This appears to be the first study to document a positive response of a slow-growing

exotic to native mycorrhizae. It is likely, however, that further study will show that there

is no link between life history and mycorrhizal dependency in exotic species as has been

found in the mycorrhizal literature as a whole (Allsopp and Stock 1992, Janos 1980,

Smith and Smith 1996, Zangaro et al. 2000). Until a greater predictive framework for

mycorrhizal response is developed, invasive plant response must be examined on a

species by species basis.

The results of our study suggest that it is difficult to predict how competitive

interactions between exotic and native plants are modified by mycorrhizae. The exotic

plant's response in isolation does not necessarily predict its response to mycorrhizae in a

competitive environment. Another study of competition between an exotic forb and

native grass found that neither species' biomass was altered by mycorrhizae when grown

in isolation, but when grown in mixture, mycorrhizae increased the growth of the exotic

plant to the detriment of the native (Marler et al. 1999).

Our results also suggest that the type of mycorrhizal inoculum must be considered

when evaluating mycorrhizal effects on competitive interaction between native and exotic

species. The results vary depending on whether the fungal inoculum is the one preferred

by the native or the exotic. Thus, studies should incorporate evaluation of both species

and the mycorrhizae of the ecosystem invaded by exotic plant species.






27


The role of AM in mediating plant invasions and competitive interactions needs to

be examined carefully. The response of exotic plants to mycorrhizae is highly variable

depending on genotype interactions both in isolation and in competitive environments.

Understanding how native and exotic plants respond to the local microbial community

will be important for understanding the mechanisms and impacts of community invasion.

Similarly, it is also imperative that we determine how exotic species potentially alter the

microbial community and its ecosystem functions.









Table 2-1: ANOVA summarizing the effects of light and soil on LAR and R/S in
experiment 1 (model P <0.05). Fungal isolate effects (control, S3029, and
S3060) were not significant and were pooled.
LAR R/S
Source F P F P
Model 4.51 0.0001 3.24 0.003
Light 9.24 0.002 0.20 0.66
P-level 3.11 0.03 3.94 0.009
Light P-level 3.37 0.02 3.24 0.02
Notes: Fungal isolate effects (control, S3029, and S3060) were not significant and were
pooled. Abbreviations are: LAR, leaf area ratio; R/S, root to shoot ratio; P, phosphorus.


Table 2-2: Means of leaf area ratio (LAR) and root:shoot ratios (R/S) from all soil-P
levels in experiment 1. Different letters in the same column signify
significant difference in Tukey HSD values (alpha = 0.05).
[P] LAR R/S RGR
(cm2 g-) (mg g-1 day-l)
Moderate Light 0 20.5 a 2.38 a 5.05
5 17.5 a 2.24 a 5.08
30 22.5 a 1.83 a 6.05
60 28.9 ab 1.52 b 6.62

Low Light 0 29.1 ab 1.91 a 5.77
5 24.9 a 2.12 a 4.75
30 38.0 b 1.66 b 7.02
60 24.2 ab 2.08 a 5.18
Notes: Abbreviations: RGR, relative growth rate; other abbreviations as in Table 2-1.


Table 2-3: Comparison of light saturated net photosynthesis rate (Amax) and dark
respiration under moderate vs. low light treatments (means + 1 SE) from gas
exchange measurements of three individuals from experiment 1.
Amax Dark Respiration
Inoculum Moderate Low Moderate Low
Control 2.02 + 0.55 2.02 + 0.53 -0.279 + 0.062 -0.310 + 0.045
S3029 NA 3.43 + 1.19 NA -0.307 + 0.132
S3060 2.65 + 0.38 2.19 + 0.35 -0.382 + 0.078 -0.253 + 0.015
HA 4.04 + 0.28 4.42 + 0.78 -0.356 + 0.046 -0.313 + 0.065
Notes: No data were collected for the moderate light treatment of isolate S3029 due to
small size of leaves in this group. Abbreviations: NA, not available; HA, host-associated.












Table 2-4: ANOVA summary of the effects of species, competition, and mycorrhizae on RGR, LAR, R/S, and colonization rates from
experiment 1 (model P < 0.05). A dash (-) indicates that the specified effect was not significant (P > 0.1) and was dropped
from the model.
RGR LAR R/S Colonization
Source F P F P F P F P
Species 29.1 <0.0001 0.129 0.73 62.5 <0.0001 2.03 0.16
Myc 2.94 0.09 1.16 0.28 2.93 0.09 38.9 <0.0001
Comp 3.19 0.08 2.20 0.14 0.059 0.81
Sp*Myc 1.10 0.30 6.31 0.01 12.0 0.001
Myc*Comp 0.481 0.49 4.54 0.04
Sp*Comp 11.9 0.001 3.36 0.07 -
Sp*Myc*Comp 3.78 0.06 -
Notes: Degrees of freedom were: RGR, 1,63; LAR, 1,65; R/S, 1, 66; colonization, 1,67. NS indicates that the specified effect was not
significant (P > 0.1), and was dropped from the model.


Table 2-5: ANOVA summary of the effect of species, competition, and mycorrhizae for P concentration and total P content from the
competition study.
P concentration (mg P g-1 tissue) Total P content (mg P)
Source F P F P
Species 43.6 <0.0001 19.1 <0.0001
Myc 0.086 0.70 24.9 <0.0001
Comp 9.6 0.003 2.57 0.11
Sp*Myc 3.81 0.06 -
Myc*Comp 4.15 0.05 4.51 0.04
Sp*Comp 7.06 0.01 5.27 0.02
Notes: Degrees of freedom were, P concentration, 1,64; total P content, 1,65. The three-way interaction was dropped due to lack of
significance in both tests as was the species mycorrhizae interaction in the P content test.














0.016
ad
0.014 b d
u 200
7 0.012 o
o 150
S0.010 150
0.008
a a a 10
I 0.006 E

0-
0.002 111
-a
80 b c U 3.5 a
E 70 3.0 ab
o o ^
-0 60 bc 2.5
-U 0) ab
50 -2.0
40 ab 0
S ab ab 1.5 b
E 30 -a T
a a 1.0
S20 -
0 0.5
10 0

3.5 C 1.4
0 3.0 1.2 a
a ab
2.5 a 2 1.0
SE ab
0 2.0 0.8 b
4 1.5 c 0.6
0 0
r, 1.0 0 0.4
0.5 0.2
0.0 0.0
Control S3029 S3060 HA Control S3029 S3060 HA
Inoculum Source Inoculum Source






Figure 2-1: Response ofArdisia to light and inoculum type at 5 mg kg-1 P. Light
treatment is distinguished only when it had a significant effect (P < 0.05). a)
Relative growth rate in response to inoculum (P < 0.0001). b) Leaf area ratio
in response to light (P = 0.003) and inoculum (P < 0.0001). Open bars =
moderate light; hatched bars = low light. c) Root:shoot ratio in response to
inoculum (P = 0.0006). d) Specific leaf area in response to inoculum (P =
0.06). e) Shoot-P concentration (mg P g-1 tissue) in response to inoculum (P
0.005). f) Shoot-P content (mg P) in response to inoculum (P = 0.04). Error
bars represent + 1 SE. Letters signify difference between Tukey HSD values
at alpha = 0.05.
















0.010

S0.008
-o
- 0.006

0.004

0.002


2 100

E
o 80
-o

2 60

c? 40
E
o
S20
J


1.2 C


1.0 a

0.8
Wc
S0.6
o
0.4


HC HC
AM NM

Ardisia


b

b






HC HC
AM NM
Prunus


S--250
U)
m
0-
I 200

w 150

C' 100
05
< 50
-j
C'
U 12
U)
U)
10
0)
0-
0) 8
E
a 6

S4
0
o
a 2
o
0
n-
a3


e c

cb

ab I ab ab
a
Ta abr


-II


10 f


HC HC
AM NM

Ardisia


ab
bc

bc






HC HC
AM NM
Prunus


Figure 2-2: Response of Ardisia and Prunus to heterospecific or conspecific competition
and mycorrhizal status. H = heterospecific competition; C = conspecific
competition. a) Relative growth rate (P < 0.0001). b) Leaf area ratio (P =
0.04). c) Root to shoot ratio (P < 0.0001). d) Specific leaf area (P = 0.18). e)
Leaf-P concentration (mg P g-1 tissue) (P < 0.0001). f) Leaf-P content (mg P)
in response to inoculum source (P < 0.0001). Error bars represent + 1 SE.
Letters signify difference between Tukey HSD values at alpha = 0.05.














CHAPTER 3
SOIL MICROBIAL COMMUNITY STRUCTURE AND FUNCTION IN FLORIDA
PLANT COMMUNITIES PRONE TO NON-NATIVE PLANT INVASION

Introduction

Microbes are a key component of ecosystems that function as pathogens, mutualists

and decomposers. Recent studies have shown that the composition of microbial

communities differs among plant communities, as influenced by biotic and abiotic factors

(Waldrop et al. 2000, Myers et al. 2001, Gallo et al. 2004, Leckie et al. 2004, Waldrop

and Firestone 2004). Microbial community metabolic potential and structure change

across climatic gradients (Staddon et al. 1998), nitrogen levels (Pennanen et al. 1999,

Gallo et al. 2004, Leckie et al. 2004), and soil moisture contents (Bossio and Scow 1998).

Microbial community composition also varies under different plant species within a

given community (Grayston et al. 1998, Bardgett et al. 1999, Saetre and Baath 2000,

Priha et al. 2001). These differences are likely due to the differences in root exudates and

turnover and the quantity and quality of aboveground litter inputs (Grayston et al. 1996).

Plant species composition is changing in many plant communities due to the

introduction of non-native species. These invasions result not only in shifts in plant

community composition, but can also have ecosystem-level effects such as altered

nutrient levels, hydrology, and soil accumulation (Gordon 1998). Invasions, then, offer a

"natural experiment" in which to examine the role of species composition in ecosystem

function. It is likely that altered substrate composition due to plant composition shifts

accompanied by altered abiotic soil environment will result in altered microbial









community composition. In a temperate forest, two non-native understory species

supported microbial communities significantly different from a native understory species

within the same forest stand (Kourtev et al. 2002). These differences extended below the

root zone of the vegetation but larger scale effects were not examined. This leaves

interesting questions on the potential impacts of exotic species on microbial community

composition: 1) Can invasive plants alter microbial community composition beyond their

crown- and root-zone of influence? 2) Do invaders have a consistent effect on microbial

community composition across a landscape scale? and, 3) Does the identity of the invader

and the community type it invades influence the potential effects of that invasion on

microbial community composition?

The state of Florida presents many opportunities to examine a number of non-

native invaders across a variety of plant communities. The Florida Exotic Pest Plant

Council lists over 120 non-native species as invasive (FLEPPC 2003). These species

range from grasses to trees and invade communities from freshwater marshes to upland

pine savannas in freeze and freeze-free climate zones. We examined the impact of five

invasive plants, ranging from trees to understory herbs, on microbial communities of

plant communities in saturated south Florida everglades marsh soils to well-drained

north-central Florida forest soils. We used phospholipid fatty acids (PLFA) to examine

the structure of microbial communities and Biolog substrate utilization to examine the

function of those communities.

Biolog and PLFA profiles have become common methods to examine microbial

community composition as they are rapid and of low enough cost to allow for the

analysis of the number of samples required for ecological studies. Biolog plates assay the









functional traits of microbial communities by testing their cumulative ability to

metabolize 95 different carbon substrates. Because Biolog depends upon the growth of

organisms within the 95 wells, it has many of the same limitations as culture-based

techniques and organisms adapted to high resource availability are likely to be over-

represented (Smalla et al. 1998). However, Biolog has been successfully used to

differentiate microbial communities in different soil types, land management treatments

and rhizospheres of different plant species (Gorlenko and Kozhevin 1994, Bossio and

Scow 1995, Garland 1996). PLFA profiles are a useful technique for examining the

structure of the microbial community at many levels. A common use is to examine

overall "fingerprints" of microbial communities by subjecting the abundance of the

various fatty acids within a sample to multivariate analysis (Frostegard et al. 1993).

Major taxonomic groups of microbes, such as fungi, microeukaryotes, and Gram-

negative and Gram-positive bacteria can be specifically examined by the use of

biomarkers (Vestal and White 1989). In addition, functional groups such as aerobes,

anaerobes, methanotrophs, and sulfate reducers can be examined by the use of fatty acid

biomarkers (Harwood and Russell 1984, Dowling et al. 1986, Parkes 1987, Hill et al.

2000).

We used these two measures of microbial community structure and function to

examine microbial communities within invaded and non-invaded areas within the habitat

prone to invasion by each of the five invasive species. As both biotic and abiotic factors

are known to influence microbial community composition, we propose two alternative

hypotheses for the composition of microbial communities. Inherent characteristics (e.g.

soil moisture, nutrient levels, physiognomy of vegetation) of habitats may drive









composition of microbial communities regardless of plant species. Alternatively, the

invasion and dominance of non-native species may override the habitat effect, resulting

in changes in the microbial community because of changes in vegtation.

Methods

Species, Sites, and Sampling

The invaders we examined included trees, Sapium sebiferum, Schinus

terebinthifolius, and Melaleuca quinquenervia, an understory shurb, Ardisia crenata, and

an herbaceous perennial, Ruellia brittoniana. In addition to being dominants in the

habitats they invade, Sapium, Schinus, and Melaleuca have been shown to have

ecosystem-level impacts on native communities such as increased soil elevation,

increased litter accumulation, or altered mineralization, disturbance or hydrology

(Woodall 1981, Cameron and Spencer 1989, Greenway 1994, Laroche 1994, Gordon

1998). Secondary defensive chemicals in Schinus are known to have allelopathic effects

on other plants (Morton 1978, Mahendra K.J. et al. 1995) and the conspecific Melaleuca

alternifolia has antibiotics that depress decomposition rates (Boon and Johnstone 1997,

Bailey et al. 2003). Thus, we felt that these three species would have the greatest

potential for altering microbial community composition. We chose the additional

understory plants, Ardisia crenata and Ruellia brittoniana as cases in which invaders do

not contribute a large percentage of the total biomass of a plant community.

The habitats that these species invade vary in their locations, physiognomy of

native vegetation, moisture regimes and susceptibility to invasion by specific invaders

(Table 3-1). For ease of discussion we will identify these habitats by the species to which

they are most prone to invasion, i.e., Sapium-prone habitat (SA), Schinus-prone habitat

(SC), Melaleuca-prone (ME), Ruellia-prone (RU), and Ardisia-prone (AR). For









consistency of presentation these habitats will appear in order of hydrological position

from wettest to driest in the results figures and tables.

We selected three sites for each habitat. Most habitats included only one native

plant community type (Myers and Ewel 1990) except for RU habitats in which two sites

were in bottomland swamp forests and one in a cabbage palm hammock and SA habitats

sampled in 2 wet prairies in Texas on clay-dominated soil and one site in Florida on

sandy muck soil. The two southern-most habitats, SC and ME, were sampled in

Cladium-dominated everglades and tropical hammock rocklands, respectively. The AR

habitats were sampled exclusively in mixed hardwood forests. We chose sites that had

areas of high invader density (invaded) and areas free of the invader (non-invaded). In

order to minimize the chance that differences between invaded and non-invaded areas

were based on site conditions prior to invasion, we chose only sites where other workers

knew the history of the site. The absence of invasion in non-invaded areas was

maintained either by human intervention or a clear "invasion front" was apparent. Except

for the presence of the invasive species, the invaded areas appeared similar to non-

invaded areas used for the study in terms of topography and soil type.

Within each site, six 5 m x 5 m plots were established-three in invaded areas and

three in non-invaded areas. In most cases, each invaded plot was spatially separated from

other distinct areas of invasion. When this was not possible, invaded plots were

separated by at least 50 m in a large patch of the invader. Corresponding non-invaded

plots were located at least 30 m from the invasion front.

Within each 5 m x 5 m plot, 20 soil cores (diameter 2 cm) were extracted

systematically, approximately 80-100 cm apart, from the top 10 cm of soil, including the









litter layer. Cores within a plot were combined to create a composite sample. Samples

were kept cool (- 4 C) during collection in the field and frozen at the end of each day to

prevent alteration of the microbial community during the remainder of field collection.

The southern most habitats, ME and SC, were collected within 3 days of each other

in the July of 2002 and Florida SA and AR in August 2002. The Texas SA sites were

collected in September 2002. In the summer of 2003, RU samples were collected. All

samples were collected within the wet season.

Microbial Community Composition and Nutrient Analysis

Microbial community structure was examined using phospholipid fatty acid

(PLFA) biomarkers. All glassware was heated at 500 OC for five hours and PTFE lined

caps rinsed with hexane to remove any organic matter. A subsample weighing a total of

5 g (dry mass) from each plot was extracted in glass centrifuge tubes using single-phase

phosphate-buffered methane chloroform solvent (White et al. 1979). After two hours, an

additional 5 mL each of chloroform and methanol were added to break the phase.

Samples were briefly shaken, vented, and allowed to separate overnight. Samples were

then centrifuged and the organic layer passed through a Whatman #2 paper into a test

tube where the solvent was driven off under N2 gas. The organic phase was resuspended

and separated on a silcic acid column. The phospholipid fraction was transesterified to

fatty acid methyl esters (FAME) by mild alkaline methanolysis (Findlay and Dobbs

1993). For identification, samples were suspended in hexane with 19:0 fatty acid as an

internal standard and analyzed with an Agilent Technologies 6890 gas chromatograph

(Palo Alto, CA) with a 25m Ultra 2 phenyl methyl silicone column. The temperature

program increased from 170 C to 270C at 5 OC per minute. Peaks were identified by









MIDI peak identification software (MIDI, Inc., Newark, DE) and co-elution with

standards.

Fatty acids were defined in terms of the ratio of total number of carbon atoms :

number of double bonds. The position of the double bond from the methyl end of the

molecule is signified by the symbol co followed by the carbon position; cis and trans

geometry are referred to by "c" and "t." The prefixes "i" and "a" signify and iso and

anteiso branching; "cy" signifies a cyclopropyl fatty acid; 10Me indicates a methyl group

on the tenth carbon atom from the carboxyl end of the fatty acid. The position of

hydroxy groups is indicated by xOH. Fungal:bacterial ratio was calculated as

(18:2co6c)/(il5:0 + a15:0 + 15:0 + i16:0 + 16:1co5c + i17:0 + a17:0 + 17:0 + 18:1om7c +

cyl9:0) (Frostegard and Baath 1996). Gram bacteria were represented by fatty acids

16:lco7c, 16:lcow7t, 17:0cy, and 18:lco7c; Gram + bacteria by il5:0, a15:0, i16:0, a17:0,

and 17:0 10 ME (O'Leary and Wilkinson 1988, Wilkinson 1988). Total PLFA (nmol g-l)

was used as a proxy for total biomass.

Microbial community function was examined as the ability to metabolize the 95

substrates in Biolog Gm- plates (Biolog, Inc., Haywood, CA). We inoculated one plate

per plot with a 10-3 dilution of one-gram subsamples. Absorbance was measured at 12,

24, 48, 72, and 96 hours with a microplate reader (Bio-Rad, Hercules, CA).

A subsample of soil from each plot was dried at 60 oC until constant weight was

achieved to determine gravimetric moisture content. Subsamples were then ground with

a mortar and pestle to homogenize. The carbon and nitrogen contents of soil were

determined by combustion on an ECS4010 elemental combustion system (Costech

Analytical Technologies, Inc, Valencia, CA).









Statistical Analyses

In statistical analyses, site was treated as the unit of replication (i.e., means of the

three plots per site), yielding an n of 30. Differences in %C, /oN, C:N, %moisture, total

PLFA (nmol g-1), and fungal:bacterial ratio among habitats and invasion-status were

tested with ANOVA. Variables were log-transformed for normality and homogeneity of

variances when necessary. When interaction terms of the two-way ANOVA were

significant, Bonferroni-corrected t-tests were used to look for differences between

invaded and non-invaded soils within each community. Correlations between soil

characteristics and fungal:bacterial ratio and total PLFA were examined with simple

linear and quadratic regressions.

We used principle components analysis (PCA) to summarize the multivariate data

sets and create microbial community fingerprints for both PLFAs and substrate utilization

from Biolog. The twenty-three most common, positively identified PLFAs comprising >

1% of the total amount of fatty acids extracted were included in the ordination as

percentage of the total fatty acids per sample after arcsine-square root transformation.

Absorbance values at 48 hours for 96 carbon sources for four of the five habitats

examined (excluding RU) were ordinated to examine substrate utilization patterns.

Average scores per site from the first two PCA axes of each ordination were subjected to

MANOVA to examine the effects of habitat, invasion and habitat by invasion interaction.

Finally, differences among habitats and invasion-status in the relative abundance of

PLFAs that were important in structuring the PCA were determined with one- and two-

way ANOVAs.

All univariate statistics were performed in JMP 4.0 (SAS Institute) and ordinations

were performed in PCOrd 4.2 (MJM Software Design, Gleneden Beach, OR).









Results

Soil characteristics varied among habitats but invasion had no significant effect.

ME habitats were most different from the other habitats examined with the highest

moisture, carbon and nitrogen contents in the soil and lowest C:N ratio (Table 3-2).

Percent C, N and moisture were highly correlated with one another (C-N r2 = 0.88, C-

moisture r2 = .61, N-moisture r2 = 0.74).

Total phospholipids and fungal:bacterial ratios differed significantly among habitat

types. Only habitat type had a significant effect on total PLFA (F4,20 = 22.29, p <

0.0001), with RU, SA, and SC habitats having the highest total PLFA (Figure 3-1).

Fungal:bacterial ratio was highest in RU and AR habitats and lowest in ME habitats (F4,

20 = 69.5, p <0.00001, Figure 3-2). Although there was no main effect of invasion, there

was a significant habitat by invasion interaction (F4, 20 = 3.06, p = 0.04). AR and ME

habitats showed a trend of increased fungal : bacterial ratio with invasion, while SC, SA

and RU habitats showed a trend towards decreased fungal : bacterial ratio with invasion

(Figure 3-2).

Both total PLFA and fungal : bacterial ratio were correlated with the soil

characteristics examined. The relationship of total PLFA with soil characteristics was

unimodal (i.e., maximum biomass at intermediate values of each soil characteristic) and

quadratic equations produced a better fit than linear equations. Total PLFA was best

correlated with percent carbon and nitrogen (Table 3-3). Relationships between

fungal:bacterial ratio and soil characteristics were monotonic and simple linear

regressions produced the best fits. Moisture, nitrogen, and carbon contents were

negatively correlated with fungal:bacterial ratio explaining 70%, 48%, and 26% of the









variance, respectively. Conversely, C:N was positively correlated with fungal:bacterial

ratio (Table 3-3).

We examined differences in relative representation of Gram and Gram + bacterial

groups as related to habitats and invasion-status. Gram + bacteria were highest in SA and

AR habitats while the inverse was true for Gram bacteria (Figure 3-3). All effects were

significant in a two-way ANOVA of Gram PLFAs (Table 3-4). Gram PLFAs

increased with invasion in SC and RU habitats, decreased in AR habitats and had no

change in ME and SA habitats. Nitrogen and carbon contents of the soil were positively

correlated with Gram bacteria and negatively correlated with Gram + bacteria.

Moisture content was positively correlated only with Gram bacteria and C:N had no

significant relationship with either Gram negative or positive bacteria (Table 3-5).

To examine overall differences in microbial community composition in relation to

habitat and invasion status, we ordinated the PLFA data by principle components

analysis. The first two components explained 22.3% and 17.2% of the variance,

respectively. Microbial communities from different habitats were primarily separated by

the first axis, although the two wettest habitats were also separated from the remaining

communities by the second axis (Figure 3-4). Within habitats, differences between

invaded and non-invaded samples were primarily on the first axis. In MANOVA, both

habitat and invasion effects were significant (Table 3-6).

The PLFA with high loadings on the first two PCA axes (loadings > 0.3) included

ill:0 30H, 16:0, i16:0, 16:lco7c, 16:lco5c, 16:1 20H, cyl7:0, and 18:2co6c (Table 3-7).

PCA 1 was negatively correlated with 16:0 (ubiquitous fatty acid) and 16:lco7c (Gram -

PLFA), and positively correlated with i16:0 (Gram + PLFA) and 16:1 2 OH (Gram -









PLFA). PCA 2 was negatively correlated with i 11:0 30H, 16: lw5c (mycorrhizal PLFA,

also often classified as Gram -), cyl7:0 (Gram PLFA, possibly an anaerobic marker),

and positively correlated with 18:2co6c (fungal PLFA). Soil characteristics were not

strongly correlated with PCA 1, but were correlated with PCA 2 (Table 3-8). Moisture

content, followed by %N and %C, was strongly negatively correlated with PCA 2. C:N

was positively correlated with PCA 2.

We then explored the effect of habitat and invasion-status on the relative

representation of these eight PLFAs with high loadings. Only habitat effect was

significant for five PLFAs (il 1:0 30H, i16:0, 16: lo5c, 16:0, and cyl7:0) whereas both

habitat and invasion had significant effects on the abundance of 16: lo7c, 16:1 20H, and

18:2co6c. Two somewhat surprising patterns did arise. A purported mycorrhizal PLFA,

16:lco5c, was higher in wetter communities and lower in drier communities (Figure 3-5).

Similarly, a purported anaerobic marker, cyl7:0, was higher in drier communities than in

the wet ME and SC habitats. It is unknown what group ill:0 30H may represent, but its

concentration appeared to decrease across the moisture gradient of habitats. The

remaining fatty acids, i16:0 and 16:0 were variable across the moisture gradient.

Three fatty acids, 16:1 20H, 16:lco7c (both Gram -PLFAs), and 18:2co6c (a fungal

PLFA) exhibited both a significant main effect of habitat and also a significant

interaction effect. There was a general decrease in 16:1 20H across the habitat moisture

gradient (Figure 3-6a). At the extremes of the gradient, 16:1 2 OH tended to decrease

with invasion, while the middle of the gradient showed no effect of invasion on its

relative abundance. In contrast, for 16: lco7c, there was no clear pattern across the habitat

gradient as SA and AR habitats had the lowest relative concentration of 16: lco7c (Figure









3-6b). The effect of invasion ranged from a decrease in 16: co7c with invasion in AR and

ME habitats, an increase in RU habitats, and no discernable effect in SC and SA habitats.

Finally, the fungal PLFA, 18:2o6c, generally increased across the habitat moisture

gradient (Figure 3-6c), increasing with invasion in AR and ME habitats.

The catabolic potential of microbial communities from ME, SC, SA, and AR

habitats was examined as the ability of those microbes to metabolize 95 carbon

substrates. Although the majority of substrates were metabolized and there was no

significant difference in the number of substrates metabolized between habitats or

invasion-status, rates of utilization differed. Substrate utilization patterns show lesser

degree of discrimination among microbial communities than PLFAs (Figure 3-8). A

MANOVA indicated only a significant effect of habitat type (Table 3-9).

Discussion

Habitat Controls of Microbial Community Composition

Our results clearly indicate that habitat was the primary control of microbial

community composition across the sites we examined. Microbial communities from

different habitats differed in total PLFA, fungal:bacterial ratio, relative representation of

Gram and Gram + bacteria, and several other individual PLFAs. Because plant

community composition is not independent of the soil characteristics in this study, it is

difficult to separate the effects of plant species composition from soil environmental

variables. The major environmental gradient across the habitats examined, however, was

water content of the soil and was apparently the major control of microbial community

composition. Total PLFA was highest at intermediate moisture lowest at the wettest and

driest habitats (ME and AR, Figure 3-1). At high water contents microbial activity is

limited by oxygen availability, while at low matric potential microbial activity is limited









by water availability (Griffin 1985). Fungi are better able to withstand low matric

potentials and therefore tend to dominate in drier soils and are at low levels or absent in

water-logged soils (Bossio and Scow 1995, 1998, Nakamura et al. 2003). This pattern

was corroborated by our results for the fungal PLFA 18:2co6c that decreased from AR

(driest) to ME (wettest) habitats (Figure 6c). Such differential responses of fungi and

bacteria resulted in a decrease in fungal:bacterial ratio across the gradient from low to

high water contents, with moisture explaining 51% of the variance in fungal:bacterial

ratio (Figure 3-2).

Water content of the soil was also associated with higher organic matter as

indicated by the high carbon and nitrogen contents in waterlogged ME habitats. Thus

although carbon and nitrogen availability was highest in ME habitats, total PLFA was

highest in communities with intermediate water, carbon and nitrogen contents.

Heterotrophic soil microbes are generally thought to be carbon- (Alden et al. 2001,

Ekbald and Nordgren 2002) or nitrogen-limited (Hart and Stark 1997). Our results

indicate oxygen availability becomes a greater limitation to microbial activity and

biomass than substrate availability under saturated conditions, however, when not under

saturated conditions, microbial biomass (as measured by total PLFA) is positively

correlated with C and N. Fungi and bacteria, however, responded in a predicted, linear

fashion to increasing carbon and nitrogen availability. Fungi (as measured by 18:2co6c)

and fungal:bacterial ratio decreased with increasing C and N concentrations, and

increased with C:N. Conversely, Gram bacteria were positively correlated with carbon

and nitrogen concentrations. Bacteria have higher nitrogen requirements and metabolic

and growth rates than fungi (Griffin 1985). This leads to the pattern of fungi dominating









in soils with low nitrogen with bacteria dominating in soils with high nitrogen (Bardgett

and McAlister 1999, Priha et al. 2001, Leckie et al. 2004). The increase in bacterial

biomass with soil carbon and nitrogen concentrations in this study was primarily due to

an increase in Gram biomass. Gram bacteria are able to quickly respond to nutrient

enrichment (Griffiths et al. 1999) and have been shown to be higher in communities with

higher soluble organic contents (Leckie et al. 2004). Gram + bacteria may be being out-

competed at high resource availability and others studies have shown this group to

decrease with increasing carbon availability (Bossio and Scow 1998).

These individual responses of particular functional groups led to an overall

differentiation of microbial communities among habitats (Figures 3-4 and 3-7). There

was greater separation of habitats by PLFA than by substrate utilization method. This

trend has been seen in other studies (Buyer and Drinkwater 1997); however, we feel that

the reduced separation discrimination of habitats by substrate utilization data could

reflect an artifact of freezing soil prior to analysis. In effect, only the catabolic potential

of microbes able to withstand the freezing and thawing process was examined. In

examining the effect of storage on the catabolic potential of plant growth promoting

bacteria, (Shishido and Chanway 1998) found that frozen soil samples were more similar

to one another than to its fresh sample counterpart. Nevertheless, different habitats

supported microbial communities distinct in both function and structure.

Alteration of Microbial Communities by Invasion

Overall, invasion-status had smaller, but significant, effects on soil microbial

structure, as measured by PLFAs, than those of habitat (Figure 3-4, Table 3-6). Four of

the five invaders (all but Schinus) appeared to alter microbial community structure. We

predicted that invaders that dominate the community to a greater extent in terms of









biomass would have larger impacts on microbial community composition. We did not

find that trend. Instead, it was only the two habitats at the extreme ends of the gradient,

AR and VME, that invasion had a significant effect on PCA scores within habitats.

Melaleuca converts wetlands dominated by Cladium (sawgrass, a sedge) to dense forest

while Ardisia adds a monodominant shrub-layer without altering the overstory. Chapin

and D'Antonio (Chapin et al. 1996, D'Antonio et al. 1999) predicted that invaders that

change the structure of a community would have larger impacts on that system. This

would seem to explain the great Melaleuca effects, but does not explain the effects of

Ardisia. Kourtev et al. (2002) did find that on a local scale shrubs could alter microbial

community composition. We believe that Ardisia may be altering microbial community

composition due to its high density of carbohydrate-rich roots (S. Bray unpublished data)

that may be providing a different or larger source of exudates in the top 10 cm of soil.

Our data indicate that invasion alters microbial community composition not only in

the zone of influence of a single plant as previously shown (Kourtev et al. 2002) but also

at a landscape level. There was a large variation in the abiotic characteristics, such as soil

moisture, that we examined across sites and likely large variation in other variables not

measured in this study. This variation across sites was likely responsible for the scatter in

microbial community composition and the relatively low total variation (39.5%)

explained by ordination of PLFA profiles.

Conclusions

Our data contribute to a growing literature base that demonstrates both soils

beneath different plants support different microbial communities (Grayston et al. 1996,

Westover et al. 1997, Grayston et al. 1998, Katajisto et al. 1999, Priha et al. 1999,

Kourtev et al. 2002, 2003, Bardgett and Walker 2004) and microbial communities differ









among habitats with contrasting types of vegetation (Waldrop et al. 2000, Myers et al.

2001, Leckie et al. 2004). Habitat is the main control of microbial community

composition, but invasion significantly modified microbial communities within a given

habitat. The direction of these changes, however, was not predictable across habitat

types. Changes in microbial community structure with invasion will likely be influenced

by abiotic conditions in the community and the identity of the invader. As microbes are

primary responsible for decomposition in most ecosystems, any alteration in microbial

community structure and function with invasion may result in altered nutrient cycling and

availability. The link between microbial community composition and process rates needs

to be explicitly examined across plant communities.












Table 3-1: Location, mean annual temperature, mean annual rainfall, soil type and dominant vegetation of the three sites in each


habitat type.
Location


Site


Mean Annual
Temperature (OC)a


Mean Annual
Rainfall (mm)a


Soil
Order


Dominant Native Vegetation


ME Habitats
1
2
3
SC Habitats
1
2
3
RU Habitats
1
2
3
AR Habitats
1
2
3

SA Habitats


26024' N, 80014' W
25055' N, 80026' W
2603' N, 80033' W

2608' N, 8103' W
2602' N, 81017' W
25056' N, 81018' W

29037' N, 82019' W
2807' N, 8209' W
28046' N, 81012' W

29037' N, 82017' W
29033' N, 82021' W
29040' N, 8209 W


29036' N, 82019' W
29023'N, 9501' W
29022'N, 9502' W


26.04
24.44
26.04

23.83
23.83
23.83

20.33
22.83
22.67

20.33
20.33
20.33


20.33
21.78
21.78


1560
1487
1560

1376
1376
1376

1228
1137
1228

1228
1228
1228


1228
1113
1113


Histosol
Histosol
Histosol

Entisol
Entisol
Entisol

Alfisol
Mollisol
Alfisol

Ultisol
Alfisol
Alfisol


Alfisol
Vertisol
Vertisol


Sedges
Sedges
Sedges


Tropical Hardwoods
Tropical Hardwoods
Tropical Hardwoods

Deciduous Soft- and Hardwoods
Deciduous and Evergreen Hardwoods
Evergreen Hardwoods and Palms

Deciduous and Evergreen Hardwoods
Deciduous and Evergreen Hardwoods
Deciduous and Evergreen Hardwoods,
Softwoods


Grasses, forbs,
Grasses, forbs
Deciduous Hardwoods


a Mean annual temperature and rainfall data are 30-year (1971-2000) means from the nearest weather monitoring station of the
National Oceanic and Atmospheric Administration.









Table 3-2: Soil characteristics of the five habitats examined. Means + S.D, different
letters within the same column signify significant differences by Tukey HSD.
Df = 1, 25.
Habitat % Moisture %C %N C:N
ME 79.29 + 18.66 a 32.82 + 8.81 a 2.05 + 0.71 a 16.54 + 1.98 c
SC 40.74+ 8.82b 8.81 + 9.49 bc 0.391 + 0.18 b 20.26 + 10.11 bc
SA 43.28+ 5.49bc 9.37 + 3.68 c 0.432 + 0.15 b 21.91+ 6.69 bc
RU 28.34 + 7.68 c 12.08 + 2.02 b 0.377 + 0.045 b 32.61 + 7.39 a
AR 22.20 + 9.03 c 3.80 + 1.42 d 0.158 + 0.076 c 27.56 + 10.86 ab


Table 3-3: Relationship of total PLFA and fungal:bacterial ratio with % moisture, %C,
%N and C:N using quadratic fit for total PLFA and simple linear regression
for fungal = biomass ratio (n = 30). All significant models had p-value <
0.005 except %C vs. fungal : bacterial ratio which was significant at p = 0.03.
Total PLFA Fungal : Bacterial Ratio
R2 Direction r2 Direction
% Moisture 0.32 Convex 0.70 Negative
% C 0.45 Convex 0.26 Negative
% N 0.48 Convex 0.48 Negative
C:N NS NS 0.52 Positive

Table 3-4: ANOVA results for the effects of habitat, invasion-status and their interaction
on the relative representation of Gram and Gram + biomarkers.
Gram Gram +
Effect Df F P F P
Model 9,20 26.07 <0.0001 17.96 <0.0001
Habitat 4 45.69 <0.0001 38.13 <0.0001
Invasion 1 4.23 0.0529 2.27 0.15
Hab x Inv 4 11.91 <0.00001 1.72 0.19


Table 3-5: Correlation between oN, %C, and %moisture and the relative representation
of Gram and Gram + biomarkers (n=30). All comparisons were significant
at p <0.01.
Gram Gram +
r2 Direction r2 Direction
%C 0.37 Positive 0.32 Negative
%N 0.31 Positive 0.19 Negative
%moisture 0.21 Positive NS NS











Table 3-6: Results of MANOVA for the effects of habitat, invasion and their interaction
on principle component axes 1 and 2 scores from the ordination of 23 PLFAs.
Effect DF F P
Model 9, 20 28.4 <0.0001
Habitat 4 61.6 <0.0001
Invasion 1 5.74 0.03
Hab Inv 4 0.92 0.47


Table 3-7: Loadings for the first two axes in a principle components analysis of 23
common PLFAs extracted from soil samples. Bold face indicates loadings >
0.3.


Fatty Acid
Gram +
il5:0
al5:0
i16:0
i17:0
a17:0
16:0 10ME
Gram -
15:0
16:1w7c
16:1w5c
cyl7:0
16:1 2 OH
17:0 10ME
18:1w7c
cyl9:0
Fungi
18:2w6c
18:1w9c
Microeukaryote
20:4
Actinomycete
18:0 10ME
No classification
ill:0 30H
15:0 30H
14:0
16:0
18:0


PCA 1 Loading

0.1311
0.1980
0.2952
0.1958
0.2146
0.2456


-0.1103
-0.2922
0.0164
-0.2140
0.3478
0.2610
-0.2419
0.1892

-0.0881
-0.1377

0.0797

0.2080

0.0480
0.1190
-0.1978
-0.3237
-0.2192


PCA 2 Loading

0.1373
-0.0118
0.1035
-0.1475
-0.1363
-0.2619

0.2806
-0.2198
-0.3532
-0.3141
0.0410
0.1541
-0.1602
0.2399

0.3527
0.2555


0.1116

0.0591

-0.3176
0.0696
-0.0207
0.1680
0.2497










Table 3-8: Correlations (r) of soil characteristics with the first two axes from the
ordination of 23 PLFAs.
Soil characteristic PCA 1 PCA 2
% N -0.135 -0.634
% C -0.250 -0.538
% Moisture -0.025 -0.737
C:N -0.331 0.388


Table 3-9: The results of MANOVA for the effects of habitat, invasion and their
interaction on principle component axes 1 and 2 scores derived from an
ordination of metabolic activity of soil microbes on 95 substrates.
Effect DF F P
Model 7, 16 2.91 0.037
Habitat 3 6.67 0.004
Invasion 1 0.32 0.57
Hab x In 3 0.007 0.99






52





160
A
140 A

., 120 A

100
E
80

E 60 -
E B

S 40 -

20

0-
ME SC SA RU AR
Wet- Habitat ->Dry


Figure 3-1: Mean soil microbial community total PLFA (nmol g1 + S.D.) for five habitats
prone to invasion by five non-native species. Black bars = non-invaded; white
bars = invaded. Habitat types marked by different letters are significantly
different from one another by Tukey's HSD (alpha = 0.05).


















C1


C




K


ME SC SA RU AR


Wet<-


Habitat


->Dry


Figure 3-2: Fungal:bacterial ratios (+ S.D.) for five habitats prone to invasion by five
non-native species. Different letters over bars indicate significant differences
between habitats as determined by Tukey's HSD (alpha = 0.05). Black bars =
non-invaded; white bars = invaded.







54

50
a A A
0
E
W 40
,, B
B
.B
E 30


20
C
20


10


0
ME SC SA RU AR
Wett- Habitat ->Dry

30
b
E A A
25

_1
20 B



C 15













Figure 3-3: Relative representation (% of total nmoles extracted) of PLFAs across
habitats and invasion-status, a) Gram + and, b) Gram PLFA biomarkers
(mean + S.D.) Black bars = non-invaded; white bars = invaded. Different
15-



















letters over bars indicate significant differences between habitats as
determined by Tukey's HSD (alpha = 0.05) after pooling invaded and non-
invaded areas within a habitat. Significant differences between invasion-
status within a community by Bonferroni-corrected t-tests are indicated by
asterisks (* = p < 0.05; ** p < 0.01).






































PCA 1 (22.3%)


Figure 3-4: Mean (+ S.D.) principle components scores by habitat and invasion-status
from ordination of 23 most common microbial PLFAs found in soil samples.
Downward triangles = ME, circles = SC, diamonds = SP, upward triangles =
RU, squares = AR habitats. Open symbols = non-invaded, closed symbols =
invaded.


T


* Schinus invaded
O Schinus non-invaded
v Melaleuca invaded
v Melaleuca non-invaded
* Ardisia invaded
E Ardisia non-invaded
* Sapium invaded
O Sapium non-invaded
A Ruellia Invaded
A Ruellia non-invaded


I






56




40 A

35
B
O 30 C
E C
25 B BC
S25 -

20 -



SAB
i10 BC B B A A B B
BC I IB B A A
C C



i11:0 30H 16:0 i16:0 16:1w5c cy17:0
Fatty Acids

Ubiquitous Gram + Gram -
Figure 3-5: Mean (+ S.D.) relative abundance of five PLFAs that had high loadings on
the principle component axes that showed significant habitat effects, but no
invasion effect. Solid bars = ME, open bars = SC, diagonal stripes = SA,
stipled bars = RU, horizontal stripes = AR. Different letters over bars indicate
significant differences between habitats as determined by Tukey's HSD (alpha
= 0.05).













a
8
A
7 -
o AB
E
6--
0
5 5- BC
Gram biomarker, b) 16:7c, Gram biomarker, and c) 18:26c, fungal
4
-u C
3

2-




0
ME SC SA RU AR

Wet<- Habitat ->Dry


Figure 3-6: Mean (+ S.D.) relative abundance of PLFAs with high loadings (>0.3) in the
PCA analysis showing significant habitat and invasion effects. a) 16:1 20H,
Gram biomarker, b) 16: lco7c, Gram biomarker, and c) 18:2co6c, fungal
biomarker. Black bars = non-invaded, open bars = invaded. Different letters
over bars indicate significant differences between habitats as determined by
Tukey's HSD (alpha = 0.05) after pooling invaded and non-invaded areas
within a habitat. Significant differences between invasion-status within a
community by Bonferroni-corrected t-tests are indicated by asterisks (* = p <
0.05; ** p <0.01).












8

o 7
E
6

r5

84
Ta
3

2


0
1



0


ME SC SA RU AR
Habitat


Wet<-


Wet<-


ME SC SA RU AR
Habitat


Figure 3-6. Continued


-Dry


->Dry






































-10 -8 -6 -4 -2 0 2 4 6 8 10

PCA 1 (26%)
Figure 3-7: Mean (+ SD) principle components scores by habitat and invasion-status
from the ordination of metabolism of 95 carbon sources by soil microbial
communities. Symbols are as described in Figure 3-4.


I














CHAPTER 4
LINKS BETWEEN LITTER QUALITY, DECOMPOSITION AND MICROBIAL
COMMUNITY COMPOSITION ON NATIVE AND NON-NATIVE PLANT LITTER

Introduction

The decomposition of plant litter is dependant on climate, substrate quality and

decomposers. Within a given climate, leaf chemical composition is an excellent predictor

or decomposition. The optimal measure of litter quality as a predictor of decomposition

depends upon the system examined and the length of the study. Nitrogen content and

carbon to nitrogen ratio (C:N) are often positively and negatively correlated with

decomposition rates, respectively, in studies of short duration and in species or soils with

a high percentage of more labile carbon fractions (Flanagan and van Cleve 1983, Pastor

et al. 1987, Taylor et al. 1989). Other studies have shown that carbon quality, as

measured by proportional representation of labile and recalcitrant substrates, to be the

major factor limiting decomposition (Meentemeyer 1978, Berg and Taum 1991, Hobbie

1996). Lignin is one such recalcitrant compound degraded by only a limited number of

organisms, predominately brown- and white-rot Basidiomycete fungi, can degrade it. In

many cases, lignin:N is the best integrator of litter quality for the duration of its

decomposition (Melillo et al. 1982).

Suites of litter quality characteristics tend to be correlated with one another as a

function of habitat quality. In low-resource habitats, species tend to have low nutrient

contents, high carbon-based structural defenses, and long-lived leaves (Chapin 1980).

Conversely, species from high-resource habitats have higher nutrient contents, lower









levels of defense, and shorter-lived leaves (Chapin 1980, Grime et al. 1996, Wright et al.

2001). Moisture levels in soil can indirectly affect resource availability. At high soil

moisture contents, oxygen becomes limiting and decreases decomposition (Haynes

1986). These lowered oxygen levels lower nutrient availability and reinforce slow

decomposition rates by leading to suites of leaf traits that result in slower decomposition

rates.

While the role of substrate quality in decomposition is well established, the role of

the microbial community has been less studied and, in general, is treated as a 'black box.'

It seems likely, however, that different communities of microbial decomposers would be

found on litter of different quality and that microbes are probably responding to the same

litter quality factors that control decomposition rates. As bacteria have higher nitrogen

requirements and faster growth and reproduction rates than fungi, bacteria might be

expected to dominate on litter with high nitrogen and low C:N. Fungi, conversely, due to

their slower growth, lower nutrient requirements and ability to decompose lignin, might

be expected to dominate litter with high C:N and lignin:N. Changes in litter quality over

the duration of decomposition will also likely control the succession of microbes on the

litter. Early studies of litter decomposers have shown that the ratio of fungi to bacteria

increased with increasing C:N (Witkamp 1963, 1966).

Previous studies of decomposer microbes have primarily relied upon culture-based

techniques. As only an estimated <1% of soil microbes are culturable (Torsvik et al.

1996, Atlas and Bartha 1998), culture-based techniques are limited in their ability to

examine the microbial community. Phospholipid fatty acid (PLFA) analysis offers a non-

culture based technique that provides a measure of the living microbial community.









PLFA analysis has the advantage of providing information on specific groups of

microbes through the use of signature fatty acids as well as providing an overall

"fingerprint." Thus, not only can functional or taxonomic groups be examined, but also

the similarities (or dissimilarities) of different microbial communities can be compared.

We applied PLFA analysis as a new approach to examine decomposer microbes on

different plant litters over time.

Differences between plant species, by way of their chemical composition affecting

litter quality, are known to affect decomposition rate and we hypothesize it will also

determine the composition of the decomposer community. Plant community composition

is changing globally with the spread of non-native species and such changes can lead to

changes in ecosystem function (Vitousek and Walker 1989, Mack and D'Antonio 2003,

Allison and Vitousek 2004). While a consensus predictive framework for determining

which invaders will be more likely to have ecosystem-level impacts has not been

developed, invaders that are qualitatively different from native species are believed to be

more likely to alter ecosystem processes (Chapin et al. 1996, D'Antonio et al. 1999).

Those species that have traits that overlap with native species are expected to have

limited or slower impacts on ecosystem traits (Mack and D'Antonio 2003). As litter

quality is a continuum in which both native and exotic species are distributed, we would

expect that the potential impacts of exotic species invasion on decomposition and

microbial community composition would be determined by its relative position on this

continuum. The evolution of increased competitive ability (EICA) hypothesis predicts

that due to the lack of herbivore pressure in their new range, invasive plants evolve to









invest less in defense (Blossey and Notzold 1995) potentially resulting in higher quality

litter in invaders.

The primary goal of this study was to examine the links between decomposition

and microbial community composition through litter quality. To that end, we examined

the decomposition of litter of twenty species of plants, both native and exotic, from a

variety of habitats varying in litter quality, in a common site. By including a variety of

plant species from a variety of habitats, we attempted to sample a broad range of litter

quality. We hypothesized that species identity of the litter (hereafter "litter species")

significantly affects decomposition rates in relation to the litter species' typical habitats,

leaf longevity, and other leaf functional traits. We hypothesized that the same litter

quality factors that control decomposition in this site should also control the composition

of the microbial community. Litters of higher quality should support a higher total

biomass with proportionally more bacteria, especially Gram-negative bacteria, while

fungi should dominate litters of lower quality. The difference in microbial community

composition between plant litters should decrease as decomposition proceeds and

remaining litter is dominated by more recalcitrant substrates. Finally, because we

hypothesize that the same litter quality factors control both decomposition rate and

microbial community composition, the correlation between them should be high.

Methods

Litter Collection and Experimental Design

We chose 20 plant species with the goals of representing 1) a broad range of litter

quality as measured by leaf habit and lifespan, nitrogen content and carbon fractions and

2) non-native and native species from a variety of habitats (Table 4-1). Leaf litter of each

species was collected from a minimum of five sites and a minimum of five individuals









per site between the months of September and December. Only leaves that fell freely

from shaken plants or had a clear abscission zone were collected; obviously green leaves

and leaves with heavy herbivore damage were excluded. Litter was pooled by species

and dried at room temperature for a minimum of four weeks prior to litter bag

construction.

Litter bags were constructed of 1-mm fiberglass window screen and filled with 5 g

of air-dried litter. Subsamples of air-dried litter were weighed, dried at 600C and re-

weighed to determine initial dry mass of litter bags. A total of 30 litter bags per species

were made to allow for 5 replicates at each of 6 collection dates. We strung a total of 20

litter bags (1 per species) onto a nylon line for each harvest date. Litter bags were placed

in a common hardwood-dominated forest (29040'N, 8209'W; mean annual rainfall 1200

mm; mean annual temperature 20.30C) in February 2004. Six lines of litter bags,

representing the six sampling dates, were placed at 5 randomly determined locations

within the study site. Lines radiated out from a central flag on a litter layer dominated by

Quercus nigra, Quercus hemispherica, and Pinus taeda.

One line from each replicate was collected at 28, 57, 112, 180, 238, and 319 days.

Litter bags were transported to the lab where the exterior of the bags were brushed free of

adhering soil. Roots, soil, invertebrates and frass were removed and the remaining litter

weighed. A subsample of litter was immediately frozen for later phospholipid analysis;

the remaining litter was weighed, dried at 60 OC and reweighed after 5 days of drying.

Dried samples were ground on a Wiley Mill (Thomas Scientific) through a 40 mesh.

Carbon and nitrogen contents of initial litter and litter bags was determined by

combustion on an elemental analyzer (Costech, Inc., Valencia, CA). Multiplying









nitrogen concentration by litter mass and dividing by the initial mass of nitrogen

determined percent initial nitrogen remaining. Carbon fractions were extracted with

increasingly acidic solutions (van Soest 1963) using an Ankom 220 Fiber Analyzer

(Ankom, Macadon, NY). Neutral detergent removed non-polar extracts (fats, oils waxes)

and soluble cell contents (carbohydrates, starch, non-bound proteins) and comprises the

non-polar fraction (NPE). Dilute acid detergent removed hemi-cellulose and bound

proteins comprising the water-soluble fraction (WS). Cellulose, the acid-soluble fraction

(AS), was separated from lignin with 72% H2SO4. The lignin fraction was corrected for

ash content by ashing samples at 500 OC after the sulfuric acid step. Carbon fractions

were expressed as a percentage of total mass.

Phospholipid Fatty Acid Analysis

Prior to analysis, frozen samples were ground with a Wiley Mill to pass through a

40-mesh. All glassware was heated at 5000C for 5 hours and PTFE caps were rinsed with

hexane to remove organic. Phospholipids were extracted from litter using the methods

of Wilkinson et al (2002). Lipids were extracted from 250 mg litter samples in two 30-

minute baths in a 37C water bath in a single-phase phosphate buffered methane-

chloroform solvent. After each extraction, the supernatant liquid was transferred to a

second test tube. The phase of the second test tube was broken by the addition of 4 ml of

chloroform and 4 ml of buffer. After vortexing, the phases were allowed to separate

overnight. The organic phase was fractionated on silicic acid columns and the

phospholipid fraction collected and transesterified to fatty acid methyl esters (FAMEs) by

mild alkaline methanolysis (Findlay and Dobbs 1993). Samples were resuspended in

hexane with 19:0 fatty acid as an internal standard and analyzed with an Agilent









Technologies 6890 gas chromatograph (Palo Alto, CA) with a 25m Ultra 2 phenyl methyl

silicone column. The temperature program increased from 1700C to 2700C at 50C per

minute. Peaks were identified with MIDI peak identification software (MIDI, Inc.,

Newark, DE) and by co-elution with standards. Fatty acids were defined in terms of total

number of carbon atoms : number of double bonds. The position of the double bond

from the methyl end of the molecule is signified by the symbol co followed by the carbon

position; cis and trans geometry are referred to by "c" and "t." The prefixes "i" and "a"

signify and iso and anteiso branching; "cy" signifies a cyclopropyl fatty acid.

Statistical Analysis

In order to compare litter decay rates among species in relation to litter quality,

decomposition rate constants (k) for each litter type were determined using a negative

exponential model:

ln(Xt) = In (Xo) kt

where Xt equals the amount of mass left at time = t, Xo is the initial mass of litter,

and t = time in years (Olson 1963). In order to improve normality, k's were log

transformed and subjected to ANOVA examine differences among species. Tukey's HSD

was used to examine differences between species. To examine difference between leaf

lifespan categories and habitats, species averages were subjected to a Kruskal-Wallace

rank test, as there were uneven sample sizes among groups and unequal variances. The

relationship between decomposition rate and litter quality was examined in two ways.

The individual effects of mean initial %N, %NPE, %WS, %AS, %lignin, C:N and

lignin:N on mean decomposition constants were examined in simple regressions. As are

auto-correlated and cannot be consider as independent of each other, we sought to create

an integrative measure of overall litter quality. Therefore, we ordinated the average









initial values of %N, C:N, %NPE, %WS, %AS, %lignin, and lignin:N for each species in

a principle components analysis. The first axis from that PCA explained 56% of the

variance in litter quality data and we use this axis as a proxy for a litter quality ranking

and refer to it as the "leaf chemistry axis" herein. Average species k's were then

regressed against this leaf chemistry axis.

Similarity of microbial communities was examined using both constrained

(canonical correspondence ordination, CCA) and unconstrained (principle components

analysis, PCA) ordination. We performed these two types of ordinations because

comparison of their result is informative. Constrained ordination is best used in

ordinations when a set of independent environmental variables is believed to be

structuring the community. As we hypothesize that microbial communities should be

structured by litter quality characteristics, constrained ordination allows us to directly

examine how litter quality is structuring microbial communities as the ordination axes are

constrained to these independent variables. Unconstrained ordination, conversely,

extracts the major variation in community data irrespective of any environmental

variables. Therefore, if a PCA and CCA give similar results and explain a similar

percentage of variance in the community data, it is assumed that the environmental

variables examined are primarily responsible for the structuring of the community.

Additionally, as CCA is subject to the same limitations as multiple regression and CCA

can be sensitive to noise in the environmental matrix (McCune 1997), use of

unconstrained ordination along with constrained ordination is often suggested

(McGarigal et al. 2000).









We ran PCA and CCA ordinations on all samples simultaneously and on samples

from each time point individually. We performed both types of analyses so that we could

examine overall changes and controls of microbial community composition over

decomposition and the differences and primary controls on microbial community

composition at a given sampling date. Only the 17 PLFAs that comprised at least 1% of

the total PLFAs extracted were included in the multivariate analyses. The PCA analysis

was performed only on the main matrix of the 17 PLFA values. The CCA analysis was

performed on the main matrix and a second matrix containing the 7 measures of litter

quality, initial mass remaining (%IMR), and percent moisture. Individual PLFAs were

log-transformed to improve normality. MANOVA was performed on PCA and CCA

scores to determine the effects of time, litter species and litter species*time. The

relationship between litter quality and functional groups of microbes (terminally

branched Gram-positive bacterial PLFAs: i14:0 + i15:0 + a15:0 + i16:0 + a16:0 + i17:0 +

al7:0; monounsaturated Gram-negative bacterial PLFAs: 16:1co7 + 17:1co7 + 18:1co7;

cyclopropyl Gram-negative bacterial PLFAs: cyl7:0 + cyl9:0; and fungi: 18:2co6),

fungal:bacterial ratio (18:2o6) / (il5:0 + a15:0 + 15:0 + i16:0 + 16:lo05c + i17:0 + a17:0

+ cyl7:0 + 17:0 + 18:1co7c + cyl9:0), and total biomass was examined in simple

regressions for each time point. When necessary, PLFAs were transformed to achieve

normality.

All regressions, analysis of variance, and non-parametric tests were performed in

JMPIN 4.0 (SAS Institute 2000, Cary, NC) while PCA and CCAs were performed in PC-

ORD 4.20 (MJM Software, Gleneden Beach, OR 1999).









Results

Decomposition of Litter

Most species showed two periods of rapid decomposition-during the initial 28

days and again between 112 and 238 days, corresponding with the wet season (Figure 4-

1). Decomposition rate constants (k) ranged from < 0.4 to nearly 2.0 yr-1 with significant

differences among species (p < 0.0001, F19,80 = 23.47; Table 2). Leaf lifespan class of the

litter species had no effect on decomposition rate (x2 = 2.19, df = 2, p = 0.33). Habitat

affiliation of the litter species, however, did have a significant effect on decomposition

rate (x = 9.72, df= 2, p = 0.0077) such that decomposition rate was lowest in species

from dry habitats and higher for litter species from wetter habitats. However, some

individual species from "wet" habitats (e.g. Juncus, Taxodium, and Acer) had relatively

slow decomposition rates (Table 4-1 and 4-2).

There was also a significant difference in decomposition constants between native

and non-native species (F1,98 = 47.09, p < 0.0001); this result, however, should be

interpreted with caution as the result would likely change with different representative

native and exotic species. When the non-native species are compared with native

dominants in the plant communities they invade, Ruellia, Causurina, Schinus, and

Sapium have higher decomposition rates, Imperata had comparable rates with native

dominants, and Ardisia, which can co-occur with several of the deciduous species with

moderate decomposition rates, was comparable with those species and faster than Pinus

with which it also co-occurs (Tables 4-1 and 4-2).

We also examined the relationship between initial litter quality and decomposition

rates. Of the seven measures of litter quality examined, the concentration of non-polar

extracts and lignin:N individually explained the greatest amount variation. Lignin and









nitrogen contents as well as the relative availability of nitrogen (C:N), were also

significantly correlated with decomposition rate (Table 4-3). According to a backwards

stepwise regression with %N, %NPE, %lignin, C:N and lignin:N, the variables %lignin

and C:N were retained and produced the following model:

Ln(k) = 0.650 0.0275(%lignin) 0.00907(C:N)

This model explained 60% of the variation (F2,17 = 12.76, p = 0.0003). Multiple

regression models can be problematic because of collinearity of predictor variables and

the order in which these variables are added or removed in stepwise regression. As litter

characteristics are known to co-vary, they were analyzed by principle components

analysis. The first PCA axis from this analysis explained 55.6% of the total variance

(eigenvalue = 3.89), with high loadings of C:N, %NPE, %WS, and lignin:N. Thus PCA

1 can be interpreted as a litter quality axis in which low values indicate high litter quality

i.e., low C:N, lignin:N, %WS, and high %NPE) and high values indicate low litter quality

(Table 4-4). Decomposition rate was negatively correlated with the leaf chemistry axis

(Fi,18 = 17.57, p = 0.0005, r2 = 0.49).

Nitrogen concentration of litter increased over time (Figure 4-2a). Approximately

half of the species showed no change or net increase of nitrogen relative to the initial total

with time, indicating net immobilization of nitrogen. The three species with the fastest

decomposition rates (Ruellia, Sapium, and Taxodium, Table 4-2) showed the largest

nitrogen mobilization (= net nitrogen loss, Fig. 4-2b). Immobilization varied over the

course of the study, with several species showing increased mobilization at time points

with higher decomposition (Figure 4-1 and 4-2b).









Microbial Community Composition

We examined the structure of microbial communities of eleven species at 28, 57,

and 238 days by multivariate analysis of PLFAs. We performed both PCA and CCA

analysis on samples from all sample dates simultaneously and then each sample date

individually. The PCA analysis of all sample dates explained a greater proportion of the

total variance (52% in the first 2 PCA axes) than the CCA analysis (32.3% of variance in

first 2 CCA axes, Figure 4-3, a-b). This indicates that while the leaf litter quality

variables we measured were important, other, unmeasured variables were also important

to the structuring of microbial communities.

In the PCA analysis, there is a large cluster of samples with poor litter quality from

t = 1 and 2 samples dates and a smaller cluster composed of higher litter quality from

primarily t = 2 samples and two t = 1 samples (Ruellia and Causurina). By t = 5,

microbial communities were beginning to converge as indicated by the ellipse indicating

9 of the 11 litter species at t = 5. The CCA analysis allows differentiation of PLFA

composition (i.e., microbial community composition) on the basis of variation in litter

quality. Like the PCA, CCA axis 1 differentiates microbial communities of the dry

periods (t = 1 and 2) with low CCA-1 scores, from t = 5 microbial communities from wet

periods with high scores. CCA axis 2, in contrast, differentiates microbial communities in

relation to litter quality, such that species with high scores on the leaf chemistry axis are

associated with low CCA axis 2 scores (Fig. 4-3b), especially among samples from t = 1

and t = 2. In MANOVA of PCA and CCA axes 1 and 2 scores, time, species and time *

species effects were all significant (Table 4-6).

When ordination analyses were done for each sampling period separately, the first

two axes of the PCA analyses of individual sampling dates again described more of the









total variation than the first two axes of the CCA analyses (time 1: 50.5% vs. 36.6%, time

2: 51% vs. 35.9%, time 5: 64% vs. 38.2%). Constraining the axes by litter quality

variables in CCA altered the distribution of litter species in species-score space (Figure 4-

4, a-f).

In the PCA oft = 1, Ruellia was most different from other litter species, but there

was no obvious pattern of microbial community composition in relation to litter quality

(Figure 4-4a). The first axis was composed of high loadings for two Gram PLFAs and

the fungal PLFA, while the second axis was composed of negative Gram + PLFA

loadings (Table 4-7). The simultaneous analysis of litter quality in the CCA for t = 1

microbial communities isolated the species with high litter quality, Ruellia, Sapium and

Schinus, from the remaining samples on the basis of their higher non-polar extract and

nitrogen contents which were negatively loaded on CCA axis 1(Figure 4-4b). Pinus and

Juncus were very different from other litter species due to their high amount of initial

mass remaining, and high lignin:N and water-soluble fiber content, respectively.

In the PCA oft = 2 microbial communities, Gram + PLFAs were again negatively

loaded on the second axis while saturated fatty acids, fungi and two Gram PLFAs were

positively loaded on the first axis (Table 4-7). Inclusion of litter quality in the CCA for t

= 2 samples, did not drastically change the relative position of litter species to one

another (Figure 4-4, c-d). The primary difference between the t = 2 PCA and CCA was a

greater separation of Acer and Ardisia microbial communities from one another and a

better separation of poor litter quality species, Aristida, Imperata and Juncus, from the

remaining litter by CCA. The separation of microbial communities in the CCA was









driven by the positive correlation of NPE on CCA axes 1 and 2, and the positive

correlation of WS, AS, and initial mass remaining on CCA axis 1 (Table 4-7).

At t = 5, there was generally less separation of microbial communities on high and

moderate litter quality (Figure 4-4, e, f). Fungi and some saturated PLFAs were

negatively loaded on PCA 1 and two Gram PLFAs and one Gram + PLFA were

negatively loaded on PCA 2 (Table 4-7). Inclusion of litter quality in the CCA better

separated low quality litters Pinus, Aristida and Juncus from one another and the

remaining litter species on the basis of initial mass remaining and C:N (Figure 4-4f).

We explored the effect of litter quality on microbial groups in simple regressions at

each collection date. In general, the strength of relationship between litter quality

variables and microbial groups increased across time (Tables 4-8 through 4-13). The

fungal PLFA and total PLFA, however, had strongest correlations with litter quality at t =

2. Three litter quality factors, oN, C:N and %IMR, explained the greatest proportion of

variance across all microbial groups. C:N and %IMR were negatively correlated with

monosaturated and cyclopropyl Gram bacteria, Gram + bacteria, fungi, and total PLFA

(nmol g-) and positively correlated with fungal : bacterial ratio. The opposite trends

were true for oN.

Some surprising trends relating to carbon fractions were identified. Lignin content

was not correlated with fungal PLFAs, but was positively correlated with Gram and

Gram + bacteria and total PLFA at t = 5, although these correlations were weak. On the

other hand, the relationship between lignin:N and microbial groups was as expected,

being negatively correlated with bacterial groups and positively correlated with

fungal:bactieral ratio. The relationship between water and acid soluble fractions and









different microbial groups was not consistent and changed depending upon collection

date examined. For example, water and acid soluble fractions were positively correlated

with Gram + bacteria at t = 2, but negatively correlated at t = 5. Similar to individual

measures of litter quality, the strength of correlation between the leaf chemistry axis

(Table 4-4) and microbial groups increased with time (Table 4-14).

Linking Litter Quality, Decomposition and Microbial Communities

Just as the leaf chemistry axis was correlated with decomposition rate (r2 = 0.49)

and microbial groups (r2 up to 0.77), it was highly correlated with microbial community

principle components axes. The strength of the correlation between leaf chemistry and

microbial community composition increased with time (Table 4-15). Leaf chemistry

explained 88% of the variation in the first microbial axis from the fifth collection date

and 65% of the variation in the first microbial axis from all samples ordinated

simultaneously. Microbial community composition also explained the greatest amount of

variation in decomposition rate constants (k) of any measures explored (r2 = 0.76 for t = 5

microbial communities). As with the relationship between microbial community and

litter chemistry, the strength of the correlation between microbial community and

decomposition increased with time (Table 4-15).

Discussion

In this study we endeavored to examine the role of plant litter quality in the

structuring of microbial communities and how those communities changed over time. In

our experimental design of using field common gardens, environmental variables other

than litter quality were standardized across litter species, although the environment did

not stay constant. In addition to change in litter quality as decomposition progressed, the

background moisture availability changed as the rainy season started in the middle of the









experiment. Indeed, some of the temporal differences in microbial community

composition as detected by PLFA biomarkers appear to be better explained by

differences in litter moisture than by chemical differences of the litter. Hence, sampling

date appeared to have the greatest effect on microbial community composition (Figure 4-

3a), but there were also significant differences among litter species due, in part, to

differences in litter characteristics (Figure 4-3b). While non-polar extracts, lignin and

lignin:N (Table 4-3) were the best individual predictors of decomposition rate, non-polar,

acid- and water-soluble carbon fractions and C:N (Figure 4-3b) were most important in

structuring microbial communities. What was particularly novel about this study was the

discovery that the best correlate of decomposition rate was not a measure of litter quality,

but the composition of the microbial community (r2 = 0.76 for correlation of PCA1 of the

microbial community at t = 5, Table 4-15).

Factors Controlling Microbial Community Composition

Time, litter species and litter species time all had a significant effect on microbial

community structure. The most obvious effect is that of time with samples collected at t

= 5 being most different from other sampling dates (Figure 4-3). The t = 5 samples were

collected on October 5, 2005 within four weeks of hurricanes Frances and Jeanne which

together produced 325 mm of rain in the area and these samples had much higher

moisture contents than samples collected at t = 1 and 2. The separation of the t = 5

group is due to an increase in Gram-negative bacterial PLFAs cyl7:0, cyl9:0, and

18:lw7c and Gram-positive bacterial PLFAs i15:0 and i16:0 in those samples. Other

studies have shown increases in Gram-positive, terminally branched, saturated fatty acids

such as i15:0, a15:0, i17:0, and a17:0 in soil and litter microbial communities under

flooded conditions (Bossio and Scow 1998, Nakamura et al. 2003). Cyclopropyl fatty









acids cyl7:0 and cyl9:0 have additionally been suggested as biomarkers for anaerobic

bacteria (Guckert et al. 1985, Vestal and White 1989), but see (Parkes and Taylor 1983,

Bossio and Scow 1998). In the only analysis of microbial communities of early

decomposition of plant litter in upland conditions of which we are aware, Wilkinson et al

(2002) found that al5:0 and cyclopropyl fatty acids were generally higher in regularly

watered samples, although the effect was dependant on litter species examined. In this

study, bacterial dispersal and activity were likely limited in the first two collections by

low moisture, while the saturating conditions of the fifth collection resulted in increased

bacterial biomarkers overall and especially putative anaerobic biomarkers.

While time had the largest effect on microbial communities in PC and CC analyses,

litter species and litter species*time interaction effects were also significant (Table 4-6).

Microbial communities on poor quality litter changed less from sample date to sample

date than did microbial communities on litter with high initial litter quality (Figure 4-3a).

Two alternative hypotheses may explain this pattern. Resource availability may be high

enough in high quality litter that decomposition and microbial succession can proceed

even under low moisture levels. Alternatively (but not necessarily exclusively), greater

changes in resource availability in high quality litter may be responsible for shifts in

microbial community composition.

Ordination of each sampling date individually shows litter species and quality

effects on microbial community structure. In early decomposition under low moisture

levels, high concentration of non-polar extracts and to a lesser extent nitrogen

concentration separated high quality litter microbial communities from moderate to low

quality litters with higher water and acid soluble fiber contents that retained a greater









proportion of the initial mass expressed as %IMR (Figure 4-4 b, d). The only litter

quality characteristic that was important to structuring microbial communities at all

samples dates was %IMR. Under the higher moisture levels oft = 5, moisture, C:N, and

%N had become important to the structuring of the microbial communities (Figure 4-4f).

Other studies of soil microbial communities have shown differentiation of microbial

communities from different plant communities differing in litter or soil C:N or nitrogen

availability (Eiland et al. 2001, Gallo et al. 2004, Leckie et al. 2004, Myers et al. 2001,

Waldrop and Firestone 2004a, b). It would appear from this study that moisture levels

and availability of labile carbon control early colonization of litter and as decomposition

proceeds, nitrogen availability becomes more important.

Examination of functional groups of microbes also shows that moisture levels

controlled colonization by microbes as litter quality had low discrimination of microbial

community composition at t = 1 and 2 (Tables 4-8 through 13). However, at t = 5 when

water was not limiting, %N, C:N and %IMR were generally good predictors for bacterial

functional groups, fungal : bacterial ratio, and total biomass as inferred by total microbial

PFLA. This is consistent with other studies of soil microbial communities with both

Gram-negative and Gram-positive bacteria generally increasing with nitrogen availability

(either through nitrogen fertilization or decrease in C:N ratio)(Eiland et al. 2001, Leckie

et al. 2004, Waldrop et al. 2004). The response of fungi, however, appears to be

equivocal with fungi or fungal:bacterial biomass having a negative response (Eiland et al.

2001, Leckie et al. 2004), no response (Waldrop et al. 2004) or, as seen in this study, a

positive response (Gallo et al. 2004) to increased nitrogen availability. Many of these

studies show that there is not always a consistent response of individual PFLA









biomarkers within a functional group to litter quality treatments (see especially Gallo et

al. 2004, Waldrop et al. 2004). We found this to be true in this study and the low amount

of variance explained for a given functional group by different carbon fractions is a result

of differential response of indicator fatty acids within a functional group.

Factors Controlling Decomposition Rate

Early decomposition is dominated by the leaching and decomposition of soluble

and low molecular weight compounds (Swift et al. 1979, Berg et al. 1982). In our study,

only 182 mm of rain fell in the first 112 days of the experiment, with 56% of that rain

falling before the first sampling date. Between samples 3 and 5 (129 days),

encompassing hurricanes Frances and Jeanne, nearly 900 mm of rain fell. This

distribution in rainfall is likely responsible for two major periods of leaching resulting in

greater relative mass losses at those points (Figure 4-1). Although a single exponential

model may not be the best model to explain mass loss, we chose this model to enable us

to have a single variable to examine the relationship between decomposition and litter

quality.

All litter quality factors examined except for acid-soluble fiber were correlated with

decomposition (Table 4-3). As with the microbial communities, it is difficult to

determine which measure of initial litter quality is most responsible for controlling

decomposition. However, across the three methods we used to examine relationship

between litter quality and decomposition (single regression, multiple regression and the

ordination of a leaf chemistry axis), lignin:N, C:N and non-polar carbon fractions were

important factors. This is consistent with a large body of literature that shows

decomposition rate is related to relative availability of carbon and nitrogen (Coulson and

Butterfield 1978, Melillo et al. 1982, Taylor et al. 1989). Recalcitrant fractions such as









lignin are not degraded in a significant way until later in decomposition when most of the

more labile fractions have been reduced significantly-more than 2 years into

decomposition in a study of Scots pine litter (Berg et al. 1982). As this study was

conducted for less than a year, it is not surprising that the non-polar fractions, which are

readily available for fast-growing microorganisms, were highly correlated with

decomposition rate.

The novel information from this study is the linking of litter quality, microbial

community composition and decomposition rate. Non-polar fractions and C:N, which

were strongly correlated with decomposition rate, were also the litter quality measures

most strongly correlated with CCA axes ordinating the microbial communities (Figure 4-

3b). Additionally, the leaf chemistry axis explained a large proportion of the variation in

decomposition rate, functional groups, and overall microbial community composition

(Tables 4-14 and 4-15). The best predictor of decomposition rate, however, was not any

measure of litter quality. Instead, microbial community composition explained the

greatest amount of variation in decomposition rate (76%). These results indicate that

treating microbes as a black box may limit our understanding of controls on

decomposition.

Implications for Invasion

We believe that our study indicates that invaders will be more likely have a large

impact on microbial community composition or nutrient release if the invader's litter

quality, especially C:N, lignin:N and non-polar extracts, is significantly different from the

native community as these were the best predictors of decomposition and microbial

community composition. Invasive species may have a greater potential to alter microbial

communities and ecosystem process if they invest less in defense. Upon introduction to a









new range, several invasive species have been shown to evolve toward less allocation to

defense (EICA hypothesis) as a result of decreased herbivore pressure in their introduced

range (Blossey and Notzold 1995, Siemann and Rogers 2003). Such reductions in

defense allocation would result in high litter quality of invaders, higher decomposition

rates and a shift from oligotrophic (i.e., able to live on low nutrient levels) to copiotrophic

(i.e. requiring high levels of high quality nutrients, sensu Ohta and Hattori 1983)

microbial communities. These shifts could lead to positive feedbacks on invasion by

increasing nutrient availability (Ehrenfeld 2003).

Invasive species need to contribute a large proportion of the total litter inputs in

order to affect nutrient cycling on an ecosystem scale. In our study, although Ardisia and

particularly Ruellia have litter characteristics different from those in the plant

communities they invade, they are understory shrubs without a large leaf biomass and

thus will be less likely to affect nutrient cycling and microbial communities at the

ecosystem scale. Sapium, Causurina, Schinus, and Imperata, conversely, can form

monodominant stands and produce large amounts of litter that would enable them to

affect ecosystem-level change.

Conclusions

We investigated the differences in microbial community composition and

decomposition of litter among different species of litter in a common site. Even though

the plant litter was subjected to the same environment, with the same land-use history and

the same potential colonizing microbial community, composition of the microbial

community differed among litter species. These differences in microbial community

composition were a function of initial litter quality and litter quality at the time of

collection. Changes in litter quality due to both decomposition and changing






81


environmental factors, predominately moisture content, resulted in a significant effect of

time on microbial community composition. Microbial community composition was more

strongly correlated with decomposition rates than any measure of litter quality. This

study suggests that explicitly examining the microbial community rather than treating it

as a black box may improve our understanding of the controls of early-phase

decomposition.












Table 4-1: Characteristics of plant species whose litter was used in this experiment including leaf lifespan class (1 = <1 year, 2 1
year, 3 = > 1 year), leaf type (broadleaf, needle leaf, monocot), position on moisture gradient (dry, mesic, wet), and native
status, along with the chemical composition of freshly collected litter (senescent leaves that were easily abscised by
shaking).
Species Lifespan Type Habitat Native? C:N NPE(%) WS(%) AS(%) Lignin(%)

Acer rubrum 1 Broad Wet Yes 55.9 60.1 10.6 12.3 14.3

Ardisia crenata 3 Broad Mesic No 44.5 52.0 19.7 20.5 7.43

Aristida strict 2 Monocot Dry Yes 128.5 12.3 40.8 22.9 23.6

Caryaglabra 1 Broad Mesic Yes 38.8 55.5 12.2 18.5 13.6
00
Casuarina glauca 1 ** Mesic No 25.1 32.7 17.2 25.6 24.2

Imperata cylindrica 2 Monocot Dry No 97.2 19.2 31.0 38.4 10.9

Juncus roemerianus 2 Monocot Wet Yes 69.2 20.1 34.9 36.0 34.9

Liquidambar styricaflua 1 Broad Mesic Yes 51.9 57.4 9.21 16.2 16.9

Magnolia virginiana 3 Broad Mesic Yes 57.5 51.8 14.1 20.3 13.1

Pinus palustris 3 Needle Dry Yes 58.2 34.6 14.5 27.7 22.5

Quercus chapmannii 2 Broad Dry Yes 59.7 50.9 11.7 19.3 17.9















Table 4-1. Continued
Species

Quercus geminata

Quercus laevis

Quercus nigra

Ruellia brittoniana

Sabalpalmetto

Sapium sebiferum

Schinus terebinthifolius

Taxodium distichum

Typha latifolia


Lifespan

2

1

1

1

3

1

2

1

2


Type Habitat

Broad Dry

Broad Dry

Broad Mesic

Broad* Wet

Monocot Mesic

Broad Wet

Broad Mesic

Needle Wet

Monocot Wet


* Ruellia brittoniana was the only non-woody broad leaf examined. **Photosynthetic tissue of Casuarina gluaca,
this study, are modified stems, needle-like in appearance.


treated as leaves in


Native?

Yes

Yes

Yes

No

Yes

No

No

Yes

Yes


C:N

74.9

36.0

47.7

32.4

30.3

32.4

37.8

28.8

54.5


NPE(%)

42.4

32.0

44.2

72.2

25.27

73.7

61.1

45.9

21.4


WS(%)

12.9

11.9

13.6

9.29

20.2

9.85

5.05

12.9

23.4


AS(%)

25.1

26.6

20.7

10.2

33.9

10.5

11.9

15.1

17.0


Lignin(%)

19.1

29.2

21.3

7.92

19.1

5.02

21.6

25.9

37.5









Table 4-2: Mean and_SE for decomposition rates of native and non-native litters (n=5).
Different letters indicate significant differences (alpha = 0.05) between
species by Tukey HSD. The leaf chemistry axis is the score from the
ordination of initial plant litter chemistry. High scores indicate poor quality
(high lignin:N, C:N), and low scores indicate high quality (high non-polar
fraction, nitrogen). DF =19, 80.
Species Native? k (year') Leaf Chemistry
Axis Score
Mean SE
Sapium No 1.91a 0.0852 -2.70
Ruellia No 1.79a 0.150 -2.51
Causurina No 1.06b 0.0546 -1.26
Ardisia No 0.990bc 0.159 -1.14
Carya Yes 0.849bc 0.0311 -1.45
Liquidambar Yes 0.789bcd 0.0451 -0.80
Schinus No 0.724bcd 0.0354 -0.38
Q. nigra Yes 0.731bcd 0.0673 -1.48
Magnolia Yes 0.723bcd 0.0853 -0.64
Typha Yes 0.710bcd 0.0708 2.37
Sabal Yes 0.668bcd 0.0449 -0.49
Q. geminata Yes 0.655cde 0.0333 0.56
Acer Yes 0.647cde 0.0357 -0.81
Taxodium Yes 0.586cdef 0.0684 -1.16
Imperata No 0.531def 0.0506 2.26
Q. chapmannii Yes 0.404ef 0.0167 -0.36
Pinus Yes 0.401fg 0.0165 1.51
Juncus Yes 0.394fg 0.0170 3.49
Q. laevis Yes 0.388fg 0.0242 -0.30
Aristida No 0.387fg 0.159 5.28


Table 4-3: Result of simple linear regression for log-transformed decomposition
constants against individual variables of litter chemical composition. DF =
1,18. %NPE = non-polar fraction, %WS = water-soluble fraction, %AS =
acid-soluble fraction.
Litter Slope r2 P
variable
%N 0.583 0.24 0.028
C:N -0.0101 0.32 0.0088
%NPE 0.0173 0.45 0.0011
%WS -0.0214 0.19 0.052
%AS NS NS NS
%Lignin -0.0305 0.34 0.0066
Lignin:N -0.0193 0.47 0.0009











Table 4-4: Factor loadings for individual variables that contribute to litter quality on the
leaf chemistry PCA axis 1. In general, high leaf chemistry score indicates low
quality litter with low %N, high C:N, lowNPE, high %WS, high %lignin,
and high lignin:N ratio.
Litter Quality Loading
Measure
%N -0.353
C:N 0.431
%NPE -0.425
%WS 0.442
%AS -0.0490
%Lignin 0.282
Lignin:N 0.481


Table 4-5: Factor loadings of individual PLFAs upon the first two main axes of PCA
(PC1 and PC2) and CCA (CC1 and CC2). Loadings > 0.3 in bold
Fatty Acid PC 1 PC 2 CC 1 CC 2
Unclassified
14:0 0.245 -0.100 -0.011 0.086
15:0 0.331 0.059 0.314 -0.118
16:0 0.114 -0.417 -0.153 -0.033
17:0 0.154 -0.318 -0.123 0.058
18:0 0.179 -0.316 -0.133 -0.060
i19:0 -0.033 -0.068 -0.182 0.512
20:1co9c 0.006 -0.262 -0.267 -0.161
Gram +
i15:0 0.348 0.191 0.536 -0.077
al5:0 0.280 -0.189 0.036 -0.124
i16:0 0.351 0.136 0.371 -0.023
Gram -
16:lco7c 0.300 -0.188 0.039 0.059
17:1co8c 0.084 -0.337 -0.143 0.100
cyl7:0 0.343 0.165 0.476 -0.145
18:1 o7c 0.306 0.211 0.520 0.080
cy19:0 0.332 0.224 0.590 -0.015
Fungi
18:2co6c 0.041 -0.422 -0.188 0.068
18:lo09c 0.076 -0.025 0.050 0.033









Table 4-6: MANOVA summarizing the effects of time, litter species and litter
species*time on PCA and CCA axes 1 and 2 scores from ordination of all
samples simultaneously.


PCA CCA
F P F P
Model 9.01 <0.0001 27.14 <0.0001
Time 25.89 <0.0001 194.55 <0.0001
Species 14.50 <0.0001 34.13 <0.0001
Sp*Time 4.73 <0.0001 3.97 <0.0001
Note: Model df = 29, 112; Species df = 9, 112; Time df = 2, 112; Spe
112.


cies*time df= 18,


Table 4-7: Loadings for the first two PCA axes for PCA run separately for individual
sampling dates (t = 1, 2, and 5).
T=i T=2 T=5
Fatty Acid PCA 1 PCA 2 PCA 1 PCA 2 PCA 1 PCA 2
Unclassified
14:0 0.1307 0.3481 0.1745 0.2407 -0.1945 -0.0666
15:0 0.3287 0.0477 0.3158 -0.0603 -0.1978 -0.3765
16:0 0.3605 0.1458 0.3440 0.2246 -0.2842 -0.2803
17:0 0.2457 -0.2206 0.3597 -0.0408 -0.2059 -0.3638
18:0 0.2188 0.0340 0.3091 0.1865 -0.2990 -0.1028
i19:0 0.1531 0.3016 NA NA NA NA
20:1co9c 0.1751 -0.0434 0.1597 0.3788 -0.1396 -.0899
Gram +
i15:0 0.1022 -0.4039 0.1825 -0.4010 -0.3037 0.1943
a15:0 0.2567 -0.3499 0.1795 -0.4548 -0.2692 0.2642
i16:0 0.1533 -0.4933 0.1843 -0.4157 -0.2742 0.2461
Gram -
16:lco7c 0.3741 -0.0441 0.3399 -0.1433 -0.2952 0.2216
17:1co8c 0.3617 -0.0086 0.3184 0.0161 -0.1413 -.1089
cyl7:0 0.0299 -0.0451 0.1439 0.1822 -0.2756 0.2736
18:1om7c 0.1860 0.1305 0.1439 0.1822 -0.3068 0.1148
cyl9:0 -0.0550 0.1330 -0.0326 -0.1557 -0.3001 0.1010
Fungi
18:2co6c 0.3870 0.1301 0.3694 -0.0044 -0.0847 -.4984
18:1o9c 0.1536 0.3576 -0.0623 0.0443 -0.2768 -.2205









Table 4-8: Correlation between each litter quality variable and the amount of PLFA
(nmol) indicating monounsaturated Gram bacteria at times 1, 2, and 5. p <
0.05, ** p <0.01, *** p < 0.0001. %NPE = non-polar fraction, %WS = water-
soluble fraction, %AS = acid-soluble fraction, %IMR = initial mass


remaining, %moisture on a dry mass basis.
T=I T=2
Direction r2 Direction
+ 0.18** NS
0.23** NS
NS NS
NS NS
NS NS
NS NS
0.29*** NS
NS -NS
NS -NS


T=5
r2 Direction r2
+ 0.41***
-0.40***
+ 0.10*
-0.09*
-0.26***
+ 0.19**
-0.12**
0.45***
+ 0.09*


Table 4-9: Correlation between each litter quality variable and the amount of PLFA
(nmol) indicating branched Gram+ fatty acids at times 1, 2, and 5. p < 0.05,
** p <0.01, *** p < 0.0001. %NPE = non-polar fraction, %WS = water-
soluble fraction, %AS = acid-soluble fraction, %IMR = initial mass
remaining, %moisture on a dry mass basis.
T=I T=2 T=5
Direction r2 Direction r2 Direction r2
%N + 0.16** NS -+ 0.40***
C:N NS -NS -- 0.41***
%NPE NS -NS -NS
%WS + 0.12** + 0.17** 0.08*


%AS
%lignin
Lignin:N
%IMR
%moisture


0.05*
0.13**


0.10*
0.16**


0.18**
0.19**
0.06
0.33***
0.11**


%N
C:N
%NPE
%WS
%AS
%lignin
Lignin:N
%IMR
%moisture










Table 4-10: Correlation between each litter quality variable and the amount of PLFA
(nmol) indicating fungal fatty acid (18:2co6c) at times 1, 2, and 5. p < 0.05,
** p <0.01, *** p < 0.0001. %NPE = non-polar fraction, %WS = water-
soluble fraction, %AS = acid-soluble fraction, %IMR = initial mass
remaining, %moisture on a dry mass basis.
T=i T=2 T=5
Direction r2 Direction r2 Direction r2
% N + 0.09* + 0.47*** NS
C:N -0.10* -0.45*** NS
%NPE NS + 0.31*** NS
%WS NS -0.20** NS
%AS NS 0.32*** NS
%lignin NS NS NS
Lignin:N -0.12** 0.25*** NS
%IMR 0.43*** NS
%moisture NS 0.06* NS 0.29***


Table 4-11: Correlation between each litter quality variable and ratio fungal : bacterial
PLFAs at times 1, 2, and 5. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE=
non-polar fraction, %WS = water-soluble fraction, %AS = acid-soluble
fraction, %IMR = initial mass remaining, %moisture on a dry mass basis
T=I T=2 T=5
Direction r2 Direction r2 Direction r2
%N NS NS 0.41***
C:N + 0.11* NS + 0.32***
%NPE NS + 0.15** NS
%WS NS -0.21** NS
%AS NS 0.11* + 0.10*
%lignin NS + 0.05 NS
Lignin:N + 0.11* + 0.05 + 0.19**
%IMR + 0.07* NS + 0.37***
%moisture NS NS 0.27***










Table 4-12: Correlation between each litter quality variable and the total PLFA (nmol) at
times 1, 2, and 5. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = non-polar
fraction, %WS = water-soluble fraction, %AS = acid-soluble fraction, %IMR
= initial mass remaining, %moisture on a dry mass basis.
T=I T=2 T=5
Direction r2 Direction r2 Direction r2
%N + 015** + 041*** + 015**


C:N
%NPE
%WS
%AS
%lignin
Lignin:N
%IMR
%moisture


0.08*


0.39***
0.25***
0.26***
0.29***

0.11**
0.27***


0.14**
0.21**
0.20**
0.17**
0.08*

0.17**


Table 4-13: Correlation between each litter quality variable and the amount of PLFA
(nmol) indicating cyclopropyl Gram- bacteria at time 5. Only t = 5 is
presented as the large number of zeros at other time points did not allow for
parametric statistics. p < 0.05, ** p <0.01, *** p < 0.0001. %NPE = non-
polar fraction, %WS = water-soluble fraction, %AS = acid-soluble fraction,
%IMR = initial mass remaining, %moisture on a dry mass basis
Direction r2
%N + 0.26***
C:N 0.18**
%NPE + 0.05
%WS NS
%AS 0.23**
%lignin + 0.07*
Lignin:N NS
%IMR 0.28***
%moisture + 0.8*









Table 4-14: Coefficient of determination (r2) of abundance of PLFAs for microbial
functional groups with leaf chemistry scores at times 1, 2, and 5. Df time 2
and 5 = 1,8; dftime 1 = 1, 7. p < 0.05, ** p <0.01, *** p < 0.0001
Time = 1 Time = 2 Time = 5
Total PLFA NS 0.54** 0.62**
Fungal:Bacterial Ratio NS NS 0.49**
Fungi NS 0.53** NS
Gram+ NS NS 0.77**
Branched Gram- NS NS 0.70**
Cyclopropyl Gram- NS NS 0.60**
Note: All significant relationships were negative correlations except for fungal:bacterial
ratio which was positive.


Table 4-15: Coefficient of determination (r2) of correlations of microbial community
PCA axes (independent) vs.decomposition rate (k) dependantt), and leaf
chemistry axis (from ordination of initial litter quality) (independent) vs.
microbial community PCA axes (dependent). All relationships were
significant at p <0.05.
T=1 PCA1 T=2 PCA1 T=5 PCA1 All PCA1 All PCA2
K 0.53 0.62 0.76 0.76 0.58
Litter chemistry 0.44 0.52 0.88 0.65 0.53
Note: Microbial PCA vs. k df = 1,9 (except T=l PCA, df = 1,8); litter quality vs. PCA df
= 1,90 (except T=l PCA, df= 1,8).