MICROBIAL FUNCTIONAL ACTIVITY and DIVERSITY PATTERNS at MULTIPLE SPATIAL SCALES a Dissertation Submitted to Kent State Universit

Home , Amphisphaeria, Amphisphaeriaceae, Atheliaceae, Coriolopsis gallica, Dermea, Filobasidiales

MICROBIAL FUNCTIONAL ACTIVITY AND DIVERSITY PATTERNS AT MULTIPLE SPATIAL SCALES

A Dissertation submitted

to Kent State University in partial

fulfillment of the requirements for the

degree of Doctor of Philosophy

Larry M. Feinstein

August, 2012

Dissertation written by Larry M. Feinstein B.S., Wright State University, 1999 Ph.D., Kent State University, 2012

Approved by

, Chair, Doctoral Dissertation Committee Christopher B. Blackwood

, Members, Doctoral Dissertation Committee Laura G. Leff

Mark W. Kershner

Gwenn L. Volkert

Accepted by

, Chair, Department of Biological Sciences James L. Blank

, Dean, College of Arts and Science John R. D. Stalvey

TABLE OF CONTENTS

LIST OF FIGURES……………………………………………...... ……………….iv

LIST OF TABLES...... x

ACKNOWLEDGMENTS...... xiii

5 CHAPTER

1. Introduction………………...... ……………………………………...……..…....1

References...... 11

2. Assessment of Bias Associated with Incomplete Extraction of Microbial DNA

from Soil...... ………………..……….18

10 Abstract...... …….………………………………………………..….…...18

Introduction...... …………………………………..……………..……19

Methods...... ……………………………………………………….…...…21

Results...... ….....27

Discussion...... ………………………………………………………….43

15 Referneces...... ……………………………………………………….....47

3. Taxa-area Relationship and Neutral Dynamics Influence the Diversity of Fungal

Communities on Senesced Tree Leaves...... 55

Abstract...... ………………………………...... …………...…...55

Introduction...... ……………………………………...... …54

iii

20 Methods...... ………………………………………...…………………59

Results...... …………………………………………...... ……………..63

Discussion...... …………………………………………...... ………85

References...... ………………………………...... ……………..90

4. The Spatial Scaling of Fungal Diversity...... 101

25 Abstract...... ……………………………...... 101

Introduction...... …………...... …………….102

Methods...... …………...... …………………....105

Results...... ………………………...... …...113

Discussion...... …………...... ……………128

30 References...... …………………………...... ……...134

5. The Impact of Litter Diversity on Microbial Enzyme Activity, Biomass, and Leaf

Litter Decomposition...... 146

Abstract...... ………………...... ………………...... 146

Introduction...... ………………...... ….147

35 Methods...... ……………...... ……………………………….…….152

Results...... …………...... ……………………………...159

Discussion...... ………...... ……………………………………175

References...... …………...... ……………………………………183

6. Synthesis...... 199

40 References...... 206 iv

LIST OF FIGURES

CHAPTER 2. Assessment of Bias Associated with Incomplete Extraction of Microbial

DNA from Soil

Figure 1: Cumulative DNA yield in successive extractions conducted using four

45 cell lysis treatments. A) organic soil; B) clay soil; C) sand soil...... 29

Figure 2. PFGE gel of 1st and 6th extractions for clay and sand soils. The top row

of numbers refers to extraction treatment, the bottom row refers to extraction step.

Lane 9 contains lambda ladder PFG marker (visible bands are 48.5 kbp and 98

kb). Lane 10 contains lambda PstI marker (visible bands are 11.5, 5.1, 4.6, 2.8,

50 and 2.6 kb)...... 32

Figure 3. Cumulative microbial gene copies/g soil in successive extractions. A)

bacterial 16S gene; B) fungal 18S gene...... 34

Figure 4. Microbial gene copies/ng DNA for each extraction (not cumulative or

weighted average). Significant differences are described in the text. A) bacterial

55 16S gene; B) fungal 18S gene...... 36

Figure 5. Weighted average ratio of fungal to bacterial ribosomal gene copies.

Significant differences are described in the text...... 38

Figure 6. Canonical principal components plot of T-RFLP profiles derived from

redundancy analysis. Soil type and extraction are significant as described in text.

60 A) bacterial profiles; B) fungal profiles. “Wted ave-3” is the weighted average of

the first three T-RFLP profiles; “wted ave-6” is the weighted average of all six T-

RFLP profiles...... 40

CHAPTER 3. Taxa-area Relationship and Neutral Dynamics Influence the Diversity of

Fungal Communities on Senesced Tree Leaves

65 Supplementary Figure S1(a). Relative abundance charts show proportion of

community occupied by dominant (multi-colored) and non-dominant (black)

OTUs on vernal pool leaves...... 65

Supplementary Figure S1(b). Relative abundance charts show proportion of

community occupied by dominant (multi-colored) and non-dominant (black)

70 OTUs on upland leaves...... 66

Supplementary Figure S1(c). Relative abundance charts show proportion of

community occupied by dominant (multi-colored) and non-dominant (black)

OTUs on riparian leaves...... 67

Supplementary Figure S2(a). Rarefaction curves for diversity indices for

75 individual leaves located at Upland Sites L6 and 14...... 69

Supplementary Figure S2(b). Rarefaction curves for diversity indices for

individual leaves located at Riparian Sites J1 and Q1...... 70

Supplementary Figure S2(c). Rarefaction curves for diversity indices for

individual leaves located at Vernal Pool Sites J3 and H6...... 71

80 Figure 1. Mean OTU richness (S), Shannon diversity (H’), and Simpson diversity

(1/D) values for large vs. small sugar maple and beech leaves gathered from 3

temperate forest habitats (n=2 leaves per bar). Analysis of sugar maple and beech

leaves together resulted in a significant species X size interaction (see text). P

values shown with each chart indicate significance of leaf size within either sugar

85 maple or beech (size X habitat interactions were never significant)...... 73

Figure 2. Fungal taxa-area relationship detected on maple leaves (P < 0.05) but

not beech. Regression lines are shown for maple upland, vernal pool, and riparian

habitat TARs. For maple leaves, model selection using AICc resulted in a model

where habitat TARs have identical z-values (slope; 0.22 ± 0.07) and different c-

90 values (Y-axis intercepts)...... 76

Figure 3. Redundancy analysis ordination of fungal community composition at

the six field sites. Habitat is indicated by symbol shape, with site names shown in

legend. Because leaf type explained little variation in community composition

(5%), each symbol represents the centroid of the four leaves sampled at each site.

95 ...... 81

CHAPTER 4. The Spatial Scaling of Fungal Diversity

Figure 1. Graphical representation of step distance matrix showing connections

between leaves collected at each Jennings Woods site...... 108

Figure 2. Distance decay plots showing the average Hellinger distance between

100 communities at increasing step distances for each of the six sites in this study.

Habitat abbreviations (VP=vernal pool, Rip=riparian, Up=upland), site ID, and

the standardized Mantel statistic (rM) are shown for each plot...... 114

vii

Figure 3. Collectors curve analysis quantifying new TRFs with each randomly

selected fungal community. Plots show the set of all randomly selected

105 communities at each site (white circles) and communities randomly selected at

each depth within the forest floor leaf pack that the communities were located in

(black symbols). Symbols for layers of the leaf pack that communities were

sampled from: top layer: X; 2nd layer: square; 3rd layer: triangle; 4th layer:

diamond; 5th layer: star; 6th layer circle...... 121

110 Figure 4. Collectors curve analysis quantifying the difference between new and

previous fungal community composition. The community composition with each

additional leaf contributes to a site centroid composition value, and the

composition of each new community is then compared to the site centroid value.

Plots show the set of all randomly selected communities at each site (white

115 circles) and communities randomly selected at each depth within the forest floor

leaf pack that the communities were located in (black symbols). Symbols for

layers of the leaf pack that communities were sampled from: top layer: X; 2nd

layer: square; 3rd layer: triangle; 4th layer: diamond; 5th layer: star; 6th layer

circle...... 123

120 Figure 5. Anderson multivariate homogeneity of group dispersion determined on

TRFLP (community composition) profiles for all fungal communities found on

beech (n=70) and maple (n=60) leaves. The central point of each “starburst” is

the centroid in multivariate space for the communities found at any given site.

The terminus of each line emanating from the centroid is the point in multivariate viii

125 space of a community on one leaf. Site identification is found in the legend (V =

vernal pool, Rip = riparian, Up = upland). The p-value tests for significant

differences between groups (sites)...... 127

CHAPTER 5. The Impact of Litter Diversity on Microbial Enzyme Activity, Biomass,

and Leaf Litter Decomposition

130 Figure 1. Significant habitat (Rip = riparian, Up= upland) x harvest (Day treated

as a categorical variable) interaction for proportion AFDM remaining in beech,

hazel, oak, and maple leaves. Values are averaged for all treatments...... 160

Figure 2. Significant AFDM leaf type (beech, hazel, oak, and maple) x lignin:N

ratio shown at each harvest date (Day treated as a categorical variable). Values

135 are averaged for all treatments...... 162

Figure 3. Significant AFDM leaf type (beech, hazel, oak, and maple) x microbial

biomass interaction shown with the proportion of AFDM remaining at each of the

five sample collection times. Values are averaged by harvest for all

treatments...... 164

140 Figure 4. Fungal biomass at each harvest for each of the litter diversity

treatments. An asterisk (*) is placed above harvest with a significant treatment x

harvest interaction...... 167

Figure 5. Bacterial biomass at each harvest for each of the litter diversity

treatments. An asterisk (*) is placed above harvest with a significant treatment x

145 harvest interaction...... 168

Figure 6. Significant nagase harvest x leaf type interaction. The activity for

recalcitrant leaves is shown with black symbols and labile leaves with white

symbols. Enzyme activities are in nmol/g/AFDM/hr and have been log-

transformed. An asterisk (*) is placed above harvest with a significant treatment

150 x harvest interaction...... 171

Figure 7. Ordination chart showing fungal community composition averaged

across all leaf types and litter mixture treatments for riparian and upland

communities at each of the five collection times (harvests). Variance explained

for RDA 1 axis: 8.2%, RDA 2 axis: 2.9%...... 174

155

LIST OF TABLES

CHAPTER 2. Assessment of Bias Associated with Incomplete Extraction of Microbial

DNA from Soil

Table 1. Soil properties. a. Data from Zak and Pregitzer (1990). ND – not

160 determined...... 21

Table 2. Alternative cell lysis procedures (DNA extraction treatments). a.

Vortexer set at maximum speed. b. GenoGrinder set at 1500 strokes per min...... 23

Table 3. Variance partitioning of bacterial sequence community composition. * P

< 0.1; ** P < 0.05; NS not significant...... 42

165 Table 4. Response of bacterial phyla composition to extraction step revealed

through redundancy analysis...... 42

CHAPTER 3. Taxa-area Relationship and Neutral Dynamics Influence the Diversity of

Fungal Communities on Senesced Tree Leaves

Table 1. Blast results and percentage of variance explained by habitat (P = 0.001)

170 or site nested within habitat (P = 0.001) for dominant OTUs (>5% relative

abundance on at least one leaf). NA= not assigned. Symbols indicate the habitat

that dominant OTUs were associated with (∆ = upland, ∞ = vernal pool, ◊ =

riparian)...... 79

Supplemental Table S1. AIC values for 6 rank abundance distribution models.

175 Lowest AIC value for each sample represents the best fit model...... 83

CHAPTER 4. The Spatial Scaling of Fungal Diversity

Table 1. Mantel standard statistic (rM) values for Mantel and partial Mantel

analysis conducted on spatial distance (step or weighted), depth of forest floor

leaf was harvested from, and leaf type for all connected leaves. Partial Mantel

180 analysis (“|”) was done to correct for effects of various distance matrices. Non-

connected leaves were removed before analysis. Statistical significance is shown

with the following symbols: * P < 0.1; ** P < 0.05; *** P < 0.01; NS: not

significant...... 116

Table 2. Proportion of variance (Adjusted R2) in fungal community composition at

185 each site explained by spatial location (PCNM vectors selected after forward

selection), depth of community location within forest floor leaf pack, and leaf

type. Values shown are from redundancy analysis of all beech & maple leaves

that were within a connected leaf network (n). * P < 0.1; ** P < 0.05; *** P <

0.01; NS = not significant. Full model shows the total amount of variance

190 explained by combined significant factors...... 118

Table 3. Redundancy analysis determining the effect of factors on community

composition for all beech (n=70) & maple (n=60) leaves across all sites and all

leaves (n=160) after singleton and doubleton species at each site were removed.

The significance (p-value) is given for each factor, as is the amount of variance in

195 community composition explained by each factor (Adj R2)...... 125

CHAPTER 5. The Impact of Litter Diversity on Microbial Enzyme Activity, Biomass,

and Leaf Litter Decomposition

Table 1. Commonly reported hydrolytic and oxidative carbon-acquiring

enzymes, the natural substrates they are known to degrade, and substrate used in

200 lab assays to determine enzyme activity...... 149

Table 2. The impact of enzyme activity on litter decomposition (% AFDM

remaining)...... 165

Table 3. The impact of main effects on enzyme activity...... 169

Table 4. The impact of resource variables on enzyme activity...... 172

205

xii

ACKNOWLEDGEMENTS

It is with much gratitude that I acknowledge those who have supported me in my

academic endeavors and enriched the fabric of my reality. At the heart of my Kent State

210 experience was Chris Blackwood who quickly became a pillar of support that continually

inspired, challenged, and encouraged me to transcend limitations. His approach towards

conducting research is one of collaboration which completely made me feel like we were

equally important team members working together trying to understand how the synthesis

of microbial patterns and processes influences ecology at multiple spatial scales.

215 Understanding patterns and processes is not a mere intellectual curiosity for me, but is a

deeply meaningful part of who and what I am as a human being. Chris shares that sort of

perspective, and having him as my advisor has truly transformed my life. Anything that I

may accomplish in science from this point forward is a tribute to Chris’s unwavering

guidance and support.

220 I also owe an immense amount of gratitude to Laura Leff and Mark Kershner.

My beginnings at Kent State were rocky to say the least, and although my actions could

have easily cause Laura to become jaded in her ongoing interactions with me she treated

me with the utmost respect. I have often turned to her for advice and every time I sought

her counsel, she was always there for me with a kind heart, casual smile, and the ideal

225 perspective to help me through whatever I was dealing with at the time. Mark easily

possesses one of the most brilliantly intelligent minds I have ever had the pleasure of

interacting with and I can honestly say that I have learned something new in virtually xiii

every conversation I have had with him about ecology. Mark is an amazing instructor,

who along with Bob Carlson (when they co-taught Populations, Communities, and

230 Ecosystems) completely transformed the study of ecology for me by infusing the

challenges and developments of this discipline throughout its history into fascinating

tales associated with the personalities and perspectives of the people proposing ongoing

shifts in ecological paradigm.

I also acknowledge vital contributions from Mary Russell and Gwenn Volkert.

235 Mary took me in as her student during a transitional time for me and although things

didn’t work out for us to continue collaborating, she served as a mentor ushering me into

the realm of molecular biology laboratory techniques and helping me obtain adjunct

teaching positions at the Trumbull branch of Kent State. Gwenn graciously served on my

committee, allowed me to sit in on her lab’s research meetings in order to learn more

240 about computer science, and brought the spirit of the Grateful Dead (who have served a

pivotal role in my life) to my dissertation defense.

There are several other influential people who provided significant contributions

during my time at Kent State. Kurt Smemo provided many inspiring and interesting

conversations about spatial aspects of biogeochemical transformation, sage advice

245 concerning the institution of science, and facilitated additional experience swimming in

the school of Phish. Donna Warner was extremely helpful every time I had questions or

issues to deal with. Jim Blank provided ongoing support as the chair of the department

by hiring me for several adjunct teaching positions, personal support with career advice,

and a hilarious take on many topics. Bob Carlson and I had many interesting and xiv

250 inspiring conversations about ecological modeling and history. I also thank Florence Hsu

for many important conversations about technical lab topics and life beyond graduate

school, Sarah Vash, Jennifer Marcinkiewicz, and Andrea Case for mentoring in the ways

of teaching and accessibility for meaningful conversation, and Joe Ortiz for serving as the

College of Arts and Sciences representative during my dissertation defense.

255 The institution of Kent State has provided resources to support my work and

enrich my life. I gratefully acknowledge financial support from an Arthur and Margaret

Herrick Research Grant, the Kent State Graduate Student Senate, a Kent State University

Fellowship, and Chris Blackwood’s startup funding. The opportunities that Kent State

offered me to develop my teaching skills as an adjunct faculty allowed me enriching

260 interactions with many students which culminated in a Faculty Recognition award as

Outstanding Educator from Kent State Student Accessibility Services. I also was

enriched by my time serving as student council president on the Biology Graduate

Studies Committee.

None of my work could have been accomplished without the help of numerous

265 fellow students. They provided indispensable contributions via helping on projects and

sympathizing about many aspects of life as a student. Listing every student who has

helped would take many pages, but I must give special acknowledgment to Oscar

Valverde who has provided a wealth of investigative and statistical knowledge and served

as a valuable collaborator. Other students who have provided important work and

270 meaningful personal conversations of support include Kelly Barribal, Andie Bender,

Melissa Brewster, Seth Brown, Suhana Chattopadhyay, Mui Clark, Curtis Clevinger, xv

Chris Dejelo, Alex Delvalle, Jael Edgerton, Meghan Fernandes, Stephanie Forte, Matt

Gacura, Alex Gradishar, Devinda Hiripitiyage, Sarah Krock, Bobby Marquardt, Lisa

Regula Meyer, Justin Reeves, Eugene Ryee, Dan Sprockett, and Patricia Wallschleager.

275 Finally, I gratefully acknowledge the unwavering support and love from my

collective family. This journey would have never been possible without my wife Sharon

who has lovingly been my companion for almost two decades. She sacrificed her well-

being for many years by working a stressful job to support my way through school,

supported whatever I attempted, and encouraged me when I felt like giving up. Jenna has

280 been an ongoing source of love, strength, and healing for me. My mother and father have

done everything they can to move me towards success. I also acknowledge the influence

during this time of Chris, Jesse, Kallisti, Rio, Jimmy, Jerry, and Patrick as well as my

brothers and sister.

It is with the upmost appreciation that I acknowledge everyone who has been a

285 part of my journey towards the successful completion of my doctoral path at Kent State

University.

xvi

Chapter 1

Introduction

290 Ecology is the investigation of patterns that emerge within the biosphere as

organisms interact with their environment (Haeckle 1866). The majority of ecological

principles were based upon observations of patterns and experiments conducted at a scale

that is visible to the naked eye (Cowles 1899, Clements 1916, Gleason 1922, Elton 1924,

Tansley 1935, Lindeman 1942, Odum 1959, Whittacker 1962, MacArthur and Wilson

295 1967, Hutchinson 1965, Paine 1969, Connell and Slayter 1977, Grime 1979). Applying

conventional ecological paradigms to microbial populations and communities has severe

limitations. Microbial communities are comprised of enormous numbers of individuals

(Torsvik et al. 2002, Fenchel and Findlay 2004), and most of these individuals are

extremely difficult to identify (Amann et al. 1995). However, recent techniques in

300 molecular-based investigation have allowed researchers to begin to quantify patterns that

exist at the microbial scale (Zak et al. 2006, Hirsch et al. 2010). The central goal of this

dissertation is to quantify microbial patterns and test whether processes that have been

documented as being applicable to macroorganisms are also applicable to

microorganisms.

305 Pattern analysis and quantification is an important investigative tool that increases

our understanding of ecological processes that influence community structure and

functional activity (Konopka 2009). Natural biological assemblages may be organized

into patterns that reflect responses to environmental selective pressures (Whittaker 1962)

or emerge as a function of their ability to disperse to new habitat (Hubbell 2001). Two

310 fundamental patterns related to microbial organisms are 1) distributions of different types

of microbial taxa within one or between multiple spatially-defined areas (communities)

(Sterelny 2006), and 2) distributions of biogeochemical functional activity performed by

the communities (Sterelny 2006, O’Donnell et al. 2007). A basic property of ecology is

that patterns and processes occur at multiple spatial scales which form a hierarchical

315 interactive matrix (Allen and Starr 1982, O’Neil et al. 1986). For example, it has been

documented that microbial metabolic activity that occurs directly at a scale of

micrometers (Young et al. 2008) indirectly influences organisms and resource pools that

range in size from micrometers to kilometers (thousands of meters) (Ducklow 2007).

Despite our awareness of ecological hierarchies of interacting processes and patterns, a

320 contemporary topic of investigation is whether the patterns and processes occurring at

spatial scales of macroorganisms and microorganisms are similar (Green et al. 2004,

Horner-Devine et al. 2004, Bell et al. 2005, Fenchel and Finlay 2005, Fierer and Jackson

2006).

The central ecosystem role of saprotrophic microbial activity

325 Fungal and bacterial biogeochemical functional activity is critically important for

all life on this planet (Fenchel 1998, Peay et al. 2008). Virtually every form of life is

dependent on microbial transformation of inorganic and organic molecules into

configurations that life can utilize (Madigan et al. 2006). Microbial transformations are a

key element of ecosystem energy and material flow (Roling et al. 2007) which provide

330 inorganic nutrients to vegetation and incorporate carbon and nutrients into microbial

biomass, retaining energy, carbon, and nutrients in an ecosystem. In doing so, microbes

become a basic trophic layer (Wardle 2002); a food resource that directly or indirectly

supports most heterotrophic life on earth (Peay et al. 2008). Within the temperate forest

ecosystem investigated in this dissertation, saprotrophic microbial metabolism releases

335 unavailable nutrients and carbon that were previously bound within senesced leaf organic

compounds (Berg and McClaugherty 2008). Due to its dramatic impact on entire food

webs, microbial functional activity is often seen as being an essential component of the

functional activity conducted by entire ecosystems (Ducklow 2007).

Ecosystems are defined by general system theory (von Bertalanffy 1968) as ‘open

340 systems’ whose functional activity is characterized by the circulation of energy and

materials into, out of, and within the system. Energy within a system that is useful and

can do work is defined as exergy (Evans 1966). Because systems that possess higher

exergy have been shown to have increased energy efficiency and complexity (Schneider

and Kay 1994, Fath et al. 2004), exergy has been proposed to be a metric to measure

345 system efficiency and has been equated with system complexity (Jørgensen et al. 2007).

General system and exergy theory describe ecosystem operational dynamics and place

saprotrophic microbial functional activity into a context that is aligned with system

operational activity. Microbial functional activity serves to cycle energy and materials

(general system theory) throughout the ecosystem (Ducklow 2007). In making energy

350 and nutrients available for other trophic layers they serve to increase exergy by

maximizing energy utilization within the system (Jørgensen et al. 2007).

Theoretical framework for assembly of saprotrophic microbial communities

Microbial functional activity that is associated with ecosystem exergy is

conducted by communities that assemble on natural patches of habitat. Each senesced

355 leaf on a forest floor provides a resource patch upon which a unique microbial

community assembles. A collection of local communities that are linked by dispersal has

been defined as a metacommunity (Hanski and Gilpin 1991, Wilson 1992, Liebold et al.

2004). Members of a local community are known as the actual species pool, and they

assemble from a larger pool of species known as the regional species pool (Cornell 1985,

360 Ricklefs 1987, Leibold 1997, Shurin et al. 2000). There are two primary perspectives

proposed to explain metacommunity assembly and composition. A “species-sorting”

perspective proposes that community assembly and composition is segregated according

to environmental conditions (i.e., the niche, Whittaker 1962). A “neutral” perspective

(Hubbell 2001) proposes that all species are essentially equal in their competitive ability,

365 fitness, and movement ability; community assembly and composition is a result of

probabilities of species movement (colonization, emigration) into or out of a locale.

There is considerable debate over whether microbial metacommunity assembly and

composition is a result of species sorting, neutral dynamics, or some combination of the

two perspectives (Horner-Devine et al. 2004, Leibold et al. 2004, Bell et al. 2005,

370 Fenchel and Findlay 2005, Martiny et al. 2006, Hovatter et al. 2011).

The focus of inquiry within this dissertation was to quantify saprotrophic

microbial patterns in order to understand processes that influence community assembly,

diversity, and functional activity Patterns and processes were compared to ecological

theory in order to add to contemporary investigation within the scientific community

375 concerning whether patterns and processes occurring at spatial scales of macroorganisms

and microorganisms are similar. A brief introduction to each chapter gives more detail

about the dissertation body of research.

Chapter 2: Assessment of Bias Associated with Incomplete Extraction of Microbial DNA

from Soil

380 Patterns of diversity have been a primary area of investigation in the discipline of

ecology (Magurran 2004). Microbial diversity has been challenging to quantify due to

the small size, limited morphology, and resistance to cultivation of the vast majority of

microbes (Amann et al. 1995). Ecologists have expanded their ability to quantify

microbial communities with molecular-based techniques, which sample a much larger

385 portion of the community than traditional culture-based methodology (Hirsch et al.

2010). However, it has been noted that information obtained with molecular techniques

may become biased during DNA extraction (Lombard et al. 2011) and gene amplification

(Thies 2007). Results skewed by bias would not accurately depict community dynamics

that exist in the environment. Chapter 1 is concerned with investigating fungal and

390 bacterial molecular-based diversity patterns in order to quantify bias associated with

DNA extraction. Specific concerns addressed in this chapter include optimizing a DNA

extraction procedure to maximize the amount of DNA extracted, assessing the damage

done to DNA during the extraction process, and quantifying bias found within bacterial

and fungal gene copies, community composition, and taxa abundance distributions due to

395 incomplete extraction.

Chapter 3: Taxa–area relationship and neutral dynamics influence the diversity of fungal

communities on senesced tree leaves

It has been frequently observed that there is a correlation between the size of

habitat patches and the number of taxa detected (Rosenzweig 1995, Connor and McCoy

400 2001, Lomolino 2001, Drakare et al. 2006). The “taxa-area relationship” (Arrhenius

1921, Gleason 1922) has been described for macroorganisms at a wide variety of scales

(centimeters to kilometers), but it is still unclear whether this pattern is operational at the

scale of microbial communities. In Chapter 3, fungal diversity was quantified with clone

library molecular sequence data derived from communities located on senesced American

405 beech (Fagus grandifolia) and sugar maple (Acer saccharum) leaves. Our experimental

design allowed us to quantify α- and β-diversity, test for the presence of a taxa-area

relationship, and contribute to a central inquiry of this dissertation regarding whether

patterns and processes that have been documented for macroorganisms are also

applicable to microorganisms.

410 We also analyzed the patterns of fungal diversity in order to understand how

diversity was being influenced by either species-sorting or neutral process. Beech and

maple leaves have differing proportions of labile (energy-rich) and recalcitrant (energy-

poor) compounds, and may influence fungal community diversity via these pools of

carbon resources, providing support for species-sorting. Other species-sorting factors

415 may include variations in environmental abiotic conditions. We selected beech and

maple leaves located in three distinct habitats (upland, riparian, vernal pool) in order to

test if fungal community diversity was influenced by environmental variability. In

contrast, community patterns of diversity may also be influenced by neutral processes.

Hence, we also compared taxa rank-abundance distributions to mathematical models

420 consistent with species-sorting or neutral processes in order to test the relative influence

of these processes on fungal diversity.

Chapter 4: The spatial scaling of saprotrophic fungal diversity

In Chapter 4, we extended our investigation to groups of 30 leaves clustered

together in different habitats, each leaf providing an area with a unique history and

425 discrete boundaries, and each group of 30 leaves providing a unique distribution of fungal

resources. In each leaf, we characterized fungal community composition using TRFLP.

This allowed us to quantify patterns of compositional change (community turnover)

across multiple communities at multiple spatial scales. We collected local assemblages

of leaves in a manner that allowed us to quantify the topology of their physical proximity.

430 We used this information to construct distance matrices that allowed us to test if fungal

community composition was correlated with spatial distance (i.e., displayed a distance-

decay relationship). If a significant distance-decay relationship is driven by spatial

proximity that is independent of environmental heterogeneity, this would provide support

for dispersal limitation influencing fungal community composition. However, the forest

435 floor leaf pack includes leaves of differing tree species and buried at differing depths

from the surface, and this creates fine-scaled environmental heterogeneity in, for

example, biochemical resources present in the leaves, moisture, and the influence of

colonization from the community in the mineral soil. We analyzed the impact of leaf

type and depth within the leaf pack using a variety of statistical techniques in an attempt

440 to quantify the relative influence of neutral versus species-sorting processes. We then

compared patterns among sites located in upland, riparian, and vernal pool habitats in

order to determine if localized diversity patterns were similar among habitats and if

regional species pools from which community composition emerges may be filtered at

the landscape scale.

445 Chapter 5: The impact of litter diversity on microbial enzyme activity, biomass, and leaf

litter decomposition

The primary focus of Chapter 5 was to investigate microbial functional activity;

specifically saprotrophic bacterial and fungal carbon-cycling activity (decomposition)

occurring on senesced leaf litter in the forest floor. The impact of regulatory factors

450 (mixtures of recalcitrant and labile leaf types, leaf nitrogen and carbon compound

proportions, time of incubation in the field, upland or riparian habitat) was quantified on

rates of leaf litter decomposition. The impact of regulatory factors was also quantified on

different aspects of the microbial community that may influence decomposition rates,

including biomass, extracellular enzyme activity, and community composition. The

455 correlation between microbial dynamics and rates of litter decomposition was also

quantified. Because decomposition is the transfer of solar energy in the form of

senesced plant tissue into the microbial base of the trophic food web, we used these

results to infer whether these regulatory factors also had an impact on forest ecosystem

exergy. Particular attention was devoted to the role of litter diversity due to the ubiquity

460 of contradictory results in numerous studies comparing rates of decomposition of litter

mixtures versus single-species litter (Hӓttenschwiler et al. 2005).

Litter diversity was also utilized to investigate whether two ecological theories

that were originally hypothesized and tested with macroorganisms also apply to

microorganisms. The species complementarity hypothesis states that increased resource

465 heterogeneity (in our case, due to leaf litter diversity) offers increased opportunity for

niche differentiation, facilitating a more extensive utilization of the resources and

resulting in more productive communities (Loreau and Hector 2001, Marquard et al.

2009). Species complementarity as applied to this study predicts that microbial biomass

would be higher in litter mixtures than in single-species treatments, and the increased

470 biomass would result in faster decomposition rates for mixed litter.

We utilized community composition profiles to test an ecological theory that was

originally proposed for microorganisms. The Guild Decomposition Model (GDM)

predicts that different microbial guilds are each specialized to decompose specific litter

substrate pools (soluble polymers, cellulose, and lignin) whose proportions change during

475 successive stages of decomposition (Moorhead and Sinsabaugh 2006). Evidence for the

GDM as applied to this study would be provided if we found significant correlations

between enzyme activity patterns and microbial community composition.

Summary

The central inquiry of this dissertation was to address knowledge gaps concerning

480 whether ecological processes that have been demonstrated to be applicable to

macroorganisms are also applicable to microorganisms. The methodology utilized in this

dissertation was to quantify molecular-based microbial patterns and apply statistical

analyses designed to test the relative amount of influence that various processes have on

pattern formation and development. The relative influence of species-sorting versus

485 neutral processes on community assembly was tested on both community composition

(TRFLP) and sequence-based molecular patterns. These processes were tested at the

scales of single communities (α-diversity) and multiple communities (β-diversity). We

also were able to test whether taxa that comprised regional pools of fungal communities

were filtered by environmental heterogeneity associated with vernal pool, upland, and

490 riparian habitats. We quantified entire fungal communities located across a range of

patch sizes in order to determine if fungi follow a taxa-area relationship. Microbial

biomass associated with litter diversity was used to determine if microbes increase

ecosystem exergy, and litter diversity decomposition rates were used to test whether the

species complementary hypothesis is applicable to microbes.

495 Two additional areas of investigation were conducted to address knowledge gaps

in the microbial ecology research community. Recent advances in molecular-based

technology have increased our capacity to investigate ecological questions but there is

debate concerning how accurately the patterns we obtain with evolving molecular

technology reflect actual in situ patterns. This was tested with a study that compared

500 community composition, sequence-based taxa abundance distributions, and taxonomic

fungal and bacterial gene copies associated with multiple DNA extractions on individual

soil samples. Finally, there have been well-documented conflicting results (reviewed in

Gartner and Cardon 2004; Hӓttenschwiler et al. 2005) in the many studies comparing

decomposition rates of single-species versus mixed litter. The litter decomposition study

505 in this dissertation was designed specifically to address the conflicting results of the

many decomposition studies that have been documented.

References

Allen THF, Starr TB. 1982. Hierarchy: Perspectives for Ecological Complexity.

Chicago: The University of Chicago Press; 310 p.

510 Amann RI, Ludwig W, Schleifer KH, Torsvik KHV, Goksoyr J. 1995. Phylogenetic

identification and in situ detection of individual microbial cells without

cultivation. Microbiological Reviews, 59: 2019-2027.

Arrhenius O. 1921. Species and area. Journal of Ecology; 9: 95–99.

Bell T, Ager D, Song J, Newman JA, Thompson IP, Lilley AK, van der Gast C. 2005.

515 Larger islands house more bacterial taxa. Science 308: 1884.

Berg B, McClaugherty C. 2008. Plant litter: Decomposition, humus formation, carbon

sequestration. New York: Springer-Verlag; 279 p.

Clements FE. 1916. Plant succession: An analysis of the development of vegetation.

Washington, DC: Carnegie Institute of Washington. 658 p.

520 Connell JH, Slayter RO. 1977. Mechanisms of succession in natural communities and

their role in community stability and organization. American Naturalist 111:

1119-1144.

Connor EF, McCoy ED. 2001. Species–area relationships. In Encyclopedia of

Biodiversity, Vol. 5. Levin, S.A. (ed.). San Diego: Academic Press, pp. 397–

525 411.

Cornell HV. 1985. Species assemblage of cynipid gall wasps are not saturated.

American Naturalist 126: 565-569.

Cowles HC. 1899. The ecological relations of the vegetation on the sand dunes of Lake

Michigan. Michigan Botanical Gazette 27: 167-202.

530 Drakare S, Lennon JJ, Hillebrand H. 2006. The imprint of the geographical,

evolutionary and ecological context on species-area relationships. Ecology Letters

9: 215–227.

Ducklow H. 2007. 10th Symposium on Aquatic Microbial Ecology (SAME 10). Faro:

Portugal, pp 13–19.

535 Elton CS. 1924. Periodic fluctuations in the numbers of animals: Thier causes and

effects. British Journal of Experimental Biology 2: 119-163.

Evans RB. 1966. A proof that exergy is the only consistent measure of potential work

[thesis]. Dartmouth college, Hanover, NH.

Fath BD, Jørgensen SE, Patten BC, Straškraba M. 2004. Ecosystem growth and

540 development. BioSystems 77: 213-228.

Fenchel T. 1998. Bacterial biogeochemistry: the ecophysiology of mineral cycling (2nd

ed.). San Diego: Academic Press. 307 p.

Fenchel, T. Findlay BJ. 2004. The ubiquity of small species: patterns of local and global

diversity. BioScience 54: 777-784.

545 Fenchel T, Finlay BJ. 2005. Bacteria and island biogeography. Science 309: 1997-

1998.

Fierer N, Jackson RB. 2006. The diversity and biogeography of soil bacterial

communities. PNAS 103: 626-631.

Gartner TB, Cardon ZG. 2004. Decomposition dynamics in mixed-species leaf litter.

550 Oikos 104: 230–246.

Gleason HA. 1922. On the relationship between species and area. Ecology 3: 158–162.

Gleason HA. 1926. The individualistic concept of plant association. Bulletin of the

Torrey Botanical Club 53: 7-26.

Green JL, Holmes AJ, Westoby M, Oliver I, Briscoe D, Dangerfield M, Gillings M,

555 Beattie AJ. 2004. Spatial scaling of microbial eukaryote diversity. Nature 432:

747-750.

Grime JF. 1979. Plant Strategies and Vegetational Processes. Chichester, UK: Wiley

Pulbishing; 222 p.

Haeckel E. 1866. Generelle Morphologie der Organismen : allgemeine Grundzüge der

560 organischen Formen-Wissenschaft, mechanisch begründet durch die von C.

Darwin reformirte Decendenz-Theorie. Berlin

Hanski I, Gilpin ME. 1991. Metapopulation dynamics: Brief history and conceptual

domain. Biological Journal of the Linnean Society 42: 3-16.

Hirsch PR, Mauchline TH, Clark IM. 2010. Culture-independent molecular techniques

565 for soil microbial ecology. Soil Biology & Biochemistry 42: 878-887.

Horner-Devine MC, Lage M, Hughes JB, Bohannan BJM. 2004. A taxa-area

relationship for bacteria. Nature 432: 750-753.

Hubbell SP. 2001. The Unified Neutral Theory of Biodiversity and Biogeography.

Princeton, NJ: Princeton University Press; 375 p.

570 Hutchinson GE. 1965. The niche: An abstractly inhabited hypervolume. The Ecological

Theatre and the Evolutionary Play. New Haven, CT: Yale University Press; pp.

26-78.

Jørgensen SE, Fath BD, Bastianoni S, Marques JC, Müller F, Nielson SN, Patten BC,

Tiezzi E, Ulanowicz RE. 2007. A New Ecology: Systems Perspective. Oxford:

575 Elsevier. 275 p.

Liebold MA. 1997. Biodiversity and nutrient enrichment in pond plankton communities.

Evolutionary Ecology Research 1: 73-95.

Liebold MA, Holyoak M, Mouquet N, Armarasekare P, Chase JM, Hoopes MF, Holt RD,

Shurin JD, Law R, Tilman D, Loreau M, Gonzalez A. 2004. The

580 metacommunity concept: a freamework for multi-scale community ecology.

Ecology Letters 7: 601-613.

Lindeman RL. 1942. The trophic-dynamic aspect of ecology. Ecology 28: 399-418.

Lombard N, Prestat E, van Elsas JD, Simonet P. 2011. Soil-specific limitations for

access and analysis of soil microbial communities by metagenomics. FEMS

585 Microbiology Ecology 78: 31-49.

Lomolino MV. 2001. The species–area relationship: new challenges for an old pattern.

Progress in Physical Geography 25: 1–21.

Loreau M, Hector A. 2001. Partitioning selection and complementarity in biodiversity

experiments. Nature 412: 72–76.

590 MacArthur RH, Wilson EO. 1967. The theory of island biogeography. Princeton, NJ:

Princeton University Press; 224 p.

Madigan MT, Martinko JM, Dunlap PV, Clark DP. 2006. Brock Biology of

Microorganisms (10th edition). San Fransisco: Pearson/Benjamin Cummings;

986 p.

595 Magurran AE. 2004. Measuring Biological Diversity. Oxford, UK: Blackwell

Publishing; 264 p.

Martiny JBH, Bohannan BJM, Brown JH, Colwell RK, Fuhrman JA, Green JL, Horner-

Devine MC, Kane M, Krumins JA, Kuske CR, Morin PJ, Naeem S, Øvreås L,

Reysenbach A, Smith V, Staley J. 2006. Microbial biogeography: putting

600 microorganisms on the map. Nature Reviews Microbiology 4: 102-112.

Marquard E, Weigelt A, Temperton VM, Roscher C, Schumacher J, Buchmann N,

Fischer M, Weisser WW Schmid, B. 2009. Plant species richness and functional

composition drive overyielding in a six-year grassland experiment. Ecology 90:

3290–3302.

605 O’Donnell AG, Young IM, Rushton SP, Shirley MD, Crawford JW. 2007.

Visualization, modelling and prediction in soil microbiology. Nature Reviews

Microbiology 5: 689-699.

Odum EP. 1959. Fundamentals of Ecology. Philadelphia, PA: W.B. Saunders; 624 p.

O’Neill RV, DeAngelis D, Waide J, Allen TFH. 1986. A Hierarchical Concept of

610 Ecosystems. Princeton, NJ: Princeton University Press; 235 p.

Paine RT. 1969. A note on trophic complexity and community stability. American

Naturalist 100: 65-75.

Peay KG, Kennedy PG, Burns TD. 2008. Fungal community ecology: A hybrid

beast with a molecular master. Biosciences 58(9): 799-810

615 Reysenbach V, Smith H, Staley JT. 2006. Microbial biogeography: putting

microorganisms on the map. Nature Reviews Microbiology 4: 102-112.

Ricklefs RE. 1987. Community diversity: Relative roles of local and regional processes.

Science 235: 167-171.

Roling WFM, van Breukelen BM, Bruggeman FJ, Westerhoff HV. 2007. Ecological

620 control analysis: being(s) in control of mass flux and metabolite concentrations in

anaerobic degradation processes. Environmental Microbiology 9: 500–511.

Rosenzweig ML. 1995. Species diversity in space and time. New York: Cambridge

University Press. 436 p.

Schneider ED, Kay JJ. 1994. Complexity and thermodynamics: towards a new ecology.

625 Futures 24: 626–647.

Shurin JB, Havel JE, Liebold MA, Pinel-Alloul B. 2000. Local and regional

zooplankton species richness: A scale-independant test for saturation. Ecology 81:

3062-3073.

Sterelny K. 2006. Local ecological communities. The British Journal for the Philosophy

630 of Science 73: 215–231.

Tansley AG. 1935. The use and abuse of vegetational concepts and terms. Ecology 16:

284-307.

Theis J. 2007. Soil microbial community analysis using terminal restriction fragment

length polymorphisms. Soil Science Society of America Journal 71: 579-591.

635 Torsvik V, Ovreas L, Thingstad TF. 2002. Prokaryotic diversity - magnitude, dynamics,

and controlling factors. Science 296: 1064-1066.

von Bertalanffy L. 1968. General System Theory: Foundations, Development,

Applications. New York, NY: George Braziller, Inc.; 295 p.

Wardle DA. 2002. Communities and Ecosystems: Linking the Aboveground and

640 Belowground Components. Princeton, NJ: Princeton University Press; 400 p.

Whittaker RH. 1962. Classification of natural communities. Botanical Review 28: 1–

239.

Wilson DS. 1992. Complex interactions in metacommunities, with implications for

biodiversity and higher levels of selection. Ecology 73: 1984-2000.

645 Young IM, Crawford JW, Nunan N, Otten W, Spiers A. 2008. Microbial distribution in

soils: physics and scaling. Advances in Agronomy 100: 81–121.

Zak DR, Blackwood CB, Waldrop MP. 2006. A molecular dawn for biogeochemistry.

Trends in Ecology and Evolution 21: 288-295.

650 Chapter Two

Assessment of Bias Associated with Incomplete Extraction of Microbial DNA from

Soil

(This chapter was published in Applied and Environmental Microbiology in August

2009)

655 Abstract

DNA extraction bias is a frequently cited but poorly understood limitation of

molecular characterizations of environmental microbial communities. To assess the bias

of a commonly-used soil DNA extraction kit, we varied the cell lysis protocol and

conducted multiple extractions on subsamples of clay, sand, and organic soils. DNA, as

660 well as bacterial and fungal ribosomal gene copies measured by quantitative PCR,

continued to be isolated in successive extractions. However, cumulative quantities

approached an asymptote after three extractions when using an improved cell lysis

protocol, except for fungal ribosomal gene copies in organic soil. Using terminal

restriction fragment length polymorphism (T-RFLP), a significant shift in community

665 composition due to extraction bias was detected for bacteria, but not for fungi. This bias

was further characterized by pyrosequencing bacterial 16S ribosomal gene amplicons.

Relative abundance of sequences from rarely-cultivated groups such as Acidobacteria,

Gemmatimonades, and Verrucomicrobia was higher in the first extraction than the sixth, 18

whereas the reverse was true for Proteobacteria and Actinobacteria, suggesting that the

670 well-known phylum-level bacterial cultivation bias may be partially exaggerated by DNA

extraction bias. We conclude that bias can be adequately reduced in many situations b y

pooling three successive extractions, and additional measures should be considered when

comparing divergent soil types or when comprehensive community analysis is necessary.

Introduction

675 The vast majority of soil bacteria (Borneman et al. 1996, Pace 1997, Torsvik and

Øvreås 2002) and fungi (Hawksworth and Rossman 1997, van Elsas et al. 2002) cannot

be cultured via traditional laboratory techniques and must be identified using molecular

methods. Successful characterization of microbial communities is therefore often

dependant on DNA that is extracted from the environment. However, extraction of high-

680 quality DNA from soil can be problematic for several reasons. Microbial cell walls can

inhibit SDS penetration and disruption of cell membranes (Moore et al. 2004). Microbes

may also be protected from physical and chemical disruption by soil aggregates and by

close association with particles through electrostatic charges. These charges can also

cause adsorption of DNA onto soil particles (Griffiths et al. 2004). Finally,

685 contamination of extracted DNA with soil humic acids can inhibit the efficiency of DNA

amplification via polymerase chain reaction (Tebbe and Vahjen 1993).

Initial research showed that direct DNA extraction from a soil sample is a more

efficient method than indirect extraction, which involves extraction of microbial cells

from the soil prior to DNA extraction (Milling et al. 2005, Robe et al. 2003). A wide

690 variety of methods to disrupt microbial cell integrity and remove humic acid

contaminants have been reported (Bürgmann et al. 2001, Frostegard et al. 1999, de

Lipthaya et al. 2004, Porteous et al. 1997, Braid et al. 2003, Sharma et al. 2007). An

important development was the appearance of commercial soil DNA extraction kits,

which resulted in convenient and standardized procedures and materials.

695 Commercial DNA extraction kits are now commonly used to assess taxonomic

and functional diversity, community composition, and population abundance (Roesch et

al. 2007, O’Brien et al. 2005, Wawrick et al. 2005, Shanks et al. 2006). Studies

comparing various kits (Lloyd-Jones & Hunter 2001, Whitehouse and Hottel 2007), or

comparing commercial kits to other methods (Roh et al. 2006, Arbeli and Fuentes 2007,

700 Carrigg et al. 2007), have shown that DNA yield and purity vary depending on

methodology and soil type. While these comparative studies are valuable, it is still

unclear to what extent these protocols yield genomic DNA which is representative of the

microbial community found within soil.

Our objective in this study was to optimize and assess the bias of a widely-used

705 commercial soil DNA extraction kit, the PowerSoil DNA Isolation Kit (MoBio, Carlsbad

CA). We hypothesized that cell lysis would be enhanced and DNA would be removed

from adsorption sites by conducting multiple extractions on a single sample, thereby

increasing genomic DNA yield and obtaining a more complete survey of microbial taxa.

This hypothesis was tested by 1) varying the extraction protocol and measuring DNA

710 yield and shearing for three soils with differing characteristics, and 2) examining

extraction bias in the genomic DNA obtained from successive extractions using an

improved method. Bacterial and fungal community composition was assessed using

terminal restriction fragment length polymorphism (T-RFLP) community profiling and

quantitative PCR (QPCR) analysis. These assays were followed by pyrosequencing of

715 16S rRNA genes from extractions with significantly-differing community profiles in

order to gain a comprehensive depiction of the bias associated with those extractions.

Materials and Methods

Soil collection

Soil was obtained from three forested locations with sterile sampling instruments, and

720 stored at -20°C. Details of soil properties are provided in Table 1.

Table 1: Soil properties. a. Data from Zak and Pregitzer (1990). ND – not determined.

% % % organic Depth, Soil pH Location Coordinates clay sand C cm

Clay 40.6 0.0 6.0 6.7 10 Toledo N 41° 38’ Metroparks, OH W 83° 26’

Sanda ND 72 4.4 3.9 10 Manistee N 44° 19’ National Forest, W 85° 54’ MI

Organic 5 70 10.9 5.9 2 Kent State N 41° 09’ University, OH W 81° 20’

DNA extraction and quantification

Soil DNA was extracted using a PowerSoil DNA Isolation Kit (MoBio

Laboratories, Carlsbad, CA), following the manufacturer’s instructions except as noted

725 below. For each soil type, we extracted DNA from 2 analytical replicate samples (250

mg fresh weight). An initial extraction, followed by 5 successive extractions, was

conducted on each replicate sample. A successive extraction involved adding new

aliquots of bead solution (not including beads) and solution C1 to the soil pellet after

initial lysis, centrifugation, and removal of supernatant containing crude DNA extract

730 (i.e. after step 7 in the manufacturer’s instructions). Lysis and centrifugation steps were

then repeated, resulting in a new supernatant that was subsequently processed separately

from previous supernatants. Four cell lysis protocols (treatments) were tested as

alternatives to step 5 in the manufacturer’s instructions (Table 2). All treatments

involved incubating soil at 70°C and then beadbeating; we also experimented with

735 extended incubation and beadbeating times, freezing at -80°C before incubation at 70°C,

and using a GenoGrinder (SPEX CertiPrep Metuchen, NJ) in place of a vortexer for

beadbeating. The volume of unused supernatant at different steps was measured so that

final soil DNA concentrations could be corrected for this loss.

740

Table 2: Alternative cell lysis procedures (DNA extraction treatments). a. Vortexer set at

745 maximum speed. b. GenoGrinder set at 1500 strokes per min.

Incubation before beadbeating Beadbeating procedure

Treatment 70°C -80°C Vortexinga GenoGrinderb

1 10 min 5 min

2 20 min 10 min

3 10 min 5 min 5 min

4 10 min 5 min 1 min

Extracted DNA concentration was quantified with Quant-iT PicoGreen

fluorescent stain (Invitrogen, Carlsbad, CA) using a Synergy 2 microplate reader

(BioTek, Winooski, VT) following the manufacturer’s instructions. Diluted lambda

750 phage DNA (Promega, Madison, WI) was used as a DNA concentration standard. All

DNA extractions were diluted to 1.25 ng/µL for subsequent analyses.

Pulsed-field gel electrophoresis (PFGE)

DNA shearing was assessed using a CHEF-DR III pulsed-field electrophoresis

system (BioRad, Hercules, CA). Samples were run in a 1.0% pulsed-field certified

755 agarose gel (BioRad) for 15 hours at 6V/cm and 120°C in 0.5x TBE buffer using a 1

second initial switch time and a 6 second final switch time. DNA fragment size was

estimated by comparison to Lambda ladder PFG marker (1018.5 - 48.5 kbp) and Lambda

DNA/PstI (11.5 - 2.4 kbp) markers (New England Biolabs, Ipswitch, MA). Gels were

stained with Gelstar nucleic acid stain (Cambrex Bio Science, Rockland, ME) and

760 photographed using a Gel Doc 2000 gel documentation system (BioRad).

Quantitative polymerase chain reaction

Copy number of bacterial and fungal small subunit ribosomal genes was

quantified by QPCR on an Mx3005P thermocycler (Stratagene, La Jolla, CA). QPCR

conditions included 0.2 ng genomic DNA/uL, 0.025 U/µL Taq DNA polymerase

765 (GeneChoice, Frederick, Maryland), 3 mM MgCl2, 1X ammonium polymerase buffer,

0.16 mM each dNTP (New England BioLabs), 10 µM each primer (Integrated DNA

Technologies, Coralville, IA), and 0.1 µg/µL bovine serum albumin. SYBR Green I was

added at a final concentration of 0.167X. Primers used to amplify small subunit

ribosomal fragments in the bacterial assay were Eub338F and Eub518R (Fierer et al.

770 2005). In the fungal assay, primers were FF390 and FR1 (Vainio and Hantula 2000).

After initial denaturation (3 min 95ºC), the PCR program consisted of 40 cycles,

including an 88ºC step for fluorescence quantification (30 sec 94ºC, 30 sec 57ºC, 90 sec

72ºC, and 33 sec 88ºC), followed by a 7 min extension at 72ºC and a melting curve

analysis. Copy number was quantified by comparing the cycle at which fluorescence

775 crossed a threshold (Ct) to a standard curve constructed using a serial dilution of a

plasmid containing an appropriate template. Assays were performed in duplicate for each

sample, with standards and negative controls included in each run. Efficiency of

amplification in each reaction was estimated using the method of Kontanis and Reed

(2006).

780 Community analysis by T-RFLP

PCR for T-RFLP was run in a DNA Engine Dyad Cycler (BioRad).

Amplification of the bacterial 16S gene was conducted with the primers 1392R and

Eub338F-0-III, which was labeled with 6-carboxyfluoroscein (Fam) (Blackwood et al.

2005). PCR was performed using 24-27 cycles under conditions identical to those

785 described above for QPCR except that Sybr Green I was not included and there was no

88°C step. Amplification of the fungal ITS region was conducted using primers NLB4R

and FAM-labeled NSI1F (Martin and Rygiewicz, 2005). Fungal PCR was conducted

using 0.02 – 0.12 ng of genomic DNA/ul, 0.03 U/µL Taq DNA polymerase, 2 mM

MgCl2, 1X ammonium polymerase buffer, 0.2 mM each dNTP, and 0.5 µg/µL bovine

790 serum albumin. The PCR program consisted of 3 min initial denaturation at 95°C, 30-35

cycles (94°C 30 sec – 60°C 30 sec – 72°C 90 sec), and a final 7 min extension at 72°C.

Cycle number was varied for samples to obtain a strong band without non-specific

amplification. PCR products were run on a 1.5% agarose gel to confirm successful

amplification. Triplicate PCRs were pooled prior to restriction digestion.

795 PCR product was digested overnight at 37°C with 10 units HaeIII (New England

Biolabs). Digests were purified with QIAquick Nucleotide Removal Kits (Qiagen,

Valencia, CA). Digested PCR products were sent to Ohio State Plant Microbe Genomics

Facility for fragment analysis on an Applied Biosystems 3730 DNA Analyzer using a

LIZ1200 size standard and minimum peak height of 50 fluorescence units. Peaks

800 between 50 and 600 bp were included in the analysis if they represented >1% of the

cumulative peak height for the sample.

Pyrosequencing

Partial bacterial 16S rRNA gene sequences were obtained from both replicates of

the first and sixth extractions of sand and clay soils using the coded primer approach to

805 multiplex pyrosequencing (Binladen et al. 2007). PCR amplification of the hypervariable

V4 region of the 16S rRNA gene was performed using 8 bp key-tagged eubacterial

primers, 563F and 802R (http://wildpigeon.cme.msu.edu/pyro/help.jsp). PCR mixtures

contained 1 μM of each primer (Integrated DNA Technologies), 1.8 mM MgCl2, 0.2 M

dNTPs, 1.5 X BSA (New England Biolabs), 1 unit of FastStart High Fidelity PCR

810 System enzyme blend (Roche Applied Science, Indianapolis, IN), and 10 ng of DNA

template. The PCR program consisted of 3 min initial denaturation at 95°C, 30 cycles

(95°C 45 sec – 57°C 45 sec – 72°C 1 min), and a final 4 min extension at 72°C. For each

sample, amplicons of three replicated PCR reactions were recovered using a QIAquick

Gel Extraction kit followed by QIAquick PCR Purification kit (Qiagen). Equimolar

815 amplicons were combined and submitted to pyrosequencing using a Genome Sequencer

FLX System (454 Life Sciences, Branford, CT) at Michigan State University-Genomics

Technology Support Facility. Sequences were excluded from analysis if the read length

was less than 150 bp or if primer sequences contained errors. Raw sequences were

processed through the Ribosomal Database Project (RDP) pyrosequencing pipeline

820 (http://wildpigeon.cme.msu.edu/pyro/index.jsp). Qualified sequences were clustered into

operational taxonomic units (OTUs) defined by 95% similarity using complete-linkage

clustering, and were assigned to phyla by the RDP-II Classifier using a 50% confidence

threshold (Wang et al. 2007). Sequences that could not be classified into a phylum at this

level of confidence were excluded from subsequent phylum composition analyses.

825 Sequences have been submitted to Genbank under accession numbers FJ240440-

FJ261916.

Statistical analysis

DNA yield per g soil in individual extractions (non-cumulative), ribosomal copies

per ng DNA, and weighted average fungal to bacterial gene copy ratio were analyzed by

830 mixed model analysis of variance using SAS Proc Mixed (SAS Institute, Cary, NC). Soil

subsamples were designated as subjects, with individual extractions treated as repeated

measurements. QPCR data was log-transformed prior to analysis in order to stabilize

variance.

The effects of extraction step, soil type, and extraction X soil interaction on

835 community composition (T-RFLP profile, phylum, and OTU composition) were analyzed

by redundancy analysis with Canoco software (Microcomputer Power, Ithaca, NY).

Relative abundances were square-root transformed resulting in analysis of Hellinger

distances between samples (Legendre and Gallagher 2001). Statistical significance was

determined using 999 random permutations of sample identity.

840 Results

DNA yield and shearing from four DNA extraction methods

Regardless of soil type or treatment, substantial quantities of DNA were isolated

with successive extraction steps (Figure 1). Cumulative DNA yields were highest for

organic soil, followed by clay and then sand soil. For organic and clay soils, treatment 4

845 (Geno Grinder + freeze-thaw) cumulative yields were higher than all other treatments for

all extraction steps. Treatment 4 cumulative yields also approached an asymptote more

quickly than in other extraction treatments, with ~80% of maximum DNA yield being

reached in 3 extraction steps (Figure 1). Liquid was entrained in the soil pellet after each

extraction (22%, 26%, and 35% for clay, sand, and organic soils, respectively), and could

850 therefore contribute DNA to the next extraction. However, DNA concentrations were

consistently twice as high as would be expected due to carryover of DNA for clay and

sand soils, and 1.5 times as high for organic soils, indicating substantial amounts of DNA

released from newly-lysed cells at each extraction step. In addition, substantially reduced

yields were obtained when, after a single, extended lysis step, repeated “washing” of the

855 soil pellet with extraction buffer was performed in place of heating or bead-beating (data

not shown).

860

Figure 1

865 Cumulative DNA yield in successive extractions conducted using four cell lysis

treatments.

A) organic soil; B) clay soil; C) sand soil.

Trt 1 80 A Trt 2 70 Trt 3 60 Trt 4 50 40 30 20

µg DNA / g Soil g / DNA µg 10 0 1 2 3 4 5 6 Extraction

60 B

µg DNA / g Soil g / DNA µg 10

0 1 2 3 4 5 6 Extraction 870

25 C

5 µg DNA / g Soil g / DNA µg

0 1 2 3 4 5 6 Extraction

875

880

885

In a mixed model analysis of variance, the following factors were found to

significantly (P < 0.05) affect quantity of DNA in single extractions (non-cumulative

yield): treatment, extraction step, soil type, and treatment X extraction step interaction.

Treatment 4 extracted a significantly higher amount of DNA than treatments 1 and 3;

890 treatment 2 was not significantly different from treatments 1, 3, or 4. DNA yield in

individual extractions declined with increasing numbers of extractions. The interaction

between treatment and extraction step was significant because of the greater amount of

DNA extracted by treatment 4 in early extraction steps and lower amounts in later

extraction steps compared to other treatments, indicating more complete extraction by

895 treatment 4 in early extraction steps. Differences between soil types were significant (P

< 0.05), with organic soil yielding the most DNA and sand soil yielding the least.

Pulsed field gel electrophoresis showed that the majority of extracted DNA was

between 4.5 and 50 kb in size (Figure 2). In first extracts, maximum size of DNA

fragments was larger for treatment 1 (~50 kb) than for other treatments (~25 kb), whereas

900 the reverse was true for sixth extracts. DNA fragment size was more consistent between

first and sixth extracts for treatments 3 and 4 than for treatments 1 and 2. Greater

shearing was apparent for final extracts than initial extracts in treatments 1 and 2 (Figure

2).

905

Figure 2

PFGE gel of 1st and 6th extractions for clay and sand soils. The top row of numbers refers

to extraction treatment, the bottom row refers to extraction step. Lane 9 contains lambda

910 ladder PFG marker (visible bands are 48.5 kbp and 98 kb). Lane 10 contains lambda PstI

marker (visible bands are 11.5, 5.1, 4.6, 2.8, and 2.6 kb).

915

More detailed characterizations, described below, were performed on DNA

920 extracts from treatment 4 because this treatment performed the best at quantitatively

extracting DNA from soil.

Quantification of ribosomal gene copy number in DNA extracts

Similar to the case for DNA yield, cumulative ribosomal gene copies/g soil

leveled off after one to three extraction steps in clay and sand soils (Figure 3). An

925 asymptote was also approached in organic soil for cumulative bacterial gene copies/g

soil, but not for cumulative fungal gene copies/g soil (Figure 3). QPCR efficiency was

not significantly affected by soil type or extraction step for either bacteria (mean = 0.88)

or fungi (mean = 0.76).

930

935

Figure 3

Cumulative microbial gene copies/g soil in successive extractions.

A) bacterial 16S gene; B) fungal 18S gene.

940

Organic A Clay

) 6 Sand 10 5 4 3 2 1 Copies /Copies g (X 10 0 1 2 3 4 5 6 Extraction

3.5 )

9 B 3.0 2.5 2.0 1.5 1.0

0.5 Copies /Copies g (X 10 0.0 1 2 3 4 5 6 Extraction

945

For the most part, ribosomal gene copies/ng genomic DNA declined with

successive extractions (Figure 4). Extraction step had a significant effect on bacterial

gene copies/ng DNA (P = 0.003), as did soil type (P = 0.019). Fungal gene copies/ng

DNA decreased with successive extractions in clay and sand soil, but increased in organic

950 soil (Figure 4B). Organic soil had significantly greater fungal gene copies/ng DNA than

other soils (P < 0.0001); however extraction step and soil X extraction step interactions

were not significant primarily due to variability between organic soil replicates.

955

960

Figure 4

965 Microbial gene copies/ng DNA for each extraction (not cumulative or weighted average).

Significant differences are described in the text.

A) bacterial 16S gene; B) fungal 18S gene

1.6 A Organic )

6 Clay 1.2 Sand

0.8

0.4 Copies /Copies ng (X 10 0.0 1 2 3 4 5 6 Extraction

970

2.0

) B 5 1.5

1.0

0.5 Copies /Copies ng (10 0.0 1 2 3 4 5 6 Extraction

The ratio of fungal to bacterial ribosomal genes weighted by the proportion of

DNA obtained in each extract represents the theoretical ratio of pooled successive

extractions. Weighted average fungal to bacterial ratio was significantly affected by soil

975 (P < 0.0001), and remained relatively constant after extraction step 2, although it

increased slightly for organic soil at extraction step 6 (Figure 5). Effects of extraction

step and soil X extraction step interaction were not significant.

980

985

990 Figure 5

Weighted average ratio of fungal to bacterial ribosomal gene copies. Significant

differences are described in the text.

Organic Clay 0.08 Sand 0.06

0.04

0.02

Weighted avg. ratio Weighted 0.00 1 2 3 4 5 6 Extraction

995

1000

Community composition profiles

Using redundancy analysis of bacterial T-RFLP profiles, soil type was shown to

1005 explain a large portion of variability in bacterial community composition (P = 0.01,

variance explained = 47%). Extraction step also explained a significant portion of

bacterial community profile variability (P = 0.012, variance explained = 13%). This is

evident in the canonical ordination plot displaying soil type and extraction step effects

(Figure 6A), where gradients in clay and sand ordination scores are directly related to

1010 extraction step. The interaction between soil and extraction step was not significant for

bacterial community profiles. When soils were analyzed individually, extraction step

was significant in clay and sand soils (P < 0.05), but not significant in organic soil.

The theoretical T-RFLP profile that would have resulted if DNA extractions were

pooled prior to PCR is represented by the average T-RFLP profile weighted by the

1015 proportion of DNA obtained in each extract. Figure 6A shows that the weighted average

profile of all six extractions lies between extractions 1, 2, and 3, and is very similar to the

weighted average profile of the first three extractions.

Soil type had a significant effect on fungal T-RFLP profiles (P = 0.001, variance

explained = 65%), whereas extraction step had no effect (P = 0.645) (Figure 6B). Soil X

1020 extraction interactions were not significant for fungal community profiles, and extraction

step was also not significant for any soil analyzed individually.

Figure 6

1025 Canonical principal components plot of T-RFLP profiles derived from redundancy

analysis. Soil type and extraction are significant as described in text. A) bacterial

profiles; B) fungal profiles. “Wted ave-3” is the weighted average of the first three T-

RFLP profiles; “wted ave-6” is the weighted average of all six T-RFLP profiles.

A Organic 1 Clay Sand 6 wted avg-3 0.6 5 wted avg-6 4 3 1 6 0.2 2 5 -0.2 5 4 Axis 2 (16%) Axis 4 6 3 3 2 2 -0.6 1 1 -1 -0.5 0 0.5 1 1.5 Axis 1 (34%) 1030

1 B

0.5

Axis 2 (27%) 2 (27%) Axis -0.5

-1 -1 -0.5 0 0.5 1 1.5 Axis 1 (42%)

Taxonomic characterization of bacterial extraction bias

1035 To obtain a detailed taxonomic understanding of the bacterial community

composition bias associated with DNA extraction, we performed pyrosequencing of 16S

rRNA gene PCR amplicons from the first and sixth DNA extracts of duplicate

subsamples of clay and sand soil. To assess bias with respect to both broad-scale and

fine-scale phylogenetic groups, sequences were classified by separate methods into phyla

1040 and 95% similarity OTUs (approximately genus-level ranking). Between 1531 and 3174

sequences were obtained per sample, resulting in 179 to 1028 OTUs per sample.

Average sequence length was 207 bp.

Soil type and extraction step affected taxonomic composition of DNA extracts at

both the phylum and OTU levels, but soil type X extraction step interactions were not

1045 significant (Table 3). There was a greater level of unexplained variation in OTU

composition than there was in phylum composition, primarily due to a greater amount of

variation explained by extraction step for phylum composition (Table 3). For both clay

and sand soil, the phyla Acidobacteria, Gemmatimonades, Nitrospira, and WS3 were

strongly affiliated with the initial extraction, whereas Proteobacteria and Actinobacteria

1050 were affiliated with the final extraction (Table 4).

1055

Table 3: Variance partitioning of bacterial sequence community composition. * P < 0.1; **

P < 0.05; NS not significant.

Phylogenetic Soil Extraction Soil X Extraction

Resolution % % %

OTU 29.4** 18.4* NS

Phylum 26.8* 32.9** NS

Table 4: Response of bacterial phyla composition to extraction step revealed through

1060 redundancy analysis.a

Ave. % abundancec Cumulative Phylab T-value Extract 1 Extract 6 fit, %d

Acidobacteria 39.1 16.8 50.8 0.309

Actinobacteria 6.6 15.4 43.9 -0.239

Bacteroidetes 1.2 2.1 6.3 -1.177

BRC1 0.1 0 31.7 0.827

Chlamydia 0.8 0.7 1.6 2.101

Chloroflexi 0.8 0.8 2.5 3.195

Fibrobacteres 0 <0.1 14.3 -1.148

Firmicutes 1.6 3.1 5.2 -1.319

Fusobacteria <0.1 0 14.3 1.148

Gemmatimonades 1.2 0.6 37.5 0.400

Nitrospira 0.1 <0.1 44.8 0.448

OD1 0.6 0.2 27.3 0.492

OP10 <0.1 <0.1 12.0 1.458

Planctomycetes 2.2 7.8 13.0 -1.465

Proteobacteria 24.4 40.8 59.3 -0.361

Spirochaetes <0.1 <0.1 0.5 -6.543

Thermomicrobia <0.1 0 14.3 1.148

TM7 0.2 0.9 5.2 -1.930

Verrucomicrobia 20.4 10.5 23.9 0.511

WS3 0.7 0.2 51.1 0.345

a. Bold phyla have a critical T-value closer to zero than the T-value for extraction step in

this redundancy analysis (0.573). This indicates that extraction step is a significant

predictor for the abundance of bolded phyla.

b. Sequence classification into phyla was based on RDP-II classifier at 50% confidence

1065 (Wang et al. 2007).

c. Relative abundances are averaged across sand and clay soil samples.

d. Percent of variability in phylum abundance explained by extraction step 1 vs. 6.

Discussion

DNA extraction bias is a frequently cited but poorly understood limitation of

1070 molecular characterizations of environmental microbial communities. We have

quantified the bias associated with incomplete extraction by one commonly-used method,

and have shown that it may be essential to consider this bias in certain cases.

Total extracted DNA has been used as an index of soil microbial biomass, but

Leckie et al. (2004) found no correlation between DNA extracted and biomass measured

1075 by phospholipid fatty acid or chloroform fumigation extraction. Our results suggest that

this lack of correlation may be due to incomplete lysis of cells in a single DNA

extraction. We found that DNA continued to be isolated with successive extractions, and

that this was dependent on multiple cell lysis steps performed in fresh extraction buffer.

Washing the soil pellet with DNA extraction buffer did not substantially improve DNA

1080 yield. However, when a tissue homogenizer was used for the bead-beating step, the total

amount of DNA and number of copies of ribosomal genes generally approached an

asymptote with three extractions. This is similar to the DNA yield results of Bürgmann

et al. (2001), who also performed multiple extractions on individual soil samples. In our

case, however, the more vigorous procedure resulted in only slightly more shearing of

1085 genomic DNA visible on PFGE gels in first extracts, and less shearing in sixth extracts,

compared to the vortexing procedure recommended by the manufacturer. PCR product

could be consistently obtained and QPCR efficiency was unaffected by extraction step.

The only indication of potential damage to DNA in later extractions was a decline in the

number of bacterial and fungal ribosomal copies/ng genomic DNA (except fungi in

1090 organic soil). However, the decline in ribosomal copies/ng genomic DNA could also

result from shifts in the microbial community from which DNA was extracted, for

example toward organisms with fewer ribosomal gene copies.

Extraction of bacterial genomic DNA was shown to be biased, particularly in clay

and sand soils. However, because the majority of the DNA was extracted within the first

1095 few extractions, the weighted average T-RFLP profile of all six extractions tended to be

between the first three profiles in a canonical ordination plot (Fig. 6A). Furthermore, the

weighted average profile of all extractions was well-approximated by the weighted

average of the first three extractions. This suggests that pooling three successive DNA

extractions from a sample before performing T-RFLP may be a good compromise to

1100 overcome extraction bias. Cumulative bacterial ribosomal copy number/g soil did not

increase substantially after three extractions, indicating that it may also be appropriate to

pool three successive extractions prior to this analysis. A complete characterization of

the bacterial diversity or taxonomic makeup of a soil sample through sequencing (e.g.

Roesch et al. 2007), however, would not be possible without analysis of additional DNA

1105 extractions. Diversity (e.g. OTU richness) present in a late, low concentration extraction

is not masked (or diluted out) by pooling with other DNA extracts in the same way as for

community profiles (Blackwood et al. 2007), because counts and presence/absence are

not weighted by abundance.

It is interesting to note that phyla representing two of the historically important

1110 cultivatable groups from soil, Actinobacteria and Proteobacteria (Janssen 2006), were

approximately twice as abundant in the sixth extraction as in the first (Table 4). In

contrast, groups often detected using molecular methods but rarely cultivated from soil

(including Acidobacteria, Gemmatimonades, and Verrucomicrobia), were approximately

half as abundant in the sixth extraction as in the first. These patterns were consistent

1115 across soils and replicates (Table 4). Although the composition of the sixth extraction

represents a low overall proportion of the total DNA that can be extracted from a soil,

this is the first suggestion that the reported severity of phylum-level cultivation bias

(Janssen 2006) may be inflated by DNA extraction bias.

Results for fungi were quite different from those for bacteria. No extraction bias

1120 in fungal community composition was detected using T-RFLP (Figure 6B). Fungal

ribosomal copy number/g soil reached an asymptote after three extractions in clay and

sand soil, but continued to increase in organic soil (Figure 3B). The increasing fungal

ribosomal copy number/ng genomic DNA extracted from organic soil indicates that there

was a bias against extraction from fungi in the early extraction steps in this soil. This

1125 effect of organic soil is unlikely to be due to binding of released DNA to humic

compounds or particulate organic matter because total DNA and bacterial ribosomal copy

number/g soil both reached asymptotes for organic soil. We hypothesize that there is a

greater degree of protection of fungi in organic soil due to growth within small organic

particles that must be disrupted before the fungal cells are lysed. Alternatively, the

1130 fungal community in organic soil may be such that their cell walls are more difficult to

disrupt compared to fungi in other soils, although there is no a priori reason to expect this

would be the case.

We interpret the ratio of fungal to bacterial ribosomal gene copies as an index of

fungal to bacterial ratio present in soil for comparative purposes, rather than indicative of

1135 the actual biomass ratio, due to differences between bacteria and fungi in PCR efficiency,

ribosomal gene copy number per genome, etc. Despite the continued increase of fungal

ribosomal copy number in organic soil, the weighted average ratio of fungal to bacterial

ribosomal copies was quite stable in all soils after the second extraction. However, the

ratio increased somewhat for the sixth extraction in organic soil, and it is unclear how

1140 additional extraction steps would alter it.

In conclusion, we have found that substantial quantities of DNA are not extracted

with a commonly-used DNA extraction procedure. This results in biased estimates of

DNA quantity, ribosomal copy number, and bacterial community composition, but in

some cases this bias can be greatly reduced by repeating the extraction on the soil sample

1145 three times and pooling the successive extractions. Comparisons of soils based on single

DNA extractions may still be valid, but should be recognized as representing an easily-

lysed portion of the community. Overcoming bias is of greatest concern in two cases: 1)

when research questions or downstream methods demand great accuracy and a

comprehensive treatment of the community (such as quantification of diversity) and 2)

1150 when DNA is extracted from a set of soils with divergent soil properties or communities.

In the latter case the bias may be inconsistent between soils, resulting in an incorrect

portrayal of differences between soil communities. This is demonstrated by differences

we obtained here between an organic soil and two mineral soils.

References

1155 Amann RI, Ludwig W, Schleifer KH. 1995. Phylogenetic identification and in situ

detection of individual microbial cells without cultivation. Microbiological

Reviews 59:143-169.

Arbeli Z, Fuentes CL. 2007. Improved purification and PCR amplification of DNA from

environmental samples. FEMS Microbiology Letters 272: 269-275.

1160 Binladen J, Gilbert MT, Bollback JP, Panitz F, Bendixen C, Nielsen R, Willerslev E.

2007. The use of coded PCR primers enables high-throughput sequencing of

multiple homolog amplification products by 454 parallel sequencing. PLoS ONE

2: e197.

Blackwood CB, Hudleston D, Zak DR, Buyer JS. 2007. Interpreting ecological diversity

1165 indices applied to T-RFLP data: insights from simulated microbial communities.

Applied and Environmental Microbiology 73: 276-283.

Blackwood CB, Oaks A, Buyer JS. 2005. Phylum- and class-specific PCR primers for

general microbial community analysis. Applied and Environmental Microbiology

71: 6193-6198.

1170 Borneman J, Skroch PW, O’Sullivan KM, Palus JA, Rumjanek NG, Jansen JL, Nienhuis

J, Triplett EW. 1996. Molecular microbial diversity of an agricultural soil in

Wisconsin. Applied and Environmental Microbiology 62: 1935-1943.

Braid MD, Daniels LM, Kitts CL. 2003. Removal of PCR inhibitors from soil DNA by

chemical flocculation. Journal of Microbiological Methods 52: 389-393.

1175 Bürgmann H, Pesaro M, Widmer F, Zeyer J. 2001. A strategy for optimizing quality and

quantity of DNA extracted from soil. Journal of Microbiological Methods 45: 7-

20.

Carrigg C, Rice O, Kavanagh S, Collins G, O’Flaherty V. 2007. DNA extraction method

affects microbial community profiles from soils and sediment. Applied

1180 Microbiology and Biotechnology 77: 955–964.

de Lipthaya JR, Enzingerb C, Johnsena K, Aamanda J, Sørensen SJ. 2004. Impact of

DNA extraction method on bacterial community composition measured by

denaturing gradient gel electrophoresis. Soil Biology and Biochemistry 36: 1607-

1614.

1185 Fierer N, Jackson JA, Vilgalys R, Jackson RB. 2005. Assessment of soil microbial

community structure by use of taxon-specific quantitative PCR assays. Applied

and Environmental Microbiology 71: 4117-4120.

Frostegard A, Courtois S, Ramisse V, Clerc S, Bernillon D, Le Gall F, Jeannin P, Nesme

X, Simonet P. 1999. Quantification of Bias Related to the Extraction of DNA

1190 Directly from Soils. Applied and Environmental Microbiology 65: 5409-5420.

Griffiths RI, Manefield M, Whiteley AS, Bailey MJ. 2004. DNA and RNA extraction

from soil, p. 3-18. In Kowalchuk G, de Bruijn FJ, Head IM, Akkermans ADL,

van Elsas JD (ed.), Molecular Microbial Ecology Manual, Dordrecht, The

Netherlands: Kluwer Academic Publishers.

1195 Hawksworth DL, Rossman AY. 1997. Where are all the undescribed fungi?

Phytopathology 87: 888-891.

Janssen PH. 2006. Identifying the dominant soil bacterial taxa in libraries of 16S rRNA

and 16S rRNA genes. Applied and Environmental Microbiology 72: 1719-1728.

Kontanis EJ, Reed FA. 2006. Evaluation of real-time PCR amplification efficiencies to

1200 detect PCR inhibitors. Journal of Forensic Sciences 51: 795–804.

Leckie SE, Prescott CE, Grayston SJ, Neufeld JD, Mohn WW. 2004. Comparison of

chloroform fumigation-extraction, phospholipid fatty acid, and DNA methods too

determine microbial biomass in forest humus. Soil Biology and Biochemistry 36:

529-532.

1205 Legendre P, Gallagher ED. 2001. Ecologically meaningful transformations for ordination

of species data. Oecologia 129: 271–280.

Lloyd-Jones G, Hunter DWF. 2001. Comparison of rapid DNA extraction methods

applied to contrasting New Zealand soils. Soil Biology and Biochemistry 33:

2053-2059.

1210 Lord NS, Kaplan CW, Shank P, Kitts CL, Elrod S. 2002. Assessment of fungal diversity

using terminal restriction fragment (TRF) pattern analysis: comparison of 18S and

ITS ribosomal regions. FEMS Microbiology Ecology 42: 327-337.

Martin KJ, Rygiewicz PT. 2005. Fungal-specific PCR primers developed for analysis of

the ITS region of environmental DNA extracts. BMC Microbiology 5: 1-11.

1215 Milling A, Gomes NCM, Oros-Sichler M, Gotz M, Smalla K. 2005. Nucleic acid

extraction from environmental samples, p. 1-24. In Osborn AM, and Smith CJ

(ed.), Molecular Microbial Ecology, New York, New York: Taylor & Francis.

Moore E, Arnscheidt A, Kruger K, Strompl S, Mau M. 2004. Simplified protocols for the

preparation of genomic DNA from bacterial cultures, p. 3-18. In Kowalchuk G, de

1220 Bruijn FJ, Head IM, Akkermans ADL, and van Elsas JD (ed.), Molecular

Microbial Ecology Manual, Dordrecht, The Netherlands: Kluwer Academic

Publishers.

O’Brien HE, Parrent JL, Jackson JA, Moncalvo JM, Vilgalys R. 2005. Fungal community

analysis by large-scale sequencing of environmental samples. Applied and

1225 Environmental Microbiology 71: 5544-5550.

Porteous EA, Seidler RJ, Watrud LA. 1997. An improved method for purifying DNA

from soil for polymerase chain reaction amplification and molecular ecology

applications. Molecular Ecology 6: 787-791.

Robe P, Nalin R, Capellano C, Vogel TM, Simonet P. 2003. Extraction of DNA from

1230 soil. European Journal of Soil Biology. 39: 183-190.

Roesch LFW, Fulthorpe RR, Riva A, Casella G, Hadwin AKM, Kent AD, Daroub SH,

Camargo FAO, Farmerie WG, Triplett EW. 2007. Pyrosequencing enumerates

and contrasts soil microbial diversity. ISME Journal 1: 283-290.

Roh C, Villatte F, Kim B-G, Schmid RD. 2006. Comparative Study of Methods for

1235 Extraction and Purification of Environmental DNA from Soil and Sludge

Samples. Applied and Environmental Microbiology 134: 97-112.

Shanks OC, Domingo JWS, Lamendella R, Kelty CA, Graham JE. 2006. Competitive

metagenomic DNA hybridization identifies host-specific microbial genetic

markers in cow fecal samples. Applied and Environmental Microbiology 72:

1240 4054-4060.

Sharma PK, Capalash N, Kaur J. 2007. An improved method for single step purification

of metagenomic DNA. Molecular Biotechnology 36: 61-63.

Tebbe CC, Vahjen W. 1993. Interference of humic acids and DNA extracted directly

from soil in detection and transformation of recombinant DNA from bacteria and

1245 a yeast. Applied and Environmental Microbiology 59: 2657-2665.

Torsvik V, Øvreås L. 2002. Microbial diversity and function in soil: from genes to

ecosystems. Current Opinion in Microbiology 5: 240-245.

Vainio J, Hantula J. 2000. Direct analysis of wood-inhabiting fungi using denaturing

gradient gel electrophoresis of amplified ribosomal DNA. Mycological Research

1250 104: 927–936.

van Elsas JD, Frois-Duarte G, Keijzer-Wolters A, Smit E. 2000. Analysis of the

dynamics of fungal communities in soil via fungal-specific PCR of soil DNA

followed by denaturing gradient gel electrophoresis. Journal of Microbiological

Methods 43: 133-151.

1255 Wang Q, Garrity GM, Tiedje JM, Cole JR. 2007. Naïve Bayesian classifier for rapid

assignment of rRNA sequences into the new bacterial taxonomy. Applied and

Environmental Microbiology 73: 5261-5267.

Wawrik B, Kerkhof L, Kukor J, Zylstra G. 2005. Effect of different carbon sources on

community composition of bacterial enrichments from soil. Applied and

1260 Environmental Microbiology 71: 6776-6783.

Whitehouse CA, Hottel HE. 2007. Comparison of five commercial DNA extraction kits

for the recovery of Francisella tularensis DNA from spiked soil samples.

Molecular and Cellular Probes 21: 92-96.

Zak DR, Pregitzer KS. 1990. Spatial and temporal variability of nitrogen cycling in

1265 northern lower Michigan. Forest Science 36: 367-380.

Chapter Three

Taxa-area relationship and neutral dynamics influence the diversity of fungal

communities on senesced tree leaves

(This chapter was published online in Environmental Microbiology in April 2012)

Abstract

This study utilized individual senesced sugar maple and beech leaves as natural sampling units within which to quantify saprotrophic fungal diversity. Quantifying communities in individual leaves allowed us to determine if fungi display a classic taxa- area relationship (species richness increasing with area). We found a significant taxa- area relationship for sugar maple leaves, but not beech leaves, consistent with Wright’s species-energy theory. This suggests that energy availability as affected plant biochemistry is a key factor regulating the scaling relationships of fungal diversity. We also compared taxa rank abundance distributions to models associated with niche or neutral theories of community assembly, and tested the influence of leaf type as an environmental niche factor controlling fungal community composition. Among rank abundance distribution models, the zero-sum model derived from neutral theory showed the best fit to our data. Leaf type explained only 5% of the variability in community composition. Habitat (vernal pool, upland, or riparian forest floor) and site of collection explained >40%, but could be attributed to either niche or neutral processes. Hence, although niche dynamics may regulate fungal communities at the habitat scale, our

evidence points towards neutral assembly of saprotrophic fungi on individual leaves, with energy availability constraining the taxa-area relationship.

Introduction

Fungal community composition has been shown to be an important factor regulating decomposition (Balser and Firestone 2005, Waldrop and Firestone 2006).

Therefore, our ability to predict decomposition rates and responses to environmental changes may be enhanced with increased understanding of the processes that regulate fungal distributions. Variation in composition of ecological communities is commonly divided into two components: alpha diversity (number and evenness of taxa within a sampling unit) and beta diversity (taxa turnover among areas) (Gaston and Blackburn

2000).

For macroorganisms, it has been frequently observed that there is a correlation between the size of habitat patches or survey areas and the most fundamental measure of alpha diversity, number of taxa detected (Rosenzweig 1995, Connor and McCoy 2001,

Lomolino 2001, Drakare et al. 2006). The “taxa-area relationship” (TAR) refers to the shape of the increase in number of taxa with increasing area, and has been most often modeled as a power law (S=cAz) where S is number of species, A is area, c is the intercept in log-log space, and z is a constant related to the rate of species turnover across space

(Arrhenius,1921, Gleason 1922). TARs have been used to extrapolate species richness

(Colewell and Coddington 1994, He and Legendre 1996, Plotkin et al. 2000), estimate regional diversity inventories (Chong and Stohlgren 2007), and compare species abundance distributions (May 1975, Harte et al. 1999, Pueyo, 2006). TARs are also used

as an important tool in conservation efforts to protect species from habitat fragmentation and destruction (Faith 2008). Quantifying fungal TARs may help us understand processes regulating fungal community assembly and determine how sampling design affects our ability to detect fungal diversity.

Communities in discrete habitat patches (e.g. islands) have proven useful in studying TAR patterns because such communities have well-defined boundaries

(MacArthur and Wilson 1967, Schoener 1976, Bell et al. 2005). Although a few studies have utilized leaves as discrete habitats for microbial communities (Kinkel et al. 1987,

Jacques et al. 1995, Newell & Fell 1997), most previous molecular studies examining forest floor fungal communities have utilized homogenized mixed “grab” samples of leaves (O’Brien et al. 2005, Neubert et al. 2006, Blackwood et al. 2007a, Keeler et al.

2009, Redford et al. 2010). However, early after leaf fall, individual leaf boundaries must limit the size of fungal communities because the leaves are each colonized individually in the canopy, during leaf fall, and on top of the existing forest floor. At later successional stages it is unclear to what extent the scale of communities coincides with individual leaves because some fungi will colonize leaves from existing mycelia in the forest floor. However, even if individual fungi extend beyond the boundaries of a single leaf, individual leaves are still a natural sampling unit because they represent unit resources with differing characteristics depending on plant species of origin, and in which a highly diverse assemblage of fungal taxa can be found (Cooke and Rayner 1984). In other words, fungi interact with leaves as discrete units, unlike, for example, soil cores or vegetation quadrats, which are arbitrary sub-samples of a larger area. Hence, taking

advantage of the individual nature of forest floor leaves may provide important insight into fungal community organization.

Fungal alpha and beta diversity may be influenced by a variety of environmental and biological mechanisms that act on the organism niche. The biochemical makeup of leaves (e.g., proportion of soluble compounds, hemicelluloses, cellulose and lignin) represents the carbon and energy resources available to saprotrophic fungi. Some compounds (e.g., solubles) are easily metabolized and support a wide range of organisms; in contrast, lignin requires oxidative enzymes predominantly produced by white-rot

Basidiomycetes. Because biochemical composition varies among tree species (Hobbie et al. 2006), leaf species type may act as an “environmental filter”, favoring microbial taxa that are able to exploit the resources present (Weand et al. 2010, Wu et al. 2011). In addition, competitive interactions may influence diversity of fungi that share similar niches. Fungal diversity may also be affected by environmental variability in abiotic conditions such as soil moisture. In general, saprotrophic fungal species richness is lower in aquatic habitats than terrestrial habitats (Gessner et al. 2010, Nickolcheva et al. 2005,

Fischer et al. 2009). This could be mediated by lower spatial and temporal variation in aquatic habitats, as well as enhanced water-dependent dispersal among microsites, leading to greater dominance by competitively superior taxa. Under these conditions, we might also expect a shallower TAR (i.e., a lower z value) because of lower beta diversity among habitat patches.

Alternatively, neutral theory seeks to explain community composition strictly by immigration dynamics, ignoring species niches (Hubbell 2001, Maurer and McGill

2004). Here, we use two methods of testing the importance of niche determinism versus neutral dynamics in structuring communities. The first method is examination of rank abundance distributions. One prediction of neutral theory is that rank abundance distributions will be consistent with the zero sum model (Hubbell 2001). In contrast, niche-based theory has been used to derive rank abundance distribution models including the pre-emption (geometric) and broken stick models (Motomuro 1932, MacArthur

1957). The lognormal and Zipf-Mandlebrot are more flexible models that have also been related to niche theory (McGill et al. 2007). The second method is to explain variation in composition among communities (i.e., beta diversity). If variation in community composition is due to environmental filtering, we would expect community composition to be related to environmental or resource gradients (Jongman et al. 1995, Chase et al.

2004), whereas this would not be the case under neutral dynamics (Hubbell 2001).

Our goals in this study were to a) characterize fungal taxa-area relationships utilizing the discrete nature of leaf habitat, b) test the importance of leaf type and habitat in determining fungal alpha and beta diversity, and c) determine whether taxa rank abundances correspond with niche-based and/or neutral models of community assembly.

To determine the effects of leaf biochemistry and habitat soil moisture, we collected two leaf types of differing recalcitrance from the forest floor of three habitats providing variable levels of moisture, vegetation type, and soil conditions: upland forest, riparian forest, and vernal pools. While upland forest floor remains dry other than during rainfall or snowmelt, vernal pools are saturated with water in late winter, spring, and periodically over the rest of the year, and riparian is occasionally flooded. Fungal diversity and

community composition were determined by PCR amplification and sequencing of the fungal internal transcribed spacer (ITS) region, a common fungal taxonomic marker

(Martin and Rygiewicz 2005).

Methods

Field site and sample collection

All leaves were collected from Jennings Woods, a 30 hectare deciduous forest in northeastern Ohio. The two most dominant tree species are sugar maple (Acer saccharum, 22% relative abundance of all forest trees) and beech (Fagus grandifolia,

19% relative abundance). Two sampling locations were selected within each of three habitats: upland forest, riparian forest (within 30-50 m of the West Branch of the

Mahoning River), and seasonally saturated vernal pools. The vernal pools were embedded within the riparian forest and ~15 m from the riparian forest sampling locations.

A total of 30 leaves were randomly selected from 26.7 x 31.8 cm areas in each habitat during June 2008. Among the 30 leaves collected at each sampling location, the largest and smallest sugar maple and beech leaves were used in this study, resulting in four leaves used per site. Leaves were photographed in the field, collected with sterile implements, and placed into individual, sterile, pre-weighed containers in coolers containing dry ice. Leaf surface area was calculated from digital images using the software WinRhizo (2007 version, Regent Instruments Inc., Canada) with settings adapted to leaves. Samples were stored at -80°C. Leaves were lyophilized (VirTis

Genesis 25EL, Biopharma Process Systems Ltd, Winchester, NH) and dry leaf weight recorded.

DNA extraction and PCR amplification

Leaves were ground in sterile collection vials in a Genogrinder 2000 (SPEX

CertiPrep, Metuchen, NJ). DNA was extracted as in Wu et al. (2011). The fungal ITS1 region was amplified using primers NSI1F (5’ - GATTGAATGGCTTAGTGAGG) and

5.8SR (5’ – GCTGCGTTCTTCATCGA) (Martin and Rygiewicz 2005). PCRs were performed with a DNA Engine Dyad Peltier Thermal Cycler (Bio-Rad, Hercules, CA) using 0.025 U/µL Taq DNA polymerase, 3 mM MgCl2, 1X ammonium polymerase buffer (B-Bridge International, Mountain View, CA), 0.2µM each primer (Integrated

DNA Technologies, Coralville, IA), 0.16 mM each dNTP, and 0.1 µg/µL bovine serum albumin (New England Biolabs, Ipswich, MA). PCR reaction conditions were: initial denaturation for 3 min at 95 ºC, 30-35 cycles of denaturation for 30 s at 94 ºC, primer annealing for 30 s at 60 ºC and extension for 90 s at 72 ºC, and a final extension for 7 min at 72 ºC. Negative controls were included in each PCR run. PCR products were run on a

1.5% agarose gel to confirm successful amplification. Cycle number was varied for each sample to obtain a strong band without non-specific amplification. Three replicate PCRs were performed for each sample, and these were pooled prior to the cloning ligation reaction.

Clone libraries

PCR product was purified using the UltraClean PCR Clean-up kit (MoBio,

Carlsbad, CA). Amplicons were then ligated into a pGEM-T vector with overnight incubation and used to transform competent E. coli cells following the manufacturer’s instructions (Promega, Madison, WI). One hundred ninety-two white colonies per sample were used to inoculate wells of LB broth with 10% glycerol. Cultures were grown overnight and stored at -80ºC. Sequencing was performed at the Genome

Sequencing Center at Washington University in St. Louis, MO using M13 primers.

Plasmid sequence was removed and amplicon sequences were trimmed for quality in Sequencher (Gene Codes Corporation, Ann Arbor, MI) using default settings.

Clustering of sequences into operational taxonomic units (OTUs) was performed using

BlastClust (Cummings et al. 2002) implemented in the MPI Bioinformatics Toolkit

(Biegert et al. 2006). Sequences were only used if they contained both forward and reverse primer sequences or contained one primer and were longer than the longest sequence containing both primers (i.e., longer than 390 bp). This criterion ensures that

Blast score variation is not affected by length variation of partially-sequenced amplicons.

A percent identity threshold of 97% was used to define OTUs. For OTUs with ≥ 5% relative abundance on a leaf, nearest neighbors in Genbank were identified using Blast searches on representative sequences (Altschul et al. 1997). Sequences generated in this study have been submitted to GenBank under accession numbers JN394655 - JN397349.

Statistical analysis

The software EstimateS 8.2 (Colwell 2009) was used to calculate sample-based rarefaction curves and alpha diversity indices (observed taxa richness [S], expected

richness [Chao 2], Shannon diversity [H’], Simpson diversity [1/D]) for each leaf (Gotelli and Colwell 2001, Colwell et al. 2004, Mao et al. 2005). The Simpson evenness index

(E1/D sensu Smith and Wilson 1996) was calculated from S and 1/D values obtained by rarefaction. Diversity indices standardized to a common clone library size by rarefaction were compared with mixed model analysis of variance using SAS Proc Mixed (SAS

Institute, Cary, NC). The effects of leaf size, leaf species, and habitat type were examined as fixed effects in a three-way ANOVA with all potential interactions.

Collection site within habitat type was designated a random effect (2 sites per habitat).

Where leaf size was significant, we calculated z by regression of log-transformed S

(obtained by rarefaction) against log-transformed surface area. Other terms initially included in the regression model (i.e., habitat as a fixed factor, habitat × ln(area) interaction, and site within habitat as a random factor) were eliminated stepwise to find the model minimizing the Akaike Information Criterion corrected for small sample size

(AICc; Burnham and Anderson 2002).

To quantify beta diversity, taxa turnover was measured with the Jaccard similarity index generated in the software R (version 2.10.1) using the vegan package (Oksanen

2010). Redundancy analysis (RDA) was used to assess the effects of habitat, site nested within habitat, leaf type, and leaf size on OTU composition. RDA was performed on 1)

Hellinger distances, an abundance-based metric that excludes joint absences and downweights the most abundant OTUs and effects of differences in sampling effort

(Legendre and Gallagher 2001) and 2) Sorenson similarity indices calculated from presence/absence of OTUs. Statistical significance of RDA results was determined with

the software Canoco (Microcomputer Power, Ithaca, NY) using 999 random permutations of sample identity.

OTU rank abundances for each leaf were fit to broken stick, preemption, lognormal, Zipf, and Zipf-Mandelbrot rank abundance models using the command “radfit” found in the R package vegan (Oksanen 2010), and to the zero-sum model using TeTame

(Jabot et al. 2008). Akaike Information Criterion (AIC) values were compared to determine which model provided the best fit to the empirical data (Burnham and

Anderson 2002). The comparison of AIC values generated by radfit and TeTame is biased against the zero-sum model (F Jabot, personal communication, and as fully described in Supplementary Table S1). However, because the zero-sum model is unambiguously supported using this conservative test (see Results), this bias only strengthens our conclusions.

Results

Sequence analysis

Leaves were collected from two forest floor sites per habitat. At each site, sequences were obtained separately from the largest and smallest beech and maple leaves. A total of 2701 sequences met our selection criteria (containing both primers or longer than the longest sequence containing both primers), and these were clustered into

416 distinct OTUs at 97% sequence similarity. The majority of OTUs (278) were singletons, and these comprised 11.1% of total OTU relative abundance. We classified

41 OTUs as “dominant” (≥ 5% relative abundance on at least one leaf), comprising

61.2% of total OTU relative abundance. Each leaf had 1-6 dominant OTUs

(Supplementary Fig. S1). This dominance could be extreme; the most dominant OTU had greater than 80% relative abundance on two leaves. Nearest neighbors of dominant

OTUs in GenBank included 27 genera; 6 in the Basidiomycetes and 21 in Ascomycetes

(Table 1). The OTUs with dominant relative abundance on the most leaves (7 out of 24 leaves) were OTU 2 (Phlogicylindrium as nearest neighbor) and OTUs 5 and 6

(Polyscytalum; Table 1). OTU 1 (Mycena) was the most dominant overall OTU with a

53% average relative abundance. Dominant OTU distributions for each leaf are shown in

Supplementary Fig. S1.

Supplementary Figure S1(a)

Relative abundance charts show proportion of community occupied by dominant (multi- colored) and non-dominant (black) OTUs on vernal pool leaves

Supplementary Figure S1(b)

Relative abundance charts show proportion of community occupied by dominant (multi- colored) and non-dominant (black) OTUs on upland leaves

Supplementary Figure S1(c)

Relative abundance charts show proportion of community occupied by dominant (multi- colored) and non-dominant (black) OTUs on riparian leaves

Alpha diversity and taxa-area relationship

Fifty-nine to 158 sequences per leaf met our analysis criteria. Rarefaction was used to reduce bias associated with unequal sample sizes from different communities by simulating diversity index values corresponding to the number of individuals in the smallest sample (Magurran 2004). The rarefaction curves for H’ and 1/D reached an asymptote by 59 sequences, although S continued to rise (Supplementary Fig. S2).

Therefore, we use S obtained from rarefaction as an indicator for comparative purposes.

Supplementary Fig S2a: Rarefaction curves for diversity indices for individual leaves located at Upland Sites L6 and 14

Large Beech Large Maple Small Beech Small Maple

Supplementary Fig S2b: Rarefaction curves for diversity indices for individual leaves located at Riparian Sites J1 and Q1.

Large Beech Large Maple Small Beech Small Maple

Supplementary Fig S2c: Rarefaction curves for diversity indices for individual leaves located at Vernal Pool Sites J3 and H6.

Large Beech Large Maple Small Beech Small Maple

For all diversity indices (S, H’, 1/D, E1/D, and Chao), there were no significant main effects of habitat, leaf species, or leaf size. However, the interaction between leaf species and leaf size was significant or marginally significant for S (P = 0.04), H’ (P =

0.01), 1/D (P = 0.10), and E1/D (P = 0.08). Separate analyses of leaf species showed that the interaction between leaf species and size was due to a significant effect of leaf size in maple leaves (P = 0.053 for S and P < 0.05 for H’, 1/D, and E1/D), but no effect of leaf size in beech leaves (P > 0.1 for all indices, Figure 1).

Figure 1

Mean OTU richness (S), Shannon diversity (H’), and Simpson diversity (1/D) values for large vs. small sugar maple and beech leaves gathered from 3 temperate forest habitats

(n=2 leaves per bar). Analysis of sugar maple and beech leaves together resulted in a significant species X size interaction (see text). P values shown with each chart indicate significance of leaf size within either sugar maple or beech (size X habitat interactions were never significant).

Sugar Maple Sobs Beech Sobs p = 0.0534 p = 0.3112

40 40 2 Large (48.5 cm2) Large (47.4 cm ) 2 Small (14.6 cm2) Small (18.8 cm ) 30 30

S S 20 20

10 10

0 0 Vernal Pool Upland Riparian Vernal Pool Upland Riparian

Sugar Maple Shannon Beech Shannon p = 0.0455 p = 0.0962

4 4 2 Large (48.5 cm2) Large (47.4 cm ) 2 Small (14.6 cm2) Small (18.8 cm ) 3 3

H' 2 H' 2

1 1

0 0 Vernal Pool Upland Riparian Vernal Pool Upland Riparian

Sugar Maple Simpson Beech Simpson p = 0.0140 p = 0.7679

40 40 Large (48.5 cm2) Large (47.4 cm2) Small (14.6 cm2) Small (18.8 cm2) 30 30

20 20

1/D 1/D

10 10

0 0 Vernal Pool Upland Riparian Vernal Pool Upland Riparian

AICc was used as a model selection criterion to determine the influence of ln(leaf area), as well as habitat, on ln(S) in maple leaves. AICc was minimized by a model including ln(leaf area) and habitat as predictors, without interactions or random effects

(i.e., a model in which habitat alters the intercept of the relationship between ln(area) and ln(S), but not the slope; Fig. 2A). The z value of the taxa-area relationship for fungal taxa on sugar maple leaves was estimated to be 0.22 (± standard error of 0.07; Figure 2A).

Using the model selection procedure on beech leaves resulted in a model containing only a habitat effect, and no effect of leaf area (Figure 2B).

Figure 2

Fungal taxa-area relationship detected on maple leaves (P < 0.05) but not beech.

Regression lines are shown for maple upland, vernal pool, and riparian habitat TARs.

For maple leaves, model selection using AICc resulted in a model where habitat TARs have identical z-values (slope; 0.22 ± 0.07) and different c-values (Y-axis intercepts).

A. Maple leaves B. Beech leaves

3.6 3.6

3.2 3.2

)

ln( 2.8 ln( 2.8

2.4 2.4

2.0 2.0 1 2 3 4 5 1 2 3 4 5

ln(Area) Upland ln(Area) Riparian Vernal Pool

Beta diversity and community composition

A high degree of taxa turnover was indicated by an average Jaccard similarity index of 0.082. Redundancy analysis of Hellinger distances (relative abundance data) indicated that habitat explained 20.7% of the variability in OTU relative abundance (P =

0.001). Sites nested within habitat type explained an additional 20.6% of the variability in OTU relative abundance (P = 0.001). Leaf type explained 5% of variation in OTU abundance (P = 0.047), while leaf size was not significant. Redundancy analysis on

Sorenson distances (presence/absence data) indicated that habitat explained 15.5% of the variance (P = 0.001) and sites nested within habitat type explained 15.4% (P = 0.003), while leaf species and leaf size were not significant.

Habitat explained the majority of variability in five dominant OTUs (Table 1).

OTUs 3 (matching sequences in GenBank from the genus Phialea; Table 1), 5

(Polyscytalum), and 14 (Cystidendron) were dominant on multiple leaves at both upland field sites, while OTUs 10 (Sclerotinia) and 21 (Altospora) were dominant on leaves at both vernal pool sites. OTUs 3, 5, and 10 were also dominant on leaves at one of the two riparian sites. Hence, the upland forest and vernal pool fungal communities were never dominated by the same OTUs (Supplementary Figure S1). Riparian forest leaves at one site (site J2) shared dominant OTUs with both upland and vernal pool communities.

This is reflected in the RDA ordination in Fig. 3 by the intermediate position of this site on the second ordination axis. Leaves from the other riparian site (site Q1) were mostly dominated by OTU 1 (Mycena), resulting in a fungal community dramatically different

from all other sites. Hence, a major pattern in Figure 3 is separation of site Q1 from all other sites on the first ordination axis.

Table 1

Blast results and percentage of variance explained by habitat (P = 0.001) or site nested within habitat (P = 0.001) for dominant OTUs (>5% relative abundance on at least one leaf). NA= not assigned. Symbols indicate the habitat that dominant OTUs were associated with (∆ = upland, ∞ = vernal pool, ◊ = riparian).

% variance OTU Accession # Score E value Phylum Order Family Genus explained by Habitat ID # Habitat Site 1 FJ553305.1 239 8e-60 Basidiomycota Agaricales Tricholomataceae Mycena 25.93 43.08 ∆◊ 2 EU040223.1 507 2e-140 Ascomycota Xylariales NA Phlogicylindrium 5.29 23.00 ∆∞◊ 3 EF596821.1 571 7e-160 Ascomycota Helotiales Sclerotiniacae Phialea 39.02 9.26 ∆◊ 4 EF029239.1 569 4e-159 Ascomycota Helotiales Helotiaceae Helicodendron 17.36 39.54 ∞◊ 5 GQ303287.1 468 1e-128 Ascomycota NA NA Polyscytalum 34.39 9.64 ∆◊ 6 GQ303287.1 592 6e-166 Ascomycota NA NA Polyscytalum 3.60 15.40 ∆∞◊ 7 FJ554434.1 440 2e-120 Basidiomycota Agaricales Hydnangiaceae Laccaria 0.28 14.25 ∆◊ 8 AF346545.1 667 0.0 Ascomycota Xylariales Amphisphaeriaceae Amphisphaeria 8.70 13.04 ∞ 9 DQ273371.1 431 1e-117 Basidiomycota Cantharellales Cantharellaceae Cantherellus 15.49 23.24 ◊ 10 AM901701.1 381 1e-102 Ascomycota Helotiales Sclerotiniacae Sclerotinia 37.29 7.27 ∆ ◊ 11 AF141164.1 420 3e-114 Ascomycota Helotiales Dermateacea Dermea 8.70 13.04 ∆ 12 EU726288.1 459 7e-126 Ascomycota Helotiales NA NA 4.81 13.60 ∆◊ 13 AF444526.1 475 6e-131 Basidiomycota Filobasidiales Filobasidiaceae Rhodotorula 26.98 40.47 ∆ 14 DQ914671.1 407 3e-110 Ascomycota Helotiales NA Cystodendron 18.17 0.02 ∆ 15 AF141189.1 466 4e-128 Ascomycota Helotiales Dermateacea Neofrabraea 17.97 26.95 ◊ 16 FJ553668.1 414 1e-112 Ascomycota Helotiales Dermateacea Pezicula 8.70 13.04 ∞ 17 EU998923.1 590 2e-165 Ascomycota Helotiales Hypogastruroidea Articulospora 17.30 25.95 ∆ 18 EF596821.1 472 8e-130 Ascomycota NA NA Phialea 17.73 26.60 ∞ 19 EU935520.1 542 7e-151 Basidiomycota Agaricales Marasmiacea Marasmius 8.70 13.04 ◊ 20 EU040235.1 424 2e-115 Ascomycota NA NA Parapleurotheciopsis 8.70 13.04 ∞ 21 AY204588.1 435 1e-118 Ascomycota Acarosporales Altosporidea Altospora 18.17 0.02 ∞ 22 FJ839639.1 385 1e-103 Ascomycota Mycosphaerellales Dothidiomycetes Xenostigmina 18.15 27.23 ∞ 23 FJ839639.1 412 4e-112 Ascomycota Mycosphaerellales Dothidiomycetes Xenostigmina 8.70 13.04 ∞ 24 EU552142.1 329 5e-87 Ascomycota Pleosporales Lophiostomataceae Massarina 18.10 27.15 ∞ 25 AF141168.1 507 2e-140 Ascomycota Helotiales Dermateacea Scleropezicula 8.70 13.04 ∆ 26 FJ554419.1 357 2e-95 Ascomycota Helotiales Dermateacea NA 8.70 13.04 ∞

27 FJ554126.1 409 7e-111 Ascomycota Pezizales Lecanoromycetidae NA 8.70 13.04 ◊ 28 FJ553703.1 435 1e-118 Ascomycota Eruotiales Trichocomaceae Pennicillum 8.70 13.04 ∆ 29 AJ301937.1 671 0.0 Ascomycota Phyllachorales Phyllachoracea Colletotrichum 8.70 13.04 ∞ 30 FJ554126.1 501 5e-138 Ascomycota Pezizales Lecanoromycetidae NA 8.70 13.04 ∞ 31 FJ553305.1 398 2e-107 Basidiomycota Agaricales Tricholomataceae Mycena 8.70 13.04 ◊ 34 EF029241.1 326 6e-86 Ascomycota Chaetosphaeriales Chaetosphaeriacea Dictyochaeta 8.70 13.04 ∞ 36 FJ553648.1 610 2e-171 Basidiomycota Agaricales Tricholomataceae Mycena 8.70 13.04 ◊ 37 AM901701.1 381 1e-102 Ascomycota Helotiales Helotiaceae Sclerotinia 8.70 13.04 ∞ 38 AB126047.1 222 9e-55 Basidiomycota Ustilaginales Ustilaginaceae Sporobolomyces 8.70 13.04 ∆ 40 GQ509692.1 425 8e-116 Non-cult. clone NA NA NA 8.70 13.04 ∆ 41 FJ554323.1 436 4e-119 Basidiomycota Agaricales Hydnangiaceae Laccaria 8.70 13.04 ∞ 44 DQ646542.1 230 7e-57 Ascomycota Pezizales Pezizacea Peziza 8.70 13.04 ∞ 47 EU040223.1 464 1e-127 Ascomycota Xylariales NA Phlogicylindrium 8.70 13.04 ∞ 51 FJ185160.1 778 0.0 Basidiomycota Agaricales Agaricacea Coprinellus 8.70 13.04 ∞ 54 AJ301970.1 584 9e-164 Ascomycota Phyllachorales Phyllachoracea Glomerella 8.70 13.04 ∆ 57 FJ554064.1 339 1e-89 Basidiomycota Polyporales Atheliaceae NA 8.70 13.04 ∆ 64 AY706329.1 494 2e-136 Ascomycota Pezizales Leotiomycetes Leohumicola 8.70 13.04 ∞ 70 AF096215.1 213 5e-52 Ascomycota Pezizales Lecaneromycetes Unbilicaria 8.70 13.04 ∆

Figure 3:

Redundancy analysis ordination of fungal community composition at the six field sites.

Habitat is indicated by symbol shape, with site names shown in legend. Because leaf type explained little variation in community composition (5%), each symbol represents the centroid of the four leaves sampled at each site.

0.6 Upland 14 Upland L6 0.4 Rip J1 Rip Q1

0.2 VP H6 VP J3 0

-0.2 Axis 2: 10.3% 2: Axis -0.4

-0.6 -0.8 -0.4 0 0.4 0.8 1.2 1.6 Axis 1: 15.3%

Species rank abundance distribution models

In a comparison of different rank abundance distribution models, AIC indicated that the zero-sum model was the closest fit with our data for all 24 leaves (Supplementary

Table S1 and Figure S3), consistent with neutral theory dynamics. The minimum difference in AIC between the zero-sum model and any other model was 32

(Supplementary Table S1), indicating no support for any model other than the zero-sum model (Burnham and Anderson, 2002). This trend was consistent across all habitats, leaf types, and leaf sizes.

Supplemental Table S1

AIC values for 6 rank abundance distribution models. Lowest AIC value for each sample represents the best fit model.

1 AIC Leaf Species Size Habitat Site broken-stick pre-Emption log-Normal Zipf Zipf-Mandelbrot ZSM 102 Beech Large Riparian J1 257.67 249.04 157.53 119.12 121.12 54.20 240 Beech Small Riparian J1 203.15 211.49 150.61 118.06 120.06 53.17 105 Maple Large Riparian J1 106.29 100.37 94.96 92.89 91.06 45.23 57 Maple Small Riparian J1 120.94 100.98 85.27 80.42 80.91 46.29 18 Beech Large Riparian Q1 363.43 227.47 128.49 93.44 95.44 46.82 216 Beech Small Riparian Q1 226.64 152.34 92.46 69.16 71.16 35.84 133 Maple Large Riparian Q1 81.07 78.83 73.20 68.96 70.10 36.08 230 Maple Small Riparian Q1 311.40 218.06 125.14 92.20 94.20 43.29 8 Beech Large Upland 14 184.87 179.71 127.58 100.34 102.34 60.71 245 Beech Small Upland 14 147.66 148.88 135.09 122.30 124.13 42.23 33 Maple Large Upland 14 98.89 99.65 91.87 83.10 85.10 28.85 250 Maple Small Upland 14 202.30 189.33 137.70 112.91 114.91 60.09 36 Beech Large Upland L6 192.43 178.17 154.29 138.17 133.30 61.27 45 Beech Small Upland L6 98.06 99.01 92.27 84.55 86.44 32.59 177 Maple Large Upland L6 133.93 139.27 107.12 88.03 90.03 37.89 223 Maple Small Upland L6 141.29 122.22 101.33 87.11 87.45 51.51 76 Beech Large Vernal H6 145.64 139.15 133.88 125.35 126.75 35.19 53 Beech Small Vernal H6 125.92 126.68 116.59 106.78 108.75 39.13 69 Maple Large Vernal H6 117.70 118.65 108.93 99.62 101.57 36.85 268 Maple Small Vernal H6 179.47 149.30 103.07 82.80 84.80 50.48 38 Beech Large Vernal J3 62.25 61.01 61.27 58.39 58.87 36.22

95 Beech Small Vernal J3 156.79 158.61 125.94 104.71 106.71 52.92 127 Maple Large Vernal J3 110.90 109.58 105.82 101.04 100.69 35.78 258 Maple Small Vernal J3 171.47 181.94 140.60 114.93 116.93 51.94

1 Statistical details: AIC for radfit-generated models was calculated with the standard equation AIC = -2*log-likelihood +

2*npar, where npar represents the number of parame ters in the fitted model. AIC was calculated in the same way for the zero-sum model, except that the likelihood directly reported by TeTame is the minimum of - log-likelihood, and so must be first multiplied by -1 to obtain the maximum lo g-likelihood value (F. Jabot, personal communication). In addition, comparison of likelihoods, or AIC, calculated by radfit and TeTame is biased against the zero-sum model (F. Jabot, personal communication). This is because the TeTame likelihood values for the zero-sum model are conditioned only on the model parame ters and J (the total number of individuals in the sample), whereas the radfit likelihood values are additionally conditioned on S (the number of taxa present in the sample). Maximum likelihoods for the zero-sum model with the additional conditio ning on S would be equal to or greater than those reported by TeTame (F. Jabot, personal communication).

Hence, AIC for the zero-sum model would be smaller than the value calculated from TeTame output, if likelihood was calculated in the same way as AIC calculated from radfit.

Discussion

Alpha diversity and taxa-area relationship

We have shown that the diversity of saprotrophic fungal communities living on sugar maple leaves fits the classic power law TAR. However, this pattern was not observed on beech leaves, in which z was not significantly different from zero, although it is possible that further sampling of beech leaves may reveal a shallower or more noisy

TAR. There are several mechanisms proposed in the literature to account for taxa-area relationships. The “passive sampling hypothesis” (Connor and McCoy 1979) proposes that larger areas support a greater variety of organisms than smaller areas because they receive more colonists than smaller areas. The “habitat diversity hypothesis” (Williams

1964) proposes instead that larger areas offer a greater variety of habitats, while the “area per se hypothesis” (Simberloff 1976) states that larger areas support higher populations of each species, decreasing the probability of any particular species going extinct. The above hypotheses do not seem adequate to explain the different TAR patterns we found in maple and beech. In contrast, Wright’s species-energy theory (Wright 1983) can be directly related to known differences between maple and beech leaves. Wright proposed that area is actually an indicator for available energy, or another resource limiting energy use, available in a habitat patch. According to Wright (1983), the proximate causes of increased species richness in larger areas are 1) the larger population sizes supported by increased resource production rates, and 2) a correlation between resource quantity and resource heterogeneity that can support a greater diversity of species niches. In this case,

one would expect stronger TARs for habitats in which a change in area causes a greater change in energy supply than in other habitats.

Energy is obtained by saprotrophic fungi from biochemicals present in plant tissue (Berg and McClaugherty 2003). In Jennings Woods, where our leaves were sampled, sugar maple leaves contain approximately 16% acid-insoluble material, whereas beech leaves contain approximately 29% (unpublished data). Microbial growth on lignin, the primary component of acid-insoluble plant compounds, is inefficient, requiring secretion of costly oxidative extracellular enzymes that release aromatic monomers that are not readily metabolized (Berg and McClaughtery 2003). Lignin also forms a protective coating over labile plant cell wall polysaccharides, further reducing the rate of resource acquisition possible by microbial saprotrophs. Hence, substantially less energy per unit area is made available for saprotrophs growing on beech leaves in a given period of time. Variation in leaf size therefore translates into less variation in available energy on beech leaves than on maple leaves, and this smaller variation in energy availability is hypothesized to result in a shallower TAR, making detection of a significant relationship more difficult to achieve statistically.

The TAR rate of increase we found (z = 0.22) is very close to the canonical value of 0.26 (Preston 1962, Drakare et al. 2006) and similar to that found by ITS sequencing for ectomycorrhizal fungi associated with tree “islands” (0.2 – 0.23; Peay et al. 2007).

Green and Bohannan (2006) report z values for microorganisms ranging from 0.02 to

0.47, but these values are difficult to compare due to inclusion of a wide variety of habitats, taxa, and sampling formats (contiguous, noncontiguous, and island).

Furthermore, there is debate over whether different molecular methods result in comparable measures of community diversity (Woodcock et al. 2006, Blackwood et al.

2007b). Horner-Devine et al. (2004) used 16S ribosomal DNA sequencing (which is similar to the method we employed here) to characterize bacterial diversity in marsh sediments and found a much lower z value (0.02-0.04) than found for fungi.

Microorganism distribution is thought to be less limited by dispersal than macroorganism distribution, leading to fewer local endemics and a lower z value (Drakare et al. 2006,

Finlay et al. 1998, Hillebrand et al. 2001). Fungal z-values being closer to those of plants than bacteria may suggest that dispersal patterns and mechanisms of community assembly are more similar between plants and fungi.

Support for neutral dynamics from rank abundance

The zero-sum model best fit our rank abundance distributions for fungal communities in individual leaves. Neutral theory has been supported using the zero-sum model in a number of empirical community assemblages including arbuscular mycorrhizal fungi (Dumbrell et al. 2010), marine intertidal invertebrates (Wootton et al.

2005), South American and European birds (Ricklefs 2006), subtropical/temperate plants

(Forster and Wharton 2007), and pacific corals (Connolly et al. 2009). In actuality, metacommunity species distribution patterns may be a result of a complicated mix of more than one factor (Leibold et al. 2010). Our analyses of beta diversity found a relatively minor effect of the potential environmental filter leaf type (5% of variance explained), however leaf type does explain a somewhat larger portion of variance for some taxa (OTU 5, 16.3%; OTU 7, 14.0%) indicating niche-based processes may be

involved in some species distributions. Habitat and site (nested within habitat) explained much more community composition variance (20.7% and 20.6%, respectively), but these habitats differ dramatically in environmental characteristics, and we do not have enough sampling sites to attribute the habitat and site effects strictly to dispersal limitation or environmental filtering.

Although neutral theory predicts that all species are equally competitive and differences in community composition at a given location are primarily due to colonization order and priority effects (Hubbell 2001), competition has been shown to alter species abundances in some cases (Connell 1983, Schoener 1983). Competition for resources has been described as a primary mechanism driving fungal establishment and succession (Boddy and Rayner 1983, Peiris et al. 2008). In our study, many of the OTUs that were dominant on one leaf were also found in lower abundance on other leaves. For example, OTU 1 was found with relative abundance values of 0.7 – 3.9% at all sites except at site Q1 where its mean relative abundance across four leaves was 81.7%.

Neutral theory would predict that dominance of OTU 1 is due to pre-emptive colonization. However, the possibility of competitive displacement of other fungi by

OTU 1 could be empirically tested by evaluating its competitive ability against OTUs that were dominant at the five sites where OTU 1 was found, but not dominant.

We determined fungal relative abundances in this study using clone library sequences, a process that is potentially affected by biases in DNA extraction and PCR amplification (Wintzingerode et al. 1997). Although quantification of taxa richness from clone libraries has generally been found to be robust (Avis et al. 2010, Tedersoo et al.

2010), it is possible that the procedural biases favored rank abundance distributions that are best fit by the zero-sum model. However, one general form of PCR bias is for a flattened rank abundance distribution (Kanagawa 2003), which would bias against the zero-sum model, rather than for it. In contrast, preferential amplification of particular taxa is also a common PCR artifact, and this could favor the zero-sum model. In our data, fungal taxa common to multiple samples were not consistently dominant or rare, indicating that preferential amplification of certain taxa was not particularly strong. Our primers amplified the ITS1 region, which is less variable in length than the ITS2 or full

ITS regions, reducing preferential amplification and chimera formation (Bellemain et al.

2010, Tedersoo et al. 2010). We also took other steps to reduce preferential amplification through pooling multiple PCR replicates, using as few PCR cycles as possible, and diluting samples to reduce the concentration of environmental PCR inhibitors (Kanagawa

2003, Tedersoo et al. 2010). Finally, several studies indicate DNA extraction bias is not a particular problem for fungi in environmental samples (Feinstein et al. 2009, Oberkofler and Peintner 2008).

Conclusions

Taking advantage of leaves as naturally occurring resource patches, we found a significant taxa-area relationship on sugar maple leaves, but not beech leaves. We hypothesize that Wright’s species-energy relationship may explain this difference in TAR patterns between leaf types. This suggests that energy availability as affected by plant biochemistry may provide a heretofore-undescribed mechanism that influences fungal diversity over a variety of spatial scales. Fungal beta diversity was high, indicating a

large degree of turnover between habitat patches. Individual leaf rank abundance distributions are consistent with zero-sum dynamics, providing support for a neutral theory of metacommunity organization for forest floor saprotrophic fungi. Further research is necessary to elucidate mechanisms driving the patterns we have described, including further investigation of the link between fungal community resource pool quality, habitat size, and taxa richness.

Acknowledgements

This work was supported by a Kent State University Graduate Student Senate research grant and a KSU Biological Sciences Department Arthur & Margaret Herrick grant. We thank the Blackwood lab members and Donald Zak for comments on our data, as well as Kurt Smemo for lyophilizing our samples and Alex Gradisher for photographic and leaf surface area advice.

References

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. 1997.

Gapped BLASTand PSI_blast: a new generation of protein database search

programs. Nucleic Acids Research 25: 3389-3402.

Arrhenius O. 1921. Species and area. Journal of Ecology 9: 95–99.

Avis PG, Branco S, Tang Y, Mueller GM. 2010. Pooled samples bias fungal community

descriptions. Molecular Ecology Resources 10: 135-141.

Balser TC, Firestone MK. 2005. Linking microbial community composition and soil

processes in a California annual grassland and a mixed-conifer forest.

Biogeochemistry 73: 395-415.

Bell T, Ager D, Song J, Newman JA, Thompson IP, Lilley AK, van der Gast C. 2005.

Larger islands house more bacterial taxa. Science 308: 1884.

Bellemain E, Carlsen T, Brochmann C, Coissac E, Taberlet P, Kauserud, H. 2010. ITS as

an environmental DNA barcode for fungi: an in silico approach reveals potential

PCR biases. BMC Microbiology 10: 189.

Berg B, McClaugherty C. 2003. Plant litter: Decomposition, humus formation, carbon

sequestration. New York: Springer-Verlag; 279 p.

Blackwood CB, Waldrop MP, Zak DR, Sinsabaugh RL. 2007a. Molecular analysis of

fungal communities and laccase genes in decomposing litter reveal differences

among forest types but no impact of nitrogen deposition. Environmental

Microbiology 9: 1306-1316.

Blackwood CB, Hudelston D, Zak DR, Buyer JB. 2007b. Interpreting Ecological

Diversity Indices Applied To T-RFLP Data: Insights From Simulated Microbial

Communities. Applied and Environmental Microbiology 73: 5276-5283.

Boddy L, Rayner ADM. 1983. Ecological roles of basidiomycetes forming decay

communities in attached oak branches. New Phytologist 93: 77–88.

Burnham KP, Anderson DR. 2002. Model Selection and Multimodel Inference: A

Practical Information-Theoretic Approach, 2nd ed. New York: Springer-Verlag;

496 p.

Chase JM, Amarasekare P, Cottonie K, Gonzalez A, Holt RD, Holyoak M, Hoopes ME,

Leibold MA, Loreau M, Mouquet N, Shurin JB, Tilman D. 2004. Competing

theories for competitive metacommunities in Metacommunities: Spatial

Dynamics and Ecological Communities. Chocacgo, USA: The University of

Chicago Press; 520 p.

Chong GW, Stohlgren TJ. 2007. Species-area curves indicate the importance of habitats’

contribution to regional biodiversity. Ecological Indicators 7: 387-395.

Colwell RK. 2009. EstimateS 8.2: statistical estimation of species richness and shared

species from samples. Guide and application published at

http://purl.oclc.org/estimates.

Colwell RK, Coddington JA. 1994. Estimating terrestrial biodiversity through

extrapolation. Philosophical Transactions of the Royal Society Series B:

Biological Sciences 345: 101–118.

Colwell RK, Mao CX, Chang J. 2004. Interpolating, extrapolating, and comparing

incidence-based species accumulation curves. Ecology 85: 2717-2727.

Connell JH. 1983. On the prevalence and relative importance of interspecific

competition: evidence from field experiments. American Naturalist 122: 661-696.

Connolly SR, Dornelas M, Bellwood DR, Hughes, TP. 2009. Testing species abundance

models: a new bootstrap approach applied to Indo-Pacific coral reefs. Ecology

90(11): 3138-3149.

Connor EF, McCoy ED. 1979. The statistics and biology of the species-area relationship.

American Naturalist 113(6): 791-833.

Connor EF, McCoy ED. 2001. Species–area relationships. Encyclopedia of biodiversity,

Vol. 5 (ed. by S.A. Levin), pp. 397–411. Academic Press, San Diego. 4666 p.

Cook RC, Rayner ADM. 1984. Ecology of saprotrophic fungi. Longman Group Ltd.,

London. 438 p.

Drakare S, Lennon JJ, Hillebrand H. 2006. The imprint of the geographical, evolutionary

and ecological context on species-area relationships. Ecology Letters 9: 215–227.

Dumbrell AJ, Nelson M, Helgason T, Dytham C, Fitter AH. 2010. Relative roles of niche

and neutral processes in structuring a soil microbial community. ISME Journal 4:

337-345.

Faith DP, Ferrier S, Williams KJ. 2008. Getting biodiversity intactness indices right:

ensuring that ‘biodiversity’ reflects ‘diversity’. Global Change Biology 14: 207-

217.

Feinstein LM, Sul WJ, Blackwood CB. 2009. Assessment of bias associated with

incomplete extraction of microbial DNA from soil. Applied and Environmental

Microbiology 75: 5428-5433.

Finlay BJ, Esteban GF, Fenchel T. 1998. Protozoan diversity: converging estimates of the

global number of free living ciliate species. Protist 149: 29–37.

Fischer H, Bergfur J, Goedkoop W, Tranvik L. 2009. Microbial leaf degraders in boreal

streams: bringing together stochastic and deterministic regulators of community

composition. Freshwater Biology 54: 2276-2289.

Forster MA, Warton DI. 2007. A metacommunity-scale comparison of species-

abundance distribution models for plant communities of eastern Australia.

Ecography 30: 449-458.

Gaston K, Blackburn T. 2000. Pattern and process in macroecology. Blackwell Sciences,

London. 377 p.

Gessner MO, Swan CM, Dang CK, McKie BG, Bardgett RD, Wall DH. Hӓttenschwiler

S. 2010. Diversity meets decomposition. Trends in Ecology and Evolution 25(6):

372-380.

Gleason HA. 1922. On the relationship between species and area. Ecology 3: 158–162.

Gotelli N, Colwell RK. 2001. Quantifying biodiversity: Procedures and pitfalls in the

measurement and comparison of species richness. Ecology Letters 4: 379-391.

Green J, Bohannan BJM. 2006. Spatial scaling of microbial biodiversity. Trends in

Ecology and Evolution 21: 501-507.

Harte J, Kinzig A, Green J. 1999. Self-similarity in the distribution and abundance of

species. Science 284: 334-336.

He F, Legendre P. 1996. On species–area relations. American Naturalist 148: 719–737.

Hillebrand H, Watermann F, Karez R, Berninger UG. 2001. Differences in species

richness patterns between unicellular and multicellular organisms. Oecologia 126:

114–124.

Hobbie SE, Reich PB, Oleksyn J, Ogdahl M, Zytkowiak R, Hale C, Karolewski P. 2006.

Tree species effects on decomposition and forest floor dynamics in a common

garden. Ecology 87(9): 2288-2297.

Horner-Devine MC, Lage M, Hughes JB, Bohannan BJM. 2004. A taxa–area relationship

for bacteria. Nature 432: 750–753.

Hubbell SP. 2001. The Unified Neutral Theory of Biodiversity and Biogeography.

Princeton

University Press, Princeton. 375 p.

Jabot F, Etienne RF, Chave J. 2008. Reconciling neutral community models and

environmental filtering: theory and an empirical test. Oikos 117: 1308-1320.

Jacques MA, Kinkel LL, Morris CE. 1995. Population sizes, immigration, and growth of

epiphytic bacteria on leaves of different ages and positions of field-grown endive.

Applied and Environmental Microbiology 61(3): 899-906.

Jongman RHG, Braak CJF, Tongeren OFR. 1995. Data analysis in community and

landscape ecology. 2nd edition. Cambridge University Press, Cambridge, Mass,

USA. 324 p.

Kanagawa T. 2003. Bias and artifacts in multitemplate polymerase chain reactions

(PCR). Journal of Bioscience and Bioengineering 4: 317-323.

Keeler BL, Hobbie SE, Kellogg LE. 2009. Effects of long-term nitrogen addition on

microbial enzyme activity in eight forested and grassland sites: Implications for

litter and organic matter decomposition. Ecosystems 12(1): 1-15.

Kinkell LL, Andrews JH, Berbee FM. 1987. Leaves as islands for microbes. Oecologia

71(3): 405-408.

Legendre P, Gallagher ED. 2001. Ecologically meaningful transformations for ordination

of species data. Oecologia 129: 271-280.

Leibold MA, Economo EP, Peres-Neto P. 2010. Metacommunity phylogenetics:

separating the role of environmental filters and historical biogeography. Ecology

Letters 13(10): 1290-1299.

Levin SA. 2000. Multiple scales and the maintenance of biodiversity. Ecosystems 3:498-

506.

Lomolino MV. 2001. The species–area relationship: new challenges for an old pattern.

Progress in Physical Geography 25: 1–21.

MacArthur R. 1957. On the relative abundance of bird species. Proceedings of the

National Academy of Science 43: 293–295.

MacArthur RH, Wilson EO. 1967. The Theory of Island Biogeography. Princeton

University Press, Princeton. 224 p.

Mao CX, Colwell RK, Chang J. 2005. Estimating species accumulation curves using

mixtures. Biometrics 61: 433–441.

Martin KJ, Rygiewicz PT. 2005. Fungal-specific PCR primers developed for analysis of

the ITS region of environmental DNA extracts. BMC Microbiology 5: 28.

Maurer BA, McGill BJ. 2004. Neutral and non-neutral macroecology. Basic and Applied

Ecology 5: 413-422.

May RM. 1975. Patterns of species abundance and diversity. In Ecology and evolution of

communities (ed. Cody ML, Diamond JM). Cambridge, MA: Harvard University

Press. 560 p.

McGill BJ, Rampal S, Etienne JS, Gray DA, Anderson MJ, Benecha HK, Dornelas M,

Enquist BJ, Green JL, He F, Hurlbert AH, Magurran AE, Marquet PA, Maurer

BA, Ostling A, Soykan CU, Ugland KI, White EP. 2007. Species abundance

distributions: moving beyond single prediction theories to integration within an

ecological framework. Ecology Letters 10: 995-1015.

Motomura I. 1932. On the statistical treatment of communities. Zoological Magazine 44:

379–383.

Neubert K, Mendgen K, Brinkmann H, Wirsel SGR. 2006. Only a few fungal species

dominate highly diverse mycofloras associated with the common reed. Applied

and Environmental Microbiology 72: 1118-11128.

Newell SY, Fell JW. 1997. Competition among mangrove oomycetes, and between

oomycetes and other microbes. Aquatic Microbial Ecology 12(1): 21-28.

Nikolcheva LG, Bourque T, Bӓrlocher F. 2005. Fungal diversity during initial stages of

leaf decomposition in a stream. Mycological Research 109(2): 246-253.

Oberkofler I, Peintner U. 2008. Detection of soil fungal communities in an alpine primary

successtional habitat: does pooling of DNA extracts affect investigations? Annals

of Microbiology 58: 585-595.

O’Brien HE, Parrent JL, Jackson JA, Moncalvo J-M, Vilgalys R. 2005. Fungal

community analysis by large-scale sequencing of environmental samples. Applied

and Environmental Microbiology 71: 5544-5550.

Oksanen P. 2010. Vegan 1.17-0 in R version 2.10.1 2009-12-14.

Peay KG, Bruns TD, Kennedy PG, Bergeman SE, Garbelotto M. 2007. A strong species–

area relationship for eukaryotic soil microbes: island size matters for

ectomycorrhizal fungi. Ecology Letters 10: 470-480.

Peay KG, Kennedy PG, Bruns TD. 2008. Fungal community ecology: A hybrid beast

with a molecular master. Biosciences 58(9): 799-810.

Peiris D, Warwick BD, Brown M, Douglas BK, Ipsita R, Hedger JN. 2008. Metabolite

profiles of interacting mycelial fronts differ for pairings of the wood decay

basidiomycete fungus, Stereum hirsutum with its competitors Coprinus micaceus

and Coprinus disseminates. Metabolomics 4: 52-62.

Plotkin JB, Potts MD, Yu DW, Bunyavejchewin S, Condit R, Foster R. Hubbell S,

LaFrankie J, Manokaran N, Seng LH, Sukumar R, Nowak MA, Ashton PS. 2000.

Predicting species diversity in tropical forests. Proceedings of the National

Academy of Sciences USA 97: 10850–10854.

Preston FW. 1962. The canonical distribution of commonness and rarity. Ecology 43:

185-215.

Pueyo S. 2006. Self-similarity in species–area relationship and in species abundance

distribution. Oikos 112: 156–162.

Redford AJ, Bowers RM, Knight R, Linhart Y, Fierer N. 2010. The ecology of the

phyllosphere: geographic and phylogenetic variability in the distribution of

bacteria on tree leaves. Environmental Microbiology 12: 2885-2893.

Ricklefs R. 2006. The Unified Neutral Theory of Biodiversity: Do the Numbers Add Up?

Ecology 87(6): 1424-1431.

Rosenzweig ML. 1995. Species diversity in space and time. Cambridge Univ Press,

Cambridge, UK.

Schoener TW. 1976. The species–area relation within archipelagos: models and evidence

from island land birds. Proc Int Ornith Cong 16: 628–642.

Simberloff D. 1976. Experimental zoogeography of islands – effects of island size.

Ecology 57: 629–648.

Smith B, Wilson JB. 1996. A consumer’s guide to evenness indices. Oikos 76: 70-82.

Tedersoo L, Nilsson RH, Abarenkov K, Jairus T, Sadam A, Saar I, Bahram M, Bechem

E, Chuyong G, Urmas K. 2010. 454 Pyrosequencing and Sanger sequencing of

tropical mycorrhizal fungi provide similar results but reveal substantial

methodological biases. New Phytologist 188: 291-301.

Waldrop MP, Firestone MK. 2006. Response of Microbial Community Composition and

Function to Soil Climate Change. Microbial Ecology 52: 716-724.

Weand MP, Arthur MA, Lovett GM, McCulley RL, Weathers KC. 2010. Effects of tree

species and N additions on forest floor microbial communities and extracellular

enzyme activities. Soil Biology and Biochemistry 42: 2161-2173.

Williams CB. 1964. Patterns in the balance of nature. Academic Press, London.

Wintzingerode FV, Göbel UB, Stackebrandt E. 1997. Determination of microbial

diversity in environmental samples: pitfalls of PCR-based rRNA analysis. FEMS

Microbiology Reviews 21: 213-229.

Woodcock S, Curtis TP, Head IM, Lunn M, Sloan WT. 2006. Taxa-area relationships for

microbes: the unsampled and the unseen. Ecology Letters 9: 805-812.

Wootton JT. 2005. Field parameterization and experimental test of the neutral theory of

biodiversity. Nature 433: 309-312.

100

Wright DH. 1983. Species-Energy Theory: An Extension of Species-Area Theory. Oikos

41(3): 496-506

Wu L, Feinstein L, Valverde-Barrantes O, Kershner M, Leff L, Blackwood C.B. 2011.

Placing the Effects of Leaf Litter Diversity on Saprotrophic Microorganisms in

the Context of Leaf Type and Habitat. Microbial Ecology 61: 399-409

Chapter Four

The Spatial Scaling of Saprotrophic Fungal Diversity

Abstract

Community composition profiles for 180 saprotrophic fungal communities were generated in order to determine the relative importance of species-sorting and neutral processes in influencing fungal beta diversity in a Northeastern Ohio deciduous forest.

Thirty senesced leaves were collected from replicate upland, riparian, and vernal pool habitats in June 2008 in order to quantify the influence of landscape-scale environmental heterogeneity on community composition. We quantified spatial distance between communities and found significant distance-decay relationships at all but one upland site.

This is the first study where changes in fungal community composition were quantified across discrete adjacent habitat patches; providing evidence that fungal distance decay is operational at a scale of centimeters as opposed to distances of up to thousands of kilometers published in other studies. The influence of small-scale heterogeneity within a local cluster of communities was apparent as depth within leaf pack community the community was located in had an influence on community composition. Environmental heterogeneity associated with depth could include moisture gradients, relative influence of soil or spore colonization, and impact of forest floor biotic community (i.e. collembola, earthworms). Significant influence of spatial distance as well as depth indicates that both species-sorting and neutral processes are embedded within the distance-decay relationships that we found. This may at least be partially explained with 101

102

neutral assembly of communities consisting of differential pools of taxa associated with various depths that are filtered by species-sorting environmental heterogeneity.

Introduction

A long-standing focus of community ecology has been to quantify and understand patterns of species diversity (Whittaker 1972, May 1976). Because community composition of macroorganisms is easier to characterize than for microorganisms, most work examining species diversity has been conducted on macroorganisms (Green and

Bohannan 2006, Nemergut et al. 2011). However, saprotrophic fungi play a critical role in global carbon and nutrient cycling (Berg and Lazkowski 2006, Peay et al. 2008), and fungal diversity has been shown to impact ecosystem function (Tiunov and Scheu 2005,

Fukami et al. 2010). In order to understand the link between fungal diversity and ecosystem function under natural conditions, we must develop an understanding of how fungal diversity changes with spatial scale (Zak and Visser 1996, Fierer et al. 2007). This would also make an important contribution to the investigation of whether diversity patterns are similar for microorganisms and macroorganisms (Green and Bohannan

2006). From the perspective of saprotrophic fungi, each senesced tree leaf is a habitat patch with a distinct boundary and resource pool (i.e., nutrients and carbon compounds), and these resources vary by tree species (Cornwell et al. 2008). Hence, leaves on the forest floor form natural mixtures of overlapping habitat patches that can be treated as a spatial network. Spatial analyses can be used to examine distance-decay relationships, which define how diversity changes with spatial scale (Nekola and White 1999), and test

103

alternative theories about what processes drive beta diversity (Holt 2002, Brose et al.

2004, Filotas et al. 2010, Jabot and Bascompte 2012).

There are two primary contrasting theories about controls over community beta diversity (the turnover of taxa among communities) (Leibold et al. 2004). Under species- sorting, community composition is a response to a variety of environmental filters that exclude community members that cannot live within the niche space provided by the habitat patch (Whittaker 1972). Differences in community structure driven by species- sorting are correlated with gradients in the environment. In contrast, neutral theory states that organisms are not sorted by niches because they all have similar competitive capabilities (Hubbell 2001). Under neutral theory, community composition is related to dispersal limitation, as colonization of a habitat patch is much more likely from taxa that are nearby compared to taxa that are not. Thus, community composition would be primarily structured by spatial distance regardless of environmental conditions (Hubbell

2001).

Evidence has been found showing that fungal communities can be influenced by species-sorting (Helgason et al. 2002, Öpik et al. 2009), neutral dynamics (Peay et al.

2010), or a combination of both processes (Lekberg et al. 2007, Dumbrell et al. 2010,

Kivlin et al. 2011). These findings indicate that communities may be influenced by different ecological processes, depending on the habitat, community, and spatial scale.

However, it has previously been pointed out that environmental variability and spatial distance can be difficult to disentangle (Gilbert and Lechowicz 2004), and most

104

communities are probably influenced to some degree by both species sorting and neutral processes (Gravel et al. 2006).

Our objective was to quantify fungal β-diversity at varying spatial scales in order to determine whether β-diversity patterns were consistent with patterns predicted by neutral dynamics, species-sorting, or both. Leaves known to differ in biochemical composition were harvested from multiple sites in each of three forest habitats (upland forest, riparian forest, and vernal pool) that were selected to provide environmental variability. We collected leaves at each site in a manner that allowed us to quantify leaf network topology, which was then used to construct spatial distance matrices. Because of the commonness of the distance-decay pattern even for microorganisms (Green and

Bohannon 2006, Martiny et al. 2006, Astorga et al. 2012), we hypothesized that we would find similar distance-decay relationships even within small leaf neighborhoods where dispersal may not be expected to be limited. We also hypothesized that β-diversity would be highest within riparian forest because it experiences relatively high seasonal moisture fluctuations (i.e. flooding from the river, runoff from upland) and is a transition zone that may be colonized from both drier upland and saturated vernal pools. We used collectors curve analysis to determine whether community composition stabilizes after sampling a certain number of habitat patches (Gotelli and Colwell 2001), and whether the stabilization is dependent on site β-diversity. Finally, we tested the relative influence of species-sorting and neutral processes within each site and across all sites. We predicted that community composition would be significantly influenced by both species-sorting and neutral processes. Leaf type and habitat type would serve as species-sorting filters,

105

with evidence of dispersal limitation indicated by additional variation in community composition explained by spatial distance among communities.

Methods

Leaf collection

Leaves were harvested in June 2008 from duplicate 850 cm2 plots located within three NE Ohio deciduous forest habitats (upland, riparian, and vernal pool). The forest habitats were selected to provide abiotic and biotic variability (e.g., variation in moisture saturation). Thirty leaves were collected at each site. In order to have consistent variation in leaf resource pools across all sites, we collected at least 10 American beech (Fagus grandifolia) and 10 sugar maple (Acer saccharum) leaves at each site, if possible. The remaining leaves collected were comprised of whatever leaves were embedded in the habitat patch network at that site. To document the in situ layout of leaves collected, a tripod-mounted camera was used to record images of the plot before and after each leaf was removed. Individual leaves were collected with sterile implements, photographed, placed in pre-weighed containers, and stored in the field on dry ice. Leaves were lyophilized at -60°C (VirTis Genesis 25EL, Biopharma Process Systems Ltd, Winchester,

NH), and dry leaf weight was recorded, followed by grinding in sterile collection vials in a Genogrinder 2000 (SPEX CertiPrep, Metuchen, NJ).

Fungal community composition

Fungal community composition was characterized in each leaf using terminal restriction fragment length polymorphism (TRFLP; Blackwood et al. 2003). After

106

grinding, DNA was extracted from entire leaves without subsampling. DNA extraction was performed using CTAB extraction buffer and bead beating as described in Wu et al.

(2011). PCR amplification of the ITS-1 region was performed using the primers NSI1F

(5’ - GATTGAATGGCTTAGTGAGG) and 5.8SR (5’ – GCTGCGTTCTTCATCGA)

(Martin and Rygiewicz 2005). The forward primer was labeled with HEX (hexachloro-6- carboxyfluorescein). PCRs were performed with a DNA Engine Dyad Peltier Thermal

Cycler (Bio-Rad, Hercules, CA) using 0.025 U/µL Taq DNA polymerase, 3 mM MgCl2,

1X ammonium polymerase buffer (B-Bridge International, Mountain View, CA), 0.2µM each primer (Integrated DNA Technologies, Coralville, IA), 0.16 mM each dNTP, and

0.1 µg/µL bovine serum albumin (New England Biolabs, Ipswich, MA). PCR reaction conditions were: initial denaturation for 3 min at 95 ºC, 30-35 cycles of denaturation for

30 s at 94 ºC, primer annealing for 30 s at 60 ºC and extension for 90 s at 72 ºC, and a final extension for 7 min at 72 ºC. Negative controls were included in each PCR run.

PCR products were run on a 1.5% agarose gel to confirm successful amplification. Cycle number was varied for each sample to obtain a strong band without non-specific amplification.

To generate TRFLP profiles, three replicate PCRs were performed for each sample, and these were pooled prior to restriction enzyme digestion. Samples were digested overnight at 37°C with 10 units HaeIII (New England Biolabs, Ipswich, MA) and cleaned with a DNA probe purification kit (Zymo Research, Irvine, CA). Digested

PCR products were sent to the Ohio State Plant Microbe Genomics Facility for fragment analysis on an Applied Biosystems 3730 DNA Analyzer using a LIZ1200 size standard

107

and minimum peak height of 50 fluorescence units. Peaks between 50 and 600 bp were included in the analysis if they represented ≥ 0.5% relative peak height abundance for the sample.

Construction of distance matrices from leaf habitat patch networks

For each site, the series of field photographs taken before and after sampling each leaf was utilized to record the location of each leaf within the forest floor. Depth of each leaf within the network was recorded as a potential explanatory variable because of probable environmental heterogeneity at the top compared to the bottom of the forest floor. Images and leaf locations were used to visually estimate the percent surface area overlap between each pair of leaves (0%, 20%, 40%, 60%, 80%, or 100%), which was recorded in a proximity matrix. The proximity matrix was non-symmetric, since overlapping leaves often differed in size. We considered leaves to be overlapping only if they physically touched each other in the field, not if they were in similar positions but in differing layers in the forest floor. Network topology was then generated from proximity matrices using the package RBGL (Long et al. 2011) in the software R (Version 2.12.1) and Graphviz (Ellson et al. 2000). All leaves were connected (touching at least one other leaf) at three sites, but we collected one or more non-connected leaves at three sites

(Figure 1). This was due to a lower abundance of beech and maple at some sites which necessitated a somewhat scattered leaf collection pattern in order to obtain at least 10 of each of those leaf types.

108

Figure 1

Graphical representation of step distance matrix showing connections between leaves collected at each Jennings Woods site.

Vernal Pool J3

Riparian J1 Upland L6

109

Figure 1 (continued)

Riparian Q1

Upland 14

Vernal Pool H6

110

We excluded non-connected leaves and calculated two types of “spatial” leaf distance matrices for each site. For the first type of distance matrix, we calculated “step distance” by finding the shortest number of steps between each pair of leaves (i.e., the distance between two adjacent or overlapping leaves was 1 step, the distance between two leaves with one other leaf between them was 2 steps, etc.). In the second distance matrix, we calculated a “weighted distance” where each step was replaced by a distance value equal to 100% minus the percent overlap that the step represented (i.e., % of area not overlapping). For overlapping leaves of differing sizes we used the smaller percent overlap value. The weighted distance between two non-overlapping leaves (i.e., leaves separated by greater than one step) equaled the sum of the distance values for all steps in the shortest path between the leaves.

Statistical analysis

Analysis of environmental and spatial factors within sites

To determine the relative influence of species sorting and neutral processes, we compared the influence of spatial versus environmental variables on fungal community composition using both Mantel analysis and redundancy analysis (RDA). Beta diversity explained solely by spatial effects provides support for dispersal limitation and neutral processes. Beta diversity explained solely by leaf type provides support for species- sorting. Beta diversity explained by depth could support either dispersal limitation or species-sorting, as would the case where spatial effects and leaf type cannot be disentangled. All statistical tests were conducted using the vegan package (Oksanen

2005) in R using 999 random permutations of sample identity. Fungal community

111

composition was represented by Hellinger-transformed TRF relative abundance data

(Legendre & Gallagher 2001).

At each site, we utilized Mantel analysis to test for a significant linear distance- decay relationship by testing for a correlation between Euclidean distance of Hellinger- transformed TRF relative abundances and spatial distance. Mantel analysis is a direct test of the effect of spatial distance on community distance (Borcard and Legendre 2002).

Mantel analysis was done using both stepwise and weighted spatial distances, excluding non-connected leaves. We also conducted Mantel analysis to assess whether fungal community composition was correlated with the environmental variables leaf type and depth within the forest floor. Finally, where multiple explanatory factors were found to be significant, we conducted partial Mantel analysis to separate the influence of the different factors.

Spatial distance matrices were also used to calculate PCNM vectors using the R package PCNM (Borcard and Legendre 2002). PCNM vectors are orthogonal explanatory variables designed to represent the entire range of potential spatial structures that could be found within a spatial distance matrix (Borcard et al. 2004). We conducted redundancy analysis (Legendre and Legendre 1998) at each site on all leaf types collected more than twice (i.e., excluding singleton or doubleton leaf types). Redundancy analysis tested whether fungal community composition was significantly explained by spatial location (represented by PCNM vectors), leaf type, or depth. If a preliminary analysis indicated that PCNM vectors explained significant variation, forward selection was conducted to select a group of parsimonious PCNM vectors to help mitigate Type I errors

112

and avoid an overestimation of the amount of explained variance (Blanchet et al. 2008).

If multiple explanatory factors were significant, variance partitioning (Peres-Neto et al.

2006) was used to determine the relative amount of variance explained by each variable and the interactive effect between the variables. Variance partitioning allowed us to determine whether the impact of variables such as depth and PCNM were independent or embedded within each other.

Collector’s curve analysis

We conducted collector’s curve analysis on all leaves at a site to determine whether community composition stabilizes with increased sampling of leaf habitat patches. Collector’s curves were generated by artificially sampling leaves at a given site using 99 randomly generated leaf addition sequences (Coleman 1981; Colwell and

Coddington 1994). Collector’s curves were analyzed using two indices. The first index was the total number of TRFs detected at the site, and the collector’s curve using this index indicates the number of leaves required to sample all T-RFs within a set of leaves.

The second index was the Hellinger distance between each new leaf and the average community composition of the previous set of leaves, which we call “sample-to-centroid distance”. This collector’s curve indicates the number of samples required to approximate the centroid community composition of a set of leaves. To determine if patterns in collector’s curves were associated with depth, we also ran the analysis using randomized leaf addition sequences stratified by depth.

Analysis of variation across sites

113

Redundancy analysis was also conducted on all communities combined into one data set to test for significant effects of the landscape-scale factors forest habitat type and site nested within habitat, as well as leaf type. This analysis was conducted after leaf types collected only once or twice were removed (resulting in n = 160) because the site factor would necessarily explain all or most of the variance associated with these species.

We also conducted this analysis using only maple & beech leaves (n = 130), which were well-represented at every site.

In order to compare β-diversity among sites, we utilized Anderson’s multivariate homogeneity of group dispersion test (Anderson 2006). The test was performed for all beech and maple leaf communities based on Hellinger distance. This analysis calculates the distance in multivariate space of each leaf community from the site centroid and uses an ANOVA framework to test if group dispersions from various sites are significantly different. Differences were visualized using non-metric multidimensional scaling plots based on Hellinger distance.

Results

Site diversity patterns & processes

Distance-decay and Mantel analyses within sites

Mantel analysis indicated that fungal community composition similarity decreased with spatial distance at all sites except one upland site (P < 0.05; Figure 2).

114

Figure 2

Distance decay plots showing the average Hellinger distance between communities at increasing step distances for each of the six sites in this study. Habitat abbreviations (VP=vernal pool, Rip=riparian, Up=upland), site ID, and the standardized Mantel statistic (rM) are shown for each plot.

1.02 1.2 1.00 VP J3 Rip J1 Up L6 1.00 r 0.98 r rM = 0.316 M = 0.289 M = 0.187 0.98 1.1 0.96 0.96 0.94 0.94 1.0 0.92 0.92 0.90 0.90 0.9 0.88 0.88 Avg. Hel. Dist. Hel. Avg. Dist. Hel. Avg. Dist. Hel. Avg. 0.86 0.86 0.84 0.8 0.84 0.82 0.82 0.80 0.7 0.80 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7

Step distance Step distance Step distance

1.2 0.92 0.94 VP H6 Rip Q1 Up 14 r 0.90 N.S. M = 0.317 rM = 0.280 1.1 0.88 0.92

0.86 1.0 0.90 0.84

0.82 0.9 0.88

Avg. Hel. Dist. Hel. Avg. Dist. Hel. Avg. 0.80 Dist. Hel. Avg.

0.8 0.78 0.86 0.76

0.7 0.74 0.84 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 8 9 10

Step distance Step distance Step distance 115

This pattern was consistent for spatial distance measured as either number of steps or weighted-distance across the leaf networks. Mantel analysis also indicated that depth had a significant effect on fungal community composition at all of the same sites that a significant distance-decay was found (Table 1), indicating that fungal community composition differs among layers within the forest floor. In contrast to the effects of spatial distance and depth, Mantel analysis for leaf type was only significant at one vernal pool site (Table 1).

Partial Mantel analysis showed that depth and spatial distance were often embedded within each other and their relative importance could not be easily disentangled (Table 1). At two sites (vernal pool-J3 and upland-L6) spatial distance was not significant after correction by depth. However, the opposite pattern was found at site riparian-J1. At two sites (vernal pool-H6 and riparian-Q1) both spatial distance and depth were still significant after correction for the other factor. These patterns were consistent for both step- and weighted-distance. Partial Mantel analysis at the one site with a significant leaf type distance-decay indicated it remained significant after correction for step-distance, but not weighted-distance.

Table 1

Mantel standard statistic (rM) values for Mantel and partial Mantel analysis conducted on spatial distance (step or weighted), depth of forest floor leaf was harvested from, and leaf type for all connected leaves. Partial Mantel analysis (“|”) was done to correct for effects of various distance matrices. Non-connected leaves were removed before analysis.

116

Statistical significance is shown with the following symbols: * P < 0.1; ** P < 0.05; ***

P < 0.01; NS: not significant

VP J3 VP H6 Rip J1 Rip Q1 Up L6 Up 14

Step Distance 0.29*** 0.32*** 0.32*** 0.28** 0.19* NS

Step |Type 0.28*** 0.32*** 0.32*** 0.29** 0.18** NS

Step | Depth NS 0.29*** 0.30*** 0.20* NS NS

Depth 0.33*** 0.19*** 0.11* 0.42*** 0.24** NS

Depth |Step 0.18** 0.13** NS 0.38*** 0.18** NS

Type NS 0.16* NS NS NS NS

Type | Step NS 0.16* NS NS NS NS

Weighted Dist 0.23*** 0.29*** 0.26*** 0.33*** 0.23*** NS

Wt | Type 0.22*** 0.29** 0.26*** 0.34** 0.17** NS

Wt | Depth NS 0.27** 0.25*** 0.24** NS NS

Depth 0.33*** 0.19*** 0.11* 0.42*** 0.24** NS

Depth | Wt 0.24*** 0.14** NS 0.36*** 0.20** NS

Type NS 0.16* NS NS NS NS

Type | Wt NS NS NS NS NS NS

n 30 29 30 15 20 30

117

Redundancy analysis of fungal community composition within leaf networks

Fungal communities at all sites except for one were affected by spatial location, as indicated by a significant effect of step-distance PCNM vectors, weighted-distance

PCNM vectors, or both (Table 2). Significant step-distance PCNM vectors explained

7.8% - 12.3% of the variance in fungal community composition among leaves at three sites, while weighted-distance PCNM vectors explained 6.8% - 16.1% of the variance at four sites (Table 2). Leaf type had a significant effect on fungal community composition at 4 sites, explaining between 2.6% - 6.5% of the variance between leaves at one riparian, one vernal pool, and both upland sites (Table 2). Depth had a significant effect on fungal community composition at 4 sites, explaining 3.3% - 15.3% of the variance. Similar to the results of partial Mantel analysis, partial redundancy analysis showed that the influence of spatial distance, depth, and leaf type could sometimes not be disentangled.

This is indicated when the variance explained by all significant factors was less than the sum of the variances explained by each factor alone (Table 2).

Table 2

Proportion of variance (Adjusted R2) in fungal community composition at each site explained by spatial location (PCNM vectors selected after forward selection), depth of community location within forest floor leaf pack, and leaf type. Values shown are from redundancy analysis of all beech & maple leaves that were within a connected leaf network (n). * P < 0.1; ** P < 0.05; *** P < 0.01; NS = not significant. Full model shows the total amount of variance explained by combined significant factors.

118

Site n Leaf type Depth PCNM-step PCNM-wt

Full model Full model

VP J3 21 NS 0.15*** 0.12*** 0.16 0.15*** 0.20

VP H6 20 0.05** 0.03* 0.08** 0.13 NS 0.07

Rip J1 25 0.03* 0.06*** 0.12*** 0.15 0.15*** 0.15 Rip Q1 6 NS NS NS 0.00 0.16** 0.16

Up L6 19 0.07** 0.03** NS 0.10 NS 0.10

Up 14 20 0.04** NS NS 0.04 0.07** 0.08

Collectors curve analysis

When fungal communities were sampled in a random order, community composition at the upland and vernal pool sites stabilized after sampling approximately

10 leaves, as indicated by both the total number of TRFs detected and the sample-to- centroid distance (Figures 3 and 4). When stratifying the sampling by depth, the trajectory of the collector’s curves shifted most dramatically at vernal pool sites. The number of T-RFs detected rapidly increased within the first and third or fourth layers, but other layers were not as distinct and resulted in little increase in the number of T-RFs detected. This is also clear from the vernal pool sample-to-centroid distance plots, in which samples from new depths are increasingly distant from the centroid of previous depths. Stratifying by depth in the upland sites did not impact the shape of the collector’s curves, whereas the impact of depth was intermediate for riparian sites. The sample-to-

119

centroid distance stabilized at the highest value at riparian site J1, indicating that the communities within this site had the highest beta diversity. This result was confirmed with Anderson homogeneity of group dispersion (Figure 5; discussed below).

120

Figure 3

Collectors curve analysis quantifying new TRFs with each randomly selected fungal community. Plots show the set of all randomly selected communities at each site (white circles) and communities randomly selected at each depth within the forest floor leaf pack that the communities were located in (black symbols). Symbols for layers of the leaf pack that communities were sampled from: top layer: X; 2nd layer: square; 3rd layer: triangle; 4th layer: diamond; 5th layer: star; 6th layer circle.

121

Vernal Pool J3 Riparian J1 Upland L6

90 80 100

80 90 70 70 80 60 60 70 50 50 60

TRFs TRFs 40 TRFs 40 50 30 30 40

20 20 30

10 10 20 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Fungal communities Fungal communities Fungal communities

Vernal Pool H6 Riparian Q1 Upland 14

80 80 100

70 70 90

80 60 60 70 50 50 60

TRFs 40 TRFs 40 TRFs 50 30 30 40 20 20 30 10 10 20 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30

Fungal communities Fungal communities Fungal communities

122

Figure 4

Collectors curve analysis quantifying the difference between new and previous fungal community composition. The community composition with each additional leaf contributes to a site centroid composition value, and the composition of each new community is then compared to the site centroid value. Plots show the set of all randomly selected communities at each site (white circles) and communities randomly selected at each depth within the forest floor leaf pack that the communities were located in (black symbols). Symbols for layers of the leaf pack that communities were sampled from: top layer: X; 2nd layer: square; 3rd layer: triangle; 4th layer: diamond; 5th layer: star; 6th layer circle.

123

Vernal Pool J3 Riparian J1 Upland L6

1.0 1.05 0.90

1.00 0.9 0.85 0.95 0.80 0.8 0.90 0.75 0.7 0.85 0.70 0.80 0.6

Hellinger distance Hellinger distance Hellinger distance Hellinger 0.65 0.75 0.5 0.70 0.60 0.4 0.65 0.55 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30

Fungal communities Fungal communities Fungal communities

Vernal Pool H6 Riparian Q1 Upland 14

1.0 0.85 0.95

0.9 0.80 0.90 0.75 0.85 0.8 0.70 0.80 0.7 0.65 0.75 0.6

Hellinger distance Hellinger distance Hellinger 0.60 distance Hellinger 0.70

0.5 0.55 0.65 0.4 0.50 0.60 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30

Fungal communities Fungal communities Fungal communities

124

Comparison of β-diversity among forest sites

Forest habitat, sites nested within habitats, and leaf type explained significant variation in fungal community composition across the forest, although the variance explained by leaf type was much lower than that explained by habitat and site (Table 3).

This pattern was consistent when quantifying across all leaves except singletons & doubletons, and consistent across beech and maple leaves only. Principal coordinates analysis (Figure 5) indicates that fungal communities on leaves in the vernal pool sites were particularly different from the upland and riparian sites. The site effect was strongest in riparian forest habitat, with the riparian forest sites separated by the second principal coordinates axis (Figure 5). The differences between beech and maple leaves were too minor to be reflected in the first two ordination axes.

Table 3

Redundancy analysis determining the effect of factors on community composition for all beech (n=70) & maple (n=60) leaves across all sites and all leaves (n=160) after singleton and doubleton species at each site were removed. The significance (p-value) is given for each factor, as is the amount of variance in community composition explained by each factor (Adj R2).

125

Beech & Maple only, All leaves, no includes site X type interaction int Test variable p-value AdjR2 test p- AdjR2 test var var value

Site 0.005 0.07 0.005 0.04

Leaf type 0.010 0.01 0.010 0.02

Site x type int 0.005 0.02 NA NA Habitat 0.005 0.18 0.005 0.18

Depth 0.005 0.02 0.005 0.02

126

Figure 5

Anderson multivariate homogeneity of group dispersion determined on TRFLP

(community composition) profiles for all fungal communities found on beech (n=70) and maple (n=60) leaves. The central point of each “starburst” is the centroid in multivariate space for the communities found at any given site. The terminus of each line emanating from the centroid is the point in multivariate space of a community on one leaf. Site identification is found in the legend (V = vernal pool, Rip = riparian, Up = upland). The p-value tests for significant differences between groups (sites).

127

P < 0.001 VP J3 VP H6 Rip Q1 Rip J1 Up L6 Up 14

128

The magnitude of β-diversity within a site, measured as the multivariate dispersion among leaves at a site (i.e., average distance to the site centroid), significantly differed among habitats and sites across the forest (P < 0.001). β-diversity (± SD) was lower in the upland (0.613 ± 0.007) and vernal pool sites (0.629 ± 0.003) than in the riparian sites (0.641 ± 0.081). One riparian site (J1), in particular, had the highest beta diversity and more variable fungal communities than all other sites, with communities resembling those found on leaves from both upland sites and the other riparian site

(Figure 5).

Discussion

Fungal distance-decay

A primary objective of this work was to quantify fungal diversity patterns at multiple spatial scales. One pattern that has been commonly described in ecology is the distance-decay pattern (Nekola & White 1999), where community similarity decreases at increasing spatial distance. We utilized the discrete nature of senesced leaf habitat patches to quantify individual fungal communities and construct spatial distance matrices in order to determine if fungi would follow a distance-decay pattern. At five out of the six sites sampled, we found a distance-decay relationship where community distance was significantly correlated with spatial distance across 30 local communities located within a

850 cm2 plot. Previous studies had quantified distance-decay in fungal communities at much larger spatial scales. Green et al. (2004) found this pattern in soil fungi in arid

Australian habitat across distances ranging from 1 m2 to 1010 m2. Distance-decay was also described for arbuscular mycorrhizal fungi colonizing Allium cepa roots growing in

129

soil cores collected from 100 m2 plots separated by up to 250 km (van der Gast et al.

2011). In contrast to these previous studies, we examined connected networks of leaves where individuals or populations had direct physical access to each habitat patch, and where dispersal limitation and habitat heterogeneity seem less likely to be potentially important factors. However, here we provide evidence for the first time that distance- decay is operational in fungal communities located on discrete, but adjacent, habitat patches.

Influencing distance-decay: Species-sorting and dispersal limitation

We attempted to quantify the relative influence of species-sorting and dispersal limitation on the distance-decay relationship and β-diversity detected in fungal communities on leaves. However, because variance in community composition explained by spatial distance could also be largely explained by depth within the forest floor, species-sorting and dispersal limitation factors could not be unambiguously disentangled. This confounding of spatial distance with environmental variation is not uncommon (Astorga et al. 2012) and has been described in the understory vegetation of a temperate forest similar to the one we worked in (Gilbert and Lechowicz 2004). Overlap in variance explained by these factors may indicate that both species sorting and dispersal limitation are important, as has been shown previously for arbuscular mycorrhizal communities (Dumbrell et al. 2010), fermentative bacterial communities (Boyd et al.

2010), atmospheric bacteria colonizing new sterile habitat (Langenheder and Székely

2011), and soil bacterial communities (Hovatter et al. 2011).

130

The impact of environmental heterogeneity on community composition

The link between environmental heterogeneity and biological diversity has been described as being operational all levels of ecological organization (Townsend and

Fuhlendorf 2010) to the point where heterogeneity should serve as a foundation for conservation and ecosystem management (Christensen 1997, Wiens 1997). We found evidence that environmental heterogeneity may influence fungal community composition at spatial scales ranging from across the landscape to within an individual sampling site.

The pool of potential organisms that are available to form local communities is known as the habitat species pool (Belyea and Lancaster 1999). In this study we found evidence

(Figure 6) supporting the existence of two habitat species pools: one located in upland and riparian habitat and the other in vernal pool habitat. This pattern provides evidence that at a landscape scale, fungal community composition is generally filtered by species- sorting processes; most likely associated with seasonally saturated conditions that typify vernal pools and are not found within either upland or riparian sites.

Environmental heterogeneity located within sampling sites is potentially associated with leaf type and the depth at which a leaf is found. We were surprised to find that leaf type had only a small effect on fungal community composition, despite sampling leaves that should differ dramatically in lignin content. If high taxa diversity

(Torsvik et al. 1990, Gans et al. 2005) and functional redundancy (Franklin and Mills

2006) are coupled with community plasticity of functional traits (extracellular enzyme production) allows a community to shift functional activity according to the relative proportions of substrate available within a leaf at any point in time, communities would

131

not be expected to be strongly filtered by leaf type. In that case, composition would likely be more influenced by other processes (abotic conditions, presence of biological competitors, dispersal limitation).

Depth of a leaf within the Oi horizon is likely related to a variety of environmental variables that may affect fungal community composition, including moisture and access to nutrients in the mineral soil (Heal et al. 1997, Bjørnlund and

Christensen 2005). Differences between fungal communities located in adjacent layers of the leaf pack were the most pronounced in the riparian and vernal pool habitats (Figures 4 and 5); habitats with greater moisture heterogeneity than the upland. An increase in moisture has been associated with increased fungal richness in tropical (Braga-Neto et al.

2008) and boreal (Przybyl et al. 2008) leaf litter. We also noted during leaf collection that vernal pool leaves were mixed with water-saturated mineral soil, which was removed during sampling, but may cause the lower layers of the leaf pack to be more compact

(Rinklebe and Langer 2006) and prone to anaerobiosis.

A relative lack of environmental heterogeneity may help explain why community composition was less influenced by distance-decay spatial effects at the upland sites. In the upland sites community composition was the most stable (Figures 4 and 5). The leaf litter was dominated by beech and maple at these sites, and in order to collect the necessary 30 leaves, we did not need to go as deep into the leaf pack as we did at vernal pool and riparian sites. The upland sites were also elevated and exposed to windier conditions than either the riparian or vernal pool sites. Wind may mix leaves more frequently in the upland, disturbing the arrangement of adjacent habitat patches. The

132

wind and location of the communities we collected near the surface of the leaf pack may have caused upland leaves to be drier than those in riparian (and certainly than those in vernal pools). Dry conditions have been described to slow the growth of fungal hyphae

(Hawkes et al. 2011).

The impact of colonization on community composition

Beta-diversity of the fungal communities at this scale is likely influenced by two different pools of fungal colonizers. Leaves deep within the leaf pack are in contact with the Oe and A horizons and are likely to be colonized by soil fungi (Smith and Bradford

2003, Carillo et al. 2011). Soil is a microscopically heterogeneous matrix of organic and inorganic material (Pajor et al. 2010), with varying water and oxygen availability creating spatially separated aerobic and anaerobic niches (Young and Crawford 2004, Or et al.

2007). These heterogeneous conditions in soil are thought to be responsible for the extremely high bacterial diversity that has been quantified in soil (Curtis and Sloan 2005,

Gans et al. 2005). In contrast, leaf litter is mostly organic and aerobic (except seasonally within vernal pools), and the fungal diversity of litter relative to soil has not been quantified (Hyde et al. 2007).

In addition to colonization via hyphal growth that we would expect fungi in adjacent leaf litter and soil, fungi may also disperse over relatively long distances via spores (Boddy et al. 2008). Some fungi have been documented to have generalist strategies and are capable of existing in very different environmental conditions due to their ability to adapt to a wide variety of environmental factors (Cantrell et al. 2011).

Spores may be ejected into the air quite forcefully (Trail 2007), and studies have shown

133

that spores could be expected to easily disperse over the distances between our sites

(Stockmarr et al. 2007, Boddy et al. 2009). It is possible that that several leaves located in very different habitats (vernal pool and upland) had very similar community compositions (Figure 6) because of colonization from spores and subsequent priority effects. In order to quantify the influence of spore colonization compared to colonization from adjacent leaf litter and soil, changes in community composition over time would have to be quantified in these different microsites with powerful next-generation pyrosequencing.

Variations in fungal beta diversity

Site beta diversity was very different in the two riparian sites, with one site

(riparian-J1) having the highest and one site (riparian-Q1) having the lowest beta diversity found among all sites (Figure 6). In a complementary study, 75% of the leaf communities that were surveyed at the site with the lowest beta diversity (Q1) were dominated by one sequence-based operational taxonomic unit within the genus Mycena

(Feinstein and Blackwood 2012). Aggressive growth of this fungus may be responsible for the low beta-diversity that we found at that site. Riparian areas in general are characterized by a high degree of environmental variability including localized patches of nutrients (Liu et al. 2008) and moisture saturation (Douglas et al. 2005), which may affect fungal community diversity (Rinklebe and Langer 2006). Hence, it is also possible that the difference in beta diversity associated with the two riparian sites in this study could be related to localized environmental heterogeneity between riparian sites.

134

Conclusion

The significant distance-decay relationships that were found in this study are the first to be quantified at the scale of networks of adjacent leaves. Although leaves of different tree species are known to have differing amounts of biochemical resources, which affects decomposition rate and microbial metabolism, leaf type had a relatively small effect on fungal community composition. However, because we found differences among communities associated with depth and habitat type, fungal community composition was shown to be potentially associated with other types of environmental heterogeneity at landscape and local scales. In our sampling design it was not possible to disentangle the influence of these factors from dispersal limitation, which should be addressed in future studies. Although fungal community composition may be characterized at upland and vernal pool sites by quantifying communities on ~10 leaves, it is clear that the sampling of depths (leaf layers) within the Oi horizon will have an impact on patterns in community composition.

References

Amend AS, Seifert KA, Samson R, Bruns TD. 2010. Indoor fungal composition is

geographically patterned and more diverse in temperate zones than in the tropics.

Proceedings of the National Academy of Science 107: 13748-13753.

Anderson MJ, Ellingsen KE, McArdle BH. 2006. Multivariate dispersion as a measure

of beta diversity. Ecology Letters 9: 683-693.

135

Astorga A, Oksanen J, Luoto M, Soininen J, Virtanen R, Muotka T. 2012. Distance

decay of similarity in freshwater communities: do macro- and microorganisms

follow the same rules? Global Ecology and Biogeography 21: 365-375.

Berg B, Laskowski R. 2006. Litter Decomposition: A Guide to Carbon and Nutrient

Turnover. Burlington, MA, USA: Elsevier 448 p.

Belyea LR, Lancaster J. 1999. Assembly rules within a contingent ecology. Oikos 86:

402–416.

Bjørnlund L, Christensen S. 2005. How does litter quality and site heterogeneity interact

on decomposer food webs of a semi-natural forest? Soil Biology and

Biochemistry 37: 203-213.

Blanchet FG, Legendre P, Borcard D. 2008. Forward selection of explanatory variables.

Ecology 89: 2623-2632.

Blackwood CB, Marsh T, Kim S, Paul E. 2003. Terminal restriction length

polymorphism data analysis for quantitative comparison of microbial

communities. Applied and Environmental Microbiology 69(2): 926-932.

Braga-Neto R, Luizao RCC, Magnusson WE, Zuquim G, De Castilho, CV. 2008. Leaf

litter fungi in a Central Amazonian forest: the influence of rainfall, soil and

topography on the distribution of fruiting bodies. Biodiversity and Conservation;

17: 2701–2712.

Boddy L, Frankland JC, van West P (eds). 2008. Ecology of saprotrophic

basidiomycetes (Vol. 28). Amsterdam: Academic Press. 386 p.

136

Boddy L, Hynes J, Bebber DP. 2009. Saprotrophic cord systems: dispersal mechanisms

in space and time. Mycoscience 50(1): 9-19.

Borcard D, Legendre P. 2002. All-scale spatial analysis of ecological data by means of

principal coordinates of neighbour matrices. Ecological Modelling 153: 51–68.

Borcard D, Legendre P, Avois-Jacquet C, Tuomisto H. 2004. Dissecting the spatial

structure of ecological data at multiple spatial scales. Ecology 85(7): 1826-1832.

Boyd E, Hamilton T, Spear J, Lavin M, Peters J. 2010. [FeFe]-hydrogenase in

Yellowstone National Park: evidence for dispersal limitation and phylogenetic

niche conservatism. ISME Journal 4(12): 1485-1495.

Brose U, Ostling A, Harrison K, Martinez ND. 2004. Unified spatial scaling of species

and their trophic interactions. Nature 428: 167–171.

Cantrell SA, Dianese JC, Fell J, Gunde-Cimerman N, Zalar P. 2011. Unusual fungal

niches. Mycologia 103(6): 1161-1174.

Carillo Y, Ball BA, Bradford MA, Jordan CF, Molina M. 2011. Soil fauna alter the

effects of litter composition on nitrogen cycling in a mineral soil. Soil Biology

and Biochemistry 43: 1440-1449.

Chase JM. 2005. Towards a really uniﬁed theory for metacommunities. Functional

Ecology 19: 182–186.

Christensen NL. 1997. Managing for heterogeneity and complexity on dynamic

landscapes. In: Pickett STA, Ostfeld RS, Shachak M, Likens, GE (eds.). The

ecological basis for conservation: heterogeneity, ecosystems, and biodiversity;

New York, New York: p. 167-186.

137

Cody ML. 1975. Towards a theory of continental species diversities: bird distributions

over Mediterranean habitat gradients. In: Cody, M. L. and Diamond, J. M. (eds),

Ecology and evolution of communities. Harvard Univ. Press, Cambridge, MA pp.

214-257.

Coleman BD. 1981. On random placement and species-area relations. Mathematical

Biosciences 54: 191–215.

Colwell RK, Coddington JA. 1994. Estimating terrestrial biodiversity through

extrapolation. Philosophical Transactions of the Royal Society Series B:

Biological Sciences; 345:101–118.

Cottenie K. 2005. Integrating environmental and spatial processes in ecological

research. Ecology Letters 8: 1175–1182.

Cornwell WK, Cornelissen JHC, Amatangelo K, Dorrepaal E, Eviner VT, et al. 2008.

Plant species traits are the predominant control on litter decomposition rates

within biomes worldwide. Ecology Letters 11: 1065–1071.

Curtis TP, Sloan WT. 2005. Exploring microbial diversity - A vast below. Science 309:

1331- 1333.

Douglas MM, Bunn SE, Davies PM. 2005. River and wetland food webs in Australia’s

wet-dry tropics: general principles and implications for management. Marine

Freshwater Research 56: 329–342.

Dumbrell AJ, Nelson M, Helgason T, Dytham C, Fitter AH. 2010. Relative roles of

niche and neutral processes in structuring a soil microbial community. ISME

Journal 4(3): 337-345.

138

Ellson J, Gansner E, Koutsofios L, North SC, Woodhull G. 2000. Graphviz— Open

Source Graph Drawing Tools, Lecture Notes in Computer Science 2265: 594-597.

Feinstein LM, Blackwood CB. 2012. Taxa–area relationship and neutral dynamics

influence the diversity of fungal communities on senesced tree leaves.

Environmental Microbiology published online April 11, 2012.

Fierer N, Breitbart M, Nulton J, Salamon P. Louzupone C, Jones R. et al. 2007.

Metagenomic and Small-Subunit rRNA Analyses Reveal the Genetic Diversity of

Bacteria, Archaea, Fungi, and Viruses in Soil. Applied and Environmental

Microbiology 73: 7059–7066.

Filotas E, Grant M, Parrott L, Rikvold PA. 2010. Positive interactions and the

emergence of community structure in metacommunities. Journal of Theoretical

Biology 266: 419–429.

Franklin RB, Mills AL. 2006. Structural and functional responses of a sewage microbial

community to dilution-induced reductions in diversity. Microbial Ecology 52:

280–288.

Fukami T, Dickie IA, Wilkie JP, Paulus BC, Park D, Roberts A, Buchanan PK, Allen

RB. 2010. Assembly history dictates ecosystem functioning: evidence from

wood decomposer communities. Ecology Letters 13: 675-684.

Gans J, Wolinsky M, Dunbar J. 2005. Computational improvements reveal great

bacterial diversity and high metal toxicity in soil. Science 309: 1387-1390.

139

Gessner MO, Swan CM, Dang CK, McKie BG, Bardgett RD, Wall DH, Hӓttenschwiler

S. 2010. Diversity meets decomposition. Trends in Ecology and Evolution 25:

373-380.

Gilbert B, Lechowicz MJ. 2004. Neutrality, niches, and dispersal in a temperate forest

understory. Proceedings of the National Academy of Sciences USA 101: 7651–

7656.

Gotelli N, Colwell RK. 2001. Quantifying biodiversity: procedures and pitfalls in the

measurement and comparison of species richness. Ecology Letters 4: 379–391.

Gravel D, Canham CD, Beaudet M. Messier C. 2006. Reconciling niche and neutrality:

the continuum hypothesis. Ecology Letters 9: 399–409.

Green JL, Holmes AJ, Westoby M, Oliver I, Briscoe D, Dangerfield M, Gillings M,

Beattie AJ. 2004. Spatial scaling of microbial eukaryote diversity. Nature 432:

747-750.

Green J, Bohannan BJM. 2006. Spatial scaling of microbial biodiversity. Trends in

Ecology and Evolution 21: 501-507.

Hawkes CV, Kivlin SN, Rocca JD, Huguet V, Thomsen MA, Suttle KB. 2011. Fungal

community responses to precipitation. Global Change Biology 17: 1637–1645.

Heal OW, Anderson JM, Swift MJ. 1997. Plant litter quality and decomposition: an

historical overview. In: Cadisch G, Giller KE. (Eds.), Driven by

Nature.Cambridge UK, CAB International; pp. 3–30.

Helgason T, Merryweather JW, Denison J, Wilson P, Young JPW, Fitter AH. 2002.

Selectivity and functional diversity in arbuscular mycorrhizas of co-occurring

140

fungi and plants from temperate deciduous woodland. Journal of Ecology 90:

371-384.

Holt RD. 2002. Food webs in space: On the interplay of dynamic instability and spatial

processes. Ecological Research 17: 261–273.

Hovatter S, Dejelo C, Case AL, Blackwood CB. 2011. Metacommunity organization of

soil microorganisms depends on habitat type defined by presence of Lobelia

siphilitica plants. Ecology 92: 57-65.

Hubbell SJ. 2001. The Unified Neutral Theory of Biodiversity and Biogeography.

Princeton, NJ, Princeton University Press; 448 p.

Hyde KD, Bussaban B, Paulus B, Crous PW, Lee S, McKenzie EHC, Photita W,

Lumyong S. 2007. Diversity of saprobic microfungi. Biodiversity and

Conservation 16: 7-35.

Jabot F, Bascompte J. 2012. Bitrophic interactions shape biodiversity in space.

Proceedings of the National Academy of Science 109: 4521-4526.

Kivlin SN, Hawkes CV, Treseder KK. 2011. Global diversity and distribution of

arbuscular mycorrhizal fungi. Soil Biology and Biochemistry 43: 2294-2303.

Legendre P, Legendre L. 1998. Numerical Ecology 2nd English edition, New York, NY,

Elsevier; 853 p.

Legendre P, Gallagher ED. 2001. Ecologically meaningful transformations for

ordination of species data. Oecologia 129: 271-280.

141

Legendre P, Borcard D, Peres-Neto PR. 2005. Analyzing beta diversity: partitioning the

spatial variation of community composition data. Ecological Monographs 75:

435–450.

Legendre P, Borcard D, Peres-Neto P.R. 2008. Analyzing or explaining beta diversity?

Comment, Ecology 89: 3238-3244.

Leibold MA, Holyoak M, Mouquet N, Amarasekare P, Chase JM, Hoopes MF. Holt RD,

Shurin JD, Law R, Tilman D, Loreau M, Gonzalez A. 2004. The

metacommunity concept: a framework for multi-scale community ecology.

Ecology Letters 7: 601–613.

Lekberg Y, Koide RT, Rohr JR, Aldrich-Wolfe L, Morton JB. 2007. Role of niche

restrictions and dispersal in the composition of arbuscular mycorrhizal fungal

communities. Journal of Ecology 95: 95–105.

Liu X, Zhang X, Zhang M. 2008. Major factors influencing the efficacy of vegetated

buffers on sediment trapping: A review and analysis. Journal of Environmental

Quality 37: 1667–1674.

Long L, Carey VJ, Gentleman R. 2011. RBGL: R interface to boost graph library.

http://www.bioconductor.org/packages/release/bioc/html/RBGL.html

Magurran AE. 2004. Measuring Biological Diversity. Oxford, UK: Blackwell

Publishing; 264 p.

Mantel N. 1967. The detection of disease clustering and generalized regression

approach. Cancer Research 27: 209-220.

142

Martiny JBH, Bohannan BJM, Brown JH, Colwell RK, Fuhrman JA, Green JL, Horner-

Devine MC, Kane M, Krumins JA, Kuske CR, Morin PJ, Naeem S, Øvreås L,

Reysenbach AL, Smith VH, Staley JT. 2006. Microbial biogeography: putting

microorganisms on the map. Nature Reviews Microbiology 4: 102–112.

Martin KJ, Rygiewic PT. 2005. Fungal-specific PCR primers developed for analysis of

the ITS region of environmental DNA extracts. BMC Microbiology 5:1-11.

May RM. (1976). Theoretical ecology: Principles and applications, Philadelphia, PA.,

Saunders; 272 p.

Nekola JC, White PS. 1999. The distance decay of similarity in biogeography and

ecology. Journal of Biogeography 26: 867–878.

Nemergut DR, Costello EK, Hamady M, Lozupone C, Jiang L, Schmidt SK, Fierer N,

Townsend AR, Cleveland CC, Stanish L, Knight R. 2011. Global patterns in the

biogeography of bacterial taxa. Environmental Microbiology 13: 135-144.

Norberg J, Swaney DP, Dushoff J, Lin J, Casagrandi R, Levin SA. 2001. Phenotypic

diversity and ecosystem functioning in changing environments: a theoretical

framework. Proceedings of the National Academy of Science 98(20): 11376–

11381.

Oksanen J. 2005. Vegan: R functions for vegetation ecologists. Available at:

http://cc.oulu.fi/_jarioksa/softhelp/vegan.html.

Öpik M, Metsis M, Daniell TJ, Zobel M, Moora M. 2009. Large-scale parallel 454

sequencing reveals host ecological group specificity of arbuscular mycorrhizal

fungi in a boreonemoral forest. New Phytologist 184: 424-437.

143

Or D, Smets BF, Wraith JM, Dechesne A, Friedman S.P. 2007. Physical constraints

affecting microbial habitats and activity in unsaturated porous media – a review,

Advances in Water Resources 30: 1505-1527.

Peay KG, Kennedy PG, Bruns TD. 2008. Fungal community ecology: A hybrid beast

with a molecular master. BioScience 58: 799-810.

Peay KG, Garbolotto M, Bruns TD. 2010. Evidence of dispersal limitation in soil

microorganisms: Isolation reduces species richness on mycorrhizal tree islands.

Ecology 91: 3631-3640.

Peres-Neto P, Legendre P, Dray S, Borcard D. 2006. Variation partitioning of species

data matrices: estimation and comparison of fractions. Ecology 87: 2614–2625.

Przybyl K, Karolewski P, Oleksyn J, Labedzki A, Reich PB. 2008. Fungal diversity of

Norway spruce litter: effects of site conditions and premature leaf fall caused by

bark beetle outbreak. Microbial Ecology 56: 332-340.

Rinklebe J, Langer U. 2006. Microbial diversity in three floodplain soils at Elbe River

(Germany). Soil Biology and Biochemistry 38: 2144-2151.

Riser PG. 1995. Biodiversity and ecosystem function. Conservation Biology 9: 742-746.

Savage VM, Webb CT, Norberg J. 2007. A general mutil-trait-based framework for

studying the effects of biodiversity on ecosystem functioning. Journal of

Theoretical Biology 247: 213-229.

Smith VC, Bradford MA. 2003. Litter quality impacts on grassland litter decomposition

are differently dependent on soil fauna across time. Applied Soil Ecology 24:

197-203.

144

Sokal RR. 1979. Testing statistical significance of geographic variation patterns.

Systematic Zoology 28: 627-632.

Stockmarr A, Andreasen V, Østergård H. 2007. Dispersal distances for airborne spores

based on deposition rates and stochastic modeling. Phytopathology; 97: 1325-

1330.

Tiunov AV, Scheu S. 2005. Facilitative interactions rather than resource partitioning

drive diversity-functioning relationships in laboratory fungal communities,

Ecology Letters 8: 618–625.

Torsvik V, Goksøyr J, Daae FL 1990. High diversity in DNA of soil bacteria. Appl

Environmental Microbiology 56: 782–787.

Townsend DE, Fuhlendorf SD. 2010. Evaluating relationships between spatial

heterogeneity and the biotic and abiotic environments. American Midland

Naturalist 163: 351–365.

Trail F. 2007. Fungal cannons: explosive spore discharge in the Ascomycota. FEMS

Microbiology Letters 276 (1): 12–18

Tuomisto H, Ruokolainen K. 2006. Analyzing or explaining beta diversity?

Understanding the targets of different methods of analysis. Ecology 87: 2697–

2708.

Tuomisto H, Ruokolainen K. 2008. Analyzing or explaining beta diversity? Reply.

Ecology 89: 3244–3256.

145

van der Gast CJ, Gosling P, Tiwari B, Bending GD. 2011. Spatial scaling of arbuscular

mycorrhizal fungal diversity is affected by farming practice, Environmental

Microbiology 13: 1462-2920.

Whittaker RH. 1972. Evolution and measurement of species diversity. Taxon 21: 213–

251.

Wiens JA. 1997. The emerging role of patchiness in conservation biology, In: Pickett

STA, Ostfeld RS, Shachak M, Likens GE (eds.). The ecological basis for

conservation: heterogeneity, ecosystems, and biodiversity; New York, New

York: p. 93–107.

Young IM, Crawford JW. 2004. Interactions and self-organization in the soil-microbe

complex, Science 304: 1634–1637.

Zak JC, Visser S. 1996. An appraisal of soil fungal biodiversity: the crossroads between

taxonomic and functional biodiversity. Biodiversity & Conservation 5: 169-183.

Chapter Five

The impact of litter diversity on microbial enzyme activity, biomass, and leaf litter

decomposition

Abstract

Recent reviews and research have noted the importance of understanding the relationship between diversity, community assembly, and ecosystem functioning. However, few studies have examined simultaneous changes in functional activity and microbial community composition during decomposition. We assembled 1, 3, and 4 species litterbags using labile (Hamamelis virginiana, Acer saccharum) and recalcitrant (Fagus grandifolia, Quercus palustris) leaf types. Litterbags were periodically harvested over a one-year period from two habitats (upland and riparian forest) in Northeastern Ohio.

We found that although decomposition rates were not significantly different in litter mixtures than in single-species litter bags, litter mixtures were documented to have an influence on microbial activity. There was an increase in bacterial and fungal beta diversity associated with litter mixtures, as well as increased fungal and bacterial biomass. We describe how litter mixtures provide evidence for the first time of saprotrophic microbe-driven increases in ecosystem exergy; indicating an important influence of litter mixtures on ecosystem energy transfer. The importance of microbial community composition was also demonstrated as community composition was significantly correlated with decomposition (% original mass remaining) and

146

147

extracellular enzyme activity. Community composition correlation with enzyme activity provides empirical support for the Guild Decomposition Model. This study illustrates that in order to obtain a comprehensive understanding of factors influencing litter decomposition, quantification and analysis of microbial community composition and extracellular enzyme activity should be included in the experimental design along with traditional variables such as habitat, time of incubation, and proportions of leaf resource compounds.

Introduction

Leaf litter decomposition is second only to photosynthesis as a global carbon flux

(Bray and Gorham 1964) and a major pathway whereby carbon is transformed into soil organic matter (Lavelle and Spain 2001, Sayer 2006). The chemical composition of leaf litter varies widely among tree species, and has been found to be highly correlated with leaf decay rates (Melillo et al. 1982, Aerts 1997, Osono and Takeda 2005, Zhang et al.

2008, Perez-Harguindeguy et al. 2008, Cornwell et al. 2008, Hӓttenschwiler et al. 2011).

Labile leaves are rich in N and carbon compounds that can be easily utilized (e.g., soluble compounds, starch), while recalcitrant leaves have less N and a higher proportion of carbon that is resistant to breakdown (e.g., lignin) (Berg and McClaugherty 2008,

Gessner et al. 2010). However, when labile and recalcitrant leaf types are mixed, the ability to predict decomposition rates based on the rates of individual species has been shown to be quite limited (Hӓttenschwiler et al. 2005). In some cases, mixed litter may decompose at the average rate of the leaf types in the mixture (additive decomposition;

148

Rusted 1994, Ball et al. 2008). However, litter mixtures have also been noted to decompose faster (Schweitzer et al. 2005, Jonsson and Wardle 2008) and slower

(Bardgett and Shine 1999) than the average rates of the leaf types in the mixtures

(collectively known as non-additive effects). Indeed, several reviews have indicated that the decomposition rate of leaves in litter mixtures may be increased, decreased, or unchanged compared to the rates of the individual species when not in a mixture (Wardle et al. 1997, Gartner and Cardon, 2004, Hӓttenschwiler et al. 2005). While it has been shown that the effects of litter diversity are primarily dependent on leaf type (Wu et al.

2011), inconsistent results in decomposition studies indicate that examination of the microbial activity that mediates decomposition may be necessary to increase our ability to predict the effects of litter diversity on rates of decomposition.

Fungi and bacteria are responsible for 90% of all organic matter decomposition

(Swift et al. 1979), with saprotrophic fungi considered to be the most efficient decomposers of the two most abundant biopolymers found in leaf litter, cellulose and lignin (Kjøller and Struwe 2002, Berg and McClaughtery 2008). Saprotrophic microbes secrete extracellular enzymes to break down complex plant litter biopolymers into smaller products that can be easily metabolized (Nannipieri et al. 2002, Ratledge 1994,

Sinsabaugh et al. 2008). Each extracellular enzyme targets a specific substrate (Table 1).

Enzyme activity is generally correlated with the decomposition of the corresponding polymer substrate (Snajdr et al. 2011, Zifcakova et al. 2011), but not always (Sinsabaugh et al. 2005, Voriskova et al. 2011). Quantifying how microbial enzyme activity is

149

affected by mixtures of litter may help us increase our capability to predict decomposition rates in litter mixtures.

Table 1

Commonly reported hydrolytic and oxidative carbon-acquiring enzymes, the natural substrates they are known to degrade, and substrate used in lab assays to determine enzyme activity.

Natural Enzyme Substrate substrate

α-1,4-glucosidase Starch 4-MUB-α-D-glucopyranoside

β-1,4-glucosidase Cellulose 4-MUB-β-D-glucopyranoside

Cellobiohydrolase Cellulose 4-MUB- β -D-cellobioside

β-1,4-xylosidase Hemicellulose 4-MUB-B-D- xylopyranoside

Laccase Laccase 2,2′-azinobis-3-ethylbenzothiazoline-6- sulfonate (ABTS)

β-1,4-N- Chitin 4-MUB-N-acetyl-β-D-glucosaminide acetylglucosaminidase (Nagase)

Phosphatase Phosphate ester 4-MUB phosphate bonds

150

The microbial biomass that accumulates during decomposition has been shown to be an important factor that influences decomposition (Papa et al. 2008) and the activity of lignin-degrading (Fioretto et al. 2007) and cellulose-degrading (Sall et al. 2003) enzymes.

However, the interactions between litter diversity, microbial biomass, and decomposition rate are complex. Ecological theory suggests that litter mixture treatments are likely to influence microbial biomass. The species complementarity hypothesis states that increased resource heterogeneity offers greater opportunity for niche differentiation, facilitating more extensive utilization of resources by communities, as well as increased productivity and biomass (Tilman et al. 1997, Hooper and Vitousek 1998, Loreau and

Hector 2001, Marquard et al. 2009). Because leaf species differ in the proportions of different types of carbon substrates that saprotrophic microbes metabolize, litter mixtures provide microbial communities with greater resource heterogeneity than single-species litter. However, microbial biomass has previously been shown to increase in mixed litter treatments in some studies (Chapman and Newman 2010, Kominoski et al. 2007), but was either weakly correlated (Chapman and Newman 2010) or uncorrelated (Vivanco and Austin 2008) with decomposition rates.

Variation in how biomass impacts decomposition may be linked to microbial community composition. Moorhead and Sinsabaugh (2006) proposed a succession of three microbial guilds during decomposition: opportunists (consuming soluble polymers and intermediate metabolites), decomposers (consuming cellulose), and miners

(consuming aromatic ring structures). Because shifts in microbial community composition are likely linked to changing proportions of leaf carbon substrates and

151

differing capabilities of taxa to produce extracellular enzymes (Moorhead & Sinsabaugh

2006), the decomposition rate of litter mixtures should be related these shifts in microbial community composition. Alternatively, community composition could be unrelated to litter decomposition because microbial communities contain a diverse mix of functionally redundant taxa (Zak et al. 1994). If communities of differing compositions can express the same enzymes, or one community can express the appropriate enzymes at differing stages of decomposition, enzyme activity would shift in relation to substrate pool but community composition would not. Quantifying and understanding how community composition varies in conjunction with litter treatments (single- and mixed-species) may therefore help us predict decomposition rates.

Our primary goals were to 1) quantify the effect of litter diversity on litter decomposition, microbial biomass, extracellular enzyme activity, and microbial community composition and 2) determine how different aspects of the microbial community interact and impact litter decomposition. To accomplish these goals, we incubated four leaf types in mixed- and single-species litter bags in two temperate deciduous forest habitats. Decomposition and the microbial community were monitored over an annual cycle of decomposition. The experimental design allowed us to compare the importance of litter diversity to the importance of leaf type and forest habitat, as well as additional environmental heterogeneity at the meter and centimeter spatial scales.

Based on the species complementarity hypothesis, we predicted that decomposition rates in litter mixtures would be greater than in single-species litter, and that microbial biomass would be higher. According to the guild decomposition model, we predicted that there

152

would be significant correlations between patterns in enzyme activity and microbial community composition, as well as decomposition (% of original organic matter remaining).

Methods

Experimental design & field site

Our study site was in Jennings Woods, a 30 ha deciduous forest in NE Ohio containing upland and riparian habitat. During October 2007, we gathered senesced leaves that had not yet dropped to the ground from two tree species expected to have labile leaves (Acer saccharum and Hamamelis virginiana) and two species expected to have recalcitrant leaves (Fagus grandifolia and Quercus palustris). Leaves were air- dried in a laboratory at room temperature and then placed into polyethylene mesh bags

(16 x 28 cm, 3 mm mesh openings; Sacramento Bag Manufacturing Co., Woodland,

CA), which were closed with plastic ties. The design included nine litter mixture treatments: each single species alone, each of the four possible 3-species mixtures, and a mixture of all four species. Each litter bag contained a total of 5g of dried leaf material, with the weight evenly divided among species in mixed-species bags. Initial percent carbon and nitrogen content of each leaf species was determined using an

Elemental Combustion System (Costech Instruments, Valencia, CA).

Litterbags were deployed in twenty replicate blocks (0.5 x 1m) in each of the two forest habitats (upland and riparian forest) in November 2007. One litterbag of each treatment was placed in a randomly selected location in each block. Litterbags were

153

tied to plastic tent stakes anchored to the ground after forest floor litter (Oi horizon) was removed to avoid initial confounding of the litter diversity treatments. At approximately 10-week intervals (days 51, 126, 212, 287, 385 after deployment), four replicate blocks from each habitat were harvested. Litterbags were individually placed into sterile plastic Whirl-Pak bags (Nasco, Ft. Atkinson, WI) and transported on ice to the lab for analysis.

Sample processing

Leaves were rinsed with autoclaved distilled water to remove attached material and separated according to leaf species in order to isolate individual species within leaf mixture treatments. Each sample was then chopped into ~0.5 x 0.5 cm pieces with sterile implements. Sub-samples were stored at -80°C for DNA extraction, enzyme analysis, and leaf tissue chemistry. Another sub-sample was dried at 60°C for 24 hours to determine percent moisture. Portions of this dried sample were then used to obtain total

C and N (as described above). Ash content was determined by combustion at 500°C for five hours.

Leaf tissue biochemical pools

Leaf tissue was fractionated into a variety of biochemical pools using a sequential extraction procedure modified from Blackwood et al. (2007). After each step, the sample was dried at 55°C and weighed to determine mass loss. Soluble polar compounds were extracted by three rounds of heating at 95°C for 10 minutes in a 50% methanol / 50%

H2O solution, followed by centrifugation at 10000 × g for 5 minutes and removal of the

154

supernatant. Non-polar compounds were extracted by three rounds of mixing with 100% dichloromethane (CH2Cl2) solution, followed by 10000 × g centrifugation for 5 minutes and removal of the supernatant. To remove acid-hydrolyzable structural carbohydrates, the sample was incubated in a 30°C water bath with 450 μL sulfuric acid for two hours, diluted with 13 mL distilled H2O, and then autoclaved for two hours at 121°C. The non- hydrolyzable tissue (i.e., lignin) was then collected on a glass microfiber filter (Grade

GF/A, Whatman©, Piscataway, NJ) using vacuum filtration, followed by drying and weighing. Non-hydrolyzable tissue was then combusted at 400°C for 5 hours to determine ash content. Lignin and cellulose values were used to calculate a lignin: cellulose index (LCI; defined as [lignin / (lignin + cellulose)], Mellilo et al. 1982, Taylor et al. 1989).

Bacterial and fungal biomass

Bacterial biomass was determined on leaf tissue preserved in phosphate-buffered saline (pH 7.2) and 10% formaldehyde. Samples were sonicated for 5 min (McNamara and Leff 2004), concentrated on 0.2-μm black polycarbonate filters, and stained for 3 min with 15 μg/ml 4′,6-diamidino-2-phenylindole (Porter and Feig 1980). Total number of cells was determined by counting under epifluorescent microscopy. Biomass was calculated by allometric equations (Loferer-Kröβbacher et al. 1998), after biovolumes of approximately 200 cells were calculated from digital images with CMIEAS/Image Tool

1.27 (Liu et al. 2001).

Fungal biomass was determined based on the extraction of ergosterol (Tank and

Webster 1998) from leaf samples preserved in 100% methanol and stored at 4°C.

155

Ergosterol content was quantified with HPLC (Alltech, IL) by comparing sample peaks with ergosterol standards (Sigma-Aldrich, Seels, Germany) at concentrations ranging from 0.25 - 20 mg/l. Fungal biomass on leaves was calculated based on an ergosterol content of 5.5 mg/g biomass (Gessner and Chauvet 1993).

Bacterial and fungal community composition

DNA from each leaf species within a single- or four-species litter mixture was extracted using a CTAB procedure as described in Wu et al. (2011). The primers

Eub338F-0-III (labeled with 6- carboxyfluorescein) and 1392R were used to amplify bacterial 16SrRNA genes, at an annealing temperature of 57°C with cycles adjusted to produce optimal bands without non-target amplification (Blackwood et al. 2005). The fungal ribosomal internal transcribed spacer (ITS) region was amplified with NSI1F

(labeled with 6-carboxyfluorescein) and NLB4R (Martin and Rygiewicz 2005) at an annealing temperature of 60°C with cycles adjusted to produce optimal bands without non-target amplification. Bacterial and fungal PCR product was digested with HaeIII prior to separation and detection of terminal restriction fragments as described previously

(Feinstein et al. 2009). TRFs 50-600 bp in size were used in statistical analyses if their fluorescence was greater 0.5% of the total fluorescence for the sample.

Extracellular enzyme analysis

Fluorescent 4-methylumbelliferone (MUB)-linked extracellular enzyme assays were conducted on 0.5 g leaf tissue homogenized with a 2-speed hand blender (Hamilton

Beach, Minneapolis, MN) in 125 ml sodium acetate buffer (pH = 5.0). The buffer/leaf

156

suspension was loaded into a 96-well black plate with negative control and quench wells as described in Saiya-Cork et al. (2002). Enzymes and substrates that were used were described in Table 1 (substrates purchased from Sigma-Aldrich, St. Louis, MO). Nagase and phosphatase assays were incubated for 30 min and all other enzymes were incubated

2 hrs (20°C), after which enzyme activity was halted with addition of 10 uL 0.5 N NaOH.

Fluorescence was then read on a Synergy HT Microplate Reader (BioTek, Winooski, VT) at 365 nm excitation and 450 nm emission. Potential laccase enzyme activity was measured using a procedure modified from Courty et al. (2005). Samples were homogenized in buffer as described above, loaded into clear plastic 96-well plates, and incubated for 1 hour with 50 μl of 2 mM ABTS (Table 1; Sigma-Aldrich) at conditions described above. Oxidation of ABTS was measured by absorbance 420 nm wavelength.

Statistical analysis

Analysis of variance (ANOVA) and regression analysis were conducted in the software R using the nlme library. We tested the impact of experimentally manipulated factors on several response variables (AFDM, fungal biomass, bacterial biomass, and log- transformed enzyme activity) with a mixed-model ANOVA design. Fixed effects included harvest (treated as a categorical variable), habitat (upland or riparian), litter diversity treatment (1-, 3- or 4-species), and leaf type (beech, hazel, maple, or oak).

Block (nested within habitat) and bag (nested within block) were treated as random effects related to spatial autocorrelation. To reduce the complexity of the models tested, we used a hierarchical model selection approach whereby first we tested the significance

157

of the main effects of experimental factors, and then we tested two-way interaction terms among the significant main effects.

We used a multiple regression approach to explore the impact of continuous variables on the response variables. Significant experimental factors found previously were also included in this analysis. Leaf biochemical tissue variables (microbial resources) included N content, C:N ratio, polar and non-polar extractives, cellulose, lignin, lignin:N ratio, and lignin:cellulose index (LCI). Type III mixed linear models in

R were used to determine which of the resource variables were significant predictors of fungal and bacterial biomass and log-transformed enzyme activities. For AFDM, resource variables were tested in one model, fungal and bacterial biomass were tested as predictors in a “biomass model”, and log-transformed enzyme activities were tested in an

“enzyme model”. As with the categorical variables, block (nested within habitat) and bag

(nested within block) were treated as random effects related to spatial autocorrelation.

After selection of significant continuous variables, we tested interactions between these variables and the significant experimental factors.

After selection of significant experimental factors, predictor continuous variables, and interactions, R2 values were calculated for individual significant terms by analysis without any other terms in the model. This is done for comparative purposes, and does not reflect the total variance explained by the model with all terms included.

To determine which experimental factors significantly affected fungal and bacterial community composition, the multivariate method redundancy analysis was conducted using the software package vegan in R version 2.0-2 (Oksanen et al. 2001).

158

TRFLP peak height relative abundances were Hellinger-transformed prior to analysis

(Legendre and Anderson 1999, Legendre and Gallagher 2001). The significance of experimental factors and their interactions was assessed by comparing the empirical variance explained by each factor to a distribution obtained by 999 random permutations.

Similar to the second type of statistical analysis described above, redundancy analysis was also used to examine the power of litter biochemical composition and N content in explaining community composition. Forward selection was used to avoid overestimating the amount of variance explained and minimize Type I errors (Blanchet et al. 2008). Mantel analysis (Sokal 1979) was conducted to compare log-transformed enzyme activities and Hellinger-transformed TRFLP to determine if enzyme activity is significantly correlated with the composition of microbial communities using the R

(version 2.0-2) software package vegan (Oksanen et al. 2001).

In order to determine if there was a significant difference in β-diversity between single- and mixed-species treatments within each block, the test for homogeneity of multivariate dispersion (Anderson et al. 2006) was conducted on Hellinger-transformed

TRFLP profiles. This test was performed independently for each block and a Z- transform test (Stouffer et al. 1949) was then run to determine if the independent P- values were collectively significant. The Z-transform test was selected because it is not subject to the asymmetry issues associated with Fisher’s combined probability test

(Whitlock 2005).

159

Results

Factors that impact litter decomposition (AFDM)

In the analysis of experimental factors, leaf type (R2 = 5.2%), harvest (R2 =

64.5%), and habitat (R2 = 2.2%) each had a significant impact (P < 0.001) on decomposition, but litter diversity treatment did not. The amount of variance due to spatial autocorrelation associated with block was 6.77% and bag was 11.4%. There was a significant habitat x harvest interaction (R2 = 69.0%) because decomposition occurred more rapidly in the upland than the riparian habitat on some harvest dates (Figure 1).

The increased rate of decomposition in upland compared to riparian habitat was driven by hazel and maple leaves, as indicated by a significant habitat x leaf type interaction (R2 =

7.9%) (Figure 1). Labile leaves decomposed faster than recalcitrant leaves, but this was especially evident in the upland habitat (Figure 1). Other interactions tested were not significant.

160

Figure 1

Significant habitat (Rip = riparian, Up= upland) x harvest (Day treated as a categorical variable) interaction for proportion AFDM remaining in beech, hazel, oak, and maple leaves. Values are averaged for all treatments.

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4 Oak Rip Beech Rip Oak Up Beech Up 0.2 Maple Rip Hazel Rip 0.2

Proportion AFDM remaining AFDM Proportion Maple Up Hazel Up remaining AFDM Proportion

0.0 0.0 51 126 212 287 385 51 126 212 287 385

Day Day

161

In the regression analysis, none of the biochemical variables (polar extractives, nonpolar extractives, cellulose, lignin, lci, leaf N, C:N) had a significant impact on

AFDM except for lci (P = 0.093, R2 = 7.1%) and the lignin:N (P = 0.008, R2 = 15.7%) ratio. LCI averaged across all leaf types and treatments increased from 0.406 at Day 51 litter collection to 0.589 at Day 385 litter collection, increasing at every time point except the third one (Day 212). In the resource pool model, 11.90% of the variance was due to spatial autocorrelation associated with block.

There was a significant harvest x lci interaction (P = 0.023, R2 = 62.4%), but it was only significant at harvest 2 (Day 121) and the variance explained by this interaction was less than the variance explained by harvest alone (R2 = 64.5%). There was also a significant leaf type x lignin:N interaction (P = 0.016, R2 = 20.4%) that was significant for beech, hazel, and oak, but not maple leaves (Figure 2). The trend of the significant leaf scatter plots was lower lignin:N associated with lower %AFDM remaining.

162

Figure 2

Significant AFDM leaf type (beech, hazel, oak, and maple) x lignin:N ratio shown at each harvest date (Day treated as a categorical variable). Values are averaged for all treatments.

Beech Hazel Oak 40 Maple

Lignin: N ratio Lignin:

10 initial 51 126 212 287 385

Day

163

In the biomass regression model, fungal (P = 0.0025, R2 = 15.5%) and bacterial biomass (P < 0.0001, R2 = 30.9%) were both significant predictors of proportion AFDM remaining. When running fungal and bacterial biomass as predictors of AFDM, 9.8% of the variance was due to spatial autocorrelation associated with block. Although biomass increased as decomposition proceeded, there was a significant leaf type x biomass interaction for both fungi and bacteria (Figure 3). Fungal biomass was highest on beech leaves (average value across all beech leaves at all time points = 17.96 mg/g AFDM), followed by maple (17.32 g/mg AFDM), hazel (16.57 g/mg AFDM), and oak (14.93 g/mg AFDM). Oak also had the smallest bacterial biomass (11.83 g/mg AFDM), followed by beech (12.41 g/mg AFDM), hazel (12.52 g/mg AFDM), and maple (13.87 g/mg AFDM).

164

Figure 3

Significant AFDM leaf type (beech, hazel, oak, and maple) x microbial biomass interaction shown with the proportion of AFDM remaining at each of the five sample collection times. Values are averaged by harvest for all treatments.

Fungal biomass Bacterial biomass

40 40 Beech 35 Hazel Oak 30 Maple 30

20 20

10 10

Fungal Biomass (mg/g AFDM) (mg/g Biomass Fungal

5 Bacterial biomass (mg /gAFDM)

0 0 1.0 0.8 0.6 0.4 0.2 1.0 0.8 0.6 0.4 0.2 Proportion AFDM remaining Proportion AFDM remaining

165

In the enzyme regression model, all of the extracellular enzymes were significant predictors of litter decomposition except for cellulose and phosphatase (Table 2).

Laccase and Aglu explained the most variance in AFDM (Table 2). In the enzyme analysis, 12.7% of the variance was due to spatial autocorrelation associated with block.

Table 2: The impact of enzyme activity on litter decomposition (% AFDM remaining)

Enzyme P R2

Bglu 0.087 0.009

Phos NS NS

Nag 0.047 0.002

Aglu 0.003 0.113

Cell NS NS

Bxyl 0.037 0.019

Lac 0.005 0.171

None of the enzyme activities had significant interactions with significant experimental factors when AFDM was a response variable with the exception of a laccase x leaf type interaction. Laccase activity was highest on hazel leaves (1.61 log- transformed nmol/g AFDM), followed by oak (1.49), maple (1.48), and beech (1.43).

166

Mantel analysis indicates that fungal community composition was significantly correlated with AFDM (P < 0.001, r = 0.173), but bacterial community composition was not.

Factors that impact microbial biomass

Harvest had a significant impact (P < 0.001) on both fungal and bacterial biomass, which increased during the course of the year-long experiment. Neither habitat nor leaf type were significant, but the litter diversity treatment significantly affected both fungal and bacterial biomass, which was increased in 3- and 4-species leaf mixtures (P <

0.001, Figures 4 and 5). There was also a significant harvest x treatment interaction for bacterial biomass, with significant differences in biomass found on single- versus mixed- litter treatments at four of the five harvest dates (collection days 126, 212, 287, and 385).

None of the plant biochemical pool variables had a significant impact on fungal biomass (P > 0.05). However, leaf N (P = 0.027), C:N ratio (P = 0.009), polar extractives (P = 0.045) and lignin (P = 0.039) all had significant impacts on bacterial biomass. There were also significant interactions affecting bacterial biomass, including

C:N ratio x diversity treatment (P = 0.056) and leaf N x diversity treatment (P = 0.037) interactions. In both of these interactions, the C:N ratio and leaf N values remained similar in mixed- versus single-litter treatments while bacterial biomass was higher in mixed- than in single-species litter treatments

167

Figure 4

Fungal biomass at each harvest for each of the litter diversity treatments. An asterisk (*) is placed above harvest with a significant treatment x harvest interaction.

* Fungal Biomass

30 1 spe trt 3 spe trt 4 spe trt 20 mg/g AFDM mg/g *

51 126 212 287 385

Day

168

Figure 5

Bacterial biomass at each harvest for each of the litter diversity treatments. An asterisk

(*) is placed above harvest with a significant treatment x harvest interaction.

* Bacterial Biomass 50

30 * 1 spe trt 3 spe trt

mg/g AFDM mg/g 20 4 spe trt *

10 *

0 51 126 212 287 385

Day

169

Factors that impact extracellular enzyme activity

The impact of the experimental main effects on enzyme activity may be found in

Table 3. Harvest had a significant impact on the activity of every enzyme, demonstrating the importance of seasonal effects as well as the amount of time litter has been incubating in the field. Enzyme activity was highest for most enzymes at the day 212 collection (in

May 2008), except for laccase and nagase (day 385, Dec 2008). Habitat and leaf type were also shown to be important; these factors each affected the activity of 5 out of the 7 enzymes assayed. When habitat was significant there was higher enzyme activity in the upland for phosphatase, alpha-glucosidase, and beta-xylosidase, and higher in the riparian for nagase and cellobiohydrolase. Enzyme activity was not consistently higher on either recalcitrant or labile leaves.

Table 3

The impact of main effects on enzyme activity.

Enzyme Harvest Habitat Treatment Leaf Type Sig.Int. P

Bglu 0.005 0.051

Phos < 0.001 0.059 0.0611 < 0.001 Trt x Type < 0.001

Nag 0.001 0.027 < 0.001 Har x Type 0.047

Aglu < 0.001 0.007 < 0.001

Cell < 0.001 0.008 < 0.001

Bxyl 0.013 0.056 Har x Hab 0.022

Lac < 0.001

170

The litter diversity treatment had a marginally significant impact on phosphatase activity (P = 0.06) with higher activity (8.13 nmol/g AFDM/hr) in single-species treatments than in mixed (7.99 nmol/g AFDM/hr). A significant (P < 0.001) treatment x leaf type interaction for phosphatase activity showed significantly higher activity in single- than in mixed-treatments for all leaf types: hazel (8.18 single, 8.06 mixed), maple

(8.39 single, 8.13 mixed), and oak (8.28 single, 8.11 mixed) except beech (which was virtually the same in both treatments).

There was a significant (P = 0.047) harvest x leaf type interaction for nagase

(Figure 6). During the first two harvests, nagase activity was higher on the recalcitrant leaves (beech and oak) than it was on the labile leaves. Nagase activity remained higher on beech leaves than other leaf types on the third harvest, but by the final harvest nagase activity for all leaf types was virtually identical.

171

Figure 6

Significant nagase harvest x leaf type interaction. The activity for recalcitrant leaves is shown with black symbols and labile leaves with white symbols. Enzyme activities are in nmol/g/AFDM/hr and have been log-transformed. An asterisk (*) is placed above harvest with a significant treatment x harvest interaction.

8.2 *

8.0 *

7.8 * 7.6

7.4

7.2

log (enzyme activity) (enzyme log 7.0 Beech Hazel 6.8 Oak Maple 6.6 51 126 212 287 385

Day

172

A significant (P = 0.022) harvest x habitat interaction for beta-xylosidase activity indicated higher activity in the floodplain during every harvest except the first (Day 51), with significant interactions during the Day 51, 287, 385 harvests.

In the regression analysis of plant biochemical pools as explanatory variables, nitrogen did not have a major impact on most enzyme activities (Table 4). The C:N ratio was not significant for any enzyme activity. Leaf N was a significant predictor for nagase (P = 0.037), cellobiohydrolase (P = 0.040), and laccase (P = 0.024) activity.

Lignin:N ratio was a significant predictor for phosphatase (P = 0.036), nagase (P =

0.012), and alpha-glucosidase (P = 0.089) activity. There were no significant interactions between any significant experimental factors and resource variables.

Table 4

The impact of resource variables on enzyme activity.

Enzyme C:N AFN LCI Lignin:N polar cellulose lignin beta-glucosidase 0.053 0.091 0.015

Phosphatase 0.036

Nagase <0.001 0.012 0.019 0.005

Alpha-glucosidase 0.020 0.089 0.012

Cellobiohydrolase 0.040

Beta-xylosidase 0.073 0.052 0.086

Laccase 0.024 0.043 0.034

173

In a biomass regression model, fungal biomass did not have a significant impact on any enzyme activity. Bacterial biomass had a significant impact on phosphatase (P =

0.068), beta-glucosidase (P = 0.082), cellobiohydrolase (P = 0.051), and laccase (P =

0.023).

Despite the lack of significance of fungal biomass influencing enzyme activity, there was a distinct importance of fungal community composition influencing enzyme activity. When a Mantel test was used to test all enzymes together as a response to

TRFLP profiles (community composition), bacterial community composition was not significant (P > 0.1) but fungal community composition was (Mantel statistic r = 0.158, P

= 0.001). Similarly, fungal community composition was significantly correlated with the individual enzymes nagase (r = 0.044, P = 0.031), phosphatase (r = 0.106, P = 0.001), alpha-glucosidase (r = 0.155, P = 0.001), cellobiohydrolase (r = 0.121, P = 0.001), and laccase (r = 0.002, P = 0.068), while bacterial community composition was not correlated with the activity of any enzyme.

Factors that impact microbial community composition

Fungal community composition was impacted by habitat (P = 0.005, Adj R2 =

0.014), harvest (P = 0.005, Adj R2 = 0.075), and leaf type (P = 0.005, Adj R2 = 0.006), but not litter diversity treatment. A redundancy analysis ordination chart comparing upland and riparian fungal community composition at each harvest (Figure 7) shows a separation between harvest 1 and all other harvests, as well as habitat sorting during each harvest. Bacterial community composition was impacted by harvest (P = 0.005, Adj R2 =

0.009), but not habitat, leaf type, or litter diversity treatment.

174

Figure 7

Ordination chart showing fungal community composition averaged across all leaf types and litter mixture treatments for riparian and upland communities at each of the five collection times (harvests). Variance explained for RDA 1 axis: 8.2%, RDA 2 axis: 2.9%.

175

Using plant biochemical resources as predictors, fungal community composition was significantly impacted by C:N ratio (P = 0.001, Adj R2 = 0.068), polar extractives (P

= 0.001, Adj R2 = 0.009), and nonpolar extractives (P = 0.028, Adj R2 = 0.008).

Bacterial community composition was significantly impacted only by C:N ratio (P =

0.010, Adj R2 = 0.012).

The Stouffer’s test of independent P-values, aggregating the results of independent Anderson dispersion tests for all forty blocks, was significant for both bacteria (P < 0.0001) and fungi (P = 0.0017). In both cases, β-diversity was higher

(dispersion among the four leaf types within a block was greater) in the 4-species mixture treatment than in the single-species treatment.

Discussion

Similar to approximately 30% of litter decomposition studies examining the effects of litter diversity (reviewed in Gartner and Cardon 2004, Hӓttenschwiler et al.

2005), this study found that the decomposition rate of leaf litter in mixtures was not significantly different from the rate when leaves were not in a mixture. However, litter mixtures were shown to play an important role in stimulating increased bacterial and fungal beta diversity and productivity (biomass), with important implications for ecosystem-scale energy cycling. This research also demonstrated the important role of fungal community composition in leaf litter carbon-cycling activity, specifically extracellular enzyme activity.

176

Factors that impact AFDM

The higher rates of decomposition associated with labile leaves was expected based on previous studies (Aber et al. 1990, Berg and McClaugherty 2008, Cornwell et al. 2008). It was somewhat unexpected to find higher rates of decomposition occurring in the upland than the riparian habitats because the upland is generally drier, which has been associated with slowed decomposition (Giardina and Ryan 2000, Bond-Lamberty et al.

2002). Unexpected results noted in numerous litter decomposition studies have been attributed to the often un-quantified contribution of invertebrates such as earthworms

(Vos et al. 2011). We did find evidence during litter collection of earthworm activity on litter bags in the upland habitat, but not in the riparian. This may be contributing to the greater % mass loss we found in the upland habitat.

In this study, our findings agree with other studies that have found significant effects of LCI (Osono and Takeda 2005, Vivanco and Austin 2008, Hӓttenschwiler et al.

2011) and lignin:N ratio (Melillo et al. 1982, Cornwell et al. 2008, Talbot and Treseder

2012) on litter decomposition rates. Lignin has been shown to influence decay rates of many other litter carbon compounds (Parton et al. 1987, Moorhead and Sinsabaugh 2006,

Berg and McClaugherty 2008), likely due to its protection of energy-rich compounds from microbial enzymatic attack (Cooke and Whipps 1993). Recent work has shown that once lignin decomposition begins, cellulose is utilized as a labile substrate that fuels lignin degradation (Talbot and Treseder 2012). The LCI value has been proposed to be a primary control on lignin decomposition, with lignin decomposition being activated when

LCI reaches ~0.4 and above (Herman et al. 2008). In this study the LCI averaged across

177

all leaf types, litter treatments, and habitats was at a value of 0.41 at the first litter collection point (Day 51) and increased during the year. As expected based on the literature, the amount of lignin degradation was related to the LCI; lignin was approximately 1.5 g per bag in the first two harvests when LCI was right around 0.40, by day 212 lignin decreased to 1.1 g (LCI increased to 0.47), day 287 (lignin .84 g, LCI

0.52), and day 385 (0.59 g lignin, 0.59 LCI).

Nitrogen often becomes limited during litter decomposition (Berg and Staaf

1980), in part because most enzymes that need to be synthesized for microbial decomposition are rich in N (Sinsabaugh et al. 1993). One mechanism to acquire N is the production of N-acquiring nagase; the significantly higher nagase levels that we found on recalcitrant leaves compared to labile leaves from Days 51-212 may be a microbial response to N-limitation. However, the N-associated variables that were found on recalcitrant versus labile leaves were not significantly different so it is difficult to determine to what degree N may have been limited.

The impact of litter diversity on microbial biomass

In this study, fungal and bacterial biomass levels were significantly higher in mixed- than in single-species litter treatments. An increase in microbial biomass associated with litter mixtures has also been documented for Arizona aspen, fir, and pine litter (Chapman and Newman 2010). The results of our study did not support the resource complementarity hypothesis because it predicts that a more complete utilization of heterogeneous resources stimulates increased microbial biomass. Instead, we found that although biomass was higher in mixed-treatments, similar amounts of carbon were

178

degraded in both treatments. This indicates that microbes living within heterogeneous litter mixtures are able to process carbon more efficiently than those living within homogeneous single-species litter. The implication is that a greater proportion of the carbon metabolized is converted into microbial biomass in mixed litter, while more carbon is lost to respiration in single-species litter.

These results imply that there is an important ecosystem-level impact of leaf litter diversity. Energy that is retained within a system and can do work is defined as exergy

(Evans 1966). Exergy has been proposed to be a metric to measure system efficiency

(Jørgensen et al. 2007) because systems that possess higher exergy have been shown to have increased energy efficiency and complexity (Schneider and Kay 1994, Fath et al.

2004). The carbon that is more efficiently processed in litter mixtures results in less carbon being respired and lost from the forest ecosystem, and more incorporated into biomass to be transferred among trophic levels. Microbial activity has been shown to increase ecosystem exergy during the summer season in a coastal lagoon (Pusceddu and

Danovaro 2009) but this study provides the first documented evidence of microbial- driven exergy associated with litter decomposition and illustrates the critical role that leaf litter diversity and microbes play in sustaining energy within an ecosystem.

Factors that impact enzyme activity

The impact of forest habitat on enzyme activity has been shown to be linked to differences in decomposer communities. Waldrop and Zak (2006) proposed that decomposer communities are affected by forest habitat types through the influence of habitat-specific dominant leaf types acting as environmental filters for community

179

composition. In this study we documented a strong impact of forest habitat, but the exact same leaf types were incubated in the different habitats, so the leaf types could not act as differential community filters correlated with forest habitat type. Our statistical tests demonstrated that enzyme activity is correlated with fungal community composition and that fungal community composition is different in the upland versus riparian habitats

(Figure 6). These results lend support to a microbial community influence on enzyme activity that is in addition to the effects of leaf type.

Leaf type significantly affected the majority of the enzymes we measured. There is extensive previous documentation of differential expression of enzyme activity being linked to leaf type (particularly recalcitrant versus labile species) (Sinsabaugh et al. 1991,

Carrerio et al. 2000, Sinsabaugh et al. 2002, Allison and Vitousek 2004, Voriskova et al.

2011). Previous studies quantified enzyme activity differences in single-species pools rather than mixtures; we designed this study to compare dynamics of single-species versus mixed-species litter treatments. All of the enzyme activities that we quantified

(except phosphatase) were not significantly impacted by litter-mixture treatments.

Hence, enzyme activity is not subject to influence from surrounding substrate, but is closely linked to the local resource pool represented by leaf types.

While harvest significantly impacted the activity of every enzyme, this effect is most likely driven by the changing microbial community and proportions of leaf nutrients associated with leaves over time. Extracellular enzyme activity has been shown to be correlated with shifting proportions of substrate compounds as decomposition proceeds

(Steffen et al. 2007, Valaskova et al. 2007, Snajdr et al. 2011). In this study, beta-

180

glucosidase was significantly impacted by LCI, polar extractable compounds, and lignin.

Beta-glucosidase releases glucose from cellulose degradation products (Sinsabaugh

2005). The activity of beta-glucosidase has been shown to be stimulated by compounds found within the polar extractable pool (Morikawa et al. 1995). In this study, beta- xylosidase was correlated with the same resource pool variables as beta-glucosidase.

This is to be expected as polar extractable compounds increase microbial activity as well as the expression of cellulose-degrading enzymes (Sinsabaugh 1994).

Laccase activity was impacted by lignin, leaf N, and the polar extractable pool.

Laccase activity has been noted to degrade tannins which are found in the pool of polar extractable compounds (Carbajo et al. 2002), and has been shown to be related to lignin levels (Scheel et al. 2000). N is in high demand during litter decomposition because nitrogent is low in structural carbon compounds (Carriero et al. 2000). The link between leaf N and laccase is demonstrated in studies that have shown that some fungi will activate lignin degrading enzymes only when N becomes limited (Keyser et al. 1978,

Kirk 1987), suggesting that lignin degradation is switched on as an N-acquiring response because lignin may surround and protect sources of N (Berg and McClaughtery 2008).

Nagase activity was impacted by lignin, leaf N, lignin:N ratio, and the polar extractable pool. Nagase is produced in large part to acquire N (Fernandez & Koide 2012) and so it is logical that nagase activity was correlated with most of the N-related resource variables in this study (as well as lignin due to the link between N and lignin degradation as discussed above).

Microbial community composition

181

This study found evidence that the composition of saprotrophic microbial communities is related in part to the functional activity the communities are engaged in.

While bacteria certainly have the ability to produce extracellular enzymes (Medie et al.

2012, Fontes and Gilbert 2010), fungi are thought to produce a wider range of enzymes than bacteria (Kirk 1987) and therefore be the major enzyme producers in leaf litter. We found support for the importance of fungi controlling enzyme activity because fungal community composition was significantly correlated with enzyme activity, while bacterial community composition was not. Because fungal community composition was shown to change during decomposition and be correlated with enzyme activity and

AFDM, this provides support for the Guild Decomposition Model. It suggests that the proportions of guilds within communities change over time enough to shift entire community composition profiles.

The importance of carbon and N resources in shaping fungal communities was evident as the C:N ratio explained more variance than any other factor except for harvest.

C:N ratio also explained more variance in bacterial community composition than any other factor. While we did not find microbial community composition to be significantly impacted by any specific carbon pool, this dynamic has also been reported for wheat root and shoot decomposition (Baumann et al. 2011). In our study, leaf type did not significantly impact bacterial or fungal community composition, although other studies have shown composition impacted by leaf type for bacteria (Dilly et al. 2004, Liu et al.

2008), fungi (Eskalinen et al. 2009, Zabed Hossain et al. 2010), or both groups (Knapp et al. 2011). However, none of the litter in the above studies was senesced deciduous forest

182

leaf litter, suggesting that perhaps this litter contains a more complex mixture of compounds than agricultural, grassland, or pine litter. The more complex litter in this study compared to the studies mentioned above may be increasing the relative importance of fungal community composition, as fungi may be needed to break down forest leaf litter than agricultural or grassland litter.

We found significantly greater fungal and bacterial β-diversity in mixed- versus single-species treatments. The partitioning of litter carbon resources has previously been shown to be common among saprotrophic fungi and proposed to be a mechanism associated with increased diversity (Kjøller and Struwe 2002, Hanson et al. 2008). If certain taxa are specialized to dominate specific resource pools, mixed litter may provide conditions that would stimulate an increase in competition via a more diverse resource pool as compared to single-species litter treatments. This could be a factor that would increase beta diversity in moxed litter compared to non-mixed.

Conclusions

This study highlights the influence of litter diversity on several aspects of microbial and ecosystem ecology, and the importance of microbial community composition on functional activity associated with litter decomposition. Litter diversity was shown to be associated with an increase in β-diversity and productivity (biomass) for both fungi and bacteria. Carbon that is retained in the ecosystem in the form of increased microbial biomass provides empirical evidence that microbial processing of mixed litter increases ecosystem exergy. The regulation of litter decomposition remains an area of active investigation and is likely related to complex interactions between abiotic,

183

resource, microbial (microbial community composition and enzyme activity), and other factors this study did not quantify such as the activity of litter-decomposing invertebrates.

In order to gain a comprehensive understanding of the regulation of litter decomposition, path analysis and structural equation modeling in a study that quantifies as many factors as possible is something to plan for future work.

References

Aber JD, Melillo JM. 1982. Nitrogen immobilization in decaying hardwood leaf litter as

a function of initial nitrogen and lignin content. Canadian Journal of Botany 60:

2263–2269.

Aber JD, Melillo JM, McClaugherty CA. 1990. Predicting long-term patterns of mass

loss, nitrogen dynamics, and soil organic matter formation from initial fine litter

chemistry in temperate forest ecosystems. Canadian Journal of Botany 68: 2201–

2208.

Aerts R. 1997. Climate, Leaf Litter Chemistry and Leaf Litter Decomposition in

Terrestrial Ecosystems: A Triangular Relationship. Oikos 79: 439-449.

Allison SD, Vitousek PM. 2004. Extracellular enzyme activities and carbon chemistry as

drivers of tropical plant litter decomposition. Biotropica 36: 285-296.

Allison SD, LeBauer DS, Ofrecio MR, Reyes R, Ta A-M, Tran TM. 2009. Low levels of

nitrogen addition stimulate decomposition by boreal forest fungi. Soil Biology &

Biochemistry 41: 293-302.

184

Anderson MJ, Ellingsen KE, McArdle BH. 2006. Multivariate dispersion as a measure of

beta diversity. Ecology Letters 9: 683-693.

Ball BA, Hunter MD, Kominoski JS, Swan CM, Bradford MA. 2008. Consequences of

non-random species loss for decomposition dynamics: experimental evidence for

additive and non-additive effects. Journal of Ecology 96:303–313.

Bardgett RD, Shine A. 1999. Linkages between plant litter diversity, soil microbial

biomass and ecosystem function in temperate grasslands. Soil Biology &

Biochemistry 31:317–321.

Baumann K, Marchner S, Kuhn TK, Smernik RJ, Baldock JA. 2011. Microbial

community structure and residue chemistry during decomposition of shoots and

roots of young and mature wheat (Triticum aestivum L.) in sand. European

Journal of Soil Science 62: 666-675.

Berg B, Staaf H. 1980. Decomposition rate and chemical changes of Scots pine needle

litter. II. Influence of chemical composition. Ecological Bulletins (Stockholm) 32:

373–390.

Berg B, McClaugherty C. 2008. Plant litter: Decomposition, humus formation, carbon

sequestration. Springer-Verlag Berlin, Heidelberg, New York. 286 p.

Blackwood CB, Oaks A, Buyer JS. 2005. Phylum- and class specific PCR primers for

general microbial community analysis. Applied and Environmental Microbiology

71: 6193–6198.

Blanchette RA, 1991. Delignification by wood-decay fungi. Annual Review of

Phytopathology 29: 381–398.

185

Blanchet FG, Legendre P, Borcard D. 2008. Forward selection of explanatory variables.

Ecology 89: 2623-2632.

Boberg J, Finlay RD, Stenlid J, Nӓsholm T, Lindahl BD, 2008. Glucose and ammonium

additions affect needle decomposition and carbon allocation by the litter

degrading fungus Mycena epipterygia. Soil Biology & Biochemistry 40: 995–999.

Bond-Lamberty B, Wang C, Gower ST. 2002. Annual carbon flux from woody debris for

a boreal black spruce fire chronosequence, Journal of Geophysical Research 107:

2002-2012.

Bray R, Gorham E. 1964. Litter production in forests of the world. Advances in

Ecological Research 2: 101-157.

Burns RG. 1982. Enzyme-activity in solid location and a possible role in microbial

ecology. Soil Biology & Biochemistry 14: 423–427.

Burns RG, Dick RP. 2002. Enzymes in the Environment: Activity, Ecology, and

Applications. Marcel Dekker, New York. 640 p.

Carbajo JM, Junca H, Terron MC. 2002. Tannic acid induces transcription of laccase

gene in the white-rot fungus Coriolopsis gallica. Canadian Journal of Micrbiology

48: 1041-1047.

Carreiro MM, Sinsabaugh RL, Repert DA, Parkhurst DF. 2000. Microbial Enzyme Shifts

Explain Litter Decay Responses to Simulated Nitrogen Deposition. Ecology 81:

2359-2365.

Chapin FS III, Matson PA, Mooney HA. 2002. Principles of terrestrial ecosystem

ecology. Springer, New York. 472 p.

186

Chapman SK, Koch GW. 2007. What type of diversity yields synergy during mixed litter

decomposition in a natural forest ecosystem? Plant and Soil 299: 153–160.

Chapman SK, Newman GS. 2010. Biodiversity at the plant-soil interface: microbial

abundance and community structure respond to litter mixing. Oecologia 162: 763-

769.

Cooke RC, Whipps JM. 1993. Ecophysiology of fungi. Blackwell, Oxford. 352 p.

Cornwell WK, Cornelissen JHC, Amatangelo K, Dorrepaal E, Eviner VT, et al. 2008.

Plant species traits are the predominant control on litter decomposition rates

within biomes worldwide. Ecology Letters 11: 1065–1071.

Courty P, Pritsch K, Schloter M, Hartmann A, Garbaye J. 2005. Activity profiling of

ectomycorrhiza communities in two forest soils using multiple enzymatic tests.

New Phytologist 167:309-319. de Boer W, Folman LB, Summerbell RC, Boddy L. 2005. Living in a fungal world:

impact of fungi on soil bacterial niche development. FEMS Microbiology

Reviews 29: 795–811.

Dilly O, Bloem J, Vos A, Munch JC. 2004. Bacterial diversity in agricultural soils during

litter decomposition. Applied and Environmental Microbiology 70: 468–474.

Entry JA. 2000. Influence of nitrogen on cellulose and lignin mineralization in

blackwater and redwater forested wetland soils. Biology and Fertility of Soils 31:

436–440.

187

Eskelinen A, Stark S, Männistö M. 2009. Links between plant community composition,

soil organic matter quality and microbial communities in contrasting tundra

habitats. Oecologia 161:113–123.

Feinstein LM, Sul WJ, Blackwood CB. 2009. Assessment of bias associated with

incomplete extraction of microbial DNA from soil. Applied and Environmental

Microbiology 75: 5428–5433.

Fernandez CW, Koide RT. 2012. The role of chitin in the decomposition of

ectomychorrhizal fungal litter. Ecology 93: 24-28.

Fog K. 1988. The effect of added nitrogen on the rate of decomposition of organic matter.

Biological Reviews 63: 433–462.

Fontes CMGA, Gilbert HJ. 2010. Cellulosomes: highly efficient nanomachines designed

to deconstruct plant cell wall complex carbohydrates. Annual Review of

Biochemistry 79: 655-681.

Frey SD, Knorr M, Parrent JL, Simpson RT. 2004. Chronic nitrogen enrichment affects

the structure and function of the soil microbial community in temperate hardwood

and pine forests. Forest Ecology and Management 196: 159–171.

Gartner TB, Cardon ZG. 2004. Decomposition dynamics in mixed-species leaf litter,

Oikos 104: 230–246.

Gessner MO, Chauvet E. 1993. Ergosterol-to-biomass conversion factors for aquatic

hyphomycetes. Applied and Environmental Microbiology 59: 502–507.

188

Gessner MO, Swan CM, Dang CK, McKie BG, Bardgett RD, Wall DH, Hӓettenschwiler,

S. 2010. Diversity meets decomposition. Trends in Ecology & Evolution. 25: 372-

380.

Giardina CP, Ryan MG. 2000. Evidence that decomposition rates of organic carbon in

mineral soil do not vary with temperature, Nature 404: 858-861.

Hӓttenschwiler S, Tiunov AV, Scheu S. 2005. Biodiversity and litter decomposition in

terrestrial ecosystems, Annual Review of Ecology, Evolution, and Systematics.

36: 191–218.

Hӓttenschwiler S. 2005. Effects of tree species diversity on litter quality and

decomposition. In Forest Diversity and Function: Temperate and Boreal Systems,

ed. M Scherer- Lorenzen, C K¨orner, E-D Schulze, 176:149–64. Heidelberg:

Springer-Verlag.

Hӓttenschwiler S, Coq S, Barantal S, Handa IT. 2011. Leaf traits and decomposition in

tropical rainforests: revisiting some commonly held views and towards a new

hypothesis. New Phytologist 189(4): 950-965.

Hӓttenschwiler S, Fromin N, Barantal S. 2011. Functional diversity of terrestrial

microbial decomposers and their substrates. Compter Rendus Biologies 334: 393-

402.

Hanson CA, Allison SD, Bradford MA, Wallenstein MD, Treseder KK. 2008. Fungal

taxa target different carbon sources in forest soil. Ecosystems 11: 1157–1167.

Hobbie SE. 2000. Interactions between litter lignin and soil nitrogen availability during

leaf litter decomposition in a Hawaiian montane forest. Ecosystems 3:484–94.

189

Högberg MN, Högberg P, Myrold DD. 2007. Is microbial community composition in

boreal forest soils determined by pH, C-to-N ratio, the trees, or all three?

Oecologia 150: 590–601.

Hooper DU, Vitousek PM. 1998. Effects of plant composition and diversity on nutrient

cycling. Ecological Monographs 68: 121–149.

Hutchinson GE. 1959. Homage to Santa Rosalia or why are there so many kinds of

animals. American Naturalist 93: 145–159.

Johnson D, Leake JR, Lee JA, Campbell CD. 1998. Changes in soil microbial biomass

and microbial activities in response to 7 years simulated pollutant nitrogen

deposition on a heathland and two grasslands. Environmental Pollution 103: 239–

50.

Jonsson M, Wardle DA. 2008. Context dependency of litter-mixing effects on

decomposition and nutrient release across a long-term chronosequence. Oikos

117: 1674–1682.

Jørgensen SE, Nielsen SN, Meher H. 1995. Emergy, environ, exergy, and ecological

modeling, Ecological Modeling 77: 99-109.

Keyser P, Kirk TK, Zeikus JG. 1978. Ligninolytic enzyme system of Phanerochaete

chrysosporium: synthesized in the absence of lignin in response to nitrogen

starvation. Journal of Bacteriology 135: 790-797.

Kirk TK. 1987. Enzymatic “combustion”: the microbial degradation of lignin. Annual

Review of Microbiology 41: 465–505.

190

Kjøller AH, Struwe S. 2002. Fungal communities, succession, enzymes, and

decomposition. In: Burns, R.G., Dick, R.P. (Eds.), Enzymes in the Environment.

Marcel Dekker, New York, pp. 267–284.

Knapp BA, Rief A, Seeber J. 2011. Microbial communities on litter of managed and

abandoned alpine pastureland. Biology and Fertility of Soils 47: 845-851.

Kjoller A, Struwe S. 2002. Fungal communities, succession, enzymes, and

decomposition. In: Burns RG, Dick RP (eds) Enzymes in the environment:

activity, ecology and applications. Marcel Dekker, New York, pp 267–284.

Lavelle P, Spain AV. 2001. Soil ecology. Kluwer, London. 264 p.

Legendre P, Anderson MJ. 1999. Distance-based redundancy analysis: testing

multispecies responses in multifactorial ecological experiments. Ecological

Monographs 69: 1–24.

Legendre P, Gallagher ED. 2001. Ecologically meaningful transformations for ordination

of species data. Oecologia 129: 271–280.

Liu Z, G Liu, B Fu, Zheng X. 2008. Relationship between plant species diversity and soil

microbial functional diversity along a longitudinal gradient in temperate

grasslands of Hulunbeir, Inner Mongolia, China. Ecological Research 23: 511–

518.

Liu J, Dazzo FB, Glagoleva O, Yu B, Jain AK. 2001. CMEIAS: a computer-aided system

for the image analysis of bacterial morphotypes in microbial communities.

Microbial Ecology 41: 173–194.

191

Loferer-Kröβbacher M, Kilma J, Psenner R. 1998. Determination of bacterial cell dry

mass by transmission electron microscopy and densitometric image analysis.

Applied and Environmental Microbiology 64: 688–694.

Loreau M. 2001. Microbial diversity, producer-decomposer interactions andecosystem

processes: a theoretical model. Proceedings of the Royal Society of London Ser.

B. 268: 303–309.

Loreau M, Hector A. 2001. Partitioning selection and complementarity in biodiversity

experiments. Nature 412: 72–76.

MacArthur RH, MacArthur JW. 1961. On bird species diversity. Ecology 42: 594–598.

Martin KJ, Rygiewicz PT. 2005. Fungal-specific PCR primers developed for analysis of

the ITS region of environmental samples. BMC Microbiology 5: 28.

Marquard E, Weigelt A, Temperton VM, Roscher C, Schumacher J, Buchmann N,

Fischer M, Weisser WW. Schmid B. 2009. Plant species richness and functional

composition drive overyielding in a six-year grassland experiment. Ecology 90:

3290–3302.

McNamara C, Leff L. 2004. Bacterial community composition in biofilms on

decomposing leaves in a Northeast Ohio stream. Journal of North American

Benthological Society 23: 677–685.

Medie FM, Davies GJ, Drancourt M, Henrissat B. 2012. Genome analyses highlight the

different biological roles of cellulases. Natural Reviews in Microbiology 10: 227-

234.

192

Melillo JM, Aber JD, Muratore JF. 1982. Nitrogen and lignin control of hardwood leaf

litter decomposition dynamics. Ecology 63: 621–626.

Mittelbach GG, Steiner CF, Scheiner SM, Gross KL, Reynolds HL, Waide RB, Willig

MR, Dodson SI, Gough L. 2001. What is the observed relationship between

species richness and productivity? Ecology 82: 2381-2396.

Moorhead D, Sinsabaugh R. 2006. A theoretical model of litter decay and microbial

interaction. Ecological Monographs 76(2): 151-174.

Morikawa Y, Mantani O, Ohashi T. 1995. Cellulase induction by Trichoderma reesei PC-

3-7. Applied Microbiology and Biotechnology 44: 106-111.

Morin PJ. 2000. The complexity of co-dependency. Nature 403: 718-719.

Nannipieri P, Kandeler E, Ruggiero P. 2002. Enzyme activities and microbiological and

biochemical processes in soil. In: Burns, R.G., Dick, R.P. (Eds.), Enzymes in the

Environment. Marcel Dekker, Inc., New York, pp. 1–34.

Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara RB, Simpson GL,

Solymos P, Stevens HH, Wagner H. 2011. Package ‘vegan’ The community

ecology package, http://vegan.r-forge.r-project.org/

Olson JS. 1963. Energy stores and the balance of producers and decomposers in

ecological systems. Ecology 44: 322–331.

Osono T, Takeda H. 2005. Decomposition of organic chemical components in relation to

nitrogen dynamics in leaf litter of 14 tree species in a cool temperate forest.

Ecological Research 20: 41–49.

193

PartonWJ, Schimel DS, Cole CV, Ojima DS. 1987. Analysis of factors controlling soil

organic-matter levels in great-plains grasslands. Soil Science Society of America

Journal 51, 1173–1179.

Pérez Harguindeguy N, Blundo CM, Gurvich DE, Díaz S, Cuevas E. 2008. More than the

sum of its parts? Assessing litter heterogeneity effects on the decomposition of

litter mixtures through leaf chemistry. Plant and Soil 303: 151–159.

Porter KS, Feig YS. 1980. The use of DAPI for identifying and counting aquatic

microflora. Limnology and Oceanography 25: 943–948.

Pusceddu A, Danovaro R. 2009. Exergy, ecosystem functioning and efficiency in a

coastal lagoon: The role of auxiliary energy. Estuarine Coastal and Shelf Science.

84: 227-236.

Ratledge C. 1994. Biochemistry of Microbial Degradation. Kluwer Academic Publishers,

Dordrecht. 584 p.

Ricklefs RE. 1977. Environmental heterogeneity and plant species diversity: a

hypothesis. American Naturalist 111: 376-381.

Rustad LE. 1994. Element dynamics along a decay continuum in a red spruce ecosystem

in Maine, USA. Ecology 75: 867–879.

Saiya-Cork KR, Sinsabaugh RL, Zak DR. 2002. The effects of long term nitrogen

deposition on extracellular enzyme activity in anAcer saccharum forest soil, Soil

Biology and Biochemistry 34: 1309-1315.

Salinas N, Malhi Y, Meir P, Silman M, Cuesta R, Roman C, Huaman J, Salinas D,

Huaman V, Gibaja A, Mamani M, Farfan F. 2011. The sensitivity of tropical leaf

194

litter decomposition to temperature: results from a large-scale leaf translocation

experiment along an elevation gradient in Peruvian forests New Phytologist 189:

967-977.

Saugier B, Roy J, Mooney HA. 2001. Estimation of global terrestrial productivity:

converging toward a single number? In: Roy J, Saugier B, Mooney HA (eds)

Terrestrial global productivity. Academic Press, London, pp 543–557.

Sayer EJ. 2006. Using experimental manipulation to assess the roles of leaf litter in the

functioning of forest ecosystems. Biological Reviews 81: 1–31.

Scheel T, Hofer M, Ludwig S. 2000. Differential expression of manganese peroxidase

and laccase in white-rot fungi in the presence of manganese or aromatic

compounds. Applied Microbiology and Biotechnology 54: 686-691.

Setӓlӓ H, McLean A. 2004. Decomposition rate of organic substrates in relation to the

species diversity of soil saprophytic fungi. Oecologia 139: 98–107.

Schweitzer JA, Bailey JK, Hart SC, Whitham TG 2005. Nonadditive effects of mixing

cottonwood genotypes on litter decomposition and nutrient dynamics. Ecology

86: 2834–2840.

Silver WL, Miya RK. 2001. Global patterns in root decomposition: comparisons of

climate and litter quality effects. Oecologia 129: 407–419.

Sinsabaugh RL, Antibus RK, Linkins AE. 1991. An emzymatic approach to the analysis

of microbial activity during plant litter decomposition. Workshop on modern

techniques in soil ecology, Atlanta GA.

195

Sinsabaugh RL, Antibus RK, Linkins AE, McClaugherty CA, Rayburn L, Repert D,

Weiland T. 1993. Wood decomposition: nitrogen and phosphorus dynamics in

relation to extracellular enzyme activity. Ecology 74: 1586–1593.

Sinsabaugh RL. 1994. Enzymatic analysis of microbial pattern and process. Biology and

Fertility of Soils 17: 69–74.

Sinsabaugh RL, Carreiro MM, Repert DA. 2002. Allocation of extracellular enzymatic

activity in relation to litter composition, N deposition, and mass loss,

Biogeochemistry 60: 1-24.

Sinsabaugh RL. 2005. Fungal enzymes at the community scale In The fungal community:

Its organization and role in the ecosystem, 3rd edition, edited by Dighton J, White

JF, Oudemans P. Taylor & Francis Group New York NY, USA. 960 p.

Sinsabaugh RL, Lauber CL, Weintraub MN, Ahmed B, Allison SD, Crenshaw C,

Contosta AR, Cusack D, Frey S, Gallo ME, Gartner TB, Hobbie SE, Holland K,

Keeler BL, Powers JS, Stursova M, Takacs-Vesbach C, Waldrop MP, Wallenstein

MD, Zak DR, Zeglin LH. 2008. Stoichiometry of soil enzyme activity at global

scale. Ecology Letters 11:1252–1264.

Snajdr J, Cajthaml T, Valaskova V, Merhautova V, Petrankova M, Spetz P, Leppanen K,

Baldrian P. 2011. Transformation of Quercus petraea litter: successive changes

in litter chemistry are reflected in differential enzyme activity and changes in the

microbial community composition. FEMS Microbiology Ecology 75(2): 291-

303.

196

Steffen KT, Cajthaml T, Snajdr J, Baldrian P. 2007. Differential degradation of oak

(Quercus petraea) leaf litter by litter decomposing basidiomycetes. Research in

Microbiology 158: 447–455.

Swift MJ, Heal OW, Anderson JM. 1979. Decomposition in terrestrial ecosystems.

Blackwell Scientific Publications, Oxford, 384 pp.

Tank JL, Webster JR 1998. Interaction of substrate and nutrient availability on wood

biofilm processes in streams. Ecology 79: 151–162.

Taylor BR, Parkinson D, Parsons WFJ. 1989. Nitrogen and lignin content as predictors of

litter decay rates: a microcosm test. Ecology 70: 97-104.

Thorn G. 1997. The fungi in soil. In: van Elsaß, J.D., Trevors, J.T., Wellington, E.M.H.

(Eds.), Modern Soil Microbiology. Marcel Dekker, New York, pp. 63–128.

Tilman D, Knops J, Wedin D, Reich P, Ritchie M, Siemann E. 1997. The influence of

functional diversity and composition on ecosystem processes. Science 277: 1300–

1302.

Tiunov AV, Scheu S. 2005. Facilitative interactions rather than resource partitioning

drive diversity-functioning relationships in laboratory fungal communities.

Ecology Letters 8: 618–625.

Valaskova V, Snajdr J, Bittner B, Cajthaml T, Merhautova V, Hoffichter M, Baldrian P.

2007. Production of lignocellulose-degrading enzymes and degradation of leaf

litter by saprotrophic basidiomycetes isolated from a Quercus petraea forest. Soil

Biology and Biochemistry 39: 2651–2660.

197

Vivanco L, Austin AT. 2008. Tree species identity alters forest litter decomposition

though long-term plant and soil interactions in Patagonia, Argentina. Journal of

Ecology 96: 727-736.

Voriskova J, Dobiasova P, Snajdr J, Vanek D, Cajthaml T, Santruckova H, Baldrian P.

2011. Chemical composition of litter affects the growth and enzyme production

by the saprotrophic basidiomycete Hypholoma fasciculare. Fungal Ecology 4(6):

417-426.

Vos VCA, van Ruijven J, Berg MP, Peetters ETHM, Berendse F. 2011. Macro-detritivore

identity drives leaf litter diversity effects. Oikos 120: 1092-1098.

Waldrop MP, Zak DR, Sinsabaugh RL, Gallo M, Lauber C. 2004. Nitrogen deposition

modifies soil carbon storage through changes in microbial enzyme activity.

Ecological Applications 14: 1172–1177.

Waldrop MP, Zak DR. 2006. Response of oxidative enzyme activities

to nitrogen deposition affects soil concentrations of dissolved organic carbon.

Ecosystems 9(6): 921-933.

Wang S, Ruan H, Han Y. 2011. Effects of microclimate, litter type, and mesh size on leaf

litter decomposition along an elevation gradient in the Wuyi Mountains,

China. Ecological Research. 25: 1113-1120.

Wardle DA, Bonner KI, Nicholson KS. 1997. Biodiversity and plant litter: experimental

evidence which does not support the view that enhanced species richness

improves ecosystem function. Oikos 79: 247–258.

198

Wittmann C, Kahkonen MA, Ilvesniemi H, Kurola J, Salkinoja-Salonen MS. 2004. Areal

activities and stratification of hydrolytic enzymes involved in the biochemical

cycles of carbon, nitrogen, sulphur and phosphorus in podsolized boreal forest

soils. Soil Biology and Biochemistry 36: 425–433.

Wieder RK, Lang GE. 1982. A critique of the analytical methods used in examining

decomposition data obtained from litter bags. Ecology 63:1636–1642.

Wohl DL, Arora S, Gladstone JR. 2004. Functional redundancy supports biodiversity and

ecosystem function in a closed and constant environment. Ecology 85: 1534–

1540.

Wu L, Feinstein LM, Valverde-Barrantes O, Kershner MW, Leff LG, Blackwood CB.

2011. Placing the effects of leaf litter diversity on saprotrophic microorganisms in

the context of leaf type and habitat. Microbial Ecology 61: 399-409.

Zabed Hossain M, Okubo A, Sugiyama S. 2010. Effects of grassland species on

decomposition of litter and soil microbial communities. Ecological Research 25:

255–261.

Zak JC, Willig MR, Moorhead DL. Wildman HG. 1994. Functional diversity of

microbial communities. Soil Biology and Biochemistry 26: 1101–1108.

Zhang D, Hui D, Luo Y, Zhou G. 2008. Rates of litter decomposition in terrestrial

ecosystems: global patterns and controlling factors. Journal of Plant Ecology 1:

85-95.

Zifcakova L, Dobiasova P, Kolarova Z, Koukol O, Baldrian P. 2011. Enzyme activities

of fungi associated with Picea abies needles. Fungal Ecology 4(6): 427-43

Chapter 6

Synthesis

The work in this dissertation was concerned with quantifying microbial diversity and biogeochemical activity patterns in order to understand processes that influenced pattern formation and variability. This is an important task because although we are aware of many important functions that microbes conduct (Fenchel 1998, Madigan et al.

2006), there is much that we do not understand concerning processes driving microbial diversity (Maron et al. 2011) and biogeochemical (carbon-cycling) activity patterns

(Ranjard et al. 2010). Microbial patterns and processes were then compared to theory that has been developed for macrorganisms. We documented evidence for common macro-scale patterns at the microbial scale including a taxa-area relationship, distance- decay, and increased resource heterogeneity associated with an increase in productivity

(the species complementary hypothesis).

Novel aspects of the research within this dissertation included quantifying microbial communities on entire, discrete habitat patches (whole individual leaves), documentation of differential fungal taxa-area relationships, documentation of distance- decay patterns across discrete adjacent habitat patches, quantification of bias associated with a common DNA extraction protocol, and recommendations for improving the ability to quantify microbial communities. Community composition was shown to be influenced by energy availability. Neutral and species-sorting community assembly processes were

199

200

tested at various spatial scales; the synthesis of these patterns is described below.

Community composition and dynamics (productivity, extracellular enzyme activity) were documented to have an important role in saprotrophic carbon-cycling activity. Finally, the interaction between community dynamics and litter mixtures was documented, providing evidence of an ecological process whereby microbial activity increases ecosystem exergy. Details regarding how each of these relates to the ecological body of literature and future directions to enhance dissertation findings may be found in the remainder of this chapter.

The impact of energy availability on microbial richness

The discrete nature of senesced leaves was utilized to document the ubiquitous

TAR that has been documented (Rosenzweig 1995, Connor and McCoy 2001, Lomolino

2001, Drakare et al. 2006) for macroorganisms for fungi on maple leaves, but not on beech. Because taxa richness is often used as a measure of diversity (Magurran 2004), the taxa-area relationship that was documented on maple but not beech leaves suggests there is a link between energy availability and saprotrophic fungal community richness.

Wright’s species-energy theory (1983) was developed based on the species-area relationship (Arrhenius 1921; Gleason 1922) and linked available energy with increased richness. Most of the work testing the species-energy relationship (SER) has been conducted on macroorganisms at a global scale (Hawkins et al. 2003; Kreft and Jetz

2007; Whittaker et al. 2007); this is the first study that has documented a pattern suggestive of a SER for microbes.

201

Wright identified one hypothesis that is common to both the species-energy and species-area relationships (SAR) that may explain a mechanism driving the SER that was documented for saprotrophic microbes: increased species richness is a positive function of an increased number of individuals found in areas with greater energy (SER) or area

(SAR). This idea became known in studies as the “more-individuals hypothesis” (e.g.,

Srivastava and Lawton 1998; Hurlbert 2004). This hypothesis could be tested in future work with saprotrophic microbes: individuals could be counted and compared via microscopy or qPCR which counts gene copies. This approach would inform researchers in determining if saprotrophic SER is linked to the more-individuals hypothesis, or is being driven by other mechanistic processes.

The “productivity hypothesis” is a SER hypothesis states that differences in global climate patterns affect richness by either constraining or increasing trophic food web energy and that SER is driven by energy cascading through a food web (Wright et al.

1993, Huston 1994, Mittelbach et al. 2001) and is typical of what has emerged as the scientific community tested the SER across global scales. The productivity hypotheses.

This hypothesis could be tested in the systems that were studied in this dissertation if the food web (and not just microbes) associated with maple versus beech senesced leaves could be quantified. Examining what has been done at a macro scale, it becomes evident that the SER documented in this study could be the beginning of decades of study that would increase understanding regarding how energy availability affects saprotrophic microbial richness. Future work could not only test the two hypotheses described above,

202

but also determine if SER exists on labile leaves besides sugar maple, and does not exist on recalcitrant leaves besides beech.

The relative impact of species-sorting and neutral processes on community composition is related to spatial scale

We found that evidence supporting neutral and species-sorting processes was related to the scale of a particular investigation. Neutral processes were found to be supported at the scale of individual leaf communities (alpha diversity). The sequence data that was used to quantify richness in the taxa-area relationship project was also used to generate community rank-abundance distributions (RADs). The RADs were compared to distributions consistent with either species-sorting or neutral theories, with overwhelming support for neutral dynamics. Community composition (TRFLP) profiles were generated for these plus 156 other leaf communities located in vernal pool, upland, and riparian habitat. At the scale of multiple communities (beta diversity across leaf mixtures), we found evidence supporting both species-sorting and neutral dynamics

(specifically, dispersal limitation). Environmental heterogeneity within the leaf packs was at least partially represented by depth within the leaf pack. Depth within the leaf pack was was shown to influence community composition by partial Mantel analysis and redundancy analysis, suggesting a degree of species-sorting. The operation of both species-sorting and neutral processes within a biological assemblage is a pattern that has recently been described as being essential to the maintenance of biodiversity (Zhang et al.

2012) and is part of an ubiquitous natural continuum (Chesson 2000; Chase 2005; Gravel et al. 2006; Leibold and McPeek 2006; Adler et al. 2007; Cadotte 2007).

203

Collectively, our data suggests the relative role of species-sorting and neutral processes across scales of 1 to 30 leaf communities. Species-sorting may serve to filter pools of taxa at different depths, while individual communities within each depth are assembled according to neutral processes. This suggests a degree of species-sorting influencing beta diversity, while neutral processes influence alpha diversity, with the relative importance of each process varying depending on the scale being quantified.

In addition, significant distance-decay relationships were found at 5 out of 6 sites sampled. The distance-decay relationship has been documented for all domains of life

(Nekola and White 1999, Green et al. 2004, Horner-Devine et al. 2004) and determining processes that influence it is the subject of many studies (Tuomisto et al. 2003, Ramette and Tiedje 2007, Buckley and Jetz 2008, Bell 2010) because understanding what influences changes in community composition helps us understand how biodiversity is maintained (Martiny et al. 2011). While distance-decay has been documented for sub- sampled fungal and bacterial communities at scales from centimeters to continents

(Martiny et al. 2011), this is the first study upon which fungal distance-decay was documented for entire adjacent communities across a scale upon which we would expect fungal hyphal dispersal to occur (Boddy et al. 2009). Neutral processes influencing alpha diversity and depth associated with species-sorting indicate that both of these processes are embedded within the distance-decay relationships that we found.

The importance of microbial communities in carbon-cycling activity

Although designed to determine the effects of the usual array of explanatory variables quantified in most litter decomposition studies (leaf type, carbon and N pools,

204

habitat, and incubation time) (Gartner and Cardon, 2004, Hӓttenschwiler et al. 2005), the litter decomposition study in this dissertation was also designed to quantify the ecological role of microbial communities during saprotrophic activity. The impact of microbial community composition on litter decomposition is a potentially important, though neglected, “black box” that has rarely been quantified in manipulative litter decomposition studies (McGuire and Treseder 2011). It has been proposed that community composition may be important in later stages of decomposition when recalcitrant-based compounds are most prevalent (McGuire and Treseder 2011) because recalcitrant breakdown requires a pool of enzymes produced by a relatively specialized group of fungi (Schimel 1995; Fontaine and Barot 2005). We found that fungal community composition was significantly correlated with the activity of all measured enzymes, and therefore important not only at the later stages of decomposition. There was no significant relationship between bacterial community composition and enzyme activity. Fungal community composition was also correlated with percent AFDM remaining, whereas bacterial community composition was not. These results provide documentation that fungal community composition likely changes during decomposition, is correlated with saprotrophic extracellular enzyme activity, and has an influence on litter decomposition.

Although leaf litter mixtures did not decompose at a different rate than single- species litter, the work in this dissertation indicated that mixtures of litter play an important role in both microbial and ecosystem ecology. Microbial biomass was significantly greater on litter mixtures than on single-species litter, although the same

205

amounts of carbon were lost from the litter. This shows that microbes process carbon more efficiently in litter mixtures because a greater proportion of the carbon was converted into microbial biomass in mixed treatments and less carbon is lost to respiration. Microbial biomass represents carbon that has been retained within the ecosystem and which can be transferred among trophic levels. This provides the first documented evidence of increased exergy driven by leaf litter diversity and microbial communities, and illustrates the critical role that microbes play in sustaining energy within an ecosystem.

How well are we able to quantify actual patterns that exist in nature?

The modern approaches to quantifying microbial diversity patterns are based on molecular methods, and there are concerns within the research community regarding the accuracy of the patterns derived from these methods (Head et al. 1998; Frostegård et al.

1999; Martin-Laurent et al. 2001; Carrig et al. 2007; Lombard et al. 2011). Studies have been conducted to investigate these concerns (Dineen et al. 2010; Jones et al. 2011,

Zammit et al. 2011), and Chapter 2 of this dissertation (also published as Feinstein et al.

2009) adds to the body of work that has been conducted. In this study we demonstrated that the quantity of DNA and soil fungal and bacterial gene copies may be increased with multiple extractions on each sample. We demonstrate significant differences in bacterial community composition associated with initial and final extractions. The study proposed that in order to accurately quantify comprehensive community analysis for soil bacterial communities, multiple extractions should be conducted. This information has proved to be useful to the scientific community: since its publication in August 2009 information

206

within the chapter has been cited in 27 peer-reviewed publications according to the

Science Citation Index.

References

Adler PB, HilleRisLambers J, Levine JM. 2007. A niche for neutrality. Ecology Letters

10: 95–104.

Arrhenius O. 1921. Species and area. Journal of Ecology 9: 95–99.

Bell T. 2010. Experimental tests of the bacterial distance-decay relationship. ISME

Journal 4:1357–1365.

Boddy L, Hynes J, Bebber DP. 2009. Saprotrophic cord systems: dispersal mechanisms

in space and time. Mycoscience 50(1): 9-19.

Buckley LB, Jetz W. 2008. Linking global turnover of species and environments.

Proceedings of the National Academy of Science USA 105: 17836–17841.

Cadotte MW. 2007. Concurrent niche and neutral processes in the competition-

colonization model of species coexistence. Proceedings of the Royal Society B:

Biological Sciences 274: 2739–2744.

Carrig C, Rice O, Kavanagh S, Collins G, O’Flaherty V. 2007. DNA extraction method

affects microbial community profiles from soils and sediment. Applied

Microbiology and Biotechnology 77: 955–964.

Chase JM. 2005. Towards a really unified theory for metacommunities. Functional

Ecology 19: 182–186.

207

Chesson P. 2000. Mechanisms of maintenance of species diversity. Annual Review of

Ecology and Systematics 31: 343–366.

Connor EF, McCoy ED. 2001. Species–area relationships. Encyclopedia of biodiversity,

Vol. 5 (ed. by S.A. Levin), pp. 397–411. Academic Press, San Diego. 4666 p.

Dineen SM, Aranda R, Anders DL, Robertson JM. 2010. An evaluation of commercial

DNA extraction kits for the isolation of bacterial spore DNA from soil. Journal of

Applied Microbiology 109: 1886-1896.

Drakare S, Lennon JJ, Hillebrand H. 2006. The imprint of the geographical, evolutionary

and ecological context on species-area relationships. Ecology Letters 9: 215–227.

Feinstein LM, Sul WJ, Blackwood CB. 2009. Assessment of bias associated with

incomplete extraction of microbial DNA from soil. Applied and Environmental

Microbiology 75: 5428–5433.

Fenchel T. 1998. Bacterial biogeochemistry: the ecophysiology of mineral cycling (2nd

ed.). San Diego: Academic Press. 307 p.

Fontaine S, Barot S. 2005. Size and functional diversity of microbe populations

control plant persistence and long-term soil carbon accumulation. Ecology

Letters 8: 1075-1087.

Frostegård A, Courtois S, Ramisse V, Clerc S, Bernillon D, Le Gall F, Jeannin P, Nesme

X, Simonet P. 1999. Quantification of bias related to the extraction of DNA

directly from soils. Applied and Environmental Microbiology 65: 5409-5420.

Gleason HA. 1922. On the relationship between species and area. Ecology; 3: 158–162.

208

Gravel D, Canham CD, Beaudet M, Messier C. 2006. Reconciling niche and neutrality:

the continuum hypothesis. Ecology Letters 9: 399–409.

Green JL, Holmes AJ, Westoby M, Oliver I, Briscoe D, Dangerfield M, Gillings M,

Beattie AJ. 2004. Spatial scaling of microbial eukaryote diversity. Nature 432:

747-750.

Hawkins BA, Field R, Cornell HV, Currie DJ, Guegan JF, Kaufman DM, Kerr JT,

Mittelbach GG, Oberdorff T, O'Brien EM, Porter EE, Turner, JRG. 2003. Energy,

water, and broad scale geographic patterns of species richness. Ecology 84:

3105–3117.

Head IM, Saunders JR, Pickup RW. 1998. Microbial evolution, diversity, and ecology: a

decade of ribosomal RNA analysis of uncultivated microorganisms. Microbial

Ecology 35: 1–21.

Horner-Devine MC, Lage M, Hughes JB, Bohannan BJM. 2004. A taxa-area

relationship for bacteria. Nature 432: 750-753.

Hurlbert AH. 2004. Species-energy relationships and habitat complexity in bird

communities. Ecology Letters 7: 714–720.

Huston MA. 1994. Biological diversity. Cambridge, UK: Cambridge University Press;

704 p.

Jones MD, Singleton DR, Sun W, Aitken MD. 2011. Multiple DNA Extractions

Coupled with Stable-Isotope Probing of Anthracene-Degrading Bacteria in

Contaminated Soil. Applied and Environmental Microbiology 77: 2984-2991

209

Kretf H., and W. Jetz. 2007. Global patterns and determinants of vascular plant diversity.

Proceedings of the National Academy of Sciences of the USA 104:5925–5930.

Leibold MA, McPeek MA (2006) Coexistence of the niche and neutral perspectives in

community ecology. Ecology 87: 1399–410.

Lombard N, Prestat E, van Elsas JD, Simonet P. 2011. Soil-specific limitations for

access and analysis of soil microbial communities by metagenomics. FEMS

Microbiology Ecology 78: 31-49

Lomolino MV. 2001. The species–area relationship: new challenges for an old pattern.

Progress in Physical Geography 25: 1–21.

Madigan MT, Martinko JM, Dunlap PV, Clark DP. 2006. Brock Biology of

Microorganisms (10th edition). San Fransisco: Pearson/Benjamin Cummings;

986 p.

Magurran AE. 2004. Measuring Biological Diversity. Oxford, UK: Blackwell

Publishing 264 p.

Maron PA, Mougel C, Ranjard L. 2011. Soil microbial diversity: methodological

strategy, spatial overview and functional interest. Comptes Rendus Biologies;

334: 403–411.

Martin-Laurent F, Philippot L, Hallet S, Chaussod R, Germon JC, Soulas G, Catroux G.

2001. DNA extraction from soils: old bias for new microbial diversity analysis

methods. Applied Environmental Microbiology 67: 2354–2359.

Martiny JBH, Eisen JA, Penn K, Allison SD, Horner-Devine MC. 2011. Drivers

of bacterial beta-diversity depend on spatial scale. Proceedings of the National

210

Academy of Sciences USA 108: 7850-7854.

McGuire, K.L., Treseder, K.K., 2010. Microbial communities and their relevance

for ecosystem models: decomposition as a case study. Soil Biology &

Biochemistry 42: 529–535.

Mittelbach GG, Steiner CE, Scheiner SM, Gross KL, Reynolds HL, Waide RB, Willig

MR, Dodson SI, Gough L. 2001. What is the observed relationship between

species richness and productivity? Ecology 82: 2381-2396.

Nekola JC, White PS. 1999. The distance decay of similarity in biogeography and

ecology. Journal of Biogeography 26: 867–878.

Ramette A, Tiedje JM. 2007. Multiscale responses of microbial life to spatial distance

and environmental heterogeneity in a patchy ecosystem. Proceedings of the

National Academy of Sciences USA 104:2761–2766.

Ranjard L, Dequiedt S, Jolivet C, Saby NA, Thioulouse J, Harmand J, Loisel P, Rapaport

A, Fall S, Simonet P, Joffre R, Chemidlin-Prévost Bouré N, Maron PA, Mougel

C, Martin MP, Toutain B, Arrouays D, Lemanceau P. 2010. Biogeography of

soil microbial communities: a review and a description of the ongoing French

national initiative, Agronomy Sustainable Development 30: 359–365.

Rosenzweig ML. 1995. Species diversity in space and time. Cambridge Univ Press,

Cambridge, UK.

Schimel J. 1995. Ecosystem consequences of microbial diversity and community

structure. In: Chapin, F.S., Korner, C. (Eds.), Arctic and Alpine Biodiversity:

211

Patterns, Causes, and Ecosystem Consequences. Berlin, Germany: Springer-

Verlag; 332 p.

Srivastava DS, Lawton JH. 1998. Why more productive sites have more species: an

experimental test of theory using tree-hole communities. American Naturalist

152: 510–529.

Tuomisto H, Ruokolainen K, Yli-Halla M. 2003. Dispersal, environment, and floristic

variation of western Amazonian forests. Science 299:241–244.

Whittaker RJ, Nogues-Bravo D, Araujo MB. 2007. Geographical gradients of species

richness: a test of the water-energy conjecture of Hawkins et al. (2003) using

European data for five taxa. Global Ecology and Biogeography 16:76–89.

Wright DH. 1983. Species-energy theory: an extension of species-area theory. Oikos;

41: 496–506.

Wright DH, Currie DJ, Maurer BA. 1993. Energy supply and patterns of species

richness on local and regional scales. In Ricklefs RE, Schluter D, editors. Species

diversity in ecological communities: historical and geographical perspectives.

Chicago, Illinois, USA: University of Chicago Press Pages 66-74.

Zammit CM, Mutch LA, Watling HR, Watkin ELJ. 2011. The recovery of nucleic acid

from biomining and acid mine drainage microorganisms. Hydrometallurgy 108:

87–92.

Zhang DY, Zhang BY, Lin K, Jiang XH, Tao Y, Hubbell S, He FL, Ostling A. 2012.

Demographic trade-offs determine species abundance and diversity. Journal of

Plant Ecology 5(1): 82-88.