. a .

1 , ,
.‘su. Jivx‘ 4
I6J‘fl?§alvr¢d.§
..‘§...Jllfivd.l I 3:: A“...
.r1 . (an... (5.51... 30. .l
rm .3 a... guy-610 may!!!
: 3.37.3.4!
h. i!!! a... as
, (‘50 xi
3

5....
n...
2%.... a,

.s
..
«vafi;u.1nns 2.. En}.

1:32.... cad-"HQ

5:1 .19)?
hi l.’ i

It.

.i... mi:
. 13:4 .

I

.. L.,....L....m

'5!

 

We};

3 LIBRARY
200 5; Michigan State
University

This is to certify that the
dissertation entitled

PLANT-SOIL FEEDBACKS IN TEMPERATE AND TROPICAL
FORESTS

presented by
Sarah McCarthy Neumann

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Forestry

and Ecology, Evolutionary
Biology and Behavior

/"‘7

{\[1\;L( \C\<~/ ML

Major Professor’ 5 Signature
May 21", 2008

Date

MSU is an Affirmative Action/Equal Opportunity Institution

._ .. -..—.-n-n-c-n-

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KilProj/Acc8Pres/CIRC/DateOue.indd

PLANT-SOIL FEEDBACKS IN TEMPERATE AND TROPICAL FORESTS
By

Sarah McCarthy Neumann

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Forestry
Program in Ecology, Evolutionary Biology and Behavior

2008

ABSTRACT
PLANT -SOIL FEEDBACKS IN TEMPERATE AND TROPICAL FORESTS
By
Sarah McCarthy Neumann

The Janzen-Connell (J -C) Model proposes that host-specific natural enemies maintain
high tropical tree diversity by reducing seed and/or seedling survivorship near
conspecific adults and/or at high conspecific densities. Such non-competitive distance or
density—dependent (N CDD) mortality would favor establishment of heterospecific
individuals, thus promoting species coexistence. Negative plant-soil feedback, whereby
individual plants “culture” the soil community in which they grow to the detriment of
themselves and other conspecific individuals, may be an important mechanism that could
create NCDD mortality and/or reduced growth. I used a wet-sieving method to filter out
biotic and water extractable chemical elements from soil that had been cultured by
conspecific and heterospecific adults and seedlings. These soil extracts were used in
greenhouse experiments with temperate and tropical tree species to examine 1)
advantages to heterospecific and disadvantages to conspecific recruitment, 2) soil
mechanisms underlying NCDD, 3) differences between common and rare species in
sensitivity to J-C processes, 4) the strength of J-C processes in tropical versus temperate

forests, 5) and the interactions of J -C processes with light availability. I found that

susceptibility to microbial extract cultured by conspecific individuals was negatively
correlated with seedling shade tolerance not a species’ local abundance, thereby
exaggerating apparent shade tolerance differences among species and likely contributing
to species coexistence through heightening niche differentiation. When comparing effects
of con- vs. hetero-specific cultured soils, 1 found that species-specific feedbacks between
adult trees (not seedlings) and soil influenced seedling performance for all temperate and
tropical species. Con- and hetero-specific effects had similar prevalence and magnitude
of influence for temperate species whereas three of the six tropical species had decreased
performance when grown with extract cultured by con- vs. all hetero-specific adults and
an additional two species had decreased performance in con- vs. two or more hetero-
specific cultured extracts. In addition, in temperate forests, soils cultured by a particular
species do not necessarily improve heterospecific seedling performance relative to
conspecific seedlings which may impede the ability of these plant-soil feedbacks to
enhance species coexistence. However, in tropical forests, heterospecific seedlings are
favored relative to conspecific seedlings in soils cultured by a given species. Thus, J -C
processes appear stronger in tropical vs. temperate forests, at least those mediated by
plant-soil feedbacks. Surprisingly, chemical factors in the soil not micro-organisms seem
to be primarily responsible for these feedbacks. Thus, my dissertation identifies a novel
mechanism (feedback between adult trees and soil abiotic factors) that creates NCDD
seedling mortality and/or reduced growth and moves the J-C Model beyond solely

focusing on natural enemies.

DEDICATED

To my husband David and my entire family

iv

ACKNOWLEDGMENTS .

I would like to thank my husband, David, for his love and support. He believed in
me and gave me the encouragement to pick myself up after set-backs and start my
experiments anew. For my son, Isaac, who is a blessing — he makes me laugh and remind
me of what is important in life. My parents, John and Fran McCarthy, have always loved
and encouraged me, but I thank them specifically for the wonderful example they have
provided in their marriage, as parents and as Christians.

My undergraduate advisor at Sewanee, Jon Evans, saw potential in me as a fiiture
scientist and teacher and not only encouraged me but taught me the skills necessary to
excel. Ithank him for mentoring me and guiding me throughout my undergraduate
career.

I thank my advisor, Rich Kobe, for all of his support, guidance, encouragement
and for giving me so much freedom to pursue and carry-out my research interests. I also
thank my other committee members, Andy J arosz, John Klironomos, David Rothstein
and Mike Walters, who each provided their unique perspective to this work and through
their constructive criticism improved this thesis greatly.

My time in graduate school would have been much more difficult without the
incredible support from the following Forestry staff: Barb Anderson, Carol Graysmith,

Paul Bloese, Randy Klevickas and Juli Kerr.

The staff at La Selva helped me with logistics and made my research in Costa
Rica so much easier. In addition, I met so many graduate students from all over the world
during my stays at La Selva and they all helped make my trips to Costa Rica both
productive and fun. i

To all of my assistants who have helped me either at MSU or in Costa Rica
(Mindy McDermott, Ted Salk, Katie Weaver, Sam Tourtellot, Renee Pereault, Jennifer
Sun, Amanda Gevens, Jennifer Hunnell, Melissa McDermott and so many more) thank
you for working so hard over the years sometimes in less than ideal conditions. In
addition, my research assistants in Costa Rica were invaluable for all of my tropical
work. I wish to thank Marisol Luna, Ademar Hurtado and Martin Cascante for all of their
hard work over the years and their patience and good humor in the face of my limited
ability to speak Spanish.

I am grateful for the support and friendship of all the Forestry graduate students
who were my colleagues during the past seven years. I would especially like to thank
Meera Iyer and Tom Baribault who were my lab mates for most of the time while I was at
MSU — they both were always willing to give advice, and to lend an car when I needed to
vent. Thank you to Laura Marx, Joseph LeBouten, Jesse Randall and Justin Kunkle who
may as well have been my lab mates for the good times we had together as well as the
great advice and help they always gave me. I also want to thank Zhanna Yermakov who
was a great friend while at MSU and has remained one to this day.

I thank my friends from Sewanee (Colleen Rye, Paige Eagan and Jaclyn

Newman) who have all struggled through graduate or medical school the same time as 1.

vi

They have given me so much encouragement and friendship throughout the years and I
cherish all of them.

This research was fimded by a grant from the National Science Foundation (DEB
0235907). Additional fellowship and grant support was provided by the Organization of
Tropical Studies, and the Graduate School, Forestry Department and Graduate Program
in Ecology, Evolutionary Biology and Behavior at Michigan State University. The
Ministerio del Ambiente y Energia (MINAE) granted me permission to conduct research

in Costa Rica.

vii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................... xii

LIST OF FIGURES ................................................................................ xiii

CHAPTER 1
INTRODUCTION ........................................................................... 1

CHAPTER 2

TOLERANCE OF SOIL PATHOGENS CO-VARIES WITH SHADE TOLERANCE

ACROSS SPECIES OF TROPICAL TREE SEEDLINGS ................................... 25
Abstract ..................................................................................... 25
Introduction ................................................................................. 26
Materials and Methods .................................................................... 29
Results ....................................................................................... 36
Discussion .................................................................................. 38

CHAPTER 3

CONSPECIFIC AND HETEROSPECIFIC TREE-SOIL FEEDBACKS INFLUENCE

SURVIVORSHIP AND GROWTH OF TEMPERATE TREE SEEDLINGS ............. 51
Abstract ..................................................................................... 51
Introduction ................................................................................. 52
Materials and Methods .................................................................... 55
Results ....................................................................................... 61
Discussion .................................................................................. 66

CHAPTER 4

CONSPECIFIC TREE—SOIL FEEDBACKS REDUCE SURVIVORSHIP AND

GROWTH OF TROPICAL TREE SEEDLINGS .............................................. 87
Abstract ..................................................................................... 87
Introduction ................................................................................. 88
Materials and Methods .................................................................... 91
Results ....................................................................................... 96
Discussion ................................................................................. 100

viii

CHAPTER 5

CONCLUSION ........................................................................... 1 23
APPENDICES

Appendix A. Kaplan-Meier analysis of seedling life span compared between

microbial vs. sterilized microbial extract for all study species ...................... 130

Appendix B. Analysis of covariance results for the effects of inoculum treatment
on seedling mass, stem height and root morphology. Inoculum = microbial vs.
sterilized microbial extract; adult = location where soil was collected for use in
extraction; bench = location that seedlings were randomly assigned to during the
experiment; covariate = initial seed mass .............................................. 132

Appendix C. Analysis of covariance results for the effects of inoculum treatment
on organ mass, stem height and root morphology allocation. Inoculum =
microbial vs. sterilized microbial extract; adult = location where soil was
collected for use in extraction; bench = location that seedlings were randomly
assigned to during the experiment; covariate = total plant mass ................... 136

Appendix D. Analysis of covariance results for the effects of inoculum treatment
on root morphology. Inoculum = microbial vs. sterilized microbial extract; adult
= location (near the bole of 4 different adults per species) where soil was collected
for use in extraction; bench = location that seedlings were randomly assigned to
during the experiment; covariate = root mass ......................................... 139

Appendix E. Meta-analysis results for the effect (size and magnitude) of inoculum
treatment on organ mass, stem height and root morphology for 21 tropical tree
species grouped into 3 categories in each of 5 different local species
characteristics. Data for meta-analysis derived from ANCOVA analysis using
initial seed mass as a covariate (Appendix B) ........................................ 141

Appendix F. Meta-analysis results for the effect (size and magnitude) of inoculum
treatment on organ mass, stem height and root morphology for 21 tropical tree
species grouped into 3 categories in each of 5 different local species
characteristics. Data for meta-analysis derived from ANCOVA analysis using
total plant mass as a covariate (Appendix C) .......................................... 142

Appendix G. Meta-analysis results for the effect (size and magnitude) of
inoculum treatment on root morphology for 21 tropical tree species grouped into 3

ix

categories in each of 5 different local species characteristics. Data for meta-
analysis derived from ANCOVA analysis using root mass as a covariate
(Appendix D) .............................................................................. 143

Appendix H. Experimental design and allocation of temperate seedlings to
treatments .................................................................................. 145

Appendix I. Hazards ratios for the effects of light availability (2% vs. 22% full
sun), soil extract (non-sterile vs. sterile), soil source (tree species culturing soil)
and initial seed mass on seedling mortality estimated by the Cox regression

model ....................................................................................... 147

Appendix J. Split-plot analysis of covariance results for the effects of light
availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source
(tree species culturing soil), bench and covariate (initial seed mass) on A. rubrum
seedling mass and stem height. ......................................................... 148

Appendix K. Split-plot analysis of covariance results for the effects of light
availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source
(tree species culturing soil), bench and covariate (initial seed mass) on A.
saccharum seedling mass and stem height ............................................ 149

Appendix L. Split-plot analysis of covariance results for the effects of light
availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source
(tree species culturing soil), bench and covariate (initial seed mass) on F.
americana seedling mass and stem height. ............................................ 152

Appendix M. Split-plot analysis of covariance results for the effects of light
availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source
(tree species culturing soil), bench and covariate (initial seed mass) on Q. rubra
seedling mass and stem height. .......................................................... 153

Appendix N. Analysis of variance results for the effect of soil source (tree species
culturing soil) on base cations (combined Ca, K, and Mg), total organic C, total N,
and C :N in the soil extracts .............................................................. 155

Appendix 0. Kaplan-Meier analysis of the effect of inoculum type (Five
F usarium morphotypes and control) on seedling life span for A. rubrum, F.
americana and Q. rubra .................................................................. 156

Appendix P. Mixed effects analysis of covariance results for the fixed effects of
inoculum type (Control vs. F usarium morphotype 1-5), random effect of bench
and covariate (initial seed mass) on seedling mass (mg) on A. rubrum seedling
mass and stem height .................................................................... 157

Appendix Q. Mixed effects analysis of covariance results for the fixed effects of
inoculum type (Control vs. F usarium morphotype 1-5), random effect of bench
and covariate (initial seed mass) on F. americana seedling mass and stem

height ....................................................................................... 159

Appendix R. Mixed effects analysis of covariance results for the fixed effects of
inoculum type (Control vs. Fusarium morphotype 1-5), random effect of bench
and covariate (initial seed mass) on Q. rubra seedling mass and stem height. . .161

Appendix S. Experimental design and allocation of tropical seedlings to
treatments ................................................................................... 163

Appendix T. Hazards ratios for the effects of light availability (1% vs. 5% full
sun), soil extract (non-sterile vs. sterile), soil source (tree species culturing soil)
and initial seed mass on seedling mortality estimated by the Cox regression

model ....................................................................................... 164

Appendix U. Split-plot analysis of covariance results for the effects of light
availability (1% vs. 5% full sun), soil extract (non-sterile vs. sterile), soil source
(tree species culturing soil), bench and covariate (initial seed mass) on seedling
mass (mg) for each species. ............................................................. 166

LITERATURE CITED ........................................................................... 170

xi

LIST OF TABLES

Table 1.1. Review of relevant published studies on J anzen—Connell processes in tropical
and temperate tree species and forest communities ............................................. 18

Table 2.1. Local community characteristics of the 21 tropical tree species used to assess
the prevalence and patterns of soil-pathogen induced mortality .............................. 46

Table 2.2. List of significant seedling responses to microbial vs. sterilized microbial
extract for all study species (condensed from Appendices A-D) ............................. 48

Table 3.1. List of significant seedling life span, total mass and integrated seedling
performance responses for all study species to: soil source (tree species culturing soil),
soil extract (non-sterile vs. sterile), and irradiance (2% vs. 22% full sun) (condensed fiom
Appendices I-M and Table 3.2). ............................................... . .................. 76

Table 3.2. Reciprocal effects (percent difference in integrated seedling performance
[(mean total mass x mean life span) / (days of experiment)]) of plant-soil feedbacks
relative to a tap water control for each study species integrated across extract treatment
and at a) high and b) low irradiance levels ....................................................... 78

Table 4.1. List of significant seedling life span and total mass responses for all study
species to: soil source (tree species culturing soil), soil extract (non-sterile vs. sterile), and
irradiance (1% vs. 5% full sun) (condensed from Appendices T & U and Table

4.2) ................................................................................................... 108

Table 4.2. Reciprocal effects (percent difference in integrated seedling performance
[(mean total mass x mean life span) / (days of experiment)]) of plant-soil feedbacks for
each study species integrated across extract treatment and irradiance level ............... 111

xii

LIST OF FIGURES

Figure 2.1. Effect size of soil pathogens on (A) life span and (B) total mass (magnitude
and sign of the effect of the microbial vs. control treatment) compared to categories of
local community characteristics for 21 tropical tree species. Data points show means

and bars show 95% CI ranges for all studies (Overall) as well as each study category.
Sample sizes and categories are indicated on the x-axis; the dotted line shows Hedges’ (1,
indicating the absence of an effect. Significance is shown as: NS, not significant; * P 5
0.05; *** P g 0.001 ................................................................................. 49

Figure 2.2. Relationship between seedling shade intolerance and change in performance
between seedlings in the microbial vs. control treatments for 21 tropical tree species.
Shade intolerance is measured as percent mortality at 1% full sun and zero conspecific
seedling density. Change in performance is measured as ([mean total mass]X[mean life
span])/(number of days in experiment). The significance value is from multiple, stepwise
regression ............................................................................................. 50

Figure 3.1. Fig. 1) Survival curves for F. americana seedlings by (a) soil source (tree
species culturing soil: Qr = Q. rubra, Ar = A. rubrum, As = A. saccharum and Fa = F.
americana) and (b) soil microbial treatment (sterile vs. non-sterile extract). Survival
curves end at 58 days since no F. americana seedlings died after that date ............... 79

Figure 3.2. Total final mass by soil source (tree species culturing soil) and extract (sterile
vs. non-sterile) for each study species in low and high light (A. rubrum (Ar) a-b, A.
saccharum (As) c-d, F. americana (Fa) e-f and Q. rubra (Qr) g-h). Dotted line represents
seedling mass when grown with tap water and is shown for reference only, and were not
included in statistical analysis ...................................................................... 80

Figure. 3.3. Relationship between chemical [sterile extract / tap water] and microbial
[(un-sterile extract / tap water) — (sterile / tap water)] effects in soil extracts “cultured” by
different species of adult on seedling performance [(mean total mass x mean life span) /
(days of experiment)]) for each study species in high and low light. Ar = A. rubrum, As =
A. saccharum, Fa = F. americana, and Qr = Q. rubra. Bootstrap devised 95% CI
included. .............................................................................................. 84

xiii

Figure 3.4. Change in performance [(mean total mass x mean life span) / Days in
Experiment] between seedlings in the sterile control versus each of the F usarium
morphotypes for each study species. Ar= A. rubrum, F a F. americana and Qr= Q.
rubra ................................................................................................... 86

Figure 4.1. Survival curves for study species [a) Apeiba membranacea, b) Colubrina
spinosa, c) Iriartea deltoidea, and d) Prestoea decurrens] with significant affect of soil
source (tree species culturing soil: Am = Apeiba membranacea, Cs = Colubrina spinosa,
Id = Iriartea deltoidea, Pm = Pentaclethra macroloba, Pd = Prestoea decurrens and Vk =
Virola koschnyi) and integrated across irradiance level and soil microbial treatment.
Arrows indicate seedling response to conspecific cultured soil source ..................... 113

Figure 4.2. Total final mass by soil source (tree species culturing soil) and extract (sterile
vs. non-sterile) for each study species in low and high light (Apez'ba membranacea (Am)
a-b, Colubrina spinosa (Cs) c-d, Iriartea deltoidea (Ia) e-f, Pentaclethra macroloba (Pm)
g-h, Prestoea decurrens (Pd) seedling mass without cotyledon mass i-j and seedling mass
with cotyledon mass k-l, and Virola koschnyi (Vk) m-n). Dotted line represents seedling
mass when grown with tap water and is shown for reference only, and were not included
in statistical analysis .............................................................................. 114

Figure 4.3. Survival curve for Pentaclethra macroloba seedlings by soil. microbial
treatment (sterile vs. non-sterile extract) and integrated across soil source and irradiance
level .................................................................................................. 120

Figure 4.4. Survival curves for study species [a) Apeiba membranacea, b) Colubrina
spinosa, c) Pentaclethra macroloba, and d) Prestoea decurrens] with significant effect of
light level (high light; 5% full sun vs. low light; 1% full sun) and integrated across soil
source and soil microbial treatment ............................................................. 121

Figure 4.5. Relationship between chemical [sterile extract / tap water] and microbial [(un-
sterile extract / tap water) — (sterile / tap water)] effects in soil extracts “cultured” by
different tree species on seedling performance [(mean total mass x mean life span) / (days
of experiment)]) for each study species integrated across irradiance levels, except for C.
spinosa and P. decurrens seedlings whose results are only from high light. Am = Apeiba
membranacea, Cs = Colubrina spinosa, Id = Iriartea deltoidea, Pm = Pentaclethra
macroloba, Pd = Prestoea decurrens and Vk = Virola koschnyi). Bootstrap devised 95%
CI included. ....................................................................................... 122

xiv

CHAPTER ONE

INTRODUCTION

Identifying the mechanisms that maintain species richness is a central question in
plant community ecology. Under the competitive exclusion principle, competitively
superior species exclude inferior species if competition for resources remains unchecked
(Gause 1934). Although niche partitioning is often invoked as an explanation for species
coexistence, most plants require the same resources. How then are there so many species
rich forests, many in the tropics containing more than one hundred species per hectare?
Out of the vast array of hypotheses that have been proposed (Palmer 1994), one of the
most influential was put forth independently by Janzen (1970) and Connell (1971)
(hereafter referred to as the J -C Model). They proposed that host-specific natural enemies
could maintain high tree diversity of tropical forests by reducing seed and/or seedling
survivorship near conspecific adults and/or at high conspecific densities. Such non-
competitive distance or density-dependent (N CDD) mortality would favor establishment
of heterospecific individuals, thus promoting species coexistence.

Although the J anzen-Connell Model has produced a vast body of literature of both
empirical and theoretical studies, there is still contention over the importance of NCDD

mortality in tropical forest community dynamics (detractors: Hubbell 1979 & 1980,

Connell et a1. 1984, Hubbell and Foster 1986, Hubbell et a1. 1990, Welden et a1. 1991,
Condit et a1. 1992, He et a1. 1997, Hyatt et a1. 2003; supporters: Wills et al. 1997, Webb
and Peart 1999, Wills and Condit 1999, Harms et a1. 2000, Wright 2002, Peters 2003). In
addition, the focus on explaining the maintenance of tropical. diversity has detracted from
testing whether similar processes are operating in temperate forests. Among the few
studies conducted in temperate forests, there is evidence of distance and density-
dependent tree seedling mortality (Packer and Clay 2000, Hille Ris Lambers et a1. 2002),
but it is not yet clear how tropical and temperate forests may differ in this regard.

Most studies on distance and density-dependent tree seedling survivorship are
structured to test whether spatial patterns of tree seedling recruitment are consistent with
the J -C Model. Conflicting results among the above studies could be due to the simple
fact that patterns of seedling recruitment arise from several different mechanisms, which
may preclude straightforward interpretations of spatial patterns. For example, in a
comparative review of seed and seedling performance at near and far distances from
conspecific adults, only 2 of 27 studies involving a vertebrate herbivore but 15 of 19
studies involving an insect herbivore showed negative effects of being close to a
conspecific adult (Hammond & Brown 1998).

My aim in this chapter is to outline the major assumptions of the Janzen-Connell
Model and to demonstrate how the J -C Model can be placed within the larger context of
negative feedbacks between plants and the soil in which they grow. Specifically my
dissertation research was designed to examine: 1) the mechanism underlying non-
competitive distance and density-dependent mortality and/or reduced growth; 2)

advantages to heterospecific and disadvantages to conspecific recruitment, 3) differences

between common and rare species in sensitivity to J -C processes, 4) the strength of J -C
processes in tropical versus temperate forests, 5) and the interactions of J -C processes
with light availability.

ELant-Soil Feedbjflgs - Individual plants not only use resources in the
environment for their survival and growth but they interact and can change the
environment in which they live. In particular, there is growing appreciation for the
potential interactions that can occur between plants and their soil environment. Plant-soil
feedbacks could be considered a two-step process: 1) the plant or population changes the
soil community, and 2) the soil community affects plant survival and/or growth.
Feedbacks can be either positive or negative but negative feedbacks can allow for plant
species coexistence by reducing establishment, growth or reproduction of individuals
under their parent plants, thereby allowing for heterospecific establishment to occur.

Although soil pathogens (e. g. firngi, bacteria and/or nematodes) have a long
history of study in forestry and horticulture, their potential role as a meChanism of NCDD
mortality in the dynamics of natural plant communities has only recently been recognized
(Gilbert and Hubbell 1996, Packer and Clay 2000, 2003, Hood et a1. 2004, Bell et a1.
2006). For instance, many of these pathogens show strong host specialization, short
generation times, high fecundity, long persistence in soil, and more limited dispersal than
their hosts (Agrios 1997, Gilbert 2002). These characteristics could underpin the
mechanisms that create negative feedbacks when individual plants “culture” the soil
microbial community in which they grow to the detriment of themselves and other
conspecific individuals (van der Putten et a1. 1993, Bever 1994, Mills & Bever 1998,

KJironomos 2002).

There are other important plant-soil feedbacks that are not mediated by soil
pathogens (Ehrenfeld et a1. 2005). For example, the presence of a particular plant species
could be associated with the formation of mycorrhizal networks (Booth 2004), production
of allelochemicals (Stinson et al 2006), alterations to soil physical properties (Rillig et a1.
2002) and nutrient availability (Finzi et a1. 1998a—b). All of these feedbacks could impact
seedling performance in a species-specific manner and result in distance and density-
dependent mortality and/or reduced growth. That is, a particular species could modify the
soil to the detriment of heterospecific vs. conspecific seedlings. However, soil pathogens
can be host-specific and are more likely to specialize on common species, two attributes
that make them ideal for creating NCDD processes.

Distance and Density-Dependent Processes - Traditionally in studies of NCDD
processes, all heterospecific individuals have been lumped into a single heterospecific
category (e. g. “far” distance) (Augspurger 1983a-b, Packer and Clay 2000 and 2003;
Hood et a1 2004). However, tree species vary in many characteristics (e.g., resource
allocation to defense vs. growth) and soil-mediated effects of mature individuals on
seedlings could be species-specific as well. In this dissertation I investigate species-
specific effects of conspecific vs. heterospecific individuals rather than ‘near’ vs. ‘far’
categories.

I will be defining density-dependent mortality, in this dissertation, as a feedback
operating between a natural enemy and the density of seedlings within a specific area.
Some studies consider juvenile mortality as a function of adult conspecific density
(Connell et a1 1984, Welden et a1. 1991, Webb and Peart 1999). Choosing between

seedling vs. adult abundance as the predictive variable for seedling mortality should

reflect the spatial scale that is most relevant for natural enemy-tree interactions. In
addition, tree and soil pathogen interactions occur at small spatial scales so seedling
density likely is a more relevant metric than adult density when soil pathogens are the
agent of mortality. I

Although distance and density-dependent mortality is at the core of the J -C
Model, determining the relative importance of these processes is difficult because seed
density is inversely related to distance from adult in many tree species. Most studies have
not distinguished between distance and density effects (Table 1). This may be due to both
mechanisms operating simultaneously or because they may operate on different life
history stages. One way to investigate these differences is to compare how seedlings
respond to soil micro-organisms cultured by adults vs. seedlings since each life history
stage may culture unique enemies or may impact the abundance or virulence of the same
enemies in a different way (Gilbert 2002). For instance, adult trees can act as a reservoir
for host-specific soil pathogens that can kill seedlings. Seedlings likely culture soil
pathogens that have long-lived resting spores and/or are also saprophytic since seedlings
have patchy spatial distributions and relatively short life span. There is some evidence
that seedlings themselves can have an impact on the biota of the soil community. Packer
and Clay (2003) found that after ~4 months of Prunus serotina seedlings interacting with
forest soil at low plant density the feedback between seedlings and the soil microbial

community changed fi'om positive to negative.

It is important to distinguish between effects of adults vs. seedlings on disease
population dynamics for a few reasons. First, the increase in mortality often associated

with high seedling density may have less to do with natural enemies than simple seedling

intra-specific competition. Pathogens may interact with high density simply through
‘self-thinning’ (intraspecific competition resulting in ‘stressed’ seedlings that are more
susceptible to infection). Alternatively, high density may increase the negative feedback
between seedling and soil firngi through increasing fungal populations and disease
transmission. Also, not all species have the majority of their progeny dispersed
underneath their crown. Seeds that are dispersed by birds or other mammals can have
high density far from the parent tree, so it is important to determine how effective NCDD
processes are on these species. Temperate species are often wind dispersed but many
tropical species are bird/mammal dispersed and may be more susceptible to areas “far”
from conspecific adult but with high conspecific seedling density (Clark et a1. 1999). By
separating out the impact that both adult and seedlings have on natural enemy
populations a more mechanistic perspective of the J -C Model can emerge.

Host Specificity — Another assumption of the model is that the natural enemies
causing NCDD are host-specific (J anzen 1970). If all species were equally vulnerable to
natural enemies then there would be a reduction in successful recruitment for all species
and coexistence would not be maintained. The more generalized the natural enemy, the
weaker these processes are in maintaining species diversity. Likewise, pathogen dispersal
distance must be more restricted than its host distribution (Gilbert 2002). If a species does
have a natural enemy that is highly host-specific but dispersal for both the host plants and
the pathogen overlap then host plant recruitment will be constrained across the host’s
range and may result in local extinction.

However, there is limited knowledge of both host-specificity and dispersal for

most soil pathogens in forests. An analysis of the polypore (Aphyllophorales) community

in a tropical forest in Panama revealed that the most common species were generalists
(Gilbert et al. unpublished data cited in Gilbert & Hubbell 1996). However, Augspurger
and Wilkinson (2007) have demonstrated that Pythium, a common soil pathogen of tree
seedlings, varies in pathogenicity among seedlings of different tropical tree species but
does not show strict host-specificity. Thus, this intermediate level of specificity suggests
that Pythium spp. have the potential to have some effect on forest community structure
and diversity. There also appears to be localized adaptations between natural enemies
(insect herbivores at least) and hosts in tropical forests where specialization occurs at the
level of individual reproductive adults resulting in only non-progeny seedlings surviving
under the crown of these adults (Langenheim & Stubblebine 1983, Sanchez-Hidalgo et a1.
1999). There is preliminary evidence that the pathogen that causes NCDD mortality in
black cherry (Prunus serotina) may be host-specific (Packer and Clay 2000). This is the
only study currently published in the J-C literature where the mechanism of distance and
density-dependent mortality patterns has been determined and the natural enemy
exhibited host-specificity (Table 1).

Lack of widespread documentation for host-specificity in research of the J-C
Model reflects a gap in our investigations rather than proof that host-specificity is rare in
natural communities. Strict host-specialization of natural enemies, along with their more
limited dispersal, would likely result in the most effective J -C process leading to species
coexistence. There is a clear need to begin conducting research on the host-specificity
and dispersal range of natural enemies in order to link J -C patterns in a particular species

to species coexistence.

Species’ Abundance - Maintaining species diversity via NCDD responses requires
that these processes are more prevalent in species that are common versus those that are
rare, thereby constraining the abundance of Common species. Rare species may “escape”
NCDD mortality because they have fewer specialist enemies due to their low abundance
and unpredictable distributions in time and space (comparable to ‘apparency’ theory for
herbivory; Feeny 1976 and Rhoades & Cates 1976). Conversely, common species, more
available as hosts, could be disproportionately targeted by enemies, leading to the
community compensatory trend posited by Connell (1971 & 1978). Thus, rare species
would have lower pathogen loads than common species regardless of distance or density
fi‘om conspecifics. Alternatively, rarity could be an advantage, regardless of the strength
of NCDD processes, simply because rare species are less likely to encounter areas
“cultured” by conspecifics than common species. Both scenarios would operate to
promote species coexistence, but their distinction has not been widely recognized. It is
also possible that rare species experience stronger NCDD mortality due to pathogens than
common species, as supported in grasslands (Klironomos 2002), providing an
explanation for species rarity.

In forests, it remains to be elucidated whether natural enemies target common
species and keep their populations in check or keep rare species rare. Few studies have
explicitly compared the incidence and strength of density-dependent mortality patterns
between common and rare species. There is conflicting evidence on the relationship
between species abundance and strength of NCCD among the few studies that have done
this type of comparison. Negative effects of distance or density on survivorship and/or

growth sometimes are reported to be more severe in common vs. rare species (Wills et a1.

1997, Webb & Peart 1999, Wills et al. 2006), sometimes the converse (He et al. 1997,
Hubbell et al. 2001, Ahumada et al. 2004), and sometimes are pervasive with no
relationship to local species abundance (Harms et al. 2000, Peters 2003). It is important
to note that most studies have investigated only common species because of inherently
low sample sizes in rare species (Wills & Condit 1999, Wills et a1. 2004). These
conflicting observational results, occasionally from the same study site and investigators,
provide strong motivation for experimentally testing specific mechanisms of NCDD and
their relationship to species abundance.

It is important to note that there are two very different approaches to testing the
relationship between NCDD processes and species’ abundance within forest community
ecology. First, many studies compare the relationship between the average population
level mortality of individuals and a species’ abundance within that local community (e.g.
Connell et al. 1984, He et al. 1997, Welden et a1. 1991, Webb and Peart 1999, Peters
2003). There are two problems with this approach: 1) these studies do not test the actual
mechanism for mortality (i.e. vertebrate predator, insect herbivore or pathogens), and 2)
all of these studies investigate the relationship between species abundance and NCDD
using current species abundance which is a static metric. A preferable approach would be
to link trajectories of tree species abundance in a community to disease pressure through
time (i.e. do common species become less common due to increased pressure and vice
versa). The second approach is to investigate performance for single species among plots
characterized by the density of the focal species (e.g. Hubbell and Foster 1986, Condit et
al. 1994, Gilbert et al. 1994, Silva Matos et al. 1999, Harms 2000). The problem with this

approach is that determining negative density-dependent performance for an individual

species does not necessarily mean that species coexistence can be maintained at the
community level. Common species (based upon abundance at the community level) still
need to be at a disadvantage in comparison to rarer species for species diversity to be
maintained. This disadvantage could be due to NCDD processes only occurring in
common species or if NCDD processes occur regardless of species abundance because
common species are more likely to encounter areas near or at high density of conspecific
individuals.

Temperate vs. Tropicaflorests — It is often assumed, but rarely tested, that
mechanisms underlying forest dynamics in tropical vs. temperate forests are different.
For instance, J anzen (1970) proposed that distance and density-dependent mortality
would be greater in tropical vs. temperate forests because warm temperatures, greater
rainfall and aseasonality in tropical forests would result in both a higher abundance of
natural enemies and a greater proportion of specialist to generalist natural enemies.
Givnish (1988) expanded on this idea, and proposed that increases in rainfall and soil
fertility as well as decreased seasonality would not only favor herbivores and pathogens
but should decrease plant investment in defenses against these natural enemies. However,
Gilbert (1995 and 2002) has proposed that firngal specificity may actually decrease in
tropical systems because as host diversity increases the selective pressure for
specialization may diminish. For instance, Gilbert et al. (2002) found an inverse
relationship between host-specificity of wood-decay fungi and tree species diversity
when comparing different tropical forests. This relationship between species diversity
and fungal host-specificity might be due to a decreased probability of successfirl

colonization of pathogens as their hosts become rare.

10

For insect herbivores, Basset (1994) showed in feeding trials that tropical insects
have greater host-specialization than temperate species. In addition, tropical leaves
appear to experience more damage from herbivores than their temperate counterparts,
even though tropical leaves tend to be more heavily defended (Coley and Aide 1991). I
am unaware of a direct comparison of the degree of soil pathogen host-specificity and the
effect of these pathogens on tree species between tropical and temperate forests.
However, there is growing evidence that both individual temperate (Packer and Clay
2000) and tropical (Augspurger 1983a—b, 1984, Augspurger and Kelly 1984, Hood et al.
2004, Bell et al. 2006) tree species have higher mortality in the presence of conspecifics
due to soil pathogens. Whether these pathogens exhibit host-specificity is unknown,
although Packer and Clay (2000) suggest that this is the case for Prunus serotina. Hille
Ris Lambers et a1. (2002) also found density-dependent mortality patterns in six of seven
species in a North Carolina temperate forest and when comparing their results to studies
conducted in tropical forests (BCI and Pasoh) concluded that density-dependent mortality
is prevalent in both systems.

Currently there is no clear consensus on whether the J -C Model operates similarly
in temperate and tropical forests even beyond our limited knowledge of host-specificity
of natural enemies in both biomes. Investigating the strength and importance of the J —C
Model for maintaining species coexistence in both temperate and tropical forests also
requires testing the following predictions in both biomes: 1) non-competitive distance
and density-dependent processes decrease seedling performance of common species more
than rare species, 2) through NCDD processes, the spatial distribution of species will

become less clumped through time, 3) there should be greater recruitment of

11

heterospecific individuals due to NCDD effects on seedlings near conspecific adults, and
4) over time, species diversity should be greater than expected with NCDD processes
compared with random survival of seeds and seedlings.

Ligtflmd Disease Imam - Disease in the early life history stages may play an
important role in maintaining species coexistence through distance and density-dependent
. mortality and through heightening light gradient partitioning. These two theories often
have been investigated separately or as competing mechanisms for species coexistence
(Itoh et al. 1997, Kobe 1999), but this dichotomy is likely more conceptual than
biological.

Light availability may mediate disease induced NCDD processes. Seeds (Dalling
2004, O’Hanlon-Manners and Kotanen 2004) and seedlings (Augspurger 1983b,
Augspurger 1984, Augspurger & Kelly 1984, Kitajima and Augspurger 1989, Hood et al.
2004) of some species experience higher disease related mortality at low than high light
in both shade house and field environments. Increased light availability. could interact
with pathogen infection in at least three ways, as summarized by Augspurger (1990): 1)
unfavorable conditions (e. g. increased temperature, decreased soil moisture and/or an
absence of conspecific adult acting as a reservoir) that lower pathogen abundance, 2)
seedlings accumulating mass more rapidly to compensate for tissue lost to disease,
whereas the same tissue lost in shade might result in death, 3) seedlings reducing
exposure to disease through faster lignification rates and/or growth to an invulnerable
size (Niinemets & Kull 1998, Seiwa 1998), and 4) seedlings are protected from disease

due to increased AMF mycorrhizal colonization in high light habitats (Lovelock & Miller

12

2002, Gehring 2003, Garnage et al. 2004, Gehring 2004) which in turn suppresses disease
(Newsham et al. 1995, Borowicz 2001).

Interspecific differences between growth and/or survival in varying light
environments result in species segregating into dominance at different light levels
(Hubbell and Foster 1992, Kitajima 1994, Kobe et al. 1995, Pacala et al. 1996, Kobe
1999) which can result in species coexistence. Historically, light gradient partitioning has
been viewed primarily based on plant carbon balance due solely to photosynthetic gain
vs. respiration cost (Baltzer and Thomas 2007). However, the interaction between light
and pathogens, as mentioned previously, may have a large impact on survival and growth
that needs to be considered. An inverse relationship between high light growth and low
light survivorship has been documented for many temperate and tropical species that
correspond to shade tolerance classifications (Kitajima 1994, Pacala et al. 1994, Kobe et
al. 1995 , Kobe 1999, Walters and Reich 2000). Species susceptibility and/or response to
disease may contribute to this trade-off. For instance, shade intolerant species tend to
invest in traits that maximize growth (Herms & Mattson 1992, Reich et al. 1998, Walters
& Reich 1999) while shade tolerant species invest more in functions that enhance
survivorship, such as defense against natural enemies (Coley et a1. 1985, Coley & Barone
1996) and carbohydrate storage (Kobe 1997, Myers and Kitaj ima 2007). Additionally,
disease susceptibility has been reported to correlate negatively to shade tolerance
classifications (Augspurger and Kelly 1984). Thus, classification of some tree species as
shade tolerant or intolerant may have less to do with photosynthesis and respiration in
low light than with characteristics that regulate the loss of tissue to all agents, including

pathogens (Walters and Reich 1999).

13

Some of the pathways discussed previously for reduced disease pressure in high
light environments may preferentially benefit shade intolerant species in high light. Shade
intolerant species can increase photosynthesis more effectively in high light than shade
tolerant species and thus they may be able to accumulate mass more rapidly to
compensate for disease (Walters et al. 1993, Kitajima 1994). In addition, data from
tropical rain forests suggest that early successional (i.e. shade intolerant) species exhibit
higher colonization rates and degree of positive growth responses to mycorrhizal
infection than late successional (i.e. shade tolerant) species (Siqueira et al. 1998, Zangaro
et al. 2003). Since AMF mycorrhizal colonization seems to increase in high light habitats
(Lovelock & Miller 2002, Gehring 2003, Garnage et al. 2004, Gehring 2004) then shade
intolerant species would likely benefit more through AMF protection from diseases than
shade tolerant species.

The interactions between light availability and disease need to be explored further
for a better understanding of the J -C Model and its role in species coexistence. For
instance, shade intolerant species may have greater susceptibility to pathogens than shade
tolerant species which would allow for greater species coexistence through increased
niche differentiation (i.e. exaggerating shade tolerance differences among species). In
addition, shade intolerant species could be more susceptible to pathogen induced NCDD
processes than shade tolerant species. How this would impact the strength of J anzen-
Connell processes, since shade intolerant species are a smaller proportion of the total
community than shade tolerant species, on species coexistence should be investigated.

Conclusion — Many purported tests for the J-C Model have only looked at the

pattern of juvenile mortality but have not determined the mechanism causing that pattern

14

(Table 1). Also, the majority of studies that have looked at the mechanism causing
NCDD have focused on single, common species, and thus do not test that species
coexistence can be maintained in the forest community through these processes. Studies
that have incorporated both common and rare species have often lacked any information
on the mechanism causing J -C patterns and thus any information on host-specificity. It is
critical that we progress from simply determining that distance and density-dependent
mortality exists to investigating whether these patterns originate fi'om host-specific
natural enemies that constrain the abundance of common species but not rare species.
Although finding J -C processes in temperate forests does not negate the importance of
this mechanism for maintaining species coexistence it is still important to compare
between tropical and temperate forests as it increases our knowledge of how communities
are formed and structured. In a global context of deforestation and conversion of forests
to agriculture and plantations, it is not a merely academic exercise to test soil pathogens
as a mechanism for maintaining tree species diversity. For example, plant diseases often
are viewed negatively in conservation reserves or managed forests even though they may
play a role in the population dynamics of their hosts and the structure and diversity of the
communities that they inhabit. Also, without adequate knowledge of plant- pathogen
interaCtions in 'natural' systems, it is more difficult to combat the spread of exotic,
invasive pathogens. Disturbance had long been viewed in ecology as a nuisance to
understanding equilibrium characteristics of ecosystems. Although disturbance is now
viewed as a key mechanism in most ecosystems, disease still is often disregarded. With
further studies, background levels of disease may turn out to be as important a

mechanism as tree fall disturbances.

15

Dissertation outline - My research investigated plant-soil feedbacks caused by
both biotic (F usarium) and abiotic factors (possibly species-induced differences in base
cation availability) in soils that had been cultured by either conspecific or heterospecific
adults and seedlings and the effect on survivorship and growth of focal conspecific
seedlings. Through a series of shadehouse and greenhouse experiments with both tropical
and temperate species, I tested the major assumptions of the J anzen-Connell Model.
Chapter 2 focuses on the prevalence and effect of pathogens, derived from soils cultured
by conspecific adults and seedlings, on a broad survey of 21 tropical tree species. I found
that 9 of the tree species negatively responded to microbial extracts, while 2 species
positively responded, and an additional 2 species experienced opposing reductions and
increases in seedling life span and/or growth. In addition, species’ seedling shade
tolerance not their local abundance co-varies with susceptibility to these micro-
organisms. This chapter is in press in Ecology. In chapter 3 I test whether 4 temperate
tree species experience NCDD mortality and/or reduced growth, and if these processes
are mediated by host-specific soil microbes and influenced by light availability. I found
that species-specific feedbacks between adult trees (not seedlings) and soil (mediated
through chemical mechanisms not soil microbes) influenced life span and/or growth for
all temperate species. Contrary to the J -C hypothesis, however, heterospecific and
conspecific effects had similar prevalence and magnitude of influence on seedling
performance. In addition, soil microbes decreased seedling performance for 3 of the 4
species and this negative effect occurred regardless of light availability for some species
and for others only in high light. This chapter is in review for publication in Ecology.

Chapter 4 describes a parallel experiment to the one conducted in chapter 3 but with 6

l6

tropical tree species. Once again I found that species-specific feedbacks between adult
trees and soil (mediated through chemical mechanisms not soil microbes) influenced life
span and/or growth for all species. Feedbacks from conspecific adults reduced seedling
performance relative to all heterospecific adult effects for three out of six tropical tree
species, and an additional two species had decreased performance in conspecific vs. two
or more heterospecific cultured extracts, supporting the J -C hypothesis. In addition,
conspecific seedlings were more likely to be disadvantaged versus heterospecific seedling
in soils influenced by a given species. Differences in NCDD processes between
temperate and tropical forests are discussed. This chapter will be revised in preparation
for review by Ecology. Finally, in chapter 5, I discuss the implications of this research
for both temperate and tropical forest community dynamics and suggest avenues of

research that may answer questions raised by this study.

17

Hugo Hp. W039! 0». 8—38: vac—area gamma 0: Héashogo: @3838 E "833— 85 85388 3.8 .6098 BE mega
888538. A+v n 3538 Seaman" as? 35598.? 3 n ciao—boo 58:38:" in. rvfiofirommm. ._. u Bong—Ema o».

303:5. 96:0qu 8&8 E wank. OOH u 838:st ooBuosmmSQ Rana.

 

 

 

 

 

 

 

33.3538.— gowns—5.
5:835
manage Ala—.55 Ema Ema—d. 6255. Eng—.8 09. 38—552: 2. =8? meg—.8
Mama Zena—5. 2.8565.
$3333.23 «Emma. moon + <onoc~w8 Z >
93mm» "3303 38m". . on . ammufimoa
wash—av moon—5m + wwwromg €365” oxen—Son Emu»
o
Eocfioe‘mzi «Emma ...
35mm» c.038— monomr moon—Em + + mewmoa 22 . >=mmvc~ma~
magma—fl AngmomO 3553 Ex?
Eocfiofits «Names.» .
35mm” #638: mean? magnum + + 390%: 39835- 22 . >=mmufimon
3:253 03 389:3 Ems.»
5 €593mua88
€098 . + a: m .. magma: >emmv=~mo~ ma
Meadow" ”0E3— ?er wade—um mp6 Agumsmomv 20 WE? Home.
BBB»
o €Ea-&m_oo~mom
muoomom . Hum—90mm: EBB—VFW 22 >emmv=~man
A823 :8on 83m". mooESm + N3 $6 + uG mg 03 $8538 Sm?
H.353
. _. - + . . .
9:85 355:». Goa moon Sm <59 commune—mo Zo _ coo

“Egg—é

235

18

Sea 2. 8::

 

 

 

 

 

 

maven—.882. 328—3.
5:835

menace Ala—:8: E8 Image—Q €2.me Ema—=8 788—552: on :8? meg—.8

manna Zola—3. mfiammaa.
mefiaaae Sammofifia. ...
35mm" "333— moaamr mag—Em + AQWWENWME ”he" . on m8: 9 a: Noom
wozwov u m BE:

. ... $9.238 85 no»

8 M886 anon—8m . + N3 m3 For m" Bone—3.“ 23 9:58: an
388%. 85 88%. mew—Em €8ka + m3 mg 393 83803. 38m 3553 Hugh Gem
38823 omooc Ba momma 83503
moxmmakogmg
“5:35:95. . . Zen wanna»: an
8858— 85 88%. moon—Sm - - $83.58 88858 58% 88
mos? spacey
\MMHMMNNWS mood + - A t» SE _.. 5830688

moon—Em - + @5839 68:3 22
P. 86% quxazSo 98353 05:.» 33

h . . mama + + + <onoc~w8

A393 ”3an manna". .

mama—am - -
moo—.5
US$48“ unanimahw .
Q338— €o~ 88¢”. mag—mam + + WoflMWMHMVGOmmHEa ”“8an WWW mu 9me

How—58$

l9

288 2.2. A025

 

 

 

 

 

25.2.8289 32.8.5.
$2833..

magma 98283 58 E221 .8:qu 5238 one 38—533— en =92. mac—.8

m88 32.8.3. macaw—.85.

>22 .3 H + -
23933 6828335. A303: 8: v

>8: + +

8:278: .

. . 22 ox . on 22 . 0252 2 ”—

bmmiohua. finanamaa. >95 $85.58 50s.
A822 2322: 882. . + +
3.2525 ABonEggs

>92 - -

Amaze—V
DERGR finaafimaflw
A322 ”3682 2.2.82. magnum + C2838 22 . U8 M88: 8
wen—BB8 22:58 38 Ems.
w 938-8882 .88
£828 22 Ca m8<8= an
3.22 :88»— 882. moon + Cu m3 Curses: 8x828; film—2 Noow
552:2

A82 222

moon 828:3. ... 8828688

MSEMNHmamaw W98“ “2 Em: + @3835 828$ 22
$8 228 28 2:. 39? . mammOmo $3
mamas MM Ewam a... £8: 2 + ... <onocnm8m 322:8
a m me @8022

22283

 

20

H88 .... A02...

 

 

 

 

 

 

 

 

33.62.8882. 32.8.3.
5.22.28
M828 $8.28.. :8 E203. 5285. 58828 on... 3.8.8222 cm .82. M288
m88 32.8.5. $82.85.
20 22:2.
9828 8.2.8.. 8wa 2. .
A222 "3.22.. 882. $6.25 + + .. 52.8%: 22: 802 2:2 9:82 2 8..
25.85 .02.
.8me». 82888
28082
088a i=8. .
A222 "3.28. 882. 82.58 + C302: 82.252. 22 . 9:8: 2 2.
32.28 3808:. 822222. moo.
mm 2.52;. 2828 mag . 22 .838 2 2.
.222 R028. .282. 82.55 + 22 32252. 952.52. N25
.8258.
q 9.02... 2822.. 82. + M... £8 + uh 8.... 22 .38 ”8
225.888 882. 22 982.22. . .8288 2 8..
:25 mgeam + 3 .8 + 3 as. 9558 88
mag. 22 282.22.
S $.02? «.828 + 2. .0 mg 3.21.2.2. 2. 22 ~58 .mm
82888 882. 82:88.82. :2 32252. 5.388 an
695 magnum - :3 m3 208.5. 0.me No8
282522..
.5288 8%.“. _._
.222. 82.58.8028 82.58 + awgwmg Z2 . 2. Woo“. 2 8..
.282. 0:28. A .8563 9825
3.38 9.5.2285...» <28688 c.2858.
.222 23.888. .282. 82. + 882888 22 . .828 2 8..
2.2252. .03

.8258.

A2838.

 

21

5% Z. Gav

 

 

 

 

 

 

 

flwucnrommua 32.95%
5.22.3.
muanmom Ala—.55 Ema 532.! 53me Ema—=8 00,—. 3855“,...— c.. :8... mac—.8
m8? . Zena—3 $.85an

38258
Gowaauhclms 5 0.9.35. . W2: 3 E.
995388 moan”. Cm>v moon Ea Ex 20" Q on 20 Noon

. anon

menace
MESS 83:6.»me . 255% 85
@3388 38m". moon—Em + + ”fireman Aguim- who” . 8 wammENES
magma meow
REQE «8&8 3Q .
man—Tagacosm manna”. magmam - 553888 22 . 28—55 a” a.
0558 Curioczan m3; ESE—:8 Boo

. . + @83 muaommmmfi Bo:—

M§m~§5 San‘ohafin . . . 20H Zanmrmcon 9
3°38. ”can". wuss moésm a8 om afiafima £5.58 a. 88

35an horohrnmav

<8
3435. .8335 moan—=5 + + cram—n9. * HUM—90mg Q9535: 98:3 ago—Sn an 93
$36088 38$. Owe m 03.83 «29v oiqoaoov Nooo
3 80m" 8:55: :8 ng§
. 353950: om

«woo—am v we 0:. 0w: :3 + Q: + - . ._ H 22 $3.05 an
25 "333: 33a. :3 - mi - .552 .am so 9558 EB: 38
>83:me Boa—5%

magmamfiflc

22

.965 _.~. 839

 

 

 

 

 

 

 

yaw—39832. 3919:?
Hana—.3
magma Ala—:25 Ema Ems—Q 62.me Ewan—.8 one zanwaumma a. :8... manna
m8? . 33.3.3 @8383
mg 28% .6038 + NmuGfi
. >3: 329.:
3.0208 manna. 7355 m3 mammoma
g mm $52? $3.98 22 9358 “WWW“? wanna noow
A505» flow—ow— monamr >95 + m3 _ mm 8&2:
Hymn—~83 mum
mama EmathSE. .
908.9088 383. magnum + + 331338 $58 We. . on WWW“ Noam
wanna
fiaxafima onnfiminmu mag + <anodnm8 20H
Mun—2% Mow—ow— moRmr mama—E - 683330: 93358 mag? 3mm
”58: m <mnnc38
A $598 c». 03335. mood + wk mg I fin «Eu WMHWwawsg 22 mEamS mu
936038 med? m 38:53 memmENawm
a 9: mafia: + #3 + NB m . Gem
mu v m £6 E Gagsméa
NENQN was? $93. .
coca—58 H338. magnum + + Zoe 88358 ””8an WMJWWWSm R
woman. wBNE .
Sagas: 8:383 333338 22
3Q 3%on manna. mac—Em + Amusinwnrn 988an mags: meow
088 803 fiwogmamamhmv

 

23

add—a Hg. 3::

 

 

 

 

 

 

 

 

mum—59533.. 732.3:5~
5:33;
magma A5535 E? 532% 6255. E358 OOH 28—532: 3. flow? mas—.8
m9.» 32.35% $.85an
Mmmgaomnmhmwofl moan + <03¢¢88 22 H055. a" a.
Hug—”v . . + 335%: 0x558 $8
@5QO 3:5. <anoc-8 Aaomgc
A23 "339: ”.38". moon - 533338 who” . an “WWW—M88 a
wage Ago—«mama 68:3 .
73 £83 $898
3883—? 3 383. moan—Em + + C3823 2o" . sag an wan:
Facsoamv axed—=3 Goo
mg £83 ".6098 .
35mm" "338.— moamr W ~03 me + «Q? + Cup—Boss 22 . $35 on a.
55ng m3. 355$ 53
woo £003 avgmam .
Guam" monomr 3:55» W ~95 me + 9:902: 22 . $55?
.W 33935 935.52— Oosa: $3
Mormfimn mega? .
35mm" "SEQ: moan". moon + 933388 Adz—0E9 29 . film? 3mm
wgfismv 68:8 SSS—gm
N‘malma mmtoflmn wag—3338 @3835
Mama + on" ov
Ah‘onndii 22 c552 an
3:332»: moon + M3888 68°35 oxaaaoa Q55: N8...
355” now—om. manna. Hammocan 938—qu

mafia

68:8

 

24

CHAPTER TWO

TOLERANCE OF SOIL PATHOGENS CO-VARIES WITH SHADE TOLERANCE

ACROSS SPECIES OF TROPICAL TREE SEEDLINGS

Abstract

A negative feedback between local abundance and natural enemies could contribute to
maintaining tree species diversity by constraining population growth of common species.
Soil pathogens could be an important mechanism of such noncompetitive distance and
density-dependent (N CDD) mortality, but susceptibility to local pathogens may be
ameliorated by a life history strategy that favors survivorship. In a shade-house
experiment (1% full sun), we tested seedling life span, growth, and mass allocation
responses to microbial extract filtered fi'om conspecific-cultured soil in 21 tree species
that varied in abundance and shade tolerance in a wet tropical forest (La Selva Biological
Station, Costa Rica). Forty-three percent of the species had significant reductions and
10% of the species significant increases in life span, growth, root length, or root surface
area when inoculated with microbial extract; 10% of the species experienced opposing
reductions and increases in these characteristics. Contrary to expectation, species’ local
abundance was not related to species-specific responses to microbial extracts from
cultured soils. Across species, seedling shade tolerance (survival at 1% full sun) was

negatively correlated with susceptibility to the microbial treatment for both survival and

25

total mass accumulation, thereby exaggerating shade tolerance differences among
species. Thus, soil pathogens may contribute to species coexistence through heightening

niche differentiation rather than through negative density dependence in common species.

Key words: common vs. rare species; density dependence; Janzen-Comte]!
hypothesis; plant—soil feedback; shade tolerance; soil pathogens; species coexistence;

tropical forests.

Introduction

Identifying the mechanisms that maintain tree species richness is a central
question in plant community ecology. Under the competitive-exclusion principle,
competitively superior species exclude inferior species in the absence of niche
partitioning (Gause 1934). Although there has been a vast array of hypotheses proposed
for how competitive exclusion can be precluded (Palmer 1994), one of the most
influential was put forth independently by J anzen (1970) and Connell (1971). They
proposed that host—specific natural enemies could maintain high tree diversity of tropical
forests by reducing seed and/or seedling survivorship near conspecific adults and/or at
high conspecific densities. Such noncompetitive distance- or density-dependent (N CDD)
mortality would favor establishment of heterospecific individuals, thus promoting species
coexistence.

Soil pathogens (e. g., fungi, bacteria, and/or nematodes) could be an important
mechanism of NCDD seedling mortality and/or reduced growth because many of these

pathogens show strong host specialization, short generation times, high fecundity, long

26

persistence in soil, and more limited dispersal than their hosts (Agrios 1997, Gilbert
2002). These characteristics could underpin the mechanisms that create negative
feedback when individual plants “culture’ ’ the soil microbial community in which they
grow to the detriment of themselves and other conspecific individuals (van der Putten et
al. 1993, Bever 1994, Mills and Bever 1998, Klironomos 2002). Although soil pathogens
have a long history of study in forestry and horticulture, their potential role as a
mechanism of NCDD mortality in the dynamics of natural plant communities has only
recently been more widely recognized (Gilbert and Hubbell 1996, Packer and Clay 2000,
2003, Hood et al. 2004, Bell et al. 2006).

Maintaining species diversity via NCDD responses requires that these processes
are more prevalent in species that are common vs. those that are rare, thereby
constraining the abundance of common species. There are at least two distinct
mechanisms through which species abundance could influence NCDD responses.

First, rare species may “escape” NCDD mortality because they have fewer specialist
enemies due to their low abundance and unpredictable distributions in time and space
(comparable to “apparency’ ’ theory for herbivory [Feeny 1976, Rhoades and Gates
1976]). From this same view, common species, more available as hosts, could be targeted
disproportionately by enemies, leading to the community compensatory trend posited by
Connell (1971, 1978). Thus common species would be expected to experience greater
impact from enemies than rare species, regardless of distance or density from
conspecifics. Under the second mechanism, rarity could be an advantage, even if rare and
common species experience similar NCDD processes, simply because rare species are

less likely to encounter areas ‘ ‘cultur ” by conspecifics than common species. Both

27

mechanisms would operate to promote species coexistence, but their distinction has not
been recognized. It is also possible that rare species experience stronger NCDD mortality
due to pathogens than common species, as supported in grasslands (Klironomos 2002),
providing an explanation for species rarity.

Among tropical trees, there is conflicting evidence on the relationship between
species abundance and strength of NCDD. Negative effects of distance or density on
survivorship and/or growth sometimes are reported to be more severe in common vs. rare
species (Wills et al. 1997, 2006, Webb and Peart 1999), sometimes the converse (He et
al. 1997 , Hubbell et al. 2001 , Ahumada et al. 2004), and sometimes are pervasive with no
relationship to local species abundance (Harms et al. 2000, Peters 2003). It is important
to note that most studies have investigated only more common species because of
inherently low samples sizes in rare species (Wills and Condit 1999, Wills et a1. 2004).
Nevertheless, these conflicting observational results, occasionally fi'om the same study
site and investigators, provide strong motivation for experimentally testing specific
mechanisms of NCDD and its relationship to species abundance.

Species-specific traits may also influence susceptibility to and/or impact of
disease. For instance, shade-intolerant species tend to invest in traits that maximize
growth (Herms and Mattson 1992, Reich et al. 1998, Walters and Reich 1999) while
shade—tolerant species invest more in functions that enhance survivorship, such as defense
against natural enemies (Coley et al. 1985, Coley and Barone 1996) and carbohydrate
storage (Kobe 1997, Myers and Kitajima 2007). Thus, the expectation is that shade-
intolerant species could be more susceptible to disease than shade-tolerant species

(Augspurger and Kelly 1984). Similarly, species with larger seed mass may offset

28

disease-related losses during germination and establishment (Foster 1986, Armstrong and
Westoby 1993).

The spatial scale of NCDD processes depends upon the biology of the particular
natural enemy and host plant. Because soil-bome pathogens have limited dispersal
(Agrios 1997 , Gilbert 2002), we expect that soil “culturing” operates at a spatial scale
commensurate with the area occupied by an individual canopy tree. From the vantage
point of understanding species coexistence, density-dependence must extend beyond the
scale of a single tree and is most relevant at the local community level.

The purpose of this study was to determine the prevalence and effect of
pathogens, derived from soils cultured by conspecific adults and seedlings, through a
broad survey of tropical tree species. This survey served as the basis for selecting species
that were included in a subsequent experiment focused on the effects of soils cultured by
conspecific vs. heterospecific individuals (McCarthy-Neumann and Kobe, in review;
Chapter 3). In the present study, we tested the following hypotheses: (l) Seedling
survival and growth decrease with soil microbial extract cultured by conspecific adults
and seedlings vs. sterilized extract. (2) A species’ vulnerability to soil pathogens
increases with its abundance. (3) Among species, vulnerability to soil pathogens declines

with increasing shade tolerance.

Materials and Methods
Field site—This research took place at La Selva Biological Station (Sarapiqur'
Region, Costa Rica) operated by the Organization for Tropical Studies. La Selva is a

1510-ha reserve of diverse (400+ tree species), wet, tropical forest receiving ~ 4000 mm

29

of rain annually with a mean armual temperature of 25.88C (Hartshom and Harnmell
1994). Per distributions of focal species, we primarily collected residual soils of volcanic
origin, which are the most common at La Selva (Sollins et al. 1994) and which are
representative of other tropical areas. We collected alluvial soil for the study species
(Castillo, Luehea, and Neea) that occur only under these conditions.

Species—In 10-week-long experiments, undertaken from April 2004 to July
2005, we assessed survivorship and growth responses of seedlings of 21 tree species
(Table 2.1) to soil pathogens in a shade-house experiment. Henceforth, we refer to
species by genus name. We randomly selected species from those encountered in a 5.5-
year field study of natural seedling dynamics (R. K. Kobe and C. F. Vriesendorp,
unpublished manuscript), with species selection stratified across local abundance (based
upon adult and seedling density), local dominance (basal area), seedling shade tolerance,
and dry seed mass (determined from ~ 20 randomly selected seeds with emergent radicles
for each species). Adult abundance was assessed as the density of 3 5 cm dbh individuals
(number/ha) within three 41 x 240 m mapped stands and seedling abundance as mean
standing seedling density over 5.5 years within a l x 200 m belt transect located in the
middle of each stand (R. K. Kobe and C. F. Vriesendorp, unpublished manuscript). We
characterized species shade tolerance as the probability of seedling survival at 1% firll
sun and zero conspecific seedling density, calculated from mortality models that were
calibrated fi'om survival time data of naturally established seedlings in l-m2 quadrats that
were censused every six weeks as part of the same 5.5-year field study (C. F. Vriesendorp
and R. K. Kobe, unpublished manuscript). Light availability was assessed as percent

canopy openness estimated from hemispherical canopy photos for each quadrat measured

30

twice during this period) and density was expressed as the mean conspecific density
experienced by a seedling over its lifetime. By evaluating the mortality models at zero
density, we are removing NCDD effects on seedlings fi'om the estimates of shade
tolerance. Sample sizes for model calibration for the species of interest ranged fi'om 13 to
6051 seedlings (median N = 119 seedlings). Survival analysis and maximum likelihood
techniques were used to estimate the parameters for the survival models, generally
following methods in Kobe (1999). Due to low sample sizes, survival models have not
been developed for four study species. For these species, shade tolerance was interpolated
to our scale based on published low-light mortality (Luehea [R. K. Kobe, unpublished
data], Miconia [Pearson et al. 2003], Sawhnodendron [Guariguata 2000], and T rophis
[Kobe 1999]).

Soil and seed collection—We derived microbial extract from areas predicted to
have the strongest negative effect, i.e., near adults and at high seedling density. For each
species, we removed a 10 cm diameter by 30 cm deep soil core from within 1 m of the
bole of four randomly selected conspecific adults, with a dbh at 3 75th percentile for that
species. For further culturing, each soil core was planted with four conspecific seedlings
at 1% full sun for eight weeks or until all seedlings had died, whichever occurred first.
Soil was stored at 4°C until seeds were available for planting. Seeds were collected
within 10 m of trails throughout La Selva. Seeds were surface sterilized (0.6% NaOCl for
three minutes), rinsed with deionized water and germinated in either ziplock bags with
peat moss or petri dishes with filter paper in partial sun. Prior to planting, seeds were
surface sterilized with NaOCl for 30 s, rinsed with deionized water, air-dried for 15 min,

and weighed. Because seeds were unavailable, we used recently germinated Coussarea

31

and Vochysia field-collected seedlings, which were sterilized for 30 s with 0.6% NaOCl
and rinsed with deionized water.

Microbial extraction and planting.—-—-Soil microorganisms (excluding arbuscular
mycorrhizal fungi, AMF [Sylvia 1994]) were extracted from cultured soil using a wet-
sieving method adapted from Klironomos (2002), which cull particles < 20 um. For each
extraction, 30 g of soil was blended with 200 ml of tap water for 30 s. The liquid
suspension was washed through 250-, 45-, and 20- um analytical sieves with tap water,
but keeping the extract to 5 450 ml. Sieves were cleaned ultrasonically between each
extraction.

On the same day, seeds with newly emerged radicles were planted in a 1:4
mixture of sterilized field soil and commercial peat moss (Nutripeat, Sun Gro
Horticulture Canada Ltd., Vancouver, British Columbia, Canada). Field soil was
collected from a common pit in a residual, secondary forest at La Selva and was
autoclaved for l h at 121°C followed by a 2-day incubation and a second autoclaving.
Each seedling was randomly assigned to an extract from one of four soil cores (each core
collected near a different conspecific adult and kept separate to test for effect of
individual tree) and received either 100 ml nonsterilized extract (microbial treatment) or
100 ml extract that was autoclaved for 20 min at 121°C (control). We did not add AMF
spores (collected on the 45-urn sieve) because they enhanced seedling mortality under
similar conditions in a previous experiment (S. McCarthy-Neumann and R. K. Kobe,
unpublished data).

Experimental treatments and seedling measurements—To summarize,

experimental treatments consisted of 21 tree species, four or two conspecific adults where

32

soil was collected, and two soil extract treatments (nonsterile vs. sterile). Soil for Dussia,
Quararibea, and Virola were collected from only two conspecific adults due to limited
seed availability, all other species had four conspecific adult locations. The average
number of seedling replicates was 7.5 replicates per extract treatment per adult because
half were replicated eight times and half were replicated seven times. The 1170 seedlings
were randomly assigned to each of the six benches with the criterion that each bench had
five replicates per extract treatment per species. To mimic understory irradiance
(Chazdon and Fetcher 1984), potted seedlings were placed in two shade houses at ~1%
full sun. We confirmed light levels with paired PAR (photosyntheticallyactive radiation)
measurements in the open and at each shade house bench with a LI-COR 250A quantum
sensor (LI-COR, Lincoln, Nebraska, USA) on a uniformly overcast day. Emergence and
survival were censused three times each week, height was measured weekly, and
seedlings were watered (~50 ml of deionized water) by hand twice weekly throughout the
experiment. We assigned date of death as the first census with total leaf and/or stem
tissue necrosis. To determine mass and mass allocation, we harvested seedlings surviving
to the end of the experiment, washed soil fi'om roots, and divided seedlings into root,
stern, and leaf fractions. Necrotic tissue was not included. Roots (except Colubrina,
Quararibea, and Virola) were scanned (Epson Perfection 1260; Epson America, Long
Beach, California, USA) at high resolution (400 dpi), colored black in Photoshop Plus
(Adobe Systems, San Jose, California, USA) for enhanced image contrast and analyzed
using WinRHIZO system version 5.0 (Regent Instruments, Blain, Quebec, Canada) for

length and surface area. Tissue was oven-dried at 70°C to constant mass and weighed.

33

To test for unintended nutrient differences between treatment and control, we
measured nitrogen concentrations (as the percentage of oven-dried leaf mass) with
a CHN Analyzer (Carlo Erba Instruments, Milan, Italy) for 12 of the 21 species (N = 10
seedlings for each species—treatrnent group).

Statistical analysis—Results for all analyses were considered significant at P <
0.10 because the study’s short duration and low light conditions limited the threshold for
treatment effects that could be detected, especially in growth and allocation.

Emergence time and life span, the latter of which includes mortality during both
pre- and post-aboveground emergence stages (preemergence mortality was estimated as
the mean emergence date for seedlings of each species), were compared between
microbial and sterile treatments using the Breslow 12 test of homogeneity in a Kaplan-
Meier survival analysis (SPSS version 14.0; SPSS, Chicago, Illinois, USA). The Breslow
x2 was adjusted for the effect of adult in soil collection and/or bench when these terms
had aP value 5 0.25.

We tested for main treatment (extract and adult) effects and their interactions on
growth, allocation and root morphology with AN COVA using bench as a blocking factor
(SPSS) and seed mass as a covariate in the growth and root morphology analyses and
total plant mass as a covariate in allocation analysis. Root mass also was used as a
covariate in root morphology analysis. We ran full models (extract, adult, bench,
covariate and their interactions) for each dependent variable and species and determined
that the covariate effects were independent of treatment effects (P > 0.05). Thus, main
treatments X covariate interaction terms were removed. If adult, bench, covariate or

extract X adult terms were insignificant beyond the threshold suggested for pooling

34

variances P > 0.25 (Bancroft 1964), then the highest order term with the highest P value
was removed and the analysis was run with the reduced model. This process was repeated
until all terms with P > 0.25 were removed. Adjusted means of treatments were compared
under AN COVA and raw means under ANOVA. We used a 2-tailed independent t test
for each species (SPSS) to test for unintended treatment differences in foliar N
concentrations.

We used fixed effects meta-analysis (MetaWin, version 1.0; Sinauer Associates,
Sunderland, Massachusetts, USA) to test effects of seedling shade tolerance, adult and
seedling abundance, species basal area, and seed size on species sensitivity to the soil
microbial extracts. Each species was placed into one of three categories for each of these
characteristics, and we tested whether effect size (mean difference between treatment and
control means, weighted by each group’s sample size) differed among categories using a
between-class homogeneity statistic (Qb; Gurevitch and Hedges 1993) for life span, total
mass, organ mass, mass allocation, and height and root morphology. In parallel, with
multiple stepwise linear regression we tested for relationships between species
characteristics (Table 2.1) and changes in seedling performance ([(mean total mass) X
(mean life span)] / [number of days of experiment] in the microbial treatment relative to
the control).

A few seedlings (2.5%) were not used for some analyses due to accidental loss of
seedling tissue prior to weighing (15 seedlings), failure to scan roots (14 seedlings), or
other factors (two seedlings were uprooted during watering; one Pentaclethra seedling

had a 40% larger seed size than the second largest seed for that species).

35

ma

Emergence and Survival—Life span in the microbial treatment was significantly
lower for seedlings of Luehea, Coussarea and Prestoea (23%, 13%, and 10%,
respectively) and was higher for Iriartea and Welfia (14% and 5%, respectively)
compared to the sterile treatment (Table 2.2 and Appendix A). Soil microorganisms did
not influence emergence time for any species.

Plant Mass and Allocation—Microbial treatments reduced total mass for Apeiba
(15%), Castilla (8%) and Pentaclethra (18%) seedlings (P< 0.10; Table 2.2 and
Appendix B) and also impacted mass of individual organs in other species, which may
influence future performance. Three species (Guatteria, Trophis, and Iriartea) had lower
root mass, two species (Coussarea and Prestoea) lower stem mass, five species (Apeiba,
Castilla, Prestoea, Quararibea, and Virola) lower leaf mass and one species (Castilla)
lower root length and surface area in response to the microbial treatment. Only two
species increased organ mass (Dussia for stem and Prestoea for cotyledon) in the
microbial treatment (Table 2.2). Distribution of mass among organs was impacted by
microbial extract for eight species but, there was considerable variation among species in
which organ mass was impacted and whether the response was positive or negative
(Table 2.2). Castilla and Euterpe seedlings decreased specific root length and Castilla
reduced specific root surface area in the microbial vs. sterile treatment (Table 2.2).
Growth and mass allocation were affected by which adult cultured the soil for 17 and 16
of the study species, respectively (Appendices B-C). Likewise, initial seed mass was a

significant covariate in most growth responses for all species except Quararibea and

36

Miconia (Appendix B). Mean foliar N concentrations did not differ between treatment
and control for any species (results not shown).

Meta-analysis: Functional Group Comparisons—Contrary to expectation,
common species (whether measured as adult density, seedling density, or species basal
area) were not more responsive to the microbial treatments than relatively rare species
with respect to emergence time (results not shown), life span (Figure 2.1A) or total plant
mass (Figure 2.1B). However, species shade tolerance was associated with seedling
response to the microbial treatment, which lengthened seedling life span for shade-
tolerant species but not intermediate and intolerant species (d = 0.33 vs. -0. 10 and -0. 14;
Q, = 13.33, P < 0.001; Figure 2.1A). The microbial treatment also reduced total mass in
shade-intolerant species and their response differed from that of tolerant species (d = -
0.32 versus 0.15; Q, = 8.31, P = 0.02; Figure 2.1B). Species with large seed mass also
had a longer life span in the microbial treatment compared to medium— or small-seeded
species ((1 = 0.27 versus -0.05 and -0.11; Q, = 7.46, P = 0.02; Figure 2.1A).

Leaf mass (d = -0.21), leaf mass fraction (d = -0.16), and stem height (d = -0.15)
were reduced in the microbial vs. sterile treatment overall but there were no differences
in responses among any of the meta-analysis groupings (Appendices E and F). In
addition, groupings did not differ in organ mass, mass allocation or root morphology
responses to the microbial treatment (Appendices E-G).

Species shade tolerance was the only tested characteristic that was related to
species sensitivity to the microbial treatment (assessed as percentage change in seedling
performance in the treatment relative to the sterile control) (F = 15.79, df = 1,19, P =

0.001, R2 = 0.45; Figure 2.2). Shade tolerance was correlated with seed size for these

37

species (F = 5.15, df = 1,19, P = 0.04, R2 = 0.21), but seed size was not related to
sensitivity to the microbial treatment even when shade tolerance was not included in the

regression.

Discussion

Species susceptibility to soil microorganisms in low light was inversely correlated
with seedling shade tolerance (Figures. 2.1 and 2.2), supporting our third hypothesis.
Shade-intolerant species generally had reduced total mass in the microbial treatment
(Figure 2.1B) whereas shade-tolerant and large seeded species had increased life span
(Figure 2.1A). Thus, NCDD via soil microorganisms exaggerated differences in seedling
shade tolerance, leading to enhanced potential for tree species coexistence through light
gradient partitioning in the presence of soil microorganisms. Similarly, plant-soil
feedbacks could be partially responsible for the segregation of tropical tree seedlings
along light gradients found in recent experimental field studies (Kobe 1999, Montgomery
and Chazdon 2002). However, because our experiment took place solely under low light,
our inferences are limited to how soil microorganisms influence species’ shade tolerance,
but not performance at high light.

Heightened vulnerability of shade-intolerant species to the microbial treatment
could reflect lower investment in defense (e. g., weak leaves, fewer secondary
metabolites, and reduced carbohydrate storage [Coley et al. 1985, Coley and Barone
1996, Myers and Kitajima 2007]). These characteristics, often found in shade-tolerant
species, have been documented to defend against insect herbivores but they may also help

protect seedlings from disease. Soil fungal pathogens are likely the agent causing the

38

negative response in shade-intolerant species to the microbial treatment. These seedlings
often had symptoms characteristic of damping-off (i.e., necrotic roots or stem tissue at
root collar and/or leaf discoloration and wilting) and we have isolated several fungal
pathogens (including species of F usarium and Rhizoctonia) from a subsequent
experiment using a subset of these species and the same extraction methodology.

Our results suggest that the negative effect of the microbial treatment outweighed
any positive microbial influences for the shade-intolerant species. In contrast, in shade—
tolerant species, the positive impacts of the soil microbial community (absence of AM
fimgi) appear to outweigh their negative impacts, for a positive net effect. We
investigated whether the positive response in shade-tolerant and large seeded species to
the microbial treatment could be due to microorganisms involved in nutrient cycling by
measuring foliar nitrogen (N). However, N concentration did not differ between
treatment and control for any species investigated and were generally high (mean = 1.7%-
7.7%), suggesting that N was not limiting. Other nutrients or microorganisms could differ
between the treatment and control such that the elimination of the soil microbial
community led to decreased performance in these species. However, a particular factor
has yet to be identified.

The covariance between shade tolerance and disease resistance documented here
is consistent with the correlation between wood density (as a proxy for shade tolerance)
and disease resistance in 18 tropical tree species (Augspurger and Kelly 1984). The soil
used in their study was from a single common soil pit suggesting either that generalist
pathogens were causing mortality or that host-specific pathogens were ubiquitous. In

contrast, the present study used soil that was cultured by conspecific trees in order to

39

assess the potential for neighborhood-scale negative feedbacks between soil pathogens
and tree seedlings. It is unknown, however, whether the study tree species vary in
susceptibility and/or response to a few soil-bome pathogen species or whether there are
many host-specific pathogen species, each with a unique feedback with a tree species.
The former scenario is probably more likely considering that Augspurger and Wilkinson
(2007) demonstrated that Pythium, a common soil pathogen, varies in pathogenicity
among seedlings of different tropical tree species but does not show strict host specificity.
A similar result was found with 4 temperate tree species and five F usarium morphotypes
(McCarthy-Neumann and Kobe, in review; Chapter 3).

We acknowledge the possibility that our estimates of shade tolerance could
include some contribution fi'om pathogens, but only to the extent that the pathogens are
independent of conspecific density (since we removed density effects from the model
estimates of low-light survivorship). However, it is highly unlikely that our estimates of
shade tolerance arise fiom species differences in susceptibility to density-independent
pathogens, which in tum led to a relationship between apparent shade tolerance and
sensitivity to pathogens. First, many soil pathogens are passively spread and thus their
transmission is density dependent. Second, species differences in shade tolerance arise
from variation in morphological and physiological traits, including leaf-level gas
exchange, whole-plant mass allocation to organs, and allocation to carbohydrate storage
(Reich et al. 1998, Poorter and Rose 2005, Myers and Kitajima 2007). Given this well-
founded body of knowledge, it is very unlikely that density-independent pathogens are
the major cause of species differences in shade tolerance. Nevertheless, interactions

between shade tolerance and species susceptibility to soil pathogens likely influences

40

species distributions across light gradients, especially since soil-home diseases on tree
seedlings appear to be higher in the shaded understory than in canopy gaps (Augspurger
and Kelly 1984, Augspurger 1984). This potential spatial covariance between shade and
soil-bome diseases together with the co-varying species traits of susceptibilities to shade
and pathogens would act in concert to exaggerate differences among species in habitat
preferences.

Species sensitivity to the microbial treatment, as assessed with seedling
performance, was not related to local species abundance. These results are similar to
findings in other tropical forests in seed to seedling transitions (Harms et al. 2000) and
adult tree mortality (Peters 2003). In our study, both rare and common species are equally
attacked by natural enemies in conspecific influenced neighborhoods, which runs counter
to our second hypothesis and suggest that plant-soil microbial feedbacks are not
consistent with the J anzen—Connell hypothesis. However, we have not eliminated the
possibility that rare species could have an advantage over common species because their
seedlings are less likely to encounter soils with host-specific pathogens due to the low
density of conspecifics “culturing” these pathogens. If this community level dynamic is
occurring then plant-soil microbial feedbacks could still facilitate tree species coexistence
through J anzen-Connell processes. Most studies, including our own, investigate the
relationship between species abundance and NCDD using current species abundance
which is a static metric. A preferable approach would be to link trajectories of tree
species abundance in a community to disease pressure through time (i.e., do common

species become less common due to increased pressure and vice versa).

41

The relationship between species abundance and susceptibility to soil
microorganisms could be dependent on the spatial scale at which abundance is
determined. The appropriate scale for determining abundance for our study was the
community since our focus was on species coexistence. Additionally, the spatial scale of
interactions between trees and natural enemies depends strongly on the mobility of the
enemies. Interactions between plants and the soil community appear to be quite local
given that there was significant variation in seedling response to soil cultured by different
trees of the same species. However, the effect of these feedbacks should be manifested at
the community level per the Janzen-Connell model. Although it is difficult to place
spatial boundaries on a community, l-ha plots are likely an adequate, albeit arbitrary,
representation. Even if we had quantified abundance fi'om smaller plots, relative species
rankings in abundance would not have changed (e.g., Pentaclethra and Welfia both
would still be common with Pentaclethra having intermediate susceptibility among
species and Welfia having a positive response to soil microorganisms). Regardless of
scale, it is unlikely that soil pathogens target common species preferentially since
susceptibility was not related to abundance.

There was considerable variation among the 21 species in the way that they
responded to the microbial treatment. Eight of these species responded to soil
microorganisms cultured by conspecific adults and seedlings through changes in their life
span or total mass. An additional five species responded solely through changes in
individual organ mass or root morphology. Although about half of the study species

exhibited some form of negative feedback with the soil microbial community cultured by

42

conspecific individuals, our first hypothesis was only weakly supported since there were
also a few species that exhibited positive feedbacks.

Our results support the accumulating evidence in field and greenhouse studies that
negative feedbacks between plants and soil microorganisms can be an important
mechanism impacting community dynamics in both tropical (Hood et al. 2004, Bell et. al
2006) and temperate (Packer and Clay 2000, 2003 and 2004, Reinhart et al. 2005,
McCarthy-Neumann and Kobe, in review; Chapter 3) forests as well as grassland (Bever
1994, Mills and Bever 1998, Klironomos 2002) and dune (van der Putten et al. 1993)
ecosystems. In addition, our results suggest that considerable spatial and temporal
heterogeneity in plant-soil interactions likely exist in tropical forests since the seedling
response to conspecific cultured soil was influenced by variability among individual
adults for a majority of species. Moreover, our multi-species study determined that life
history characteristics of the plant species (e. g., shade tolerance and to a lesser extent
seed size) are important factors in how the plant responds to the feedback and thus, links
plant-soil feedbacks with light gradient partitioning.

Our results are likely conservative due to limitations of our experimental design.
Seedling growth was highly constrained by low light and the experiment’s short duration,
thus constricting differentiation between treatments. However, the low-light condition
may have enhanced the mortality response of seedlings to the microbial treatment since
low light has been shown to increase the negative effects of pathogen infection
(Augspurger and Kelly 1984, Augspurger 1984). In addition, soil storage at 4°C prior to
microbial extract filtration may have decreased the abundance and/or influenced the

composition of the soil microbial community. We also did not consider seed density as a

43

metric for species abundance even though soil pathogens can cause high mortality at the
preemergence seed state (Forget et al. 2004). Similarly, we focused on soil pathogens as
the natural enemy and the strength of numerous other density-dependent processes (e.g.,
foliar pathogens, insect herbivores, and vertebrate predators) could still covary with
species abundance and create NCDD patterns in forest communities.
Conclusions.—Among species, there were widespread negative feedback
responses for seedlings inoculated with soil microorganisms cultured by conspecifics, but
some species also responded positively. Contrary to expectations, however, local species
abundance did not impact soil microbial susceptibility. The impact of soil
microorganisms, however, was most strongly related to shade tolerance, which critically
influences the dynamics of forest communities (Kobe et al. 1995, Pacala et al. 1996).
Thus, rather than soil microorganisms causing negative feedbacks that constrain the
abundance of common species, our results suggest that soil microorganisms may
exaggerate seedling shade tolerance differences among species which in turn may

influence species coexistence through enhancing light gradient partitioning.

Acknowledgements

We thank Marisol Luna, Ted Salk, Katie Weaver, Ademar Hurtado and Martin
Cascante for their invaluable assistance in the field, greenhouse and laboratory. We also
thank Mike Walters, John Klironomos, Andy J arosz, David Rothstein, Kurt Reinhart and
two anonymous reviewers for comments on earlier versions of the manuscript. The
Organization for Tropical Studies and the staff at La Selva Biological Station provided

logistic support. This research was funded by grants fi'om National Science Foundation

(DEB 0235907) and Andrew W. Mellon Foundation awarded as a research fellowship

through the Organization for Tropical Studies.

45

H88 N. 9. 589 83:83. 07308938 em 98 N9 @2388 :88 $8098 :89 8 888 98 95888388 39 3838 em 896388:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

59889 98395..
588m 9898 >5: 9835.4 mag—5m 9835.5. >8: 889 8884 m89 38%.
88858
5888 8.333% 58883 0338: 988 0033 m8»:
8N5 Emu. 23 88. 23 Chm. 23 C. S
288 888888.... 58883 0358: 00838: 5838988 m8»:
88.8 89.3. 23 fidu. 23 8.8. 2.3 88. 3
3883 3835.383 58883 5838988 5838988 5838988 m3»:
883 8.8. 0.3 8~.w. and 8.3. 953 bx. 3
0282.3 .3893 58883 5838988 0030: 8 wow. 58 78983
893 88. «3 ~33 8.3. Pod 8m. 3
Q3828 9833.338 58883 5838988 5838988 988 7.8958
89% 8.9. P8 89b. 3.3 8.8.. 8.83 8a. .0
5883.8 3:33838 58883 5838988 5838988 ~88 78988
88.3 8.9 98 89m. 3.8 8.2. 925 88. 98
08.38 889.33 5883: ~88 988 5838988 28988
80.8 88. 2.3 8.o. 23 8.8. 23 b 8. nd
0859.3 5:858 5838988 00:38: 0.358: waw. 5838988 28983
$93 89$ :3 :33 8.5. 983 Cum. wuv
3838883 58383 5838988 00:38: 00:38: 00338: Ema
A83 898. 0.8 898. 933 CPS. N3 883. 33
3.883 825.3... 5838988 00:38: 5838988 33.? 988 7.8988
39¢ C So. 83 38.3 88m. 903 P3. NNV
@3383 838.8083” 5830988 988 988 958 3%
298 8a. N3 8b. #3 8.3. ome 83. ed
$615. 383930 5838988 W3 W88 ”88 790983
CNS 8a.. 0.3 8.m. 93 8.8. obs 8m. 98
V6388 83.8.33 5830988 958 00330: 5838988 m3»:
993 88. 93 838. $88 8.8. 9ro 8w. 3

 

46

H888 8.8 A033

 

 

 

 

 

 

 

 

0.03003. 3.3.0:. H0883 00330: 00330: 5838988 8.880
8.3 8:8. 88.83 83.3. 888.8 8.8. 8.33 889. 833
8.2830 3.008828 H0883 00330: 5838988 5838988 H.380
A83 . 888.8. Nd 88.8. 88.3 8.98. 8.33 838. 83
3.3.80 8088880 H 0883 003300 5838988 G 8.8. 0030: 9.880
8.3 88.3. 88.3 3.3 $88. 8.33 883. 883
3080 ~38 H0883 00330: 00330: 00330: 8.080
8.3 88.8. 8.83 A9888. USN 8.3. 8.83 A380. 33
3.38 800838 H0883 5838988 5838988 00330: 3m0
. 8.3 8.8. 8.3 88.8. 88.3 8.88. 8.83 2388. N85
65.08 380333888 H0883 82.0 5830988 5838888 5.80
8.3 8.3. wa 88.8. 3.3 8.88. 8.83 8888. 885
380:8 0 588 88.0 82.0 830 m3»:
8. 33.: 80.8 9.0. 0.3 Eu. 00.00 8.8. 0.80 e. 8
8.3808033: H0883 C 8.3 830 988 00308 28883

 

5830308 3

  

N8. :8

 

8. N9

  

 

2080.. 38080 8.0 00089 8% 000858 0890 88300 08005038: 38 88: ~83 0.839300. $8000 8 8883:0000 30 303 H
mU. 9890 888800 80 30 80080330 338:3~ 08. 0009380 3 18 3:03 39 33 N03 0030:0050 000955 90305.” 0890
8883. A8808“ 0880 8838988. 8.8-8808“ 38 0880 38883. v8.8. >98: 80:05. 80 88 0383 08. 39788380 €13 88: W
8 033 :08. 88088“ 88. A u 398883038“ 8830888. .0790 8988839088“ 0030:. v8 5883888038. 7803 083:3 000858
90:05. 80 88 3383 0». 0008580 80309 80 .A. 8 v8 088V :00 80088 8009. <03” 88. A88 00085808353 8838888. No-88
0008—83333.“ 00330:. vNoo 00093808354. >98: 8803 2.00. 80 88— 30» A3093” 3.0. A 8.88 338“ 3830988. 8.88 I 8.88
3088“ 0030:. v 8.88 338. M008 3800 80 83 0008 3800 A38” 038:. A98 3% 30883. 8.8 I 888 3% 880. v 8.88 3m. 2>
888080 8000 8:80 030 8.03 88 0:5~ 382009 08.88 33 003388 3000 088080. 00 :0 MD 03 80 00.88888. 4 U88 3000308
08 3030 38 08:83 8988030 W03 38388 08:90. 8 8.080330 328:3~ 3 1x. 3: 03“ $8803 30 0009—93 80:05~
3.83080 80 8 030: 0388 080. a U38 0030 0008 8 38:80:83 808338 0890 888500 80. 08080 $8000 338:3. 3090—0
£08 :3 8:08.88 88 8 038: 0330 0800” 8:088 9. W. 8080. 3808:0808 833. 5.033. 8080 833. 3.8.0038 80300: 08 8..
Noowv. 833883080383: 8.33888 N83.

47

Table 2.2. List of significant seedling responses to microbial vs. sterilized microbial
extract for all studxspecies (condensed from Appendices A-lfl.

 

 

 

Species Variables with Significant Response
Luehea seemannii life span (-23% *), root surface area fraction (-47% T)
Neea psychotroides root length fraction (+21% :)

 

Apeiba membranacea

total mass (-15% T), leaf mass (-23% *), root mass fraction
(+21% T)

 

Colubrina spinosa

no significant response

 

Guatteria diosplroides

root mass (-12% T)

 

Psychotria panamensis

no significant response

 

Castilla elastica

total mass (~8% 1'), leaf mass (-12% *), root length (-15%
1'), root surface area (-1 1% **), root mass fiaction (+9% T),
stem mass fi'action (+7% *), leaf mass fraction (-4% **),

specific root length (-12% **), specific root surface area (-
8% *)

 

Coussarea handensis

Pentaclethra macroloba

life span (-13% T), stem mass (-12% 1‘), leaf mass fi'action
(+13% T)
total mass (-18% T)

 

Prestoea decurrens

Quararibea bracteolosa

life span (-10% 1'), stern mass (-18% **), leaf mass (-21%
*), cotyledon mass (+31% *), stein mass fraction (~20% *),
leaf mass fraction (-12% 1‘), cotyledon mass fraction (+11%
*)

leaf mass (-29% I)

 

T rophis racemosa

root mass (-13% *)

 

 

 

Vochysiaferruflea no significant response
Capparis pittieri no significant response
Euterpe precatoria specific root length (-9% 1')

 

Iriartea deltoidea

Welfia regia
Virola koschnyi

Dussia macrophyllata
Miconia aflinis
Stryphnodendron
microstachyum

life span (+14% *), root mass (~23% *) (-l6% *), root mass
fraction (-18% *), stem mass fraction (-12% 1'), cotyledon
mass fraction (+1% *), root length fi'action (-l6% T), root
surface area fi'action (-15% *)

life span (+5% 1')

leaf mass {-15% 1'), stem mass fraction (+10% *), leaf mass
fraction (-11% 1')

stern mass (+45% *)

no sigificant response

no significant response

Note: Values given are the negative (-) or positive (+) percent difference in seedling
response between the microbial and control treatment. Significance is shown as: T P 5
0.10; * P 5 0.05; ** P g 0.01.

48

A) Life span

 

 

0'75 Shade Adult Seedling Basal Seed
()5 .. tolerance density density area size !
*1" NS NS NS * l
0.25 e l
l ++w+++~+lw ------

Effect Size

.6

N

on o
HIH

t—.—E-t

r—O—E—d

 

 

 

 

 

 

B) Total Mass

 

 

 

 

 

 

 

 

0'75 Shade Adult Seedling Basal Seed
()5 4 tolerance density density area size
79 ;
i “l l l l - i ll
f ----------- -------------- W}
-0.51
‘0.75 l I l l l l 1 fi’ 1 I r— __.
h l\ ’\ ‘0 F) l\ I\
,.gg...gg.,g§ggg
\ C C in §
- E K . k L . .
é" a ~ s a? 1‘5 a 1‘5 a s E 2 35
a a ‘0 g?
d g a
§ :5 3's" 1:“

Figure 2.1. Effect size of soil pathogens on (A) life span and (B) total mass (magnitude
and sign of the effect of the microbial vs. control treatment) compared to categories of
local community characteristics for 21 tropical tree species. Data points show means
and bars show 95% CI ranges for all studies (Overall) as well as each study category.
Sample sizes and categories are indicated on the x-axis; the dotted line shows Hedges’ d,
indicating the absence of an effect. Significance is shown as: NS, not significant; * P _<_
0.05; *** P _<_ 0.001.

49

J;
O

y = -0.965x + 14.749

  

 

3 30-11 . ’ R2=0.43;p=0.001

V l

g 204; . . .

g

a: 1---.--“ - --- -- ....................

a O 1 O 9'. 0 .

,5 -10 J.

a) l O 9

21” -20 4‘

‘° 5 o

'5 -3O +| o

'40 1 fl T l —‘

0 10 20 3O 40 50

Shade intolerance

Figure 2.2. Relationship between seedling shade intolerance and change in performance
between seedlings in the microbial vs. control treatments for 21 tropical tree species.
Shade intolerance is measured as percent mortality at 1% fiill sun and zero conspecific
seedling density. Change in performance is measured as ([mean total mass]X[mean life
span] )/(number of days in experiment). The significance value is from multiple, stepwise

regression.

50

CHAPTER THREE

CONSPECIFIC AND HETEROSPECIFIC TREE-SOIL FEEDBACKS INFLUENCE

SURVIVORSHIP AND GROWTH OF TEMPERATE TREE SEEDLINGS

Abstract

The J anzen-Connell (J -C) hypothesis proposes that host-specific enemies could maintain
high tree species diversity by reducing seedling survivorship near conspecific adults
and/or at high conspecific seedling densities. Negative feedback between plant and soil
communities could be an important mechanism of such non-competitive distance and
density-dependent (N CDD) mortality. In a greenhouse experiment, we assessed: 1) life
span and grth responses of Acer rubrum, Acer saccharum, F raxinus americana and
Quercus rubra seedlings to extracts taken from soils that had been cultured by adults of
each of these species; 2) whether these relationships were influenced by additional
culturing of soil by different species and density of seedlings; 3) soil microbes as the
mechanism creating these plant-soil feedbacks; and 4) whether low light availability
increased species vulnerability to pathogens. Species-specific feedbacks between adult
trees (but not seedlings) and soil influenced life span and/or growth for all species.
Conspecific and heterospecific feedbacks had similar prevalence and magnitude of
influence. In addition, heterospecific seedlings were not necessarily favored by these

feedbacks. Chemical factors in the soil mediated these plant-soil feedbacks, whereas

51

microbial factors (primarily F usarium) simply reduced seedling performance regardless
of which tree species cultured the soil. Five Fusarium morphotypes were the primary
infectious agents responsible for killing seedlings in the non-sterile extract treatment.
Disease reduced seedling performance for some species regardless of light availability
and for others only in high light. Species-specific adult-soil feedbacks impacted seedling
performance, and thus have the potential to influence forest community dynamics.
However, the idiosyncratic nature of these interactions likely diminishes their ability to

enhance species diversity via J -C processes.

Key words: Community structure; density-dependence; distance-dependence;
F usarium; irradiance; Janzen—Connell; plant-soil feedback; soil microbes; species

coexistence.

Introduction

Identifying the mechanisms that maintain tree species richness is a central
problem in plant community ecology because competitively dominant species are
expected to exclude inferior species (Gause 1934). Janzen (1970) and Connell (1971)
hypothesized that competitive exclusion could be precluded by host-specific enemies that
reduce seed and/or seedling survivorship near conspecific adults and/or at high
conspecific densities. Such non-competitive distance or density-dependent (N CDD)
mortality would favor establishment of heterospecific individuals, thus promoting species

coexistence. Although, NCDD was proposed as a mechanism operating in tropical

52

forests, there is some evidence of NCDD occurring in temperate forests as well (Packer
and Clay 2000, Hille Ris Lambers et al. 2002, Packer and Clay 2003).

Natural enemies such as soil pathogens are likely effective agents of NCDD
mortality and/or reduced grth because many of them Ashow host specialization, short
generation times, high fecundity, long persistence in soil, and more limited dispersal than
their hosts (Gilbert 2002). These characteristics could enhance the potential for
“culturing” the local soil microbial community by the resident plant species, leading to a
potential negative feedback for the plant species that cultured the microbes (van der
Putten et al. 1993, Mills and Bever 1998, Klironomos 2002, Bezemer et a12006, Casper
and Castelli 2007, Kardol et al 2007).

Distance and density-dependent mortality is at the core of the J anzen-Connell (J —
C) Model, but determining the relative importance of these processes is difficult because
seed density often is inversely related to distance from adult. One way to investigate
these differences is to compare how seedlings respond to soil micro-organisms cultured
by adults vs. seedlings since each life history stage may culture unique enemies or may
impact abundance or virulence of the same enemies in a different way (Gilbert 2002).

Irradiance may also mediate disease-induced NCDD processes. Seedlings of
many tropical species (Augspurger and Kelly 1984, Kitajima and Augspurger 1989,
Hood et al. 2004) experience higher disease related mortality at low than high light. Four
hypotheses have been proposed for mitigated influence of disease in high light: 1)
compensation for tissue lost to disease by accumulating biomass more rapidly 2) reduced
exposure to disease through faster lignification, 3) unfavorable conditions (e. g. higher

temperature or decreased moisture) that lower pathogen abundance and 4) increased

53

AMF colonization that suppresses disease (Augspurger 1990, Borowicz 2001, Gehring
2003)

Although the primary focus in this paper is on soil pathogens in NCDD processes,
there are other important plant-soil feedbacks (Ehrenfeld et al. 2005). The presence of a
particular plant species could be associated with formation of mycorrhizal networks
(Booth 2004), production of allelochemicals (Stinson et al 2006), alterations to soil
physical properties (Rillig et al. 2002) and nutrient availability (Finzi et al. 1998a—b). All
of these feedbacks could impact seedling performance in a species-specific manner, i.e., a
particular species could modify soil to the detriment or benefit of conspecific or
heterospecific seedlings (Bezemer et al 2006). The potential for complex relationships
among plant species mediated through soil feedbacks challenges lumping all non-
conspecific species into a single heterospecific category (e. g. Augspurger and Kelly
1984, Packer and Clay 2000 and 2003, Hood et al 2004).

The purpose of this study was to examine mortality and grOwth responses of
seedlings of four temperate tree species to: species of adult culturing soil in the field
(“source” effects), species and density of seedlings further culturing field soil in a
greenhouse, presence of microbial pathogens, and light level. Specifically, we tested the
following hypotheses: H1 & H2) Soil cultured by conspecific adults (Hl) or seedlings
(HZ) reduces seedling survival and/or growth more than soil cultured by heterospecific
adults (H1) or seedlings (H2). H3) High seedling density during soil culturing increases
the magnitude of seedling responses. H4) Sterilization of soil extracts enhances seedling
survival and growth due to the elimination of soil pathogens. H5) Higher irradiance

reduces negative effects of pathogen infection.

54

Materials and Methods

Soil culturing by conspecrfic versus heterospecific adults (HI )— To test the effect
on seedling performance of soils cultured by conspecific vs. heterospecific mature trees
(Hl), we collected soil beneath each of the study species (Acer rubrum, Acer saccharum,
F raxinus americana and Quercus rubra) in November 2004. For each species used in the
seedling response experiment, two soil cores (7.5 cm diameter x 25-cm depth) were taken
within 1 m from the bole of three trees of each of the 4 species for a total of 96 soil cores
(4 species of seedling x 4 species of adult culturing x 3 trees x 2 samples/ tree). Sampled
trees were randomly selected from adults with a diameter at breast height at 3 75th
percentile for that species located in 4 mapped stands on moraines in the Manistee
National Forest, MI. To minimize the potential for multi-species cultru'ing of soil, we
took soil under trees that were at least 2 crown diameters away from adults of the other
species. Sampling locations for culturing by heterospecific adults had the additional
criterion that no conspecific adults were closer than 20 m, except that we used a 10 m
distance'for A. saccharum due to its high local abundance. Soil was stored for ~ 2 wks at
4°C until seeds were available for planting in intact soil cores for the seedling culturing
step.

Soil culturing by conspecific vs. heterospecific seedlings (H2) and seedling
density (H3)— To test for effects of seedling soil culturing in addition to previous
culturing by adults in the field, we planted conspecific vs. heterospecific seeds at high
density in intact soil cores collected from each heterospecific adult (H2). To test effects

of seedling density, paired soil cores taken fi'om each conspecific and heterospecific adult

55

location were planted with one versus four conspecific seeds (H3). A. rubrum and F.
americana seeds were from Sheffield’s Seed Co (Locke, NY, USA) and A. saccharum
and Q. rubra seeds from the Wisconsin Department of Natural Resources (Hayward, WI,
USA). Seeds were surface sterilized (0.6% NaOCl solution), rinsed with DI water and
weighed prior to planting. Seedlings cultured soil for 14 weeks in a greenhouse at 2% full
sun; additional seedlings were planted as needed to maintain desired density for the entire
culturing period. Before soil extraction, roots were cut and mixed with the soil and
aboveground seedling portions were discarded. All culturing treatments were kept
separate through the soil extraction process. Extracting soils that had been cultured at two
seedling densities avoided potential confounding of seedling competition with NCDD
effects.

Eflect of sterilization on pathogen infection (H4)— Soil micro-organisms < 20 mm
were extracted from cultured soil using a wet-sieving method adapted from Klironomos
(2002). For each extraction, 50 g of soil was blended with 250 m1 of tap water for 30
seconds. The liquid suspension was washed through 250, 45 and 20-p.m analytical sieves
with tap water, keeping the extract to 5 l L. Sieves were cleaned ultrasonically for 5 min
between each extraction, which at least minimized if not eliminated contamination
between treatments. To test for microbial effects on seedling performance (H4), planted
seedling pots were amended with autoclaved (30 min at 121 °C) versus unsterilized soil
extract.

Planting methods and eflect of irradiance on pathogen infection (H5)— Seeds with
newly emerged radicles were planted in a 1:4 mixture of sterilized field soil and

commercial peat moss (F afard Mix #2, Conrad Fafard Inc., Agawarn, MA USA). Field

56

soil was collected from a common pit in a mixed Fagus-Acer stand at Michigan State
University’s Tree Research Center and was autoclaved for l h at 121°C followed by 2 d
incubation and a second autoclaving. Lethal temperatures (2 121°C) were confirmed at
the center of each soil bag. Each seedling received 100 ml of non-sterilized or sterilized
extract. We did not add arbuscular mycorrhizal spores (collected on the 45 um sieve)
because they enhanced seedling mortality under similar conditions in a previous
experiment (McCarthy-Neumann and Kobe, unpublished data). To test for irradiance
effects (H5), seedlings were grown at two light levels (2% vs. 22% fiill sun).

Experimental Treatments, Seedling Measurements and Harvesting— To
summarize, experimental treatments consisted of species of adult, species of seedling and
density of seedlings culturing soil, sterilization, and irradiance level. Density and species
of seedling were tested only in low light. The 1,668 seedlings were randomly assigned to
8 benches (6 for low and 2 for high light) and were allocated among treatments per
Appendix H.

Emergence and survival were censused thrice weekly and seedlings were watered
(~50 m1 of DI) by hand twice weekly for 12 weeks, from March to June 2005. We
assigned date of death as the first census with total leaf and/or stem tissue necrosis, at
which time dead seedlings were harvested for pathogen isolation. To determine live mass,
we harvested seedlings surviving to the end of the experiment, washed soil fiom roots,
divided seedlings into organ fractions, and oven-dried living tissue at 70°C to constant
mass. Stem length was also measured for aboveground height.

Chemical analysis of soil extracts— To test for potential chemical differences, we

measured exchangeable base cations (Ca, K, and Mg), total organic C, total N, C:N ratios

57

and protein-precipitable phenolics in extracts for each soil source (species of adult
culturing). Extract samples were stored at 4°C for two years prior to analysis since testing
for chemical effects was not part of our original research plan. Soil extracts were filtered
with Whatrnan # 2 papers and exchangeable base cations (Ca, K, and Mg) were measured
using a Perkin-Elmer Optima 2100 DV Optical Emission Spectrometer (Perkin Elmer,
Norwalk, Connecticut, USA). For the total organic C, total N and C :N analysis, soil
extracts were filtered through pre-rinsed Whatrnan GF/F papers prior to being analyzed
with a TOC-V CPN Total Organic Carbon Analyzer equipped with a total nitrogen
measuring unit (Shimadzu Co., Kyoto, Japan). We measured protein-precipitable
phenolics with a spectrophotometer, following Makkar et al. (1988).

Pathogen Isolation and Host-Specificity of F usarium Morphotypes - Upon
seedling death, roots were rinsed of soil in DI water, surface sterilized for 1 min with
0.6% NaOCl and rinsed with sterilized water. Cross-sections from the leading edge of the
disease lesion were plated on water agar amended with Ampicillin. Isolates were
subcultured and maintained on water agar and potato dextrose agar, both amended with
Ampicillin. Isolates were identified to genus using morphological characteristics; over
75% of isolates were in the genus F usarium and were classified into five morphotypes

To determine host-specificity, seeds with emerged radicles were inoculated with
each of the five morphotypes and a control of pure water agar (A. saccharum seeds were
unavailable for inclusion in this experiment). Inoculum was obtained from pure cultures
of the Fusarium morphotypes stored on silica beads. Individual silica beads were placed
on multiple water agar plates amended with Ampicillin for conidia propagation and

germination. The treatment consisted of five,10 mm disks of water agar containing

58

F usarium isolates that were placed directly into the soil (a 1:5 mixture of sterilized field
soil and commercial peat moss), and the control consisted of an equal number of pure
water agar disks. Inoculated seedlings were assessed for original symptoms. This
experiment consisted of 360 seedlings (3 species x (5 Fusarium morphotypes + control) x
20 replicates) all grown in low (~2% full sun) light for 8 wks from April-June 2006. Seed
cleaning, soil sterilization and seedling measurements were identical to methods used in
the main experiment.

Statistical Analysis— We analyzed seedling species and density separately from
other treatments since they were carried out at low light only. In the absence of
significant differences, data were pooled across seedling species and density treatments.

Life span was analyzed with survival analysis (SPSS v. 14.0, SPSS Inc, Chicago,
IL) and includes both pre- and post- aboveground emergence stages (pre-emergence
mortality date was estimated as the mean emergence date for seedlings of that species).
The Breslow )6 test of homogeneity in a Kaplan-Meier analysis tested effects of seedling
species and density on mean life span. Cox proportional hazards regression (Cox and
Oakes 1984) tested relative effects of soil source, sterilization, light availability, and
initial seed mass on mortality. »

For growth responses, we tested for main treatment effects and their interactions
with AN COVA (treatments = species and density of seedling culturing the soil) and with
a split-plot AN COVA, split for light (treatments = soil source, sterilization and irradiance
level), using bench as a blocking factor (SPSS). Because seed mass can influence
seedling size, estimated dried embryo mass (based on regressions of dry embryo mass to

fresh seed mass developed from ~30 randomly selected seeds for each species) was a

59

covariate. Data were natural-log transformed when errors were not normal. We ran full
models (main treatments, bench, covariate and interactions) for each dependent variable
and species to test the assumption that covariate effects were independent of treatment
effects; interaction terms were removed when P > 0.05. If either terms for bench,
covariate or the interaction between main treatments had P > 0.25 (Bancroft 1964), then
the highest order term with the highest P value was removed and the analysis was run
with a reduced model. This process was repeated until all terms with P > 0.25 were
removed. Adjusted means were compared when the covariate was retained and raw
means when the model reduced to ANOVA. When the main effect of soil source was
significant (analyses were considered significant at P < 0.10 because of experiments short
duration) a Holrn adjustment was used to compare the conspecific to each of the 3
heterospecific adult soil sources for each species. Differences between treatment means
were assessed and P values calculated through degree of overlap in 95% confidence
intervals (Austin and Hux 2002).

To summarize reciprocal effects, we compared percent difference in integrated
seedling performance [(mean total mass x mean life span) / (days of experiment)] for
each species pair in soil extract (combined sterile and non-sterile treatments) relative to
tap water in high and low light. To determine the relative effects of microbial versus
chemical factors, we also calculated percent difference in integrated seedling
performance for each species pair as: adult culturing through chemical effects alone [=
seedling performance in sterile extract / seedling performance in tap water] and adult
culturing through soil micro-organisms alone [= (seedling performance in non-sterile

extract — seedling performance in sterile extract) / seedling performance in tap water]. We

60

estimated 95% confidence intervals for integrated seedling performance metrics by

bootstrapping 3000 data sets for each study species (sampling with replacement), and

analyzing each data set as described above using R (R Development Core Team 2008).
We used similar statistical methods to analyze Seedling responses to the five

Fusarium morphotypes.

m

Conspecific and heterospecific culturing have similar influence on seedling
performance (HI )— Species of adult culturing the soil (i.e., soil source) affected
survivorship of F. americana seedlings, with reduced life span in soil extract cultured by
Q. rubra vs. F. americana adults (Figure 3.1a, Table 3.1). Soil source did not affect life
span for any other study species (Appendix I).

Seedling total mass for all species except F. americana were influenced by soil
source (Figure 3.2, Table 3.1, see Appendix J -M for growth responses for all species), but
conspecific soil did not disproportionately affect mass relative to heterospecific soil.
Total seedling mass for A. rubrum was lower with soil extract cultured by conspecific vs.
A. saccharum and Q. rubra adults. In contrast, seedlings of Q. rubra had greater mass in
conspecific vs. A. rubrum and F. americana soil sources. For A. saccharum seedlings, the
influence of soil source was mediated by irradiance. In low light, seedling mass was
lower in conspecific vs. F. americana soil source, whereas in high light, mean seedling
mass was greater in conspecific vs. F. americana and Q. rubra soil sources. Sterilization
of the soil extract had no effect on whether soil source influenced life span and/or total

mass for any species.

61

Soil culturing by conspecific and heterospecific seedlings (HZ) and seedling
density (H3) had negligible eflects— In general, seedling culturing did not alter effects of
soil source on seeding responses (results not shown). In the few significant responses
effects were minor, with increased stem height for A. rubrum (F = 2.76, df= 1,118, P 5
0.10, mean = 57.9 vs. 54.5 cm) and life span for Q. rubra seedlings (x = 3.11, df= 1,134,
P < 0.10, mean 82.5 vs. 77.4 days) in conspecific vs. heterospecific culturing by
seedlings. The density of seedlings culturing the soil did not affect life span or grth for
any of the species (results not shown).

Sterilization (H4) broadly influences seedling performance; irradiance (H5)
interacts with sterilization— Sterilization of soil extracts increased life span and/or growth
for all species, except A. saccharum (Table 3.1; Appendices I & J -M). F. americana
seedlings had significantly reduced life span in the non-sterile extract across all soil
sources (Figure 3.1b; Table 3.1). The non-sterile extract also reduced total seedling mass
for Q. rubra regardless of soil source or irradiance level; A. rubrum and F. americana
seedling mass was reduced across all soil sources but only in high light (Figure 3.2 and
Table 3.1). Under higher irradiance, total mass was greater for all species (Figure 3.2;
Table 3.1; Appendices J -M), but life span did not vary (Table 3.1 and Appendix I).
Sterilization of soil extracts and efforts to minimize cross contamination were effective;
70% of dying seedlings in non-sterile treatments were infected versus 6% of dying
seedlings in sterile treatments. 75% of infected seedlings harbored at least one of five
F usarium morphotypes.

Individual organ mass and stem height responses to the microbial extract were

generally consistent with total mass responses for all of the species (Appendices J -M).

62

For all species, mass and height increased with initial seed mass (except for A. saccharum
height). Random assignment of seedlings to certain low-light benches affected growth
responses for all species; this effect was eliminated except for A. saccharum when a
bench closest to the greenhouse fan was excluded fi’om‘the analysis. The exclusion of this
bench did not change the significance or percent differences for any growth response so
all data were used in the analysis.

Integrated seedling performance— Plant-soil feedbacks can influence performance
of a given species in conspecific vs. heterospecific cultured soils, (within rows of Table
3.2; significant comparisons highlighted in Table 3.1), as well as which species of
seedling is most affected by soil cultured by a given species (within columns of Table
3.2). Taking the conventional approach of comparing performance in con- vs. hetero-
specific soils, heterospecific soils were as likely to be detrimental as conspecific soils.
For instance, Q. rubra culturing least affected Q. rubra seedlings compared to extracts
cultured by the three heterospecific species (significant at low light), and under high light
A. saccharum culturing affected conspecific seedlings less than F. americana soil source.
However, under low light, A. saccharum culturing most affected conspecific seedlings
compared to extracts cultured by F. americana. In addition, conspecific culturing tended
to be the most detrimental for A. rubrum seedlings compared to culturing by other species
(significant difference between A. rubrum and Q. rubra culturing at low light).

From an alternative perspective of interspecific comparisons of performance in
soils cultured by a given study species, conspecific seedlings were not disadvantaged
relative to heterospecific seedlings (Table 3.2). For instance, conspecific seedlings were

less affected than at least some heterospecific seedlings for both A. saccharum (F.

63

americana, P 5 0.05) and Q. rubra cultured soils (A. saccharum, P 5 0.05 and F.
americana, P _<_ 0.001). In addition, A. rubrum seedlings tended to experience less severe
feedbacks than seedlings of other species in heterospecific cultured soils (except for F.
americana cultured soil at low light). Only F. americana cultured soil was more
detrimental to performance of conspecific than heterospecific seedlings under high light
(A. rubrum, P _<_ 0.05 and Q. rubra, P 5 0.05).

Both microbial and chemical factors in the soil extracts influenced seedling
performance (Figure 3.3). Under high light, the response of Q. rubra seedlings to all soil
extracts arose from both chemical and microbial factors; under low light, the response to
con- and hetero-specific culturing was dominated by microbial and chemical factors,
respectively, with stronger chemical than microbial effects. For A. saccharum seedlings,
chemical factors from most soil sources were detrimental, except for F. americana source
soils under low light; but conspecific vs. F. americana cultured soil had less (14%, P =
0.02) and greater (18%, P = 0.06) negative chemical factors in high and low light
respectively. Under high light, A. rubrum responded negatively to biotic factors. Under
low light, the negative effect of conspecific soil source on A. rubrum seedlings arose
from both microbial and chemical factors; however, both microbial and chemical factors
in Q. rubra cultured extract were neutral for A. rubrum seedlings. Soil extracts decreased
F. americana is performance (which was similar regardless of soil source) through
combined detrimental microbial and chemical factors (significant in high light only).

Chemical analysis of soil extracts— Soil extracts cultured by Q. rubra adults had
lower base cation availability (combined Ca, K, and Mg) than extracts cultured by A.

saccharum and F. americana adults, but had greater total C and C:N ratios than sources

64

from the other 3 species (Appendix N). A. rubrum source soils also had lower base
cations and greater C:N ratios than A. saccharum and F. americana source soils
(Appendix N). Total N (Appendix N) and protein-precipitable phenolics (F = 0.42, df=
3,19, P = 0.74) did not differ among soil sources.

Host-Specificity of F usarium Morphotypes - F usarium morphotypes reduced
seedling life span for A. rubrum (x = 17.99, df= 5,120, P 5 0.01) but not for the other
species (Appendix 0). Fusarium morphotypes also influenced total mass for A. rubrum
(F = 2.7, df= 5,91, P 5 0.05), F. americana (F = 2.5, df= 5,49, P 5 0.05) and Q. rubra
seedlings (F = 2.2, df= 5,107, P 5 0.10) (Appendices P-R). Reductions in seedling mass
were due to proportionate decreases in stem and leaf mass for all species and root mass
for A. rubrum seedlings (Appendices P-R). Total seedling mass was positively related to
initial seed mass for F. americana and Q. rubra seedlings; for all species, the random
bench assignment affected total seedling mass due to a Thysanoptera outbreak that
occurred on the second bench (Appendices P-R). The exclusion of this bench did not
change the significance or percent differences for any grth response so all data were
used in the analysis.

We also evaluated effects of Fusarium morphotype in terms of integrated seedling
performance relative to the sterile treatment (Figure 3.4). Fusarium morphotypes 3-5
reduced A. rubrum seedling performance relative to the control and morphotype 2.
Fusarium morphotype 3 reduced seedling performance relative to the control primarily
through decreased total mass (22%; P S 0.05, Holm adjusted P = NS), F usarium 4
through decreased life span (19%; P 5 0.01, Holm adjusted P _<_ 0.10) and Fusarium 5

through decreased total mass (32%; P 5 0.01, Holm adjusted P 5 0.05). Despite

65

significant F-statistics in the overall grth models for each species, seedling total mass
did not differ in the control versus the various F usarium morphotypes for F. americana
and Q. rubra after adjusting for multiple comparisons. The following results should be
viewed as suggestive trends. For F. Americana, seedling performance was reduced (due
to a decrease in total mass) by morphotypes l (19%; P = NS) and 3 (17%; P = NS)
relative to the control and morphotype 2, 4 and 5. For Q. rubra, F usarium morphotype 1
had the lowest seedling performance (due to lower total mass) relative to morphotype 5
and to a lesser extent the control (15%; P _<_ 0.10, Holm adjusted P = NS) and morphotype
2. In addition, Q. rubra seedlings had lower seedling performance (due to lower total

mass) when grown with morphotype 3 and 4 than morphotype 5.

Discussion

Soil-mediated species —- specific feedbacks between tree adults and seedlings were
ubiquitous in this study. However, feedbacks were largely idiosyncratic and were not
consistent with the J anzen-Connell Model. Most importantly, conspecific and
heterospecific feedbacks occurred at similar strength across the study species.
Furthermore, conspecific seedlings were not disadvantaged by these feedbacks, thus
weakening the potential for these feedbacks to constrain populations and thereby
contribute to species coexistence. Lastly, species-specific feedbacks tended to be
chemical rather than biotic; microbial factors reduced seedling performance regardless of
which tree species cultured the soil.

Conspecific and heterospecific culturing have similar influence on seedling

performance (HI)- A. rubrum was the only species for which conspecific soil source

66

more negatively affected seedling performance than heterospecific soil sources in both
high and low light; A. saccharum had a similar response but only in low light. For F.
americana and Q. rubra at both light levels and A. saccharum at high light, at least one or
more heterospecific soil sources were more detrimental than conspecific sources. Thus,
for 3 of 4 species under high light and 2 of 4 species under low light, conspecific soil
sources had the lowest adverse effect among all soil sources, however not all of these
ranks were statistically significant. Similarly, the relative effect of con— versus hetero-
specific cultured soils in grasslands depends upon particular species pair-wise
interactions, with heterospecific cultured soils sometimes having more negative impacts
than conspecific cultured soils (Bezemer et a1 2006). Furthermore, none of the soil
sources here were uniformly beneficial or detrimental to seedlings of all species.
Likewise, the relative influences of biotic and chemical factors from a given soil source
could change depending upon the species of responding seedlings. As a whole, these
results support fairly specific interactions between pairs of species that are mediated
through both chemical and biotic factors.

Although there are limited data to test whether similar species pair-wise
interactions are operating in the field, the one result that we could test with field data was
supported. We used seedling demography data spanning 6 years fi'om northwest lower
Michigan (Kobe, unpublished data) to test if F. americana seedlings had a shorter life
span with Q. rubra vs. conspecific source soils. Average one year survival of newly
germinated F. americana seedlings was 15% lower when both Q. rubra and F.
americana adults were within 10 m of focal seedlings versus when only conspecific

adults were present. The 15% reduction in survivorship was consistent with the present

67

study’s results (Figure 1, 10% reduction at 77 days). Lower F. americana survivorship in
the presence of Q. rubra is unlikely to arise from species differences in canopy light
transmission since both tree species transmit similar irradiance (Canharn et al. 1994).
Support for the J anzen-Connell Model often comes from decreased seedling
performance for one or more focal species at near versus far distances fi'om conspecific
adults (Augspurger 1984, Augspurger and Kelly 1984, Hood et al. 2004, Bell et al. 2006).
The implicit assumption of these studies is that a given species, whose seedlings are
disadvantaged in the presence of con- versus hetero-specific adults, would more likely
have its adults replaced by hetero- than con-specific seedlings. However, this assumption
may not be valid if heterospecific seedlings are affected more than conspecific seedlings
in a given soil source, even if the given species’ seedlings perform relatively poorly in
con- versus hetero- specific soil. In the present study, for example, even though A.
rubrum seedlings were most negatively affected in conspecific source soils, both F.
americana and A. saccharum seedling performance was more strongly negatively
affected (under high light); Q. rubra seedlings were non-significantly less affected than A
rubrum seedlings. Under low light, seedling performance of all four species was similarly
affected (—17 to -20%) in A. rubrum source soils. Based on these feedbacks and not taking
into account other aspects of species life history, the expectation is that Q. rubra and A.
rubrum seedlings would be least affected by A. rubrum source soils and thus the most
likely species to replace an A. rubrum canopy tree when it dies and creates a gap (with
higher irradiance). Thus, knowledge of the relative benefits or costs of ‘escape’ from
conspecific adults on seedling recruitment does not translate into knowledge of which

species’ seedlings are favored for recruitment near a particular tree species. But in

68

combination these two approaches provide a better understanding of how plant-soil
feedbacks (or other mechanisms producing NCDD) contribute to tree species
coexistence.

In general, A. rubrum seedlings were less affected by soil culturing than other
species’ seedlings. By absolute rank, A. rubrum was least affected in 5 of 8 soil source by
light combinations (Table 2). Together with other factors (deer browse, fire exclusion),
these feedbacks may be contributing to the documented increase in A. rubrum abundance
throughout the eastern United States (F ie and Steiner 2007).

Contrary to expectation, the effect of adult cultru'ing on seedling life span and/or
growth occurred in both non-sterile and sterile soil extracts. Ifeffects of adult culturing
had been due only to biotic agents, then we would have expected negligible differences
among soil sources for sterilized extracts. Thus, adult culturing reflects both chemical and
biotic processes. Chemical analyses of the extracts suggest that base cation availability
may be partially responsible for the species-specific feedbacks between tree adults and
seedlings. Extracts derived from soil cultured by Q. rubra had significantly lower base
cation availability than extract cultured by F. americana and A. saccharum, but greater
total organic C and C:N ratios than extract cultured by any of the other study species. In
addition, extracts derived fiom soil cultured by A. rubrum had significantly lower base
cation availability and greater C:N ratios than extract cultured by F. americana and A.
saccharum. Exchangeable cation availability differs under canopies of the same species
(F inzi et al. 1998b) and manipulation of exchangeable cations can affect seedling and
sapling performance (Kobe et al. 2002, Bigelow and Canham 2007). In this study,

however, base cation availability was not significantly related to total mass for any

69

species, but only ~50% of the seedlings could be used for this analysis since only a sub-
sarnple of the extracts were analyzed for cations. Variation in C:N ratios among soil
sources (which was largely a function of differences in total organic C) is a less likely
mechanism for observed plant—soil feedbacks since N immobilization is unlikely under
the observed ratios of < 25: 1. Chemical analysis of soil extracts were post-hoe and were
conducted after two years of cold storage. Thus we consider these results to be
suggestive. Experiments linking soil chemical factors with species-specific feedbacks
between tree adults and seedlings are still needed to fully ascertain which chemical
factors are responsible for creating these feedbacks.

Our soil source results would not have been interpretable or significant for F.
americana lifespan if we had lumped all soils cultured by heterospecific adults into one
treatment (e. g. “ ar” treatments in Augspurger 1984, Packer and Clay 2000 and 2003,
Hood et al 2004). For other species, the general relationship between soil source and
seedling life span or grth is consistent whether or not heterospecific species are
aggregated. However, which heterospecific species influences performance relative to a
conspecific adult would be unknown.

Soil culturing by conspecrfic vs. heterospecific seedlings (H2) and seedling
density (H3) had negligible eflects— Soil culturing by seedlings did not alter the effect
that soil culturing by adult trees had on seedling responses, except for Q. rubra. Contrary
to hypothesis 2, Q. rubra seedlings had increased life span under conspecific versus
heterospecific seedling culturing. Contrary to hypothesis 3, density of seedlings during
the culturing step did not affect subsequent seedling performance, presumably because

higher seedling density did not change the abundance or composition of soil microbes.

70

Seedlings cultured soil for 14 weeks, a similar duration as other studies where seedling
culturing did result in negative feedbacks (Packer and Clay 2004, Stinson et al. 2006).
Similar to our results, Milicia regia seedling survival did not differ in soil from female
versus male conspecific adults, the latter of which should have lower seedling densities
(Hood et al. 2004). However, seedling density appears to mediate the impact of disease
on seedling performance in studies that simultaneously vary experimental seedling
density and pathogen presence (Packer and Clay 2000, Bell et al. 2006). Different
methodologies could lead to conflicting results among studies, especially if density
enhances disease transmission through close proximity of roots, which was eliminated in
our extraction-based study.

Sterilization broadly influences seedling performance (H4) — Three of four
species responded negatively to non-sterile extract, the majority of which were infected
by Fusarium. A. saccharum seedlings were not affected by the non-sterile treatment,
consistent with a lack of pathogen effect on the germination and viability of A.
saccharum seeds (O’Hanlon-Manners and Kotanen 2006). A. saccharum ’s shade
tolerance, the highest among the study species, also could convey tolerance to soil
pathogens, as shade and pathogen tolerances are positively correlated across tropical
species (Augspurger and Kelly 1984, McCarthy-Neumann and Kobe, in press; Chapter
2).

Five F usarium morphotypes were the primary seedling mortality agents in the
non-sterile extracts, and these morphotypes are likely widespread since we were able to
culture the majority of them at least once from seedlings grown with soil extracts cultured

by each tree species (results not shown). However, when directly inoculated with these

71

five Fusarium morphotypes they differentially affected A. rubrum, F. americana and Q.
rubra seedling survival and/or growth. Similarly, Augspurger and Wilkinson (2007)
demonstrated that Pythium, another common soil pathogen, varies in pathogenicity
among species of tr0pical tree seedlings but was not strictly host-specific.

Irradiance interacts with sterilization (H5)— Contrary to expectation, increased
irradiance did not ameliorate the negative effect of soil microbes on seedling
performance. Indeed F. americana life span and Q. rubra growth were reduced by soil
microbes without regard to irradiance level and growth of A. rubrum and F. americana
seedlings were reduced by soil microbes only at high light availability. Seedlings have
been hypothesized to be less affected by disease in high light environments through four
mechanisms: tissue compensation, faster lignification, unfavorable conditions for soil
micro-organisms, and through increased AMF colonization (Augspurger 1990, Borowicz
2001, Gehring 2003). The first two mechanisms likely are not operating here since they
would have led to lower disease effects at high light with our methodology. However,
since we minimized abiotic differences between light treatments and excluded AMF
spores from microbial extracts, this study may not have adequately tested the third or
fourth potential mechanisms, which may explain contradictory results with studies that
allow abiotic factors (e. g. soil moisture and temperature) and/or AMF colonization to
vary between irradiance levels (Augspurger 1984, Augspurger and Kelly 1984, Hood et
al. 2004). Additionally, grth for some species (e.g., A. rubrum and F. americana here)
may have been so severely constrained in shade that effects of soil microbes were not

manifested. Similarly, interspecific competition for other resources also can influence the

72

magnitude of seedling responses in plant-soil feedbacks (Casper and Castelli 2007 ,
Kardol et al 2007).

Although tree-soil chemical and microbial feedbacks occur, spatial heterogeneity
in light availability has greater potential to affect seedling performance. The two
strongest soil source effects were a 43% reduction in A. saccharum performance in F.
americana soil under high light and a 42% reduction in F. americana performance in Q.
rubra soil. In contrast, mean seedling mass increased 90-3000% across species from 2 to
22% full sun, which encompasses the typical range of variation of understory to large tree
fall gap light levels in Michigan deciduous forests (Schreeg et al. 2001). However, the
potential influence of spatial heterogeneity in light levels rarely is realized. Along 200 m
belt transects at 4 moraine sites where adult cultured soils were originally sampled,
irradiance in 1m2 quadrats averaged 1.5% canopy openness, ranging from 0.5 to 6%
(Kobe, unpublished data). In contrast, spatial heterogeneity in soil culturing likely is
ubiquitous. Thus, effects of irradiance and soil cultruing on seedling performance in the
field may be more similar than suggested by potential irradiance effects

Caveats— Excluding the filtrate component > 20 pm in soil extracts may have
underestimated the role of soil microbes in plant-soil feedbacks, by eliminating some
pathogens that could be present in field soil. Nevertheless, sterilization of extracts likely
benefited seedling grth because infection rates were much higher in non-sterile than
sterile treatment seedlings. Sterilization of soil extracts also may have resulted in a
nutrient pulse, which would have heightened differences between sterilization treatments.
A nutrient pulse arising from sterilization is consistent with increased grth for two

species in sterilized soil extract under high light, where nutrients could constrain growth

73

(Kobe 2006). Counter to this interpretation, however, the nutrient pulse from sterilization
of field soil (20% of the potting medium for all seedlings) would have overwhelmed any
nutrient contribution from extract sterilization. Extract culturing by mature trees of
different species may be confounded with an underlying template of soil chemistry that
determines tree species occurrence. However, soils were collected from a narrow range of
fertility conditions (moraine sites). The study species also occur across a similar range of
mineral-bound nutrients (presumably uninfluenced by species occupancy) with divergent
exchangeable nutrient pools (F inzi et al 1998a), which could be modified by species
occupancy.

Conclusions- Our results add to accumulating evidence that feedbacks between
plants and soil micro-organisms could influence plant community dynamics, as supported
in temperate (Packer and Clay 2000 and 2003, Reinhart et al. 2005) and tropical (Hood et
al. 2004, Bell et. a1 2006, McCarthy-Neumann and Kobe, in press; Chapter 2) forest as
well as grassland (Mills and Bever 1998, Klironomos 2002, Bezemer 2006, Casper and
Castelli 2007, Kardol et a1 2007) and dune (van der Putten et al. 1993) ecosystems.
However, feedbacks in this study occurred primarily between mature trees (not seedlings)
and soil and encompassed both biotic (Fusarium) and abiotic (possibly through
differences in base cation availability) factors that have the potential to differentially
impact seedling performance. Additionally, these plant-soil interactions were specific to
species pairs with conspecific and heterospecific feedbacks occurring at similar strength
across species. Contrary to expectation, effects of the microbial treatment on seedling
performance manifested more strongly in high rather than low light. Plant-soil

interactions affected seedling performance by as much as 43%, a magnitude of effect that

74

would be expected to influence community dynamics. The complex nature of these
interactions, however, likely diminishes their ability to enhance tree species coexistence
via J anzen-Connell processes, especially since the establishment of heterospecific

seedlings was not always favored.

Acknowledgements

We thank Mindy McDermott, Sam Tourtellot, Renee Pereault, Amanda Gevens,
Jennifer Hunnell, Melissa McDermott, David Neumann and many others for their
assistance in the field, greenhouse and laboratory. We also thank Mike Walters, John
Klironomos, Andy J arosz and David Rothstein for comments on earlier versions of this
manuscript. This research was funded by a grant from the National Science Foundation

(DEB 0235907).

75

Hugo mg. Cm" em mmmimoma mama—Em 5.0 meme. 8:: 89%. Ba gamma—Ha moan—Em Ramona—“Sod 3.60:me mg m: weave £393 8”
we: mos—.8 3.8 2898 nae—815m 85. me: 9:83 Aegean—o <m. mama—av. Ba Seaweed 3.x. <m. BX. we: 33 ASE—958 ”.83

369538 ES E5 .365 PB.

 

 

moaned om muoomam Ema mega Hoe: Emma“ gagged moon—5m mammoamsoow
$530:

3 s... >... 2%... w u Poe

. . .. e H
>H<m.OHAl—m.xe.~.vflo.od >n<m.0~5P—_WEAHM\8~V OLOV

he. 20 amines“: managed

>m <m. mm 5 H. mmE Ammo? % .l. ome >m <m. m.» E H. mm! A-3e\e. w n ome
he. 20 «5:585 Emmougoo >m <m. m.» E I am? A+u~£9 w .l. 988 >~ <m. m.» E I am! A+NAe\e. w u
>m <m. On 5 m mmE A+~ue§ me n obwv 0.83

 

we: mefiod
he mmmxwdnw. can. me 555 n 20 animus“: @5338 20 awn—Bomb" Engage
e M Q <m. >n E r mmrziofie. w n 0.95
be. 20 mmmimoma Engage W H”. WM mum“: Wu ”WW 0.. <m. >m 5 h mmwxi fix: w u 0.3V
. e . 02a. memew=m=2+:.x..wue.§
we: magma" Se. 20 mmmamoma Engage Z-m <m m 5 I mmE Ammo? me n ob: Z>

me. 20 mmmimoma E3538 Ze amines“: “managed Z>

76

Ame—o up. AOSV
Z-m <m. m fibm. Ho fined u

 

fin Pod Z-m <m m 5 I mmE Abueé. Ho u 9?: . 75
me. 20 finances” Ema—.38 Z-m <m m $5; Ho u ob: Z>
me. 20 mmmamoma @5998 H. <m. m Abflx. He n 983 Z...»
. . . - e u
H :58 x? 20 mmemoB: Emma—.38 H. <m. H: mu \9 He oboe Z>
Hun Zo amamoma Engage H. <m. E Abfix: He u 0.83 2.?
be. 20 finance: qugoa H. <m. HH $33. He u oboe Z>

 

 

229. misused 338 V a Eamon—8 39820: 5 93$ 8 Bone—3. e». 35v. madame? “55—8 26 9a somemé 3 2. @833 Hi v9.82
Engage E moon—Em «emcee—ma caged: 0256850 mum refinemudommo add me: 8:83. between—0 85 £an 98.99. BE 52 gm
Em: mmE. Unmonvmoa om again—moan H. n _92 swan 25 m n Em: swan Z-m u Bondage 85 m n man—ow .3 n 38‘. 33:3. >m
"meme. “censuses. mm H waning aimless? Ba 9 n @365. 396“ Z> n :9 36:82.“ £58 mam—«mom 2% so" 8:988? He

$58 men BEES—d oeauamoam :98 and: 2: 88m 2:: m mew: 2:588?

Table 3.2. Reciprocal effects (percent difference in integrated seedling performance

[(mean total mass x mean life span) / (days of experiment)]) of plant-soil feedbacks

relative to a tap water control for each study species integrated across extract treatment

and at a) high and b) low irradiance levels.

 

 

 

a) High Light Species of Adult Culturing the Soil
Species of seedling Acer Acer Fraxinus Quercus
responding to soil rubrum saccharum americana rubra
Acer rubrum -22% -5% -13% -6%
(-33, -12%) (-20, 10%) (-23, -3%) {-20, 7%)
-30% -19% -43% -37%

Acer saccharum
F raxinus americana

Quercus rubra

(-43, -13%)
-3796

(-46, -27%)
-l6%

(-26, -8%)

(.29, -11%)
-38%

(.49, -27%)
-15%

(-28, -5%)

{-54, -33%)
-3396

(-43, -22%)
-15%

(-26, -5%)

(.53, -20%)
42%

(-54, -30%)
-1093
(-21, -1%)

 

 

 

 

 

b) Low Light Species of Adult Culturing the Soil
Species of seedling Acer Acer Fraxinus Quercus
responding to soil rubrum saccharum americana rubra
Acer rubrum -18% -4% -13% -l%
(-28, -7%) (-17, 10%) (~22, -3%) (-13, 13%)
Acer saccharum '1 7% 40% 3% 42%
(-28, -7%) Q28, -11%) (-14, 5%) (-21, -4%)
F raxinus americana 49% 46% -8% 41%
(-30, -6%) (-27, -4%) (-21, 10%) (-34, -5%)
-20% -20% -21% -10%

Quercus rubra

S-27, 44%! 1-28, 44%) 1-28, -14%! 1-18, -3%)

Note: Effect of conspecific cultured soil on seedlings is in bold. Bootstrap devised 95%

C1 are in parentheses.

78

a) Soil source

 

 

 

 

 

 

 

 

 

100“ Soil
Source
33 95‘ J'IAr
1’ .J‘lAs
g 90— -' fFa
g - T ’Qr
U) 85— g
80— ’
I I I I I
0 15 30 45 60
Life span (day)
b) Soil microbial treatment
1 00*
-, Extract
g; 95“ 1 J'18terile
:5 ‘..l_ ‘ '- 1N0"-
.5 90‘ ' 'Sterile
3 - ' ' '
(D 35—
80-
I I I I I
O 15 30 45 60
Life span (day)

Figure 3.1. Survival curves for F. americana seedlings by (a) soil source (tree species
culturing soil: Qr = Q. rubra, Ar = A. rubrum, As = A. saccharum and Fa = F.
americana) and (b) soil microbial treatment (sterile vs. non-sterile extract). Survival

curves end at 58 days since no F. americana seedlings died after that date.

79

Figure 3.2. Total final mass by soil source (tree species culturing soil) and extract (sterile
vs. non-sterile) for each study species in low and high light (A. rubrum (Ar) a—b, A.
saccharum (As) c-d, F. americana (Fa) e-f and Q. rubra (Qr) g-h). Dotted line represents
seedling mass when grown with tap water and is shown for reference only, and were not

included in statistical analysis.

80

Total Mass (mg) + 95% Cl Total Mass (mg) + 95% Cl

Total Mass (mg) + 95% Cl

20 7 a) A. rubrum (2% full sun)

16
12

8
4
0

 

 

J,

. 1 I T --

7 I Sterile

I CI Non-Sterile
Ar As Fa Qr

Species of Adults Culturing Soil

 

 

 

625 _‘ b) A. rubrum (22% full sun)
Sterile
I3 Non-Sterile
Ar As Fa Qr
Species of Adults Culturing Soil
125 fi c) A. saccharum (2% full sun)
100 I " I. ' I
75 ‘ I Sten'le
50 .‘ I3 Non-Sterile
25 I
O «.

Ar As Fa Qr
Species of Adults Culturing Soil

81

Figure 3.2. (Ctd)

1250 l d) A. saccharum (22% full sun)

1000 I
750 ‘1 lSterile
500 -l IZINon-Sterile
1m
0 7H.
Ar As Fa Qr

Species of Adults Culturing Soil

 

Total Mass (mg) + 95% Cl

 

 

 

g 100 T e) F. americana (2% full sun)
1.3") 80 i t 1' I I
+
’6: 60 lSterile
E, 40 EINon-Sterile
«n
a 20
2
E 0 , _
8 Ar As Fa Qr
Species of Adults Culturing Soil
2 2000 _‘ f) F. amen’cana (22% full sun)
33 1600 A
+
31200 I' ISterile
E 300 I DNon-Sterile
tn
8 400 I
2
a 0 4*
.3 Ar As Fa Qr

Species of Adults Culturing Soil

82

Figure 3.2. (Ctd)

 

  

 

 

 

6 1500 19) Q. rubra (2% full sun)
°\O ‘ t ........
g 1200 I
+
E": 900 " lSterile
:0, 600 _. EINon-Stenle
8
E 300 a
E
19 o H ,

Ar As Fa Qr

Species of Adults Culturing Soil

°\°
3 4000 - . - -
4.
033000 A lSterile
3 2000 g E] Non-Sterile
E 1000 A
E
.2 0 ~

Ar As Fa Qr

Species of Adults Culturing Soil

83

Figure 3.3. Relationship between chemical [sterile extract / tap water] and microbial [(un-
sterile extract / tap water) — (sterile / tap water)] effects in soil extracts “cultured” by
different species of adult on seedling performance [(mean total mass x mean life span) /
(days of experiment)]) for each study species in high and low light. Ar = A. rubrum, As =
A. saccharum, Fa = F. americana, and Qr = Q. rubra. Bootstrap devised 95% CI

included.

84

Microbial Effect (% Difference in Seedling Performance)

a) A. rubmm in High Light b) A. rubrum in Low Light 9 Ar

 

 

 

 

 

 

30 j I AS
15 —~' I i] A Fa
o _— L, ———-+ l . , e e 1 —~~ ---———. XQr

-15 l 1 . *

~30 j 92‘ . l

45 4‘

c) A. saccharum in High d) A. saccharum in Low Light

30
15

0 MT—

-15 i

 

fl- , f. _ #F:+=J ____

 

 

 

  

 

 

 

 

 

 

-30 7
45 J i
e) F. americana in High f) F. americana in Low Light
30
15 l -
o e I. T 1 ~ . , .
-15 % 1' ‘
-30 4 —~
45 l J
g) Q. rubra in High Light h) Q. rubra in Low Light
30 1 i
15 l
0 fia—“I‘” T i i_—" ""‘" L i
'15 l l i I +
.30 i i
45 —' 4

-45 -30 -15 0 15 3o -45 -30 -15 0 15 30
Chemical Effect (% Difference in Seedling Performance)

85

60—

c3", 4O «
g .
E A 20 ‘ ClFus. 1
:5 8, ‘ I .FUS. 2
‘t g 0 “ ' 7 ' 71 .FUS. 3
a) J: / I IFus. 4
CL 0 /
g) _20 ‘ 4 [:1 FUS. 5
E
g _40 ,. Ar Fa Or

-60 ~

Study Species

Figure 3.4. Change in performance [(mean total mass x mean life span) / Days in
Experiment] between seedlings in the sterile control versus each of the Fusarium
morphotypes for each study species. Ar = A. rubrum, Fa = F. americana and Qr = Q.

rubra.

86

CHAPTER FOUR

CONSPECIFIC TREE-SOIL FEEDBACKS REDUCE SURVIVORSHIP AND

GROWTH OF TROPICAL TREE SEEDLINGS

Abstract

The J anzen-Connell (J -C) Model proposes that host-specific enemies could maintain high
tree species diversity by reducing seedling survivorship near conspecific adults and/or at
high conspecific seedling densities. Negative feedback between plant and soil
communities could be an important mechanism of such non-competitive distance and
density-dependent (NCDD) mortality. In a shade-house experiment, we assessed: 1) life
span and growth responses for six species of tropical tree seedlings (Apeiba
membranacea, Colubrina spinosa, Pentaclethra macroloba, Prestoea decurrens, Iriartea
deltoidea and Virola koschnyi) to extracts taken from soils that had been cultured by each
of these species; 2) soil microbes as the mechanism creating these plant-soil feedbacks;
and 3) whether low light availability increased species vulnerability to pathogens.
Species-specific feedbacks between trees and soil influenced life span and/or grth for
all species. Supporting the J -C Model, three of the six species had decreased seedling
performance when grown with extract cultured by con- vs. all hetero-specific individuals.

I. deltoidea had mixed results, performing better in two and worse in one hetero-versus

87

con-specific cultured soil. C. spinosa performed worse in con- vs. one of the hetero-
specific soils. Chemical rather than biotic soil factors primarily mediated these plant-soil
feedbacks. Shading did not increase vulnerability to soil micro—organisms in seedlings of
any species. These results, along with parallel prior reSearch in temperate forests, suggest
that plant-soil feedbacks are an important component of seedling dynamics in both
temperate and tropical forests, but would be more likely to enhance species coexistence
in tropical forests because negative conspecific feedbacks were more pronounced for

tropical than temperate tree species.

Key words: Density-dependence; distance-dependence; irradiance; host-
specificity; Janzen-Cantrell; plant-soil feedback; soil microbes; species coexistence;

tropical forests.

Introduction

Identifying the mechanisms that maintain species richness is a central question in
plant community ecology because competitively dominant species are expected to
exclude inferior species (Gause 1934). J anzen (1970) and Connell (1971) hypothesized
that competitive exclusion could be precluded by host-specific enemies that reduce seed
and/or seedling survivorship near conspecific adults and/or at high conspecific densities.
Such non-competitive distance or density-dependent (NCDD) mortality would favor
establishment of heterospecific individuals, thus promoting species coexistence.

Negative feedback, whereby individual plants “culture” the local soil microbial

community in which they grow to the detriment of themselves and other conspecific

88

individuals (van der Putten et al. 1993, Mills and Bever 1998, Klironomos 2002,
Bezemer et al 2006, Casper and Castelli 2007, Kardol et a1 2007), may be an important
mechanism that could create NCDD mortality and/or reduced growth. Natural enemies
such as soil pathogens are likely effective agents in creating this type of feedback since
many of these pathogens show host specialization, short generation times, high fecundity,
long persistence in soil, and more limited dispersal than their hosts (Gilbert 2002). In a
prior study, we found that a majority of 21 tropical tree species experienced reductions in
seedling life span and/or growth when inoculated with non-sterile vs. sterile soil extract
cultured by conspecific adults and seedlings (McCarthy-Neumann and Kobe in press;
Chapter 2). However, whether these soil microbe-mediated feedbacks between individual
plants were more detrimental in con- than hetero-specific interactions was not
investigated and thus host-specificity of these feedbacks is unknown.

Traditionally in studies of NCDD, all heterospecific individuals have been
lumped into a single heterospecific category (6. g. “ ar” distance) (Augspurger and Kelly
1984, Packer and Clay 2000 and 2003; Hood et al 2004). However, tree species vary in
many characteristics (e.g., resource allocation to defense vs. growth) and soil-mediated
effects of mature individuals on seedlings could be species specific as well (McCarthy-
Neumann and Kobe, in review; Chapter 3).

Irradiance may also mediate disease-induced NCDD processes. Seedlings of
many tropical species experience higher disease related mortality at low than high light
(Augspurger 1983, Augspurger 1984, Augspurger and Kelly 1984, Hood et al. 2004).
Four mechanisms (tissue compensation, faster lignification, unfavorable conditions for

soil micro-organisms, and through increased AMF colonization ) have been proposed for

89

how seedlings could be less affected by disease in high light (Augspurger 1990,
Borowicz 2001, Gehring 2003). However, for temperate species, disease reduced
seedling performance for some species regardless of light availability and for others only
in high light (McCarthy-Neumann and Kobe, in review; Chapter 3).

Although the primary focus in this paper is on soil pathogens in NCDD processes,
there are other important plant-soil feedbacks (Ehrenfeld et al. 2005). For example, in a
temperate greenhouse experiment (McCarthy-Neumann and Kobe, in review; Chapter 3)
species of adult culturing soil influenced seedling performance, but the effect was
manifested in both non-sterile and sterile soil extracts, suggesting a species-induced
change in soil chemistry (possibly base cation availability). Additionally, plant-soil
feedbacks could be created when a particular plant species is associated with the
formation of mycorrhizal networks (Booth 2004), production of allelochemicals (Stinson
et a1 2006), alterations to soil physical properties (Rillig et al. 2002) and nutrient
availability (F inzi et al. 1998a—b).

It is often assumed, but rarely tested, that mechanisms underlying community
dynamics in tropical versus temperate forests are different. In particular, the J anzen-
Connell Model assumes that NCDD processes are less pronounced in temperate versus
tropical forests because temperate forests have a less diverse tree community along with
lower rainfall and greater seasonality, which could result in an overall lower abundance
of enemies and disproportionately fewer specialist enemies. Over the past 30+ years, the
focus of NCDD studies has been primarily on tropical tree species and communities.
Among the few studies that have been conducted in temperate forests, there is

accumulating evidence of NCDD effects on seedling mortality and growth (Packer and

90

Clay 2000, Hille Ris Lambers et al. 2002, Packer and Clay 2003, McCarthy-Neumann
and Kobe, in review; Chapter 3). However, we are not aware of any studies that use
parallel experiments in both temperate and tropical forests, which would enable more
direct comparisons.

The purpose of this study was to examine mortality and growth responses of
seedlings of six tropical tree species to: species culturing the soil (adult species from
which soil was collected with further culturing by a high density of seedlings of the same
species in the greenhouse), presence of microbial pathogens, and light level. Specifically,
we tested the following hypotheses: H1) Soil cultured by conspecific individuals reduces
seedling survival and/or grth more than soil cultured by heterospecific individuals.
H2) Sterilization of soil extracts enhances survival and growth due to the elimination of
soil microbes. H3) Higher irradiance reduces effects of pathogen infection. Additionally,
by comparing results here to a parallel study (McCarthy-Neumann and Kobe, in review;
Chapter 3) we assess whether tropical species exhibit greater sensitivity to NCDD

processes than temperate species.

Materials and Methods

Species - To test these hypotheses, we assessed seedling survivorship and growth
responses for six tropical tree species (Apeiba membranacea, Colubrina spinosa, Iriartea
deltoidea, Pentaclethra macroloba, Prestoea decurrens and Virola koschnyi) to extracts
taken from soils that had been cultured by each of the six species. These species were
selected because they vary in abundance class and seedling shade tolerance; we

previously investigated their seedling responses to microbes extracted from soil cultured

91

by conspecific but not heterospecific adults and seedlings (McCarthy-Neumann and
Kobe, in press; Chapter 2).

Field Site — This research was conducted at La Selva Biological Station
(Sarapiqui Region, Costa Rica) which is operated by the Organization for Tropical
Studies. La Selva is a 1,510 ha reserve of diverse (> 400 tree species), wet tropical forest
receiving ~ 4000 mm of rain annually with a mean annual temperature of 25.8° C
(Hartshom and Hammell 1994).

Soil culturing by conspecific versus heterospecific individuals (H1 )-To test the
effect on seedling performance of soils cultured by conspecific vs. heterospecific trees we
collected soil cores beneath each of the study species and further cultured the soil by
planting a high seedling density of the same species in intact soils cores in shadehouses.
We removed a lO-cm diameter by 30-cm deep soil core within 1 m fi'om the bole of four
adults for each species in December 2005. Sampled trees were randomly selected from
adults with a diameter at breast height at _>_ 75th percentile for that species located in 5
mapped stands. To minimize potential for multi-species culturing of soil, we randomly
selected adults for which no individuals >5, or >9, or > 20 cm DBH of the other 5 species
were located within 5, 10, or 20 m, respectively, fiom the focal tree. Seeds were collected
within 5 m from the trail system throughout La Selva. Seeds were surface-sterilized
(0.6% NaOCl solution), rinsed with DI water and weighed prior to planting. To further
culture the soils, each core was planted with 4 germinating conspecific seeds, a density
that was maintained for 13 weeks in a shadehouse at 1% full sun. Before microbial

extractions, roots were cut and mixed with the soil and aboveground seedling portions

92

were discarded. All culturing treatments were kept separate through the extraction
process. A

Effect of sterilization on pathogen infection (H2)—Soil micro-organisms < 20 um
were extracted from cultured soil using a wet-sieving method adapted from Klironomos
(2002). For each extraction, 40 g of soil was blended with 250 m1 of water for 30
seconds. The liquid suspension was washed through 250, 45 and 20-um analytical sieves
with tap water, keeping the extract to 5 800 ml. Sieves were cleaned ultrasonically for 5
min between each extraction, which at least minimized if not eliminated contamination
between treatments. To test for microbial effects on seedling performance (HZ), planted
seedling pots were inoculated with autoclaved (20 min at 121 °C) versus unsterilized soil
extract.

Planting methods and effect of irradiance on pathogen infection(H3)— Within 5
days of obtaining extract (which was stored at 4 °C), seeds with newly emerged radicles
were planted in a 1:4 mixture of sterilized field soil and commercial peat moss
(Nutripeat, Sun Grow Horticulture Canada Ltd, Vancouver, BC, Canada). Field soil was
collected from a common pit in a residual, secondary forest at La Selva and was
autoclaved for 2 h at 121°C followed by 2 d incubation and a second autoclaving. Lethal
temperatures (_>_ 121°C) were confirmed at the center of each soil bag. Each seedling
received 100 ml of non-sterilized or sterilized extract. We did not add arbuscular
mycorrhizal spores (collected on the 45 um sieve) because they enhanced seedling
mortality under similar conditions in a previous experiment (McCarthy-Neumann and
Kobe, unpublished data). To test for irradiance effects (H3) seedlings were grown at 1%

and 5% filll sun. Light levels were confirmed with paired PAR measurements in the Open

93

and at each bench in the shade houses. PAR was measured on a uniformly overcast day
with a LI-COR 250A quantum sensor (LI-COR, Lincoln, NE USA).

Experimental Treatments, Seedling Measurements and Harvesting — To
summarize, experimental treatments consisted of species of tree culturing the soil,
sterilization and, irradiance level. Seedlings were also grown with just tap water addition
as an additional control. The 3,084 seedlings were randomly assigned to 10 benches and
were allocated among treatments as detailed in Appendix S.

Twice a week for 11 weeks (March to July 2006), emergence and survival were
censused, and seedlings were watered (~50 ml of DI) by hand. We assigned date of death
as the first census with total leaf and/or stem tissue necrosis, at which time dead seedlings
were harvested. To isolate pathogens, roots were rinsed of soil in DI water, surface
sterilized for l min with 0.6% NaOCl and rinsed with sterilized water. Cross-sections
from the leading edge of the disease lesion were plated on water agar amended with
Ampicillin. Isolates were identified to genus. To determine live mass, we harvested
seedlings surviving to the end of the experiment, washed soil from roots, and oven-dried
living tissue at 70°C to constant mass.

Statistical Analysis — Life span was analyzed with survival analysis (SPSS v.
15.0, SPSS Inc, Chicago, IL) and includes both pre- and post- aboveground emergence
stages. Cox proportional hazards regression (Cox and Oakes 1984) tested relative effects
of soil source (species culturing soil), sterilization, light availability, and initial seed mass
on mortality.

For growth responses, we tested for main treatment effects and their interactions

with split-plot AN COVA, split for light (treatments = soil source, sterilization and

94

irradiance level), using bench as a blocking factor (SPSS). Because seed mass can
influence seedling size, estimated dried embryo mass (based on regressions of dry
embryo mass to fresh seed mass developed fiom 20-40 randomly selected seeds for each
species) was a covariate. Data were natural-log transformed when errors were not normal.
We ran firll models (main treatments, bench, covariate and interactions) for each
dependent variable and species to test the assumption that covariate effects were
independent of treatment effects; interaction terms were removed when P > 0.05. If either
terms for bench, covariate or the interaction between main treatments had P > 0.25
(Bancroft 1964), then the highest order term with the highest P value was removed and
the analysis was run with a reduced model. This process was repeated until all terms with
P > 0.25 were removed. Adjusted means were compared when the covariate was retained
and raw means when the model was reduced to ANOVA. When the main effect of soil
source was significant (analyses were considered significant at P < 0.10 because of the
experiment’s short duration) a Hohn adjustment was used to compare the conspecific to
each of the 5 heterospecific soil sources for each species. Differences between treatment
means were assessed and P values calculated through degree of overlap in 95%
confidence intervals (Austin and Hux 2002).

To summarize reciprocal effects, we compared percent difference in integrated
seedling performance [(mean total mass x mean life span) / (days of experiment)] for
each species pair in soil extract (combined sterile and non-sterile treatments) relative to
tap water averaged across irradiance levels. To determine the relative effects of microbial
versus chemical factors in non-sterile soil extracts, we also calculated percent difference

in integrated seedling performance for each species pair: tree species culturing through

95

chemical effect alone [ = seedling performance in sterile extract / seedling performance in
tap water] and tree species culturing though soil micro-organisms alone [ = seedling
performance in non-sterile extract — seedling performance in sterile extract / seedling
performance in tap water]. We estimated 95% confidence intervals for integrated seedling
performance metrics by bootstrapping 3000 data sets for each study species (sampling
with replacement), and analyzing each data set as described above using R (R

Development Core Team 2008).

m

Conspecific culturing had greater influence on seedling performance than
heterospecific culturing (HI) — Tree species culturing soil (i.e., soil source) affected
survivorship in 4 (Figure 4.1a-d and Table 4.1; Appendix T) and total mass in 5 of 6
species (Figure 4.2, Table 4.1; Appendix U). Conspecific cultured soil disproportionately
decreased seedling survival and/or growth relative to heterospecific soil.

Seedlings of both C. spinosa and P. decurrens had shorter life span with
conspecific versus heterospecific soil extract. Results for I. deltoidea were mixed with
life span either longer or shorter in conspecific versus heterospecific cultured soil extract
depending upon the particular species of heterospecific. In contrast, life span for A.
membranacea seedlings was longer with soil extract cultured by conspecific vs.
heterospecific individuals. Life span for P. macroloba and V. koshnyi seedlings did not
vary based on species culturing the soil (Table 4.1).

Total mass for C. spinosa seedlings was greater with soil extract cultured by

conspecific vs. I. deltoidea individuals but only in high light. In contrast, seedlings of I.

96

deltoidea, P. macroloba, and V. koschnyi all had lower mass with soil extract cultured by
conspecific vs. at least some if not all heterospecific individuals (Table 4.1 and Figure
4.2). However, the difference between V. koshnyi seedling mass in extracts cultured by
conspecific vs. I. deltoidea and P. decurrens individuals only occurred in non-sterile
extract. The AN COVA model assumptions for P. decurrens could not be met when
cotyledon mass was included in total mass since > 10% of seedlings lost cotyledons prior
to harvest. We present results for P. decurrens seedlings without cotyledon mass for all
seedlings and with cotyledon mass for the subset of seedlings retaining cotelydons.
Excluding cotyledon mass, P. decurrens total mass was lower in conspecific vs.
heterospecific (I. deltoidea and V. koschnyi) cultured non-sterile soil extract in high light.
When including cotyledon mass, however, total mass for P. decurrens was greater in
conspecific vs. heterospecific (C. spinosa and I. deltoidea) soil sources in low light.
These apparently conflicting results are due to greater retention of cotyledon mass in P.
decurrens cultured soil extract (results not shown). The inclusion or exclusion of
cotyledon mass did not change results for species.

Sterilization minimally influences seedling performance for one species (H2) —
Sterilization of soil extracts did not influence life span or growth in most species, with the
exception of P. macroloba, where the non-sterile extract decreased life span (Figure 4.3,
Table 4.1; Appendix T), but increased total mass of surviving seedlings relative to
sterilized extract across all soil sources (Figure 4.2g-h, Table 4.1; Appendix U). The
majority of dead seedlings that were identified as diseased were infected by Fusarium

(36%), Rhizoctonia (29%) and Phoma (22%).

97

Irradiance does not interact with pathogen infection (H3) - Light availability did
not influence pathogen effects (i.e., no significant light x sterilization interactions) on life
span or growth for any species. Under higher light, life span was longer for all species
except I. deltoidea and V. koschnyi, which were the most shade tolerant of the species
(Figure 4.4, Table 4.1; Appendix T) and mass was greater for all species except I.
deltoidea (Figure 4.2; Appendix U).

Within each species, initial seed mass did not influence life span but was
positively associated with total seedling mass for P. macroloba, P. decurrens, and V.
koschnyi. Random assignment of seedlings to the high-light bench nearest to the
shadehouse door affected final seedling mass for C. spinosa, P. macroloba and P.
decurrens; exclusion of this bench (with slightly higher light levels) eliminated the bench
effect, but did not change significance or percent differences for seedling mass, Thus, all
data were-used in the analysis and are reported.

Integrated seedling performance — Plant-soil feedbacks can influence
performance of a given species in con— vs. hetero-specific cultured soils (within rows of
Table 4.2; significant comparisons highlighted in Table 4.1), as well as which species of
seedling is most affected by soil cultured by a given species (within columns of Table
4.2). Taking the conventional approach of comparing performance in con- vs. hetero-
specific soils, conspecific soils were never beneficial and were more likely to be
detrimental than at least four heterospecific soils for P. macroloba, P. decurrens and V.
koschnyi seedlings. I. deltoidea seedlings performed better with culturing by P.
macroloba and P. decurrens, worse under culturing by C. spinosa, and performed equally

well under A. membranacea and V. koschnyi culturing relative to conspecific soils. C.

98

spinosa seedlings performed better under I. deltoidea than conspecific culturing, but
heterospecific culturing by the other four species did not differ from conspecific
culturing.

From an alternative perspective of interspecific comparisons of seedling
performance in soils cultured by a given study species, conspecific seedlings were not
necessarily disadvantaged relative to heterospecific seedlings (Table 4.2). For instance,
three species had mixed results with conspecific seedlings less affected than some
heterospecific seedlings (C. spinosa vs. A. membranacea, P = 0.03; I. deltoidea vs. A.
membranacea, P = 0.03 and C. spinosa, P = 0.007 and P. macroloba vs. A.
membranacea, P = 0.02) and more affected than other heterospecific seedlings (C.
spinosa vs. P. macroloba, P < 0.001, P. decurrens, P 5 0.001 and V. koschnyi P g 0.001;
I. deltoidea vs. P. macroloba, P 5 0.001 and V. koschnyi, P 5 0.001 and P. macroloba vs.
I. deltoidea, P 5 0.001 and V. koschnyi, P 5 0.001). Whereas conspecific seedlings had
better performance than heterospecific seedlings (A. membranacea P 5 0.001 and I.
deltoidea P = 0.02 in V. koschnyi cultured soils. Only A. membranacea and P. decurrens
cultured soils were more detrimental to performance of conspecific than heterospecific
seedlings [(C. spinosa, P 5 0.001, P. macroloba, P g 0.001, P. decurrens, P = 0.007, and
V. koschnyi, P 5 0.001 vs A. membranacea) and (C. spinosa, P = 0.001, I. deltoidea, P 5
0.001, P. macroloba, P 5 0.05, and V. koschnyi, P _<_ 0.001 vs. P. decurrens)]

In general, chemical factors in soil extracts influenced seedling performance more
than biotic factors (Figure 4.5). The response of P. macroloba seedlings to all soil
extracts (non-significant for C. spinosa soils) arose from chemical factors, with greater

negative response to con- vs. hetero-specific culturing. For P. decurrens seedlings, both

99

chemical and microbial factors were detrimental in conspecific cultured soils, but
heterospecific cultured soils benefited performance through (either chemical or microbial
factors, depending on soil source) V. koschnyi seedlings performed relatively well in
heterospecific soils due to chemical or microbial factors. Chemical factors primarily
influenced how I. deltoidea seedlings performed; P. macroloba and P. decurrens soils
were beneficial whereas C. spinosa soils were detrimental relative to conspecific soils.
Likewise, chemical factors were detrimental to C. spinosa performance with the greatest
negative effect fiom I. deltoidea soils. Soil extracts decreased A. membranacea
performance (which was similar regardless of soil source) primarily through chemical
factors.

Seedling performance for A. membranacea, I. deltoidea and P. macroloba are
based on effects of soil source regardless of extract treatment or light. For the other
species either extract or irradiance treatments interacted with soil source effects. Thus,
differences in performance among the different sources of soil extracts are restricted to
high light for C. spinosa, non-sterile extract treatment for V. koschnyi, and non-sterile
extract treatment in high light for P. decurrens seedlings (based on mean mass excluding

cotyledon mass).

Discussion

Our experiment revealed a complex web of interactions among tree species,
where the dominant mode of soil feedbacks were mediated through chemical rather than
biotic influences, and where negative feedbacks from conspecific individuals were

stronger than all heterospecific feedbacks for half of the species. In addition, most

100

heterospecific seedlings had better performance than conspecific seedlings in soils
influenced by a given species. However, we only investigated the species-specific effects
of plant-soil feedbacks for six species; if we had added more species, we may have found
other species with greater negative influence, usurping the primacy of conspecific effects
on the three most common species.

Conspecrfic culturing had larger influence on seedling performance than
heterospecific culturing (H I ) — Three of the six species (P. macroloba, P. decurrens and
V. koschnyi) had decreased seedling performance when grown with extract cultured by
con- vs. all hetero-specific individuals. Whereas I. deltoidea had mixed results,
performing better in two and worse in one hetero-versus con-specific cultured soil, and
only C. spinosa performed worse in con- vs. one of the hetero-specific soils. None of the
soil sources here were unifome beneficial or detrimental to seedlings of all species.
Likewise, the relative influences of biotic and cherrrical factors from a given soil source
could change depending upon the species of responding seedlings; however, chemical
effects in soil extracts generally influenced seedling performance much more than biotic
effects. As a whole, these results support fairly specific interactions between pairs of
species that are mediated mainly through chemical factors with conspecific culturing
tending to be more detrimental than heterospecific soil culturing.

An assumption of the J anzen-Connell (J -C) Model is that maintaining species
diversity requires that NCDD processes are more prevalent in species that are common
versus those that are rare, thereby constraining the abundance of common species. In this
study, species that are common as adults from the site where soil was collected (e. g.

Pentaclethra macroloba, Prestoea decurrens, and Virola koschnyi) tended to perform

10]

worse in conspecific vs. heterospecific cultured soil extracts. Thus, tree species
coexistence could possibly be maintained with these processes even though chemical
factors are the main mechanism for reduced performance in the soil and not host-specific
pathogens as assumed by the J -C Model.

There have been numerous studies, like this one, supporting the J -C prediction
that a focal species seedling performance is reduced near versus far distances fiom
conspecific adults (Augspurger 1983a-b, 1984, Augspurger and Kelly 1984, Packer and
Clay 2000, Hood et a1. 2004, Bell et a1. 2006). In contrast, there have been no studies in
tropical forests that have provided direct evidence that a given species, whose seedlings
are disadvantaged in the presence of con- versus hetero-specific adults, would more
likely have its adults replaced by hetero- than con-specific seedlings; leading to greater
local plant diversity than expected in the absence of NCDD processes. In temperate
forests, there is some evidence that heterospecific seedlings have lower mortality due to
soil pathogens than P. serotina seedling in soils cultured by P. serotina adults (Packer
and Clay 2000). In our study, heterospecific seedlings were more likely to have better
performance than conspecific seedlings in soils influenced by a given individual.
However, for many species the results were mixed so that some heterospecific seedlings
were favored whereas others were disfavored relative to the conspecific seedling. The
lack of a consistent disadvantage to conspecific versus heterospecific seedling
recruitment, in this study, may place some limits on the extent that these chemically
mediated plant-soil feedbacks can enhance tree species coexistence in this wet tropical
forest. In addition, although V. koschnyi seedlings did worse in conspecific cultured soils

the heterospecific seedlings had relatively poorer performance. Thus, knowledge of the

102

relative benefits or costs of ‘escape’ from conspecific adults on seedling recruitment does
not translate into knowledge of which species’ seedlings are favored for recruitment near
a particular tree species. But in combination these two approaches provide a better
understanding of how plant-soil feedbacks (or other mechanisms producing NCDD)
contribute to tree species coexistence.

Contrary to expectation, the effect of soil culturing on seedling performance
presumably reflects a chemical rather than a biotic process, since responses to soil
culturing were similar in non-sterile and sterile soil extracts. Ifeffects of culturing had
been due only to biotic agents, then we would have expected negligible differences
among soil origins for the extracts that had been sterilized. In a parallel temperate
experiment, differences in extract cation availability may have been partially responsible
for observed species — specific feedbacks between tree adults and seedlings (McCarthy-
Neumann and Kobe, in review; Chapter 3). However, there are other possible chemical
factor(s) that could mediate plant-soil feedbacks. For instance tannins in leaf litter can
decrease nutrient cycling through decreased decomposition, inhibition of nitrification
and/or changes in microbial activity (Kraus et a1 2003). Additionally, leaf litter leachates
of some species can suppress seedling performance of other species’ seedlings (Stinson et
al. 2006). Our results provide strong motivation for investigating the chemical factors that
could vary among soils cultured by different species of mature trees and their potential
impact on seedling dynamics.

Our soil source results would not have been interpretable or significant for half of
our study species if we had lumped all soils cultured by heterospecific individuals into

one treatment (e.g. “ ar” treatments in Augspurger 1984, Packer and Clay 2000 and 2003,

103

Hood et al 2004). For instance, A. membranacea life span and C. spinosa mass do not
appear to be influenced by soil source when comparing conspecific to a lumped
heterospecific grouping (results not shown). However, soil source does influence seedling
performance for these species but the effect varies among the heterospecific species in
magnitude and whether the effect is beneficial or detrimental relative to conspecific
derived soil (Figure 4.1 and 4.2c-d).

Sterilization minimally influences seedling performance for one species (H2) —
Diseased dead seedlings for all species were infected by a variety of soil fungal
pathogens (F usarium, Rhizoctonia and Phoma). However, only one species, P.
macroloba, responded to the non-sterile extract with a decrease in life span and a slight
increase in mass (Figure 4.4 and 4.6g-h). This result was surprising, because in a prior
study seedling performance for the majority of 21 tropical tree species differed
significantly between sterile and non-sterile extract treatments at low light (McCarthy-
Neumann and Kobe, in press; Chapter 2). In particular, the non-sterile extract from
conspecific soil reduced either life span or total mass for three of the species represented
here (A. membranacea, P. macroloba and P. decurrens) and increased life span for I.
deltoidea. We obtain consistent results between these two studies when we restrict our
current data-set to compare life span and total mass only for conspecific cultured extract
at low light. For instance, life span is reduced in the non-sterile vs. sterile extract
treatment for P. macroloba (3%, x2 = 3.94, df = 1,48, P < 0.05) and mass is reduced for
A. membranacea (25%, F1,” = 2.06, P < 0.10) and P. decurrens (25%, F133 = 3.98, P <

0.10) seedlings. The response of seedlings to non-sterile extract cultured by various

104

species was highly diverse (Figure 4.5), and may be partially responsible for the lack of a
widespread effect of soil microbes on seedling performance in this study.

Irradiance does not interact with pathogen infection (H3) — Contrary to
expectation, increased irradiance did not ameliorate the negative effect of soil microbes
on seedling performance. Indeed P. macroloba ’s life span and grth were affected by
soil microbes without regard to irradiance level, whereas soil origin differences for P.
decurrens seedling growth were only different in the non-sterile treatment at high light.
Tissue compensation and faster 1i gnification are likely not operating here since they
would have led to lower disease effects at high light with our methodology. However,
this study may not have adequately tested whether unfavorable conditions for soil micro-
organisms or increased AMF colonization reduce disease effects at high light since we
minimized abiotic differences between light treatments and excluded AMF spores from
microbial extracts. Differences in methodology may explain why our results contradict
studies that did allow abiotic factors (e. g. soil moisture and temperature) and/or AMF
colonization to vary between irradiance levels (Augspurger 1984, Augspurger and Kelly
1984, Hood et al. 2004).

Plant-soil interactions affected seedling performance by as much as 60%, which is
similar in magnitude to the effect (~0-130%) that an increase in light from 1% to 5% full
sun had on mean seedling mass for the same species. This moderate increase in light
encompasses the endpoints typically encountered from understory to small tree fall gaps
in forests at La Selva Biological Station (Chazdon and Fetcher 1984). Thus, plant-soil
feedbacks could have a large influence (perhaps similar in magnitude to that of varying

irradiance) on community dynamics and composition in tropical forests.

105

Caveats. — Excluding the filtrate component > 20 pm in soil extracts may have
underestimated the role of soil microbes in plant-soil feedbacks, by eliminating some
pathogens that could be present in field soil. In addition, extract culturing by trees of
different species may be confounded with an underlying template of soil chemistry that
determines tree species occurrence. Lastly, our study tested only the simultaneous effect
of plant-soil feedbacks and light availability. Interspecific competition, another process
critical in recruitment dynamics, has been shown to influence whether plant-soil
feedbacks occur for grassland species (Casper and Castelli 2007, Kardol et a1 2007), and
a variety of other biological processes may interact with plant-soil feedbacks as well.

Comparison between tropical and temperate forests.—There is accumulating
evidence of NCDD seedling mortality and reduced growth (Packer and Clay 2000, Hille
Ris Lambers et al. 2002, Packer and Clay 2003) in temperate systems which have called
into question the long-standing assumption of the Janzen-Connell Model that NCDD
processes are less pronounced in temperate vs. tropical forests. We wanted to explicitly
test this assumption by directly comparing seedling responses to NCDD processes
between tropical and temperate species with parallel experiments designed to enable
direct comparison. We found that soil-mediated (likely through chemical factors in the
soil) species — specific feedbacks between tree adults and seedlings were ubiquitous in
both temperate and tropical systems. However, the feedbacks between temperate species
were largely idiosyncratic and were not consistent with the Janzen-Connell Model since
soil cultured by heterospecific species were more likely to decrease seedling performance
than soil cultured by conspecific individuals. In addition, soils cultured by a particular

species did not necessarily improve heterospecific seedling performance relative to

106

conspecific seedlings (McCarthy-Neumann and Kobe, in review; Chapter 3). In contrast,
in the tropical forest, species that were common as adults fiom the site where soil was
collected tended to perform worse in conspecific vs. heterospecific cultured soil extract,
and soils cultured by a particular species were more likely to favor heterospecific
performance relative to conspecific seedling performance. These results suggest that
chemical mediated plant-soil feedbacks are an important component of seedling
dynamics in both temperate and tropical forests; that these feedbacks can create complex
interactions between tree species, and currently there is more evidence for these

feedbacks enhancing species coexistence in tropical than temperate forests.

Acknowledgements

We thank Mindy McDermott, Marisol Luna, Ademar Hurtado, Martin Cascante,
Edgar Vargas and David Neumann for their invaluable assistance in the field, greenhouse
and laboratory. We also thank Mike Walters, John Klironomos, Andy Jarosz, David
Rothstein and Tom Baribault for comments on earlier drafts of this manuscript. The
Organization for Tropical Studies and the staff at La Selva Biological Station provided
logistic support. This research was firnded by the National Science Foundation (DEB

0235907)

107

.305 A. _. Cm” om mmmimoma moon—Em 5.0 when 25 88.— 358 480268 wen u: m8?~ muoomom 8” me: 858 A38 muoowam 9:835

 

mo: . me: 9389 boa-Amado <m. man—o . ~50 535:8 HRH <m. max. me: 25 850953 0.08 > 0:058 a me C 25 Hugo ab .

 

 

 

moEdo e». muoomom Ema mvgm Hog: 88m“ Sandman woman—m memo—.8883
<~1mmo=
:5 >5 <m. Sn AH .00. w -I- 0.03 20 awn—Bomb» 03.0858 20 amines“: Emma—dean
Om <m. >3 A000. do H 0.008
on <m. 55 A000. m n 0.08 am <m. E 5 E =er A+0m.x: w n . . o n
9. Om. <m. 00 A0. m 0. me n 0.03 0.08 Om <m. E S m 0%: A+>0 \o. m 0.003
Om. 5... Sn A000. w H 0.008
E <m. 9 APR»: Nu u 0.0000 c I E <m. Om 13.x: me H 0.008
a a a... we 8.5. w n 0.88 M M”. wmsAflwxswn-oowws a a. we 29x. w u 988
a s... E 3.3. w u. 98: . a. . a s... 2 Sex. w u 983
wB <m. >8 A-EX: m u 0.08 5: <m. >8 A-HwoA: do n 0.00 v
we: . . . ~08 <m. Om A-EX: me u 0.03 T: <m. 9 AL 1x: Nu H 0.000
mafia as zo 2838.: 94388 as a. E E E: w u 980 we a. E 03... w u 98 0
53%. <wA-_~.x..wu0.03 35%. <wA-:.xien0._0v
PM <m. >3 E Z-m 8x55 86 I .53
A-uufi: w u 0.03
mm <m. Om m: Z-m 353 ES E mm! A-
a <m. E E Z-m exam..." 85 m 00:? w n 00$
we? ma <m. Sn A0.~0. me H 0.003 0%: Tax: do .I- 0.0: mm <m. E S Z-m 9:82 85 m 0%: A-

00 <m. Sn 5 Z-m 8239 85 m
smE A-wflx: w H 0.000

max: w n 0.0000
mm <m. 3: m: Z-w 95.2 3.0 3 mm?
ALSO? m n 0.03

~00 <m. Sn 3 Z-m 9989 85 E mm! A-

te... w u 980

108

0.8.0 A. 0. A0000

2 a. Q a 0. :8 3:. w n
8:

 

 

000 00 <m. Sn 8.3. w H 0.008 000 S... 00 8 H. 0%.: A+0o\o. me n 7:?
0.0000
Sn <m. >8 8 Z-m 880 A-mmfi: w H
0.008
Sn <m. >3 A-NmX. do u 0.0000 . a n
Sn s... o... 3 ex. 8 u 983 Sn «wows z-m 982 08 x... 0.
Sn <m. 00 8 Z-m 80.8» A-wfix: . . c n
Ba 20 $8508" 05.988 0c M 0.0000 SA “40%“: Z-m 80.80 A-Nq \9 0e
. Sn <m. wB A-NAX. w H 0.003 . . o u
§ 5.. E a z-m gene 08:. SA “new“: 5 z-m axes, to \o. w
w u 8: <w s... E a z-m 983 03:. w n m
0.0: 1
05 20 08508" 05.988 20 08583 002988 2.?
0» 2o «0800080 05.988 20 4.080008" 05.988 Z>
m 0 H0 20 080083 0000.988 20 £80009: 05.988 Z>
mmefi es z-m a. m 3.9. w n 98.0 z-m 5.. m as. w n 83 z>
«.00 2o «08509: 05.988 20 38588 0000988 Z>
E0 20 £8508" 000.0988 20 885080 05988 Z>
§ 20 «0800080 000.0988 20 380008. 05.988 Z>
03 H- <m. m Amdm. m u 0.0000 H. 8. m A-0_.x.. do u 0.0000 Z>
9. H. <m. m ANLw. w u 0.0000 H. <m. m ALEX. do n 0.0000 Z>
H0 20 £80.88" 05.988 20 8800080018988 Z>
0.80088 3: H. <m. m APmu. w n 0.00: H. <m. m A-quo. w H 0.005 Z>
we? r <m. I ANS. m H 0.003 - <m. m 30.x. w H 0.008 Z>
$0 H. <m. 3 APS. 0e H 0.003 20 Ema—00.880 05.988 Z>
SA 20 38588 05988 H. <m. 00 AL 0.x: 00 u 0.0000 Z>

 

Hugo 3.0. “000w

288.. 9.0390 808 v 0 30038 80803 3 03m 8 82.35. o0. 9:00. 3809. 3.080 39m 839 98 En 8308 3 9 088008
A+v @983 000.0938 3 83:3 8338 8283 033850 30 80980850 98 3: 3589. 53-9908 30 £908 398". 30
092 30 Em: mmE. o 3.983 833.89 800:5 39m 000 so” 0:00:08 83.0003 :89. 3.983 833.89 88055 :89 000 no"
08500 830803 :59. b99803 00. 30838039 H. n 00$ :3" 30 E n Em: :mrn Z..m N 33.39008 30 m n 39:9 >8 n
9888 5983383. Om "080883 .3883. 00 H 0.833 808803. 3: u $388883 38883. 00 H 3.983 833.89. 30
SA n 3.88 088038.. Z> u no" 330320 308 38¢.me £9 :2 830880. my $009 0.9 8:538 833038 038 83 800580
:09 8 H0008 80.33830.

 

110

Hugo #9. “#8632: 3.?on 6383 83.0338 5 58%“:ko monazsm Ramona—goo :88: 88. 89% x 385 an mama \ 38% cm

£31833: c». 28.78: $2?»on man go: macaw muaomam mammamfioa 88%. 9:89 nag—:38 Ba wan—Egon 53—.

 

 

magma. om >9: 0:555 :5 mo:

 

 

 

 

 

 

muaomom om maaqmam Womuoambm AER? 022698 Niazwn wmafifimurwa 3.8.83 $.on
8 mo: 583963an 9.3398 mmNBEma 382683 mmgx‘maa range“.
Lcfi. Lug Lufi. -uofi. -wmfc LSX.
3‘»me Swivwaanndn
Tmu. -Naoé A-mh. -Bfio A-mu. -395 TE. booby TS. -prov A-mo. -mNfib
mfg Lufi. .mm..\.. --.x. .m..\.. Lug
00:535 €303
TC. 3.5 ALWN. «.5 $3. $15 Aam. -mfiv Ahm. 3.x; Aha. fix;
-Nmfi. -mmfi. buck. fax. 3.x. -NQX.
Nialma mmNBRma
ALE. Lax; A-a©. Lifev 3%. Lwfio Ow. max; 3. 3.x; A53. Lmfio
-mfo LOX. Luck. -Bfi. LNR. -ofi.
wmzuaanSE Sawing
Tr Max; TE. -15 Tub. bake TS. Lax; ALP -m..\ov $3. Leno
1x. L3 3.x. Lfi. use} 5.x.
3.38% «@2363
Tub. ~15 AL? 2.5 Go. unfiov Two. 3.5 A-mu. Lugv $9 3.x;

 

111

6min Pm. A93

 

8.x. Nofi. Nix. 3.x. 5.x. .m..\:
3.33 ‘88::va
Co. Nmfov C H. ~95 Co. ”3.x; 3. MEX; G. Nocé ALP Mono

 

 

205. Wag—3 m9. 0. 9.3398 @83on m3 #2: Em: _mmE. V. wougai $330me mdB 505-“..81—0 9:39 @833». 85 my.
«@3365 333on 26 W08 “anion—a 8282 9.8333 F Em: _mmE. mayo" om 836830 Ash—88¢ mom on mogmnmm mm 5 :05.

wooafifi mafia 3.x. 9 m3 5 33:983.

112

a) Apeiba membranacea 5) Colubrina spinosa

 

 

   

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

1 Soil 100— Soil
Source Source
$5 8“ .rIAm g 30" .nAm
:: .nCs — I105
g 60'— “:Hz’d g 60— ":":Id
g ..="=Pm 3 ..1"=Pm
w 40- _!.,Pd :0 4o— _flpd
" £in
20‘ I T I I -. Wk 20‘ ' ' * ‘ '
o 20 4o 60 so 0 ?° 4° 50 80
Life span (day) LIfe span (day)
c) Iriartea deItoidea d) Prestoea decurrens
100" Soil 1oo~ Soil
Source Source
35 30‘ .nAm g3 80— mm
717 J‘le 2’ .rICs
E 60. .:":[d g 60" .,:":Id
3 .5”me 3 “HP”?
0) 40— .3”; Pd (0 40“ ..'”;Pd
-’"IVk 5'in
20— I I I I I 20‘ , , I . .
0 20 40 60 80 0 20 40 60 80
Life spa" (day) Life span (day)

Figure 4.1. Survival curves for study species [a) Apeiba membranacea, b) Colubrina
spinosa, c) Iriartea deltoidea, and d) Prestoea decurrens] with significant effects of soil
source (tree species culturing soil: Am = Apeiba membranacea, Cs = Colubrina spinosa,
Id = Iriartea deltoidea, Pm = Pentaclethra macroloba, Pd = Prestoea decurrens and Vk =

Virola koschnyi). Arrows indicate seedling response to conspecific cultured soil source.

113

Figure 4.2. Total final mass by soil source (tree species culturing soil) and extract (sterile
vs. non-sterile) for each study species in low and high light (Apeiba membranacea (Am)
a-b, Colubrina spinosa (Cs) c-d, Iriartea deltoidea (Ia) e-f, Pentaclethra macroloba (Pm)
g-h, Prestoea decurrens (Pd) seedling mass without cotyledon mass i-j and seedling mass
with cotyledon mass k-l, and Virola koschnyi (Vk) m-n). Dotted line represents seedling
mass when grown with tap water and is shown for reference only, and were not included

in statistical analysis.

114

Figure 4.2. (Ctd)

7. 51 a) Apeiba membranacea (1% full sun)

6.01

4 5 _ IISterile
DNon-
1 5 H Sterile

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

 

Total Mass (mg) + 95% Cl

 

 

5 20 j b) Apeiba membranacea (5% full sun)
$ I
O)
i -lSterile
E’
V DNon-
8 Sterile
m
2
E . , r.
.2 Am Cs Id Pm Pd Vk

Species of Adults Culturing Soil
5 25 c) Colubrina spinosa (1% full sun)
o\°
8 20 :L
+ . ' ‘ 'ISterile
g 15 ‘
7,; 10 ’4 [ZlNon-
g 5 W; Sterile
E l
I9 Am Cs Id Pm Pd Vk

Species of Adults Culturing Soil

115

Figure 4.2. (Ctd.)

0 1d) Colubn'na spinosa (5% full sun)

‘I’ I I
I Sterile
El Non-
Sterile

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

0

 

41003-501
00
1

Total Mass (mg) + 95% Cl
C O

00 7 e) Iriartea deltoidea (1%full sun)

18° * lSterile
120 - EINon_-
60 - Sterile

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

NO)
A
O

 

Total Mass (mg) + 95% Cl

6 300 Hf) Iriartea deltoidea (5% full sun)

 

3\: 240

o: L}.

E3180 1‘ lSterile
i3 120 1‘ EINon-
g 60 7‘ Sterile
E

I9 o I

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

116

Figure 4.2. (Ctd.)

5 2250 g) Pentaclethra macroloba (1% full sun)
\°

0

 
  

    

8 1800 f‘ " ----------- --

+ I _

g, 1350 a“ I Sterlle
l

70’ 900 3 El Non_—

3 Sterile

E

E

.2

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

3000 jh) Pentaclethra macroloba (5% full sun)

6

e\°

ID ________

G)

+ .
A I Sterlle
O)

E

1;; El Non-
3 Sterile
2

T!

,9

‘I

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

100 ,1 i) Prestoea decumans (1% full sun)

  

so 3

lSterile
4o ,,
20 i

0*

El Non-
Sterile

Total Mass (mg) + 95% CI
o:
0

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

117

Figure 4.2. (Ctd.)

6120 .1 j) Prestoea decunens (5% full sun)

 

  
  

$100 —
O) 1' - T I -
i 80 i St V'l
l er

g 60 l Ie
g 40 s‘ CINonl-
2 [ Sterile
_ 20
9 l
O .
l— ‘I T .

Am Cs Id Pm Pd Vk

Species of Adults Culturing Soil
5 175 g k) Prestoea decun'ens (1 % full sun)
°\O ........................ -- -
3 140 .3
+ l
E3 105 ‘ lSterile
g 70 " DNon-

I Ster'le
E 35 3 '
E
['2 0 *' “r I i”

 

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

175 A I) Prestoea decun'ens (5% full sun)

  

5

°\° __ - _ _

g: 140 ~‘

+

3105 “I flSterile
g 70 j EJNon-
:2“ 35 M Sterile
g I

{3 0

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

118

Figure 4.2. (Ctd)

 

   

61000 m) Virola koschnyi (1% full sun)
33 l
3 800 f
j; 600 ._ __ __ _- _ lSterile
m 7.
E
v .. EINon-
g 400 Sterile
g 200 .
'2 0 4' T 7" I

Am Cs Id Pm Pd Vk

Species of Adults Culturing Soil

51 1250 n) Virola koschnyi (5% full sun)
e\°
31000 -I
i 750 _ ------- lSterile
g [ZINon-
g 500 7 Sterile
E 250
9:.
.9

Am Cs Id Pm Pd Vk
Species of Adults Culturing Soil

119

Pentaclethra macroloba

 

 

100- .. Extract
_\ _nSteriIe

A 95—d ‘-..... .--.NOI"l-
9‘5 "'-..,. 'sterile
f? 90-1 "
E
3
(O 85...

80"

 

 

 

l l l l l
0 20 40 60 80
Life span (day)

Figure 4.3. Survival curve for Pentaclethra macroloba seedlings by soil microbial

treatment (sterile vs. non-sterile extract) and integrated across soil source and irradiance

level.

120

a) Apeiba membranacea

Survival (96)

Survival (96)

100-
80‘
60— .
4o—
20~
I I I I I
0 20 40 6O 80
Life span (day)
0) Pentaclethra macroloba
100_ "'0.. T
8%
60—
40"
20-
l l l T l
0 20 4O 60 80
Life span (day)

 

 

 

 

 

 

 

 

 

 

Light
_I'IHigh
. i' ELow

Light
_rIHigh
- ' fLow

b) Colubrina spinosa ,

Survival (96)

Survival (96)

 

 

 

 

 

 

 

 

100- _. Light
J'IHigh
80— "a... -"!Low
40.4
20-
f I I I I
O 20 40 60 80
Life span (day)
d) Prestoea decurrens
100- \ Light
J'IHigh
80~ ”‘ -.' Low
60-
40-
20—
I I I I I
0 20 40 60 80
Life span (day)

Figure 4.4. Survival curves for study species [a) Apeiba membranacea, b) Colubrina

spinosa, c) Pentaclethra macroloba, and d) Prestoea decurrens] with significant effect of

light level (high light; 5% full sun vs. low light; 1% full sun) and integrated across soil

source and soil microbial treatment.

121

a) A. membranace b) C. spinosa 0 Am

 

 

 

 

 

 

 

 

 

30 '1. I Cs
A L I A Id
8 0 ‘ a:- " ‘flrtmr _. ,mmg me
g -30 I 4' Pd
E I o Vk
o I
E -60 -
CL
8’
% (0)0" deltoidea d) P. macroloba
$
U) I i
. Aer—Ii ea. ~

I

C ' 5
a) -30 ,
*6 l I
a: . :
5 -60 . .
.3
‘8' e; P. decurrens f) V. koschnyi
it: 0 '1
Lu If
g o . n... I. _._.§
9 i
.2 _ _ ‘
2 30 l

-60 ~ J

 

-30 -30 0 30 30 -60 -30 .0 30 60
Chemical Effect (% Difference in Seedling Performance)

Figure 4.5. Relationship between chemical [sterile extract / tap water) and microbial [(un-
sterile extract / tap water) — (sterile / tap water)] effects in soil extracts “cultured” by
different tree species on seedling performance [(mean total mass x mean life span) / (days
of experiment)]) for each study species integrated across irradiance levels, except for C.
spinosa and P. decurrens seedlings whose results are only from high light. Am = Apeiba
membranacea, Cs = Colubrina spinosa, Id = Iriartea deltoidea, Pm = Pentaclethra
macroloba, Pd = Prestoea decurrens and Vk = Virola koschnyi). Bootstrap devised 95%
CI included.

122

CHAPTER FIVE

CONCLUSION

My goal in this dissertation was to test the major assumptions of the J anzen-
Connell (J-C) Model using plant-soil feedbacks as the particular agent for creating
differences in tree seedling performance. In this chapter I will briefly outline the major
results from my dissertation and suggest avenues of research that may answer questions
raised by the research.

BLant-Soil Feedbac_ks - The preceding chapters have documented that species-
specific plant-soil feedbacks influence both temperate and tropical seedling performance
which may in turn affect tree community dynamics and composition. The Janzen-Connell
Model can be placed within the larger context of negative feedbacks between plants and
the soil in which they grow. Surprisingly, the effect of adult culturing on seedling
performance occurred in both non-sterile and sterile soil extracts presumably reflecting a
chemical rather than a microbial process. This effect may be mediated by tree species
induced differences in base cation availability, at least for the temperate species since the
amount of base cations in the temperate soil extracts varied depending upon the species
of tree that cultured the soil. This is a novel mechanism for producing NCDD mortality
and/or reduced growth. Experiments linking soil chemical factors with species-specific

feedbacks between tree adults and seedlings are still needed to fully ascertain which

123

chemical factors are responsible for creating these feedbacks and documenting the full
extent of these feedbacks on a variety of tree species.

Distancefi and Densig-Dgpendent Processes - In the temperate study (chapter 3), I
determined that plant-soil feedbacks created by different species of adults culturing the
soil were generally not influenced by the species of seedling culturing the soil nor the
density of those seedlings. Thus, my results suggest that plant-soil feedbacks are more
important in creating distance (e. g. conspecific vs. heterospecific) rather than density-
dependent effects on seedling performance. The lack of a density effect on seedling
performance in my study differs from the positive relationship between seedling density
and disease induced seedling mortality documented by other researchers (Packer and
Clay 2000, Bell et a1. 2006). Conflicting results among studies might arise because of
different methodologies, especially if density enhances disease transmission through
close proximity of roots, which was eliminated in our extraction-based study.

For temperate species heterospecific and conspecific adult interactions with the soil had
similar prevalence and magnitude of influence on seedling performance, contrary to J -C
expectations. For tropical species, however, negative feedbacks from conspecific
individuals were stronger than heterospecific feedbacks for the majority of the species
(chapter 4).

I find evidence in both chapters 3 and 4 that a species-specific approach was
more appropriate than simply comparing seedling responses to near vs. far conditions. In
addition, by comparing species-pair interaction to these plant-soil feedbacks, I was able
to assess whether reduced seedling performance from non-competitive distance-

dependent mortality and/or growth in a particular species contributed to greater success

124

of heterospecific seedlings. Surprisingly, the large negative impact that conspecific
cultured soil had on a focal species’ seedlings did not necessarily result in heterospecific
seedlings having relative better performance. Future work should concentrate not only on
the relative benefits to seedling recruitment of ‘escape’ from conspecific adults, but also
the relative recruitment success among species dispersed into neighborhoods influenced
by heterospecific adults. In combination, these two approaches to NCDD processes will
provide a better understanding of how well plant-soil feedbacks enable species
coexistence.

Host-Specificity — Soil pathogens decreased seedling performance for three
temperate and one tropical species, but this negative biotic effect was consistent
regardless of which species of adult had cultured the soil, which indicates that these soil
pathogens may not exhibit host-specificity. When investigating pathogen host-specificity
in the temperate species I found that infection by five F usarium morphotypes, isolated
from roots of dead seedlings, affected seedling performance differentially among these
species but were neither strictly host-specific nor strictly generalist.

Seedlings of the study species did not react uniformly to any of the soils cultured
by a particular species (i.e. no soil extract was either the most beneficial or the most
detrimental for all of the tree species). This result suggests that the chemical factors
mediating the plant-soil feedbacks documented in my dissertation differentially influence
seedlings of different species, which is akin to pathogen host-specificity. Future research
is necessary not only to determine which chemical factors are responsible for creating

these feedbacks, but also how they affect these species differently and whether there are

125

any species traits that make them more likely to be susceptible to these chemically
mediated plant-soil feedbacks.

Light and Disease Interaction - Irradiance levels did not influence response or
susceptibility to disease for my tropical study species, but the negative effect of soil
microbes on growth occurred only in high light for two of my temperate study species.
These results suggest that growth for some species is so severely constrained in shade
that the effect of soil microbes were not manifested, but when light limitation is
alleviated microbial impacts can become apparent. Light availability did not affect how
seedlings responded to soil extract cultured by conspecific vs. heterospecific adults in a
consistent manner; for most species feedbacks were consistent regardless of light
availability, for a few species there were greater negative effects on seedling mass from
conspecific than heterospecific cultured soils in low light and for a few other species this
pattern occurred only in high light.

By simultaneously testing the effects of plant-soil feedbacks and light availability
on seedling performance in both temperate and tropical species, I can compare the likely
impact that these two mechanisms have on community dynamics and composition. Plant-
soil interactions affected seedling performance by as much as 40% in the temperate and
60% in the tropical forest, a magnitude of effect that would be expected to influence
community dynamics and composition. Light availability, primarily through changes in
mean seedling mass, had a much larger impact (between 90-3000% increase from 2% to
22% full sun with temperate species and O—l30% increase fi'om 1% to 5% full sun with
tropical species) on seedling performance. Thus, plant-soil feedbacks may be similar in

impact to moderate increases in light availability that occur fi'equently in the forest

126

understorey, but play a secondary role to the increases in light availability created from
less frequently forming large canopy gaps in tree community dynamics.

Species’ Abundance — In the second chapter working with 21 tropical tree
species, I demonstrated that susceptibility to soil microbes cultured by conspecific
individuals co-varied with a species’ seedling shade intolerance rather than local
abundance. Thus, rather than soil micro-organisms causing negative feedbacks that
constrain the abundance of common species, my results suggest that soil micro-organisms
may exaggerate seedling shade tolerance differences among species which in turn may
influence species coexistence through enhancing light gradient partitioning. I then used
six of these tropical species (which varied in shade tolerance and abundance) to test the
relative effect of conspecific vs. heterospecific cultured soils on seedling performance.
My results (outlined in chapter 4) suggest that common species are more likely to
experience greater negative effects on seedling performance from conspecific than
heterospecific cultured soils than species that are locally rare. A major assumption of the
J -C Model is that maintaining species diversity requires that NCDD processes are more
prevalent in species that are common versus those that are rare. Thus, tree species
coexistence could possibly be maintained through plant-soil feedbacks even though
chemical factors are the main mechanism for reduced performance in the soil and not
host-specific pathogens as assumed by the Janzen-Connell model. Additional species
(ranging in local species abundance) need to be tested for the relative effect of plant-soil
feedbacks by conspecific vs. heterospecific adults before a general consensus can be

reached that these feedbacks are more detrimental to locally common species.

127

Temperate vs. Tropical Forests — Implicit in the J -C Model is the assumption that
NCDD processes are less pronounced in temperate vs. tropical forests. I explicitly tested
this assumption in this dissertation and that soil-mediated (likely through chemical
factors in the soil) species - specific feedbacks between tree adults and seedlings were
ubiquitous in both temperate and tropical systems. However, the feedbacks between
temperate species were largely idiosyncratic and were not consistent with the J anzen-
Connell hypothesis since soil cultured by a heterospecific species was more likely to
decrease seedling performance than soil cultured by conspecific adults. In addition, soils
cultured by a particular temperate species do not necessarily improve heterospecific
seedling performance relative to conspecific seedlings (chapter 3). In contrast, tropical
species that were common as adults fi'om the site where soil was collected tended to
perform worse in conspecific vs. heterospecific cultured soil extracts (chapter 4). In
addition, with tropical species, heterospecific seedlings were more likely to have better
performance than conspecific seedlings in soils influenced by a given individual. These
results suggest that chemical mediated plant-soil feedbacks are an important component
of community seedling dynamics in both temperate and tropical forests; that these
feedbacks can create complex interactions between tree species, and currently there is
more evidence for these feedbacks enhancing species coexistence in tropical than
temperate forests. However, more work needs to focus on the relative success in seedling
recruitment among species dispersed into neighborhoods influenced by adults of different
tree species. In addition, studies need to begin explicitly testing whether, over time, plant

diversity is greater than expected with plant-soil feedbacks (either biotic or abiotic

128

mediated) compared with random survival of seeds and seedlings in both temperate and
tropical forests.

In summary, this work has demonstrated that negative plant-soil feedbacks are
important in both temperate and tropical forests, but these feedbacks primarily are
mediated not from host-specific enemies, but from chemical feedbacks that differentially
affect seedlings of different species. Not all species do better with soils cultured by
conspecific versus heterospecific adults, but tropical species that are locally common
appear to be more likely to be affected in this manner. My dissertation research has raised
as many questions as it has answered. I plan on investigating some of these questions,
and I hope that the results presented in this dissertation will also stimulate more research
into the role that plant-soil feedbacks (particularly those mediated by chemical factors)

have on tree community dynamics.

129

Appendix A. Kaplan-Meier analysis of seedling life span compared between microbial
vs. sterilized microbial extract for all study species.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Species Extract at: 22:30 Breslow, x P value
Luehea seemannii Microbial 38.90 (4.51) 4 25* P = *
(Intolerant / Common) Sterile 50.60 (4.13) '

Neea psychotroides Microbial 60.53 (3.92) 1 70° P = NS
(Intolerant / Common) Sterile 65.72 (2.29) '

Apeiba membranacea Microbial 49.51 (4.40) 0 43. P = NS
(Intolerant / Intermediate) Sterile 52.47 (4.10) °

Colubrina spinosa Microbial 63.77 (2.65) 0 31. P = NS
(Intolerant / Intermediate) Sterile l 63.10(2.84) '

Guatteria diospyroides Microbial 66.61 (2.03) 1 19 P = NS
(Intolerant / Intermediate) Sterile 63.33 (2.70) '

Psychotria panamensis Microbial 66.64 (2.41) 0 35 P = NS
(Intolerant / Intermediate) Sterile l 68.00 (1.97) °

Castilla elastica Microbial 68.05 (1.95) 1 00 P = NS
(Intolerant / Rare) Sterile 70.00 (0.00) '

Coussarea hondensis Microbial 47.57 (4.90) 3 19. P = I
(Intermediate / Common) Sterile 54.60 (4.19) '

Pentaclethra macroloba Microbial 70.00 (0.00) 1 00 P = NS
(Intermediate / Common) Sterile 68.90 (1.08) °

Prestoea descurrens Microbial 59.02 (4.61) 2 83° P = 1'
(Intermediate / Common) Sterile 65.81 (2.87) '

Quararibea bracteolosa Microbial 43.47 (6.47) 0 08- P = NS
(Intermediate / Rare) Sterile 47.73 (7.64) '

T rophis racemosa Microbial 59.78 (4.18) 0 31. P = NS
(Intermediate / Rare) Sterile 61.92 (3.78) '
Vochysiaferruginea Microbial 50.97 (5.08) 0 28- P = NS
(Intermediate / Rare) Sterile 52.07 (4.76) '

Capparis pittieri Microbial 63.70 (2.54) 0 07- P = NS
(Tolerant / Common) Sterile 63.17 (2.83) '

Euterpe precatoria Microbial 64.07 (2.73) 1 37 P = NS
(Tolerant / Common) Sterile 55.30 (4.45) '

Iriartea deltoidea Microbial 65.45 (2.52) 4 32* P ___ *
(Tolerant / Common) Sterile 57.23 (3.94) '

Welfia regia Microbial 70.00 (0.00) 2 81* P = I
(Tolerant / Common) Sterile 66.70 (2.12) '

Virola koschnyi Microbial 70.00 (0.00) 1 00 P = NS
(Tolerant / Intermediate) Sterile 65.892 (3.97) '

Dussia macrophyllata Microbial 53.79 (6.89) 0 29 P = NS
(Tolerant / Rare) Sterile 48.93 (6.88) '

Miconia afiim’s Microbial 96.90 (3.91) 1 31* P = NS
(Tolerant / Rare) Sterile 85.92 (6.42) '

 

130

Appendix A. (Ctd)

 

Stryphnodendron microstachyum Microbial 56.67 (4.35)
(Tolerant / Rare) Sterile 47.07 (4.73)
Notes: Seedlings that did not emerge are part of the survival analysis with their death
date given as the mean number of days prior to emergence. “0” indicates that the
Breslow statistic for effect of microbial extract was adjusted for the effect of adult
culturing the soil. ”*” indicates that the Breslow statistic for effect of microbial extract
was adjusted for the effect of bench. Degrees of freedom for all species are 1.
Significance is shown as: NS, not significant; 1' P g 0.10; * P _<_ 0.05; *"‘ P 5 0.01; ***
P 5 0.001. When the extract treatment is significant (P 5 0.10) the means between the
microbial and sterile treatment means are in bold.

1.140 P=NS

131

Appendix B. Analysis of covariance results for the effects of inoculum treatment on seedling
extract; adult = location where soil was collected for use in extraction; bench = location 039

g

 

 

 

 

 

 

 

Species
(Shade Tolerance / Adult Dependent Variable Inoculum (1)
Abundance Classrfrcatron) N F P Means (Std. err
Microbial vgj
Luehea seemannii Root Mass 22 0.44 NS 0.06 (0.09) vs.l
(Intolerant / Common) Stem Mass 23 0.79 NS 0.34 (0.03) v3.1
Leaf Mass 23 0.48 NS 0.42 (0.06) vs.(
Total Mass 23 0.16 NS 0.83 (0.16) vs.(
Root Length 23 0.06 NS 0.46 (0.16) vs. (
Root Surface Area 23 0.29 NS 0.06 (0.02) vs.(
Height 23 0.52 NS 16.33 (1.21) vs.
Neea psychotroides Root Mass 48 0.03 NS 2.38 (0.23) vs]
(Intolerant / Common) Stem Mass 48 0.92 NS 3.98 (0.21) vs.4
Leaf Mass 48 2.44 NS 12.72 (0.84) vs.‘
Total Mass 48 2.09 NS 19.06 (1.13) vs?
Root Length 48 0.29 NS 12.56 (1.03) vs:
Root Surface Area 48 0.08 NS 2.08 (0.18) vs. 2
Height 48 0.67 NS 32.17 (1.15) vs;
Apeiba membranacea Root Mass 20 0.10 NS 0.68 (0.07) vs. 0
(Intolerant/ Intermediate) Stem Mass 21 0.69 NS 1.54 (0.08) vs. 1.
Leaf Mass 21 6.38 * 2.26 (0.22) vs. 2
Total Mass 20 4.26 1' 4.61 (0.30) vs. 5
Root Length 20 0.00 NS 3.27 (0.50) vs. 3
Root Surface Area 20 0.08 NS 0.54 (0.07) vs. 0
Height 21 0.02 NS 23.15 (2.27) xi
Colubrina spinosa Root Mass 47 0.06 NS 2.87 (0.18) vs. 2
(Intolerant / Intermediate) Stem Mass 47 0.79 NS 5.70 (0.31) vs. 6
Leaf Mass 47 0.59 NS 13.62 (1.00) vs.
Total Mass 47 0.87 NS 22.09 (1.39) vs.f
Height 47 2.38 NS 56.58 (2.96) vs.)
Guatteria diospyroides Root Mass 36 3.02 1' 5.20 (0.28) vs. 5.
(Intolerant / Intermediate) Stem Mass 36 0.18 NS 10.19 (0.31) vs.‘
Leaf Mass 36 0.85 NS 5.58 (0.88) vs. 6
Total Mass 36 0.59 NS 21.34 (0.97) vs.1
Root Length 36 1.81 NS 19.89 (1.30) vs.i
Root Surface Area 36 1.43 NS 3.75 (0.21) vs. 4.
Height 36 0.34 NS 74.38 (2.79) vs;
Psychotria panamensis Root Mass 47 0.05 ' NS 3.63 (0.16) vs. 3.
(Intolerant/ Intermediate) Stem Mass 47 0.28 NS 5.80 (0.27) vs. 6.
Leaf Mass 47 0.16 NS 9.49 (0.60) vs. 9.
Total Mass 47 0.00 NS 18.87 (0.64) vs. i
Root Length 46 0.00 NS 26.09 (1.15) vsl

132

£13: stem height and root morphology. Inoculum = microbial vs. sterilized microbial
gigs were randomly assigned to during the experiment; covariate = initial seed mass.

Source of Variation (df)

 

'—

 

 

 

 

 

 

 

Adult (3) Bench (5) Covariate (1) Cgvlirtriirte (1)

115:- ” F P F P F P F P

1'01".

1.. 1.07) — — — —— 2.45 NS — —
.102) —— — ,— .— 5455 *** .— _
1'! "-05) — — — —— 23.98 111* _ _
1:11.12) — —— — — 17.33 *** _. _
13913) — — — — 41.74 *1. __ __
I1.1.02) — —— —— —— 67.97 *** _ _
(o. 97) —— — — — _ _ _ _
” 22) 4 25 ** —— —- 2 37 NS — _—
23.0) . 3.19 * 1.66 NS _ _ _ __
(0 79) 9.66 *** — — —_ _ _ _
6 :: (1.06) 9.65 *** —— — _. _. __ _
(0. 97) 8.07 *** — — — _ _ _
. ~ 17) 2.62 1 — — ._ _ __ .—
(1.10) —— — _ _ ._ _

/ 06) 2.31 NS — — — _ _. _—
.1111 07) — —— —— — 26.53 *** _ _—
.11; 20) 1 72 NS — —— 2.04 NS — _
90-925) — — — —— 10 33 1' — —
if 41) — -— — — — — — —
1- .05) _. _. __ _ _ _ _ _
-.(2 04) 3.82 * — — _ _ _ _

1/ 19) — —— 4 69 ** 13.77 ** —_ _
W .32) — —— — 4.41 * _ _
1:1 1(1.06) — — — — — __ _ _
WW. 1‘“ .49) — -— — — 1.95 NS — _—
(3 16) —- — — — 1.58 NS ._ _—
mw) — — — — 15.38 *** _ _—
101 W 0. 32) —— — — — 22.80 *** —— _—
W .92) 3.07 * — — _ _ _ _
9; (103) —- — — — 16.34 *** — _
W (1.38) — —— —— —— 13.45 *** — _—
9 ,W‘W‘W’ .23) — — —— — 16.56 *** — _—
‘W‘. (2. 89) 4.40 * — _. __ _ _ _
‘/‘W 16) — -— 6.47 *** 42. 55 *** —’ _
311:1". 6) _ _ _ _ 344 T — _
"‘41 — — ~ — — 5. 29 * —— _—
9'99 (0 63) —— — 3.47 ** 20.11 *** — —

5331.011) 1.98 NS 2.99 * 57.57 ** __ _

Appendix B. (Ctd)

 

 

 

 

 

 

Root Surface Area 46 0.94 NS 4.15 (0.18) v!
Height 47 0.55 NS 48.46 (1.57) j
Castilla elastica Root Mass 59 0.27 NS 17.78 (0.86)1
(Intolerant / Rare) Stem Mass 59 0.03 NS 34.35 (1.01)!
Leaf Mass 58 5.83 * 79.56 (3.14”
Total Mass 58 3.76 1' 131.54 (4.16!
Root Length 58 7.07 ** 103.41 (4.92)!
Root Surface Area 58 3.07 1' 17.06 (0.85)‘
Height 59 0.20 NS 114.33 (2.593
Coussarea hondensis Root Mass 36 0.05 NS 15.43 (1.55) 11
(Intermediate / Common) Stem Mass 36 3.84 1' 36.37 (1.75) ‘1
Leaf Mass 36 0.31 NS 42.09 (1.64) 11
Total Mass 36 1.37 NS 93.43 (2.15) 11
Root Length 36 0.10 NS 22.84 (2.40) VI
Root Surface Area 36 0.31 NS 4.22 (0.38) vs]
Height (final - initial) 36 0.91 NS 10.52 (1.51”:
Pentaclethra macroloba Root Mass 58 0.28 NS 299.17 (16.87]
(Intermediate / Common) Stem Mass 58 0.72 NS 569.90 (40.48)
Leaf Mass 58 1.59 NS 726.10 (51.47)
Cotyledon Mass* 58 1.41 NS 176.82 (96.54)
Total Mass 58 3.92 1‘ 1760.56 (138.1
Root Length 58 0.34 NS 206.64 (15.27)
Root Surface Area 58 0.33 NS 41.48 (2.72) vs
Height 58 0.02 NS 370.3 (16.76_)j
Prestoea descurrens Root Mass 48 0.06 NS 16.52 (0.87) vs
(Intermediate / Common) Stem Mass 48 7.52 ** 13.51 (0.82) 111
Leaf Mass 48 7.00 * 19.73 (1.56) v:
Cotyledon Mass 47 6.74 * 60.70 (4.25) 111
Total Mass 48 1.89 NS 109.26 (2.83)‘
Root Length 48 0.99 NS 32.76 (1.89) v
Root Surface Area 48 0.54 NS 5.92 (0.32) vs.
Height 48 2.32 NS 53.91 (2.80) g
Quararibea bracteolosa Root Mass 15 0.05 NS 38.90 (5.05) v
(Intermediate / Rare) Stem Mass 15 1.09 NS 47.20 (5.37) v
Leaf Mass 15 3.21 1' 82.08 (14.22)
Total Mass 15 3.04 NS 175.56 (18.02
Height 15 0.07 NS 64.45 (3.84) Vj
Trophis racemosa Root Mass 51 4.05 * 4.09 (0.21) vs
(Intermediate / Rare) Stem Mass 51 0.14 NS 10.29 (0.41) v
Leaf Mass 51 0.12 NS 22.35 (0.94) V
Cotyledon Mass 51 0.15 NS 13.94 (1.38) V
Total Mass 51 0.00 NS 50.68 (2.02) V
Root Length 51 0.64 NS 16.92 (0.78) V
Root Surface Area 51 1.88 NS 2.98 (0.13) vs
Height 51 2.38 NS 76.40 (2.283

 

133

 

 

 

 

 

 

4.1510(019) 1 83 NS 2.50 * 15.35 *** — —
Limo (1.54) —— — ._ _ _ _ __ _
17.11.36 (0.78) 4.52 ** — — 54.11 *** _ .—
34.3:'.=;..12 (1.01) 2.92 * _ _ 76.03 44111 _ _
795111.42 (3.21) — — — — 48.91 *** _‘ _
131.44.43.35 (4.24) —— _ _. _ 6797 444* __ _
111141 121.73 (4.83) 1.74 NS — — 46.47 *** _ _
17011114 (0 83) 1.61 NS — —— 41.14 11411 _ _
114: 15 95 (2.59) 2.01 NS — — 22.13 *** — _
15.431193 (1.47) — — _ _ 20,10 *** _ _
363111.14 (1.67) _ _ _ _ 6537 441* _ _
42.091185 (1.53) — — 2_43 1- 88.02 4014* _ _
93,413.88 (2.01) — — 4 52 ** 228.81 *** — _
22.543180 (2.27) — — — — 12.62 *** _ _
4,22 113 (0.37) — — 1.53 NS 4.57 * _ _
1052.51.19 (1.36) 2.67 T 2.42 T _ _ _ _
M31181 (16.87) —- — 1.59 NS 19.32 44111 _ _
569.9;9-‘L618.55 (40.48) — — __ _ 1995 444:4: _ _
726.14: 817.94 (51.47) — — — — 11.93 *** _ _
17682437461 (110.87) _ _ _ _ _ _ _
1760:4141. 2148.90 (138.90) 1.67 NS 1.53 NS 12.24 4441 _ _
306.511.219.18 (15.27) — — _ 2,73 NS _ _
411314.69 (2. 73) 1 83 NS — — 3.98 1 _ _
3703l"--‘734(1676) —- — — _ 1191 *4“): _ _
165: ”/1482 (0 80) — — — — 3.34 ’r _ _
1. :1 41140 (0. 72) 1.69 NS — — 9.29 ** _ _
197311903 (1 36) 2.52 T — — 1.58 NS _ _
1418435 (3.79) 2.32 T ._ _ 1135 =4 _ _
411-03 90 (2 66) — — — 48.23 *** _ _
34141.9- 16 (1.66) 2.61 T _ _ _ _ _ _
5911 2 (0 29) 1.66 NS _ _ _ _ _ _
53914.45 (2.51) 4.50 ** _ _ __ __ _ _
3899: 4144 (4. 73) _ _ _ _ _ _ _ _
41294 430 (5. 02) — — — —— _ _ _ _
3108 1115 22 (13.66) 2.71 NS _ _ _ _ _ __
115.9 218 55(16 86) — ._ _ _ _ __ _ _
134/.) 415-. )11 (3. 69) 4.96 * _ _ _ _ _ _
4099 4m 20) 1.80 NS 213 7 12.45 ** _ _
10 411 1.08 (0. 39) 4.75 ** — 19.96 *** 5.99 **

WWW... .79 (0 90) 5.34 ** — — 37.97 *** 8.34 ***

4411.919 (1 33) 4.89 ** —— — 6.77 * 6.35 ***

069» 0"“473 (1. 95) 9.56 *** — — 36.4 *** 13.21 ***
16939 4 .77 (0. 74) 4.64 ** 2.12 W _ _ _ _
WWWWWNO 13) — — — — 3.22 NS _ _
49 (2.19) 4.56 ** — — 16.39 *** _ _

 

Appendix B. (Ctd)

 

 

 

 

 

 

Vochysiaferruginea Root Mass 40 0.42 NS 2.92 (0.17) vs. 2.1
(Intermediate / Rare) Stem Mass 40 0.01 NS 7.27 (0.22) vs. 7.1
Leaf Mass 40 0.00 NS 15.78 (0.31) vs. 1!

Total Mass 40 0.19 NS 26.04 (0.50) vs. 21

Root Length 40 2.38 NS 4.00 (0.32) vs. 3.3

Root Surface Area 40 2.06 NS 0.86 (0.05) vs. 0.7

Height (final — initial) 40 1.00 NS 11.15 (.847) vs. 15

Capparis pittieri Root Mass 43 1.94 NS 76.53 (6.23) vs. 8';
(Tolerant / Common) Stem Mass 43 0.57 NS 109.636 (6.81) vs. 1
Leaf Mass 43 0.77 NS 395.57 (20.68) vs.

Total Mass 43 1.01 NS 580.16 (29.26) vs.

Root Length 42 0.28 NS 26.95 (2.32) vs. 21

Root Surface Area 42 1.65 NS 9.37 (0.57) vs. 10.

Height 43 0.18 NS 79.82 (2.75) vs. 81

Euterpe‘ precatoria Root Mass 45 0.83 NS 36.08 (1.65) vs. 31
(Tolerant / Common) Stem Mass 45 0.88 NS 13.25 (0.55) vs. 12
Leaf Mass 45 0.43 NS 80.18 (5.37) vs. 85

Cotyledon Mass 45 0.37 NS 137.49 (10.81) vs.

Total Mass - 45 0.07 NS 266.45 (5.22) vs. 2
Root Length 45 1.77 NS 32.99 ( 1.75) vs. 36
Root Surface Area 45 1.55 NS 6.97 (0.32) vs. 7.51
Height 45 0.03 NS 100.55 (6.02) v1.41
Iriartea deltoidea Root Mass 48 5.56 * 116.94 (6.19) vs. 1
(Tolerant / Common) Stern Mass 48 2.59 NS 96.10 (4.55) vs. 10
Cotyledon Mass 48 1.76 NS 2821.98 (32.34) vs
Total Mass 48 0.49 NS 3034.63 (28.83) 11
Root Length 48 2.49 NS 42.53 (2.90) vs. 49
Root Surface Area 48 3.69 NS 14.56 (0.78) vs. 16
Height 48 1.05 NS 33.04 (1.33) vs};
Welfia regia Root Mass 56 0.94 NS 187.10 (6.94) vs. 1
(Tolerant / Common) Stem Mass 56 0.22 NS 47.23 (2.40) vs. 48
Leaf Mass 56 0.19 NS 296.00 (15.97) vs.
Cotyledon Mass 56 1.56 NS 25.92 (3.10) vs. 20

Total Mass 56 1.04 NS 555.79 (21.41) vs.
Root Length 56 0.97 NS 211.78 (11.03) vs.
Root Surface Area 56 1.23 NS 37.31 (1.79) vs. 34
Height 56 1.16 NS 146.89 (5.93) 113;]

Virola koschnyi Root Mass 28 1.83 NS 176.93 (13.50) vs.
(Tolerant / Intermediate) Stem Mass 28 1.54 NS 483.66 (24.71) vs.
Leaf Mass 28 3.68 1' 263.11 (16.57) vs.
Cotyledon Mass 28 0.95 NS 7.90 (2.09) vs. 4.9'

Total Mass 28 0.12 NS 933.05 (39.19) vs.
Height 28 1.71 NS 216.67 (6.52) VS’. 2

 

134

x

 

 

 

 

 

 

20.13515) 1.89 NS —— 18.70 *** —— .—
71,0.31='-2) — — 2.19 13.83 *** _ _
7816.3: 1.28) 2.15 NS 108.75 *** _ _—
041053146) 2.55 1 — 105.56 191* __ ._
010.3 1:12) —— — — —— _ _ _
610‘ 5) —— — — 2.06 NS —- —
139847) _ _ _ _ _ __ __
5311:1109) 8.27 *** 2.80 54.88 *** 11.08 ***
9.661.653.3(681) 8.06 W — 57.02 *** 7.33 **
S.57126*7(20.66) 10.80 *** — 67.07 *** 12.10 ***
0.1613240923) 11.43 *** — 80.29 W 12.68 ***
9505:132) — — — 14.36 191* _ _—
17105157) 1.54 NS _. 18.88 *** _. _
821275169) — — — 27.53 m — _—
7557? .76) ._ __ _ _ _ _ __
251151—581 1.80 NS —- —- -- -- —
.1815574) —- - - -— -- — --
75 55 5511 (11.56) — —— — 5.36 * — _
655.515. 58) — , .— — 34.57 *** —— _—
9911137) — — — — — — —
971033 3 ‘— — — — — — _—
505555 (9.37) 1.69 NS — 3.43 1 _. _—
6955/1702) — —— — 11.06 691* — _—
555 5555. 1.27) 3.02 * — 8.59 ** — —
.5595 5550907. 51) 3.81 * —. 100.33 *** — —
535555556003 .45) 3.10 * —— 152.50 *** — _—
5 55 555, 2.42 1 —— 4.26 * — _—
5565.91) 264 ’r — 6.60 * —— —
)01J 53) — — “ — — —‘ —
/57.(7. 47) —— —- — 43.19 *** —— _—
’3 3333.518) ' —- — —— 16.22 *** —- _-
73" 333‘ .1(17.18) — —— — 13.84 *** — —
1.600133 53) _ __ __ __ __ __ _
5933335123 .03) —— —— —— 38.40 *** — —
553333 51(11. 86) — —— — 21.82 *** — _—
1-331 53.13) — — — 3 24.97 *** — —
7-333 537) —- — — 12.11 ** — —
km 98) — —— — 6.38 * —— _—
76 3‘33 '(26. 09) 2.91 1 —— 2.74 NS 3.55 *
8133‘ 55m. .15) — — —— 5.16 * — —
113-313 — — —. 3.17 ’r -— —
9033314137) 3.46 * — 8.62 * 4.02 *
33333 5.74) — — -— —— —— — -—

166 '3

 

Appendix B. (Ctd)

 

 

 

Dussia macrophyllata Root Mass 19 0.26 NS 209.45 923.03)
(Tolerant / Rare) Stem Mass 19 4.86 * 306.00 (27.59) '1‘
Leaf Mass 19 0.00 NS 269.38 (33.69)‘
Total Mass 19 1.03 NS 784.17 (7194)
Root Length 19 0.71 NS 70.28 (7.68) vs.
Root Surface Area 19 0.01 NS 21.34 (1.80) vs.
Height 19 0.30 NS 263.04 (17.07)‘
Miconia affinis Root Mass 40 1.50 NS 0.02 (0.01) vs.(
(Tolerant / Rare) Stem Mass 44 1.24 NS 0.06 (0.01) vs.(
Leaf Mass 46 0.04 NS 0.16 (0.02) vs.(
Total Mass 38 0.51 NS 0.26 (0.03) vs.(
Root Length 36 0.47 NS 0.64 (0.09) vs. I
Root Surface Area 36 0.94 NS 0.07 (0.01) vs. (
Height 46 0.66 NS 7.64 (0.42) vs. I
Stryphnodendron Root Mass 35 0.44 NS 4.05 (0.24) vs. 3
microstachyum Stem Mass 35 0.15 NS 12.32 (0.72) vs:
(Tolerant / Rare) Leaf Mass 35 1.86 NS 20.02 (1.30) vs-
Total Mass 35 1.57 NS 36.37 (1.69) vsl
Root Length 34 0.11 NS 11.86 (0.80) vs;
Root Surface Area 34 0.29 NS 2.00 (0.16) vs. 1
Height 35 0.01 NS 110.72 (6°52)L‘

 

Notes: Tests were performed using the Type III sum of squares from SPSS version 14. N = to
contain that term due to a non-significant (P > 0.10) effect on the dependent variable. Root,s
and root surface area in cmz. Means have been rounded up. Significance is shown as: NS, not
inoculum is significant (P g 0.10) the microbial and sterile treatment means are in bold.

13S

 

20945919150 (24.43) —— —— —— — 16.55 *** — —

 

 

 

30611015171137 (29.32) 3.39 1' _ _ 28.26 am: _ _.
269381337080 (35.74) — _ _ _ 545 at _ _
784.17177IJ4 (76.33) — — _ _ 1393 an“: _ _
70.281789 (8.15) — — — — 4.42 1' _ ' _
213411314091) — _ _ _ 5.59 =1: __ _
2630414859 (18.14) 5.64 * — — 6.59 * _ __
0.02 122.1 (0.00) 1.80 NS — — — _ __ _
006111121001) 4.14 * 2.49 T — _ _ _
0161011002) — _ 2_47 * _ _ _ _
0261531003) 3.29 * 2.55 T — __ _ _
06-) 113,140.08) — — — — — _ _ _
0.0711111'I0-01) — — — — — __ _ _
7.6411? 41(0-43) — — — — — — — —
11131754030) — — 5.26 ** _ _ _ _
12.331138 (0.89) — — — — 2.14 NS — _—
301131130 (159) — — 2.92 * 4.47 * — —
36.37114) (2.07) — — 2.98 * 4.84 * _ _—
11.86133 7 (0-94) — — 3.88 ** _ _ _ _
2.0010110”) — — 1.63 NS 3.84 1- — —
110.12.11.42 (7.98) — — — —— 8.17 .4. _. _.
313100 11 \umber of seedlings in the model. “—“denotes that the ANCOVA model does not

{311311133 leaf, cotyledon and total mass measurements are in mg; root length and height in cm

01m gsiificant; 'I' P S 0-10; * P S 0.05; ** P g 0.01; *** P g 0.001. When the main effect of

11? in h

Appendix C. Analysis of covariance results for the effects of inoculum treatment on organ mass, 1;
microbial extract; adult = location where soil was collected for use in extraction; bench =locati01ri

 

 

 

 

 

 

 

 

 

plant mass. .
~E
Spec1es Treatment (1) :.
(Shade Tolerance I De endent Variable
Adult Abundance p
Classification) N F P Means (Std- error)
Microbial vs. Sterile A
Luehea seemanm'i Root Mass 22 0.07 NS 0.09 (0.04) vs. 0.10 (0.03) .4
(Intolerant/ Stem Mass 22 0.82 NS 0.34 (0.02) vs. 0.32 (0.02) :7
Common) Leaf Mass 22 0. 94 NS 0.48 (0.02) vs. 0.51 (0.02) 3
Root Length 22 1. 60 NS 0.34 (0.12) vs. 0.53 (0.09) A
Root surface Area 22 3.06 1' 0.04 (0.02) vs. 0.08 (0.01) -
Neea psychotroides Root Mass 48 0. 55 NS 2. 55 (0.16) vs. 2.39 (0.15) f
(Intolerant/ Stem Mass 48 0. 00 NS 4. 20 (0.14) vs. 4.19 (0.13)
Common) Leaf Mass 48 0.41 NS 13.80 (0. 20) vs. 14.00 (0.19)
Root Length 48 5.66 * 13.35 (0. 70) vs. 11.03 (0. 66)
Root Surface Area 48 1. 64 NS 2.26 (0.12) vs. 2.06 (0.11) ,
Apeiba membranacea Root Mass 20 3. 32 1 0.79 (0.06) vs. 0.66 ( 0.05) 7
(Intolerant/ Stem Mass 20 0.28 NS 1.61 (0.10) vs. 1.54 (0.08)
Intermediate) Leaf Mass 20 1.76 NS 2.68 (0.12) vs. 2.89 (0.09)
Root Length 20 0.02 NS 3.62 (0.46) vs. 3.04 (0.37)
Root Surface Area 20 0.12 NS 0.58 (0.07) vs. 0.55 (0.05) .
Colubrina spinosa Root Mass 47 2.14 NS 3. 02 (0.13) vs. 2. 74 (o. 14) 3
(Intolerant/ Stem Mass 47 0.05 NS 5.93 (0.17) vs. 5.98 (0.18)
Intermediate) Leaf Mass 47 0.70 NS 14.03 (0. 23) vs. 14. 32 (0. 25)
Guatteria Root Mass 36 0.21 NS 5. 56 (0.24) vs. 5.72 (0. 25) .j
diospyroides Stem Mass 36 2.03 NS 10.40 (0.39) vs. 9.60 (0.41)
(Intolerant/ Leaf Mass 36 1.29 NS 5.73 (0.48) vs. 6.52 (0.50)
Intermediate) Root Length 36 0.00 NS 21.44 (1.24) vs. 21.32 (1.28) 1
Root Surface Area 36 0.09 NS 3.98 (0.19) vs. 3.90 (0.20L
Psychotria Root Mass 55 0.05 NS 3.79 (0.16) vs. 3.74 (0.15)
panamensis Stem Mass 55 0.22 NS 5.69 (0.20) vs. 5.82 (0.19)
(Intolerant/ Leaf Mass 55 0.11 NS 9.68 (0.21) vs. 9.58 (0.21) 1
Intermediate) Root Length 55 0.16 NS 27.29 (1.09) vs. 26.69 (1.07) .
Root Surface Area 55 0. 00 NS 4.46 (0.21) vs. 4.44 (0.20);
Castilla elastica Root Mass 58 3. 73 1' 18.71 (0.54) vs. 17.18 (0.58) 3‘
(Intolerant / Rare) Stem Mass 58 6. 00 * 35.59 (0.67) vs. 33.25 (0.67)
Leaf Mass 58 7. 40 ** 83.17 (0.80) vs. 86.28 (0.81) ,
Root Length 57 2.34 NS 108.40 (4.64) vs. 118.45 (4.51
Root Surface Area 57 0.55 NS

17.85 (0.79) vs. 18.68 (0.181

 

136

eight and root morphology allocation. Inoculum = microbial vs. sterilized
eedlings were randomly assigned to during the experiment; covariate = total

 

If

 

 

 

 

 

 

 

 

fiSource of Variation (df)
Adult (3) Treatment Bench (5) Covariate ( 1) Adult x
X Adlllt (3) Covariate (1)
F P F P F P F P F P
3.04 1' — — __ _ 76.63 *** _ _
4.57 * — — — — 18.74 ** 14.30 ***
5.23 * —— — — — 63.37 *** 3.95 *
1.84 NS — — _ _ 44.96 us: _ _
5““ —— — —— —- — 119.56 *** — —
— — — — — — 58.09 *** — —
— -—— —— — — — 89.58 *** — —
— — — — — — 1220.4 *** — —
— — — — 2.56 * 71.30 *** — —
.184 ** —— — — — 62.91 *** — —
1 — — —— — —— — 15.48 *** — —
— — — — — — 19.16 *** — —
— — — — — — 75.93 *** — —
— — — — — — 6.12 * — —
;— — — — ~—-— —— 2.72 NS —— ——
7.05 ** -— — 3.15 * 52.67 *** — —
—- — — — — 107.06 *** — —
2.90 * —- — —— — 763.09 *** — —
1.42 NS 1.57 NS — — 22.57 *** — —
— — — — — — 3.91 1' — —
1.00 NS 1.97 NS — — 68.92 *** — —
1.64 NS — __ _ _ 23.07 mm: _ _
1.44 NS — —— — —— 32.86 *** _ _
2.27 1' 1.86 NS — — 41.33 M»: _ _
- — -— — — — 18.51 *** — —
0.90 NS 1.66 NS — -— 260.73 *** — —
1.66 NS — — — — 37.28 *** _ _
1.39 NS 1.85 NS — — 24.14 ** —- —
4.33 ** 1.99 NS 1.56 NS 107.59 *** — —
4.63 ** — — 2.05 1' 186.30 *** — —
0.32 NS 1.49 NS 3.25 * 834.78 ** * — —
2.05 NS — — — — 56.30 Mus _ _
—— — —- —— 2.48 * 45.66 *** — —

 

Appendix C. (Ctd)

 

 

 

 

 

 

 

 

Coussarea hondensis Root Mass 36 0.37 NS 15.79 (1.39) vs. 14.61 (1.31
(Intermediate/ Stem Mass 36 2.22 NS 38.19 (2.42) vs. 41.71 (2.12;
Common) Leaf Mass 36 4.13 'I' 43.45 (1.19) vs. 38.14 (1.45
Root Length 36 0.01 NS 21.14 (2.30) vs. 18.72 (2.93:
Root Surface Area 35 0.95 NS 4.35 (0.36) vs. 3.86 (0.35L
Pentaclethra Root Mass 58 0.53 NS 313.29 (15.36) vs. 297.28 (1
macroloba ' Stem Mass 58 0.12 NS 602.67 (34.76) vs. 585.78 (3
(Intermediate / Leaf Mass 58 0.05 NS 764.80 (43.36) vs. 779.23 (4
Common) Cotyledon Mass 58 0.03 NS 266.49 (75.55) vs. 284.95 (7
Root Length 58 0.06 NS 215.22 (13.71) vs. 210.60 (1
Root Surface Area 58 0.23 NS 43.49 (2.43) vs. 41.82 (2.43;
Prestoea descurrens Root Mass 45 0.42 NS 17.47 (0.74) vs. 16.83 (0.66]
(Intermediate / Stern Mass 45 5.97 * 14.40 (0.77)vs. 16.92 (0.69]
Common) Leaf Mass 45 3.30 1' 22.62 (1.29) vs. 25.77 (1.161
Cotyledon Mass 45 3.10 1’ 50.97 (2.13) vs. 45.93 (1.91]
Root Length 45 0.08 NS 36.42 (1.56) vs. 35.83 (1.39)
Root Surface Area 45 0.38 NS 6.49 (0.26) vs. 6.28 (0.23L
Quararibea Root Mass 14 0.01 NS 37.95 (2.54) vs. 37.64 (2.17)
bracteolosa Stem Mass 14 0.00 NS 51.31 (4.50) vs. 51.51 (3.86)
(Intermediate / Rare) Leaf Mass 14 0.00 NS 115.41 (3.03) vs. 115.52 (21
Trophis racemosa Root Mass 51 2.69 NS 4.13 (0.21) vs. 4.59 (0.20)
(Intermediate / Rare) Stem Mass 51 0.46 NS 10.09 (0.31) vs. 9.79 (0.31)
Leaf Mass 51 0.01 NS 22.11 (0.65) vs. 22.03 (0.63)
Cotyledon Mass 51 0.15 NS 13.00 (1.44) vs. 12.20 (1.41)
Root Length ' 51 0.86 NS 16.75 (0.76) vs. 17.73 (0.74)
Root Surface Area 51 1.84 NS 3.01 (0.12) vs. 3.25 (0.12L
Vochysiaferruginea Root Mass 40 0.27 NS 2.88 (0.16) vs. 2.78 (0.14)
(Intermediate / Rare) Stem Mass 40 0.06 NS 7.23 (0.17) vs. 7.29 (0.17)
Leaf Mass 40 0.25 NS 15.67 (0.19) vs. 15.79 (0.17)
Root Length 40 0.66 NS 3.91 (0.31) vs. 3.58 (0.29)
Root Surface Area 40 1.33 NS 0.85 (0.05) vs. 0.78 (0.04L
Capparis pittierin Root Mass 43 2.33 NS 59.49 (2.09) vs. 63.81 (2.05)
(Tolerant / Common) Stem Mass 43 0.05 NS 110.96 (3.02) vs. 111.88 (2.9
Leaf Mass 43 0.19 NS 376.64 (3.53) vs. 374.56 (3.3J
Root Length 42 0.05 NS 27.44 (2.37) vs. 28.20 (2.37)
Root Surface Area 42 0.46 NS 9.52 (0.48) vs. 9.98 (0.48Lq
Euterpe precatoria Root Mass 45 0.83 NS 36.08 (1.65) vs. 38.28 (1.76)
(Tolerant / Common) Stem Mass 45 0.68 NS 13.09 (0.54) vs. 12.45 (0.56)
Leaf Mass 45 0.95 NS 78.92 (4.82) vs. 85.86 (5.13)
Cotyledon Mass 45 0.33 NS 136.12 (6.05) vs. 131.11 (6.34
Root Length 45 1.77 NS 32.99 (1.75) vs. 36.39 (1.87)
Root Surface Area 45 1.55 NS 6.97 (0.32) vs. 7.56 (0:15.)...

 

137

 

 

 

 

 

 

 

 

— — — — — — 33.46 *** — ——
1.87 NS — — — — 74,42 *** _ _
1.02 NS 2.39 NS — — 267.66 *** _ _
0.85 NS 1.57 NS —— —— 15.34 . ** __ _
— — — — —- — 11.57 ** —— _—
— — — —— — — 39.20 *** —— —
— —— —— — — —— 49.28 *** -— —
— — —— — —- — 4.71 *** — —
—— —- —— -— — —— 53.88 *** — —
— —_ — — — — 18.40 *** — —
__.— -— — — — —— 26.78 *** — _—
— —— —— — — — 5.80 * — —
— — — —- — — 14.34 *** — —
——- -— — —— —— — 2.81 NS — _
— — — — — — 95.25 *** _ _
— — — — — — 5.19 * _ _
— —— — —- —— — 3.81 1 _ _.
— —- — — — — 28.36 *** _ _
— — — — —— —- 8.96 * _ _
— — — — — — 174.03 *** — _—
— -— -— — 250 * 14.57 *** — —
— — — — — — 60.57 *** — —
2-62 T — — — — 120.68 *** _ _
— — — —— — —- 10.19 ** —— _—
3.81 * 1.58 NS — — 7.46 ** _ _
0.45 NS 1.51 NS — — 11.00 ** _ _—
1.50 NS — —— — — 29.14 :44 __ _
— — — -— 189 NS 43.87 *** — _—
1.78 NS — — —— —- 360.73 =64 _ _
3.39 * — — — — 5.65 * 3.91 *
3.09 1 —— — 1.67 NS 13.63 * 3.54 ’r
4.56 * — — 8.54 *** 225.11 *** 5.94 **
2.43 ’r — —— 114.50 *** _ _ _ _
1.87 NS —— —— 1.53 Ns 1095.0 *** — _—
—- — — — — — 12.29 *** — ——
— — — — — — 43.08 *** — _—
1.57 NS 1.91 NS —- —— 1.75 NS — —
0.47 NS 1.64 NS 1.48 NS 12.34 ** _ _
0.46 NS 1.50 NS — — 109.99 *** _ _—

 

Appendix C. (Ctd)

 

 

 

 

 

 

' Iriartea deltoidea Root Mass 48 5.84 * 115.90 (6.70) vs. 140.42 (7
(Tolerant / Common) Stem Mass 48 3.15 1' 95.60 (4.69) vs. 108.36 (5.4
Cotyledon Mass 48 5.95 * (21821331 (10°40) vs. 2778'0
Root Length 48 3.13 1' 42.02 (2.99) vs. 50.13 (3.47
Root Surface Area 48 4.21 * 14.45 (0.83) vs. 17.04 (0.5!
Welfia regia Root Mass 56 0.17 NS 183.64 (4.50) vs. 180.89 (4
(Tolerant / Common) Stem Mass 56 0.37 NS 46.86 (2.18) vs. 48.86 (2.37
Leaf Mass 56 1.04 NS 283.89 (6.06). vs. 293.39 (6
Cotyledon Mass 56 1.56 NS 25.92 (3.10) vs. 20.24 (3.33
Root Length 56 0.37 NS 207.22 (9.40) vs. 199.22 (9.
Root Surface Area 56 1.49 NS 36.77 (1.31) vs. 34.35 (1.4;
Virola koschnyi Root Mass 29 1.06 NS 181.62 (11.13) vs. 198.18 (.
(Tolerant / Stem Mass 29 4.45 * 491.07 (14.77) vs. 445.97 (2
Intermediate) Leaf Mass 29 3.55 1' 269.61 (11.85) vs. 301.91 (I
Cotyledon Mass 29 1.63 NS 8.30 (2.03) vs. 4.55 (2.10L
Dussia macrophyllata Root Mass 19 0.00 NS 201.19 (21.53) vs. 200.68 (2
(Tolerant / Rare) Stem Mass 19 1.83 NS 290.61 (28.72) vs. 232.06 (3
Leaf Mass 19 0.01 NS 271.73 (24.49) vs. 268.19 (2
Cotyledon Mass 19 0.58 NS 220.43 (59.01) vs. 288.08 (1
Root Length 19 0.53 NS 69.09 (7.06) vs. 61.37 (7.50
Root Surface Area 19 0.06 NS 20.90 (1.77) vs. 21.53 (1.87
Miconia afi‘inis Root Mass 38 1.96 NS 0.02 (0.00) vs. 0.03 (0.00)
(Tolerant / Rare) Stem Mass 38 0.40 NS 0.06 (0.00) vs. 0.06 (0.00)
Leaf Mass ‘ 38 0.02 NS 0.16 (0.00) vs. 0.16 (0.01)
Root Length 30 0.66 NS 0.66 (0.07) vs. 0.73 (0.06)
Root Surface Area 30 0.94 NS 0.07 (0.01) vs. 0.08 (0.00fl
Stryphnodendron Root Mass 35 0.49 NS 4.22 (0.37) vs. 3.86 (0.32)
microstachyum Stem Mass 35 0.32 NS 11.52 (0.81) vs. 12.06 (0.99)
(Tolerant / Rare) Leaf Mass 35 0.16 NS 18.74 (0.56) vs. 18.38 (0.69)
Root Length 34 0.11 NS 11.86 (0.80) vs. 12.27 (0941
Root Surface Area 34 0.00 NS 1.97 (0.21) vs. 1.96 (0.23)‘

 

Notes: Tests were performed using the Type III sum of squares from SPSS version 14. N = totalt
contain that term due to a non-significant (P > 0.10) effect on the dependent variable. Root, stem
cmz. Means have been rounded up. Significance is shown as: NS, not significant; T P g 0.10; * P

g 0.10) the microbial and sterile treatment means are in bold.

138

— - — — 2.73 NS —- _

4:.
8
* I
l

l

I

I

 

 

 

 

 

5.26 * — _

2.66 1‘ — — — — 1327.6 *** —-— —
2.61 '1' — —— — —— _. _. __ __
3-11 * —- — — — 1.65 NS __ _
—— — —— — — — 177.62 *** —— —
1.98 NS —— -— —- —- 21.19 *** — —
1.21 NS 1.63 NS —— — 434.77 *** __ _
1.63 NS — — —— _ 59.27 *** _ _

- 2.10 NS —— — 2.31 1' 66.40 *** V—— _—
1 — — — — — —— 22.12 ** —— —
15' — — — -- — — 76.37 ** — _—
' — — — — —- -— 5.13 1' — —
—‘ — -— — — -- 5.38 1’ _. _.
— —— — — — — 19.11 *** — —
3.02 NS 1.69 NS —- — 21.31 *** __ _
__ _ _ _ _ _ 22.26 M8: _ _
_ _ __ __ _ _ 23.06 us: _ _
0.80 NS 1.75 NS —- — 5.90 * _ _
—— —-— — —— — —- 5.21 * — ——
— — — — — —— 23.90 *** — —
-— —— — — —— — 96.36 *** ~— ——
1.63 NS — —— — — 501.61 #44: _ _
4.00 *' — — —— — 53.47 *** — —
2.47 ‘I' — — __ _ 37.77 *** __ __
0.21 NS 1.80 NS 3.58 * -— _ __ _
1.48 NS —— -— -— — 11.57 *4: __ _
— — —- — — — 202.68 *** — —
—— -— -— — 3.88 ** — — — —
4.15 * — — 2.84 * — — — —

 

er of seedlings in the model. “——“ denotes that the ANCOVA model does not
'and cotyledon measurements are in mg; root length in cm and root surface area in
)5; ** P _<_ 0.01; *** P g 0.001. When the main effect of inoculum is significant (P

Appendix D. Analysis of covariance results for the effects of inoculum treatment on root morphol
bole of 4 different adults per species) where soil was collected for use in extraction; bench = local

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mass.
S ecies a
(Sphade Tolerance / Adult Dependent Variable Treatment (1)
Abundance Classification) N F P Mean (Std. error)
Microbial vs. Sterij
Neea psychotroides Root Length 48 1.12 NS 12.87 (0.80) vs. 11.'.
(Intolerant / Common) Root Surface Area 48 0.01 NS 2.14 (0.11) vs. 2.13
Apeiba membranacea Root Length 20 0.01 NS 3.44 (0.42) vs. 3.50
(Intolerant/ Intermediate) Root Surface Area 20 0.33 NS 0.57 (0.04) vs. 0.61)
Guatteria diospyroides Root Length 36 0.07 NS 21.29 (0.94) vs. 20.9
(Intolerant / Intermediate) Root Surface Area 36 0.40 NS 4.00 (0.14) vs. 3Q
Psychotria panamensis Root Length 57 0.28 NS 27.12 (1.01) vs. 26.3
(Intolerant / Intermediate) Root Surface Area 57 0.02 NS 4.38 (0.18) vs. 4.fl
Castilla elastica Root Length 58 7.65 ** 105.73 (3.63) vs. 11
(Intolerant / Rare) Root Surface Area 58 4.10 * 17.42 (0.53) vs. 18;!
Coussarea hondensis Root Length 34 0.49 NS 22.94 (2.33) vs. 20.6
(Intermediate / Common) Root Surface Area 34 0.66 NS 3.65 (0.50) vs. 3.26)
Pentaclethra macroloba Root Length 58 0.24 NS 209.83 (11.87) vs. 2
(Intermediate / Common) Root Surface Area 58 0.08 NS 42.32 (1.97) vs. 431
Prestoea descurrens Root Length 48 0.63 NS 33.93 (1.24) vs. 35.2
(Intermediate / Common) Root Surface Area 48 0.44 NS 6.07 (0.20) vs. 635
Trophis racemosa Root Ifingth 51 0.21 NS 16.93 (0.75) vs. 17.4
(Intermediate / Rare) Root Surface Area 51 0.28 NS 3.06 (0.108) vs. fl
Vochysiaferruginea Root Length 40 0.94 NS 4.00 (0.37) vs. 3.55!
(Intermediate / Rare Root Surface Area 40 0.64 NS 0.83 (0.03) vs. 0.721
Capparis pittierin Root Length 42 0.09 NS 27.27 (2.53) vs. 28.1
(Tolerant / Common) Root Surface Area 42 0.53 NS 9.51 (0.47) vs. 9.294i
Euterpe precatoria Root Length 45 3.41 1' 33.09 (1.13) vs. 36.1
(Tolerant / Common) Root Surface Area 45 2.47 NS 7.03 (0.18) vs. 7 45}
Iriartea deltoidea Root Length 48 0.69 NS 46.44 (1.61) vs. 44.3
(Tolerant / Common) Root Surface Area 48 1.99 NS 15.76 (0.23) vs. L5;
Welfia regia Root Length 56 0.35 NS 208.21 (9.53) vs. 19
(Tolerant / Common) Root Surface Area 56 1.10 NS ' 36.78 (1.41) vs. 3i:
Dussia macrophyllata Root Length 19 0.70 NS 69.07 (5.98) vs. 61.1
(Tolerant / Rare) Root Surface Area 19 0.02 NS 21.08 (1.22) vsgL.
Miconia afiinis Root Length 32 0.30 NS 0.73 (0.07) vs. 0.681
(Tolerant / Rare) Root Surface Area 32 0.02 NS 0.08 (0.01) vs. M

 

139

inoculum = microbial vs. sterilized microbial extract; adult = location (near the
=Itt seedlings were randomly assigned to during the experiment; covariate = root

L
r

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Source of Variation (df)
Adult (3) Treatment x Bench (5) Covariate. (1)
Adult (3)
F P F P F P F p
T75) 3.78 * — _ _ _ 30.37 am:
L 2.38 1' — — — _ 68.44 ***
1 2.53 ‘I' — —— __.— __ 9.69 *5:
) 7.53 ** — _ _ _ 25.94 ***
'99) — — — — — — 52.47 ***
) 0.06 NS 1.80 NS —— — 74.34 m
'00) — — — — — —- 64.80 ***
l — — — —— — —- 73.99 ***
1350) 1.78 NS —- — — — 127.64 444
51) 3.73 * — —- 2.24 1 143.49 ***
'19) — -.—- — — — — 8.84 **
) 2.58 '1' — __ __ _ 1941 ***
201.94) 0.17 NS 1.77 NS — — 36.73 *4...
98) 5.36 ** 1.93 NS 1.77 NS 52.29 ***
'10) 3-32 * — — — — 55.48 ***
l 2.74 1 1.66 NS — — 71,33 ***
72) 3.97 * 1.66 NS — — 9.14 *4
92 — —- —— — — — 24.49 ***
) 0.58 NS 1.63 NS — — 17.59 ***
P — — — — — — 53.67 ***
53) — — — — — — 6.04 *
L — — — — — — 47.03 ***
31) — — _ __ 4.64 ** 48.45 ***
L — — — — 4.42 ** 80.71 ***
38) 3.12 * — _ _ _ 11733 sum:
27) 10.33 *** — — __ _ 573.32 #4:»:
1526) — — — — — — 47.90 ***
55) 1.57 NS — —— 1.43 NS 51.54 ***
i3 — — —- — — — 14.78 ***
19) — — — — — — 27.24 ***
7 1.50 NS — — — — 37.15 ***
1.52 NS — — — — 37.72 ***

 

 

I. -3 illr" .t’.
III .11» ,E4illllll’LII-1L. '1... II: 19.1.”! i. 1.111..

Appendix D. (Ctd)

 

Stryphnodendron Root Length 34 0.71 NS 11.31 (0.72) vs. 12
microstachyum
(Tolerant / Rare) Root Surface Area 34 1.34 NS 1.80 (0.14) vs. 1.91

 

Notes: Tests were performed using the Type III sum of squares from SPSS version 14. N = total
contain that term due to a non-significant (P > 0.10) effect on the dependent variable. Root lengtl
Significance is shown as: NS, not significant; 1' P _<_ 0.10; * P g 0.05; ** P 5 0.01; *** P _<_ 0.001
treatment means are in bold.

140

 

(0.86) — — — — — — 23.41 ***

12) 8.81 *** 1.68 NS — — 34.04
:ber of seedlings in the model. “—“ denotes that the ANCOVA model does not

measured in cm and root surface area in cmz. Means have been rounded up.
hen the main effect of inoculum is significant (P g 0.10) the microbial and sterile

***

Appendix E. Meta-analysis results for the effect (size and magnitude) of inoculum treatment cg anmass stel
categories in each of 5 different local species characteristics. Data for meta-analysis derivedllNCOVA at

 

 

 

 

 

 

 

 

 

Root Mass , Stem Mass , Leaf Mass ‘; R0
Species Nbrc Effect d,e P Effect d,e P Effect d,e . éEffect
Categoriesa Size value Size value Size t: 33121:
(2 Std 5 (2 Std 5 (2 Std (2 Std
error) 5 error) 5 error) # Senor)
-O.15 E -0.08 E -0.21 50.13
overall 21 (0.14) 5 (0.14) 5 (0.14) 50.15)
Shade 7 -014 5 -0.08 5 034 £0.27
Intolerant (0.23) g (0.23) g (0.23) .1026)
Shade -0.12 E -0.29 E -0.26 , {.004
Intermediate 6 (0.25) 0'09 NS (0.25) 4'71 T (0.26) 3'38 (0.26)
Shade 8 -0.17 0.09 -0.04 $0.13
Tolerant (0.22) E (0.22) E (0.24) .7024)
_ -0.12 § 0.15 § -0.15 3-1T
Adult Rare 7 (0.25) g (0.24) g (0.24) ' (026)
Adult— -0.20 5 -0.04 ,, 5 -0.28 4). 15
Intermediate 5 (0.29) 0'21 NS (0.29) 6'00 (0.29) 0'50 i (0 (37)
Adult— 9 -0.14 E -0.24 E -0.22 011
Common (0.20) (0.20) i (0.21) #‘ 3(020)
Seedling — 6 -022 g 0.12 g -020 “W
Rare (0.27) 5 (0.26) 5 (0.26) gmg)
Seedling — 026 g -0.07 § -0.34 a .023
Intermediate 8 (0.23) 3'07 NS (0.23) 3'85 NS (0.25) 1'64 ‘ 5(024)
Seedling— 7 0.00 5 -0.22 5 -0.12 1010
Common (0.22) (0.22) (0.22) # £10.24)
Basal Area —— 7 -0.25 5 -0.11 5 -023 \W
Rare (0.23) § (0.23) § (0.23)-1026)
Basal Area— -0.05 E —0.05 E -0.27 :0
I i 0
Intermediate 8 (0.22) 1'50 NS 5 (0.23) 0'17 NS E (0.23) 0'90 (0 2:
Basal Area— 6 -0.15 -0.07 -0.10 ! 350'10)
Common (0.25) E (0.25) 5 (0.28) #‘ 3(0'77
Seed Size — 4 -0.07 5 0.02 5 -0.28 W
Small (0.30) (0.30) (0.30) :(0. 31
Seed Size — -0.15 i -0.16 i -0.18
. : . : , r 0 9
Medium 8 (0.21) O 37 NS 5 (0.21) 105 NS 1 (021) 0 27 ' 2mg:
Seed Size — 9 -0. 19 g -0. 04 5 -0. 21 g 01“)
Large (0. 22) : (0. 22) : (0. 24) ' ( ‘ 2

 

Notes: aLocalized species Characteristics for all 21 species (Table 1). Shade tolerance classif%
seedling mortality at 1% full- -sun and with zero conspecific seedling density. Adult abundan
Seedling abundance classification (rare <20, intermediate 20- 200, common >200 standing dd ,5 e
0. 06— 0. 25, common > 0. 25 total adult mass (m 2/ha)). Seed size classification (small <0. 03. in] 0

category. cNumber of species is reduced for leaf mass since lriratea seedlings had no leased gt ‘

used 1n the analysis. dA statistics that measures the homogeneity of effect sizes between edicts 520‘" 16
0.10;*P50.05;**P50.;01***PSOHOOI -§2

141

iii-organ mass, stem height and root morphology for 21 tropical tree species grouped into 3
:lm ANCOVA analysis using initial seed mass as a covariate (Appendix B).

 

 

 

 

 

 

 

, Root Length 5 Root Surface Area Height
5'- iEffect QBd e P EEffect QBd’ ac P gEffect QBd’ ,6 P
lue S Size value S Size value 5 Size value
5(2 Std 5(2 Std 5(2 Std '
5 error) 5 error) 5 error)
5 -0.13 g —0.17 g -015
g 5 (0.15) 5 (0.14) 5 (0.14)
5 -0.21 ‘5 -0.28 7018
5 (0.26) 5(0.26) g (0.23)
5004 5006 50.19
6 5 (0.26) 0.79 NS 5 (0.26) 1.31 NS 5 (0.25) 0.51 NS
5 -013 g -0.17 g -0.08
5(0.24) 5(0.24) 5 (0.22)
’ §-0.15 §-0.14 §-0.18
5 (0.26) 5(0.26) 5 (0.24)
: -0.15 : -O.28 : —0.28
§ (0.37) 0.08 NS 5 (0.37) 0.41 NS 5 (0.29) 1.65 NS
5 -0.11 5 016 5 -0.06
5(0.20) i(0.20) g (0.20)
’ 5032 5-028 5014
5(0.28) 5(0.28) 5 (0.26)
g -023 * g -0.30 g -0.201
E (0.24) 6.15 5 (0.24) 4.78 ’r E (0.23) 0.60 NS
§ 0.10 50.04 5 -009
5(0.24) 5(0.24) i (0.22)
’ 5021 50.30 5030
§(0.26) §(0.26) § (0.23)
E —0.08 a -0.14 E -0.09
5(0.23) 0.62 NS 5 (0.23) 1.56 NS 5 (0.22) 2.86 NS
g -0.10 g -0.08 g -0.03
5 (0.27) 5 (0.27) 5 (0.25)
/ 50.10 7004 50.17
g (0.31) 5(031) 5 (0.30)
5 -025 5 -022 5 -022
“022) 3.22 NS 5 (0 22) 0.94 NS 5 (0 21) 1.27 NS
5.0. 12 5-0. 19 5-0. 05
i (0. 24) :(0. 24) :.(0 22)

 

’in (shade tolerant <10. 5%, shade intermediate 10. 5- 20% arid shade intolerant >20%
issification (rare < 3, intermediate 3- 10, common >10 individuals (>5 cm DBH)/ha).
of seedling (__ <5 yrs old) /ha/year. Basal area classification (rare < 0. 06, intermediate
311m 0. 03— 0. 30, large > 0. 30 mg of dry seed mass. l’Number of species in each

“i for root length and root surface area since Colubrina, Quararibea and Virola are not

{is edf = 2 for all comparisons. Significance is shown as: NS, not significant; 1' P g

Appendix F. Meta-analysis results for the effect (size and magnitude) of inoculum treatment (15111111355, Ste;
categories in each of 5 different local species characteristics. Data for meta-analysis derived i‘lCOVA an

 

 

 

 

 

 

 

 

 

 

 

Root Mass Stem Mass '1an855

Species Nbrc Effect Size QBd'e P- Effect Size QBd’e P Effects Q3“
Categoriesa (2 Std value (2 Std value (2 Std

error) error) erroj;

0.01 0.03 -0.16'
Overall 21 (0.14) (0.14) (0&1
Shade 7 0.24 0.19 —0.31
Intolerant (0.23) (0.23) (0. 24)’
Shade 0.05 * -0.13 -0. 01
Intermediate 6 (0.25) 8'36 (0.25) 3'30 NS (0. 25) 290

-0.24 0.01 -0.14
Shade Tolerant 8 (0.22) (0.22) (0.24)

0.02 0.19 -0.15 '
Adm ' Rare 7 (0.25) (0.25) (0.25)-
Adult — 0.11 0.17 -0.27 ‘
Intermediate 5 (0.29) 0'85 NS (0.29) 5‘51 T (0.29). 0'37
Adult — 9 -0.05 -O.14 -0.10 -
Common (0.20) (0.20) (0.21)‘
Seedling —- 6 -0.01 0.21 -O.l8
Rare (0.27) (0.27) (0.27)
Seedling — -0. 12 0.01 -0. 30
Intermediate 8 (0.23) 2'30 NS (0.23) 2'73 NS (0. 25) 256
Seedling — ' 7 0.13 -0.08 -0. 03
Common (0.22) (0.22) (0.22)
Basal Area — 7 -0.04 -0.05 -0.15
Rare (0.23) (0.23) (0.23) '
Basal Area — 0.11 0.14 —0.14 .
Intermediate 8 (0.23) 1'45 NS (0.23) 1'77 NS (0.23)‘. 0.10
Basal Area — 6 0.08 -0.03 —0.20 '
Common (0.25) (0.25) (0.28 '
Seed Size — 4 0.07 0.04 -030}‘«\
Small (0.31) (0.31) (0.31) 5
Seed Size — 0.13 0.01 -0.13 .
Medium 8 (0.21) 4'14 NS (0.21) 0'05 NS (0.21) ~ 0.15
Seed Size — 9 -0.18 0.04 -O.17 1'
Large (0. 22) (0. 22) (0.24)-

 

Notes. aLocalized species characteristics for all 21 species (Table 1). Shade tolerance classified?
seedling mortality at 1% full- -sun and with zero conspecific seedling density. Adult abundance 4,551“: ‘
Seedling abundance classification (rare <20, intermediate 20- 200, common >200 standing dCD‘Jf 10'

Seedl
0. 06- 0. 25, common > 0. 25 total adult mass (mz/ha)). Seed size classification (small <0. 03, mo 03]

category. CNumber of species is reduced for leaf mass since lriratea seedlings had no leaves 805,! r00t

d=f

used In the analysis. A statistics that measures the homogeneity of effect sizes between catequ °
0.10; * P5005; ** P_<_O.;01 *** PgOHOOI

142

organ mass, stern height and root morphology for 21 tropical tree species grouped into 3
E“ ANCOVA analysis using total plant mass as a covariate (Appendix C).

 

 

 

 

 

 

 

 

Leaf Mass Root Lendgth : Root Surface Area
size Q];Le P iEffectSize QB P 5 Effect Size die P
value :(2 Std value §(2 Std error) value
fi :ierror) 5
:0.02 €0.04
e 5(015) i (0.15)
50.06 50.00
i (0.26) i (0.26)
50.05 50.09
2.90 NS 3 (0.26) 1.79 NS § (026) 2.23 NS
§-0.15 §-0.17
g g (0.24) g (0.24)
50.14 50.13
i (0.26) § (0.26)
50.13 50.07
0.87 NS i (0.38) 1.53 NS i (0.37) 0.83 NS
g0.00 §-0.01
a (0.20) a (0.20)
{-0.31 €0.29
g (0.29) g (0.29)
5-0.06 50.11
2.56 NS 5 (0.24) 7.74 * i (0.24) 7.77 *
50.21 50.21
i (0.24) i (0.24)
g-0.05 §-0.08
g (0.27) g (0.27)
:0.07 :0.02
0.10 NS § (0.23) 1.15 NS § (0.23) 0.45 NS
50.12 5-0.08
:(027) 3 (0.27)
50.20 50.05
i (0.32) § (0.32)
50.07 50.02
0.15 NS § (0.22) 2.34 NS § (0.22) 0.65 NS
§-0.09 §—0.11
s (0.24) s (0.24)

 

fin (shade tolerant <10.5%, shade intermediate 105-20% and shade intolerant >20%
tssification (rare < 3, intermediate 3- 10, common >10 individuals (>5 cm DBH)/ha).
.1 of seedling (_ <5 yrs old) lha/year. Basal area classification (rare < O. 06, intermediate

um 0.03— 0.30, large > 0. 30 mg of dry seed mass.

bNumber of species in each

. for root length and root surface area since Colubrina, Quararibea and Virola are not

edf = 2 for all comparisons. Significance is shown as: NS, not significant; 1' P _<_

inoculum
categories in
rived from

 

urffce Area

,e

213 P
value

 

 

).55 NS

 

3.73 NS

 

2.74 NS

 

2.22 NS

 

1.19 NS

 

ade tolerance
shade

:ific seedling
nmon >1 0

<20,

3 old) /ha/year.

Appendix G. Meta-analysis results for the effect (size and magnitude) of inoculum
treatment on root morphology for 21 tropical tree species grouped into 3 categories in
each of 5 different local species characteristics. Data for meta-analysis derived from
AN COVA analysis using root mass as a covariate (Appendix D).

 

 

 

 

 

 

 

b Root Lgngth Root Surface Area
Species Categoriesa N 25:01 ,6 falue 183:?t ,e falue
(2 SE) (2 SE)
Overall 17 3(1):) 23(1):)
Shade Intolerant 5 233;) 23%;)
Shade Intermediate 5 23:32) 0.11 NS 233;) 0.55 NS
Shade Tolerant 7 €331) €003)“
Adult ' Rare 6 (322) (322)
Adult — Intermediate 3 2’6???” 1.04 NS $317) 0.73 NS
Adult— Common 8 23:33) €30)
. Seedling — Rare 4 233(9)) (333)
Seedling — Intermediate 7 23:33) 4.26 NS 232(2):.) 2.74 NS
Seedling -— Common 6 33214) V $00284)
Basal Area — Rare 5 3’0; 6) 23(2):)
Basal Area — Intermediate 8 23:3?) 0.42 NS 233;) 2.22 NS
Basal Area — Common 4 20022 8) $01238)
Seed Size — Small 4 $334) 30933)
Seed Size — Medium 7 (33:) 3.39 NS 23:53) 1.19 NS
Seed Size - Large 6 23(2):) 30034)

 

Notes: aLocalized species characteristics for all 21 species (Table l). Shade tolerance
classification (shade tolerant <10.5%, shade intermediate 105-20% and shade
intolerant >20% seedling mortality at 1% full-sun and with zero conspecific seedling
density. Adult abundance classification (rare < 3, intermediate 3-10, common >1 0

individuals (>5 cm DBH)/ha). Seedling abundance classification (rare <20,
intermediate 20—200, common >200 standing density of seedling (5 5 yrs old) /ha/year.

143

Appendix G. (Ctd)

Basal area classification (rare < 0.06, intennediate 0.06 - 0.25, common > 0.25 total adult
mass (mz/ha)). Seed size classification (small <0.03, medium 0.03 - 0.30, large > 0.30

mg of dry seed mass. Number of species in each category. cNumber of species is
reduced for root length and root surface area since Colubrina, Luehea, Quararibea and

. . d . . .
Virola are not used In the analys1s. Statistlc that measures the homogenelty of effect

sizes between categories. edf= 2 for all comparisons. Significance is shown as: NS, not
significant; T P 5 0.10; * P _<_ 0.05; ** P 5 0.01; *** P _<_ 0.001.

144

>233? E. meaBQE mommmn Ba 858:8: om 88388 mama—5% 8 Emma—mam.

3 En: Em!
mac—o — m8“. N mac—o u
me: 2:83.— ? =5 dramas.»— 8: 2:31..“ me: 3383:: 2:—
35 5a 5 =3 5.85.2.3 :5 538335
08.6850 35:: N l Em... 6033 cm 00333?" AZosmfinma Gov
\ moon—3mm Q8 8:: 83v manna Q3
S. may X lllv. u 2083,6855" >3: x IIV Em: Ugaa‘ om noavgmmo Zosrmdonza C 3 n moo moon—5%
/ moon—5% ANM 88— 8qu mama—o :3
.56 €39 0953 Go

 

Zen ”3:820 it 88— 83v

145

>233? I A98

3 has. Em... r02 63me o», nosmvaommo A 28.23? 08 V

moomsamm Go 88. 83v
nonmvoommo >95 X A

. . . Zo=-m81_o 08
gm: 6083 o», Osage—mo man—a C8
maoasnmm 3o 88— 83V
r02 U255 o». OoBuaommo Z - H ._ _ m

\ $2:me Cm 8E REV A o: m a: «A Vv
& m N IIIV u $22.8 ammo >3: X . H {mm 8253mm
EU 3 / Em: @3me cm Oonmvaommo Ava—7935 CMV
K magma—mm 3m 88. 8me man—a 28
Em: UQEQ o».
mafiamvoommo moan—5mm Go A Zo=-m81_003 V
8E :63

. r .56 €32. 0033— CN
22 318%? ‘ _ 88— 83v

146

Appendix I. Hazards ratios for the effects of light availability (2% vs. 22% full sun),
soil extract (non-sterile vs. sterile), soil source (tree species culturing soil) and initial

seed mass on seedling mortality estimated by the Cox regession model.

 

 

 

 

Species Source of variation df Parameter SE Wald P Ha?“
estlmate )6 value ratio
Ligthow) 1 0.25 ‘ 0.37 0.44 0.51 1.28
ExtracRNomStefile) 1 0.41 0.45 0.85 0.36 1.51
A. rubrum Source“. saccharum) 1 -0.25 0.42 0.35 0.55 0.78
Source“: americana) 1 -O.22 0.42 0.29 0-59 0.80
Source“; mm) 1 -O.36 0.43 0.71 0.40 0.70
Seed Mass 1 0.20 0.23 0.72 0.40 1.22
Ligthow) 1 1.67 1.04 2.59 0.11 5.32
Extractmomstm-le) l 1.04 0.76 1.90 0.17 2.83
A. Source“, rubrum) 1 0.54 0.62 0.76 0.25 1.72
5‘1“”an Source“: amen-cm) 1 -0.33 0.72 0.22 0.38 0.72
Source(Q. rubra) 1 0.70 0.61 1.32 0-64 2.01
Seed Mass 1 -0.06 0.02 9.87 0.002 0.94
Ligthow) 1 0.02 0.33 0.01 0.94 1.02
Extractmommme) 1 0.87 0.48 3.25 0.07 2.38
F. Source“, rubrum) 1 -0.13 0.43 0.09 0.77 0.88
“WWW“ Source“. saccharum) 1 -0.34 0.47 0.52 0.47 0.72
Source(Q. rubra) 1 0.61 0.37 2.67 0.10 1.84
Seed Mass 1 0.03 0.03 1.15 0.29 1.03
Light(Low) 1 -0.32 0.49 0.42 0.52 0.73
Extractmonsmlc) l -0.72 0.51 2.02 0.16 0.49
Q. m bra Source“ rubrum) 1 -0.75 0.81 0.86 0.36 0.47
SOUICC(A_ saccharum) 1 0.07 0.59 0.02 0-90 1.08
Source“; amen-cam) l -O.41 0.70 0.35 0-56 0.66
Seed Mass 1 -0.00 0.00 0.82 0.37 1.00

Notes: Seedlings that did not emerge are part of the analysis with their death date given
as the mean number of days prior to emergence. Total number of seedlings in model
for each species was 395 except for Q. rubra (N=304). Effects of low light vs. high
light, non-sterile vs. sterile extract, soil source (reference category is always the study
species) and initial seed mass on mortality. Significant P—values are shown in bold.

Hazard ratios > 1 indicate reduction in days to mortality of study species. For

quantitative covariates (e. g. initial seed mass), subtracting 1 from the hazard ratio and

multiplying by 100 (i.e. 100(eB - 1)) gives the per cent change in the hazard of

mortality associated with a l-unit increase in the covariate, controlling for effects of
other covariates.

147

  

Appendix J. Split-plot analysis of covariance results for the effects of light availability (2% vs. 22% full sun), soil extract (non—sterile vs. sterile), soil source (tree species culturing soil),
bench and covariate (initial seed mass) on A. rubrum seedling mass and stem height.

Source of Variation (df)

 

 

 

 

 

 

Light(1) Extract(1) Soil Source (3) Light*Extract(1) Bench (6) Covariate
Dependent (1)
Vamble F P Means F P Means F P Means (95% CI) F P Means (95% on F P F P
(95% CI) (95% CI) Ar vs. As, Fa, Qr L/N-S, L/S, H/N—S,
L vs. H N-S vs. S H/S
29.1 (26.2-32.4) vs. 13.3 (125142)
Total Mass 33.8 (30.2—37.7) 13.3 (11.4—15.5)
**** ** *>l< ** *
(mg) ”398 NA 3'9 NA 2'7 30.1(270—335) 4‘6 371.7(331.3-417.4) 4 * 8'4 ”M
34.3 (30.8—38.2) 483.5 (4149—5634)
3.4 (3.039) vs.
3.9 (3642)
Root Mass 1.5 (1.4-1.6) vs. 4.3 (3.8-4.9)
**** ** -_ _ _ **
(mg) 1020.5 68.7(61.0—77.4) 0'0 NS 38964-45) 3'6 3.6(3.2—4.2) 2'7 153 ****
‘ ‘ ' 4.3 (3.8—4.9)
7.7 (6.98.5) vs. 3.9 (3.7—4.2)
Stern Mass * ** 8.8 (7.99.9) *** 3.7 (3.244) M
(mg) 1117.4 **** NA 36 NA 3.1 ”(6.581) 7.0 63.2(56.3-71.0) 0 5.9 **
8.4 (7.69.4) 84.9 (72.8-99.1)
17.3 (154-194) vs. 7.7 (728.2)
Leaf Mass M * 19.8 (17.6-22.3) M 7.9 (6.79.3) **
(mg) “674 *** NA 5'9 NA 2'1 18.7(16.7—21.0) 4'0 238.4(211.0—269.3) 2'6 7'4 ***
20.6 (184232) 323.1 (2748—3803)
65.2 (61.8—68.8) vs. 114.5 (108.1-121.4)
Height * 66.6 (62.5-71.1) ***:1: 138.0 (1279—1489) _
(mm) 378.0 *** NA 3.7 NA 0.7 NS 666625709) 13.8 55.8(53.5-58.1) 1.7 NS —

69.0 (64.7—73.6)

52.6 (48.9-56.8)

 

Notes: Tests were performed using the Type III su

m of squares from SPSS version 14. Total number of seedlings in model = 269 (height) and 349 (all other dependent variables). All

 

  
   

—“denotes that the ANCOVA model does not contain that term due to a non-significant (P > 0.25 for covariate, bench and
abbreviations: L = low light and H = high light; N-S = non—sterile and S = sterile; Ar = A. rubrum, As
ows: NS, not significant, P > 0.10; *P g 0.10; **P g 0.05; *** P g 0.01; ****P g 0.001.

dependent variables were natural log transformed. “
interactions between main terms) effect on the dependent variable. Description of
= A. saccharum, Fa = F. americana and Qr = Q. rubra. Significance shown as foll

148

 

 

Appendix K. Split-plot analysis of covariance results for the effects of light availability (2% vs. 22% full sun), soil extract (non—sterile vs. sterile), soil source (tree species culturing
s011). bench and covariate (initial seed mass) on A. SUCC/larlllll seedling mass and stem height.

Dependent
Variable

Total Mass
(mg)

Root Mass
(mg)

Stern Mass
(mg)

Leaf Mass
(mg)

Height
(mm)-

Light (1)
F

253.9

112.3

189.9

442.4

P

****

****

****

****

****

Means (95% C1)
L vs. H

NA

NA

NA

NA

86.7 (83.5-89.9) vs.
106.4 (101.5-111.5)

149

Extract (1)

F

0.7

0.9

0.1

0.4

P

NS

NS

NS

NS

NS

Source of Variation (df)

Means (95% CI)
S vs. S

N-

154.9 (1472—1629) vs.

147.7 (1339—1629)

34.4 (31.9—37.0) vs.
31.7 (27.3-36.8)

43.9 (39.9-48.4) vs.
44.5 (42.4-46.8)

70.5 (67.1-74.0) vs.
68.1 (61.9-74.9)

95.7 (92.7—98.8) vs.
96.4 (91.4—101.7)

Soil Source (3)

F

5.7

6.8

2.6

P

****

NS

****

****

Means (95% CI)
As vs. Ar, Fa, Qr

NA

NA

NA

NA

92.2 (87.6-97.0) vs.
90.0 (61.7-131.4)
97.0 (64.5—139.4)
92.7 (54.9-148.8)

Light * Source (3)

F

6.5

4.7

5.2

4.8

P

Means (95% CI)

L/As,

L/Ar, L/FA, L/Qr vs. H/As,

H/Ar, H/Fa, H/Qr
83.2 (74.7-92.7)
87.5 (78.1—98.2)
101.2 (89.9-113.9)

92.3 (82.4—103.4) vs.

850.6 (741.7—975.5)
727.8 (600.0—881.8)
590.5 (5012—6965)
653.3 (5506—7751)
16.7 (14.2—19.7)
18.4 (15.5—21.8)
19.7 (167233)
18.6 (15.6-22.1) vs.
269.6 (219.6—331.3)
186.4 (142.5—242.0)
161.1 (125.3—207.3)
173.3 (1344—2232)
27.6 (24.8-30.7)
28.6 (25.6-32.1)
34.4 (30.7-38.7)
30.4 (27.2—34.0) vs.
166.7 (1456—1908)
133.0 (109.9—160.8)
126.2 (107.3-148.4)
135.6 (114.5-160.5)
37.4 (33.6-41.6)
38.8 (347433)
43.8 (39.0—49.2)
41.9 (37.5-46.8) vs.
395.8 (346.2—452.6)
381.1 (315.8—460.4)
294.7 (250.9-346.2)
332.6 (2815—3935)

NA

 

 

 

 

Appendix K. (Ctd)
Notes: Tests were performed using the Type III sum of squares from SPSS version 14. Total number of seedlings in model = 369. All dependent variables were natural log transformed.

“—“denotes that the AN COVA model does not contain that term due to a non—significant (P > 0.25 for covariate, bench and interactions between main terms) effect on the dependent
“-” indicates that linear regression analysis

variable or due to the assumption of homogenous slopes not being violated (P > 0.05 for interactions between main terms and covariate).
was used rather than ANCOVA due interaction term between the treatment and covariate terms being significant. Means provided are from the linear regression analysis. Description of
abbreviations: L 2 low light and H = high light; N—S = non—sterile and S = sterile; Ar: A. rubrum, As : A. saccharum, Fa = F. americana and Qr = Q. rubra. Significance shown as

follows: NS. not significant, P > 0.10; *P g 0.10; **P g 0.05; *** P g 0.01; ****P g 0.001.

150

 

Appendix K Ctd. Split-plot analysis of covariance results for the effects of light
availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source (tree
species culturing soil), bench and covariate (initial seed mass) on A. saccharum

seedling mass and stem heiat.

 

 

 

 

 

 

Dependent Source of Variation (dfi

Variable Bench (6L Covariate (1) Source * Covariate (3)
F P F i P F P

TOtal Mass 6.2 **** 104.5 **** 5.3 ****

(mg)

Rom Mass 3.1 ***-I: 57.9 ***1: _ _

(mg)

Stem Mass 4.8 ***... 91.9 ***-1: 5.6 ****

(mg)

Leaf Mass 41 **** 98.7 ***:e 6.0 ****

(mg)

Height _ _ 15.4 **** 3.1 **

flag—__—
Notes: Tests were performed using the Type HI sum of squares fi'om SPSS version 14.
Total number of seedlings in model = 369. All dependent variables were natural log
transformed. “—“denotes that the AN COVA model does not contain that term due to a
non-significant (P > 0.25 for covariate, bench and interactions between main terms)
effect on the dependent variable or due to the assumption of homogenous slopes not
being violated (P > 0.05 for interactions between main terms and covariate).
Significance shown as follows: NS, not significant, P > 0.10; *P 5 0.10; **P _<_ 0.05;
*** P 5 0.01; ****P 5 0.001.

151

Appendix L) Split—plot analysis of covariance results for the effects of light availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source (tree species

culturing soil), bench and covariate (initial seed mass) on F. americana seedling mass and stem height.

Source of Variation (df)

  

 

 

 

 

 

 

Dependent Light (1) Extract (1) Soil Source (3) Light * Extract (1) Bench (6) Covariate (1)
Variable P Means F P Means F P Means (95% CI) F P Means (95% CI) F P F P
(95% C1) (95% CI) Fa vs. Ar, As, Qr L/N-S, US vs. H/N-S,
L vs. H N—S vs. S H/S

161.1 (1418-1831) vs. 77.9 (72.1-84.4)

Total 142.6 (1251-1626) 71.5 (59.6—86.1) vs.

>1<>1<>1<>1< >1<>1< >1:

Mass (mg) 6674 NA 1‘4 NS NA 0‘9 NS 144.4 (1265—1649) 5'2 1026.6 (8925-11809) 2'0 4'4 **
147.3 (1282—1692) 1340.8 (11166-16100)
33.4 (27.9-39.6) vs. 13.6 (12.2—15.2)

Root Mass 28.2 (23.6—33.8) 11.8 (9.1—15.2) vs.

$10101: * * ** :1:

(mg) 4584 NA 0'3 NS NA 2‘1 28.5 (23.8—34.5) 3'7 286.9 (2375—3472) 2'2 5'7 *
24.5 (20.3-30.0) 373.5 (2900-4830)
41.7 (36.6-47.0) VS. 21.2 (19.7-22.8)

Stem 35.2 (31.2-39.6) 19.7 (16.6-23.5) vs.

* >1< ** ***

Mass (mg) 7043 *** NA 2'2 NS NA 2'1 37.0 (32.8-42.1) 6‘0 188.7 (1653—2153) 1'6 NS 7'3
34.5 (30.3-39.3) 250.9 (2112—2983)
81.4 (71.2—93.0) vs. 41.1 (37.9-44.6)

Leaf Mass 74.9 (65.4-85.9) ** 38.4 (31.70465) vs. *

(mg) 637.9 **** NA 1.8 NS NA 0.8 Ns 74.2 (646852) 4.7 5248 (45306073) 10 2.0 NS
83.7 (72.4—96.7) 696.5 (5754—8430)
95.3 (88.8-102.3) vs. 75.4 (71.6-79.5)

Height 85.6 (78.5-93.2) M 73.6 (67.3-80.6) vs. *

(mm) 237.3 **** NA 2.3 NS NA 1.5 NS 90.8 (84.6—97.4) 4.5 1557 (1453-1668) 1.7 NS 3.2

93.6 (87.3-100.3) 179.5 (164.0-196.4)

Notes: Tests were performed using the Type III sum of squares from SPSS version 14. Total number of seedlings in model = 256 (height) and 334 (all other dependent
variables). All dependent variables were natural log transformed. Description of abbreviations: L 2 low light and H = high light; N —S = non-sterile and S = sterile; Ar = A.

rubrum, As = A. saccharum, Fa = F. americana and Qr = Q. rubra.

 

152

 

 

 

Appendix M. Split-plot analysis of covariance results for the effects of light availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source (tree species
culturing soil), bench and covariate (initial seed mass) on Q. rubra seedling mass and stem height.

Source of Variation (of)

 

 

 

 

 

 

Dependent Light (1) Extract (1) Soil Source (3) Light * Extract (1)
Var‘able F P Means (95% CI) L F P Means (95% C1) F P Means (95% CI) F P Means (95% CI)
vs. H N-S vs. S Qr vs. Ar, As, Fa L/N-S, L/S, H/N—S, H/S
1642.5 (15284-17652)
Total Ma“ 1156.3 (10977-12180) 1430.8 (13775-14862) vs.
(m ) “ 373.9 **** vs. 7.4 *** vs. 2.6 ** 1451.0 (1346.1-15624) — — NA
g 3470.3 (32487-37071) 1618.1 (14937-17529) 1523.9 (14123-16442)
1477.3 (13706-15940)
642.9 (5846-7070) vs. 395.8 (3735-4191)
Root Mass 568.5 (5144-6289) 421.6 (3735-4191)
**** *** *>1< >1<
(mg) 2967 NA 8'7 NA 2'6 543.5 (4913-6018) 3'6 1435.1 (12960-15892)
642.9 (5846-7070) 1916.0 (16574-22128)
Stem Mass 309.8 (2971-3231) vs.
**** ** >1< _ _
(mg) 147.7 NA 4.7 3387 (31513636) 2.2 NA NA
777.9 (7199-8405) vs.
Leaf Mass 441.0 (4165-4668) vs. 683.1 (6514-7164) vs. *1- 678.6 (6250-7374) _ _
(mg) 4717 H“ 1131.2 (10526-12156) ‘9 NS 731.1 (6713-7963) 3'7 717.0 (6589-7802) NA
658.8 (6062-7180)
Height 168.9 (1634-1747) vs. *** _ _
(mm) 27.6 **** NA 0.3 NS 1660 (15654762) 4.1 NA NA

 

sum of squares from SPSS version 14. Total number of seedlings in model = 284 (stern) and 364 (all other dependent variables).
All dependent variables were natural log transformed. “—“denotes that the ANCOVA model does not contain that term due to a non-significant (P > 0.25 for covariate, bench
and interactions between main terms) effect on the dependent variable or due to the assumption of homogenous slopes not being violated (P > 0.05 for interactions between main
terms and covariate). Description of abbreviations: L 2 low light and H = high light; N-S = non-sterile and S = sterile; .Ar = A. rubrum, As = A. saccharum, Fa = F. americana
and Qr = Q. rubra. Significance shown as follows: NS, not significant, P > 0.10; *P g 0.10; **P g 0.05; *** P g 0.01; ****P g 0.001.

Notes: Tests were performed using the Type III

153

//

 

 

 

Appendix M Ctd. Split—plot analysis of covariance results for the effects of light availability (2% vs. 22% full sun), soil extract (non-sterile vs. sterile), soil source (tree species
culturing s01l), bench and covariate (initial seed mass) on Q. rubra seedling mass and stem height.

Source of Variation ((if)

 

 

 

 

 

 

Dependent nght * Source (3) Extract * Source (3) Bench (6) Covariate (1) Source * Covariate (3)
Variable F P Means (95% CI) F P Means (95% CI) F P F P F P
L/Qr. L/Ar, L/As, L/Fa vs. N-S/Qr, N—S/Ar, N—
H/Qr, H/Ar, H/As, H/Fa S/As. N—S/Fa vs. S/Qr,
S/Ar, S/As, S/Fa
Total Mass (mg) — —— NA —— — NA 2_() * 3434 **>1<>1< __ _
Root Mass (mg) — — NA - .— NA 21 * 241.9 ***a: _ _
274.0 (2537-2962)
228.8 (207.1-296.2)
261.9 (2377-2886)
Stem Mass (mg) 2.7 * (26913354(%612.3—)324.1)vs. 676.5 — — NA 2.9 >1<>1< 278.7 ***>l< 2.8 **
519.0 (4604-5858)
646.] (570.2-732.2)
575.4 (5112—6481)
Leaf Mass (mg) — — NA — — NA _ _ 260.8 >1:>1<>1<=1< _ _
153.3 (142.1—1654)
136.7 (126.3—147.9)
154.0 (1421-1668)
Height (mm) 5.0 *** 177.0 (1631—1921) vs. 196.4 1.9 NS NA _ _ 1135 MM _ _

Notes: Tests were performed using the Type III sum of sq
All dependent variables were natural log transformed. “——“
and interactions between main terms) effect on the dependent variable or due to the assumption of h

terms and covariate). Description of abbreviations: L =

(1797—2147)

166.2 (1498—1843)
196.4 (176.2—218.7)
169.2 (1527—1877)

and Qr = Q. rubra. Significance shown as follows: NS, not significant, P > 0.10; >“P _<_ 0.10; **P g 0.05; *** P g 0.01; ****P g 0.001.

154

uares from SPSS version 14. Total number of seedlings in model = 284 (stem) and 364 (all other dependent variables).
denotes that the ANCOVA model does not contain that term due to a non-significant (P > 0.25 for covariate, bench
omogenous slopes not being violated (P > 0.05 for interactions between main
low light and H = high light; N—S = non—sterile and S = sterile; Ar = A. rubrum. As = A. saccharum, Fa = F. americana

 

 

Appendix N. Analysis of variance results for the effect of soil source (tree species
culturing soil) on base cations (combined Ca, K, and Mg), total organic C, total N, and
C:N in the soil extracts.

 

Source of F P Means (95% CI)
Variation

Ar = 259.39 (24596-27283)
Base Cations 11 15 0 000 As = 288.79 (27536-30222)
(ppm) ' ' Fa = 302.14 (28816-31612)

Ql' = 255.49 (242.05-268.92)

Ar = 5.47 (4.46-6.48)
Total Organic C As = 4.22 (3.21-5.23)
(mg/L) ”'24 0'0"" Pa = 4.28 (3.21-5.23)
Qt = 8.17 (7.16-9.18)
Ar = 2.30 (1 .68-2.91)
As = 2.54 (1.68-2.91)
Fa = 2.54 (1.92-3.15)
Q1 : 2.22 (1 .61-2.84)
Ar = 2.37 (1 922.91)
, As= 1.71 (1.39-2.11)
ON 1691 0'0”” Pa = 1.48 (1.18-1.86)

g = 3.81 (3.10-4.69)

Notes: Tests were performed using the Type III sum of squares from SPSS version 15.
Degree of freedom for all cation models is 3,47 and for total organic C, total N, and
C:N ratios are 3,19. Description of abbreviations: Ar = A. rubrum, As = A. saccharum,
Fa = F. americana and Qr = Q. rubra.

Total N (mg/L) 1.00 0.41

155

Appendix 0. Kaplan-Meier analysis of the effect of inoculum type (Five Fusarium
morphotypes and control) on seedling life span for A. rubrum, F. americana and Q.
rubra.

 

 

 

 

Species Treatment 2:186:21; Breslow, 12° P value
Control 56.30 (1.66) '
Fusarium 1 52.30 (3.19)
Fusarium 2 56.10 (1.85)

A' ”‘me Fusarium 3 55.55 (2.39) ”'99 0'0”
F usarium 4 45.85 (4.19)
Fusarium 5 51.35 (3.26)
Control 39.65 (4.91)
F usarium l 31.50 (4.46)

. Fusarium 2 42.90 (4.19)

F. americana Fusarium 3 36.70 (4.38) 4.06 0.54
F usarium 4 36.80 (4.46)
Fusarium 5 41.75 (4.99)
Control 58.00 (0.00)
Fusarium 1 58.00 (0.00)

Q. ru bra Fusarium 2 57.40 (0.63) 5.00 0.42

F usarium 3 58.00 (0.00)
Fusarium 4 58.00 (0.00)

Fusarium 5 58.00 (0.00)

Notes: Seedlings that did not emerge are part of the survival analysis with their death
date given as the mean number of days prior to emergence. df= 5, 120 for each species.

156

Appendix P. Mixed effects analysis of covariance results for the fixed effects of
inoculum type (Control vs. Fusarium morphotype 1-5), random effect of bench and

covariate (initial seed mass) on A. rubrum seedling mass and stem heigt.

Source of Variation (df)

Inoculum (5) Bench (1) Covariate
Dependent (1)
Variable F P Treatment Means F P F P
(95% CI)
14.6 (12.3-17.3)
13.6 (11.4-16.3)
14.4 (12.1-17.0)
11.4 (9.6-13.6)
13.0 (10.4-16.3)
9.9 (8.1-12.0)

 

Control

F usarium 1
F usarz'um 2
F usarium 3
F usarium 4
F usarz'um 5

Total (mg) 2.7 0.03

\O

.8 0.002 — —

Root (mg)

Stem (mg)

Leaves
(mg)

Cotyledon
(mg)

3.1 0.01

2.6 0.03

2.7 0.03

1.5 0.21

Control

F usarium 1
F usarium 2
F usarium 3
F usarium 4
F usarium 5
Control

F usarium 1
F usarium 2
F usarium 3
F usarium 4
F usarium 5
Control

F usarium 1
F usarium 2
F usarium 3
F usarium 4
F usarium 5
Control

F usarium 1
F usarium 2
F usarium 3
F usarium 4
F usarium 5
Control

1.7 (1.42.1)
1.5 (1.2-1.9)
1.4(1.1-1.8)
1.0 (0.7-1.3)
1.4 (1.0-1.9)
1.0 (0.7-1.3)
3.3 (2.9-3.7)
2.9 (2.4-3.3)
3.2 (2.7-3.6)
2.6 (2.1-3.0)
2.8 (2.3-3.3)
2.4 (1.9-2.8)
7.4 (5.8-9.4)
6.8 (5.2-8.8)
7.4 (5.79.4)
5.1 (3.9-6.6)
6.5 (4.7-9.0)
4.2 (3.1-5.7)
2.2 (1.8-2.5)
2.1 (1.72.4)
2.4 (2.12.7)
2.7 (2.3-3.0)
2.3 (1.92.8)
2.6 (2.22.9)

9.4

5.8

9.5

0.007

0.02

0.003

2.6 0.11

46.5 (42.6-50.4)
Fusarium 1 43.6 (39.5-47.8)
1 5 0 19 Fusarium 2 47.6 (43.7-51.5) _ __ _ _
' ' Fusarium 3 42.9 (38.9
Fusarium 4 42.5 (37.2
Fusarium 5 40.8 (36.4

Height
(cm)

 

157

Appendix P. (Ctd)

Notes: Tests were performed using the Type III sum of squares from SPSS version 14.
Total number of seedlings in model = 99. All dependent variables except stem and
cotyledon mass and stem height were transformed with natural log function. “—
“denotes that the AN COVA model does not contain that term due to a non—significant
(P > 0.25 for covariate and for random factor of bench) effect on the dependent
variable.

158

Appendix Q. Mixed effects analysis of covariance results for the fixed effects of

inoculum type (Control vs. F usarium morphotype 1-5), random effect of bench and

covariate (initial seed mass) on F. americana seedling mass and stem heigt.

 

 

 

 

 

 

 

Source of Variation (df)
Inoculum (5) Bench (1) Covariate
Dependent (1)
Variable F P Treatment Means F P F P
(95% CI)
Control 40.5 (30.4-53.9)
Fusarium 1 32.5 (22.1—47.6)
Fusarium 2 49.3 (37.2-65.2)
Total (mg) 2.5 0.04 Fusarium 3 33.4 (23.8-46.6) 3.8 0.06 2.9 0.10
Fusarium 4 58.3 (41.8-81.0)
Fusarium 5 56.8 (43.8-73.4)
Control 4.8 (3.7-6.3)
Fusarium l 3.7 (2.4-5.3)
Fusarium 2 6.6 (5.1-8.5)
Root (mg) 1.8 0.13 Fusarium 3 4.2 (3.0-5.7) 5.2 0.03 5.5 0.02
Fusarium 4 5.3 (3.8-7.2)
Fusarium 5 5.0 (3.9-6.3)
Control 8.8 (7 .0—1 1.0)
Fusarium 1 7.4 (5.4-10.0)
Fusarium 2 9.8 (7.9-12.2)
Stem (mg) 2.9 0.02 Fusarium 3 7.9 (6.1-10.3) 4.4 0.04 5.3 0.03
Fusarium 4 12.6 (9.8-16.2)
Fusarium 5 11.9 (9.8-14.5)
Control 19.0 (12.3-28.9)
Fusarium 1 13.6 (7.4-24.2)
Leaves Fusarium 2 22.8 (14.9-34.6)
(mg) 2'8 0'03 Fusarium 3 10.2 (5.9-17.1) 4'8 0°03 _ _—
Fusarium 4 28.1 (17.2-45.8)
Fusarium 5 27.5 (18.7—40.3)
Control 8.6 (6.9-10.2)
Fusarium 1 8.6 (6.3-10.8)
Cotyledon Fusarium 2 9.8 (8.2-11.5)
(mg) 1'2 0'33 Fusarium 3 9.1 (7211.0) _ _ 6'4 0'02
Fusarium 4 10.4 (8.4-12.3)
Fusarium 5 10.8 (9.3-12.4)
Control 74.1 (63.9-84.3)
Fusarium 1 60.2 (46.6-73.8)
Height F usarium 2 72.4 (62.3-82.4)
(cm) 1'7 0'15 Fusarium 3 67.4 (55.6-79.1) 2'7 0'11 — —
Fusarium 4 81.4 (69.4-93.3)
Fusarium 5 80.2 (70.8-89.5)

159

Appendix Q. (Ctd)

 

Notes: Tests were performed using the Type III sum of squares from SPSS version 14.
Total number of seedlings in model = 57. All dependent variables except cotyledon
mass and stem height were transformed with natural log function. “—“denotes that the
AN COVA model does not contain that term due to a non-significant (P > 0.25 for
covariate and for random factor of bench) effect on the dependent variable.

160

>255? W. 2:28 «@005 25272me ow 855258 Hams—5 men :8 98a 03.0on ow Boos—E: 38 A0033— <m. $525.25 Bone—5368 T

 

u . 55908 03.8" cm ago: Ba 032128 532— 88 82mm on

 

 

 

 

 

mofioo om <mnmmo= A28
5095:: CV manor A: 0925.28 A: 509:5: ..
mam—fig" 092528 CV
B a 2.. 5 $83.82 :88 32. as 2.. w 2.. 2. a. 2.
02:3— 5:; C 39%.:qu
$2525.25 H 38.5 GSNLNELV
52522.25 N 85b 3 585.3293 I: I
2.08— 35 PM 93 5525.25 u Emma CommuLumcbv PM Pom 59¢ 98:
$2525.25 A Emma. G _orflioobv
22.252125 m 33.5 232539:
025.c— Nomu $39323
25.252125 H Mme; $099398
282 38. N2 25 Mmmmumw WWW.“ mmwmwmw 2 E: :25 2.8 No 98
5525.25 5 man APTQGPMV
29523.25 m Ht; @893me
9520— Nawu Amumuhoouv
M52125 _ Noah Cmoubwkuv
was 25$ .3 98 mmmmumw WWW mwummww 3 9.: 8; 98.. l 1
5525.25 5 Nth figbhopov
h52125 m was Awuobbgbv
0952 $2; 35.7393
52525.25 — uNmb $295-23an
525188 _b 93 fi52125 N Eco 03.0533 II I :23 P25 |. |
512525.25 w 93b Am:.o-mmo.$
$522.25 2 mama Am3.o-amo.£

161

>285? Hw. A02:

@8225 m 80.0 90.0-3.8
0952 58 $888.3
$285.25 H Hawb AHahHmwuv

. 2.82%.: N a; wadé

mama Ea 0.0 SK @8225 0 :8 2338.3
@8225 2 a; c3932.:
2.8225 0 52 5.380

I I H w .o 925 I I

 

2.080.. H.008 €08 001.0809 Sam 90 .360 E 88 OH. 3580 $08 mHumm «.088: HP .HdSH 5:809 0H. 0003880 5 Song u H Hm.
:I.do:o8m 98 E0 >ZOO<> 8098 @000 :2 02:8: 58 83. 98 8 0 38-88888: av v 93 man 0928.88 85 man 8.8508
808.. 0», 60:05 3.89 o: 90 800308201020 9. 98 8 90 0008838: 0H. con—£05080 80000 :2 605m $0589 AHV v Pom mom
5833080 009200: H.809 H.880 85 008188. a... SEQ—80 98 H908. 88.00805 8858 £8 :03 390—. as: >ZOO<> 98
588030: 8:: H8380: :8 #08302 88 038.58 835 H00H=m 05:508.». 3085 H5258 8.0 m3:— 90 :88. 8m8mmH0=

85580.

162

Appendix S. Experimental design and allocation of tropical seedlings to treatments.

a) High Light

6spr—>

\

b) Low Light

6spr—>

Step 1 - Soil cultured in
the field and greenhouse
by: '

Conspecific Adult X

5 Heterospecific Adult X

Not applicable +
Step 1 - Soil cultured in

the field and greenhouse

by:

Conspecific Adult X

5 Heterospecific Adult X

Not applicable

163

Step 2 - Soil extraction:

(Non-Sterile (16)

Sterile (16)

(Non-Sterile (16) = 1,243 sdlg
Sterile (l6)

Tap Water Control
(16 total reps)

Step 2 - Soil extraction:

(Non-Sterile (30)
Sterile (18)

(Non-Sterile (30) = 1,836 sdlg

Sterile (18)

Tap Water Control
(1 8 total reps)

Appendix T. Hazards ratios for the effects of light availability (1% vs. 5% full sun),
soil extract (non-sterile vs. sterile), soil source (tree species culturing soil) and initial

seed mass on seedling mortalitz estimated bx the Cox rcEession model.

 

 

 

 

 

Species Source of d f Parameter SE Wald P Hazard
Variation estlmate )6 value ratio
Ligthow) 1 0.97 0.15 44.76 0.000 2.65
Extractmonsmle) l -0.03 0.13 0.05 0.83 0.97
Source(Co,ub,,-na) 1 0.26 0.23 1.33 0.25 1 .30
Apeiba SourceUn-am, 1 0.15 0.21 0.48 0.49 1.16
membranacea Source(pe,,,ade,;,,a) 1 -004 0.22 0.04 0.85 0.96
Source(p,es,0ea) 1 0.27 0.22 1.51 0.22 1.30
Source(y,-,01a) 1 0.46 0.21 4.97 0.03 1.59
Seed Mass 1 0.04 0.03 1.87 0.17 1.05
Ligthow) 1 0.89 0.20 19.97 0.000 2.43
Extract(Non-sten-le) 1 -0.28 ‘ 0.17 2.76 0.10 0.75
Sourceupeiba) 1 -1.06 0.34 10.01 0.002 0.35
Sourcegfiama) 1 0.34 0.23 2.12 0.15 1.40
Célubrina Source(pe,,,acle,hm) 1 -0.69 0.29 5.77 0.02 0.50
31”"05“ Source(p,es,0ea) 1 -053 0.30 3.88 0.05 0.56
Source(y,-,ola) l -l.03 0.33 9.83 0.002 0.36
Seed Mass 1 0.01 0.00 2.22 0.15 1.01
Ligthow) 1 -0.21 0.16 1.72 0.19 0.81
Extractmomsteme) l -0.01 0.16 0.01 0.93 0.99
Sourceupeiba) l -0.01 0.24 ~ 0.00 0.97 0.99
Iriartea Sourcewolubn-M ) 1 0.93 0.23 16.00 0.000 2.52
deltoidea Source(pe,,,acle,hm) 1 -2.08 0.51 16.41 0.000 0.13
Source(p,.es,oea) 1 -l .30 0.38 1 1.78 0.001 0.27
Source(V,-,o,a) 1 0.10 0.23 0.19 0.66 1.1 1
Seed Mass 1 0.00 0.00 1.70 0.19 1.00
Ligthow) 1 1.36 0.41 11.03 0.001 3.87
Extractmommme) 1 1.02 0.35 8.29 0.004 2.78
Pentaclethra Sourceupeiba) 1 -0.27 0.54 0.25 0.62 0.77
m a a, 0 l 0 b a Source(colub,im) 1 0.56 0.45 1.55 0.21 1.75
Sourcegfiama) 1 0.23 0.47 0.23 0.63 1.25
Source(p,es,oea) 1 -0.00 0.50 0.00 0.99 1 .00
Source( 14,010) 1 -0.50 0.57 0.77 0.38 0.61

164

Appendix T (Ctd)

 

 

Ligthow) 1 0.78 0.26 9.24 0.002 2.17
Extractmommfile) 1 0.26 0.26 1.07 0.30 1.30
Sourceupeiba) 1 -0.23 0.33 0.50 0.48 0.79
Prestoea Sourcewolubn-M) 1 -040 0.34 1.42 0.23 0.67
decurrens Source(1,,-a,.,ea) 1 -053 0.35 2.27 0.13 0.59
Source(pe,,,acle,h,a) 1 -035 0.34 1.04 0.31 0.71
Source(y,-,01a) 1 -1.34 0.46 8.71 0.003 0.26
Exfactmm'sm‘fl 1 -0.69 0.51 1.83 0.18 0.50
lghtflow)
Ligthow) 1 -019 0.46 0.17 0.68 0.83
EXtraCRNon-Sterile) 1 -0.1 8 0.46 O. 16 0.69 0.83
Sourceupeiba) l 0.90 0.72 1.56 0.21 2.45
Virola Source(colub,m) 1 0.00 0.96 0.00 1.00 1.00
koschnyi Source(1,,-a,,ea) 1 -0.82 1.17 0.50 0.48 0.44
Source(pe,,,ac,e,,,m) 1 0.27 0.83 0.1 l 0.75 1.31
SOUTCC(p,-estoea) l 0.43 0.77 0.31 058 154
Seed Mass 1 0.00 0.00 2.34 0.13 1.00

Notes: Effects of low light vs. high light, non-sterile vs. sterile extract, soil source
(reference category is always the study species) and initial seed mass on mortality.

Total number of seedlings in model for each species was 480 except for Apeiba

membranacea = 469, Colubrina spinosa = 471 and Iriartea deltoidea = 454. Hazard
ratios > 1 indicate reduction in days to mortality of study species. For quantitative
covariates (e. g. initial seed mass), subtracting 1 from the hazard ratio and multiplying

by 100 (i.e. 100(eB - 1)) gives the per cent change in the hazard of mortality associated

with a l-unit increase in the covariate, controlling for effects of other covariates.

165

Appendix U. Split-plot analysis of covariance results for the effects of light availability
(1% vs. 5% full sun), soil extract (non-sterile vs. sterile), soil source (tree species
culturing soil), bench and covariate (initial seed mass) on seedling mass (mg) for each

species.

1) A. membranacea

 

 

Source of Variation

df

F

'Means (95% CI)

 

Light

Extract

Soil Source

Bench
Covariate

95.80

0.06

0.41

1.30
4.62

0.000

0.82

0.84

0.25
0.03

L = 4.5 (3.9 — 5.2) vs.

H = 11.6 (10.5 — 12.8)

N-S = 8.7 (6.0 — 7.5) vs.
s = 6.6 (5.8 — 7.5)

Am = 6.4 (5.2 - 7.7) vs.
Cs = 6.7 (5.4 — 8.3)

Id = 6.2 (5.1 — 7.7)

Pm = 6.5 (5.4 — 7.9)

Pd = 7.5 (6.1 — 9.0)

Vk = 6.6 (5.2 — 8.3)

NA

NA

 

2) C. spinosa
Source of Variation

Light
Extract

Soil Source

F
42.12

1.15

3.38

0.000
0.29

0.005

Means (95% CI)

NA

N-S = 21.3 (20.2 — 22.5) vs.
S = 22.3 (20.9 — 23.9)

NA

 

Light * Soil Source

Bench

2.89

4.96

0.01

0.000

166

L/Cs = 16.5 (13.9 — 19.6) vs.
L/Am= 17.2 (15.1 — 19.6)
1de = 16.5 (12.2 — 20.1)
L/Pm= 17.0 (14.9— 19.3)
L/Pd= 16.6 (14.5 —19.0)
L/Vk = 17.7 (15.7 —20.1)

I-I/Cs = 34.6 (29.8 — 41.0) vs.
I-I/Am = 36.0 (31.5 — 41.0)
H/Id = 20.5 (17.3 — 24.4)
H/Pm = 34.1 (29.2 — 39.9)
H/Pd = 35.6 (30.9 — 41.1)
HNk = 32.8 (28.5 - 37.7)
NA

Appendix U. (Ctd)
3) I. deltoidea

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Source of Variation df F P Means (95% CI)
. L = 181.9 (169.2 — 194.5) vs.
1‘13” 1 0'30 0'58 H = 186.9 (171.3 — 202.6)
N-S = 190.4 (177.1 — 203.8) vs.
Ema“ 1 1'78 0'18 S = 178.4 (163.6 - 193.2)
Id = 176.2 (154.0 — 198.4) vs.
Am = 162.9 (140.3 -— 185.6)
Soil Source Cs = 138.1 (95.1 — 182.0)
5 ”'32 0'00" Pm = 254.6 (237.7 — 271.5)
Pd = 206.5 (188.8 — 224.1)
Vk = 168.1 (144.5 — 191.8)
4) P. macroloba
Source of Variation jf F P Means (95% CD
. L = 1729.4 (1661.1 — 1800.5) vs.
“gm 1 (24°08 0'0"] H = 2256.0 (2155.6 — 2361.1)
N-S = 1967.0 (1889.1 — 2048.0) vs.
Ema“ 1 2°98 0'09 S = 1880.8 (1797.6 — 1967.9)
Pd = 1741.9 (1641.1 - 1879.8) vs.
Am = 2015.7 (1876.1 -— 2165.8)
. Cs=2020.1(1871.8—2180.0)
3"“ 30m“ 5 2'06 0'07 Id = 1847.4 (1713.7 — 1991.6)
Pm = 1957.6 (1818.9 — 2106.8)
Vk = 1974.0 (1835.7 — 2122.6)
Light * Extract 1 2.12 0.15 NA
Light * Soil Source 5 1.34 0.25 NA
Extract * Soil Source 5 1.24 0.29 NA
. * * .
Light Extract 8011 5 1.41 0.22 NA
Source
Bench 8 3.06 0.002 NA
Covariate 1 278.61 0.000 NA
3 OP. decurrens (without cotyledon mass)
Source of Variation df F P Means (95% CI)
Light 1 18.76 0.002 NA
Extract 1 1.93 0.17 NA
Soil Source 5 6.08 0.000 NA
Light * Extract 1 3.41 0.07 NA
Light * Soil Source 5 0.78 0.56 NA
Extract * Soil Source 5 0.62 0.69 NA

 

167

Appendix U. (Ctd)

L/N-S/Pd = 69.3 (62.3 — 76.4)
L/N-S/Am = 71.0 (63.7 — 78.2)
L/N-S/Cs = 70.4 (63.0 — 77.8)
L/N-S/lId = 69.4 (62.2 — 76.6)

_ L/N-S/Pm = 60.4 (52.9 — 68.0)

L/N-SN'k = 76.5 (69.7 - 83.3)

L/S/Pd = 16.5 (13.9 —19.6)
L/S/Am = 73.0 (63.8 — 82.3)
L/S/Cs = 61.6 (52.7 - 70.6)
L/S//Id = 70.6 (61.0 — 80.1)
L/S/Pm = 58.2 (48.5 — 67.8)
L/SNk = 62.7 (53.8 — 71.7)

 

 

. * .1. .

185251111126 Extract $011 5 2.53 0.03
H/N-S/Pd = 74.7 (65.8 — 83.6)
H/N-S/Am = 81.1 (71.6 — 90.7)
H/N-S/Cs = 74.5 (64.9 — 84.1)
H/N-S//Id = 87.4 (78.8 — 96.1)
H/N-S/Pm = 74.7 (65.8 - 83.6)
H/N-SNk = 81.6 (72.7 — 90.5)
H/S/Pd = 69.6 (60.4 — 78.8)
H/S/Am = 82.5 (72.9 — 92.1)
H/S/Cs = 85.1 (76.2 - 94.0)
H/S//Id = 75.5 (66.6 — 84.4)
H/S/Pm = 70.2 (61.2 — 79.1)
H/SNk = 81.2 (72.6 — 89.9)

Bench 8 1.71 0.09 NA

Covariate 1 75.56 0.000 NA

 

 

6) §P. decurrens (seedlings without cotyledon’s at harvest were excluded from

 

 

 

 

 

 

analysis)
Source of Variation df F P Means (95% CI)
Ligt 1 0.85 0.36 NA
N-S = 150.4 (148.2 — 152.6) vs.
Ema“ 1 0'07 0'79 S = 150.0 (147.6 — 152.4)
Soil Source 5 6.1 1 0.000 NA
Light * Extract 1 1.39 0.24 NA
L/Pd = 156.7 (151.4 - 162.0) vs.
L/Am = 149.6 (144.4 — 154.7)
Light * Soil Source 5 2.78 0.02 ”CS = ”3'5 (”6'6 ‘ 1472)

168

Md = 141.9 (143.9 - 155.0)
L/Pm = 149.5 (143.9 —155.0)
L/Vk = 155.5 (150.8 - 160.3)

Appendix U. (Ctd)

 

H/Pd = 152.8 (145.8 — 159.7) vs.
H/Am = 160.9 (155.1 — 166.6)
H/Cs = 147.0 (141.2 -— 152.7)
H/Id = 148.5 (142.9 —154.0)

, H/Pin = 144.4 (138.3 — 150.5)

HNk = 152.3 (146.8 — 157.8)

 

 

 

 

 

 

 

 

 

 

 

Covariate 1 582.71 0.000 NA
7) V. koshnyi
Source of Variation df F P Means (95% CI)
. L = 739.6.0 (712.7 — 766.5) vs.
L131“ 1 ”'76 0'00” H = 867.5 (835.0 — 900.1)
Extract 1 0.66 0.42 NA
3°“ SW“ 5 3.70 0.003 NA
Light * Soil Source 5 1.40 0.23 NA
N—SNk = 667.4 (601.1— 733.8) vs.
N—S/Am = 843.7 (777.1 — 910.4)
N-S/Cs = 842.9 (771.0 — 914.8)
N-S/Id = 895.1 (829.4 — 960.9)
N-S/Pm = 792.7 (725.6 — 856.8)
. N-S/Pd = 829.1 (763.1 — 895.1)

Extract * Soil Source 5 2.23 0.05
SNR = 663.5 (579.5 — 747.2) vs.
S/Am = 826.0 (748.8 — 902.9)
S/Cs = 902.3 (819.1 — 985.1)
S/Id = 774.6 (702.2 — 846.7)
S/Pm = 856.5 (799.8 — 928.8)
S/Pd = 748.9 (673.4 — 824.1)

Covariate 1 70.31 0.000 NA

S0“ 59““ * 1 4.27 0.001 NA

Covariate

Notes: Tests were performed using the Type HI sum of squares from SPSS version 15.
Total number of seedling in model for A. membranacea = 204, C. spinosa = 333, I.
deltoidea = 277, P. macroloba = 425, OP. decurrens (without cotyledon mass) = 390,
§P. decurrens (seedlings without cotyledon’s at harvest were excluded from analysis) =
345, and V. koschnyi = 460. Seedling mass for A. membranacea, C. spinosa, and P.
macroloba were natural log transformed. I. deltoidea seedling mass did not include
cotyledon mass. Description of abbreviations: L = low light and H = high light; N-S =
non-sterile and S = sterile; Am = Apeiba membranacea, Cs =Colubrina spinosa, Id =
Iriartea deltoidea, Pm = Pentaclethra macroloba, Pd = Prestoea decurrens, and Vk =

Virola koschnyi.

169

LITERATURE CITED

Agrios, G. N. 1997. Plant pathology. 4th edition. Academic Press, San Diego, California,
USA.

Ahumada, J ., S. P. Hubbell, R. Condit, and R. B. Foster. 2004. Long-term tree survival in
a neotropical forest. Pages 408—432 in E. C. Losos and E. G. Leigh, Jr, editors.
Tropical forest diversity and dynamism. The University of Chicago Press,
Chicago, Illinois, USA.

Armstrong, D. P., and M. Westoby. 1993. Seedlings from large seeds tolerate defoliation
better: a test using phylogenetically independent contrast. Ecology 74:1092—1100.

Augspurger, C. K. 1983a. Offspring recruitment around tropical trees: changes in cohort
distance with time. Oikos 40: 189-196.

Augspurger, C. K. 1983b. Seed dispersal of the tropical tree, Platypodium elegans, and
the escape of its seedlings from fungal pathogens. Journal of Ecology. 71: 7 59-77.

Augspurger, C. K. 1984. Seedling survival of tropical tree species: interactions of
dispersal distance, light-gaps, and pathogens. Ecology 65:1705—1712.

Augspurger, C. K. 1990. Spatial patterns of damping-off disease during seedling
recruitment in tropical forests. Pages 131-144 in J. Burdon, and S. Leather,
editors. Pests, Pathogens, and Plant Communities. Blackwell Scientific
Publications, Oxford, UK.

Augspurger, C. K., and C. K. Kelly. 1984. Pathogen mortality of tropical seedlings:
experimental studies of the effects of dispersal distances, seedling density, and
light conditions. Oecologia 61 :21 1—217.

170

Augspurger, C. K., and H. T. Wilkinson. 2007. Host-specificity of pathogenic Pythiurn
species: implications for tree species diversity. Biotropica 39(6):702—708.

Austin, P. C ., and J. E. Hux. 2002. A brief note on overlapping confidence intervals.
Journal of Vascular Surgery 36(1): 194-195.

Baltzer, J. L. and S. C. Thomas. 2007. Determinants of whole-plant light requirements in
Bomean rain forest tree seedlings. Journal of Ecology. 95(6): 1208-1221.

Bancroft, T. A. 1964. Analysis and inferences for incompletely specific models involving
the use of preliminary tests of significance. Biometrics 20:427-442.

Barot, S., J. Gignoux, and J. C. Menaut. 1999. Seed shadows, survival and recruitment:
how simple mechanisms lead to dynamics of population recruitment curves. Oikos
86: 320-330.

Basset, Y. 1994. Palatability of tree foliage to chewing insects: a comparison between a
temperate and a tropical site. Acta Oecologica 15: 181-191.

Blundell, A. and D. Peart. 1998. Distance-dependence in herbivory and foliar condition
for juvenile Shorea trees in Bomean dipterocarp rain forest. Oecologia 117: 151-160.

Bell, T., R. P. Freckleton, and O. T. Lewis. 2006. Plant pathogens drive density-
dependent seedling mortality in a tropical tree. Ecology Letters 9(5):569—5 74.

Bever, J. D. 1994. Feedback between plants and their soil communities in an old field
community. Ecology 75: 1965—1977.

Bezemer, T. M., C. S. Lawson, K. Hedlund, A. R. Edwards, A. J. Brooks, J. M. Igual, S.
R. Mortimer, and W. H. van der Putten. Plant species and fimctional group effects
on abiotic and microbial soil properties and plant-soil feedback responses in two
grasslands. Journal of Ecology 94: 893-904.

Bigelow, S. W. and C. D. Canham. 2007. Nutrient limitation of juvenile trees in a
northern hardwood forest: Calcium and nitrate are preeminent. Forest Ecology
and Management 243: 310-319.

171

Booth, M. G. 2004. Mycorrhizal networks mediate overstorey-understorey competition in
a temperate forest. Ecology Letters 7(7): 538-546.

Borowicz, V. 2001. Do arbuscular mycorrhizal finigi alter plant-pathogen interactions.
Ecology 82(11): 3067-3068.

Boudreau, S. and M. J. Lawes. 2008. Density- and distance-dependent seedling survival
in a ballistically dispersed subtropical tree species Philenoptera sutherlandii.
Journal of Tropical Ecology 24:1-8.

Canham, C. D., A. C. Finzi, S. W. Pacala, and D. H. Burbank. 1994. Causes and
consequences of resource heterogeneity in forests: interspecific variation in light
transmission by canopy trees. Canadian Journal of Forest Research 24: 337-349.

Casper, B. B., and J. P. Castelli. 2007. Evaluating plant-soil feedback together with
competition in a serpentine grassland. Ecology Letters 10: 394-400.

Chazdon, R. L., and N. Fetcher. 1984. Photosynthetic light environments in a lowland
tropical rain-forest in Costa Rica. Journal of Ecology 72(2):553—564.

Cintra, R. 1997. A test of the J anzen-Connell model with two common tree species in
Amazonian forest. Journal of Tropical Ecology 13: 641-658.

Clark, D. A. and D. B. Clark. 1984. Spacing dynamics of a tropical rain forest tree:
evaluation of the J anzen-Connell model. American Naturalist. 124(6): 769-788.

Clark, J. S., M. Silrnan, R. Kern, E. Macklin, J. HilleRisLambers. 1999. Seed dispersal
near and far: patterns across temperate and tropical forests. Ecology 80(5): 1475-
1494.

Coley, P. D. and T. M. Aide. 1991. Comparison of herbivory and plant defenses in
temperate and tropical broad-leaves forests. Pages 25-49 in P. W. Price, T. M.
Lewinsohn, G. W. Femandes, and W. W. Benson, editors. Plant-Animal
Interactions: Evolutionary Ecology in Tropical and Temperate Regions. John
Wiley and Sons, New York, NY, USA.

172

Coley, P. D., and J. A. Barone. 1996. Herbivory and plant defenses in tropical forests.
Annual Review of Ecology and Systematics 27:305—335.

Coley, P. 1)., J. P. Bryant, and F. s. Chapin, 111. 1985. Resource availability and plant
antiherbivore defense. Science 230:895—899.

Condit, R., S. P. Hubbell, and R. B. Foster. 1992. Recruitment near conspecific adults
and the maintenance of tree and shrub diversity in a neotropical forest. American
Naturalist 140: 261-286.

Connell, J. H. 1971. On the role of natural enemies in preventing competitive exclusion
in some marine animals and in rain forest trees. Pages 298—3 12 in P. J. den Boer
and G. R. Gradwell, editors. Dynamics of populations. Centre for Agricultural
Publishing and Documentation, Wageningen, The Netherlands.

Connell, J. H. 1978. Diversity in tropical rain forests and coral reefs: high diversity of
trees and corals is maintained only in a non-equilibrium state. Science 199:1302—
1 3 10.

Connell, J. H., J. G. Tracey, and L. J. Webb. 1984. Compensatory recruitment, growth,
and mortality as factors maintaining rain forest tree diversity. Ecological
Monograph 54: 141-164.

Cox, D. R., and D. Oakes. 1984. Analysis of Survival Data. Chapman and Hall, New
York.

Dalling J. W. 2004. The fate of seed banks: factors affecting seed survival for light-
demanding species in tropical forests. Pages 31-44 in Seed Fate: Predation,
dispersal and seedling establishment, P. M. Forget, J. E. Lambert, P. E. Huhne,
and S. Vander Wall (Editors). CABI, Wallingford, UK

De Steven, D. and F. E. Putz. 1984. Impact of mammals on early recruitment of a tropical
canopy tree, Dipteryx panarnensis, in Panama. Oikos 43(2):207-216.

173

De Steven, D. and J. Wright. 2002. Consequences of variable reproduction for seedling
recruitment in three neotropical tree species. Ecology 83(8): 2315-2327.

Ehrenfeld, J. G., B. Ravit, K. Elgersma. 2005. Feedbacks in the plant-soil system. Annual
Review of Environment and Resources 30: _75-1 15.

Feeny, P. 1976. Plant apparency and chemical defenses. Recent Advances in
Phytochemistry 10:1—40.

Fie, S. L., and Steiner K. C. 2007. Evidence for increasing red maple abundance in the
eastern United States. Forest Science 53(4): 473-477.

F inzi, A. C., N. Van Breemen, and C. D. Canham. 1998. Canopy tree soil interactions

within temperate forests: Species effects on soil carbon and nitrogen. Ecological
Applications 8: 440-446.

Finzi, A. C., C. D. Canham, and N. Van Breemen. 1998. Canopy tree soil interactions

within temperate forests: Species effects on pH and cations. Ecological
Applications 8: 447-454.

Forget, P. M., J. E. Lambert, P. E. Hulrne, and S. Vander Wall. 2004. Seed fate:
predation, dispersal and seedling establishment. CABI, Wallingford, UK.

Foster, S. A. 1986. On the adaptive value of large seeds for tropical moist forest trees: a
review and synthesis. Botanical Review 52(3):260—299.

Fragoso, J. 1997 . Tapir- generated seed shadows: scale-dependent patchiness in the
Amazon rain forest. Jom'nal of Ecology 85:519-529.

Garnage, H., B. M. P Singhakumara, and M. S. Ashton. 2004. Effects of light and
fertilization on arbuscular mycorrhizal colonization and growth of tropical rain-
forest Syzygium tree seedlings. Journal of Tropical Ecology 20: 525-534.

Gause, G. F. 1934. The struggle for existence. Lippincott Williams and Wilkins,
Baltimore, Maryland, USA.

174

Gehring, C. 2003. Grth responses to arbuscular mycorrhizae by rain forest seedlings
vary with light intensity and tree species. Plant Ecology 167: 127—139.

Gehring, C. 2004. Seed reserves and light intensity affect the growth and mycorrhiza
development of the seedlings of an Australian rain-forest tree. Journal of Tropical
Ecology 20: 345-349.

Gilbert, G. S. 1995. Rain forest plant diseases: the canopy-understory connection.
Selbyana 16: 75-77.

Gilbert, G. S. 2002. Evolutionary ecology of plant diseases in natural systems. Annual
Review of Phytopathology 40:13—43.

Gilbert, G., S. P. Hubbell, and R. Foster, R. 1994. Density and distance-to-adult effects of
a canker disease of trees in a moist tropical forest. Oecologia 98: 100-108. ,

Gilbert, G. S., and S. P. Hubbell. 1996. Plant diseases and the conservation of tropical
forests. BioScience 46:98—106.

Gilbert, G., K. Harms, D. Hammill, and S. P. Hubbell. 2001.‘Effects of seedling size, E1
Nifio drought, seedling density, and distance to nearest conspecific adult on 6-year
survival of 0cotea whitei seedlings in Panama Oecologia 127(4): 509-516.

Guariguata, M. R. 2000. Seed and seedling ecology of tree species in neotropical
secondary forests: Management implications. Ecological Applications 10(1): 145—
1 54.

Gurevitch, J ., and L. V. Hedges. 1993. Meta-analysis: combining the results of
independent experiments. Pages 378—425 in S. Scheiner and J. Gurevitch, editors.
Design and analysis of ecological experiments. Chapman and Hall, London, UK.

Hammond, D.S. and V. K. Brown. 1998. Disturbance, phonology and life-history
characteristics: factors influencing distance/density-dependent attack on tropical
seeds and seedlings. Pages 51-78 in: D. M. Newberry, H. H. T. Prins and N. D.
Brown, editors. Dynamics of Tropical Communities. Blackwell Scientific
Publications, Oxford, UK.

175

Harms, K. E., S. J. Wright, 0. Calderon, A. Hernandez, and E. A. Herre. 2000. Pervasive
density-dependent recruitment enhances seedling diversity in a tropical forest.
Nature 404: 493—495.

Hartshom, G. S., and B. E. Hammell. 1994. Vegetation types and floristic patterns. Pages
73—89 in L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G. S. Hartshom,

editors. La Selva: ecology and natural history of a neotropical rain forest.
University of Chicago Press, Chicago, Illinois, USA.

He, F., P. Legendre, and J. V. LaFrankie. 1997. Distribution patterns of tree species in a
Malaysian tropical rain forest. Journal of Vegetation Science 8(1): 105—1 14.

Herrns, D. A., and W. J. Mattson. 1992. The dilemma of plants: to grow or defend.
Quarterly Review of Biology 67(3):283—335.

Hille Ris Lambers, J ., J. S. Clark, and B. Beckage. 2002. Density-dependent mortality
and the latitudinal gradient in species diversity. Nature 417: 732-735.

Hille Ris Lambers, J ., and J. S. Clark. 2003. Effects of dispersal, shrubs, and density-
dependent mortality on seed and seedling distributions in temperate forests. Canadian
Journal of Forest Research 33:783-795.

Hood, L. A., M. D. Swaine, and P. A. Mason. 2004. The influence of spatial patterns of
damping-off disease and arbuscular mycorrhizal colonization on tree seedling
establishment in Ghanaian tropical forest soils. Journal of Ecology 92:816—823.

Howe, H. F., E. W. Schupp and L. C. Wesley. 1985. Early consequences of seed
dispersal for a neotropical tree (V irola-Surinamensis). Ecology 66(3): 781-791.

Hubbell, SP. 1979. Tree dispersion, abundance, and diversity in a tropical dry forest.
Science 203: 1299-1309.

Hubbell, SP. 1980. Seed predation and the coexistence of tree species in tropical forests.
Oikos 35: 214-229.

176

Hubbell, SP. and R. B. Foster. 1986. Biology, chance, and history and the structure of
tropical rain forest tree communities. Pages 314-329 in: J. Diamond, and T. J.
Case, editors. Community Ecology. Harper and Row, New York, NY, USA.

Hubbell, S.P., R. Condit, and R. B. Foster. 1990. Presence and absence of density
dependence in a neotropical tree community. Philosophical Transactions of the
Royal Society of London B 330: 269—281.

Hubbell, S. P., J. A. Ahumada, R. Condit, and R. B. Foster. 2001. Local neighborhood
effects on long-term survival of individual trees in a neotropical forest. Ecological
Research 16(5):859—875.

Hyatt, L., M. Rosenberg, T. Howard, G. Bole, W. Fang, J. Anastasia, K. Brown, R.
Grelia, K. Himman, J. Kurdziel, and J. Gurevitch. 2003. The distance dependence
prediction of the J anzen-Connell hypothesis: a meta-analysis. Oikos 103:590-602.

Itoh, A., T. Yamakura, K. Oginao, H. S. Lee, and P. S. Ashton. 1997 . Spatial distribution
patterns of two predominant emergent trees in a tropical rainforest in Sarawak,
Malaysia. Plant Ecology 132: 121-136.

J anzen, D. H. 1970. Herbivores and the number of tree species in tropical forests.
American Naturalist 104:501—528.

Kardol, P., N. J. Cornips, M. M. L. van Kempen, J. M. T. Bakx-Schotman, and W. H. van
der Putten. 2007. Microbe-mediated plant-soil feedback causes historical

contingency effects in plant community assembly. Ecological Monographs 77(2):
147-162.

Kitaj ima, K. 1994. Relative importance of photosynthetic traits and allocation patterns as
correlates of seedling shade tolerance of 13 tropical trees. Oecologia 98: 419-428.

Kitajima, K., and C. K. Augspurger 1989. Seed and seedling ecology of a monocarpic
tropical tree. Ecology 70 (4): 1102-1114.

Klironomos, J. N. 2002. Feedback with soil biota contributes to plant rarity and
invasiveness in communities. Nature 417:67—70.

177

Kobe, R. K. 1997 . Carbohydrate allocation to storage as a basis of interspecific variation
in sapling survivorship and growth. Oikos 80(2):226—233.

Kobe, R. K. 1999. Light gradient partitioning among tropical tree species through
differential seedling mortality and growth. Ecology 80: 1 87—201.

Kobe, R. K., S. W. Pacala, and J. A. Silander, Jr. 1995. Juvenile tree survivorship as a
component of shade tolerance. Ecological Applications 5:517—532.

Kobe, R. K., G. E. Likens, and C. Eagar. 2002. Tree seedling growth and mortality
responses to manipulations of calcium and aluminum in a northern hardwood
forest. Canadian Journal of Forest Research 32(6): 954-966.

Kobe, R. K. 2006. Sapling growth as a function of light and landscape-level variation in
soil water and foliar nitrogen in northern Michigan. Oecologia 9(49): 119-133.

Kraus, T. E. C., R. A. Dahlgren, and R. J. Zasoski. 2003. Tannins in nutrient dynamics of
forest ecosystems — a review. Plant and Soil 256: 41-66.

Kwit, C., D. Levey, and C. Greenberg. 2004. Contagious seed dispersal beneath
heterospecific fi'uiting trees and its consequences. Oikos 107: 303-308.

Langenheim J. H. and W. H. Stubblebine. 1983. Variation in leaf resin composition
between parent tree and progeny in Hymenea — Implications for herbivory in the
humid tropics. Biochemical Systematics and Ecology 11(2): 97-106.

Lovelock, C. and R. Miller. 2002. Heterogeneity in inoculum potential and effectiveness
of arbuscular mycorrhizal fungi. Ecology 83(3): 823-832.

Makkar, H. P. S., R. K. Dawra, and B. Singh.1988. Determination of both tannin and
protein in a tannin-protein complex. Journal of Agricultural and Food Chemistry
36: 523—525.

178

Masaki, T., and T. Nakashizuka. 2002. Seedling demography of Swida controversa:
effect of light and distance to conspecifics. Ecology 83(12): 3497-3507.

Mills, K. E., and J. D. Bever. 1998. Maintenance of diversity within plant communities:
soil pathogens as agents of negative feedback. Ecology 79:1595—1601.

Montgomery, R. A., and R. L. Chazdon. 2002. Light gradient partitioning tropical tree
seedlings in the absence of canopy gaps. Oecologia 131 :165—174.

Myers, J. A., and K. Kitaj ima. 2007. Carbohydrate storage enhances seedling shade
tolerance in a neotropical forest. Journal of Ecology 95(2):383—395.

Newsham, K. K., A. H. Fitter, and A. R. Watkinson. 1995. Arbuscular mycorrhiza protect
an annual grass from root pathogenic fungi in the field. Journal of Ecology 83:
991-1000.

Nichols, J ., V. Agyeman, F. Agurgo, M. Wagner, and J. Cobbinah. 1999. Patterns of
seedling survival in the tropical African tree Milicia excelsa. Journal of Tropical
Ecology 15: 451-461.

Niinemets, U. and O.Kull. 1998. Stoichiometry of foliar carbOn constituents varies along
light gradients in temperate woody canopies: implications for foliage
morphological plasticity. Tree Physiology 18(7):467-479.

Norghauer, J. M., J. R. Malcolm, and B. L. Zimmerman. 2006. Juvenile mortality and
attacks by a specialist herbivore increase with conspecific adult basal area of

Amazonian Swietenia macrophylla (Meliadea). J oumal of Tropical Ecology
22 :45 1-460.

O'Hanlon-Manners, D. L., and P. M. Kotanen. 2004. Evidence that fungal pathogens
inhibit recruitment of a shade-intolerant tree, white birch (Betula papyrifera), in
understory habitats. Oecologia 140: 650-653.

O'Hanlon-Manners, D. L., and P. M. Kotanen. 2006. Losses of seeds of temperate trees to
soil fungi: effects of habitat and host ecology. Plant Ecology 187: 49-58.

179

Pacala S. W., C. D. Canham, J .A. Silander and R. K. Kobe. 1994. Sapling growth as a
fiinction of resources in a north temperate forest. Canadian Journal of Forest
Research 24(11): 2172-2183.

Pacala, S. W., C. D Canham, J. Saponara, J. A. Silander, Jr., R. K. Kobe, and E. Ribbens.
1996. Forest models defined by field measurements: estimation, error analysis and
dynamics. Ecological Monographs 66:1—43.

Packer, A., and K. Clay. 2000. Soil pathogens and spatial patterns of seedling mortality in
a temperate tree. Nature 404: 278—281.

Packer, A., and K. Clay. 2003. Soil pathogens and Prunus sertotina seedling and sapling
growth near conspecific trees. Ecology 84: 108—1 19.

Packer, A., and K. Clay. 2004. Development of negative feedback during successive
growth cycles of black cherry. Proceedings of the Royal Society 271 :317—324.

Palmer, M. W. 1994. Variation in species richness: towards a unification of hypotheses.
Folia Geobot Phytotaxon 29:511-530.

Pearson, T. R. H., D. F. R. P. Burslem, R. E. Goeriz, and J. W. Dalling. 2003.
Regeneration niche partitioning in neotropical pioneers: effects of gap size,
seasonal drought and herbivory on growth and survival. Oecologia 13(3):456—
465.

Penfold, G., and D. Lamb. 1999. Species co-existence in an Australian subtropical rain
forest: evidence for compensatory mortality. Journal of Ecology 87: 316-329.

Peters, H. A. 2003. Neighbor-regulated mortality: the influence of positive and negative
density dependence on tree populations in species-rich tropical forests. Ecology
Letters 6:757—765.

Pi got, A. L. and S. R. Leather. 2008. Invertebrate predators drive distance-dependent
patterns of seedling mortality in a temperate tree Acer pseudoplatanus. Oikos
117(4):521-530.

180

Poorter, L., and S. A. Rose. 2005. Light-dependent changes in the relationship between
seed mass and seedling traits: a meta-analysis for rain forest tree species.
Oecologia 142:378-387.

Reich, P. B., M. G. Tjoelker, M. B. Walters, D. W. Vanerklein, and C. Bushena. 1998.
Close association of RGR, leaf and root morphology, seed mass and shade

tolerance in seedlings of nine boreal tree species grown in high and low light.
Functional Ecology 12(3):327—338.

Reinhart, K. O., A. A. Royo, W. H. van der Putten, and K. Clay. 2005. Soil feedback and
pathogen activity in Prunus serotina throughout its native range. Journal of
Ecology 93(5):890—898.

Rhoades, D. F., and R. G. Cates. 1976. Toward a general theory of plant antiherbivore
chemistry. Recent Advances in Phytochemistry 10: 168—213.

Rillig, M., S. Wright, and V. T. Eviner. 2002. The role of arbuscular mycorrhizal fungi

and glomalin in soil aggregration: comparing effects of five plant species. Plant
and Soil 238: 325-333.

Sanchez-Hidalgo, M.E., M. Martinez-Ramos, F. J. Espinosa-Garcia. 1999. Chemical
differentiation between leaves of seedlings and spatially close adult trees from the
tropical rain-forest species Nectandra ambigens (Lauraceae): an alternative test of
the J anzen-Connell model. Functional Ecology 13: 725-732.

Schreeg, L. A., R. K. Kobe, and M. B. Walters. 2005. Tree seedling growth, survival, and
morphology in response to landscape-level variation in soil resource availability
in northern Michigan. Canadian Journal of Forest Research 35 (2): 263-273.

Schupp, E. 1988. Seed and early seedling predation in the forest understory and in treefall
gaps. Oikos 51: 71-78.

Seiwa K. 1998. Advantages of early germination for growth and survival of seedlings of
Acer mono under different overstory phenologies in deciduous broad-leaved
forests. Journal of Ecology 86:219—228.

181

Shibata, M. and T. Nakashizuka. 1995. Seed and seedling demography of four co-
occurring Carpinus species in a temperate deciduous forest. Ecology 76(4): 1099-
1 108.

Silva Matos, D., R. Freckleton, and A. Watkinson. 1999. The role of density dependence
in the population dynamics of a tropical palm. Ecology 80(8): 2635-2650.

Siqueira, J ., M. Carneiro, N. Curi, S. Rosado, A. David. 1998. Mycorrhizal colonization
and mycotrophic growth of native woody species as related to successional
groups in Southeastern Brazil. Forest Ecology and Management 107: 241-252.

Sollins, P., F. M. Sancho, R. Mata, and R. L. Sanford. 1994. Soils and soil process
research. Pages 34—53 in L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G.
S. Hartshom, editors. La Selva: ecology and natural history of a neotropical rain
forest. University of Chicago Press, Chicago, Illinois, USA

Sullivan, J. 2003. Density-dependent shoot-borer herbivory increases the age of first
reproduction and mortality of neotropical tree saplings. Oecologia 136: 96-106.

Sylvia, D. M. 1994. Vesicular-arbuscular mycorrhizal fungi. Page 360 in R. Weaver, J.
Angle, P. BottomLey, D. Bezdicek, S. Smith, A. Tabatabai, and A. Wollum,
editors. Methods of soil analysis. Part 2, Microbiological and biochemical
properties. Soil Science Society of America, Madison, Wisconsin, USA.

Stinson, K. A., S. A. Campbell, J. R. Powell, B. E. Wolfe, R. M. Callaway, et a1. 2006.
Invasive Plant Suppresses the Growth of Native Tree Seedlings by Disrupting
Belowground Mutualisms. Public Library of Science - Biology 4(5): e140 DOI:
10.1371/iourn_al.nbio.0040140.

Tomita, M., Y. Hirabuki, and K. Seiwa. 2002. Post-dispersal changes in the spatial
distribution of Fagus crenata seeds. Ecology 83(6): 1560-1565.

van der Putten, W. H., C. Van Dijk, and B. A. M. Peters. 1993. Plant-specific soil-home
diseases contribute to succession in foredune vegetation. Nature 362:53—56.

182

von Alhnen, C., P. Morellato, and M. Pizo. 2004. Seed predation under high seed density
condition: the pahn Euterpe edulis in the Brazilian Atlantic Forest. Journal of
Tropical Ecology 20: 471-474.

Walter, M. B., E. L. Kruger, and P. B. Reich. 1993. Growth, biomass distribution and
C02 exchange of northern hardwood seedlings in high and low light:
relationships with successional status and shade tolerance. Oecologia 94:7 -l6.

Walters, M. B., and P. B. Reich. 1999. Low light carbon balance and shade tolerance in
the seedlings of woody plants: do winter deciduous and broad-leaves evergreen
species differ? New Phytologist 143(1):143—154.

Walters, M.B., and P. B. Reich. 2000. Seed size, nitrogen supply, and growth rate affect
tree seedling survival in deep shade. Ecology. 81: 1887-1901.

Webb, C. O., and D. R. Peart. 1999. Seedling density dependence promotes coexistence
of Bomean rain forest trees. Ecology 80:2006—2017.

Welden, C. W., S. W. Hewett, S. P. Hubbell, and R. B. Foster. 1991. Sapling survival,
growth, and recruitment: relationship to canopy height in a neotropical forest.
Ecology 72: 35-50.

Wills, C., R. Condit, R. B. Foster, and S. P. Hubbell. 1997. Strong density- and diversity-
related effects help to maintain tree species diversity in a neotropical forest.
Proceedings of the National Academy of Sciences (USA) 94:1252—1257.

Wills, C., and R. Condit. 1999. Similar non-random processes maintain diversity in two
tropical rainforests. Proceedings of the Royal Society of London B 266: 1445-
1452.

Wills, C., R. Condit, S. P. Hubbell, R. B. Foster, and N. Manokaran. 2004. Comparable
nonrandom forces act to maintain diversity in both a new world and an old world
rainforest plot. Pages 384—407 in E. C. Losos and E. G. Leigh, Jr, editors.
Tropical forest diversity and dynamism. The University of Chicago Press,
Chicago, USA.

183

Wills, C., et a1. 2006. Nonrandom processes maintain diversity in tropical forests. Science
311:527—531.

Wright, S. J. 1983. The dispersion of eggs by a bruchid beetle beetle among Scheelea
palm seeds and the effect of distance to the parent palm. Ecology 64(5): 1016-
1 021 .

Wright, S. J. 2002. Plant diversity in tropical forests: a review of mechanisms of species
coexistence. Oecologia. 130: 1-14.

Wyatt, J. and M. Sihnan. 2004. Distance-dependence in two Amazonian pahns: effects of
spatial and temporal variation in seed predator communities. Oecologia 140: 26-35.

Zangaro, W., S. M. A. Nisizaki, J. C. B Domingos, and E. M. Nakano. 2003. Mycorrhizal
response and successional status in 80 woody species from south Brazil. Journal
of Tropical Ecology 19: 315-324.

184

 

 

 

 

a

'v’ERSlTV LIBRARIE

U
E
T
A
1

AN 5

HIGHW-

11;

“ll

”11

l
l

llll’llilllll
1293

ll

02956