Citation
Seedling Carbohydrate Storage, Survival, and Stress Tolerance in a Neotropical Forest

Material Information

Title:
Seedling Carbohydrate Storage, Survival, and Stress Tolerance in a Neotropical Forest
Creator:
MYERS, JONATHAN ANDREW ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Biomass ( jstor )
Cotyledons ( jstor )
Defoliation ( jstor )
Ecology ( jstor )
Pretreatment ( jstor )
Seedlings ( jstor )
Species ( jstor )
Trees ( jstor )
Tropical forests ( jstor )
Understory ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Jonathan Andrew Myers. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/1/2005
Resource Identifier:
436098659 ( OCLC )

Downloads

This item has the following downloads:

myers_j ( .pdf )

myers_j_Page_43.txt

myers_j_Page_27.txt

myers_j_Page_53.txt

myers_j_Page_12.txt

myers_j_Page_04.txt

myers_j_Page_19.txt

myers_j_Page_11.txt

myers_j_Page_60.txt

myers_j_Page_50.txt

myers_j_Page_02.txt

myers_j_Page_36.txt

myers_j_Page_16.txt

myers_j_Page_70.txt

myers_j_Page_08.txt

myers_j_Page_17.txt

myers_j_pdf.txt

myers_j_Page_69.txt

myers_j_Page_31.txt

myers_j_Page_62.txt

myers_j_Page_39.txt

myers_j_Page_46.txt

myers_j_Page_57.txt

myers_j_Page_71.txt

myers_j_Page_01.txt

myers_j_Page_09.txt

myers_j_Page_33.txt

myers_j_Page_25.txt

myers_j_Page_03.txt

myers_j_Page_38.txt

myers_j_Page_34.txt

myers_j_Page_15.txt

myers_j_Page_45.txt

myers_j_Page_41.txt

myers_j_Page_23.txt

myers_j_Page_07.txt

myers_j_Page_40.txt

myers_j_Page_68.txt

myers_j_Page_72.txt

myers_j_Page_35.txt

myers_j_Page_54.txt

myers_j_Page_14.txt

myers_j_Page_29.txt

myers_j_Page_63.txt

myers_j_Page_10.txt

myers_j_Page_64.txt

myers_j_Page_05.txt

myers_j_Page_66.txt

myers_j_Page_20.txt

myers_j_Page_61.txt

myers_j_Page_73.txt

myers_j_Page_13.txt

myers_j_Page_06.txt

myers_j_Page_24.txt

myers_j_Page_59.txt

myers_j_Page_74.txt

myers_j_Page_51.txt

myers_j_Page_65.txt

myers_j_Page_22.txt

myers_j_Page_44.txt

myers_j_Page_55.txt

myers_j_Page_47.txt

myers_j_Page_32.txt

myers_j_Page_42.txt

myers_j_Page_37.txt

myers_j_Page_49.txt

myers_j_Page_48.txt

myers_j_Page_58.txt

myers_j_Page_52.txt

myers_j_Page_56.txt

myers_j_Page_21.txt

myers_j_Page_67.txt

myers_j_Page_30.txt

myers_j_Page_18.txt

myers_j_Page_28.txt

myers_j_Page_26.txt


Full Text












SEEDLING CARBOHYDRATE STORAGE, SURVIVAL, AND STRESS
TOLERANCE IN A NEOTROPICAL FOREST
















By

JONATHAN ANDREW MYERS


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Jonathan Andrew Myers

































I dedicate this thesis to my parents, Roy and Susan Myers, whose support and
encouragement of my pursuits in ecology know no bounds.















ACKNOWLEDGMENTS

I would first like to thank my advisor, Kaoru Kitajima, for her unwavering

mentorship in both academia and field research. Her passion for seedling ecology

and dedication to strong work ethic is constantly inspiring. I would also like to thank my

committee member, Jack Putz, for his critical and insightful suggestions on all aspects of

the thesis and for his overall guidance in my development as an ecologist.

Numerous individuals have generously contributed field and laboratory assistance.

I would like to thank Marta Vargas for her assistance with data collection throughout the

entirety of the project, Roberto Cordero for his help and advice during the early stages of

project development, Sebastian Bernal for his spirited assistance in the field and

laboratory (and for keeping us dry and motivated in even the heaviest of rains), and

Momoka Yao for her help with many of the laboratory measurements and analyses. I

would also like to thank the following undergraduate students for assistance with data

collection: Sarah Tarrant, Lisa Cowart, and Sergio Montes. Laboratory assistance was

also provided by Zachary Tyser, who participated in the project through the 46th Annual

Pre-collegiate Summer Science Training Program at the University of Florida.

Finally, I would like to thank the many friends and colleagues who made my time

at the University of Florida more than memorable. I would especially like to thank

Teresa Kurtz, Silvia Alvarez, Camila Pizano, Jennifer Schaefer, Joseph Veldman, and

Eddie Watkins for constantly reminding me that the beauty of life extends well beyond

science.
















TABLE OF CONTENTS

page

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

LIST OF TABLES ....................................................... ............ .............. .. vii

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

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

CHAPTER

1 STRESS TOLERANCE ENHANCES SEEDLING SURVIVAL IN THE
UNDERSTORY OF A TROPICAL FOREST: EXPERIMENTAL EVIDENCE .......1

In tro du ctio n ...................................... ............................. ..... ......... ...... .
M eth od s ......................................................................... . 3
Study Site and Species................ ............. ...... ...... .............. ........... .. 3
Field and Laboratory M ethods ........................................ ......................... 4
Statistical A n aly ses........... ............................................................ ........ .. ....
R e su lts ........................................................................ ........................ 9
Seedling Survival .................. .................. .... ...... ........ ................ 9
Effects of Defoliation and Light Reduction on Seedling Biomass and Growth..10
D iscu ssion .......................................... .. .... .. .............. ......................... 12
Stress Tolerance as a Component of Seedling Survival in Shade....................12
Seedling Tolerance to Defoliation: The Importance of Storage Reserves ..........13
Seedling Tolerance to Temporal Variation in Understory Light Availability ....14

2 CARBON ALLOCATION TO STORAGE IN TROPICAL FOREST SEEDLINGS:
EFFECTS ON GROWTH, SURVIVAL AND STRESS TOLERANCE ..................24

Introdu action ...................................... ................................................. 24
M eth o d s ..............................................................2 7
Study Site and Species............ .... ........................................ .... .. ........... 27
Field and Laboratory M ethods ........................................ ........................ 28
Quantification of TNC in Stems and Roots...................................................31
Statistical A n aly ses........... ................................................ .......... ..... .... .... 32









R e su lts ............... ...... .. .... ...... ... ...... ......................................... 3 4
Variation in Pretreatment Biomass and TNC Reserves ......................................34
Effects of Defoliation and Deep Shade on Biomass and TNC Reserves ............35
Influence of Seedling Size, Growth, and TNC Reserves on Seedling Survival..36
D iscu ssion ........................................ ........... ..... .... ................................. .38
Role of TNC Reserves in Seedling Survival and Stress Tolerance...................38
Tradeoffs between Seedling Growth and Storage.............................................40
TNC Allocation, Seed Size, and Seedling Functional Morphology ..................41
C o n clu sio n .................................................. ................ 4 2

APPENDIX

A SUMMARY STATISTICS FOR SEEDLING BIOMASS AND LEAF AREA........ 52

B SUMMARY STATISTICS FOR TOTAL NONSTRUCTURAL
CARBOHYDRATE (TNC) RESERVES IN STEMS AND ROOTS......................55

L IST O F R E F E R E N C E S ........................................................................ .....................57

BIO GRAPH ICAL SK ETCH .................................................. ............................... 63
















LIST OF TABLES


Table pge

1-1 Seed and seedling characteristics of the study species...........................................17

1-2 Proportional hazards (Cox regression) survival time analysis of seedling
survival over 1 year................... ............................ .. ........ .. ............ 18

1-3 Analysis of variance for total seedling biomass................... ................... ...............18

2-1 Analysis of variance for stem and root biomass. ................ ................ ..............43

2-2 Analysis of variance for TNC concentrations and pool sizes ...............................44

A-i Statistics for pre- and post-treatment biomass and leaf area..................................52

A-2 Statistics for total seedling biomass and leaf area 1 year after treatment ................54

B-l Statistics for pre- and post-treatment TNC concentrations and pool sizes..............55
















LIST OF FIGURES


Figure pge

1-1 Effects of light reduction and defoliation on 1st-year seedling survival ................. 19

1-2 Two-month survival of stress treatment seedlings plotted against 1st-year
survival of control treatm ent seedlings.. ....................................... ............... 20

1-3 Effects of light reduction and defoliation on seedling size and growth 2 months
after treatm ent.. ........................................................................2 1

1-4 Leaf area recovery 2 m months after defoliation.. ........................................ ............22

1-5 Effects of light reduction and defoliation on relative growth rate (RGR) from 2
months to 1 year. ............. .. ............ ............................. ....... 23

2-1 Pretreatment biomass and TNC concentrations......................................................45

2-2 Three potential correlates of interspecific variation in lst-year control seedling
survival .................................. .................................... ........... 46

2-3 Pretreatment TNC pool sizes in stems and roots.................... ........................... 47

2-4 Post-treatment biomass and TNC in stems and roots.............................................48

2-5 Post-treatment cotyledon biomass.................................................. ............... 49

2-6 Two-month relative growth rate (RGR) of control seedlings plotted against
pretreatm ent TN C pool size.. ............................. ..............................................50

2-7 Survival of stressed seedlings plotted against total TNC pool size and total
seedling biom ass.. ...................................................................... 5 1















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

SEEDLING CARBOHYDRATE STORAGE, SURVIVAL, AND STRESS
TOLERANCE IN A NEOTROPICAL FOREST

By

Jonathan Andrew Myers

May 2005

Chair: Kaoru Kitajima
Major Department: Botany

I investigated the role of stem and root total nonstructural carbohydrate (TNC)

reserves in seedling growth, survival, and stress tolerance in the shaded understory of a

tropical moist forest in central Panama. Seven woody species that ranged widely in

seedling shade tolerance were selected for study: Aspidosperma cruenta and Lacmelia

panamensis (Apocynaceae), Coussarea curvigemnia (Rubiaceae), Callichlamys latifolia

and Tabebuia rosea (Bignoniaceae), Castilla elastica (Moraceae), and Platypodium

elegans (Fabaceae). Seedlings of each species were raised from seed in a shade house,

transferred to common-gardens enclosed by wire fencing in the shaded forest understory

(ca. 1% full sun), and then randomly assigned to one of three experimental treatments:

extreme light reduction (ca. 0.06% of full sun for 2 months), complete defoliation at 0

months, and control (no manipulation). The goal of the light-reduction and defoliation

stress treatments was to force seedlings into negative carbon balance, to determine the

role of TNC reserves for seedling survival over the period of 1 year.









First-year survival of control seedlings-which I used as a quantitative index of

shade tolerance-varied significantly among species, ranging from 96% in Aspidosperma

to 19% in Tabebuia. The light-reduction and defoliation treatments caused significant

decreases in seedling survival. Two-month survival after the stress treatments was

positively correlated with 1st-year control seedling survival among species. Seedlings of

the two species with lowest 1st-year control seedling survival (Tabebuia and

Platypodium) were unable to survive either stress treatment past 60 days.

Pretreatment TNC concentrations and pool sizes in stems and roots differed

significantly among species. Stem concentrations were higher than root concentrations in

all species, and comprised between 6 to 25% of the total stem biomass. Variation in

1st-year control seedling survival was positively related to pretreatment TNC pool size

(combined stem and root pools) among species; but not to pretreatment stem or root TNC

concentrations, total seedling biomass, or cotyledon biomass. Survival of the stress

treatment seedlings during the first 2 months (and from 2 months to 1 year) was also

positively related to interspecific variation in TNC pool size. Relative growth rates of

control seedlings from 0 to 2 months were negatively related to TNC pool size among

species, suggesting a tradeoff between carbon allocation to growth versus storage.

My results highlight the importance of TNC reserves for seedling survival and

stress tolerance in the shaded understory of tropical forests. Allocation-based tradeoffs

between seedling growth and storage provide a mechanistic basis for species differences

in shade tolerance; and have likely played important roles in the evolution of life-history

diversity and maintenance of tree-species diversity in species-rich tropical tree

communities.














CHAPTER 1
STRESS TOLERANCE ENHANCES SEEDLING SURVIVAL IN THE
UNDERSTORY OF A TROPICAL FOREST: EXPERIMENTAL EVIDENCE

Introduction

Differential seedling survival among coexisting species has long been considered to

play important roles in determining tree-species composition and diversity in tropical

forests (Janzen 1970; Connell 1971; Hubbell 1979; Harms et al. 2000; Wright 2002).

Seedling mortality is especially high in the shaded forest understory (Augspurger

1984a,b; Osunkoya et al. 1992; Li et al. 1996) where light availability is low (ca. 1% of

full sun) and variable in time and space (Chazdon and Fetcher 1984; Montgomery and

Chazdon 2002; Montgomery 2004). Under these conditions, low potential for

photosynthetic carbon gain limits the ability of seedlings to maintain positive carbon

balance, a prerequisite for survival in shade. Diurnal and seasonal fluctuations in

understory light availability (Chazdon and Fetcher 1984; Chazdon and Pearcy 1986)

and/or deep shading of seedlings by fallen litter or understory foliage (Vazquez-Yanes et

al. 1990; Wang and Augspurger 2004; Farris-Lopez et al. 2004; Montgomery 2004)

further constrain carbon balance by forcing seedlings to cope with stressful periods, when

light levels drops below the photosynthetic compensation point. In addition to

light-limitation, seedlings risk mortality when tissues are lost or damaged by disease

(Augspurger 1984a), herbivores (Osunkoya et al. 1992; Asquith et al. 1997; Kitajima

2004), and physical disturbance (Clark and Clark 1989; Scariot 2000). Identifying the

extent and factors that contribute to species differences in their ability to cope with









stresses due to light limitation and tissue loss is essential for predicting and evaluating

community-level patterns of seedling recruitment and relative species abundance.

Survival in light-limited habitats (shade tolerance) is contingent on the ability of

seedlings to maintain positive carbon balance. Some researchers have hypothesized that

shade tolerance is achieved by morphological and physiological traits that maximize the

rate of carbon capture in low light at the whole-plant level (Loach 1967; Givnish 1988;

Woodward 1990; Walters and Reich 1996). According to this hypothesis, fast rates of

net carbon gain facilitate both high seedling growth rates and enhanced survival in shade,

by allowing seedlings to quickly increase in size and out-compete slower growing

neighbors (Walters and Reich 2000).

In contrast, others researchers have hypothesized that fast growth in shade could

have deleterious consequences on survival, because high carbon allocation to growth

must occur at the expense of allocation to storage and defense (Kitajima 1994, 1996;

Coley et al. 1985; Kobe et al. 1995; Kobe 1997; Veneklaas and Poorter 1999), thereby

reducing seedling tolerance and/or resistance to biotic and abiotic stresses. However, few

studies have directly tested whether interspecific variation in seedling survival and

growth in the shaded forest understory is linked to species-specific differences in seedling

stress tolerance (e.g., Augspurger 1984a). Quantifying the degree to which different

species are able to tolerate and recover from abiotic and biotic stresses (especially those

posed by light-limitation and tissue loss) could help to clarify the functional traits

underlying seedling shade tolerance and life history diversity in the seedling community

of species-rich tropical forests.









In this study, I examined seedling stress tolerance, growth, and survival among

seven coexisting woody species in a tropical moist forest in central Panama. The

overarching goal of my study was to test whether interspecific variation in seedling shade

tolerance, measured quantitatively as 1st-year seedling survival in the shaded understory

(see also Augspurger 1984a,b; Kitajima 1994; Boot 1996), was linked to variation in

survival and growth responses to stresses caused from light limitation and tissue loss. I

tested the following two hypotheses: (1) interspecific variation in lst-year seedling

survival in the shaded understory is positively correlated with the ability to survive stress

due to light limitation and tissue loss; and (2) seedling growth is less affected by stress in

species with higher lst-year seedling survival. To test these hypotheses, I conducted a

1-year common-garden field experiment in the shaded understory (- 1% full sun), using

transplanted seedlings of an equivalent ontogenetic stage, defined by expansion of the

first photosynthetic organs (leaves or cotyledons). Growth and survival were monitored

on seedlings assigned to one of three treatments: (1) extreme light reduction; (2)

complete defoliation; (3) control (no manipulation). The goal of the light-reduction

treatment was to expose seedlings to light levels below those required for whole-plant

carbon balance, while the purpose of the defoliation treatment was to force seedlings to

rely on storage reserves after removal of photosynthetic leaves and cotyledons.

Methods

Study Site and Species

The study was conducted on Buena Vista peninsula, an area of ca. 60-year-old

secondary lowland tropical forest, located in the Barro Colorado Nature Monument

(BCNM), Panama (9010' N, 7951' W). The BCNM forest is semi-deciduous, with a









pronounced 4-month dry season that usually lasts from mid-December to mid-April.

Annual rainfall on the BCNM averages 2,700 mm (Rand and Rand 1982).

Seven woody species with sufficient seed availability were chosen for study (Table

1-1). All of the species are canopy trees as adults, except for Callichlamys latifolia

(Bignoniaceae), which is a woody liana. Species were selected to span a range of shade

tolerance, which was quantified by estimating the percentage of seedlings surviving for 1

year in experimental plots located in the shaded understory (Table 1-1). Small-seeded,

pioneer species that are generally unable to establish seedling populations in the shaded

forest understory were deliberately excluded from the study.

The seven species differ in seed mass (24 to 453 mg) and cotyledon functional

morphology (Table 1-1). Aspidosperma cruenta, Lacmeliapanamensis, Castilla elastica,

and Platypodium elegans have hypogeal cotyledons that remain below or just above the

soil surface, and that function primarily for storage of seed reserves. The three other

species have epigeal cotyledons that become elevated above the ground after

germination; Coussarea curvigemnia and Tabebuia rosea have foliaceous photosynthetic

cotyledons, and Callichlamys latifolia has thick green cotyledons that function primarily

for storage (Kitajima 2002). For brevity, species are henceforth referred to by genus.

Field and Laboratory Methods

Depending on seed availability, 200 to 400 seedlings per species were raised from

seed in a shade house on the BCNM, under a light level similar to that found in the forest

understory (1.5% full sun, based on total daily photon flux density). When possible,

seeds were collected from at least 3 nonadjacent parent trees, to increase genetic diversity

within each species. Seedlings were raised in plastic trays containing a 1:1 mixture of

forest soil and sand, until all seedlings reached an equivalent ontogenetic stage, defined









by the full expansion of the first photosynthetic organs (leaves or cotyledons, depending

on the species; Table 1-1), at which time they were transplanted into the forest (described

below). Seedlings were transplanted at a standardized ontogenetic stage, rather than at a

common chronological age, so that all species had initiated autotrophic growth by the

onset of the study. Because of interspecific variability in timing of seed dispersal,

germination, and development, seedlings were transplanted over a 3-month period during

the early-mid rainy season (May to July).

Seedlings of all species were transplanted into four 7 x 7 m2 common-gardens

located in the shaded forest understory on Buena Vista Peninsula. Each garden was

separated by at least 50 m from the nearest adjacent garden, and enclosed by 1 m tall

fencing to exclude large herbivores. For each species, equal numbers of seedlings were

transplanted into 1 x 1 m2 cells in four equally sized, stratified sections in each garden

(50 to 100 total seedlings garden-1 species-1). All vegetation < 1.5 m tall was removed

from each garden before seedling transplantation, and all seedlings were positioned a

minimum of 25 cm apart. When necessary, leaf litter was also removed to avoid burial of

seedlings. Seedlings dying within the first 2 weeks after transplantation were replaced

(< 10 seedlings total).

Seedlings were randomly assigned to one of three treatments 2 weeks after

transplantation: (1) defoliation; (2) light reduction; (3) control (no manipulation). For the

defoliation treatment, all leaves and photosynthetic cotyledons (if present) were clipped

at the base of the petiole; cotyledons were retained on all of the species with hypogeal

storage cotyledons (Table 1-1). For the light-reduction treatment, cylindrical wire cages

covered with 90% shade cloth were placed over individual seedlings. Shade cage









dimensions ranged from 6 x 12 cm (height x diameter) to 15 x 30 cm (depending on

initial seedling size); and were constructed to be large enough so as not to hinder growth

over a 2-month period, after which the shade cages were removed.

Light transmission was measured inside and outside of the shade cages with Li-Cor

quantum sensors during the early-mid rainy season. Percent light transmission

underneath the cages, and within each common-garden, was determined using both

instantaneous and continuous total daily photon flux density (PFD) measurements.

Instantaneous PFD measurements were taken inside and outside of shade cages at 8 to 15

stratified positions in each garden under mid-day, cloudy sky conditions. Continuous

PFD was measured by placing a single censor in the middle of each common-garden for

2 days. To calculate % light transmission, measurements within the forest were

referenced to a sensor placed in completely open sky on a laboratory rooftop. Mean light

reduction ( 1 SD) by the shade cages was 89 ( 2)%. Mean light levels outside of the

shade cages, as determined from the instantaneous PFD measurements, ranged among

gardens from 0.38 to 0.68% (overall mean = 0.55%); mean light levels from the

continuous PFD measurements taken over 2 full days in each garden were higher, and

ranged among gardens from 0.68 to 1.44% of full sun (overall mean = 1%). Seedlings

inside the shade cages thus experienced only ca. 0.06% of full sun over the 2-month

light-reduction treatment.

To estimate pre- and post-treatment biomass, I harvested randomly selected

sub-samples of seedlings from each common-garden. For the pretreatment samples, 6 to

12 seedlings per species were harvested from each common-garden just prior to

treatment. Post-treatment samples were harvested 2 months (coinciding with the end of









the light-reduction treatment) and 1 year after treatment; samples sizes at these harvests

ranged from 2 to 32 total seedlings per species x treatment combination, depending on

species-specific differences in survival. Seedlings were harvested by carefully

excavating the root system with a small trowel. Roots, stems, leaves, and cotyledons

were separated within 12 hours after harvesting, and then oven dried at 100C for 1 hour.

All samples were then dried for an additional 48 hours at 60C before weighing. Leaf

and cotyledon area (for epigeal cotyledons only) were measured before drying using a

Li-Cor leaf area meter.

Seedling survival was measured by conducting censuses every week for the first 2

months, and then every 2 weeks from 2 months to 1 year. Missing seedlings were treated

as dead. Seedlings killed by branch fall and locally heavy herbivory (e.g., outbreak of

stem cutting insects in one common-garden) were excluded from the study.

Statistical Analyses

Differences in survival among species and treatments were analyzed using

semi-parametric proportional hazards modeling (i.e., Cox regression; Fox 2001). Cox

regression was selected over alternative parametric tests because it does not assume

a particular probability distribution for survival times (the non-parametric component of

the model; Fox 2001), which varied widely among species and treatments. Seedlings

harvested for biomass measurements and those that survived beyond the final survival

census were 'right-censored' in the analysis. I then used the Kaplan-Meier approach to

estimate the proportion of seedlings surviving over time for each species x treatment

combination. Differences between Kaplan-Meier estimates among species and among

treatments within species were analyzed using log-rank tests, which tests for

heterogeneity in survival times among groups by comparing observed versus expected









number of deaths between each mortality event (Fox 2001). Percent survival was

estimated from 0 to 2 months (i.e., the time of shade cage removal), 2 months to 1 year

(stress treatments only), and from 0 to 1 year (control treatment only) from the Kaplan-

Meier survival curves. Survival data were pooled across common-gardens for all

analyses and garden was included as a main factor in the Cox regression model.

Differences among species and treatments in total seedling biomass at 2 months

and 1 year were analyzed using ANOVA. In both analyses, common-garden means were

used as replicates for each species x treatment combination (n = 1-4 per species x

treatment combination; Appendix A). Because all seedlings had died before the

2-month harvest for some species x treatment combinations, I used two separate ANOVA

models to analyze seedling biomass. In the first model ("saturated model"), I tested for

species, treatment, and species x treatment interaction effects on biomass, but only

included data for the species that had living seedlings present in all treatments at the time

of harvest (n = 5 species at 2 months; n = 4 species at 1 year). In the second model

("Reduced main factor model"), I included data for all 7 species and tested for the main

effects of species and treatment on biomass, but excluded species x interaction terms.

Data from the 1-year harvest were loglo-transformed before analysis to improve

normality and homogeneity of variance.

Simple least squares linear regression was used to test for correlations between

interspecific variation in seedling survival, growth, and size. Relative growth rates

(RGR) of total seedling biomass, stem height, and leaf area were estimated using the

following formula:

RGR (wk-1) = [ln (X2) In (Xl)] / time









where X2 = mean size at time 2; X1 = mean size at time 1; and time = number of
weeks between times 1 and 2.

RGR for LACP from 0 to 2 months could not be calculated due to missing pretreatment

biomass samples. Leaf and cotyledon mass were included in all RGR calculations.

Cotyledon area was included in all leaf area calculations for species with photosynthetic

cotyledons (Coussarea and Tabebuia). All statistical analyses were performed using

JMP software (SAS Institute, 1997).

Results

Seedling Survival

Seedling survival differed substantially among species and treatments (Table 1-2;

Figure 1-1). First-year survival of control seedlings differed significantly among species

(P < 0.0001 for log-rank test of survival time distributions), and ranged from 19 to 96%

at the end of one year (Figure 1-1; Table 1-1). For control seedlings, survival from 0 to 1

year was positively correlated with survival from 0 to 2 months among species

(P < 0.0001, r2 = 0.96, n = 7), when mortality was generally highest for most species.

Thus, species maintained their survival ranks from 2 months to 1 year. Survival also

differed significantly among common-gardens (Table 1-2), and was generally higher in

gardens with higher light levels, even though the range in light levels among gardens was

small (0.6-1.4% full sun).

Defoliation and light reduction significantly reduced survival in all species

(Figure 1-1; P < 0.0001 for log rank tests), and there was a significant interaction

between species and treatment on survival times (Table 1-2). Species that had high

survival in the light-reduction treatment also had high survival in the defoliation

treatment (P = 0.0001, r2 = 0.94, n = 7 for proportional survival from 0 2 months). For









some species (Aspidosperma, Coussarea, Callichlamys) defoliation had a larger overall

impact on survival than light reduction, while others showed similar survival in the two

stress treatments (Figure 1-1).

The impact of the stress treatments on seedling survival varied widely among

species (Figure 1-1). Survival in the stress treatments from 0 to 2 months was positively

correlated with 1st-year control seedling survival among species (Figure 1-2). Two-

month survival for the three species with highest 1st-year survival exceeded 85% in both

stress treatments. In contrast, of the four species with the lowest 1st-year survival of

control seedlings, three had reached 100% seedling mortality in one or both of the stress

treatments before 60 days. There were no significant relationships between seedling

survival and seed mass or pretreatment cotyledon mass among species for any treatment

(P > 0.2 for all regressions), even though the species with the largest seed size

(Aspidosperma) had the highest survival in all treatments at 2 months, and the smallest-

seeded species (Tabebuia) the lowest survival after defoliation and the second lowest

survival after light reduction.

Effects of Defoliation and Light Reduction on Seedling Biomass and Growth

Total seedling biomass at 2 months differed significantly among species and

treatments (Figure 1-3a; Table 1-3). The effect of the stress treatments on seedling

biomass varied widely among species (Figure 1-3a) and there was a significant

interaction between species and treatment on biomass at 2 months, but not at 1 year

(Table 1-3). Whole-seedling relative growth rate (RGR) from 0 to 2 months also varied

widely among species and treatments (Figure 1-3b). Control seedling RGR ranged 2.5

fold between species and was highest for Tabebuia (which had the lowest 1st-year control

seedling survival) and lowest for Aspidosperma (highest 1 t-year survival), but there was









no significant correlation between control seedling survival and RGR among species (P >

0.4, r2 = 0.16, n = 6). Two-month survival of control seedlings was, however, negatively

correlated to both stem height and leaf area RGR among species (P = 0.01, r2 = 0.77 for

stem height; P = 0.03, r2 = 0.68 for logo leaf area).

Light reduction had a noticeable effect on RGR, which varied 5.5 fold among

species at the end of 2-month treatment (Figure 1-3b). RGR in the light-reduction

treatment was less affected (relative to the control treatment) for species with higher

1st-year control seedling survival: RGR was reduced by 0.013 to 0.015 wk-1 for the two

species with higher 1st-year survival (Aspidosperma and Coussarea) compared to 0.042

to 0.050 wk-1 in the three species with lower 1st-year survival (Callichlamys, Castilla, and

Tabebuia; no Platypodium seedlings survived the light-reduction treatment). As with

control seedlings, whole-seedling RGR was not a good correlate of survival among

species in the light-reduction treatment (P > 0.7, r2 = 0.04, n = 5). However, the three

species with the lowest survival exhibited the highest whole-seedling RGR in the control

treatment, and among species, survival in the light-reduction treatment was negatively

correlated to height RGR of control seedlings (P = 0.06, r2 = 0.59, n = 6). Whole-

seedling RGR during the light-reduction treatment was highest for the three smallest-

seeded species, all of which had epigeal cotyledons (Figure 1-3b; Table 1-1).

RGR following defoliation was greater for species that were able to produce new

leaves (Figure 1-3b,c). Four of the seven species, including the three species with highest

1st-year control seedling survival, were able to develop new and fully expanded leaves

within 2 months (Figure 1-3c). Furthermore, species with relatively higher 1st-year

survival produced more new leaf area following defoliation than was produced by control









seedlings over the same 2-month period (Figure 1-5): leaf production for Aspidosperma,

the species with the highest 1st-year survival, was six times greater than in controls. In

contrast, leaf production in Castilla, which had the lowest 1st-year survival among the

species that produced new leaves, was lower relative to controls. It is also worth noting

that leaf production following defoliation was not restricted to species with relatively

large seed mass or hypogeal storage cotyledons. Coussarea, despite having the second

smallest seed mass and photosynthetic epigeal cotyledons (Table 1-1), was able to

recover more leaf area relative to control seedlings following defoliation than Castilla

(Figure 1-5), which had three-fold higher seed mass and hypogeal cotyledons.

Stress treatments had significant negative effects on total seedling biomass at 1

year (Table 1-3). RGR from 2 months to 1 year varied widely among species and

treatments (Figure 1-6). The three species with the highest lst-year control seedling

survival (Aspidosperma, Lacmelia, Coussarea) experienced greater reduction in RGR by

stress relative to controls, while the species with the lowest lst-year survival

(Callichlamys, Castilla) showed overcompensation of RGR.

Discussion

Stress Tolerance as a Component of Seedling Survival in Shade

My results support the hypothesis that enhanced seedling survival in the shaded

understory is related to the ability of species to tolerate stress. Among species, survival

in both the defoliation and light-reduction treatments was strongly related to interspecific

variation in lst-year control seedlings survival (i.e., shade tolerance). The ability to

tolerate stress also varied widely among species: the two species with the lowest lst-year

survival reached 100% mortality in both stress treatments before 60 days, while the three

species with the highest 1st-year survival showed substantial tolerance to both treatments









(> 85% survival). The stress tolerance of the latter three species was considerable given

the severity of the experimental stresses imposed upon them: only 0.06% of full sun for 2

months or complete defoliation. These results support the notion that shade tolerance is

dependent upon functional traits that both allow seedlings to maintain a positive carbon

balance in low light, and to survive through and recover from periods when they

experience negative carbon balance due to stress.

Survival following the defoliation and light-reduction treatments was highly

correlated among species, indicating that the most stress tolerant species are able to cope

with a wide range of biotic and abiotic hazards. Earlier studies have also shown negative

interspecific correlations between 1st-year survival of tropical tree seedlings and

susceptibility to pathogens in shade (Augspurger and Kelly 1983; Augspurger 1984a,b).

In concert, these results suggest that the traits conferring stress tolerance may be broad

spectrum, potentially allowing seedlings to cope with other stresses caused by drought

(Engelbrecht and Kursar 2003; Khurana and Singh 2004), nutrient limitation (Gunatilleke

et al. 1997), and physical/mechanical damage due to litterfall (Clark and Clark 1989;

Scariot 2000).

Seedling Tolerance to Defoliation: The Importance of Storage Reserves

The goal of the defoliation treatment was to force seedlings to temporarily rely on

storage reserves for maintenance of a positive carbon balance and subsequent leaf tissue

recovery. My results show that post-defoliation seedling survival was dependent on both

the capacity of a species to develop and fully expand new leaves, and the relative degree

of leaf area recovery among species. Total leaf area and relative leaf area recovery was

most impacted in species with lower 1st-year control seedling survival, confirming my

initial hypothesis. The three species that were unable to develop new leaves did not









survive past 60 days, while survival among the other four species was highest for species

with greater leaf area recovery. For example, leaf area production in Aspidosperma, the

species with the highest post-defoliation survival over the first 2 months, was 64 times

greater than in Castilla (relative to leaf production in control seedlings), which had 50%

lower survival.

Recovery of photosynthetic tissue lost to herbivores, disease, or disturbance is

ultimately dependent on the amount of carbon reserves within remaining vegetative

tissues and storage cotyledons at the time of tissue loss. In this study, three of the four

species that were able to develop new leaves following defoliation had hypogeal storage

cotyledons, consistent with the idea that large seed reserves may enhance seedling

recovery following tissue damage (Harms and Dalling 1997; Green and Juniper 2004).

Leaf area recovery, however, was not restricted to species with this particular cotyledon

functional morphology. Coussarea, which had epigeal photosynthetic cotyledons, as

well as other epigeal-photosynthetic species in this forest, are also capable of substantial

leaf area recovery following complete defoliation (Kitajima 2004). For species with

epigeal cotyledons, storage reserves within stem and roots in the form of total

nonstructural carbohydrate (TNC), provide the carbon source necessary for tissue

recovery following damage (McPherson and Williams 1998; Canham et al. 1999;

Hoffman et al. 2004). TNC reserves in stems and roots may play particularly important

roles in seedling survival after seedlings have depleted seed reserves, or when storage

cotyledons are prematurely lost to herbivores or disease (Kitajima 2004; Chapter 2).

Seedling Tolerance to Temporal Variation in Understory Light Availability

The seven species varied widely in their ability to survive through drastic

reductions in light availability in the shaded understory. Three of the species showed









substantial tolerance to light reduction, with > 92% survival during the 2-month

treatment. Platypodium suffered the highest mortality during the first 2 months (all

seedlings died before 55 days). The low survival of this species was influenced by

damping-off disease (Augspurger 1983, 1984a,b), confirming the idea that fungal attack

can limit seedling establishment in deeply shaded habitats (e.g., Grime 1965). Reduction

in light availability also increased variability among species in seedling growth rates.

These results support the notion that small-scale heterogeneity in light availability, even

in the closed forest understory, can have dramatic impacts on seedling growth, survival,

and recruitment processes in tropical forests (Montgomery and Chazdon 2002; Wang and

Augspurger 2004; Montgomery 2004).

Interspecific variation in 2-month seedling survival in both the light-reduction and

control treatments was not related to whole-seedling RGR. However, species that

maintained high RGR in the control treatment also tended to have the lowest survival in

the light-reduction treatment, suggesting that the overall carbon balance of species that

normally exhibit fast growth rates may be most affected by temporal reductions in light

availability in the shaded understory. Post-light reduction RGR from 2 months to 1 year,

following removal of the shade cages at 2-months, was higher relative to control

seedlings for species with lower 1st-year control seedling survival. This result suggests

that overcompensation of RGR during seedling recovery from stress can have negative

consequences on survival, potentially because increased allocation to traits that maximize

growth rates come at the cost of allocation to traits such as defense and storage that

enhance survival (Kitajima 1994; 1996; Kobe 1997). The functional mechanisms that






16


allow seedlings to acclimate to short-term deficits and increases in light availability in the

shaded understory deserve future study.












Table 1-1. Seed and seedling characteristics of the study species.
Species Family Code Cotyledon function 1st-year survival in shade Seed mass
(position) n Mean (%) SE (mg)2
Aspidosperma cruenta Apocynaceae ASPC Storage (hypogeal) 111 96.6 1.9 453
Lacmelia panamensis Apocynaceae LACP Storage (hypogeal) 78 91.8 3.5 244
Cousarrea curvigemnia Rubiaceae COUC Photosynthesis (epigeal) 107 89.1 3.4 97
Callichlamys latifolia Bignoniaceae CALL Storage (epigeal) 55 79.9 6.1 198
Castilla elastica Moraceae CASE Storage (hypogeal) 101 57.6 5.6 314
Platypodium elegans Fabaceae PLAE Storage (hypogeal) 46 37.9 8.7 332
Tabebuia rosea Bignoniaceae TABR Photosynthesis (epigeal) 71 19.3 5.9 24
1Survival data from this study (= 1 year survival of control treatment seedlings in the shaded forest understory)
2Estimated from 6 to 37 dried seeds per species, after removing seed coats









Table 1-2. Proportional hazards (Cox regression) survival time analysis of seedling
survival over 1 year. P = significance from Wald's x2 test statistic.
Variable df Wald X p P
Species 6 806.9 <0.0001
Treatment 2 389.3 <0.0001
Species x Treatment 12 73.2 <0.0001
Garden 3 93.7 <0.0001


Table 1-3. Analysis of variance for total seedling biomass 2 months and 1 year after
treatment.
Variable 2-month biomass 1-year biomass
df F P df F P
Saturated model1
Species 4 96.1 <0.0001 3 65.2 <0.0001
Treatment 2 23.9 <0.0001 2 15.3 <0.0001
Species x Treatment 8 2.2 0.04 6 0.7 0.62
Error 57 45

Reduced main factor model
Species 6 65.3 <0.0001 6 48.4 <0.0001
Treatment 2 24.9 <0.0001 2 16.4 <0.0001
Error 67 56
The following species were removed from the saturated models because of 100%
mortality in some treatments (see Methods): Platypodium and Tabebuia
(2-month analysis); Platypodium, Tabebuia, Callichlamys (1-year analysis).






















0.4

0.2


Aspidosperma cruentum


0.0
1.0

0.8

0.6

0.4

0.2
Castilla elastica
0.0
0 40 80 120 160 200 240 280 320 360


Lacmelia panamensis










SPlatypodium elegans


Coussarea curvigemnia


Callichlamys latifolia
0 40 80 120 160 200 240 280 320 360



Control
Light reduction
Defoliation


Tabebuia rose


0 40 80 120 160 200 240 280 320 360 0 40 80 120 160 200 240 280 320 360


Time (days)


Figure 1-1. Effects of light reduction and defoliation on 1st-year seedling survival. The proportion of seedlings surviving was
estimated by the Kaplan-Meier method. Species are shown in order of 1st-year survival of control seedlings as summarized
in Table 1-1.


0l
01

0










S1 0 Defoliation: r2 = 0.67, P = 0.02
100 Light reduction: r2 = 0.75, P < 0.01

C. 80


60

60 //

20 -
4o
20


0 20 40 60 80 100

1st-year control seedling survival (%)

Figure 1-2. Two-month survival of stress treatment seedlings plotted against 1st-year
survival of control treatment seedlings. Each point is a species mean. Dashed
and solid lines show the best-fit linear regressions for the light-reduction and
defoliation treatments, respectively.













400


^ 300


S200


100


0
0.05


0.00


-0.05


-0.10
0 -o.lo

-0.15


-0.20
60


o 40

30

S20o

10
^L


- Control
Light reduction
Defoliation


u r----------
ASPC LACP COUC CALL CASE PLAE TABR
Highest Species Lowest
survival survival


Figure 1-3. Effects of light reduction and defoliation on seedling size and growth 2
months after treatment. A) Total seedling biomass. B) Relative growth rate
(RGR) from 0 to 2 months. C) Leaf area. Vertical bars = +1 SD; n = 1-4
common-garden replicates per bar (see Methods; Table A-i). Species are
ordered from highest to lowest 1st-year control seedling survival (Table 1-1).
Arrows in C) indicate complete lack of new leaf growth. RGR was not
calculated for LACP due to missing pretreatment biomass data; otherwise, a
missing bar indicates that no seedlings survived a treatment past 2 months.


* *











12

10

^ 8

0 6



S0


-2
-2 I I

ASPC COUC CASE

Species

Figure 1-4. Leaf area recovery 2 months after defoliation. Data are shown for 3 of the 4
species that developed new leaves following defoliation; leaf area recovery
was not calculated for LACP due to missing pretreatment data. Vertical
bars = +1 SD; n = 4 common-garden replicates for each species. Leaf area
recovery index = (AH Ac)/Ac, where AH = leaf area of defoliated seedlings at
2 months, and Ac = leaf area production in control seedlings from 0 to 2
months.










0.04 -



0.03 -



0.02 -



0.01 -



0.00


Highest
survival


< Species


Lowest
survival


Figure 1-5. Effects of light reduction and defoliation on relative growth rate (RGR) from
2 months to 1 year. Shade cages for the light-reduction treatment were
removed at 2 months. Vertical bars = +1 SD; n = 2-4 replicates per bar
(Appendix A).


SControl
II Light reduction
II Defoliation














ASPC LACP COUC CALL CASE PLAE TABR


CL














CHAPTER 2
CARBON ALLOCATION TO STORAGE IN TROPICAL FOREST SEEDLINGS:
EFFECTS ON GROWTH, SURVIVAL AND STRESS TOLERANCE

Introduction

For many tropical trees, successful recruitment into the canopy hinges on survival

and long-term persistence of seedlings in the shaded forest understory (e.g., Welden et al.

1991; Connell and Green 2000). Survival in the understory is dependent on the ability of

seedlings to both maintain positive net carbon balance under light-limited conditions and

cope with a host of additional biotic and abiotic stresses (reviewed in Moles and Westoby

2004; Chapter 1). The capacity of coexisting species to cope with light-limitation and

other stresses during seedling establishment ultimately influences patterns of adult tree

distributions, species composition, and coexistence (Janzen 1970; Connell 1971; Hubbell

1979; Harms et al. 2000).

What are the mechanisms that allow seedlings to maintain carbon balance and

tolerate stresses in the shaded understory? Some have hypothesized that enhanced

survival in shade (shade tolerance) is characterized by morphological and physiological

traits that maximize the rate of net carbon capture in low light at the whole-plant level

(Loach 1967; Givnish 1988), thereby increasing seedling size and competitive ability

under low light conditions. However, maintenance of a positive carbon balance is not

necessarily dependent on maximization of carbon gain, especially if seedlings experience

stress caused from tissue loss or temporal fluctuations in resource availability, both of

which may severely constrain seedling establishment in tropical forests (Augspurger









1984b; Clark and Clark 1989; Asquith et al. 1997; Scariot 2000; Montgomery and

Chazdon 2002; Engelbrecht and Kursar 2003; Khurana and Singh 2004; Kitajima 2004;

Wang and Augspurger 2004). An alternative hypothesis is that high survival is achieved

through traits that allow seedlings to maintain a positive carbon balance necessary for

long-term persistence in the understory, and which allow seedlings to cope with and

recover from biotic and abiotic stress (Kitajima 1994, 1996). Several studies from both

temperate and tropical forests provide strong support for this latter hypothesis, by

showing that: (1) interspecific variation in low light seedling and sapling survival can be

explained through tradeoffs between fast growth rates and high survival (Kitajima 1994;

Kobe et al. 1995; Pacala et al. 1996; Veneklaas and Poorter 1998; Walters and Reich

1999; Kobe 1999); (2) species with seedlings that grow quickly in low light also do so in

high light (Osunkoya et al. 1994; Grubb et al. 1996; Poorter 1999), suggesting that fast

growth rate is not necessarily an adaptive strategy for increased survival in shade.

Collectively, these studies suggest that species do not change growth ranks between low

and high light environments and that ecological tradeoffs between growth and survival

may contribute to niche partitioning across the forest light gradient. Functional traits that

enhance seedling tolerance to stress and that contribute to growth-survival tradeoffs may

therefore constitute an important, but previously little-explored aspect of seedling

regeneration and diversity in tropical forests.

Carbon allocation to storage has been proposed as one mechanism that allows

species to tolerate low light and other stresses in the shaded forest understory (Kitajima

1994, 1996; Kobe et al. 1995; Kobe 1997). Energy reserves, in the form of total

nonstructural carbohydrate (TNC, total of starch and soluble sugars), enhance recovery of









tissue lost to consumers and disturbance (McPherson and Williams 1998; Canham et al.

1999; Kabeya and Sakai 2003). TNC reserves can also aid in the maintenance of positive

carbon balance during periods when light is limited (Bloom et al. 1985) and may be

especially important for seedling survival in the shaded understory of tropical forests

where light availability can be both extremely low (0.2% full sun) and highly dynamic in

time and space (Chazdon and Pearcy 1986; Montgomery and Chazdon 2002;

Montgomery 2004). In the tropics, TNC reserves have previously been shown to vary as

a function of seasonality and light availability in adult trees (Bullock 1992; Marenco et

al. 2001; Newell et al. 2002), aid in seasonal leaf flush and reproduction of understory

shrubs and palms (Tissue and Wright 1995; Cunningham 1997; Marquis et al. 1997), and

facilitate tissue resprouting of woody plants after fire disturbance in savanna-forest

ecosystems (Hoffman et al. 2003, 2004). Despite the many important roles of TNC,

apparently no study has quantified interspecific variation in seedling TNC storage in the

understory of a tropical forest, and the ecological role of TNC in determining species-

specific susceptibility to stress-induced mortality and the associated consequences for

community-level patterns of seedling establishment remains unknown.

In this study, I examine the role of TNC reserves in the growth, survival, and stress

tolerance of seven coexisting woody species in a tropical moist forest in central Panama.

I chose species that spanned a wide range of seedling shade tolerance, which as in

previous studies (Augspurger 1984a,b; Kitajima 1994; Boot 1996), was measured as

1st-year seedling survival in the shaded understory. To quantify interspecific variation in

TNC storage, I measured TNC concentrations and pool sizes in both stems and roots at

standardized ontogenetic stages the time at which all species fully expanded their first









photosynthetic organ cotyledonss or leaves) and 2 months after leaf or cotyledon

expansion. I focused on TNC in stems and roots, and not in cotyledons and leaves,

because stem and root reserves function as more permanent and stable sites for long-term

carbohydrate storage; cotyledon reserves tend to be short-lived and most important for

seedlings before they have initiated autotrophic growth, while reserves in leaf tissue

represent both short-term and temporally dynamic storage sites for carbohydrates. I

tested three hypotheses: (1) lst-year seedling survival is higher in species with greater

allocation to TNC reserves in stems and roots; (2) species with high allocation to TNC

storage exhibit slow growth rates (i.e., a tradeoff between growth and storage); and, (3)

interspecific variation in tolerance to tissue loss and temporal reduction in light

availability are linked to species differences in TNC storage. Finally, given the

hypothesized importance of seedling size (a strong correlate of seed size in tropical

forests; Rose and Poorter 2000) in seedling survival, I also examined the relative

importance of total seedling biomass versus TNC reserve size as predictors of seedling

survival and stress tolerance among the seven species.

Methods

Study Site and Species

The study was conducted on Buena Vista peninsula, an area of ca. 60-year-old

secondary lowland tropical forest, located in the Barro Colorado Nature Monument

(BCNM), Panama (9010' N, 7951' W). The BCNM forest is semi-deciduous, with a

pronounced 4-month dry season that usually lasts from mid-December to mid-April.

Annual rainfall on the BCNM averages 2,700 mm (Rand and Rand 1982).

Seven woody species with sufficient seed availability were chosen for study

(Table 1-1). All of the species are canopy trees as adults, except for Callichlamys









latifolia (Bignoniaceae), which is a woody liana. Species were selected to span a range

of shade tolerance, which was quantified by estimating the percentage of seedlings

surviving for 1 year in experimental plots located in the shaded understory (Table 1-1).

Small-seeded, pioneer species that are generally unable to establish seedling populations

in the shaded forest understory were deliberately excluded from the study.

The seven species differ in seed mass (24 to 453 mg) and cotyledon functional

morphology (Table 1-1). Aspidosperma cruenta, Lacmeliapanamensis, Castilla elastica,

and Platypodium elegans have hypogeal cotyledons that remain below or just above the

soil surface, and that function primarily for storage of seed reserves. The three other

species have epigeal cotyledons that become elevated above the ground after

germination; Coussarea curvigemnia and Tabebuia rosea have foliaceous photosynthetic

cotyledons, and Callichlamys latifolia has thick green cotyledons that function primarily

for storage (Kitajima 2002). For brevity, species are henceforth referred to by genus.

Field and Laboratory Methods

Depending on seed availability, 200 to 400 seedlings per species were raised from

seed in a shade house on the BCNM under a light level similar to that found in the forest

understory (1.5% full sun, based on total daily photon flux density). When possible,

seeds were collected from at least 3 nonadjacent parent trees, to increase genetic diversity

within each species. Seedlings were raised in plastic trays containing a 1:1 mixture of

forest soil and sand, until all seedlings reached an equivalent ontogenetic stage, defined

by the full expansion of the first photosynthetic organs (leaves or cotyledons, depending

on the species; Table 1-1), at which time they were transplanted into the forest (described

below). Seedlings were transplanted at a standardized ontogenetic stage, rather than at a

common chronological age, so that all species had initiated autotrophic growth by the









onset of the study. Because of interspecific variability in timing of seed dispersal,

germination, and development, seedlings were transplanted over a 3-month period during

the early-mid rainy season (May to July).

Seedlings of all species were transplanted into four 7 x 7 m2 common-gardens

located in the shaded forest understory on Buena Vista Peninsula. Each garden was

separated by at least 50 m from the nearest adjacent garden, and enclosed by 1 m tall

fencing to exclude large herbivores. For each species, equal numbers of seedlings were

transplanted into 1 x 1 m2 cells in four equally sized, stratified sections in each garden

(50 to 100 total seedlings garden-1 species-1). All vegetation < 1.5 m tall was removed

from each garden before seedling transplantation, and all seedlings were positioned a

minimum of 25 cm apart. When necessary, leaf litter was also removed to avoid burial of

seedlings. Seedlings dying in the first 2 weeks after transplantation were replaced (< 10

seedlings total).

Seedlings were randomly assigned to one of three treatments 2 weeks after

transplantation: (1) defoliation; (2) light reduction; (3) control (no manipulation). For the

defoliation treatment, all leaves and photosynthetic cotyledons (if present) were clipped

at the base of the petiole; cotyledons were retained on all of the species with hypogeal

storage cotyledons (Table 1-1). For the light-reduction treatment, cylindrical wire cages

covered with 90% shade cloth were placed over individual seedlings. Shade cage

dimensions ranged from 6 x 12 cm (height x diameter) to 15 x 30 cm (depending on

initial seedling size); and were constructed to be large enough so as not to hinder growth

over a 2-month period, after which the shade cages were removed.









Light transmission was measured inside and outside of the shade cages with

Li-Cor quantum sensors during the early-mid rainy season. Percent light transmission

underneath the cages, and in each common-garden, was determined using both

instantaneous and continuous total daily photon flux density (PFD) measurements.

Instantaneous PFD measurements were taken inside and outside of shade cages at 8 15

stratified positions in each garden under mid-day, cloudy sky conditions. Continuous

PFD was measured by placing a single censor in the middle of each common-garden for

2 days. To calculate % light transmission, measurements in the forest were referenced to

a sensor placed in completely open sky on a laboratory rooftop. Mean light reduction (

1 SD) by the shade cages was 89 (+ 2)%. Mean light levels outside of the shade cages, as

determined from the instantaneous PFD measurements, ranged among gardens from 0.38

to 0.68% (overall mean = 0.55%); mean light levels from the continuous PFD

measurements taken over two full days in each garden were higher, and ranged among

gardens from 0.68 to 1.44% of full sun (overall mean = 1%). Seedlings inside the shade

cages thus experienced only ca. 0.06% of full sun over the 2-month light-reduction

treatment.

To estimate pre- and post-treatment biomass, I harvested randomly selected sub-

samples of seedlings from each common-garden at two stages: (1) 2-weeks after

seedlings were transplanted to the gardens and just before treatment (pretreatment

harvest); and, (2) 2-months after treatment, coinciding with the end of the light-reduction

treatment (post-treatment harvest). For the pretreatment harvest, 6-12 seedlings per

species were harvested from each common-garden, depending on initial seed availability.

Total samples sizes for the post-treatment harvest ranged from 2 to 32 total seedlings per









species x treatment combination, depending on species-specific differences in survival.

Seedlings were harvested by carefully excavating the root system with a small trowel.

Samples were then placed in polyethylene bags and packed inside a cooler with ice for

transport to the laboratory, where bags were stored at 40C in a refrigerator until they were

processed for biomass measurements. Roots, stems, leaves, and cotyledons were

separated within 12 hours after harvesting, than oven dried at 1000C for 1 hour. All

samples were then dried for an additional 48 hours at 600C before weighing. Leaf and

cotyledon area (for epigeal cotyledons only) were measured before drying using a

Li-Cor leaf area meter. All samples were stored in sealed polyethylene bags until

samples were ground for TNC analysis.

Quantification of TNC in Stems and Roots

TNC was analyzed separately in stems and roots by determining the concentration

of soluble sugars and starch, measured as glucose equivalents, with a colorimetric assay

(Dubois et al. 1956; modified by Ashwell 1966). Due to the small initial sizes of the

seedlings, tissue samples had to be pooled from multiple individuals to obtain adequate

tissue biomass for TNC analysis (min. of 10 mg per sample). For the pretreatment

analyses, tissues samples were pooled from 6 individuals per species in each common-

garden (n = 1-2 pooled replicates species-1 garden-'). For the post-treatment analyses,

tissue samples were pooled from 2 to 6 individuals per species x treatment combination.

The pooled tissue samples were ground using either a Wig-L-Bug bead pulverizer or a

Specs-Mill, and a 10 to 16 mg subsample was used for TNC analysis.

Starch and soluble sugars were extracted using the method of Marquis et al. (1997),

with slight modifications. Soluble sugars were first extracted by adding 1.5 mL of 80%

ethanol to the dry samples in microcentrifuge tubes, which were then placed in a shaking









water bath overnight at 270C. After shaking, tubes were centrifuged at 10,000 RPM for

10 minutes and the supernatant was decanted into 10 mL volumetric flasks. This process

was repeated for two additional, 2-hour shaking baths. The final combined solution was

diluted to 10 mL with deionized water and then stored in a refrigerator at 40C until

colorimetric analysis.

Starch content was determined from the pellet after the ethanol extractions of

simple sugars. The pellets were transferred to 15 mL test tubes and incubated with 2.5

mL of sodium acetate buffer (0.2 M; pH 4.5) in a boiling water bath for 1 hr. After

cooling, 2 mL of sodium acetate buffer and 0.5 mL of amyglucosidase solution were

added and starch was digested overnight at 550C. Solutions were filtered and then

diluted with deionized water to 25 mL in volumetric flasks. The phenol-sulfuric acid

colorimetric assay (Dubois et al. 1956; modified by Ashwell 1966) was used to determine

glucose concentrations using a spectrophotometer set at 487 nm. Glucose concentrations

were calculated from standard curves using appropriate standards and blanks. The TNC

concentration of each tissue was estimated as the sum of the glucose concentrations from

the soluble sugar and starch extractions. TNC pool sizes were calculated by multiplying

the average biomass of the pooled tissues by the sample concentrations, and then dividing

by the number of individuals in the pooled sample.

Statistical Analyses

Differences among species and treatments in biomass and TNC were analyzed

using ANOVA. Separate analyses were conducted for the pretreatment and post-

treatment (2-month) harvest, using species means from the four common-gardens as

replicates in both analyses (n = 4 replicates per species for the pretreatment analysis; n =









1-4 per species x treatment combination for the post-treatment harvest; Table A-i, B-l).

Due to lost samples, pretreatment biomass and TNC data for LACP had to be estimated

from samples collected in a separate, concurrent study in the same understory site

(K. Kitajima, unpublished data); mean stem and root biomass from this study were

calculated from 9 individuals, which were then pooled into a single sample for TNC

measurements. Due to the lack of replication for TNC, and differences in sample size for

biomass from the other 6 species, LACP was excluded from the pretreatment ANOVA.

Because all seedlings had died for some species x treatment combinations in the

first 2 months, I used 2 separate ANOVA models to analyze post-treatment biomass and

TNC. In the first model, I tested for species, treatment, and species x treatment

interaction effects on biomass and TNC response variables, but only included data for the

species that had living seedlings present in all treatments at the time of harvest (n = 5

species). The results from this model indicated that interactions had no significant effect

(P > 0.12) on any of the response variables. I therefore conducted a second analysis

using a model that contained data from all 7 species and that tested for the main effects of

species and treatment on biomass and TNC, but excluded species x interaction terms. For

brevity, I present only the results from the second model. All data except pretreatment

stem and total (stem + root) TNC pool sizes were logio-transformed before analysis to

improve normality and homogeneity of variance.

Simple least squares linear regression was used to test for relationships between

seedling biomass, growth, survival, and TNC storage. Relative growth rates (RGR) of

total seedling biomass and leaf area from 0 to 2 months was estimated using the

following formula:









RGR (wk1) = [In (size at 2 months) In (size at 0 months)]/8

Leaf and cotyledon mass was included in all RGR calculations of total seedling biomass.

Survival estimates for control seedlings over the full year (Table 1), and for stress-treated

seedling from 0 to 2 months and 2 months to 1 year, were taken from a separate study

(Chapter 1). All statistical analyses were performed using JMP software (SAS Institute,

1997).

Results

Variation in Pretreatment Biomass and TNC Reserves

Pretreatment stem and root biomass varied widely and significantly among species

(Figure 2-la; Table 2-1). Biomass was higher in stems relative to roots in all species.

Pretreatment cotyledon biomass also varied widely, ranging from 18 to 209 mg among

species (Figure 2-la). In four of the seven species (Aspidosperma, Lacmelia, Castilla,

and Tabebuia), biomass was higher in cotyledons than in stems or roots combined.

Pretreatment TNC concentrations also differed significantly among species

(Table 2-2). TNC concentrations varied more widely in stems relative to roots among

species (Figure 2-1b). TNC concentrations were also substantially higher in stems than

in roots. TNC comprised between 6 to 25% of the total stem biomass (mean = 15% for all

species combined), whereas root concentrations ranged from 4 to 21% (mean = 8%).

There were no significant correlations between 1st-year control seedling survival

(Table 1-1) and total pretreatment biomass (Figure 2-2a) or TNC concentration among

species (Figure 2-2b), despite high interspecific variation in both these variables. First

year survival was also not related to pretreatment cotyledon biomass, stem biomass, root

biomass, or root:shoot ratio among species (P > 0.3 for all regressions using species

means).









Pretreatment TNC pool sizes also differed significantly among species, especially

in stems, which were the dominant storage location for TNC in all species (Figure 2-3;

Table 2-2). Total TNC pool size (stem + root pools combined) was a better correlate of

1st-year control seedling survival among species (Figure 2-2c) than was total pretreatment

biomass or TNC concentration (Figure 2-2a,b): Aspidosperma and Lacmelia, the two

species with the highest 1st-year survival, had the largest total TNC pools (12.9 and 15

mg, respectively), while Tabebuia, the species with lowest 1st-year survival, had the

smallest pool size (1.7 mg). The other 4 species had similar, intermediate levels of TNC

ranging between 4.1 and 6.5 mg (Figure 2-3).

Seedling size had a stronger effect on TNC pool size than on concentration. The

four species with the largest total seedling biomass (Aspidosperma, Lacmelia, Castilla,

Platypodium), all of which had hypogeal storage cotyledons, had greater TNC pool sizes

than species the two smallest species (Coussarea, Tabebuia) with epigeal photosynthetic

cotyledons (Figure 2-3; Table 1-1). In contrast, there were no clear trends between stem

or root TNC concentrations and total seedling biomass among species. Coussarea and

Tabebuia, the species with the lowest total seedling biomass, had two of the highest stem

TNC concentrations.

Effects of Defoliation and Deep Shade on Biomass and TNC Reserves

There were significant species and treatment effects on post-treatment stem and

root biomass at 2 months (Figure 2-4a,b; Table 2-1). For most species, both defoliation

and light reduction caused a decrease in stem and root biomass relative to controls. The

combined biomass of stems and roots (not including leaf or cotyledon biomass) was

reduced more substantially by defoliation in species with lower 1st-year control seedling

survival (Figure 9a,b): biomass reductions (relative to control treatment biomass) ranged









from 5 to 16% in the three species with highest 1st-year survival (Aspidosperma,

Lacmelia, Coussarea), compared to 19-46% in the three species with lower 1st-year

survival (Callichlamys, Castilla, Platypodium). Species with lower 1st-year survival also

suffered greater reductions in biomass in the light-reduction treatment (range = 11-29%

for Callichlamys, Castilla, and Tabebuia) relative to species with higher l1t-year survival

(range = 1-16% for Aspidosperma, Lacmelia, and Coussarea).

Stress treatments had significant negative effects on both TNC pools and

concentrations (Table 2-2). TNC concentrations were reduced proportionally less by the

stress treatments than TNC pools in all species (Figure 2-4c-f). Reductions in TNC pool

sizes (Figure 2-2e,f) generally mirrored reductions in biomass (Figure 2-2a,b) among

species. However, total TNC pool size was reduced more substantially than was total

biomass: TNC reductions in the three species with the lowest lst-year control seedling

survival ranged from 47 to 64% and 23 to 54% for the defoliation and light-reduction

treatments, respectively. Tabebuia, the species with lowest 1st-year survival and the

smallest TNC pools, suffered the highest mortality after defoliation.

Post-treatment cotyledon biomass differed among species and treatments

(Figure 2-5). Two of the species (Platypodium and Callichlamys) had exhausted all

cotyledon reserves in all treatments in the first 2 months. Control treatment values

among the other five species ranged from 0.61 to 28 mg. Cotyledon biomass of stressed

seedlings was either higher or only slightly reduced relative to control seedlings in all

species (Figure 2-5).

Influence of Seedling Size, Growth, and TNC Reserves on Seedling Survival

Interspecific variation in TNC pool size was a better correlate of seedling survival

than seedling size for both control and stress-treated seedlings. For control seedlings,









1st-year survival was more strongly correlated with TNC pool size than to TNC

concentration (Figure 2-2). Furthermore, whole-seedling RGR from 0 to 2 months was

negatively correlated with pretreatment TNC pool size among species (Figure 2-6).

RGR, however, was not significantly correlated with either pretreatment TNC

concentrations or total seedling biomass (P > 0.8, r2 = 0.01, n = 6 for stem and root TNC

concentrations; P = 0.11, r2 = 0.50 for total seedling biomass). Two-month RGR of total

structural biomass (seedling biomass TNC pools), as well as absolute total biomass

growth, were also negatively, and more significantly, correlated with pretreatment TNC

pool size among species (P = 0.03, r2 = 0.69 for RGR of structural biomass; P = 0.01,

r2 = 0.80 for absolute biomass growth). RGR of leaf area growth from 0 to 2 months was

also negatively, but less strongly, correlated with pretreatment TNC pool size among

species (P = 0.08, r2 = 0.56). Thus, large initial investments in TNC storage pools

enhanced 1st-year seedling survival, but at the cost of lower seedling growth rates.

Survival of stressed seedlings was also significantly correlated with TNC pool size,

but not with total seedling biomass (Figure 2-7). TNC pools had the greatest influence on

survival during the first 2 months, when seedlings were most impacted by the stress

treatments (Figure 2-7a); 2-month survival was also related, though less strongly, to stem

TNC concentration among species (P = 0.01, r2 = 0.41, n = 14). Five of the seven species

had seedlings that survived the stress treatments past 2 months. Survival of these species

from 2 months to 1 year, when species were presumably recovering from the stress

treatments, was also significantly related to TNC pool size at the end of 2 months (Figure

2-7c). Seedling survival was not significantly related to total seedling biomass during the

first 2 months or from 2 months to 1 year (Figure 2-7b,d).









Discussion

My results support the hypotheses that TNC storage enhances seedling survival and

stress tolerance in the shaded forest understory. Overall, the seven species displayed

wide variation in their carbon allocation strategies to seedling growth and storage. Most

importantly, interspecific variation in TNC pool size, but not seedling size, was a good

predictor of both 1st-year seedling survival in the shaded understory, and seedling

tolerance to stress caused from both tissue loss and temporary reduction in light

availability. TNC concentrations in stems and roots ranged widely among species

(4.5-25 %) and generally varied independently of seedling size, cotyledon type, and

seedling shade tolerance. Finally, my study provides evidence for a tradeoff between

allocation to growth and carbohydrate storage among coexisting species in a tropical

forest. These results provide a mechanistic explanation for why species with high

seedling survival in the shaded forest understory also tend to exhibit slow growth rates

(Kitajima 1994, 2002; Osunkoya et al. 1994), and support the notion that tradeoffs

between growth and survival may play an important role in maintaining high tree species

diversity in tropical forests (Kitajima 1994; Kobe 1999).

Role of TNC Reserves in Seedling Survival and Stress Tolerance

My results show a strong linkage between interspecific variation in TNC storage

and survival of stressed seedlings. TNC reserves had the greatest impact on seedling

survival during the first 2 months after the stress treatments. During this time, both TNC

concentrations and pools were reduced in all species, underscoring the importance of

storage reserves for survival when seedlings experience negative carbon balance during

the early seedling establishment. However, even after TNC levels were reduced

experimentally, TNC pool size continued to be a significant predictor of variation in









seedling survival through the year. TNC reserves therefore played an important role in

both stress tolerance and post-stress recovery.

TNC reserves may allow tropical tree seedlings to cope with and recover from a

host of biotic and abiotic stresses. Several of my study species showed substantial

tolerance to defoliation, confirming previous results from seedling defoliation

experiments conducted in temperate (Canham et al. 1999) and tropical (Becker 1983;

Kitajima 2004) forests. The ability to survive defoliation was dependent on the relative

capacity of a species to develop new and fully expanded leaves, and species with greater

leaf recovery exhibited higher survival (Chapter 1). These results indicate that large TNC

reserves may be essential for seedling recovery from tissue loss due to herbivores and

pathogens, or shoot damage from falling canopy debris, all of which are major causes of

seedling mortality in tropical forests (Augspurger 1984a,b; Osunkoya et al. 1992;

Asquith et al. 1997; Clark and Clark 1989; Scariot 2000).

The results of my study also demonstrate that TNC reserves play a critical role

in maintaining carbon balance when seedlings experience short-term reductions in light

availability. The aim of the light-reduction treatment was to expose seedlings to

irradiance levels well below their light compensation point, thereby forcing seedlings into

a negative carbon balance. My results indicate that species can vary widely in their

survival and growth responses to extreme temporal reductions in light availability in the

shaded understory. The most shade-tolerant species in my study exhibited extremely

high tolerance to experimental light levels of ca. 0.06% of full sun for 2 months, a level

far below what has been used in most previous experimental studies and 1/10 of the

average understory values typically reported for tropical forests. While the degree of









light reduction in this experiment was extreme, it is probably representative of the light

limitations imposed on seedlings during periods of intensive cloud cover (daily and

seasonally) and when seedlings establish in microsites deeply shaded by litter

accumulation or understory foliage.

Tradeoffs between Seedling Growth and Storage

The results supported the prediction of a tradeoff between seedling growth rate and

allocation to carbohydrate storage reserves. Across species, TNC pool size was

negatively correlated with RGR of total seedling biomass, RGR of total structural

seedling biomass, and absolute seedling growth during the first 2 months of study. TNC

pool size was also positively correlated with 1st-year seedling survival, explaining

approximately half of the variation in survival among species. These results support the

hypothesis that seedling survival in the shaded understory of tropical forests is achieved

through a balance between allocation to growth and carbohydrate storage (Kitajima 1994,

1996), rather than by maximizing net carbon gain for fast seedling growth rates (e.g.,

Givnish 1988).

TNC pool size was a better correlate of seedling RGR and survival than either TNC

concentration or seedling size. Furthermore, TNC pools were generally more reduced by

the stress treatments than were TNC concentrations and total biomass. These results

suggest that the total pool of TNC, irrespective of seedling size per se, may serve as the

best quantitative indicator of interspecific variation in seedling survival and stress

tolerance in the shaded forest understory. For example, TNC pools in Aspidosperma, the

species with highest total seedling biomass and lowest RGR, were seven times higher

than in Tabebuia, the species with lowest seedling biomass and highest RGR. In

contrast, TNC concentrations differed only slightly between these species. These results









emphasize the importance of distinguishing between structural versus nonstructural

biomass when linking morphological traits to seedling growth, biomass allocation, and

survival (e.g., Canham et al. 1999).

TNC Allocation, Seed Size, and Seedling Functional Morphology

My results suggest that stems may be the dominant storage location for TNC in

understory tropical forest seedlings. Both TNC concentrations and pool sizes were higher

in stems than in roots in all of the 7 species. Higher TNC storage in stems could reflect

physiological constraints related to the relatively small size of seedling root systems in

the shaded understory. TNC allocation to stems versus roots might also be related to the

relative susceptibility of these tissues to damage from herbivory or disturbance during

development. In forested ecosystems that experience large-scale disturbance such as fire,

or where seedlings must survive though winter dormancy, storage reserves are typically

concentrated in roots rather than stems (Canham et al. 1999; Hoffman et al. 2004).

Variation in TNC storage was linked to species differences in both seed size and

cotyledon functional morphology. The four largest-seeded species, all of which had

hypogeal storage cotyledons, had the largest seedlings and TNC pools in stems and roots

(but not necessarily the higher TNC concentrations). Total TNC pools in these species

undoubtedly represent conservative estimates, as cotyledon reserves were not measured

in this study. In contrast, the two smallest seeded species, both of which had epigeal

photosynthetic cotyledons, had small seedlings and lower TNC pool sizes, but two of the

highest TNC concentrations. These results show that large-seeded tropical tree species,

in addition to having high amounts of energy reserves in seeds to enhance short-term

survival, also maintain large pools of longer-term storage reserves in stems and roots.

However, because all species had become dependent on autotrophic carbon gain before









my TNC measurements, it is unknown to what degree initial seed reserves contributed to

the more permanent stem and root TNC pools in these species. Future experiments that

quantify how seed reserves contribute to long-term TNC storage in seedlings will likely

shed further light on the evolutionary and ecological role of seed size variation in tropical

forest tree communities.

Conclusion

In this paper, I highlight the importance of TNC reserves for seedling survival and

stress tolerance in the shaded understory of a tropical forest. Tradeoffs between traits

such as storage that enhance seedling stress tolerance versus those that maximize growth

have likely contributed to the evolution of life history diversity and the maintenance of

species coexistence in species-rich tropical forests. Interspecific variability in TNC

storage constitutes a central component of seedling recruitment dynamics in this and

likely many other tropical forests.






43


Table 2-1. Analysis of variance for stem and root biomass.
Variable Stem biomass Root biomass
df F P F P
Pre-treatment'
Species 5 260.7 <0.0001 36.7 <0.0001
Error 18

Post-treatment (2 months)
Species 6 125.8 <0.0001 127.3 <0.0001
Treatment 2 7.7 0.001 16.8 <0.0001
Error 59
'LACP excluded from the analysis (see Methods)












Table 2-2. Analysis of variance for TNC concentrations and pool sizes.
Variable TNC concentration TNC pool size Total TNC pool size
Stem Root Stem Root (stem + root)
df F P F P F P F P F P
Pre-treatment
Species 5 115.0 <0.0001 14.4 <0.0001 116.1 <0.0001 27.0 <0.0001 113.2 <0.0001
Error 18

Post-treatment
Species 6 49.0 <0.0001 41.9 <0.0001 43.8 <0.0001 37.0 <0.0001 52.1 <0.0001
Treatment 2 12.3 <0.0001 6.8 0.002 20.0 <0.0001 16.1 <0.0001 23.8 <0.0001
Error 57 59










Z/// Cotyledon
SStem
I Root


A


ASPC
Highest
survival


230 -

220 -

210 -
c /

100 -

50

0


LACP COUC CALL CASE PLAE
< Species >


TABR
Lowest
survival


Figure 2-1. Pretreatment biomass and TNC concentrations. A) Cotyledon, stem, and
root biomass. B) Stem and root TNC concentrations. Species are ordered
from highest to lowest 1st-year survival of control seedlings (Table 1-1). Stars
in (A) indicate species with storage cotyledons. Vertical bars = + 1 SD; n = 4
common-garden replicates for each species (see Methods); values for LACP
were estimated from another experiment (n = 9 for biomass, pooled into a
single sample for TNC analysis).


*e




In I?~ flI


in S


I I I I I I I

















100 E C

80

60 -


P>0.3
r 0.16


A P>0.4
r =0.13


0
Ea Z


P= 0.09
r2 = 0.46


0 100 200 300 400 500 0 5 10 15 20 25 0 2 4 6 8 10 12 14 16


Total biomass (mg)


Stem TNC concentration (%)


Total TNC pool size (mg)


Figure 2-2. Three potential correlates of interspecific variation in 1st-year control seedling survival. A) Pretreatment total seedling
biomass (stems, roots, cotyledons, and leaves). B) Pretreatment stem TNC concentration. C) Pretreatment total TNC pool
size (stem + root pools combined).


0

iuj


40

20


* 0^^


O ASPC
A LACP
E COUC
* CALL
* CASE
* PLAE
A TABR

















N

O
O
r/s


ASPC LACP COUC CALL CASE PLAE TABR


Highest
survival


< Species ->


Lowest
survival


Figure 2-3. Pretreatment TNC pool sizes in stems and roots.







Roots


in


ILIr


hn ii


At


- Control
Light reduction
Defoliation


i I Ih


B in


7- F
6
5
4
3
2
1
0
ASPC LACP COUC CALL CASE PLACE TABR


Species


Figure 2-4. Post-treatment biomass and TNC in stems and roots. A,B) Total biomass.
C,D) TNC concentrations. E,F) TNC pool sizes. TNC for TABR in both
stress treatments and for PLAE in the light-reduction treatment could not be
analyzed due to low survival. Vertical bars = + 1 SD; n = 1-4 common-
garden replicates per bar (Table A-i, B-l).


120 A


IT


g 25 C
.2 20
15
10
5


16 E
14
12
10
8
6
4
2
0


IFFI


ASPC LACP COUC CALL CASE PLAE TABR


Species


Stems












100
Control
0I I I Light reduction
8 I Defoliation
80

ct
E 60


o 40


0 20



ASPC LACP COUC CALL CASE PLAE TABR

Species

Figure 2-5. Post-treatment cotyledon biomass. Cotyledons in COUC, CALL and TABR
were removed as part of the defoliation treatment. All cotyledon reserves in
PLAE and for the control and light-reduction treatments in CALL were
exhausted before the post-treatment harvest at 2 months. Vertical bars =
+ 1 SD; n = 2-4 common-garden replicates per bar (Table A-i).











0.03 A P = 0.03

0.02 r2 = 0.70

0.01 -

0.00 -

-0.01

-0.02 0 ASPC
E COUC
-0.03 CALL
CASE
-0.04 U PLAE
A TABR
-0.05 -
0 2 4 6 8 10 12 14

TNC pool size (mg)

Figure 2-6. Two-month relative growth rate (RGR) of control seedlings plotted against
pretreatment TNC pool size. Data are shown for the 6 species in which
complete biomass data were available.













100 A


S/

0


0.002
0.54


P= 0.20
r = 0.12


0 2 4 6 8 10 12 14 16 0 50 100 150 200 250 300 350 400 450


100 C


0 0
S o





P= 0.05
Sr2 = 0.39


P= 0.14
r = 0.24


0 2 4 6 8 10 12 14 0

Total TNC pool size (mg)


50 100 150 200 250 300

Total seedling biomass (mg)


Figure 2-7. Survival of stressed seedlings plotted against total TNC pool size and total
seedling biomass. A,B) Survival from 0 to 2 months as a function of
pretreatment total TNC pool size and total seedling biomass. C,D) Survival
from 2 months to 1 year as a function of post-treatment (2-month) total TNC
pool size and total seedling biomass. Each point is a species mean; white
circles = defoliation treatment, dark circles = light-reduction treatment.


o















APPENDIX A
SUMMARY STATISTICS FOR SEEDLING BIOMASS AND LEAF AREA

Table A-i. Statistics for pre- and post-treatment biomass and leaf area.
Species Pre-treatment Post-treatment (2 months)
Control Defoliation Light reduction
n mean SD n mean SD n mean SD n mean SD
Cotyledon mass
(mg)
ASPC 4 209.2 16.9 4 28.6 25.5 4 59.6 33.0 4 39.2 30.2
CALL 4 22.4 16.9 4 0.0 0.0 2 0.0 0.0 4 0.0 0.0
CASE 4 91.4 23.9 4 0.6 1.2 4 12.0 8.4 4 3.9 3.3
COUC 4 18.1 1.2 4 16.3 1.6 4 0.0 0.0 4 16.4 2.5
LACP 9 64.4 65.5 4 20.9 20.8 4 37.6 29.5 4 39.3 39.4
PLAE 4 33.9 12.5 4 0.0 0.0 1 0.0 0 -
TABR 4 18.6 2.8 3 17.9 2.4 0 2 16.6 0.3

Root mass (mg)
ASPC 4 48.1 3.4 4 57.7 5.7 4 50.3 4.5 4 47.8 4.4
CALL 4 20.9 4.5 4 25.6 7.2 2 16.0 2.8 4 15.4 5.0
CASE 4 20.4 5.2 4 20.3 2.6 4 14.4 1.1 4 16.5 5.1
COUC 4 8.1 0.8 4 8.5 1.2 4 6.2 0.8 4 6.7 0.6
LACP1 9 14.7 5.8 4 17.2 3.1 4 14.7 3.3 4 15.3 1.5
PLAE 4 26.6 11.0 4 27.1 9.5 1 10.0 0 -
TABR 4 5.5 0.7 3 5.8 0.9 0 2 4.7 1.1









Table A-1. Continued
Species Pre-treatment Post-treatment (2 months)
Control Defoliation Light reduction
n mean SD n mean SD n mean SD n mean SD


Stem mass (mg)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR

Total mass (mg)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR

Leaf area (cm2)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR


4 70.2 4.7 4 84.5 8.1 4 84.5 3.0 4 75.2 3.7
4 38.3 5.1 4 57.3 10.8 2 30.8 3.5 4 43.0 12.3
4 44.9 3.8 4 64.0 11.9 4 53.4 25.1 4 58.1 7.6
4 22.0 1.4 4 20.3 1.6 4 18.0 2.3 4 17.4 1.3
9 46.8 4.1 4 46.5 9.0 4 44.3 6.4 4 47.2 5.4
4 79.9 7.1 4 99.4 5.9 1 57.4 0 -
4 8.5 1.2 3 11.9 3.2 0 2 7.7 0.4



4 442.7 12.2 4 331.3 20.4 4 221.3 34.6 4 297.5 48.1
4 146.8 28.5 4 171.9 28.4 2 46.7 0.7 4 117.2 19.8
4 219.8 34.2 4 171.7 23.9 4 99.3 36.7 4 115.7 34.3
4 48.2 1.5 4 45.7 1.6 4 25.4 1.7 4 40.6 3.7
9 185.3 79.1 4 156.1 39.0 4 122.6 48.1 4 157.4 46.4
4 213.3 9.9 4 234.0 24.6 1 67.4 0 -
4 33.5 5.7 3 39.6 9.6 0 2 30.0 0.1



4 28.1 2.1 4 29.8 1.8 4 7.3 1.7 4 28.7 3.5
4 37.2 2.0 4 43.8 5.1 2 0.0 0.0 4 32.9 7.5
4 32.9 3.0 4 47.2 7.6 4 11.6 4.7 4 22.8 13.2
4 4.5 0.5 4 5.0 0.4 4 0.6 0.1 4 4.9 0.6
9 32.4 7.3 4 29.8 5.2 4 11.6 5.1 4 27.1 2.4
4 32.8 8.0 4 39.7 6.7 1 0.0 0 -
4 6.5 0.9 3 9.7 3.0 0 2 7.6 1.6









Table A-2. Statistics for total seedling biomass and leaf area 1 year after treatment.
Species Post-treatment (1 year)
Control Defoliation Light reduction
n mean SD n mean SD n mean SD
Total mass (mg)
ASPC 4 536.8 143.9 4.0 248.1 58.9 4 466.2 94.8
CALL 4 519.6 190.8 0.0 2 404.6 49.7
CASE 4 272.4 74.0 3.0 179.8 85.3 3 243.7 154.7
COUC 4 96.0 29.1 4.0 38.5 9.9 4 77.9 16.8
LACP 4 313.8 99.7 4.0 200.9 52.9 4 263.6 120.0
PLAE 3 371.6 122.0 0.0 0 -
TABR 2 41.4 20.2 0.0 0 -

Leaf area (cm2)
ASPC 4 44.5 12.2 4.0 13.6 5.7 4 38.8 4.8
CALL 4 89.4 24.2 0.0 2 61.6 7.6
CASE 4 52.6 13.9 3.0 32.6 17.1 3 37.1 38.4
COUC 4 11.8 3.7 4.0 3.3 1.5 4 9.4 2.2
LACP 4 46.8 12.3 4.0 29.0 6.9 4 40.5 17.8
PLAE 3 43.2 37.5 0.0 0 -
TABR 2 5.9 8.3 0.0 0 -















APPENDIX B
SUMMARY STATISTICS FOR TOTAL NONSTRUCTURAL CARBOHYDRATE
(TNC) RESERVES IN STEMS AND ROOTS

Table B-1. Statistics for pre- and post-treatment TNC concentrations and pool sizes.
Species Pre-treatment Post-treatment (2 months)


Control


n mean SD


6.20 0.49
4.60 0.66
5.26 0.97
10.42 1.47
21.91 -
4.38 0.85
6.11 1.25


2.95 0.38
0.97 0.30
1.04 0.16
0.85 0.19
3.23 -
1.11 0.37
0.34 0.09


14.20 0.34
8.19 0.56
12.31 0.60
21.06 2.00
25.12 -
6.10 0.54
17.38 2.42


Defoliation


n mean SD n mean SD


8.22
9.22
4.04
13.05
17.10
7.29
9.35


4.80
2.52
0.89
1.12
2.84
2.03
0.73


13.08
12.09
8.12
21.79
21.51
7.25
12.63


1.85
3.28
1.48
1.23
1.84
1.86
1.66


1.53
1.69
0.41
0.25
0.65
0.92
0.38


3.25
4.43
2.28
4.41
1.37
0.67
1.84


6.28 1.76
6.25 1.56
3.99 0.55
9.93 0.63
12.02 3.28
7.15 -




3.19 1.08
1.02 0.43
0.57 0.09
0.60 0.06
1.74 0.68
0.71 -




10.91 2.42
7.73 2.98
5.24 0.47
16.84 5.68
16.89 3.62
5.11 -


Light reduction
n mean SD


5.92 1.58
6.85 1.94
2.95 0.44
10.21 0.67
15.67 4.25





2.80 0.72
1.06 0.19
0.50 0.26
0.69 0.03
2.35 0.65





11.53 1.45
9.48 1.20
7.08 1.23
16.84 1.12
20.84 2.20


Root (%)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR

Root (mg)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR

Stem (%)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR










Table B-1. Continued
Species Pre-treatment Post-treatment (2 months)
Control Defoliation Light reduction
n mean SD n mean SD n mean SD n mean SD


9.99 0.88
3.15 0.58
5.50 0.56
4.63 0.45
11.77 -
4.85 0.24
1.45 0.12


12.94 1.24
4.11 0.87
6.55 0.66
5.49 0.40
15.00 -
5.96 0.18
1.79 0.18


4 11.18 3.77
4 6.86 2.87
4 5.51 1.92
4 4.35 0.58
4 10.28 2.60
4 7.19 0.66
4 1.58 0.32


4 15.98 5.28
4 9.38 4.54
4 6.40 2.32
4 5.47 0.82
4 13.12 3.22
4 9.22 1.20
4 2.31 0.60


9.25 2.27
2.32 0.64
2.81 1.23
2.62 0.44
7.77 2.61
2.93 -




12.44 3.15
3.34 1.07
3.37 1.32
3.22 0.40
9.51 3.29
3.64 -


8.71 1.25
3.55 0.47
4.42 1.49
2.94 0.30
10.10 1.83





11.51 1.62
4.73 0.47
4.92 1.66
3.62 0.31
12.45 2.43

1.06


Stem (mg)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR

Total (mg)
ASPC
CALL
CASE
COUC
LACP
PLAE
TABR















LIST OF REFERENCES


Ashwell, G. (1966) New colorimetric methods of sugar analysis. VII. The phenol-sulfuric
acid reaction for carbohydrates. Methods in Enzymology, 8, 93-95.

Asquith, N.M., Wright, S.J. & Clauss, M.J. (1997) Does mammal community
composition control recruitment in neotropical forests? Evidence from Panama.
Ecology, 78, 941-946.

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

Augspurger, C.K. (1984b) Light requirements of neotropical tree seedlings: A
comparative study of growth and survival. Journal ofEcology, 72, 777-795.

Augspurger, C.K. (1983) Seed dispersal of the tropical tree, Platypodium elegans, and the
escape of its seedlings from fungal pathogens. Journal of Ecology, 71, 759-771.

Augspurger, C.K. & Kelly, C.K. (1983) Pathogen mortality of tropical tree seedlings:
Experimental studies of the effects of dispersal distance, seedling density, and light
conditions. Oecologia, 61, 211-217.

Becker, P. (1983) Effects of insect herbivory and artificial defoliation on survival of
.V\/in ea seedlings. Tropical Rain Forest: Ecology and Management (eds. S.L.
Sutton, T.C. Whitmore, & A.C. Chadwick), pp. 241-252. Blackwell Scientific
Publishers, Oxford.

Bloom, A.J., Chapin, F.S. III & Mooney, H.A. (1985) Resource limitation in plants: An
economic analogy. Annual Review of Ecology and Systematics, 16, 363-392.

Boot, R.G.A. (1996) The significance of seedling size and growth rate of tropical rain
forest tree seedlings for regeneration in canopy openings. The Ecology of Tropical
Forest Tree Seedlings (ed M.D. Swaine), pp. 267-282. UNESCO, Paris.

Bullock, S.H. (1992) Seasonal differences in nonstructural carbohydrates in two
dioecious monsoon-climate trees. Biotropica, 24, 140-145.

Canham, C.D., Kobe, R.K., Latty, E.F. & Chazdon, R.L. (1999) Interspecific and
intraspecific variation in tree seedling survival: effects of allocation to roots versus
carbohydrate reserves. Oecologia, 121, 1-11.









Chazdon, R.L. & Fetcher, N. (1984) Photosynthetic light environments in a lowland
tropical rain forest in Costa Rica. Journal ofEcology, 72, 553-564.

Chazdon, R.L. & Pearcy, R.W. (1986) Photosynthetic responses to light variation in
rainforest species. II. Carbon gain and photosynthetic efficiency during lightflecks.
Oecologia, 69, 524-531.

Clark, D.B. & Clark, D.A. (1989) The role of physical damage in the seedling mortality
regime of a neotropical rainforest. Oikos, 55, 225-230.

Coley, P.D., Bryant, J.P., & Chapin, F.S. III (1985) Resource availability and plant
antiherbivore defense. Science, 230, 895-899.

Connell, J.H. (1971) On the role of natural enemies in preventing competitive exclusion
in some marine animals and in rain forest trees. Dynamics ofPopulations (eds P.J.
den Boer & G.R. Gradwell), pp. 298-312. Center for Agricultural Publication and
Documentation, Wageningen, The Netherlands.

Connell, J.H. & Green, P.T. (2000) Seedling dynamics over thirty-two years in a tropical
rain forest tree. Ecology, 81, 568-584.

Cunningham, S.A. (1997) The effect of light environment, leaf area, and stored
carbohydrates on inflorescence production by a rain forest understory palm.
Oecologia, 111, 36-44.

Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A. & Smith, F. (1956) Colorimetric
method for determination of sugars and related substances. Analytical Chemistry,
28, 350-356.

Engelbrecht, M.J. & Kursar, T.A. (2003) Comparative drought-resistance of seedlings of
28 species of co-occurring tropical woody plants. Oecologia, 136, 383-393.

Farris-Lopez, K., Denslow, J.S., Moser, B. & Passmore, H. (2004) Influence of a
common palm, Oenocarpus mapora, on seedling establishment in a tropical moist
forest in Panama. Journal of Tropical Ecology, 20, 429-438.

Fox, G.A. (2001) Failure-time analysis: Studying times to events and rates at which
events occur. Design and Analysis of Ecological Experiments (eds S.M. Scheiner &
J. Gurevitch), pp.235-266. Oxford University Press, Inc., New York.

Givnish, T.J. (1988) Adaptation to sun and shade: A whole-plant perspective. Australian
Journal ofPlant Physiology, 15, 63-92.

Green, P.T. & Juniper, P.A. (2004) Seed mass, seedling herbivory and the reserve effect
in tropical rainforest seedlings. Functional Ecology, 18, 539-547.

Grime, J.P. (1965) Shade tolerance in flowering plants. Nature, 208, 161-163.









Grubb, P.J., Lee, W.G., Kollmann, J. & Wilson, J.B. (1996) Interaction of irradiance and
soil nutrient supply on growth of seedlings of ten European tall-shrub species and
Fagus sylvatica. Journal ofEcology, 84, 827-840.

Gunatilleke, C.V.S., Gunatilleke, I.A.U.N., Perera, G.A.D., Burslem, D.F.R.P., Ashton,
P.M.S. & Ashton, P.S. (1997) Responses to nutrient addition among seedlings of
eight closely related species of ,l/,,,ea in Sri Lanka. Journal ofEcology, 85, 301-
311.

Harms, K.E. & Dalling, J.W. (1997) Damage and herbivory tolerance through
resprouting as an advantage of large seed size in tropical trees and lianas. Journal
of Tropical Ecology, 13, 617-621.

Harms, K.E., Wright, S.J., Calder6n, O., Hernandez, A. & Herre, E.A. (2000) Pervasive
density-dependent recruitment enhances seedling diversity in a tropical forest.
Nature, 404, 493-495.

Hoffman, W.A., Orthen, B. & Franco, A.C. (2004) Constraints to seedling success of
savanna and forest trees across the savanna-forest boundary. Oecologia, 140, 252-
260.

Hoffman, W.A., Orthen, B. & Nascimento, P.K.V. (2003) Comparative fire ecology of
tropical savanna and forest trees. Functional Ecology, 17, 720-726.

Hubbell, S.P. (1979) Tree dispersion, abundance, and diversity in a tropical dry forest.
Science, 203, 1299-1309.

Janzen, D.H. (1970) Herbivores and the number of tree species in tropical forests.
American Naturalist, 104, 501-528.

Kabeya, D. & Sakai, S. (2003) The role of roots and cotyledons as storage organs in early
stages of establishment in Quercus crispula: A quantitative analysis of the
nonstructural carbohydrate in cotyledons and roots. Annals ofBotany, 92, 537-545.

Khurana, E. & Singh, J.S. (2004) Germination and seedling growth of five tree species
from tropical dry forest in relation to water stress: Impact of seed size. Journal of
Tropical Ecology, 20, 385-396.

Kitajima, K. (2004) Impact of cotyedon and leaf removal on seedling survival in three
tree species with contrasting cotyledon functions. Biotropica, 35, 429-434.

Kitajima, K. (2002) Do shade-tolerant tropical tree seedlings depend longer on seed
reserves? Functional growth analysis of three Bignoniaceae species. Functional
Ecology, 16, 433-444.

Kitajima, K. (1996) Ecophysiology of tropical tree seedlings. Tropical Forest Plant
Ecophysiology (eds S.S. Mulkey, R.L. Chazdon, & A.P. Smith), pp. 559-596.
Chapman and Hall, New York, NY.









Kitajima, K. (1994) Relative importance of photosynthetic traits and allocation patterns
as correlates of seedling shade tolerance of 13 tropical trees. Oecologia, 98, 419-
428.

Kobe, R.K. (1999) Light gradient partitioning among tropical tree species through
differential seedling mortality and growth. Ecology, 80, 187-201.

Kobe, R.K. (1997) Carbohydrate allocation to storage as a basis of interspecific variation
in sapling survivorship and growth. Oikos, 80, 226-233.

Kobe, R.K., Pacala, S.W., Silander, J.A., Jr. & Canham, C.D. (1995) Juvenile tree
survivorship as a component of shade tolerance. Ecological Applications, 5, 517-
532.

Li, M., Leiberman, M. & Leiberman, D. (1996) Seedling demography in undisturbed
tropical wet forest in Costa Rica. The Ecology of Tropical Forest Tree Seedlings
(ed M.D. Swaine), pp. 285-314. UNESCO, Paris.

Loach, K. (1967) Shade tolerance in tree seedlings. I. Leaf photosynthesis and respiration
in plants raised under artificial shade. New Phytologist, 66, 607-621.

Marenco, R.A, de C. Gongalves, J.F. & Vieira, G. (2001) Leaf gas exchange and
carbohydrates in tropical trees differing in successional status in two light
environments in central Amazonia. Tree Physiology, 21, 1311-1318.

Marquis, R.J., Newell, E.A. & Villegas, A.C. (1997) Non-structural carbohydrate
accumulation and use in an understory rain-forest shrub and the relevance for the
impact of leaf herbivory. Functional Ecology, 11, 636-643.

McPherson, K. & Williams, K. (1998) The role of cabbage palm seedlings in the growth,
resilience, and persistence of cabbage palm seedlings (Sabalpalmetto). Oecologia,
117, 460-468.

Moles, A.T. & Westoby, M. (2004) What do seedlings die from and what are the
implication for the evolution of seed size? Oikos, 106, 193-199.

Montgomery, R.A. (2004) Effects of understory foliage on patterns of light attenuation
near the forest floor. Biotropica, 36, 33-39.

Montgomery, R.A. & Chazdon, R.L. (2002) Light gradient partitioning by tropical tree
seedlings in the absence of canopy gaps. Oecologia, 131, 165-174.

Newell, E.A., Mulkey, S.S., & Wright, S.J. (2002) Seasonal patterns of carbohydrate
storage in four tropical tree species. Oecologia, 131, 333-342.

Osunkoya, 0.0., Ash, J.E., Hopkins, M.S. & Graham, A.W. (1994) Influence of seed size
and seedling ecological attributes on shade-tolerance of rain-forest tree species in
northern Queensland. Journal ofEcology, 82, 149-163.









Osunkoya, 0.0., Ash, J.E., Hopkins, M.S., & Graham, A.W. (1992) Factors affecting
survival of tree seedlings in North Queensland rainforests. Oecologia, 91, 569-578.

Pacala, S.W., Canham, C.D., Saponara, J., Silander, J.A., Kobe, R.K. & Ribbens, E.
(1996) Forest models defined by field measurements: Estimation, error analysis and
dynamics. Ecological Monographs, 66, 1-43.

Poorter, L. (1999) Growth responses of 15 rain-forest trees to a light gradient: The
relative importance of morphological and physiological traits. Functional Ecology,
13, 396-410.

Rand, A.S. & Rand, W.M. (1982) Variation in rainfall on Barro Colorado Island. The
Ecology ofa Tropical Forest (eds E.G. Leigh, Jr., A.S. Rand & D.M. Windsor), pp.
47-66. Smithsonian Institution Press, Washington, DC.

Rose, S.A. & Poorter, L (2002) The importance of seed mass for early regeneration in
tropical forest: A review. Long term changes in Composition and Diversity: case
studies from the Guyana Va\l/,/, Africa, Borneo and Melanesia (ed H. ter Steege),
pp. 5-19. Tropenbos Foundation, Wageningen, The Netherlands.

Scariot, A. (2000) Seedling mortality by litterfall in Amazonian forest fragments.
Biotropica, 32, 662-669.

Tissue, D.T. & Wright, S.J. (1995) Effects of seasonal water availability on phenology
and the annual shoot carbohydrate cycle of tropical forest shrubs. Functional
Ecology, 9, 518-527.

Vazquez-Yanes, C., Orozco-Segovia, A., Rincon, E., Sanchez-Coronado, M.E., Huante,
P., Toledo, J.R., & Barradas, V.L. (1990) Light beneath the litter in a tropical
forest: Effect on seed germination. Ecology, 71, 1952-1958.

Veneklaas, E.J. & Poorter, L. (1998) Growth and carbon partitioning of tropical tree
seedlings in contrasting light environments. Inherent Variation in Plant Gi 1,11 th.
Physiological Mechanisms and Ecological Consequences (eds H. Lambers, H.
Poorter & M.M.I. Van Vuuren), pp.337-361. Backkhuys Publishers, Leiden, The
Netherlands.

Walter, M.B. & Reich, P.B. (2000) Seed size, nitrogen supply, and growth rate effect tree
seedling survival in deep shade. Ecology, 81, 1887-1901.

Walters, M.B. & Reich, P.B. (1999) Low-light carbon balance and shade tolerance in the
seedlings of woody plants: Do winter deciduous and broad-leaved evergreen
species differ? New Phytologist, 143, 143-154.

Walters, M.B. & Reich, P.B. (1996) Are shade tolerance, survival, and growth linked?
Low light and nitrogen effects on hardwood seedlings. Ecology, 77, 841-853.






62


Wang, Y.H. & Augspurger, C.K. (2004) Dwarf palms and cyclanths strongly reduce
Neotropical seedling recruitment. Oikos, 107, 619-633.

Welden, C.W., Hewett, S.W., Hubbell, S.P. & Foster, R.B. (1991) Sapling survival,
growth, and recruitment: Relationship to canopy height in a neotropical forest.
Ecology, 72, 35-50.

Woodward, F.I. (1990) From genes to ecosystems: The importance of shade tolerance.
Trends in Ecology and Evolution, 5, 111-115.

Wright, S.J. (2002) Plant diversity in tropical forests: A review of mechanisms of species
coexistence. Oecologia, 130, 1-14.















BIOGRAPHICAL SKETCH

Jonathan's interest in the natural world began when he was a boy growing up in the

forests and fields of western New York State. His formal education in the natural

sciences began during his junior year in high school, where he enrolled in and completed

a 2-year vocational program in natural resources conservation. Jonathan went on to

pursue a 2-year degree in forestry from Paul Smith's College, Paul Smiths, New York,

graduating in May of 1999. As part of his Paul Smith's education, he participated in a

tropical biology field course in Belize, which sparked his interest in both tropical

ecosystems and a professional career in ecology. He continued his undergraduate

education at Cornell University, Ithaca, New York, majoring in biological sciences, with

a concentration in ecology and evolutionary biology. He graduated with a B.S. degree in

May of 2002 and was awarded high honors for an undergraduate thesis on seed dispersal

by white-tailed deer, later published in Oecologia (2004). Jonathan spent the following

summer working as research assistant for the Institute of Ecosystem Studies, studying

hurricane effects on seedling regeneration at the Luquillo Experimental Forest in Puerto

Rico.

Jonathan began his graduate work in the fall of 2002 in the Department of Botany

at the University of Florida, Gainesville, Florida, studying tropical rainforest seedling

ecology and ecophysiology. During his second semester, he participated in an 8-week

tropical ecology field course in Costa Rica, hosted through the Organization for Tropical

Studies at Duke University. He then traveled northwest to Panama to begin the field






64


research for his M.S. thesis, which focused on the role of nonstructural carbohydrate

reserves in seedling growth, survival, and stress tolerance in a tropical wet forest. He

received his M.S. degree in interdisciplinary ecology, with a concentration in botany, in

May of 2005. Jonathan is currently working towards a Ph.D. in tropical forest

community ecology in the Department of Biological Sciences at Louisiana State

University, Baton Rouge, Louisiana, and plans to pursue an academic career in research

and teaching.




Full Text

PAGE 1

SEEDLING CARBOHYDRATE STORAGE, SURVIVAL, AND STRESS TOLERANCE IN A NEOTROPICAL FOREST By JONATHAN ANDREW MYERS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Jonathan Andrew Myers

PAGE 3

I dedicate this thesis to my parents, Roy and Susan Myers, whose support and encouragement of my pursuits in ecology know no bounds.

PAGE 4

ACKNOWLEDGMENTS I would first like to thank my advisor, Kaoru Kitajima, for her unwavering mentorship in both academia and field research. Her passion for seedling ecology and dedication to strong work ethic is constantly inspiring. I would also like to thank my committee member, Jack Putz, for his critical and insightful suggestions on all aspects of the thesis and for his overall guidance in my development as an ecologist. Numerous individuals have generously contributed field and laboratory assistance. I would like to thank Marta Vargas for her assistance with data collection throughout the entirety of the project, Roberto Cordero for his help and advice during the early stages of project development, Sebastian Bernal for his spirited assistance in the field and laboratory (and for keeping us dry and motivated in even the heaviest of rains), and Momoka Yao for her help with many of the laboratory measurements and analyses. I would also like to thank the following undergraduate students for assistance with data collection: Sarah Tarrant, Lisa Cowart, and Sergio Montes. Laboratory assistance was also provided by Zachary Tyser, who participated in the project through the 46th Annual Pre-collegiate Summer Science Training Program at the University of Florida. Finally, I would like to thank the many friends and colleagues who made my time at the University of Florida more than memorable. I would especially like to thank Teresa Kurtz, Silvia Alvarez, Camila Pizano, Jennifer Schaefer, Joseph Veldman, and Eddie Watkins for constantly reminding me that the beauty of life extends well beyond science. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................ix CHAPTER 1 STRESS TOLERANCE ENHANCES SEEDLING SURVIVAL IN THE UNDERSTORY OF A TROPICAL FOREST: EXPERIMENTAL EVIDENCE.......1 Introduction...................................................................................................................1 Methods........................................................................................................................3 Study Site and Species...........................................................................................3 Field and Laboratory Methods..............................................................................4 Statistical Analyses................................................................................................7 Results...........................................................................................................................9 Seedling Survival...................................................................................................9 Effects of Defoliation and Light Reduction on Seedling Biomass and Growth..10 Discussion...................................................................................................................12 Stress Tolerance as a Component of Seedling Survival in Shade.......................12 Seedling Tolerance to Defoliation: The Importance of Storage Reserves..........13 Seedling Tolerance to Temporal Variation in Understory Light Availability....14 2 CARBON ALLOCATION TO STORAGE IN TROPICAL FOREST SEEDLINGS: EFFECTS ON GROWTH, SURVIVAL AND STRESS TOLERANCE..................24 Introduction.................................................................................................................24 Methods......................................................................................................................27 Study Site and Species.........................................................................................27 Field and Laboratory Methods............................................................................28 Quantification of TNC in Stems and Roots.........................................................31 Statistical Analyses..............................................................................................32 v

PAGE 6

Results.........................................................................................................................34 Variation in Pretreatment Biomass and TNC Reserves......................................34 Effects of Defoliation and Deep Shade on Biomass and TNC Reserves............35 Influence of Seedling Size, Growth, and TNC Reserves on Seedling Survival..36 Discussion...................................................................................................................38 Role of TNC Reserves in Seedling Survival and Stress Tolerance.....................38 Tradeoffs between Seedling Growth and Storage...............................................40 TNC Allocation, Seed Size, and Seedling Functional Morphology....................41 Conclusion...........................................................................................................42 APPENDIX A SUMMARY STATISTICS FOR SEEDLING BIOMASS AND LEAF AREA........52 B SUMMARY STATISTICS FOR TOTAL NONSTRUCTURAL CARBOHYDRATE (TNC) RESERVES IN STEMS AND ROOTS........................55 LIST OF REFERENCES...................................................................................................57 BIOGRAPHICAL SKETCH.............................................................................................63 vi

PAGE 7

LIST OF TABLES Table page 1-1 Seed and seedling characteristics of the study species.............................................17 1-2 Proportional hazards (Cox regression) survival time analysis of seedling survival over 1 year..................................................................................................18 1-3 Analysis of variance for total seedling biomass.......................................................18 2-1 Analysis of variance for stem and root biomass......................................................43 2-2 Analysis of variance for TNC concentrations and pool sizes..................................44 A-1 Statistics for preand post-treatment biomass and leaf area....................................52 A-2 Statistics for total seedling biomass and leaf area 1 year after treatment.................54 B-1 Statistics for preand post-treatment TNC concentrations and pool sizes...............55 vii

PAGE 8

LIST OF FIGURES Figure page 1-1 Effects of light reduction and defoliation on 1 st -year seedling survival...................19 1-2 Two-month survival of stress treatment seedlings plotted against 1 st -year survival of control treatment seedlings....................................................................20 1-3 Effects of light reduction and defoliation on seedling size and growth 2 months after treatment..........................................................................................................21 1-4 Leaf area recovery 2 months after defoliation.........................................................22 1-5 Effects of light reduction and defoliation on relative growth rate (RGR) from 2 months to 1 year.......................................................................................................23 2-1 Pretreatment biomass and TNC concentrations.......................................................45 2-2 Three potential correlates of interspecific variation in 1 st -year control seedling survival.....................................................................................................................46 2-3 Pretreatment TNC pool sizes in stems and roots......................................................47 2-4 Post-treatment biomass and TNC in stems and roots...............................................48 2-5 Post-treatment cotyledon biomass............................................................................49 2-6 Two-month relative growth rate (RGR) of control seedlings plotted against pretreatment TNC pool size.....................................................................................50 2-7 Survival of stressed seedlings plotted against total TNC pool size and total seedling biomass......................................................................................................51 viii

PAGE 9

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SEEDLING CARBOHYDRATE STORAGE, SURVIVAL, AND STRESS TOLERANCE IN A NEOTROPICAL FOREST By Jonathan Andrew Myers May 2005 Chair: Kaoru Kitajima Major Department: Botany I investigated the role of stem and root total nonstructural carbohydrate (TNC) reserves in seedling growth, survival, and stress tolerance in the shaded understory of a tropical moist forest in central Panama. Seven woody species that ranged widely in seedling shade tolerance were selected for study: Aspidosperma cruenta and Lacmelia panamensis (Apocynaceae), Coussarea curvigemnia (Rubiaceae), Callichlamys latifolia and Tabebuia rosea (Bignoniaceae), Castilla elastica (Moraceae), and Platypodium elegans (Fabaceae). Seedlings of each species were raised from seed in a shade house, transferred to common-gardens enclosed by wire fencing in the shaded forest understory (ca. 1% full sun), and then randomly assigned to one of three experimental treatments: extreme light reduction (ca. 0.06% of full sun for 2 months), complete defoliation at 0 months, and control (no manipulation). The goal of the light-reduction and defoliation stress treatments was to force seedlings into negative carbon balance, to determine the role of TNC reserves for seedling survival over the period of 1 year. ix

PAGE 10

First-year survival of control seedlingswhich I used as a quantitative index of shade tolerancevaried significantly among species, ranging from 96% in Aspidosperma to 19% in Tabebuia. The light-reduction and defoliation treatments caused significant decreases in seedling survival. Two-month survival after the stress treatments was positively correlated with 1 st -year control seedling survival among species. Seedlings of the two species with lowest 1 st -year control seedling survival (Tabebuia and Platypodium) were unable to survive either stress treatment past 60 days. Pretreatment TNC concentrations and pool sizes in stems and roots differed significantly among species. Stem concentrations were higher than root concentrations in all species, and comprised between 6 to 25% of the total stem biomass. Variation in 1 st -year control seedling survival was positively related to pretreatment TNC pool size (combined stem and root pools) among species; but not to pretreatment stem or root TNC concentrations, total seedling biomass, or cotyledon biomass. Survival of the stress treatment seedlings during the first 2 months (and from 2 months to 1 year) was also positively related to interspecific variation in TNC pool size. Relative growth rates of control seedlings from 0 to 2 months were negatively related to TNC pool size among species, suggesting a tradeoff between carbon allocation to growth versus storage. My results highlight the importance of TNC reserves for seedling survival and stress tolerance in the shaded understory of tropical forests. Allocation-based tradeoffs between seedling growth and storage provide a mechanistic basis for species differences in shade tolerance; and have likely played important roles in the evolution of life-history diversity and maintenance of tree-species diversity in species-rich tropical tree communities. x

PAGE 11

CHAPTER 1 STRESS TOLERANCE ENHANCES SEEDLING SURVIVAL IN THE UNDERSTORY OF A TROPICAL FOREST: EXPERIMENTAL EVIDENCE Introduction Differential seedling survival among coexisting species has long been considered to play important roles in determining tree-species composition and diversity in tropical forests (Janzen 1970; Connell 1971; Hubbell 1979; Harms et al. 2000; Wright 2002). Seedling mortality is especially high in the shaded forest understory (Augspurger 1984a,b; Osunkoya et al. 1992; Li et al. 1996) where light availability is low (ca. 1% of full sun) and variable in time and space (Chazdon and Fetcher 1984; Montgomery and Chazdon 2002; Montgomery 2004). Under these conditions, low potential for photosynthetic carbon gain limits the ability of seedlings to maintain positive carbon balance, a prerequisite for survival in shade. Diurnal and seasonal fluctuations in understory light availability (Chazdon and Fetcher 1984; Chazdon and Pearcy 1986) and/or deep shading of seedlings by fallen litter or understory foliage (Vzquez-Ynes et al. 1990; Wang and Augspurger 2004; Farris-Lopez et al. 2004; Montgomery 2004) further constrain carbon balance by forcing seedlings to cope with stressful periods, when light levels drops below the photosynthetic compensation point. In addition to light-limitation, seedlings risk mortality when tissues are lost or damaged by disease (Augspurger 1984a), herbivores (Osunkoya et al. 1992; Asquith et al. 1997; Kitajima 2004), and physical disturbance (Clark and Clark 1989; Scariot 2000). Identifying the extent and factors that contribute to species differences in their ability to cope with 1

PAGE 12

2 stresses due to light limitation and tissue loss is essential for predicting and evaluating community-level patterns of seedling recruitment and relative species abundance. Survival in light-limited habitats (shade tolerance) is contingent on the ability of seedlings to maintain positive carbon balance. Some researchers have hypothesized that shade tolerance is achieved by morphological and physiological traits that maximize the rate of carbon capture in low light at the whole-plant level (Loach 1967; Givnish 1988; Woodward 1990; Walters and Reich 1996). According to this hypothesis, fast rates of net carbon gain facilitate both high seedling growth rates and enhanced survival in shade, by allowing seedlings to quickly increase in size and out-compete slower growing neighbors (Walters and Reich 2000). In contrast, others researchers have hypothesized that fast growth in shade could have deleterious consequences on survival, because high carbon allocation to growth must occur at the expense of allocation to storage and defense (Kitajima 1994, 1996; Coley et al. 1985; Kobe et al. 1995; Kobe 1997; Veneklaas and Poorter 1999), thereby reducing seedling tolerance and/or resistance to biotic and abiotic stresses. However, few studies have directly tested whether interspecific variation in seedling survival and growth in the shaded forest understory is linked to species-specific differences in seedling stress tolerance (e.g., Augspurger 1984a). Quantifying the degree to which different species are able to tolerate and recover from abiotic and biotic stresses (especially those posed by light-limitation and tissue loss) could help to clarify the functional traits underlying seedling shade tolerance and life history diversity in the seedling community of species-rich tropical forests.

PAGE 13

3 In this study, I examined seedling stress tolerance, growth, and survival among seven coexisting woody species in a tropical moist forest in central Panama. The overarching goal of my study was to test whether interspecific variation in seedling shade tolerance, measured quantitatively as 1 st -year seedling survival in the shaded understory (see also Augspurger 1984a,b; Kitajima 1994; Boot 1996), was linked to variation in survival and growth responses to stresses caused from light limitation and tissue loss. I tested the following two hypotheses: (1) interspecific variation in 1 st -year seedling survival in the shaded understory is positively correlated with the ability to survive stress due to light limitation and tissue loss; and (2) seedling growth is less affected by stress in species with higher 1 st -year seedling survival. To test these hypotheses, I conducted a 1-year common-garden field experiment in the shaded understory (~ 1% full sun), using transplanted seedlings of an equivalent ontogenetic stage, defined by expansion of the first photosynthetic organs (leaves or cotyledons). Growth and survival were monitored on seedlings assigned to one of three treatments: (1) extreme light reduction; (2) complete defoliation; (3) control (no manipulation). The goal of the light-reduction treatment was to expose seedlings to light levels below those required for whole-plant carbon balance, while the purpose of the defoliation treatment was to force seedlings to rely on storage reserves after removal of photosynthetic leaves and cotyledons. Methods Study Site and Species The study was conducted on Buena Vista peninsula, an area of ca. 60-year-old secondary lowland tropical forest, located in the Barro Colorado Nature Monument (BCNM), Panama (910 N, 7951 W). The BCNM forest is semi-deciduous, with a

PAGE 14

4 pronounced 4-month dry season that usually lasts from mid-December to mid-April. Annual rainfall on the BCNM averages 2,700 mm (Rand and Rand 1982). Seven woody species with sufficient seed availability were chosen for study (Table 1-1). All of the species are canopy trees as adults, except for Callichlamys latifolia (Bignoniaceae), which is a woody liana. Species were selected to span a range of shade tolerance, which was quantified by estimating the percentage of seedlings surviving for 1 year in experimental plots located in the shaded understory (Table 1-1). Small-seeded, pioneer species that are generally unable to establish seedling populations in the shaded forest understory were deliberately excluded from the study. The seven species differ in seed mass (24 to 453 mg) and cotyledon functional morphology (Table 1-1). Aspidosperma cruenta, Lacmelia panamensis, Castilla elastica, and Platypodium elegans have hypogeal cotyledons that remain below or just above the soil surface, and that function primarily for storage of seed reserves. The three other species have epigeal cotyledons that become elevated above the ground after germination; Coussarea curvigemnia and Tabebuia rosea have foliaceous photosynthetic cotyledons, and Callichlamys latifolia has thick green cotyledons that function primarily for storage (Kitajima 2002). For brevity, species are henceforth referred to by genus. Field and Laboratory Methods Depending on seed availability, 200 to 400 seedlings per species were raised from seed in a shade house on the BCNM, under a light level similar to that found in the forest understory (1.5% full sun, based on total daily photon flux density). When possible, seeds were collected from at least 3 nonadjacent parent trees, to increase genetic diversity within each species. Seedlings were raised in plastic trays containing a 1:1 mixture of forest soil and sand, until all seedlings reached an equivalent ontogenetic stage, defined

PAGE 15

5 by the full expansion of the first photosynthetic organs (leaves or cotyledons, depending on the species; Table 1-1), at which time they were transplanted into the forest (described below). Seedlings were transplanted at a standardized ontogenetic stage, rather than at a common chronological age, so that all species had initiated autotrophic growth by the onset of the study. Because of interspecific variability in timing of seed dispersal, germination, and development, seedlings were transplanted over a 3-month period during the early-mid rainy season (May to July). Seedlings of all species were transplanted into four 7 x 7 m 2 common-gardens located in the shaded forest understory on Buena Vista Peninsula. Each garden was separated by at least 50 m from the nearest adjacent garden, and enclosed by 1 m tall fencing to exclude large herbivores. For each species, equal numbers of seedlings were transplanted into 1 x 1 m 2 cells in four equally sized, stratified sections in each garden (50 to 100 total seedlings garden -1 species -1 ). All vegetation < 1.5 m tall was removed from each garden before seedling transplantation, and all seedlings were positioned a minimum of 25 cm apart. When necessary, leaf litter was also removed to avoid burial of seedlings. Seedlings dying within the first 2 weeks after transplantation were replaced (< 10 seedlings total). Seedlings were randomly assigned to one of three treatments 2 weeks after transplantation: (1) defoliation; (2) light reduction; (3) control (no manipulation). For the defoliation treatment, all leaves and photosynthetic cotyledons (if present) were clipped at the base of the petiole; cotyledons were retained on all of the species with hypogeal storage cotyledons (Table 1-1). For the light-reduction treatment, cylindrical wire cages covered with 90% shade cloth were placed over individual seedlings. Shade cage

PAGE 16

6 dimensions ranged from 6 x 12 cm (height x diameter) to 15 x 30 cm (depending on initial seedling size); and were constructed to be large enough so as not to hinder growth over a 2-month period, after which the shade cages were removed. Light transmission was measured inside and outside of the shade cages with Li-Cor quantum sensors during the early-mid rainy season. Percent light transmission underneath the cages, and within each common-garden, was determined using both instantaneous and continuous total daily photon flux density (PFD) measurements. Instantaneous PFD measurements were taken inside and outside of shade cages at 8 to 15 stratified positions in each garden under mid-day, cloudy sky conditions. Continuous PFD was measured by placing a single censor in the middle of each common-garden for 2 days. To calculate % light transmission, measurements within the forest were referenced to a sensor placed in completely open sky on a laboratory rooftop. Mean light reduction ( + 1 SD) by the shade cages was 89 ( + 2)%. Mean light levels outside of the shade cages, as determined from the instantaneous PFD measurements, ranged among gardens from 0.38 to 0.68% (overall mean = 0.55%); mean light levels from the continuous PFD measurements taken over 2 full days in each garden were higher, and ranged among gardens from 0.68 to 1.44% of full sun (overall mean = 1%). Seedlings inside the shade cages thus experienced only ca. 0.06% of full sun over the 2-month light-reduction treatment. To estimate preand post-treatment biomass, I harvested randomly selected sub-samples of seedlings from each common-garden. For the pretreatment samples, 6 to 12 seedlings per species were harvested from each common-garden just prior to treatment. Post-treatment samples were harvested 2 months (coinciding with the end of

PAGE 17

7 the light-reduction treatment) and 1 year after treatment; samples sizes at these harvests ranged from 2 to 32 total seedlings per species x treatment combination, depending on species-specific differences in survival. Seedlings were harvested by carefully excavating the root system with a small trowel. Roots, stems, leaves, and cotyledons were separated within 12 hours after harvesting, and then oven dried at 100C for 1 hour. All samples were then dried for an additional 48 hours at 60C before weighing. Leaf and cotyledon area (for epigeal cotyledons only) were measured before drying using a Li-Cor leaf area meter. Seedling survival was measured by conducting censuses every week for the first 2 months, and then every 2 weeks from 2 months to 1 year. Missing seedlings were treated as dead. Seedlings killed by branch fall and locally heavy herbivory (e.g., outbreak of stem cutting insects in one common-garden) were excluded from the study. Statistical Analyses Differences in survival among species and treatments were analyzed using semi-parametric proportional hazards modeling (i.e., Cox regression; Fox 2001). Cox regression was selected over alternative parametric tests because it does not assume a particular probability distribution for survival times (the non-parametric component of the model; Fox 2001), which varied widely among species and treatments. Seedlings harvested for biomass measurements and those that survived beyond the final survival census were right-censored in the analysis. I then used the Kaplan-Meier approach to estimate the proportion of seedlings surviving over time for each species x treatment combination. Differences between Kaplan-Meier estimates among species and among treatments within species were analyzed using log-rank tests, which tests for heterogeneity in survival times among groups by comparing observed versus expected

PAGE 18

8 number of deaths between each mortality event (Fox 2001). Percent survival was estimated from 0 to 2 months (i.e., the time of shade cage removal), 2 months to 1 year (stress treatments only), and from 0 to 1 year (control treatment only) from the Kaplan-Meier survival curves. Survival data were pooled across common-gardens for all analyses and garden was included as a main factor in the Cox regression model. Differences among species and treatments in total seedling biomass at 2 months and 1 year were analyzed using ANOVA. In both analyses, common-garden means were used as replicates for each species x treatment combination (n = 1 per species x treatment combination; Appendix A). Because all seedlings had died before the 2-month harvest for some species x treatment combinations, I used two separate ANOVA models to analyze seedling biomass. In the first model (saturated model), I tested for species, treatment, and species x treatment interaction effects on biomass, but only included data for the species that had living seedlings present in all treatments at the time of harvest (n = 5 species at 2 months; n = 4 species at 1 year). In the second model (Reduced main factor model), I included data for all 7 species and tested for the main effects of species and treatment on biomass, but excluded species x interaction terms. Data from the 1-year harvest were log 10 -transformed before analysis to improve normality and homogeneity of variance. Simple least squares linear regression was used to test for correlations between interspecific variation in seedling survival, growth, and size. Relative growth rates (RGR) of total seedling biomass, stem height, and leaf area were estimated using the following formula: RGR (wk -1 ) = [ln (X2) ln (X1)] / time

PAGE 19

9 where X 2 = mean size at time 2; X 1 = mean size at time 1; and time = number of weeks between times 1 and 2. RGR for LACP from 0 to 2 months could not be calculated due to missing pretreatment biomass samples. Leaf and cotyledon mass were included in all RGR calculations. Cotyledon area was included in all leaf area calculations for species with photosynthetic cotyledons (Coussarea and Tabebuia). All statistical analyses were performed using JMP software (SAS Institute, 1997). Results Seedling Survival Seedling survival differed substantially among species and treatments (Table 1-2; Figure 1-1). First-year survival of control seedlings differed significantly among species (P < 0.0001 for log-rank test of survival time distributions), and ranged from 19 to 96% at the end of one year (Figure 1-1; Table 1-1). For control seedlings, survival from 0 to 1 year was positively correlated with survival from 0 to 2 months among species (P < 0.0001, r 2 = 0.96, n = 7), when mortality was generally highest for most species. Thus, species maintained their survival ranks from 2 months to 1 year. Survival also differed significantly among common-gardens (Table 1-2), and was generally higher in gardens with higher light levels, even though the range in light levels among gardens was small (0.6.4% full sun). Defoliation and light reduction significantly reduced survival in all species (Figure 1-1; P < 0.0001 for log rank tests), and there was a significant interaction between species and treatment on survival times (Table 1-2). Species that had high survival in the light-reduction treatment also had high survival in the defoliation treatment (P = 0.0001, r 2 = 0.94, n = 7 for proportional survival from 0 2 months). For

PAGE 20

10 some species (Aspidosperma, Coussarea, Callichlamys) defoliation had a larger overall impact on survival than light reduction, while others showed similar survival in the two stress treatments (Figure 1-1). The impact of the stress treatments on seedling survival varied widely among species (Figure 1-1). Survival in the stress treatments from 0 to 2 months was positively correlated with 1 st -year control seedling survival among species (Figure 1-2). Two-month survival for the three species with highest 1 st -year survival exceeded 85% in both stress treatments. In contrast, of the four species with the lowest 1 st -year survival of control seedlings, three had reached 100% seedling mortality in one or both of the stress treatments before 60 days. There were no significant relationships between seedling survival and seed mass or pretreatment cotyledon mass among species for any treatment (P > 0.2 for all regressions), even though the species with the largest seed size (Aspidosperma) had the highest survival in all treatments at 2 months, and the smallest-seeded species (Tabebuia) the lowest survival after defoliation and the second lowest survival after light reduction. Effects of Defoliation and Light Reduction on Seedling Biomass and Growth Total seedling biomass at 2 months differed significantly among species and treatments (Figure 1-3a; Table 1-3). The effect of the stress treatments on seedling biomass varied widely among species (Figure 1-3a) and there was a significant interaction between species and treatment on biomass at 2 months, but not at 1 year (Table 1-3). Whole-seedling relative growth rate (RGR) from 0 to 2 months also varied widely among species and treatments (Figure 1-3b). Control seedling RGR ranged 2.5 fold between species and was highest for Tabebuia (which had the lowest 1 st -year control seedling survival) and lowest for Aspidosperma (highest 1 st -year survival), but there was

PAGE 21

11 no significant correlation between control seedling survival and RGR among species (P > 0.4, r 2 = 0.16, n = 6). Two-month survival of control seedlings was, however, negatively correlated to both stem height and leaf area RGR among species (P = 0.01, r 2 = 0.77 for stem height; P = 0.03, r 2 = 0.68 for log 10 leaf area). Light reduction had a noticeable effect on RGR, which varied 5.5 fold among species at the end of 2-month treatment (Figure 1-3b). RGR in the light-reduction treatment was less affected (relative to the control treatment) for species with higher 1 st -year control seedling survival: RGR was reduced by 0.013 to 0.015 wk -1 for the two species with higher 1 st -year survival (Aspidosperma and Coussarea) compared to 0.042 to 0.050 wk -1 in the three species with lower 1 st -year survival (Callichlamys, Castilla, and Tabebuia; no Platypodium seedlings survived the light-reduction treatment). As with control seedlings, whole-seedling RGR was not a good correlate of survival among species in the light-reduction treatment (P > 0.7, r 2 = 0.04, n = 5). However, the three species with the lowest survival exhibited the highest whole-seedling RGR in the control treatment, and among species, survival in the light-reduction treatment was negatively correlated to height RGR of control seedlings (P = 0.06, r 2 = 0.59, n = 6). Whole-seedling RGR during the light-reduction treatment was highest for the three smallest-seeded species, all of which had epigeal cotyledons (Figure 1-3b; Table 1-1). RGR following defoliation was greater for species that were able to produce new leaves (Figure 1-3b,c). Four of the seven species, including the three species with highest 1 st -year control seedling survival, were able to develop new and fully expanded leaves within 2 months (Figure 1-3c). Furthermore, species with relatively higher 1 st -year survival produced more new leaf area following defoliation than was produced by control

PAGE 22

12 seedlings over the same 2-month period (Figure 1-5): leaf production for Aspidosperma, the species with the highest 1 st -year survival, was six times greater than in controls. In contrast, leaf production in Castilla, which had the lowest 1 st -year survival among the species that produced new leaves, was lower relative to controls. It is also worth noting that leaf production following defoliation was not restricted to species with relatively large seed mass or hypogeal storage cotyledons. Coussarea, despite having the second smallest seed mass and photosynthetic epigeal cotyledons (Table 1-1), was able to recover more leaf area relative to control seedlings following defoliation than Castilla (Figure 1-5), which had three-fold higher seed mass and hypogeal cotyledons. Stress treatments had significant negative effects on total seedling biomass at 1 year (Table 1-3). RGR from 2 months to 1 year varied widely among species and treatments (Figure 1-6). The three species with the highest 1 st -year control seedling survival (Aspidosperma, Lacmelia, Coussarea) experienced greater reduction in RGR by stress relative to controls, while the species with the lowest 1 st -year survival (Callichlamys, Castilla) showed overcompensation of RGR. Discussion Stress Tolerance as a Component of Seedling Survival in Shade My results support the hypothesis that enhanced seedling survival in the shaded understory is related to the ability of species to tolerate stress. Among species, survival in both the defoliation and light-reduction treatments was strongly related to interspecific variation in 1 st -year control seedlings survival (i.e., shade tolerance). The ability to tolerate stress also varied widely among species: the two species with the lowest 1 st -year survival reached 100% mortality in both stress treatments before 60 days, while the three species with the highest 1 st -year survival showed substantial tolerance to both treatments

PAGE 23

13 (> 85% survival). The stress tolerance of the latter three species was considerable given the severity of the experimental stresses imposed upon them: only 0.06% of full sun for 2 months or complete defoliation. These results support the notion that shade tolerance is dependent upon functional traits that both allow seedlings to maintain a positive carbon balance in low light, and to survive through and recover from periods when they experience negative carbon balance due to stress. Survival following the defoliation and light-reduction treatments was highly correlated among species, indicating that the most stress tolerant species are able to cope with a wide range of biotic and abiotic hazards. Earlier studies have also shown negative interspecific correlations between 1 st -year survival of tropical tree seedlings and susceptibility to pathogens in shade (Augspurger and Kelly 1983; Augspurger 1984a,b). In concert, these results suggest that the traits conferring stress tolerance may be broad spectrum, potentially allowing seedlings to cope with other stresses caused by drought (Engelbrecht and Kursar 2003; Khurana and Singh 2004), nutrient limitation (Gunatilleke et al. 1997), and physical/mechanical damage due to litterfall (Clark and Clark 1989; Scariot 2000). Seedling Tolerance to Defoliation: The Importance of Storage Reserves The goal of the defoliation treatment was to force seedlings to temporarily rely on storage reserves for maintenance of a positive carbon balance and subsequent leaf tissue recovery. My results show that post-defoliation seedling survival was dependent on both the capacity of a species to develop and fully expand new leaves, and the relative degree of leaf area recovery among species. Total leaf area and relative leaf area recovery was most impacted in species with lower 1 st -year control seedling survival, confirming my initial hypothesis. The three species that were unable to develop new leaves did not

PAGE 24

14 survive past 60 days, while survival among the other four species was highest for species with greater leaf area recovery. For example, leaf area production in Aspidosperma, the species with the highest post-defoliation survival over the first 2 months, was 64 times greater than in Castilla (relative to leaf production in control seedlings), which had 50% lower survival. Recovery of photosynthetic tissue lost to herbivores, disease, or disturbance is ultimately dependent on the amount of carbon reserves within remaining vegetative tissues and storage cotyledons at the time of tissue loss. In this study, three of the four species that were able to develop new leaves following defoliation had hypogeal storage cotyledons, consistent with the idea that large seed reserves may enhance seedling recovery following tissue damage (Harms and Dalling 1997; Green and Juniper 2004). Leaf area recovery, however, was not restricted to species with this particular cotyledon functional morphology. Coussarea, which had epigeal photosynthetic cotyledons, as well as other epigeal-photosynthetic species in this forest, are also capable of substantial leaf area recovery following complete defoliation (Kitajima 2004). For species with epigeal cotyledons, storage reserves within stem and roots in the form of total nonstructural carbohydrate (TNC), provide the carbon source necessary for tissue recovery following damage (McPherson and Williams 1998; Canham et al. 1999; Hoffman et al. 2004). TNC reserves in stems and roots may play particularly important roles in seedling survival after seedlings have depleted seed reserves, or when storage cotyledons are prematurely lost to herbivores or disease (Kitajima 2004; Chapter 2). Seedling Tolerance to Temporal Variation in Understory Light Availability The seven species varied widely in their ability to survive through drastic reductions in light availability in the shaded understory. Three of the species showed

PAGE 25

15 substantial tolerance to light reduction, with > 92% survival during the 2-month treatment. Platypodium suffered the highest mortality during the first 2 months (all seedlings died before 55 days). The low survival of this species was influenced by damping-off disease (Augspurger 1983, 1984a,b), confirming the idea that fungal attack can limit seedling establishment in deeply shaded habitats (e.g., Grime 1965). Reduction in light availability also increased variability among species in seedling growth rates. These results support the notion that small-scale heterogeneity in light availability, even in the closed forest understory, can have dramatic impacts on seedling growth, survival, and recruitment processes in tropical forests (Montgomery and Chazdon 2002; Wang and Augspurger 2004; Montgomery 2004). Interspecific variation in 2-month seedling survival in both the light-reduction and control treatments was not related to whole-seedling RGR. However, species that maintained high RGR in the control treatment also tended to have the lowest survival in the light-reduction treatment, suggesting that the overall carbon balance of species that normally exhibit fast growth rates may be most affected by temporal reductions in light availability in the shaded understory. Post-light reduction RGR from 2 months to 1 year, following removal of the shade cages at 2-months, was higher relative to control seedlings for species with lower 1 st -year control seedling survival. This result suggests that overcompensation of RGR during seedling recovery from stress can have negative consequences on survival, potentially because increased allocation to traits that maximize growth rates come at the cost of allocation to traits such as defense and storage that enhance survival (Kitajima 1994; 1996; Kobe 1997). The functional mechanisms that

PAGE 26

16 allow seedlings to acclimate to short-term deficits and increases in light availability in the shaded understory deserve future study.

PAGE 27

Table 1-1. Seed and seedling characteristics of the study species. SpeciesFamilyCodeCotyledon function Seed mass(position)nMean (%)SE(mg)2Aspidosperma cruentaApocynaceaeASPCStorage (hypogeal)11196.61.9453Lacmelia panamensisApocynaceaeLACPStorage (hypogeal)7891.83.5244Cousarrea curvigemniaRubiaceaeCOUCPhotosynthesis (epigeal)10789.13.497Callichlamys latifoliaBignoniaceaeCALLStorage (epigeal)5579.96.1198Castilla elasticaMoraceaeCASEStorage (hypogeal)10157.65.6314Platypodium elegansFabaceaePLAEStorage (hypogeal)4637.98.7332Tabebuia roseaBignoniaceaeTAB R Photosynthesis (epigeal)7119.35.9241Survival data from this study (= 1 year survival of control treatment seedlings in the shaded forest understory)2Estimated from 6 to 37 dried seeds per species, after removing seed coats1st-year survival in shade1 17

PAGE 28

18 Table 1-2. Proportional hazards (Cox regression) survival time analysis of seedling survival over 1 year. P = significance from Walds 2 test statistic. able 1-3. Analysis of variance for total seedling biomass 2 months and 1 year after Variable VariabledfWal d 2 PSpecies6806.9<0.0001Treatment2389.3<0.0001Species x Treatment1273.2<0.0001Garde n 393.7<0.0001 T treatment. 2-month biomass1-year biomass dfF P dfF P Saturated model1 Species496.1<0.0001365.2<0.0001 Treatment223.9<0.0001215.3<0.0001 Species x Treatment82.20.0460.70.62 Error5745Reduced main factor mode l Species665.3<0.0001648.4<0.0001 Treatment224.9<0.0001216.4<0.0001 Error67561The following species were removed from the saturated models because of 100% mortality in some treatments (see Methods): Platypodiumand Tabebuia (2-month analysis); Platypodium, Tabebuia, Callichlamys(1-year analysis).

PAGE 29

Aspidosperma cruentum Proportion surviving 0.00.20.40.60.81.0 Control Light reduction Defoliation Lacmelia panamensis Coussarea curvigemnia Callichlamys latifolia 04080120160200240280320360 Castilla elasticaTime (days) 04080120160200240280320360 0.00.20.40.60.81.0 Platypodium elegans 04080120160200240280320360 Tabebuia rosea 04080120160200240280320360 19 Figure 1-1. Effects of light reduction and defoliation on 1 st -year seedling survival. The proportion of seedlings surviving was estimated by the Kaplan-Meier method. Species are shown in order of 1 st -year survival of control seedlings as summarized in Table 1-1.

PAGE 30

20 1st-year control seedling survival (%) 020406080100 Stress treatment survival (%) 020406080100 Defoliation: r2 = 0.67, P = 0.02 Light reduction: r2 = 0.75, P < 0.01 Figure 1-2. Two-month survival of stress treatment seedlings plotted against 1 st -year survival of control treatment seedlings. Each point is a species mean. Dashed and solid lines show the best-fit linear regressions for the light-reduction and defoliation treatments, respectively.

PAGE 31

21 CSpecies ASPCLACPCOUCCALLCASEPLAETABR Leaf area (cm2) 0102030405060 A Biomass (mg) 0100200300400 Control Light reduction Defoliation HighestsurvivalLowestsurvival B RGR (wk-1) -0.20-0.15-0.10-0.050.000.05 Figure 1-3. Effects of light reduction and defoliation on seedling size and growth 2 months after treatment. A) Total seedling biomass. B) Relative growth rate (RGR) from 0 to 2 months. C) Leaf area. Vertical bars = +1 SD; n = 1 common-garden replicates per bar (see Methods; Table A-1). Species are ordered from highest to lowest 1 st -year control seedling survival (Table 1-1). Arrows in C) indicate complete lack of new leaf growth. RGR was not calculated for LACP due to missing pretreatment biomass data; otherwise, a missing bar indicates that no seedlings survived a treatment past 2 months.

PAGE 32

22 Species ASPCCOUCCASE Leaf area recovery index -2024681012 Figure 1-4. Leaf area recovery 2 months after defoliation. Data are shown for 3 of the 4 species that developed new leaves following defoliation; leaf area recovery was not calculated for LACP due to missing pretreatment data. Vertical bars = +1 SD; n = 4 common-garden replicates for each species. Leaf area recovery index = (A H A C )/A C where A H = leaf area of defoliated seedlings at 2 months, and A C = leaf area production in control seedlings from 0 to 2 months.

PAGE 33

23 LowestsurvivalHighestsurvival Species ASPCLACPCOUCCALLCASEPLAETABR RGR (wk-1) 0.000.010.020.030.04 Control Light reduction Defoliation Figure 1-5. Effects of light reduction and defoliation on relative growth rate (RGR) from 2 months to 1 year. Shade cages for the light-reduction treatment were removed at 2 months. Vertical bars = +1 SD; n = 2 replicates per bar (Appendix A).

PAGE 34

CHAPTER 2 CARBON ALLOCATION TO STORAGE IN TROPICAL FOREST SEEDLINGS: EFFECTS ON GROWTH, SURVIVAL AND STRESS TOLERANCE Introduction For many tropical trees, successful recruitment into the canopy hinges on survival and long-term persistence of seedlings in the shaded forest understory (e.g., Welden et al. 1991; Connell and Green 2000). Survival in the understory is dependent on the ability of seedlings to both maintain positive net carbon balance under light-limited conditions and cope with a host of additional biotic and abiotic stresses (reviewed in Moles and Westoby 2004; Chapter 1). The capacity of coexisting species to cope with light-limitation and other stresses during seedling establishment ultimately influences patterns of adult tree distributions, species composition, and coexistence (Janzen 1970; Connell 1971; Hubbell 1979; Harms et al. 2000). What are the mechanisms that allow seedlings to maintain carbon balance and tolerate stresses in the shaded understory? Some have hypothesized that enhanced survival in shade (shade tolerance) is characterized by morphological and physiological traits that maximize the rate of net carbon capture in low light at the whole-plant level (Loach 1967; Givnish 1988), thereby increasing seedling size and competitive ability under low light conditions. However, maintenance of a positive carbon balance is not necessarily dependent on maximization of carbon gain, especially if seedlings experience stress caused from tissue loss or temporal fluctuations in resource availability, both of which may severely constrain seedling establishment in tropical forests (Augspurger 24

PAGE 35

25 1984b; Clark and Clark 1989; Asquith et al. 1997; Scariot 2000; Montgomery and Chazdon 2002; Engelbrecht and Kursar 2003; Khurana and Singh 2004; Kitajima 2004; Wang and Augspurger 2004). An alternative hypothesis is that high survival is achieved through traits that allow seedlings to maintain a positive carbon balance necessary for long-term persistence in the understory, and which allow seedlings to cope with and recover from biotic and abiotic stress (Kitajima 1994, 1996). Several studies from both temperate and tropical forests provide strong support for this latter hypothesis, by showing that: (1) interspecific variation in low light seedling and sapling survival can be explained through tradeoffs between fast growth rates and high survival (Kitajima 1994; Kobe et al. 1995; Pacala et al. 1996; Veneklaas and Poorter 1998; Walters and Reich 1999; Kobe 1999); (2) species with seedlings that grow quickly in low light also do so in high light (Osunkoya et al. 1994; Grubb et al. 1996; Poorter 1999), suggesting that fast growth rate is not necessarily an adaptive strategy for increased survival in shade. Collectively, these studies suggest that species do not change growth ranks between low and high light environments and that ecological tradeoffs between growth and survival may contribute to niche partitioning across the forest light gradient. Functional traits that enhance seedling tolerance to stress and that contribute to growth-survival tradeoffs may therefore constitute an important, but previously little-explored aspect of seedling regeneration and diversity in tropical forests. Carbon allocation to storage has been proposed as one mechanism that allows species to tolerate low light and other stresses in the shaded forest understory (Kitajima 1994, 1996; Kobe et al. 1995; Kobe 1997). Energy reserves, in the form of total nonstructural carbohydrate (TNC, total of starch and soluble sugars), enhance recovery of

PAGE 36

26 tissue lost to consumers and disturbance (McPherson and Williams 1998; Canham et al. 1999; Kabeya and Sakai 2003). TNC reserves can also aid in the maintenance of positive carbon balance during periods when light is limited (Bloom et al. 1985) and may be especially important for seedling survival in the shaded understory of tropical forests where light availability can be both extremely low (0.2% full sun) and highly dynamic in time and space (Chazdon and Pearcy 1986; Montgomery and Chazdon 2002; Montgomery 2004). In the tropics, TNC reserves have previously been shown to vary as a function of seasonality and light availability in adult trees (Bullock 1992; Marenco et al. 2001; Newell et al. 2002), aid in seasonal leaf flush and reproduction of understory shrubs and palms (Tissue and Wright 1995; Cunningham 1997; Marquis et al. 1997), and facilitate tissue resprouting of woody plants after fire disturbance in savanna-forest ecosystems (Hoffman et al. 2003, 2004). Despite the many important roles of TNC, apparently no study has quantified interspecific variation in seedling TNC storage in the understory of a tropical forest, and the ecological role of TNC in determining species-specific susceptibility to stress-induced mortality and the associated consequences for community-level patterns of seedling establishment remains unknown. In this study, I examine the role of TNC reserves in the growth, survival, and stress tolerance of seven coexisting woody species in a tropical moist forest in central Panama. I chose species that spanned a wide range of seedling shade tolerance, which as in previous studies (Augspurger 1984a,b; Kitajima 1994; Boot 1996), was measured as 1 st -year seedling survival in the shaded understory. To quantify interspecific variation in TNC storage, I measured TNC concentrations and pool sizes in both stems and roots at standardized ontogenetic stages the time at which all species fully expanded their first

PAGE 37

27 photosynthetic organ (cotyledons or leaves) and 2 months after leaf or cotyledon expansion. I focused on TNC in stems and roots, and not in cotyledons and leaves, because stem and root reserves function as more permanent and stable sites for long-term carbohydrate storage; cotyledon reserves tend to be short-lived and most important for seedlings before they have initiated autotrophic growth, while reserves in leaf tissue represent both short-term and temporally dynamic storage sites for carbohydrates. I tested three hypotheses: (1) 1 st -year seedling survival is higher in species with greater allocation to TNC reserves in stems and roots; (2) species with high allocation to TNC storage exhibit slow growth rates (i.e., a tradeoff between growth and storage); and, (3) interspecific variation in tolerance to tissue loss and temporal reduction in light availability are linked to species differences in TNC storage. Finally, given the hypothesized importance of seedling size (a strong correlate of seed size in tropical forests; Rose and Poorter 2000) in seedling survival, I also examined the relative importance of total seedling biomass versus TNC reserve size as predictors of seedling survival and stress tolerance among the seven species. Methods Study Site and Species The study was conducted on Buena Vista peninsula, an area of ca. 60-year-old secondary lowland tropical forest, located in the Barro Colorado Nature Monument (BCNM), Panama (910 N, 7951 W). The BCNM forest is semi-deciduous, with a pronounced 4-month dry season that usually lasts from mid-December to mid-April. Annual rainfall on the BCNM averages 2,700 mm (Rand and Rand 1982). Seven woody species with sufficient seed availability were chosen for study (Table 1-1). All of the species are canopy trees as adults, except for Callichlamys

PAGE 38

28 latifolia (Bignoniaceae), which is a woody liana. Species were selected to span a range of shade tolerance, which was quantified by estimating the percentage of seedlings surviving for 1 year in experimental plots located in the shaded understory (Table 1-1). Small-seeded, pioneer species that are generally unable to establish seedling populations in the shaded forest understory were deliberately excluded from the study. The seven species differ in seed mass (24 to 453 mg) and cotyledon functional morphology (Table 1-1). Aspidosperma cruenta, Lacmelia panamensis, Castilla elastica, and Platypodium elegans have hypogeal cotyledons that remain below or just above the soil surface, and that function primarily for storage of seed reserves. The three other species have epigeal cotyledons that become elevated above the ground after germination; Coussarea curvigemnia and Tabebuia rosea have foliaceous photosynthetic cotyledons, and Callichlamys latifolia has thick green cotyledons that function primarily for storage (Kitajima 2002). For brevity, species are henceforth referred to by genus. Field and Laboratory Methods Depending on seed availability, 200 to 400 seedlings per species were raised from seed in a shade house on the BCNM under a light level similar to that found in the forest understory (1.5% full sun, based on total daily photon flux density). When possible, seeds were collected from at least 3 nonadjacent parent trees, to increase genetic diversity within each species. Seedlings were raised in plastic trays containing a 1:1 mixture of forest soil and sand, until all seedlings reached an equivalent ontogenetic stage, defined by the full expansion of the first photosynthetic organs (leaves or cotyledons, depending on the species; Table 1-1), at which time they were transplanted into the forest (described below). Seedlings were transplanted at a standardized ontogenetic stage, rather than at a common chronological age, so that all species had initiated autotrophic growth by the

PAGE 39

29 onset of the study. Because of interspecific variability in timing of seed dispersal, germination, and development, seedlings were transplanted over a 3-month period during the early-mid rainy season (May to July). Seedlings of all species were transplanted into four 7 x 7 m 2 common-gardens located in the shaded forest understory on Buena Vista Peninsula. Each garden was separated by at least 50 m from the nearest adjacent garden, and enclosed by 1 m tall fencing to exclude large herbivores. For each species, equal numbers of seedlings were transplanted into 1 x 1 m 2 cells in four equally sized, stratified sections in each garden (50 to 100 total seedlings garden -1 species -1 ). All vegetation < 1.5 m tall was removed from each garden before seedling transplantation, and all seedlings were positioned a minimum of 25 cm apart. When necessary, leaf litter was also removed to avoid burial of seedlings. Seedlings dying in the first 2 weeks after transplantation were replaced (< 10 seedlings total). Seedlings were randomly assigned to one of three treatments 2 weeks after transplantation: (1) defoliation; (2) light reduction; (3) control (no manipulation). For the defoliation treatment, all leaves and photosynthetic cotyledons (if present) were clipped at the base of the petiole; cotyledons were retained on all of the species with hypogeal storage cotyledons (Table 1-1). For the light-reduction treatment, cylindrical wire cages covered with 90% shade cloth were placed over individual seedlings. Shade cage dimensions ranged from 6 x 12 cm (height x diameter) to 15 x 30 cm (depending on initial seedling size); and were constructed to be large enough so as not to hinder growth over a 2-month period, after which the shade cages were removed.

PAGE 40

30 Light transmission was measured inside and outside of the shade cages with Li-Cor quantum sensors during the early-mid rainy season. Percent light transmission underneath the cages, and in each common-garden, was determined using both instantaneous and continuous total daily photon flux density (PFD) measurements. Instantaneous PFD measurements were taken inside and outside of shade cages at 8 15 stratified positions in each garden under mid-day, cloudy sky conditions. Continuous PFD was measured by placing a single censor in the middle of each common-garden for 2 days. To calculate % light transmission, measurements in the forest were referenced to a sensor placed in completely open sky on a laboratory rooftop. Mean light reduction ( + 1 SD) by the shade cages was 89 ( + 2)%. Mean light levels outside of the shade cages, as determined from the instantaneous PFD measurements, ranged among gardens from 0.38 to 0.68% (overall mean = 0.55%); mean light levels from the continuous PFD measurements taken over two full days in each garden were higher, and ranged among gardens from 0.68 to 1.44% of full sun (overall mean = 1%). Seedlings inside the shade cages thus experienced only ca. 0.06% of full sun over the 2-month light-reduction treatment. To estimate preand post-treatment biomass, I harvested randomly selected sub-samples of seedlings from each common-garden at two stages: (1) 2-weeks after seedlings were transplanted to the gardens and just before treatment (pretreatment harvest); and, (2) 2-months after treatment, coinciding with the end of the light-reduction treatment (post-treatment harvest). For the pretreatment harvest, 6 seedlings per species were harvested from each common-garden, depending on initial seed availability. Total samples sizes for the post-treatment harvest ranged from 2 to 32 total seedlings per

PAGE 41

31 species x treatment combination, depending on species-specific differences in survival. Seedlings were harvested by carefully excavating the root system with a small trowel. Samples were then placed in polyethylene bags and packed inside a cooler with ice for transport to the laboratory, where bags were stored at 4C in a refrigerator until they were processed for biomass measurements. Roots, stems, leaves, and cotyledons were separated within 12 hours after harvesting, than oven dried at 100C for 1 hour. All samples were then dried for an additional 48 hours at 60C before weighing. Leaf and cotyledon area (for epigeal cotyledons only) were measured before drying using a Li-Cor leaf area meter. All samples were stored in sealed polyethylene bags until samples were ground for TNC analysis. Quantification of TNC in Stems and Roots TNC was analyzed separately in stems and roots by determining the concentration of soluble sugars and starch, measured as glucose equivalents, with a colorimetric assay (Dubois et al. 1956; modified by Ashwell 1966). Due to the small initial sizes of the seedlings, tissue samples had to be pooled from multiple individuals to obtain adequate tissue biomass for TNC analysis (min. of 10 mg per sample). For the pretreatment analyses, tissues samples were pooled from 6 individuals per species in each common-garden (n = 1 pooled replicates species -1 garden -1 ). For the post-treatment analyses, tissue samples were pooled from 2 to 6 individuals per species x treatment combination. The pooled tissue samples were ground using either a Wig-L-Bug bead pulverizer or a Specs-Mill, and a 10 to 16 mg subsample was used for TNC analysis. Starch and soluble sugars were extracted using the method of Marquis et al. (1997), with slight modifications. Soluble sugars were first extracted by adding 1.5 mL of 80% ethanol to the dry samples in microcentrifuge tubes, which were then placed in a shaking

PAGE 42

32 water bath overnight at 27C. After shaking, tubes were centrifuged at 10,000 RPM for 10 minutes and the supernatant was decanted into 10 mL volumetric flasks. This process was repeated for two additional, 2-hour shaking baths. The final combined solution was diluted to 10 mL with deionized water and then stored in a refrigerator at 4C until colorimetric analysis. Starch content was determined from the pellet after the ethanol extractions of simple sugars. The pellets were transferred to 15 mL test tubes and incubated with 2.5 mL of sodium acetate buffer (0.2 M; pH 4.5) in a boiling water bath for 1 hr. After cooling, 2 mL of sodium acetate buffer and 0.5 mL of amyglucosidase solution were added and starch was digested overnight at 55C. Solutions were filtered and then diluted with deionized water to 25 mL in volumetric flasks. The phenol-sulfuric acid colorimetric assay (Dubois et al. 1956; modified by Ashwell 1966) was used to determine glucose concentrations using a spectrophotometer set at 487 nm. Glucose concentrations were calculated from standard curves using appropriate standards and blanks. The TNC concentration of each tissue was estimated as the sum of the glucose concentrations from the soluble sugar and starch extractions. TNC pool sizes were calculated by multiplying the average biomass of the pooled tissues by the sample concentrations, and then dividing by the number of individuals in the pooled sample. Statistical Analyses Differences among species and treatments in biomass and TNC were analyzed using ANOVA. Separate analyses were conducted for the pretreatment and post-treatment (2-month) harvest, using species means from the four common-gardens as replicates in both analyses (n = 4 replicates per species for the pretreatment analysis; n =

PAGE 43

33 1 per species x treatment combination for the post-treatment harvest; Table A-1, B-1). Due to lost samples, pretreatment biomass and TNC data for LACP had to be estimated from samples collected in a separate, concurrent study in the same understory site (K. Kitajima, unpublished data); mean stem and root biomass from this study were calculated from 9 individuals, which were then pooled into a single sample for TNC measurements. Due to the lack of replication for TNC, and differences in sample size for biomass from the other 6 species, LACP was excluded from the pretreatment ANOVA. Because all seedlings had died for some species x treatment combinations in the first 2 months, I used 2 separate ANOVA models to analyze post-treatment biomass and TNC. In the first model, I tested for species, treatment, and species x treatment interaction effects on biomass and TNC response variables, but only included data for the species that had living seedlings present in all treatments at the time of harvest (n = 5 species). The results from this model indicated that interactions had no significant effect (P > 0.12) on any of the response variables. I therefore conducted a second analysis using a model that contained data from all 7 species and that tested for the main effects of species and treatment on biomass and TNC, but excluded species x interaction terms. For brevity, I present only the results from the second model. All data except pretreatment stem and total (stem + root) TNC pool sizes were log 10 -transformed before analysis to improve normality and homogeneity of variance. Simple least squares linear regression was used to test for relationships between seedling biomass, growth, survival, and TNC storage. Relative growth rates (RGR) of total seedling biomass and leaf area from 0 to 2 months was estimated using the following formula:

PAGE 44

34 RGR (wk -1 ) = [ln (size at 2 months) ln (size at 0 months)]/8 Leaf and cotyledon mass was included in all RGR calculations of total seedling biomass. Survival estimates for control seedlings over the full year (Table 1), and for stress-treated seedling from 0 to 2 months and 2 months to 1 year, were taken from a separate study (Chapter 1). All statistical analyses were performed using JMP software (SAS Institute, 1997). Results Variation in Pretreatment Biomass and TNC Reserves Pretreatment stem and root biomass varied widely and significantly among species (Figure 2-1a; Table 2-1). Biomass was higher in stems relative to roots in all species. Pretreatment cotyledon biomass also varied widely, ranging from 18 to 209 mg among species (Figure 2-1a). In four of the seven species (Aspidosperma, Lacmelia, Castilla, and Tabebuia), biomass was higher in cotyledons than in stems or roots combined. Pretreatment TNC concentrations also differed significantly among species (Table 2-2). TNC concentrations varied more widely in stems relative to roots among species (Figure 2-1b). TNC concentrations were also substantially higher in stems than in roots. TNC comprised between 6 to 25% of the total stem biomass (mean = 15% for all species combined), whereas root concentrations ranged from 4 to 21% (mean = 8%). There were no significant correlations between 1 st -year control seedling survival (Table 1-1) and total pretreatment biomass (Figure 2-2a) or TNC concentration among species (Figure 2-2b), despite high interspecific variation in both these variables. First year survival was also not related to pretreatment cotyledon biomass, stem biomass, root biomass, or root:shoot ratio among species (P > 0.3 for all regressions using species means).

PAGE 45

35 Pretreatment TNC pool sizes also differed significantly among species, especially in stems, which were the dominant storage location for TNC in all species (Figure 2-3; Table 2-2). Total TNC pool size (stem + root pools combined) was a better correlate of 1 st -year control seedling survival among species (Figure 2-2c) than was total pretreatment biomass or TNC concentration (Figure 2-2a,b): Aspidosperma and Lacmelia, the two species with the highest 1 st -year survival, had the largest total TNC pools (12.9 and 15 mg, respectively), while Tabebuia, the species with lowest 1 st -year survival, had the smallest pool size (1.7 mg). The other 4 species had similar, intermediate levels of TNC ranging between 4.1 and 6.5 mg (Figure 2-3). Seedling size had a stronger effect on TNC pool size than on concentration. The four species with the largest total seedling biomass (Aspidosperma, Lacmelia, Castilla, Platypodium), all of which had hypogeal storage cotyledons, had greater TNC pool sizes than species the two smallest species (Coussarea, Tabebuia) with epigeal photosynthetic cotyledons (Figure 2-3; Table 1-1). In contrast, there were no clear trends between stem or root TNC concentrations and total seedling biomass among species. Coussarea and Tabebuia, the species with the lowest total seedling biomass, had two of the highest stem TNC concentrations. Effects of Defoliation and Deep Shade on Biomass and TNC Reserves There were significant species and treatment effects on post-treatment stem and root biomass at 2 months (Figure 2-4a,b; Table 2-1). For most species, both defoliation and light reduction caused a decrease in stem and root biomass relative to controls. The combined biomass of stems and roots (not including leaf or cotyledon biomass) was reduced more substantially by defoliation in species with lower 1 st -year control seedling survival (Figure 9a,b): biomass reductions (relative to control treatment biomass) ranged

PAGE 46

36 from 5 to 16% in the three species with highest 1 st -year survival (Aspidosperma, Lacmelia, Coussarea), compared to 19% in the three species with lower 1 st -year survival (Callichlamys, Castilla, Platypodium). Species with lower 1 st -year survival also suffered greater reductions in biomass in the light-reduction treatment (range = 11% for Callichlamys, Castilla, and Tabebuia) relative to species with higher 1 st -year survival (range = 116% for Aspidosperma, Lacmelia, and Coussarea). Stress treatments had significant negative effects on both TNC pools and concentrations (Table 2-2). TNC concentrations were reduced proportionally less by the stress treatments than TNC pools in all species (Figure 2-4c-f). Reductions in TNC pool sizes (Figure 2-2e,f) generally mirrored reductions in biomass (Figure 2-2a,b) among species. However, total TNC pool size was reduced more substantially than was total biomass: TNC reductions in the three species with the lowest 1 st -year control seedling survival ranged from 47 to 64% and 23 to 54% for the defoliation and light-reduction treatments, respectively. Tabebuia, the species with lowest 1 st -year survival and the smallest TNC pools, suffered the highest mortality after defoliation. Post-treatment cotyledon biomass differed among species and treatments (Figure 2-5). Two of the species (Platypodium and Callichlamys) had exhausted all cotyledon reserves in all treatments in the first 2 months. Control treatment values among the other five species ranged from 0.61 to 28 mg. Cotyledon biomass of stressed seedlings was either higher or only slightly reduced relative to control seedlings in all species (Figure 2-5). Influence of Seedling Size, Growth, and TNC Reserves on Seedling Survival Interspecific variation in TNC pool size was a better correlate of seedling survival than seedling size for both control and stress-treated seedlings. For control seedlings,

PAGE 47

37 1 st -year survival was more strongly correlated with TNC pool size than to TNC concentration (Figure 2-2). Furthermore, whole-seedling RGR from 0 to 2 months was negatively correlated with pretreatment TNC pool size among species (Figure 2-6). RGR, however, was not significantly correlated with either pretreatment TNC concentrations or total seedling biomass (P > 0.8, r 2 = 0.01, n = 6 for stem and root TNC concentrations; P = 0.11, r 2 = 0.50 for total seedling biomass). Two-month RGR of total structural biomass (seedling biomass TNC pools), as well as absolute total biomass growth, were also negatively, and more significantly, correlated with pretreatment TNC pool size among species (P = 0.03, r 2 = 0.69 for RGR of structural biomass; P = 0.01, r 2 = 0.80 for absolute biomass growth). RGR of leaf area growth from 0 to 2 months was also negatively, but less strongly, correlated with pretreatment TNC pool size among species (P = 0.08, r 2 = 0.56). Thus, large initial investments in TNC storage pools enhanced 1 st -year seedling survival, but at the cost of lower seedling growth rates. Survival of stressed seedlings was also significantly correlated with TNC pool size, but not with total seedling biomass (Figure 2-7). TNC pools had the greatest influence on survival during the first 2 months, when seedlings were most impacted by the stress treatments (Figure 2-7a); 2-month survival was also related, though less strongly, to stem TNC concentration among species (P = 0.01, r 2 = 0.41, n = 14). Five of the seven species had seedlings that survived the stress treatments past 2 months. Survival of these species from 2 months to 1 year, when species were presumably recovering from the stress treatments, was also significantly related to TNC pool size at the end of 2 months (Figure 2-7c). Seedling survival was not significantly related to total seedling biomass during the first 2 months or from 2 months to 1 year (Figure 2-7b,d).

PAGE 48

38 Discussion My results support the hypotheses that TNC storage enhances seedling survival and stress tolerance in the shaded forest understory. Overall, the seven species displayed wide variation in their carbon allocation strategies to seedling growth and storage. Most importantly, interspecific variation in TNC pool size, but not seedling size, was a good predictor of both 1 st -year seedling survival in the shaded understory, and seedling tolerance to stress caused from both tissue loss and temporary reduction in light availability. TNC concentrations in stems and roots ranged widely among species (4.5 %) and generally varied independently of seedling size, cotyledon type, and seedling shade tolerance. Finally, my study provides evidence for a tradeoff between allocation to growth and carbohydrate storage among coexisting species in a tropical forest. These results provide a mechanistic explanation for why species with high seedling survival in the shaded forest understory also tend to exhibit slow growth rates (Kitajima 1994, 2002; Osunkoya et al. 1994), and support the notion that tradeoffs between growth and survival may play an important role in maintaining high tree species diversity in tropical forests (Kitajima 1994; Kobe 1999). Role of TNC Reserves in Seedling Survival and Stress Tolerance My results show a strong linkage between interspecific variation in TNC storage and survival of stressed seedlings. TNC reserves had the greatest impact on seedling survival during the first 2 months after the stress treatments. During this time, both TNC concentrations and pools were reduced in all species, underscoring the importance of storage reserves for survival when seedlings experience negative carbon balance during the early seedling establishment. However, even after TNC levels were reduced experimentally, TNC pool size continued to be a significant predictor of variation in

PAGE 49

39 seedling survival through the year. TNC reserves therefore played an important role in both stress tolerance and post-stress recovery. TNC reserves may allow tropical tree seedlings to cope with and recover from a host of biotic and abiotic stresses. Several of my study species showed substantial tolerance to defoliation, confirming previous results from seedling defoliation experiments conducted in temperate (Canham et al. 1999) and tropical (Becker 1983; Kitajima 2004) forests. The ability to survive defoliation was dependent on the relative capacity of a species to develop new and fully expanded leaves, and species with greater leaf recovery exhibited higher survival (Chapter 1). These results indicate that large TNC reserves may be essential for seedling recovery from tissue loss due to herbivores and pathogens, or shoot damage from falling canopy debris, all of which are major causes of seedling mortality in tropical forests (Augspurger 1984a,b; Osunkoya et al. 1992; Asquith et al. 1997; Clark and Clark 1989; Scariot 2000). The results of my study also demonstrate that TNC reserves play a critical role in maintaining carbon balance when seedlings experience short-term reductions in light availability. The aim of the light-reduction treatment was to expose seedlings to irradiance levels well below their light compensation point, thereby forcing seedlings into a negative carbon balance. My results indicate that species can vary widely in their survival and growth responses to extreme temporal reductions in light availability in the shaded understory. The most shade-tolerant species in my study exhibited extremely high tolerance to experimental light levels of ca. 0.06% of full sun for 2 months, a level far below what has been used in most previous experimental studies and 1/10 of the average understory values typically reported for tropical forests. While the degree of

PAGE 50

40 light reduction in this experiment was extreme, it is probably representative of the light limitations imposed on seedlings during periods of intensive cloud cover (daily and seasonally) and when seedlings establish in microsites deeply shaded by litter accumulation or understory foliage. Tradeoffs between Seedling Growth and Storage The results supported the prediction of a tradeoff between seedling growth rate and allocation to carbohydrate storage reserves. Across species, TNC pool size was negatively correlated with RGR of total seedling biomass, RGR of total structural seedling biomass, and absolute seedling growth during the first 2 months of study. TNC pool size was also positively correlated with 1 st -year seedling survival, explaining approximately half of the variation in survival among species. These results support the hypothesis that seedling survival in the shaded understory of tropical forests is achieved through a balance between allocation to growth and carbohydrate storage (Kitajima 1994, 1996), rather than by maximizing net carbon gain for fast seedling growth rates (e.g., Givnish 1988). TNC pool size was a better correlate of seedling RGR and survival than either TNC concentration or seedling size. Furthermore, TNC pools were generally more reduced by the stress treatments than were TNC concentrations and total biomass. These results suggest that the total pool of TNC, irrespective of seedling size per se, may serve as the best quantitative indicator of interspecific variation in seedling survival and stress tolerance in the shaded forest understory. For example, TNC pools in Aspidosperma, the species with highest total seedling biomass and lowest RGR, were seven times higher than in Tabebuia, the species with lowest seedling biomass and highest RGR. In contrast, TNC concentrations differed only slightly between these species. These results

PAGE 51

41 emphasize the importance of distinguishing between structural versus nonstructural biomass when linking morphological traits to seedling growth, biomass allocation, and survival (e.g., Canham et al. 1999). TNC Allocation, Seed Size, and Seedling Functional Morphology My results suggest that stems may be the dominant storage location for TNC in understory tropical forest seedlings. Both TNC concentrations and pool sizes were higher in stems than in roots in all of the 7 species. Higher TNC storage in stems could reflect physiological constraints related to the relatively small size of seedling root systems in the shaded understory. TNC allocation to stems versus roots might also be related to the relative susceptibility of these tissues to damage from herbivory or disturbance during development. In forested ecosystems that experience large-scale disturbance such as fire, or where seedlings must survive though winter dormancy, storage reserves are typically concentrated in roots rather than stems (Canham et al. 1999; Hoffman et al. 2004). Variation in TNC storage was linked to species differences in both seed size and cotyledon functional morphology. The four largest-seeded species, all of which had hypogeal storage cotyledons, had the largest seedlings and TNC pools in stems and roots (but not necessarily the higher TNC concentrations). Total TNC pools in these species undoubtedly represent conservative estimates, as cotyledon reserves were not measured in this study. In contrast, the two smallest seeded species, both of which had epigeal photosynthetic cotyledons, had small seedlings and lower TNC pool sizes, but two of the highest TNC concentrations. These results show that large-seeded tropical tree species, in addition to having high amounts of energy reserves in seeds to enhance short-term survival, also maintain large pools of longer-term storage reserves in stems and roots. However, because all species had become dependent on autotrophic carbon gain before

PAGE 52

42 my TNC measurements, it is unknown to what degree initial seed reserves contributed to the more permanent stem and root TNC pools in these species. Future experiments that quantify how seed reserves contribute to long-term TNC storage in seedlings will likely shed further light on the evolutionary and ecological role of seed size variation in tropical forest tree communities. Conclusion In this paper, I highlight the importance of TNC reserves for seedling survival and stress tolerance in the shaded understory of a tropical forest. Tradeoffs between traits such as storage that enhance seedling stress tolerance versus those that maximize growth have likely contributed to the evolution of life history diversity and the maintenance of species coexistence in species-rich tropical forests. Interspecific variability in TNC storage constitutes a central component of seedling recruitment dynamics in this and likely many other tropical forests.

PAGE 53

43 Table 2-1. Analysis of variance for stem and root biomass. VariabledfF P F P Pre-treatment1 Species5260.7<0.000136.7<0.0001 Error18Post-treatment (2 months) Species6125.8<0.0001127.3<0.0001 Treatment27.70.00116.8<0.0001 Error59 1 LACP excluded from the analysis (see Methods)Stem biomassRoot biomass

PAGE 54

Table 2-2. Analysis of variance for TNC concentrations and pool sizes. VariabledfF P F P F P F P F P Pre-treatment Species5115.0<0.000114.4<0.0001116.1<0.000127.0<0.0001113.2<0.0001 Error18Post-treatment Species649.0<0.000141.9<0.000143.8<0.000137.0<0.000152.1<0.0001 Treatment212.3<0.00016.80.00220.0<0.000116.1<0.000123.8<0.0001 Error57 59(stem + root)Total TNC pool sizeTNC concentrationTNC pool sizeStemRootStemRoot 44

PAGE 55

45 A Biomass (mg) 050100210220230 Cotyledon Stem Root B ASPCLACPCOUCCALLCASEPLAETABR TNC concentration (%) 0510152025 Species LowestsurvivalHighestsurvival ***** Figure 2-1. Pretreatment biomass and TNC concentrations. A) Cotyledon, stem, and root biomass. B) Stem and root TNC concentrations. Species are ordered from highest to lowest 1 st -year survival of control seedlings (Table 1-1). Stars in (A) indicate species with storage cotyledons. Vertical bars = + 1 SD; n = 4 common-garden replicates for each species (see Methods); values for LACP were estimated from another experiment (n = 9 for biomass, pooled into a single sample for TNC analysis).

PAGE 56

ATotal biomass (mg) 0100200300400500 1-year survival (%) 020406080100 BStem TNC concentration (%) 0510152025 CTotal TNC pool size (mg) 0246810121416 P > 0.3r2 = 0.16P = 0.09r2 = 0.46P > 0.4r2 = 0.13ASPC LACP COUC CALL CASE PLAE TABR Figure 2-2. Three potential correlates of interspecific variation in 1 st -year control seedling survival. A) Pretreatment total seedling biomass (stems, roots, cotyledons, and leaves). B) Pretreatment stem TNC concentration. C) Pretreatment total TNC pool size (stem + root pools combined). 46

PAGE 57

47 ASPCLACPCOUCCALLCASEPLAETABR TNC pool size (mg) 0246810121416 Root Stem Species LowestsurvivalHighestsurvival Figure 2-3. Pretreatment TNC pool sizes in stems and roots.

PAGE 58

48 B 010203040506070 A Biomass (mg) 020406080100120 Control Light reduction Defoliation D 05101520 C TNC concentration (%) 0510152025 ESpecies ASPCLACPCOUCCALLCASEPLAETABR TNC pool size (mg) 0246810121416 FSpecies ASPCLACPCOUCCALLCASEPLAETABR 01234567 StemsRoots Figure 2-4. Post-treatment biomass and TNC in stems and roots. A,B) Total biomass. C,D) TNC concentrations. E,F) TNC pool sizes. TNC for TABR in both stress treatments and for PLAE in the light-reduction treatment could not be analyzed due to low survival. Vertical bars = + 1 SD; n = 1 common-garden replicates per bar (Table A-1, B-1).

PAGE 59

49 Species ASPCLACPCOUCCALLCASEPLAETABR Cotyledon biomass (mg) 020406080100 Control Light reduction Defoliation Figure 2-5. Post-treatment cotyledon biomass. Cotyledons in COUC, CALL and TABR were removed as part of the defoliation treatment. All cotyledon reserves in PLAE and for the control and light-reduction treatments in CALL were exhausted before the post-treatment harvest at 2 months. Vertical bars = + 1 SD; n = 2 common-garden replicates per bar (Table A-1).

PAGE 60

50 P = 0.03r2 = 0.70TNC pool size (mg) 0246810121 4 RGR (wk-1) -0.05-0.04-0.03-0.02-0.010.000.010.020.03 ASPC COUC CALL CASE PLAE TABR Figure 2-6. Two-month relative growth rate (RGR) of control seedlings plotted against pretreatment TNC pool size. Data are shown for the 6 species in which complete biomass data were available.

PAGE 61

51 A 0246810121416 Survival (0 2 mo) (%) 020406080100 P = 0.002r2 = 0.54 CTotal TNC pool size (mg) 02468101214 Survival (2 mo 1 yr) (%) 020406080100 P = 0.05r2 = 0.39 B 050100150200250300350400450 P = 0.20r2 = 0.12 DTotal seedling biomass (mg) 050100150200250300 P = 0.14r2 = 0.24 Figure 2-7. Survival of stressed seedlings plotted against total TNC pool size and total seedling biomass. A,B) Survival from 0 to 2 months as a function of pretreatment total TNC pool size and total seedling biomass. C,D) Survival from 2 months to 1 year as a function of post-treatment (2-month) total TNC pool size and total seedling biomass. Each point is a species mean; white circles = defoliation treatment, dark circles = light-reduction treatment.

PAGE 62

APPENDIX A SUMMARY STATISTICS FOR SEEDLING BIOMASS AND LEAF AREA Table A-1. Statistics for preand post-treatment biomass and leaf area. Speciesnmea n SDnmea n SDnmea n SDnmea n SDCotyledon mass(mg) ASPC4209.216.9428.625.5459.633.0439.230.2 CALL422.416.940.00.020.00.040.00.0 CASE491.423.940.61.2412.08.443.93.3 COUC418.11.2416.31.640.00.0416.42.5 LACP964.465.5420.920.8437.629.5439.339.4 PLAE433.912.540.00.010.00 TAB R 418.62.8317.92.40216.60.3Root mass (mg) ASPC448.13.4457.75.7450.34.5447.84.4 CALL420.94.5425.67.2216.02.8415.45.0 CASE420.45.2420.32.6414.41.1416.55.1 COUC48.10.848.51.246.20.846.70.6 LACP1914.75.8417.23.1414.73.3415.31.5 PLAE426.611.0427.19.5110.00 TAB R 45.50.735.80.9024.71.1Pre-treatmentControlDefoliationPost-treatment (2 months)Light reduction 52

PAGE 63

53 Table A-1. Continued Speciesnmea n SDnmea n SDnmea n SDnmea n SDStem mass (mg) ASPC470.24.7484.58.1484.53.0475.23.7 CALL438.35.1457.310.8230.83.5443.012.3 CASE444.93.8464.011.9453.425.1458.17.6 COUC422.01.4420.31.6418.02.3417.41.3 LACP946.84.1446.59.0444.36.4447.25.4 PLAE479.97.1499.45.9157.40 TAB R 48.51.2311.93.2027.70.4Total mass (mg) ASPC4442.712.24331.320.44221.334.64297.548.1 CALL4146.828.54171.928.4246.70.74117.219.8 CASE4219.834.24171.723.9499.336.74115.734.3 COUC448.21.5445.71.6425.41.7440.63.7 LACP9185.379.14156.139.04122.648.14157.446.4 PLAE4213.39.94234.024.6167.40 TAB R 433.55.7339.69.60230.00.1Leaf area (cm2) ASPC428.12.1429.81.847.31.7428.73.5 CALL437.22.0443.85.120.00.0432.97.5 CASE432.93.0447.27.6411.64.7422.813.2 COUC44.50.545.00.440.60.144.90.6 LACP932.47.3429.85.2411.65.1427.12.4 PLAE432.88.0439.76.710.00 TAB R 46.50.939.73.0027.61.6Pre-treatmentPost-treatment (2 months)ControlDefoliationLight reduction

PAGE 64

54 Table A-2. Statistics for total seedling biomass and leaf area 1 year after treatment. Speciesnmea n SDnmea n SDnmea n SDTotal mass (mg) ASPC4536.8143.94.0248.158.94466.294.8 CALL4519.6190.80.02404.649.7 CASE4272.474.03.0179.885.33243.7154.7 COUC496.029.14.038.59.9477.916.8 LACP4313.899.74.0200.952.94263.6120.0 PLAE3371.6122.00.00 TAB R 241.420.20.00Leaf area (cm2) ASPC444.512.24.013.65.7438.84.8 CALL489.424.20.0261.67.6 CASE452.613.93.032.617.1337.138.4 COUC411.83.74.03.31.549.42.2 LACP446.812.34.029.06.9440.517.8 PLAE343.237.50.00 TAB R 25.98.30.00Post-treatment (1 year)ControlDefoliationLight reduction

PAGE 65

APPENDIX B SUMMARY STATISTICS FOR TOTAL NONSTRUCTURAL CARBOHYDRATE (TNC) RESERVES IN STEMS AND ROOTS Table B-1. Statistics for preand post-treatment TNC concentrations and pool sizes. Speciesnmea n SDnmea n SDnmea n SDnmea n SDRoot (%) ASPC46.200.4948.221.8546.281.7645.921.58 CALL44.600.6649.223.2826.251.5636.851.94 CASE45.260.9744.041.4843.990.5542.950.44 COUC410.421.47413.051.2349.930.63410.210.67 LACP121.91417.101.84412.023.28415.674.25 PLAE44.380.8547.291.8617.150 TAB R 46.111.2549.351.6600Root (mg) ASPC42.950.3844.801.5343.191.0842.800.72 CALL40.970.3042.521.6921.020.4331.060.19 CASE41.040.1640.890.4140.570.0940.500.26 COUC40.850.1941.120.2540.600.0640.690.03 LACP13.2342.840.6541.740.6842.350.65 PLAE41.110.3742.030.9210.710 TAB R 40.340.0940.730.3800Stem (%) ASPC414.200.34413.083.25410.912.42411.531.45 CALL48.190.56412.094.4327.732.9839.481.20 CASE412.310.6048.122.2845.240.4747.081.23 COUC421.062.00421.794.41416.845.68416.841.12 LACP125.12421.511.37416.893.62420.842.20 PLAE46.100.5447.250.6715.110 TAB R 417.382.42412.631.8400Pre-treatmentControlDefoliationPost-treatment (2 months)Light reduction 55

PAGE 66

56 Table B-1. Continued Speciesnmea n SDnmea n SDnmea n SDnmea n SDStem (mg) ASPC49.990.88411.183.7749.252.2748.711.25 CALL43.150.5846.862.8722.320.6433.550.47 CASE45.500.5645.511.9242.811.2344.421.49 COUC44.630.4544.350.5842.620.4442.940.30 LACP111.77410.282.6047.772.61410.101.83 PLAE44.850.2447.190.6612.930 TAB R 41.450.1241.580.3200Total (mg) ASPC412.941.24415.985.28412.443.15411.511.62 CALL44.110.8749.384.5423.341.0744.730.47 CASE46.550.6646.402.3243.371.3244.921.66 COUC45.490.4045.470.8243.220.4043.620.31 LACP115.00413.123.2249.513.29412.452.43 PLAE45.960.1849.221.2013.640 TAB R 41.790.1842.310.60011.06Pre-treatmentPost-treatment (2 months)ControlDefoliationLight reduction

PAGE 67

LIST OF REFERENCES Ashwell, G. (1966) New colorimetric methods of sugar analysis. VII. The phenol-sulfuric acid reaction for carbohydrates. Methods in Enzymology, 8, 93-95. Asquith, N.M., Wright, S.J. & Clauss, M.J. (1997) Does mammal community composition control recruitment in neotropical forests? Evidence from Panama. Ecology, 78, 941-946. Augspurger, C.K. (1984a) Seedling survival of tropical tree species: Interactions of dispersal distance, light-gaps, and pathogens. Ecology, 65, 1705-1712. Augspurger, C.K. (1984b) Light requirements of neotropical tree seedlings: A comparative study of growth and survival. Journal of Ecology, 72, 777-795. Augspurger, C.K. (1983) Seed dispersal of the tropical tree, Platypodium elegans, and the escape of its seedlings from fungal pathogens. Journal of Ecology, 71, 759-771. Augspurger, C.K. & Kelly, C.K. (1983) Pathogen mortality of tropical tree seedlings: Experimental studies of the effects of dispersal distance, seedling density, and light conditions. Oecologia, 61, 211-217. Becker, P. (1983) Effects of insect herbivory and artificial defoliation on survival of Shorea seedlings. Tropical Rain Forest: Ecology and Management (eds. S.L. Sutton, T.C. Whitmore, & A.C. Chadwick), pp. 241-252. Blackwell Scientific Publishers, Oxford. Bloom, A.J., Chapin, F.S. III & Mooney, H.A. (1985) Resource limitation in plants: An economic analogy. Annual Review of Ecology and Systematics, 16, 363-392. Boot, R.G.A. (1996) The significance of seedling size and growth rate of tropical rain forest tree seedlings for regeneration in canopy openings. The Ecology of Tropical Forest Tree Seedlings (ed M.D. Swaine), pp. 267-282. UNESCO, Paris. Bullock, S.H. (1992) Seasonal differences in nonstructural carbohydrates in two dioecious monsoon-climate trees. Biotropica, 24, 140-145. Canham, C.D., Kobe, R.K., Latty, E.F. & Chazdon, R.L. (1999) Interspecific and intraspecific variation in tree seedling survival: effects of allocation to roots versus carbohydrate reserves. Oecologia, 121, 1-11. 57

PAGE 68

58 Chazdon, R.L. & Fetcher, N. (1984) Photosynthetic light environments in a lowland tropical rain forest in Costa Rica. Journal of Ecology, 72, 553-564. Chazdon, R.L. & Pearcy, R.W. (1986) Photosynthetic responses to light variation in rainforest species. II. Carbon gain and photosynthetic efficiency during lightflecks. Oecologia, 69, 524-531. Clark, D.B. & Clark, D.A. (1989) The role of physical damage in the seedling mortality regime of a neotropical rainforest. Oikos, 55, 225-230. Coley, P.D., Bryant, J.P., & Chapin, F.S. III (1985) Resource availability and plant antiherbivore defense. Science, 230, 895-899. Connell, J.H. (1971) On the role of natural enemies in preventing competitive exclusion in some marine animals and in rain forest trees. Dynamics of Populations (eds P.J. den Boer & G.R. Gradwell), pp. 298-312. Center for Agricultural Publication and Documentation, Wageningen, The Netherlands. Connell, J.H. & Green, P.T. (2000) Seedling dynamics over thirty-two years in a tropical rain forest tree. Ecology, 81, 568-584. Cunningham, S.A. (1997) The effect of light environment, leaf area, and stored carbohydrates on inflorescence production by a rain forest understory palm. Oecologia, 111, 36-44. Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A. & Smith, F. (1956) Colorimetric method for determination of sugars and related substances. Analytical Chemistry, 28, 350-356. Engelbrecht, M.J. & Kursar, T.A. (2003) Comparative drought-resistance of seedlings of 28 species of co-occurring tropical woody plants. Oecologia, 136, 383-393. Farris-Lopez, K., Denslow, J.S., Moser, B. & Passmore, H. (2004) Influence of a common palm, Oenocarpus mapora, on seedling establishment in a tropical moist forest in Panama. Journal of Tropical Ecology, 20, 429-438. Fox, G.A. (2001) Failure-time analysis: Studying times to events and rates at which events occur. Design and Analysis of Ecological Experiments (eds S.M. Scheiner & J. Gurevitch), pp.235-266. Oxford University Press, Inc., New York. Givnish, T.J. (1988) Adaptation to sun and shade: A whole-plant perspective. Australian Journal of Plant Physiology, 15, 63-92. Green, P.T. & Juniper, P.A. (2004) Seed mass, seedling herbivory and the reserve effect in tropical rainforest seedlings. Functional Ecology, 18, 539-547. Grime, J.P. (1965) Shade tolerance in flowering plants. Nature, 208, 161-163.

PAGE 69

59 Grubb, P.J., Lee, W.G., Kollmann, J. & Wilson, J.B. (1996) Interaction of irradiance and soil nutrient supply on growth of seedlings of ten European tall-shrub species and Fagus sylvatica. Journal of Ecology, 84, 827-840. Gunatilleke, C.V.S., Gunatilleke, I.A.U.N., Perera, G.A.D., Burslem, D.F.R.P., Ashton, P.M.S. & Ashton, P.S. (1997) Responses to nutrient addition among seedlings of eight closely related species of Shorea in Sri Lanka. Journal of Ecology, 85, 301-311. Harms, K.E. & Dalling, J.W. (1997) Damage and herbivory tolerance through resprouting as an advantage of large seed size in tropical trees and lianas. Journal of Tropical Ecology, 13, 617-621. Harms, K.E., Wright, S.J., Caldern, O., Hernndez, A. & Herre, E.A. (2000) Pervasive density-dependent recruitment enhances seedling diversity in a tropical forest. Nature, 404, 493-495. Hoffman, W.A., Orthen, B. & Franco, A.C. (2004) Constraints to seedling success of savanna and forest trees across the savanna-forest boundary. Oecologia, 140, 252-260. Hoffman, W.A., Orthen, B. & Nascimento, P.K.V. (2003) Comparative fire ecology of tropical savanna and forest trees. Functional Ecology, 17, 720-726. Hubbell, S.P. (1979) Tree dispersion, abundance, and diversity in a tropical dry forest. Science, 203, 1299-1309. Janzen, D.H. (1970) Herbivores and the number of tree species in tropical forests. American Naturalist, 104, 501-528. Kabeya, D. & Sakai, S. (2003) The role of roots and cotyledons as storage organs in early stages of establishment in Quercus crispula: A quantitative analysis of the nonstructural carbohydrate in cotyledons and roots. Annals of Botany, 92, 537-545. Khurana, E. & Singh, J.S. (2004) Germination and seedling growth of five tree species from tropical dry forest in relation to water stress: Impact of seed size. Journal of Tropical Ecology, 20, 385-396. Kitajima, K. (2004) Impact of cotyedon and leaf removal on seedling survival in three tree species with contrasting cotyledon functions. Biotropica, 35, 429-434. Kitajima, K. (2002) Do shade-tolerant tropical tree seedlings depend longer on seed reserves? Functional growth analysis of three Bignoniaceae species. Functional Ecology, 16, 433-444. Kitajima, K. (1996) Ecophysiology of tropical tree seedlings. Tropical Forest Plant Ecophysiology (eds S.S. Mulkey, R.L. Chazdon, & A.P. Smith), pp. 559-596. Chapman and Hall, New York, NY.

PAGE 70

60 Kitajima, K. (1994) Relative importance of photosynthetic traits and allocation patterns as correlates of seedling shade tolerance of 13 tropical trees. Oecologia, 98, 419-428. Kobe, R.K. (1999) Light gradient partitioning among tropical tree species through differential seedling mortality and growth. Ecology, 80, 187-201. Kobe, R.K. (1997) Carbohydrate allocation to storage as a basis of interspecific variation in sapling survivorship and growth. Oikos, 80, 226-233. Kobe, R.K., Pacala, S.W., Silander, J.A., Jr. & Canham, C.D. (1995) Juvenile tree survivorship as a component of shade tolerance. Ecological Applications, 5, 517-532. Li, M., Leiberman, M. & Leiberman, D. (1996) Seedling demography in undisturbed tropical wet forest in Costa Rica. The Ecology of Tropical Forest Tree Seedlings (ed M.D. Swaine), pp. 285-314. UNESCO, Paris. Loach, K. (1967) Shade tolerance in tree seedlings. I. Leaf photosynthesis and respiration in plants raised under artificial shade. New Phytologist, 66, 607-621. Marenco, R.A, de C. Gonalves, J.F. & Vieira, G. (2001) Leaf gas exchange and carbohydrates in tropical trees differing in successional status in two light environments in central Amazonia. Tree Physiology, 21, 1311-1318. Marquis, R.J., Newell, E.A. & Villegas, A.C. (1997) Non-structural carbohydrate accumulation and use in an understory rain-forest shrub and the relevance for the impact of leaf herbivory. Functional Ecology, 11, 636-643. McPherson, K. & Williams, K. (1998) The role of cabbage palm seedlings in the growth, resilience, and persistence of cabbage palm seedlings (Sabal palmetto). Oecologia, 117, 460-468. Moles, A.T. & Westoby, M. (2004) What do seedlings die from and what are the implication for the evolution of seed size? Oikos, 106, 193-199. Montgomery, R.A. (2004) Effects of understory foliage on patterns of light attenuation near the forest floor. Biotropica, 36, 33-39. Montgomery, R.A. & Chazdon, R.L. (2002) Light gradient partitioning by tropical tree seedlings in the absence of canopy gaps. Oecologia, 131, 165-174. Newell, E.A., Mulkey, S.S., & Wright, S.J. (2002) Seasonal patterns of carbohydrate storage in four tropical tree species. Oecologia, 131, 333-342. Osunkoya, O.O., Ash, J.E., Hopkins, M.S. & Graham, A.W. (1994) Influence of seed size and seedling ecological attributes on shade-tolerance of rain-forest tree species in northern Queensland. Journal of Ecology, 82, 149-163.

PAGE 71

61 Osunkoya, O.O., Ash, J.E., Hopkins, M.S., & Graham, A.W. (1992) Factors affecting survival of tree seedlings in North Queensland rainforests. Oecologia, 91, 569-578. Pacala, S.W., Canham, C.D., Saponara, J., Silander, J.A., Kobe, R.K. & Ribbens, E. (1996) Forest models defined by field measurements: Estimation, error analysis and dynamics. Ecological Monographs, 66, 1-43. Poorter, L. (1999) Growth responses of 15 rain-forest trees to a light gradient: The relative importance of morphological and physiological traits. Functional Ecology, 13, 396-410. Rand, A.S. & Rand, W.M. (1982) Variation in rainfall on Barro Colorado Island. The Ecology of a Tropical Forest (eds E.G. Leigh, Jr., A.S. Rand & D.M. Windsor), pp. 47-66. Smithsonian Institution Press, Washington, DC. Rose, S.A. & Poorter, L (2002) The importance of seed mass for early regeneration in tropical forest: A review. Long term changes in Composition and Diversity: case studies from the Guyana Shield, Africa, Borneo and Melanesia (ed H. ter Steege), pp. 5-19. Tropenbos Foundation, Wageningen, The Netherlands. Scariot, A. (2000) Seedling mortality by litterfall in Amazonian forest fragments. Biotropica, 32, 662-669. Tissue, D.T. & Wright, S.J. (1995) Effects of seasonal water availability on phenology and the annual shoot carbohydrate cycle of tropical forest shrubs. Functional Ecology, 9, 518-527. Vzquez-Ynes, C., Orozco-Segovia, A., Rincon, E., Sanchez-Coronado, M.E., Huante, P., Toledo, J.R., & Barradas, V.L. (1990) Light beneath the litter in a tropical forest: Effect on seed germination. Ecology, 71, 1952-1958. Veneklaas, E.J. & Poorter, L. (1998) Growth and carbon partitioning of tropical tree seedlings in contrasting light environments. Inherent Variation in Plant Growth. Physiological Mechanisms and Ecological Consequences (eds H. Lambers, H. Poorter & M.M.I. Van Vuuren), pp.337-361. Backkhuys Publishers, Leiden, The Netherlands. Walter, M.B. & Reich, P.B. (2000) Seed size, nitrogen supply, and growth rate effect tree seedling survival in deep shade. Ecology, 81, 1887-1901. Walters, M.B. & Reich, P.B. (1999) Low-light carbon balance and shade tolerance in the seedlings of woody plants: Do winter deciduous and broad-leaved evergreen species differ? New Phytologist, 143, 143-154. Walters, M.B. & Reich, P.B. (1996) Are shade tolerance, survival, and growth linked? Low light and nitrogen effects on hardwood seedlings. Ecology, 77, 841-853.

PAGE 72

62 Wang, Y.H. & Augspurger, C.K. (2004) Dwarf palms and cyclanths strongly reduce Neotropical seedling recruitment. Oikos, 107, 619-633. Welden, C.W., Hewett, S.W., Hubbell, S.P. & Foster, R.B. (1991) Sapling survival, growth, and recruitment: Relationship to canopy height in a neotropical forest. Ecology, 72, 35-50. Woodward, F.I. (1990) From genes to ecosystems: The importance of shade tolerance. Trends in Ecology and Evolution, 5, 111-115. Wright, S.J. (2002) Plant diversity in tropical forests: A review of mechanisms of species coexistence. Oecologia, 130, 1-14.

PAGE 73

BIOGRAPHICAL SKETCH Jonathans interest in the natural world began when he was a boy growing up in the forests and fields of western New York State. His formal education in the natural sciences began during his junior year in high school, where he enrolled in and completed a 2-year vocational program in natural resources conservation. Jonathan went on to pursue a 2-year degree in forestry from Paul Smiths College, Paul Smiths, New York, graduating in May of 1999. As part of his Paul Smiths education, he participated in a tropical biology field course in Belize, which sparked his interest in both tropical ecosystems and a professional career in ecology. He continued his undergraduate education at Cornell University, Ithaca, New York, majoring in biological sciences, with a concentration in ecology and evolutionary biology. He graduated with a B.S. degree in May of 2002 and was awarded high honors for an undergraduate thesis on seed dispersal by white-tailed deer, later published in Oecologia (2004). Jonathan spent the following summer working as research assistant for the Institute of Ecosystem Studies, studying hurricane effects on seedling regeneration at the Luquillo Experimental Forest in Puerto Rico. Jonathan began his graduate work in the fall of 2002 in the Department of Botany at the University of Florida, Gainesville, Florida, studying tropical rainforest seedling ecology and ecophysiology. During his second semester, he participated in an 8-week tropical ecology field course in Costa Rica, hosted through the Organization for Tropical Studies at Duke University. He then traveled northwest to Panama to begin the field 63

PAGE 74

64 research for his M.S. thesis, which focused on the role of nonstructural carbohydrate reserves in seedling growth, survival, and stress tolerance in a tropical wet forest. He received his M.S. degree in interdisciplinary ecology, with a concentration in botany, in May of 2005. Jonathan is currently working towards a Ph.D. in tropical forest community ecology in the Department of Biological Sciences at Louisiana State University, Baton Rouge, Louisiana, and plans to pursue an academic career in research and teaching.