 

a. P. momqu

 

‘l «.41! .H \MIJ‘.'IJ,J|.H‘J‘J."V: 1‘ r t ‘ ‘ x 1.‘ 1 I! \: 1.L\|l“ ‘ ‘{
"Muir's“! vblfuu- ~. ‘ ‘ ‘EE.EIIHL ‘I'III n13!ln|l.ll"lralclmuﬂb\l {11/ -

THESIS

HISGAN STATEU

IIIIIIIIIIIIIIIII IIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIII

1293 01050 3229

 

 

This is to certify that the

dissertation entitled

PLANT COMPETITION IN RELATION TO PRODUCTIVITY
IN OLD-FIELD PLANT COMMUNITIES

presented by

Bryan L. Foster

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degreein Botany and Plant Path.

 

vﬂm 4%)

Major professor

Date 12-18-96

 

MS U i: an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY

MiChlgan State
University

 

 

 

URN BOX to romovo this chodtom from your record.

PLACE ll BET
3 return on or before one due.

TO AVOID FINE

DATE DUE DATE DUE DATE DUE

”a E.“ ‘1'.“ ﬁf‘: {If}
I__ __

———

_‘_

*4I333
j _I_
IeI:fﬂ
III;
;JLJ;3

MSU Is An Nﬂnnotm WM Opportunﬂv Intuition

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PLANT COMPETITION IN RELATION TO PRODUCTIVITY
IN OLD-FIELD PLANT COMMUNITIES

By

Bryan L. Foster

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

W. K. Kellogg Biological Station
and
Department of Botany and Plant Pathology

1996

ABSTRACT

PLANT COMPETITION IN RELATION TO PRODUCTIVITY
IN OLD-FIELD PLANT COMMUNITIES

By

Bryan L. Foster

I conducted three manipulative ﬁeld experiments: 1) to examine the
effects of living plant neighbors and accumulated plant litter on the
recruitment and growth of target plants across gradients of plant productivity
and community biomass, and 2) to examine the impact of nutrient
enrichment, increased productivity and litter accumulation on species
richness in southwest Michigan old-fields.

Plant removal experiments conducted at eleven old-field sites showed
that the surrounding plant community suppressed the recruitment and
growth of three native grass species (Andropogon gerardi, Schizachyrium
scoparium, Sorghastrum nutans) across a wide range of productivity. The net
suppressive effect of the surrounding plant community on two of the three
target species increased in magnitude with community biomass as predicted
by biomass-dependent theories of plant competition. Non-linear dependence
of net community effects on community biomass indicated that biomass-
dependent theories of competition may be most applicable to narrow ranges
of low productivity. The degree to which the effects of plant litter contributed
to the net effect of the plant community on target plant performance
depended on the life history stage examined. Inhibitory litter effects on
seedling recruitment increased with community biomass and contributed to

the net effect of the community on recruitment at sites of intermediate

community biomass, but litter had little impact on established transplants at
any of the sites.

In a two-year experiment carried out in a grass-dominated old-field,
nitrogen enrichment increased living plant and litter biomass and reduced
species richness by inhibiting the colonization of forb species. Community
responses to litter removal and litter addition treatments that were carried
out in conjunction with nitrogen enrichment showed that declines in species
richness and forb density were due to effects associated with increases in both
living plant and litter biomass.

The results highlight the potential importance of effects associated with
both living plant biomass and accumulated plant litter in inﬂuencing plant
recruitment, species diversity, and the distribution of native grasses across
old-field productivity gradients. The general significance of these findings are

discussed with respect to existing theories of plant community organization.

ACKNOWLEDGMENTS

First, I wish to acknowledge and thank my advisor, Kay Gross, for her
generous support and advise throughout my entire graduate career. Without
her encouragement, patience, and insight, this work would not have been
possible. I am grateful to the other members of my committee, Mike Klug,
Gary Mittelbach, and Phil Robertson for all of their help along the way and for
their critical review of my work. I thank Steve Tonsor, a former committee
member, for valuable advise early on in the development of this work. I
would also like to acknowledge Scott Wilson and Heather Reynolds for
providing insightful comments and suggestions on portions of this
dissertation.

I greatly value the interactions I've had with the graduate students and
faculty of K38 and thank them all for there inﬂuence and support over the
years. I especially thank fellow graduate students Paco Moore, Lisa Huberty,
and Casey Huckins for their assistance, warm friendship, and spirited
conversation.

A number of people provided invaluable technical and field assistance
during the course of my research. My sincere thanks goes to Char Adams,
Alice Gillespie, and John Gorentz for saving me on numerous occasions. I
thank Jay Leversee and Paco Moore for assistance in the field, Allison Porter
for critical help in the greenhouse, and Olivia Damon for her skills in

running the Alpkem.

iv

For encouragement and support in countless ways I thank Bill and
Betty Foster, and Jack and Renee O'Neill. Most of all I would like to express
my deepest appreciation and love to Ann and Thomas Foster. Their
unwavering encouragement, patience, love, and laughter have made this all
possible.

Generous financial support for this work was provided by the K38
graduate Research Training Group (RTG) funded by NSF grant DIR-09113598,
NSF LTER grant BSR 8702332, two George H. Lauff Research Awards, the
Department of Botany and Plant Pathology, and the Graduate Program in

Ecology and Evolutionary Biology (EEB).

‘7

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................... viii

LIST OF FIGURES ................................................................................................... x

CHAPTER 1 INTRODUCTION ............................................................................ 1
Organization of the dissertation ............................................................... 5
Study species ................................................................................................. 6
Study sites ..................................................................................................... 7

CHAPTER 2 REMOVAL EXPERIMENT I: PARTITIONING THE
EFFECTS OF PLANT BIOMASS AND LITTER ON ANDROPOGON

GERARDI IN OLD-FIELD VEGETATION ......................................................... 12
INTRODUCTION ........................................................................................ 13
METHODS .................................................................................................... 15

Study sites ........................................................................................ 15
Experimental design and methods .............................................. 17
Resource measurements ............................................................... 19

Data analysis .................................................................................... 20
RESULTS ....................................................................................................... 24
Community biomass and species composition ......................... 24
Resources .......................................................................................... 24

Seed addition .................................................................................... 28
Transplants ....................................................................................... 37
DISCUSSION ................................................................................................ 41
Magnitude of plant community effects ....................................... 43

CHAPTER 3 REMOVAL EXPERIMENT II: PARTITIONING THE
EFFECTS OF PLANT BIOMASS AND LITTER ON THREE NATIVE

GRASSES IN OLD-FIELD VEGETATION .......................................................... 47
INTRODUCTION ........................................................................................ 48
METHODS .................................................................................................... 51

vi

Study sites ........................................................................................ 51

Experimental design and methods .............................................. 52
Light measurements ....................................................................... 57
Data analysis .................................................................................... 57
RESULTS ....................................................................................................... 60
Community biomass and species composition ......................... 60
Light penetration ............................................................................. 61
Seed addition .................................................................................... 61
Transplants ....................................................................................... 74
DISCUSSION ................................................................................................ 81
Partitioning the effects of living neighbors and litter .............. 83
Implications for native grass distributions and species

diversity ............................................................................................ 86

CHAPTER 4 SPECIES RICHNESS IN A SUCCESSIONAL GRASSLAND:
EFFECTS OF NITROGEN ENRICHMENT AND PLANT LITTER ........... 88
INTRODUCTION ........................................................................................ 89
METHODS .................................................................................................... 92
Study site .......................................................................................... 92
Experimental treatments ............................................................... 93
Seed addition ................................................................................... 95
Light measurements ....................................................................... 96
Vegetation sampling and data analysis ...................................... 96
RESULTS ....................................................................................................... 97
Community responses ................................................................... 97
AndrOpogon recruitment .............................................................. 104
DISCUSSION ................................................................................................ 107
Community responses ................................................................... 107
Plant responses and light availability ......................................... 110
Andropogon recruitment .............................................................. 111
Productivity, diversity, and plant litter ...................................... 112
CHAPTER 5 CONCLUSIONS .............................................................................. 115
LIST OF REFERENCES ........................................................................................... 120

vii

LIST OF TABLES

Table 1. Locations, identification numbers, and vegetation characteristics

of the eleven study sites. Letters correspond to site locations shown on

the Kellogg Biological Station map (Fig. 1). Biomass values are presented

as means a: 1 SE. Total community biomass is the sum of aboveground

living plant biomass and plant litter. Methods for assessing plant biomass

and species composition are presented in the methods sections of chapters

2 and 3. Nomenclature follows Fernald (1950). ..................................................... 9

Table 2. Community characteristics of the five study sites. Biomass values

are presented as means :t: 1 SE (n = 5 for each site). For each biomass type,
values within a column with the same letter are not significantly different

(P > 0.05; Tukey’s HSD). Nomenclature follows Fernald (1950). ..................... 16

Table 3. Treatment effects and means square ratios used in the mixed-
model, within-subjects ANOVA. S = Sites, B(S) = Blocks within Sites,
N = Neighbors, L = Litter. ......................................................................................... 22

Table 4. Results of ANOVA for site and treatment effects on mean
seasonal light penetration. ....................................................................................... 25

Table 5. Results of ANOVA for site and treatments effects on mean

seasonal soil nitrogen concentration. Nitrogen concentrations were
measured only in the two treatments where all plant material (neighbors

and litter) was either removed or left intact. ....................................................... 29

Table 6. Results of ANOVA for site and treatment effects on Andropogon
density and total shoot biomass in the seed addition experiment. ................. 30

Table 7. Results of ANOVA for site and treatment effects on the RGR of
transplants in the seedling transplant experiment. ........................................... 38

Table 8. Vegetation characteristics of the nine study sites. Biomass values

are shown as means 1 1 SE (n = 6 for each site). Biomass values within a
column with the same letter are not signiﬁcantly different (P > 0.05;

Tukey’s HSD). ............................................................................................................. 53

viii

Table 9. Results of two-way ANOVA for community effects on density
and total shoot biomass (seed addition). ............................................................... 69

Table 10. Results of two-way ANOVA for community effects on the RGR
of transplants. ............................................................................................................. 77

Table 11. Mean biomass of grass species (g/ m2) in the ﬁve experimental
treatments. ANOVA's were performed for species that were present in

all treatments. Species are listed from greatest to least in biomass. Means
within rows with different letters are signiﬁcantly different (P < 0.05,

Tukey's HSD). See methods section for treatment codes. ............................... 101

Table 12. Forb biomass (g/ m2) and density (no./ m2) in the ﬁve

experimental treatments. Means are presented for life history groups and

for the ten most common forb species. ANOVA's were performed only

on life history groups and species that were present in all treatments.

Species are listed from greatest to least in density. Means within rows

with different letters are signiﬁcantly different (P < 0.05, Tukey's HSD).

See methods section for treatment codes. .......................................................... 102

ix

LIST OF FIGURES

Figure 1. Map of the Kellogg Biological Station with letters showing
the locations of the 11 study sites. See Table 1 for a description of these
sites. ............................................................................................................................... 11

Figure 2. A) Mean seasonal light penetration (mean + 1 SE) in relation to

the experimental treatments. Treatments with neighbors removed are
indicated by - N, and treatments with neighbors left intact are indicated

by +N. Treatments with litter removed are indicated by - L, and treatments
with litter left intact are indicated by shading and by +L. Brackets above and
below the bars specify within-site mean comparisons made at each level of
the two treatment factors (ns = P > 0.05; *= P < 0.05). B). Mean seasonal
nitrogen concentrations (mean + 1 SE) in the two treatments where all
vegetation (neighbors and litter) was either removed or left intact. Sites are
ranked from lowest to highest community biomass. ........................................ 26

Figure 3. A) Andropogon density and B) total shoot biomass (mean + 1
SE) in relation to the experimental treatments (seed addition
experiment). ................................................................................................................ 31

Figure 4. Relationships between the community effects on Andropogon
density (dependent variables) and community biomass (independent
variable) in the seed addition experiment. A) Mean community effects
plotted against mean community biomass (n = 5). B) Community effects
calculated for all blocks plotted against block community biomass (n = 25).
Dashed line represents the no effect line. Note log scale for community
biomass. ........................................................................................................................ 34

Figure 5. Relationships between the community effects on Andropogon

total shoot biomass (independent variables) and community biomass
(dependent variable) in the seed addition experiment. A) Mean

community effects plotted against mean community biomass (n = 5).

B) Community effects calculated for all blocks plotted against block
community biomass (n = 25). .................................................................................. 36

Figure 6. RGR (g-g'l-d'l) of Andropogon transplants (mean + 1 SE) in
relation to the experimental treatments. .............................................................. 39

Figure 7. Relationships between the community effects on the RGR of
transplants (dependent variables) and community biomass (independent
variable) in the transplant experiment. A) Mean community effects

plotted against mean community biomass (n = 5). B) Community effects
calculated for all blocks plotted against block community biomass

(n = 30). .......................................................................................................................... 40

Figure 8. Experimental block layout illustrating the arrangement of the
experimental treatment plots and the transplant and seed addition

subplots. T1 and T2 refer to the position of the two transplants in each
transplant subplot. .................................................................................................... 54

Figure 9. Mean seasonal light penetration (mean + 1 SE) in relation to

the experimental treatments. Treatments with neighbors removed are
indicated by - N, and treatments with neighbors left intact are indicated

by +N. Treatments with litter removed are indicated by - L, and

treatments with litter left intact are indicated by shading and by +L.

Letters above the bars specify the signiﬁcance of neighbor (N) and litter

(L) effects within the sites (ns = P > 0.05 ; * = P < 0.05). At sites with a
signiﬁcant neighbor x litter interaction, brackets above and below the

bars specify treatment mean comparisons made at each level of the two
treatment factors (ns = P > 0.05 ; * = P < 0.05). ...................................................... 62

Figure 10. Andropogon density (A) and total shoot biomass (B; mean + 1
SE) in relation to the experimental treatments (seed addition). ..................... 63

Figure 11. Schizachyrium density (A) and total shoot biomass (B; mean
+ 1 SE) in relation to the experimental treatments (seed addition). ............... 66

Figure 12. Relationships between the community effects on Andropogon
density (A) and total shoot biomass (B; dependent variables) and

community biomass (independent variable). Dashed line represents the

no effect line. ............................................................................................................... 70

Figure 13. Relationships between community effects on Schizachyrium
density (A) and total shoot biomass (B; dependent variables) and
community biomass (independent variable). ..................................................... 72

Figure 14. RGR (g-g'Ld‘l) of transplants (mean + 1 SE) in relation to the
experimental treatments. NT = not tested statistically. ..................................... 75

Figure 15. Relationships between the community effects on Andropogon

transplant RGR (g-g'l-d'l) and community biomass (independent variable).
ns P > 0.05, ** P < 0.001. ............................................................................................. 78

xi

Figure 16. Relationships between the community effects on Schizachyrium
transplant RGR (g-g'l-d'l) and community biomass (independent variable).
ns P > 0.05, ** P < 0.001, "* P < 0.0001. .................................................................. 79

Figure 17. Relationships between the community effects on Sorghastrum
transplant RGR (g-g‘l-d'l) and community biomass (independent variable).
ns P > 0.05. .................................................................................................................... 80

Figure 18. Diagrammatic representation of the experimental block and
treatment plot layout. Block numbers are indicated by roman numerals.

The litter addition (LA) plots that were added after the initiation of the
experiment are indicated by heavy lines. ............................................................ 94.

Figure 19. Treatment effects on living biomass (A), litter biomass (B), grass
biomass (C), forb biomass (D), forb density (E), and species richness (F).

Means (+ 1 SE) with different letters are signiﬁcantly different (P < 0.05).

See methods section for treatment codes. ANOVA F430 values:

ns P > 0.05, *** P < 0.001, W” P < 0.0001. .............................................................. 98

Figure 20. Treatment effects on light penetration. Means (+ 1 SE) with
different letters are significantly different (P < 0.05). ANOVA F430
value: W” P < 0.0001. ............................................................................................... 99

Figure 21. Treatment effects on Andropogon density (A), and biomass per
plant (B). Means (+ 1 SE) with different letters are signiﬁcantly different
(P < 0.05). ANOVA F430 values: W” P < 0.0001. .............................................. 105

Figure 22. Treatment effects on Andropogon germination (A), and
survivorship (B). Means (+ 1 SE) with different letters are signiﬁcantly
different (P < 0.05). ANOVA F430 values: *** P < 0.001, W” P < 0.0001. 106

xii

CHAPTER 1

INTRODUCTION

INTRODUCTION

A major goal of plant ecologists is to develop models of community
organization that predict community composition and diversity at any
location in the landscape. Attaining this goal is not only of interest to
ecologists, but is also of value to those that seek to manage or restore the
diversity of ecological communities. As a result, much work in plant ecology
has centered on describing and explaining ecological processes across
environmental gradients, including gradients in plant productivity, soil
fertility, disturbance, and physical stress (Newman 1973, Grime 1979, Tilman
1988, Bertness and Callaway 1991, Wilson and Tilman 1991, Campbell and
Grime 1992).

There has been considerable interest and debate focused on the
unimodal relationship that is often observed between plant species diversity
and plant productivity. This "hump-shaped" relationship, which has been
documented in a number of ecosystems (Dix and Smiens 1967, Al-Mufti et a1.
1977, Puerto et a1. 1990), exhibits an increase in species diversity up to some
intermediate level of plant productivity, followed by a decline in diversity
with further increases in productivity. Although there is some debate as to
the generality of the unimodal diversity curve (Abrams 1995), several
mechanisms have been proposed to explain it (Tilman 1988, Tilman and
Pacala 1993, Huston and DeAngelis 1994). A great deal of attention has
centered on the mechanisms responsible for reduced diversity at high levels
of productivity (declining portion of the unimodal diversity curve), not only
due to theoretical interest, but also because of the potential negative
consequences that long-term eutrophication has for the conservation of plant

diversity. Grime (1973, 1979) and others (Huston 1979, Thompson and Grime

1988, Keddy 1990) have argued that diversity often declines with increases in
productivity because the overall magnitude of competition occurring among
plants increases, thus leading to dominance by a few superior competitors in
the most productive habitats. Tilman (1988) and Tilman and Pacala (1993)
proposed that species coexistence depends upon the spatial heterogeneity of
resources and suggest that low resource heterogeneity in highly productive
habitats may limit the number of species that can coexist there, thus
promoting low diversity. Goldberg and Miller (1990) argued that strong
competition for light limits the number of species in productive habitats by
increasing the mortality of seedlings and species of small stature.

These theories of plant community organization all have in common
the idea that increased competitive exclusion leads to reduced diversity in
highly productive habitats. However, there continues to be disagreement in
the literature regarding just how competition varies across gradients in plant
productivity. Productivity-dependent theories of competition advocated by
Grime (1973, 1979) and others (Huston 1979, Southwood 1988, Keddy 1990)
suggest that the overall magnitude of competition occurring among plants is
lowest in unproductive habitats, but increases with productivity as plant
biomass and plant demand for both above- and below-ground resources
increase. Others have argued that competition should be very strong in
habitats of low productivity due to strong competition for soil resources, and
that competition will change more in quality (shift from below- to above-
ground) than in quantity as productivity increases (Newman 1973, Tilman
1988, Taylor et a1. 1990). This has led to a productivity-independent View of
competition which predicts that the over-all magnitude of competition
should be unrelated to productivity. This latter prediction is often associated

with the resource ratio hypothesis of competition (Tilman 1982, 1988).

However, it should be noted that the resource ratio hypothesis makes no
explicit prediction regarding the over—all magnitude of competition and
relates only to how the outcome of competition between species changes as
resource supply rates change. The prediction that the over-all magnitude of
competition between plants is unrelated to productivity is a verbal extension
of this hypothesis.

A number of recent experimental studies have been conducted to
examine how the over-all magnitude of plant competition varies across
gradients of plant productivity or soil resource availability. These studies
have provided mixed results with some studies supporting the productivity-
dependent view of competition and showing increasing competition with
productivity (Gurevitch 1986, Wilson and Keddy 1986, Bonser and Reader
1995, Kadmon 1995), while some results are consistent with the productivity-
independent View and report no relationship (Fowler 1990, Wilson and Shay
1990, Wilson and Tilman 1991, 1993). Inconsistency in these previous results
has been attributed to different measures of competition (absolute versus
relative measures; Grace 1993, Kadmon 1995), differences in the range of
productivity examined (Goldberg and Barton 1992, Bonser and Reader 1995,
Kadmon 1995), whether or not the gradient is natural or experimentally
created (Goldberg and Barton 1992, Kadmon 1995), and differences in target
species (Wilson and Tilman 1995). At the present time there are too few
studies encompassing these different conditions to adequately assess which of
the opposing views is most applicable to natural plant communities.

While most of the focus in these studies has been on resource
competition occurring among living plants and its role in inﬂuencing plant
species diversity, these studies do not address other potential mechanisms of

plant-plant interaction that may affect individual plant performance and

plant species diversity across productivity gradients. One mechanism which I
focus on in this dissertation involves the role of accumulated plant litter in
herbaceous old-ﬁeld communities. A number of studies have shown that
plant litter can either promote or inhibit plant performance in natural
communities (Fowler 1986, Wilms et a1. 1986, Facelli and Pickett 1991a). As a
whole these studies indicate that the magnitude (intensity or strength) of
litter effects and their direction (interference versus facilitation) may vary
signiﬁcantly among sites that differ in plant productivity and litter
accumulation.

In this dissertation I describe the results of three manipulative ﬁeld
experiments designed to examine how the effects of accumulated plant litter
can combine or interact with the exploitative effects of living plant neighbors
to limit species diversity and affect the performance of individual target
species across gradients of productivity, community biomass and soil fertility

in old-ﬁeld vegetation.

Organization of the Dissertation

In addition to this Introductory chapter (Chapter 1), there are three
empirical chapters and a Concluding chapter. Chapter 2 describes the results
of an experimental field study that I conducted to examine variation in the
effects of living plant neighbors and litter on the recruitment and growth of a
native perennial grass across five herbaceous old-ﬁeld communities that
differ in community biomass and soil fertility. Chapter 3 presents the results
of a second ﬁeld experiment that I conducted the following year which
extends the objectives of the ﬁrst experiment to incorporate more ﬁeld sites, a

greater range of plant productivity, and additional target species. In chapter 4 I

present the results of a two-year ﬁeld experiment where I manipulated plant
litter levels and added nitrogen fertilizer to ﬁeld plots to examine the short-
term effects of increased fertility, increased living plant biomass, and
increased litter biomass on species richness in a grass-dominated old-ﬁeld. In
the concluding chapter (Chapter 5) I brieﬂy summarize the findings of the
three ﬁeld experiments and discuss there general significance with respect to
existing theories of plant community organization and ecological

applications.

Study species

A major part of this dissertation examines the effects of living plant
neighbors and accumulated plant litter on the recruitment and growth of
three target grass species that were experimentally introduced into a number
of old-ﬁeld sites as both seeds and transplants. These species include:
Andropogon gerardi Vitman (Big bluestem), Schizachyrium scoparium Nash
(Little bluestem), and Sorghastrum nutans (L.) Nash (Indian grass). These
species are all perennial C4 bunchgrasses native to southwest Michigan and
are the dominant species of the tallgrass prairies of North America (Bazzaz
and Parrish 1982). I chose these species for study because one of the goals of
the research was to examine potential biotic constraints to the re-colonization
of current successional communities in southwest Michigan by historically
dominant native species (Gotshall 1972). These species are similar with
respect to photosynthetic pathway, life history, growth form, growth rate, and
ability to compete for soil resources (Parrish and Bazzaz 1982, Tilman and
Wedin 1991), but are distributed somewhat differently across environmental

gradients. In intact prairies, Andropogon and Sorghastrum are distributed

similarly across broad gradients of plant productivity, soil moisture, and soil
fertility (Weaver 1954, Smeins and Olson 1970), while Schizachyrium is
typically more restricted in distribution to the more xeric, infertile sites
(Weaver 1954, Bazzaz and Parrish 1982). The more limited distribution of
Schizachyrium appears to be related to it adaptation to xeric conditions
(Weaver 1954, Knapp 1985). Schizachyrium has lighter seeds (Rabinowitz
1978, Foster, unpublished data) than Andropogon and Sorghastrum and is
considered to be a poor competitor for light (L. Wallace, personal
communication), both factors which may restrict this species from
establishing and persisting in the most productive sites.

In the current successional landscape of southwest Michigan, all three
of these species are largely restricted to low to mid- productivity old-ﬁeld sites.
These species are generally absent from the most productive sites that are
typically dominated by introduced perennial grasses. Site pre-emption by the
early-colonizing introduced grasses and dispersal limitation likely interact to
limit the current distribution of these native grasses. Because ﬁre is currently
rare in this landscape, litter accumulation may be an important factor
limiting the re-colonization of productive sites by these native prairie species
which evolved in the presence of periodic ﬁre and which are not likely to be

adapted for establishment in dense litter (Tilman 1993).

Study sites

The three ﬁeld experiments presented in this dissertation were all
conducted in herbaceous old-fields communities at the W. K. Kellogg
Biological Station (KBS) located in southwest Michigan (Kalamazoo County;

42° 24' N, 85° 24' W). A total of eleven sites were used in these studies (Table

1, Fig. 1). All of these sites had been abandoned from row crop agriculture for
at least 15 years and were dominated by perennial grasses at the time of study.
The sites varied substantially in species composition (see Burbank et al. 1992),
above-ground living plant biomass, and litter biomass (Table 1). Soils
underlying the sites are classiﬁed as sandy loam, but differ in the amount of
top-soil lost to agricultural-related erosion.

Five of the eleven study sites were used in the ﬁrst competition
experiment (Chapter 11; sites A-E; Table 1, Fig. 1). Three of the sites used in the
ﬁrst competition experiment (B, C and E) plus six additional ones (F-K; 9 sites
total) were used in the second competition experiment (Chapter III). The

nitrogen addition experiment described in chapter IV was conducted at site C.

Table 1. Locations, identification numbers, and vegetation characteristics of
the eleven study sites. Letters correspond to site locations shown on the
Kellogg Biological Station map (Fig. 1). Biomass values are presented as
means i 1 SE. Total community biomass is the sum of aboveground living
plant biomass and plant litter. Methods for assessing plant biomass and
species composition are presented in the methods sections of chapters 2 and 3.
Nomenclature follows Fernald (1950).

Site ID # Site ID #

 

Site 1994 Exp. 1995 Exp. Total community Four most abundant Relative
Location Chapter 1 Chapter III biomass (g/mz) species biomass (%)
A 1 - 207.9:319 Centaurea maculosa 24.0

Poa compressa 22.5
Achillea millefolium 11.4
Aster pilosis 8.7
Rubus alleghaniensis 6.3
B 2 4 45461334 Agropyron repens 85.2
Achillea millefolium 7.3
Bromus inermis 6.4
C 3 5 527.5 239.4 Bromus inermis 58.1
Poa pratensis 10.3
Achillea millefolium 10.1
Agropyron repens 5.1
D 4 - 658.5 1: 73.7 Bromus inermis 88.3
Part pratensis 7.5
E 5 7 1205.5 1 48.4 Agropyron repens 71.0
Arrhenatherum elatius 28.3
F - 1 79.1 :t 16.1 Andropogon virginicus 50.0
Rubus alleghaniensis 13.9
Danthonia spicata 9.9
Solidago nemoralis 5.7
G - 2 129.1 143.2 Rubus alleghaniensis 27.8
Danthonia spicata 17.2
Centaurea maculosa 14.7
Andropogon virginicus 11.0
H - 3 221.4:148 Andropogon virginicus 49.8
Danthonia spicata 9.8
Rubus alleghaniensis 6.0
Desmodium rotundifolium 4.8
I - 6 626.9 :1: 115.7 Bromus inermis 78.0
Agropyron repens 10.3
Poa pratensis 3.8

Daucus carota 3.2

Table 1. (Cont'd)

8

9

10

1470.9 1 102.4

1973.3 i 223.8

Bromus inermis

Poa pratensis
Solidago canadensis
Agropyron repens

Bromus inermis
Poa pratensis
Agropyron repens
Barbarea vulgaris

44.2
19.1
17.4
15.4

90.6
4.9
2.5
0.1

11

336 $2: «0 28:98me 5

H8 H 2an mum .mmzm 363m 2 of mo 98382 2.: wcdgocm Beam: 5:: ~8sz Euwwoﬁomm wwosmx mﬁ mo 932 A 85me

 

 

25-532 MMHA

81..— 58985?

33 3.8.23 D

 

 

 

 

 

8:02 852 D

 

 

 

 

CHAPTER 2
REMOVAL EXPERIMENT I: PARTITION IN G THE EFFECTS OF PLANT

BIOMASS AND LITTER ON ANDROPOGON GERARDI
IN OLD-FIELD VEGETATION

12

13

INTRODUCTION

Experimental ﬁeld studies designed to examine variation in plant
interactions among habitats or along environmental gradients often use plant
removal experiments to compare the performance of target plants growing in the
presence and absence of the surrounding plant community (Wilson and Keddy
1986, Wilson and Tilman 1991, 1993, Bertness and Hacker 1994, Kadmon 1995).
When negative effects of the surrounding community on target plants are
observed in these experiments, the effect is most often attributed to competition
with plant neighbors for limiting resources (Wilson and Shay 1990, Wilson and
Tilman 1991, 1993, Bonser and Reader 1995). However, in many types of
vegetation, especially grassland and old-ﬁeld communities, the performance of
plants can also be affected by accumulated dead plant biomass or plant litter
(Hulbert 1969, Goldberg and Werner 1983, Fowler 1986, Hamrick and Lee 1987,
Facelli and Pickett 1991a, Facelli 1994). As a result, measures of competition
obtained from removal experiments, and variation in these measures observed
across different habitats, may reﬂect the effects of both living plant neighbors
and litter. In this study I investigated how these two different types of effects
may combine to determine the net effect of the surrounding plant community on
target plants in old-ﬁeld vegetation. In addition, I examined how the magnitude
of these effects vary across old-ﬁeld sites that differ in community biomass and
soil nitrogen availability.

A number of ﬁeld studies in herbaceous vegetation have shown that plant
litter can signiﬁcantly affect the performance of individual plants, and as a result
inﬂuence the structure of plant communities. Large quantities of litter can inhibit
the establishment, growth, and survival of plants, and thus contribute to low

species diversity in highly productive communities (Haslam 1971, Goldberg and

14

Werner 1983, Bergelson 1990, Carson and Peterson 1990, Tilman 1993). Other
studies suggest that litter may often facilitate the establishment and growth of
plants and enhance species diversity in unproductive environments by
improving moisture conditions (Fowler 1986, Willms et al. 1986). As a whole,
these studies suggest that the role of plant litter may vary across environments
and that changes in species composition and diversity along gradients of
community biomass and plant productivity may reﬂect variation in the balance
between the effects of plant neighbors and litter.

In this study, I examined the effects of living plant neighbors and litter on
a single target species, the C4 grass Andropogon gerardi Vitman (Big bluestem), in
herbaceous old-ﬁelds in southwest Michigan. Andropogon gerardi (hereafter
referred to by genus) is a dominant species in the tall-grass prairies of the United
States where it is distributed widely across gradients of soil resource availability
and plant productivity (Bazzaz and Parrish 1982). Andropogon was a dominant
species in many prairie and oak savanna communities of southwest Michigan at
the time of settlement by Europeans (Gotshall 1972). However, in the current
successional landscape of southwest Michigan, Andropogon is largely restricted to
low fertility sites, and is infrequent in the more fertile sites that are commonly
dominated by non-native C3 grasses such as Agropyron repens L. Beauv. and
Bromus inermis Leysser. The apparent poor ability of native C4 grasses, such as
Andropogon, to colonize and occupy fertile sites after abandonment from
agriculture, is likely related to a combination of: 1) delayed seed input, due to the
isolation of source populations in the landscape, and inherently low allocation to
seed production (Tilman and Wedin 1991a); and 2) site pre-emption and
inhibition by fast-growing C3 grasses which are effective at rapidly colonizing
and dominating fertile sites shortly after abandonment (Tilman 1987, Tilman and

Wedin 1991b, Foster 1992). Because ﬁre is largely absent from this landscape,

15

effects associated with litter accumulation in the most productive sites may be
important in restricting the distribution of native C4 grasses to lower
productivity sites. In this study, I addressed the following questions regarding
the effects of the surrounding plant community on Andropogon at a ﬁve old-ﬁeld

sites:

1) How do living plant neighbors and litter affect the recruitment and growth of

Andropogon in old ﬁelds?

2) How do the separate effects of neighbors and litter contribute to the net effect

of the surrounding plant community on Andropogon?

3) Do these effects vary in magnitude among the different study sites, and if so,

are they correlated with above-ground community biomass.

METHODS

Study sites

This study was conducted in ﬁve herb-dominated old-ﬁeld sites at the W.
K. Kellogg Biological Station (KBS) of Michigan State University in southwest
Michigan (Kalamazoo County; 42° 24' N, 85° 24' W). All ﬁve sites had been
abandoned from row crop agriculture for at least 15 years and were largely
dominated by non-native perennial grasses. (Table 2; Burbank et al. 1992).
Naturally occurring individuals of Andropogon were found growing at sites 1 and
3. Soils underlying the sites are classiﬁed as sandy loam, but differ in the amount

of top-soil lost to agricultural-related erosion. Assignment of site numbers (1-5)

16

Table 2. Community characteristics of the ﬁve study sites. Biomass values are
presented as means 1 1 SE (n = 5 for each site). For each biomass type, values
within a column with the same letter are not signiﬁcantly different (P > 0.05;
Tukey’s HSD). Nomenclature follows Fernald (1950).

 

Total
Neighbor Litter community
biomass biomass biomass Dominant Species Relative
Site (g /m2) (g/mz) (g/mz) (> 5% Rel. biomass) biomass (%)
1 184.2 1 24.83 23.7 1 8.0a 207.9 1 31.9a Centaurea maculosa 24.0
Poa compressa 22.5
Achillea millefolium 11.4
Aster pilosis 8.7
Rubus alleghaniensis 6.3
2 222.9 4 15.63 231.7 4 21.6b 454.6 4 33.4b Agmpyron repens 85.2
Achillea millefolium 7.3
Bromus inermis 6.4
3 310.8 4 11.9b 216.7 4 31.1b 527.5 4 39.4bc Bromus inermis 58.1
Poa pratensis 10.3
Achillea millefolium 10.1
Agropyron repens 5.1
4 337.4 4 20.2b 321.0 4 81.0b 658.5 4 73.7c Bromus inermis 88.3
Part pratensis 7.5
5 598.4 1 31.0C 607.1 1 34.2C 1205.5 1 48.4d Agropyron repens 71.0

Arrhenatherum elatius 28.3

 

17

was based on a ranking of total community biomass (1 = lowest and 5 = highest
biomass). In this study, I deﬁne total community biomass (or community

biomass) as the sum of above-ground living plant biomass and litter biomass.

Experimental design and methods

In 1994 both seeds and seedling transplants of Andropogon were planted
into a 2 x 2 factorial arrangement of treatments established at each site in a
randomized block design. The treatments included two levels of living plant
neighbors (intact and removed) and two levels of plant litter (intact and
removed). At all sites except site 2, two parallel rows of ﬁve 2 x 2 m experimental
blocks were established in early May. One of the rows was used for seed
additions and the other row was used for transplants. At site 2, the same block
layout was used with the exception that ten blocks were established for the
transplant experiment instead of ﬁve. In each ﬁeld, the rows of blocks were
separated by buffer zones of 3 to 6 m, and blocks within rows were separated by
buffer zones ranging from 2 to 4 m.

The four experimental treatments were applied to 70 x 70 cm plots located
at the four corners of each block. Neighbors were removed by applying a
Glyphosate herbicide solution (Roundup), and clipping the dead material after
one week Litter was removed by hand from the litter removal treatments so that
the soil was left undisturbed. To minimize root encroachment into the two
neighbor removal treatments, plot perimeters were trenched with a ﬂat blade
shovel to a depth of 20 cm three times during the growing season. Neighbor
removal plots were periodically hand-weeded during the growing season to

eliminate re-growth.

18

Seed additions - Seeds of Andropogon were collected from a local population
in the fall of 1993 and stored outside over winter in a plastic container to expose
the seeds to winter temperature ﬂuctuations. Seeds were sown by hand into the
treatment plots at a rate of 1500/ plot (3030/ m2) in mid-May 1994. Because the
goal of this experiment was to examine recruitment from seed in the absence of
dispersal limitation, a seeding rate was chosen that mimicked the seed rain
within natural stands of Andropogon (about 3000 seeds/ m2, unpublished data).
Density of AndrOpogon plants in each plot was determined in June, July and late
August by counting individuals within two randomly placed 10 x 20 cm
quadrats. Final density determined in late August is reported here and is used
throughout as a measure of Andropogon recruitment. In late August, all shoot
biomass of Andropogon was harvested from each plot. In addition, above-ground
biomass of plant neighbors and litter was harvested from the intact plots
(neighbors and litter intact). All plant material was oven-dried at 60° C to
constant mass and then weighed.

Seedling transplants - Seedlings for the transplant experiment were
obtained by germinating seeds in the greenhouse in late April 1994.
Approximately 68 days after germination, 1000 seedlings were transferred into
individual seedling plug containers (2.54 cm diameter x 10.16 cm deep)
containing a 3:1:1 mixture of peat moss, commercial potting soil, and old-ﬁeld
soil. After 14-17 days of additional growth, 480 seedlings were randomly selected
for the experiment from a group of approximately 800 healthy seedlings. The
initial shoot biomass of each seedling was estimated from a regression of
seedling biomass on longest leaf length (L) obtained from a subsample of the
remaining healthy seedlings (Initial transplant biomass = 0.004L - 0.011, R2: 0.65,
P < 0.0001, n = 100).

19

In late May, four seedlings were transplanted into each 70 x 70 cm ﬁeld
plot by inserting seedling plugs into pre-formed holes made with a narrow
bladed trowel. Within each plot, transplants were separated by approximately 20
cm. All individual transplants were watered (0.2 L/ day) for eight consecutive
days following planting to encourage initial establishment. Only 16 of the 480
transplants died and there were no plots where less than three transplants
survived until the end of the experiment. The shoots of all surviving transplants
were harvested in early September. Above-ground biomass of neighbors and
litter was harvested from the intact plots. Neighbors were sorted by species to
document the species composition of each site. All plant tissue was dried and

weighed as described previously.
Resource measurements

To determine how light penetration to the soil surface varied among the
sites and was affected by the experimental treatments, photosynthetically active
radiation (PAR) was measured above the vegetation and at the ground surface of
each plot in the transplant experiment with a PAR ceptometer probe (Decagon
Devices Inc. ). PAR measurements were taken three times over the course of the
experiment at approximately monthly intervals (early June - late August). In
plots where litter was left intact, light measurements were made by inserting the
ceptometer beneath the litter layer. In the neighbor removal plots, measurements
were made at plot edges between the hours of 11 am and 2 pm when the sun was
directly overhead to minimize the effects of transplants on light readings. Light
penetration was expressed as the percentage PAR reaching the ground surface.
Mean seasonal light penetration was calculated for each plot by averaging over

the three sampling dates.

20

Soil nitrogen concentration (N) was also measured at three times over the
course of the experiment in two treatments in the seedling transplant experiment:
neighbors and litter removed; neighbors and litter left intact. At each sampling
date, two soil cores (2.5 cm in diameter x 15 cm deep) were taken from each plot.
All cores were taken at least 10 cm away from the base of transplant shoots and
care was taken to minimize disturbance to the plots. The two cores from each
plot were composited for analysis. Subsamples (10g) were extracted for 24 hours
in a 2 mol/ L KCL solution, ﬁltered, and then analyzed colorimetrically with an
Alpkem auotoanalyzer for ammonium and nitrate. Soil nitrogen concentrations
are expressed as the sum of extractable ammonium and nitrate concentrations
(ug N / g dry soil). Mean seasonal nitrogen concentrations were calculated for

each plot by averaging over the three sampling dates.

Data analysis

Community biomass - To examine whether sites differed in neighbor
biomass, litter biomass and total community biomass (sum of neighbor and litter
biomass), I used one-way analysis of variance (AN OVA) and Tukey HSD tests.
These analyses were performed on biomass data from the intact treatment
(neighbor and litter biomass left intact).

Resources and target plant performance — Percent light penetration, total soil
nitrogen, Andropogon seedling density and total shoot biomass (seed addition
experiment), and transplant growth (transplant experiment) were analyzed using
a mixed-model, within-subjects AN OVA (Zar 1996) with one among-subjects
grouping factor (site). In this model the experimental blocks are considered to be
the subjects of the analysis with the experimental treatments applied to each.

Variation due to blocks within sites is considered random, while site and

21

treatment effects are considered ﬁxed. Error terms used to test the signiﬁcance of
main effects and interactions all contain the random component of variance due
to blocks within sites (see Table 3). After testing the signiﬁcance of the interaction
terms, unplanned comparisons were used to make the appropriate comparisons
of treatment means (Sokal and Rohlf 1981). Analyses of transplant growth were
performed on plot mean values (mean of all surviving transplants). Relative

growth rate (RGR, Chiariello et al. 1989) was calculated for each transplant as:

RGR = [111032) -1n(31)l/(t2 - t1)

where B2 is the transplant biomass at harvest, B1 is the estimated transplant
biomass at planting, and t2 - t1 is the number of days growth between harvest (t2)
and planting (t1). Light penetration data were arcsine-square root transformed to
reduce heteroscedasticity. Andropogon total shoot biomass and density data were
log transformed and square-root transformed respectively to improve normality
and reduce heteroscedasticity (Sokal and Rohlf 1981). Untransformed data are
presented in the ﬁgures.

Magnitude of community effects - The magnitude of the effect of the
surrounding plant community on Andropogon performance (density, total shoot
biomass, and RGR of transplants) was examined using a relative community
effect index (diffuse competition, Wilson and Keddy 1986; competition intensity,
Wilson and Tilman 1993). This index estimates the degree to which Andropogon
was inhibited or facilitated by the surrounding plant community by comparing
plant performance in plots cleared of all plant material (neighbors and litter) to
that in plots where all plant material was left intact. Positive values indicate
inhibition, negative values indicate facilitation, and values of zero indicate no net

effect of the surrounding community. To assess the importance of litter in

22

Table 3. Treatment effects and means square ratios used in the mixed-model,
within-subjects AN OVA. S = Sites, B(S) = Blocks within Sites, N = Neighbors, L =
Litter.

 

Source .
of variation Effect MS/ Error MS

Among Blocks

S S/B(S)
Within Blocks

N N / N x B(S)

NxS NxS/NxB(S)

L L/ L x B(S)

LxS LxS/LxB(S)

NxL NxL/NxLxB(S)

NxLxS NxLxS/NxLxB(S)

 

23

contributing to the net effect of the community on AndrOpogon performance, a
separate index was calculated for plots that were left intact and for plots where

only litter was removed. These indices were calculated for each experimental

block as:

Net Effect = (PC - PNL)/ PC

Neighbor Effect = (PC - PN)/ PC

where PC is plant performance in plots cleared of neighbors and litter, PNL is
performance in plots with neighbors and litter left intact, and PN is performance
in plots with only neighbors left intact (litter removed). An additional index was

calculated to measure the effect of litter alone and was calculated as:

Litter Effect = (PC - PL)/Pc

where PL is plant performance in plots with only litter left intact (neighbors
removed).

I used one-way AN OVA and Tukey HSD tests to examine site differences
for each community effect index (net, neighbor and litter effects). I then
examined correlations between each index and community biomass using linear
regression in two ways. First, I analyzed the site means of each community effect
index and of community biomass (n = 5). Second, to examine relationships across
the entire range of community biomass, thus including both within and among
site variation, I analyzed the community effect indices and community biomass
measured in each individual block (n = 25, seed addition experiment; n = 30,

transplant experiment). Preliminary data inspection indicated that the

24

community effect indices tended to vary non-linearly with community biomass.
As a result, community biomass was log transformed prior to linear regression
analysis. I carried out similar analyses using total soil nitrogen measured in the
intact plots as the independent variable in place of community biomass. The
results of these analyses were very similar, so for brevity I report only the
analyses using community biomass as the independent variable. SYSTAT

statistical software (version 5.2.1; Wilkinson 1992) was used for all analyses.

RESULTS

Community biomass and species composition

Neighbor biomass, litter biomass, and total community biomass varied
signiﬁcantly among the five sites (neighbor biomass, 134,25 = 55.59, P < 0.0001;
litter biomass, 134,25 = 26.28, P < 0.0001; total community biomass, F4, 25 = 57.91, P
< 0.0001 ; Table 2). There was a six-fold increase in mean community biomass
from site 1 to site 5. Mean litter biomass increased 25-fold across these sites and
was positively correlated with community biomass (R2 = 0.96, P < 0.05, n = 5).
Most sites were dominated by perennial grasses. A perennial forb (Centaurea

maculosa Lam.) and perennial grass (Poa compressa L.) were co-dominant at site 1.

Resources

Light penetration to the soil surface varied signiﬁcantly among the sites,
with neighbors, litter, and with interactions between litter and sites, and between
neighbors and litter (Table 4, Fig. 2A). Neighbors reduced light penetration to the

soil surface at all sites in plots where litter was removed, but in plots where litter

25

Table 4. Results of AN OVA for site and treatment effects on mean seasonal light
penetration .

 

Source
of variation df MS F
Among Blocks
Site 4 0.4946 52.61””
Block ' 20 0.0094
Within Blocks
Neighbors 1 1.7493 164.77*’”
Neighbors x Site 4 0.0029 0.28“5
Neighbors x Block 20 0.0106
Litter 1 3.3350 297.40’“m
Litter x Site 4 0.1817 1620*"
Litter x Block 20 0.0112
Neighbors x Litter 1 0.4847 4917*“

Neighbors x Litter x Site 4 0.0420 4.26"

Neighbors x Litter x Block 20 0.0099
“5 P > 0.05; * P < 0.05; ’” P < 0.01; *** P < 0.001.

 

26

Figure 2. A) Mean seasonal light penetration (mean + 1 SE) in relation to
the experimental treatments. Treatments with neighbors removed are
indicated by - N, and treatments with neighbors left intact are indicated by
+N. Treatments with litter removed are indicated by - L, and treatments
with litter left intact are indicated by shading and by +L. Brackets above
and below the bars specify within-site mean comparisons made at each
level of the two treatment factors (ns = P > 0.05; * = P < 0.05). B). Mean
seasonal nitrogen concentrations (mean + 1 SE) in the two treatments
where all vegetation (neighbors and litter) was either removed or left intact.
Sites are ranked from lowest to highest community biomass.

27

 

 

 

 

 

+N

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 "

80 '
60
40
20 ‘

$3 cozmbmcmm Em: 383% :82

 

 

 

 

 

 

 

 

 

 

 

 

 

T.

[:I -Vegetation

 

 

_jlll

 

10"

u a
6

8' I +Vegetation

4
2.1
0-

38 be 32 mg

cosmbguce 5onme 3:033 :32

 

Site

Figure. 2

28

was present, neighbors reduced light penetration to the soil surface only at site 1.
At site 1, litter levels were so low that litter had no effect on light penetration.
However, in the other four sites, light penetration in plots containing litter only
was similar to that in plots containing both neighbors and litter. This suggests
that light levels at the soil surface in the intact vegetation of these sites was
determined more by litter than by neighbors. Mean seasonal light penetration
measured in the intact plots declined logrithmically with community biomass
(light penetration = -0.193 (log community biomass) + 1.315, R2 = 0.52, P < 0.0001,
n = 25) and mean seasonal nitrogen concentration (light penetration = -0.282 (log
N) + 0.423, R2 = 0.51, P < 0.0001, n = 25).

The removal of all plant material (neighbors and litter) increased mean
seasonal nitrogen concentrations signiﬁcantly across all sites (Table 5, Fig. 2B),
ranging from 56% at site 5 to 218% at site 1. Mean seasonal nitrogen
concentration measured in the intact plots was positively correlated with

community biomass (R2 = 0.83, P < 0.0001, n = 25).

Seed addition

Seedling recruitment - The number of Andropogon seedlings that established
in the plots did not differ signiﬁcantly among the ﬁve sites, but did vary in
response to neighbors and litter, and with the interaction between litter and sites
(Table 6). Across all sites, the removal of neighbors led to a signiﬁcant increase in
Andropogon density (Fig. 3A). The signiﬁcant interaction between sites and litter
indicates that the degree to which litter affected Andropogon density varied
among sites. At sites 1 and 2, litter removal had no signiﬁcant effect on

Andropogon density. However, at the three sites with the greatest community

29

Table 5. Results of ANOVA for site and treatments effects on mean seasonal soil
nitrogen concentration. Nitrogen concentrations were measured only in the two
treatments where all plant material (neighbors and litter) was either removed or
left intact.

 

Source
of variation df MS F
Among Blocks
Site 4 17.360 778*“
Block 20 2.229
Within Blocks
Vegetation 1 123.884 7317*“
Vegetation x Site 4 0.280 0.16“5
Block 20 1.693

 

“5 P > 0.05; * P < 0.05; ** P < 0.01; W‘ P < 0.001.

30

Table 6. Results of AN OVA for site and treatment effects on Andropogon density
and total shoot biomass in the seed addition experiment.

 

 

Density Total shoot biomass
Source
of variation df MS F MS P
Among Blocks
Site 4 41.438 0.95“5 12.542 15.86’“m
Block 20 43.425 0.791
Within Blocks
Neighbors 1 3659.415 112.52’”m 101.729 171.84***
Neighbors x Site 4 18.252 0.56“ 9.182 15.51"M
Neighbors x Block 20 32.523 0.592
Litter 1 1325.331 4357*“ 27.010 48.40“”
Litter x Site 4 322.504 1060”“ 8.156 1462*“
Litter x Block 20 30.416 0.558
Neighbors x Litter 1 96.930 3.79"5 5.732 9.67“
Neighbors x Litter x Site 4 24.481 0.96“5 2.649 4.47“
Neighbors x Litter x Block 20 25.556 0.593

 

“5 P > 0.05; * P < 0.05; ’” P < 0.01; “* P < 0.001.

31

Figure 3. A) Andropogon density and B) total shoot biomass (mean + 1 SE)
in relation to the experimental treatments (seed addition experiment).

N”, L:

32

-Litter
+Litter

  

IIIIIIIIII

\\\\\\\\\

IIIIIIIIII
\\\\\\\\\

I54|l§lll| I~ ININ’5I N

   

   
  

IIIIIlI’IIII'
\\\\\\\\\\\\

IIIIIIIIIIIIP

\\\\\\‘\\\\\

IIIIIIIIIIIII
1\\‘§\~\\\\\\ ”N

    

     
    

 

 

 

 

 

+N

 

     

IIIIIIIIIIIIIIIIIIIII

‘\\\\\\\\\\\\\‘\\\\\\

IIIIIIIIIIIIIIIIIIIII

\\\\\\\\\\\\\\\\\\\\‘ \\\\\\\\\\\\\\

IIIIIIIIIIIIIIOIIII’I llllllll’llllr
\\\\\\\\\\\\\\

LLLLLLLLLLPDIPLLL‘L
Illllillllllll
~L§t§hh§§hh|~ * N

  

IIIIII’I'I’IOI

      

 

 

N», La»

 

 

 

       

‘\\‘\\\\h
III’IIII’
\\\‘\\\\k

Illllll’
‘\\\\\\\n

    

 

   

 

+N
+N

    

Italrllltlololtrlll
\\\\\\\\\\\\\\s\\\‘
Illlllloltlolllrll
\\\\\\\\\s\\\\§s\\x
IIIIIIIIIIIIIIIIIII

   

    

    

   

    

    

      

III’IIIIIIIIIIIII’IIOI”
\\\\\\\\\\\\\\\\\\\\\\\\

I
I
III I I If OIIIIIIIIII
I\\\I\\I\I\\I\I\\\I\\\\\\\\\\\\4 —

 

 

3
Site

         

III!!!
\\\\\\\

III!!!
\\\\\‘\
IIIIIII

 

III”'I’IO"IIII’I’III”

    

N», L»
-N

 

 

 

+N
Fig. 3

    

 
   
  
 

     

\\\\\\\\\\\\\\\\\\\\\\§\
IIIIIIIIII’IIII

I
IIIIIIAIIIIIIIII
\\\\\\\\\s\\\\\\ \
IIIIIIIOIIIIIIII

\\\\\\\‘

 

IIIIIIII
IIIIIIII
III’II”

 

NI, 1.. ns
-N

 

        

   

   

\\
\\\\\\\\\\
Iflll’lll’l
\\\\\\\\\\

 

 

+N

 

 

II
II

ns

IIllil’lltlllllllll’lllll N

 

 

 

 

1800 '

 

 

 

400 'I
00
0

1200 ‘
600 '
3
2
1

A E \65 56ch zomanEvN A E \wv mmmEoB 805m :32
N N :QMQAQEE.‘

33

biomass (sites 3-5), the removal of litter signiﬁcantly increased Andropogon
density.

The magnitude of the net and litter effects on Andropogon density differed
signiﬁcantly among the 5 sites, whereas the magnitude of the neighbor effect did
not (net, P4, 20 = 4.97, P < 0.01; neighbor, F4, 20 = 0.88, P = 0.49; litter, F4, 20 = 3.63,
P < 0.05; Fig. 4A). The mean net effect on density ranged from 57% at the site
supporting the lowest community biomass (site 1) to 97% at the site with the
greatest community biomass (site 5), and was positively correlated with mean
community biomass (Fig. 4A). The mean neighbor effect ranged from 49% at site
4 to 72% at site 2 and was not signiﬁcantly correlated with mean community
biomass. The mean litter effect on Andropogon density ranged from -30% at site 1
to 80% at site 5, and was positively correlated with mean community biomass.

There was considerable within-site variation in the magnitude of the
community effects, especially at the site with the lowest community biomass (site
1; Fig. 4B). Regressions performed on data that included this variation (all blocks,
n = 25) were consistent with those performed on site means; the community
effects on seedling recruitment were correlated with community biomass only
when litter was present (net and litter effects, Fig 43).

Total shoot biomass - The total shoot biomass of Andropogon in the seed
addition experiment differed signiﬁcantly among the sites, and in response to
neighbors and litter, and with interactions among all of these factors (Table 6).
The removal of neighbors increased Andropogon shoot biomass at all sites, both in
plots where litter was removed and in plots where litter was left intact (Fig. 33).
At sites 1 and 2, the removal of litter did not signiﬁcantly affect Andropogon shoot
biomass, either in plots where neighbors were removed or in plots where

neighbors were left intact. At site 3, litter removal signiﬁcantly increased

Net effect

Neighbor effect

Litter effect

 

 

 

 

 

 

 

0.4 I I I f ' I
22 24 26 22 30 32
L0“ 13=oorp=os4
094
034
- o
07 .
o
06-
054 o
0.4 I ' l I ' I ' I
22 24 26 22 30 32
02-
06-
o
04- o
02-
001- ------
03- 13:094P<001
.004 ' I I ' I ' I ' I
22 24 26 22 30 32

Log (total community biomass)

34

  

 

 

 

 

 

 

 

 

02- 2
R=oszp<00m
0.0 ' I ' I ' I ' l ' I ' I
20 22 24 26 23 30 32
10-
1'. +
02-‘3 A
”D“ M? -++
0.6“ +
1| +
04- "'
A
02- 2 I
n .R=ODLP=056
0.0 ' I ' [#17' I ' 17' I
20 22 24 26 23 30 32
10—
05-
00-
05-
40- n
- J 2
L5 .R=03&P<001
“2.0 w I ' I ' T ' l f I ' I
20 22 24 26 23 30 32

Log (total community biomass)

Figure 4. Relationships between the community effects on Andropogon density
(dependent variables) and community biomass (independent variable) in the
seed addition experiment. A) Mean community effects plotted against mean
community biomass (n = 5). B) Community effects calculated for all blocks
plotted against block community biomass (n = 25). Dashed line represents the no
effect line. Note log scale for community biomass.

35

Andropogon shoot biomass only in plots where neighbors were left intact. In
contrast, litter removal signiﬁcantly increased Andropogon shoot biomass at sites
4 and 5, both in plots where neighbors were removed and left intact.

The magnitude of all community effects on Andropogon shoot biomass
(net, neighbor, and litter effects) differed signiﬁcantly among the sites (net, 134,20
= 12.35, P < 0.0001; neighbor, 134,20 = 11.83, P < 0.0001; litter, F4, 20 = 4.49, P < 0.01;
Fig 5A). The mean net effect ranged from a low of 69% at site 1 to a high of 99%
at site 5, and was positively correlated with community biomass (Fig 5A). The
mean neighbor effect ranged from 65% at site 1 to 93% at site 5, and exhibited a
tendency to increase with community biomass, but the relationship was not
significant. The mean litter effect on shoot biomass ranged from -41% at site 1 to
71% at site 5 and was positively correlated with community biomass.

Regressions performed on data that included within-site variation were
largely consistent with those performed on site means, except that this analysis
revealed a signiﬁcant positive correlation between the neighbor effect and
community biomass (Fig. SB). This conﬁrms the non-signiﬁcant positive trend
observed among site means. The observation that the neighbor effect appeared to
increase with community biomass in the case of Andropogon shoot biomass, but
was unrelated to community biomass in the case of Andropogon density, suggests
that there may have been an increase in the effects of neighbors on the growth of
individual seedlings. However, because I did not measure the growth of
individual seedlings, and because Andropogon shoot biomass is confounded by
both recruitment and growth, I focus on the transplant experiment to better

assess the effects of the surrounding community on individual plant growth.

Net effect

 

 

Neighbor effect

9
\O
l

9
on
I

.o
\I
1

R2: 0.51, P = 0.19

 

.9
ox

1.0 -
0.8 -
.. 06 -
5; 0.4 -
H—l
«2 0.2 4
H
g 0.0 -
33-0.2 4
-04 ~

 

2.2

R2: 0.93, P < 0.01

    

 

 

-0.6

2.2

2.4

2.6

T I

2.8

I

3.0

3.2

Log (total community biomass)

36

  

 

 

 

R2 = 0.79, P < 0.001

 

1.0 1
0.9 -
03 .
0.7 -
06 -
0.5 -
0.4 7

2.0 2.2

I T I ' I ' I ' I ' I

2.6

2.4 2.8 3.0 3.2

 

E R2: 041,13 < 0.001

 

 

0.3

1.0 ‘
0.5 T
0.0 "
-0.5 ‘
-1.0 "
-1.5 "

2.0 2.2 2.4 2.6

ﬂ I ' I I ‘ I ' I ' I

2.8 3.0 3.2

 

2
R = 0.45, P < 0.01

 

 

-2.0 ‘
2

.0 2.2 2.4 2.6 2.8

I I r I ' I ' I V I ' I

3.0 3.2

Log (total community biomass)

Figure 5. Relationships between the community effects on Andropogon total shoot
biomass (independent variables) and community biomass (dependent variable)
in the seed addition experiment. A) Mean community effects plotted against
mean community biomass (n = 5). B) Community effects calculated for all blocks
plotted against block community biomass (n = 25).

37

Transplants

In the seedling transplant experiment, the RGR of Andropogon transplants
differed among the sites, and varied in response to neighbors and litter, and with
interactions between sites and neighbors, and between neighbors and litter
(Table 7). The RGR of transplants responded positively to the removal of
neighbors at all sites, both in the absence and presence of litter (Fig. 6). At sites 2,
4, and 5 the removal of litter signiﬁcantly increased RGR in plots where
neighbors were removed, but had no effect on RGR at any of the sites where
neighbors were left intact, reﬂecting a signiﬁcant interaction between neighbors
and litter.

The magnitude of the net effect on the RGR of transplants differed
signiﬁcantly among the 5 sites, whereas the magnitude of the neighbor and litter
effects did not (net, 134,25 = 4.28, P < 0.01; neighbor, 174,25 = 1.88, P = 0.15; litter, F4,
25 = 1.52, P = 0.22, Fig. 7A). The mean net and neighbor effects were similar in all
ﬁelds and ranged from a low of 36% (net effect) at site 5 to a high of 51% (net
effect) at site 3. Neither of these effects were signiﬁcantly correlated with
community biomass (Fig. 7A). The mean litter effect on transplant RGR was
small relative to the net and neighbor effects, ranging from 2% at site 1 to 16% at
site 5. However, the mean litter effect was positively correlated with community
biomass.

There was considerable variation in the community effects on RGR within
the sites (Fig. 7B). Regressions performed on data that included this variation
were consistent with regressions performed on site means; the magnitude of the
net and neighbor effects were unrelated to community biomass, but the

magnitude of the litter effect was positively correlated with community biomass.

38

Table 7. Results of ANOVA for site and treatment effects on the RGR of
transplants in the seedling transplant experiment.

 

Source
of variation df MS F

Among Blocks
Site 4 0.00019 1203*“
Block 25 0.00002

Within Blocks
Neighbors 1 0.01699 1689.18***
Neighbors x Site 4 0.00005 5.25"
Neighbors x Block 25 0.00001
Litter 1 0.00028 19.38“”
Litter x Site 4 0.00002 1.68
Litter x Block 25 0.00001
Neighbors x Litter 1 0.00026 19.25””
Neighbors x Litter x Site 4 0.00002 1.74"5

Neighbors x Litter x Block 25 0.00001

 

"5 P > 0.05; * P < 0.05; " P < 0.01; *** P < 0.001.

39

 

Illﬁtlll’llllllilllllllIllllllll
\\\\\\\\‘\\\\\\\\\\\\\\\\\\\\\\\

 

I’llﬂlllI”I'III”IIIIIOII'IIIII
\\\\\\\\\\\\\\\\‘\\\\\\\\\\\\\\\
s III!IIIIII’I’IIIIIII’IIIIIIIIII,

tbbllblbllbhblhhtblhbllblhlh

 

n _

 

 

IIIIIIOIIIIIIIIIIllllllllllfilllll III!!!
§\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\I\I\\\\§\\

 

 

 

IIIIIIIIIIIIIIIII’IIIIIIIIIOIIIIIIIIOIIII
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\§\\\\\\\\
IllllllllllllllllllllI’Ill\IIOIlIIIIlIIIIII

 

  

 

  

IIIOIIIIIII’IIIIIIIIIII
\\§\\\\\\\\\\\§\\\\\\\\\

 

       
   

 

 

Illtlllllllilllllllllll
\§\\\\§\\\\\\\\\\\\‘\\\\
Illllllllllll’l'llllllll

  

 

IIIIIIIIIIIIIIIIIIIIIIIIII'IIIIIII’II’III
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

 

IIIII’ll’IIIIIIIIIIII’I’II’IIIIIIIIIIIIII
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\§\\\\\\\\\
IIIIIIIIIIIIIIllillll’llllI’lllllllllllll

 

    

       

   

 

      

      

    

  

\ \

   

\ \\ \

 

      

 

 

 
  

\
I’ll’lll’lllllll’l'l’l
‘\\\\\\\\\\\\\\\\§\\\\
II’IIIIIIIIIIIIII’IIII
§\‘\\\\\\\\\\\\\\\\\\\
S IIIIIIIIIIIIIIIIIIIIII

 

         

      

      

     

    

 

    

    

IIIIIIIIII’IIIIIII

      

 

 

     

IIIIIIIIIIIII’IIIIIIIIIIII
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\§\\\\\\\\\\

IIIIIIIII’IIII’OI’I’Illllllll’ll’llll'll’ll
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

IIIIIIIIIIII'IIIIIIIIIIII’IIIII’I’IIIIIIIIII

     
   
   

   

   

 

 

\\\\\\\\\\
Il’lIIIIIIIII’IIIIIII’II’

  

 

        

(III/IIIIII

\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\\\
’IIIIIIIIII’

\\\\\\\\\\\\\\\\\\\\\\\\

   
  
  
  
   
 
 
 

   

   
   
 

 
   
   
 
 

 

I'll/I’IIIIIIIIIIIIIIIII

  

 

 

IIII’IIIIIII’IIII
\‘\\\\\\\\\\\\\\‘\\\\\\\\\\\\\\\\\\\\\\\\

IIIIIII’II’llllllllllll’lfll(IIIIIIIII’I
\\\\\\\\\\\\\\\\\\\\\‘\\\\\\\\\\\\\\\‘\\\

I’ll!)IIIIIIIfIIIIIIIIII’IIIIII’I’IIIIII

 

 

 

\\\\§\\\\§\\\\\‘\\\\\\\\\
IIIIIIIIIIIIIIIIIIIIIIIIII
\\\\\\‘\\\\\\\\\\\\\\\\\\

 

’IIIIIIIalrpilvroo;lotalr
S \s\\\\\\\\\\\\\\\\\\\\\\\

 

 

\5\\\\\‘\\\\\\\\\‘\\\\\\\\\\\\\\\\\\\\\\\\
III'IVIIIIIIIIIIIIIIIIIIIII’IIIIIIIIII’I’II
\\\\\\\‘\\\\\\\\\\\\‘\\\\\\\\\\\\\\\\\\\\\

 

IIIII’III’III’IIIIIIIIIII’IIIIIIIIIIIIIII
\\\\\\‘\\\\\\\\‘\\\\\\\\\\\\\\\\\\\\\\\\\\
till/IIIIIIIIIIIIIII’IIIIIII’IIII’IIIIIIII

-Litter
+Litter
ns

 

 

 

 

0.04 ‘
0.02 ‘

q
6
0.
0

0.08 '

Bamamcmb mo MUM

0.

+N

~N

Site

'1) of Andropogon transplants (mean + 1 SE) in relation to

g-l

RGR (g- -d
tal treatments

rimen

6

Figure
the expe

0.6 ‘

0.5 ‘

0.4 "'

Net effect

0.3 "

 

R2: 0.04, P = 0.77

 

0.2

0.6 "

0.5 "

Neighbor effect

2.2

I

2.4

—r

I

2.6

2.8

3.0

3.2

 

 

Litter effect
O
’1‘

 

2.4

 

   

R2: 0.93, P < 0.01

 

0.0
2.2

I

2.4

V

2.6

2.8

3.0

3.2

40

-0.1 " U
-0.2

Log (total community biomass)

 

 

 

0.6 - B Q I
f
0.5 ‘ ~
:1 (h.
_ U ++
04 U n u 0 I
+
R2: 0.02, P = 0.50
0.2 ' I ' ﬁ I I I l I l u |
2.0 2.2 2.4 2.6 2.8 3.0 3.2
0.6 “
O
u i‘
0.5 "' d] A I
‘g 5‘ +
0.4 ' +
u u 0' "“'
C
0.3 “
R2: 0.01, P = 0.55
0.2 I ' I ' I ' I ' Iﬁ I

 

2.0 2.2 2.4 2.6 2.8 3.0 3.2

 

R2: 0.15, P < 0.05

 

 

I ' I ' I ' I ' I ' I

2.0 2.2 2.4 2.6 2.8 3.0 3.2

Log (total community biomass)

Figure 7. Relationships between the community effects on the RGR of transplants
(dependent variables) and community biomass (independent variable) in the
transplant experiment. A) Mean community effects plotted against mean
community biomass (n = 5). B) Community effects calculated for all blocks
plotted against block community biomass (n = 30).

41

DISCUSSION

In this study, living plant neighbors negatively affected all aspects of
Andropogon performance at all ﬁve old-ﬁeld sites. This suggests that resource
exploitation by living plants reduced Andropogon seedling recruitment and
growth at all of the sites and across the entire range of community biomass and
soil nitrogen availability examined. Plant litter inhibited recruitment at the sites
with the greatest neighbor and litter biomass, but was of little consequence to the
growth of established plants at any of the sites in the presence of both living
neighbors and litter. Across sites, the net effect of the surrounding community on
recruitment was positively correlated with community biomass, largely due to
increasing interference by litter. However, the net effect of the community on the
growth of established plants was unrelated to community biomass. These results
suggest that both living plants and litter can limit the recruitment success of
Andropogon in old-ﬁelds, and that the negative impact of litter on recruitment is
likely to be most pronounced at productive sites that support a continuous layer
of dense litter.

Although numerous removal experiments have shown that the
surrounding plant community can affect target plant performance, this study is
the ﬁrst to measure the separate and combined effects of living neighbors and
litter on target plants across a range of environments. Although the effects of
living neighbors were generally greater in magnitude than the effects of litter,
litter signiﬁcantly reduced the recruitment of Andropogon at sites with the
greatest community biomass. This observation is consistent with the ﬁndings of a
number of studies investigating the effects of plant litter on individual species
(Werner 1975, Goldberg and Werner 1983, Fowler 1986, Hamrick and Lee 1987,
Carson and Peterson 1990). Fowler (1986) and Hamrick and Lee (1987) found

42

that, although small amounts of litter tended to enhance seedling establishment
of selected herbaceous species, large amounts of litter strongly inhibited seedling
establishment. Werner (1975) found that germination of the herbaceous biennial,
Dipsacus sylvestris, was strongly inhibited by grass litter in old-ﬁelds, and that the
degree of this inhibition increased with litter cover. Goldberg and Werner (1983)
showed that the removal of litter in the most productive areas of a 30 year old-
field greatly enhanced the seedling emergence of two Solidago species.

There are a number of mechanisms by which a dense litter layer can
inhibit the recruitment of plants from seed. Litter can act as a physical barrier,
preventing the movement of seeds to the soil surface or by impeding the
emergence of newly germinated seedlings (Sydes and Grime 1981, Facelli and
Pickett 1991 a). A large amount of litter can greatly reduce light penetration to the
soil surface, limiting germination and seedling growth (Facelli and Pickett
1991b). Other indirect effects associated with the dark moist conditions under the
litter mat, such as increased susceptibility of seedlings to fungal attack and
invertebrate herbivores, have been documented (Facelli, 1994).

In this study, the mechanisms by which litter inhibited seedling
recruitment are not known. The fates of individual seedlings were not followed
throughout the growing season, so I cannot distinguish whether the negative
effects of the surrounding plant community on Andropogon recruitment were due
to reduced germination, increased mortality, or a combination of these factors.
However, ﬁeld observations made over the growing season suggested that litter
inhibited both germination and seedling survival at the site with the greatest
litter biomass (site 5) due to mechanical impedance. Many of the seeds added to
plots containing litter at site 5, remained lodged in the litter layer throughout
much of the growing season. A large number of these lodged seeds germinated

in the litter, but subsequently died within a few days, likely due to the inability

43

of roots to reach the soil. Extreme reduction of light by litter may have been an
important factor limiting the emergence and growth of Andropogon seedlings in
the sites with the greatest litter biomass. Light levels at the soil surface in the
intact plots of sites 3-5 were extremely low, generally less than 5% of full sun
(Fig. 1). Light levels in plots where only litter was present were similar,
indicating that litter reduced light levels at the soil surface to a greater degree
than the living plants.

For established plants in the transplant experiment, the removal of plant
litter from plots where neighbors were left intact did not lead to any signiﬁcant
changes in growth. This suggests that the reduced growth of established plants in
these communities was due almost entirely to the exploitative effects of living
neighbors rather than inhibition by litter. However, in three of the ﬁve sites (sites
2, 4 and 5) litter signiﬁcantly reduced transplant growth in plots where living
neighbors had been removed, accounting for the signiﬁcant neighbor by litter
interaction. This indicates that the observed effects of living neighbors and litter
were non-additive, suggesting that litter has the potential to reduce the growth of
established plants, but that these effects are inconsequential in the intact
communities due to the much stronger exploitative effects of living neighbors.
The causes of reduced transplant growth in this study in plots where only litter

was present is unknown.

Magnitude of plant community effects

The relative community effect indices used in this study allowed me to
compare the magnitude of community effects across the different study sites. The
results show that the net community effect on both the recruitment of Andropogon

and the growth of established Andropogon transplants differed signiﬁcantly

44

among the sites, but that effects on recruitment were more spatially variable.
Although a number of site factors may have contributed to strong site differences
in the net effect on recruitment, the strong positive dependence of the net
community effect index on community biomass, as well on soil nitrogen
concentration (results not presented here), suggests that this variation reﬂects the
impact of plant biomass, soil fertility or productivity on the magnitude of plant
interactions as predicted by Grime (1979) and others (Southwood 1988, Keddy
1990). Interestingly, the effect of the surrounding community on recruitment was
not correlated with community biomass when litter had been removed (neighbor
effect). This, and the observed positive correlation between the effect of litter
alone (litter effect) and community biomass, suggests that the increase in the net
effect on Andmpogon recruitment was more the result of an increase in the
interference effects of plant litter rather than an increase in the effects of living
plants.

Unlike what I found for Andropogon recruitment, both the net and
neighbor effects on established transplant growth were unrelated to community
biomass. This result is inconsistent with theories of community organization
which predict an increase in competition with productivity (Grime 1979,
Southwood 1988, Keddy 1990). As was the case with Andropogon recruitment, the
effect of litter alone on transplant growth, although small relative to the effect of
living neighbors, increased signiﬁcantly with community biomass. Again, this
appears to reflect the potential effects of litter that were not realized in the intact
communities due to the much greater effects of living neighbors.

My interpretations of the community effect indices assume that I properly
standardized interspeciﬁc interactions among target plants and neighbors so as
to make them comparable among the different sites. In this experiment,

intraspeciﬁc interactions among target plants were possible because each

45

experimental plot contained more than one individual target plant. Miller (1996)
pointed out that because intraspeciﬁc and interspeciﬁc competition may both
vary along gradients, indices of competition that standardize interspeciﬁc
competition by plant performance in plots where intraspeciﬁc interactions are
occurring may result in an insensitive measure of competition. In this study
intraspeciﬁc competition was likely important in the neighbor removal plots of
the seed addition experiment where Andropogon densities were quite high. If
intraspeciﬁc competition did vary among the sites in these plots, it likely
increased with community biomass and soil nitrogen availability because total
shoot biomass increased signiﬁcantly with these factors. This would probably
have led to underestimates of recruitment in the removal plots at the most
productive sites due to a greater likely-hood of self-thinning. If this was the case
then I may have underestimated the magnitude of community effects in the seed
addition experiment at the sites with the greatest community biomass and soil
nitrogen availability. Despite the possibility of an insensitive measure, I found
that the net and litter effects on recruitment increased strongly with community
biomass, suggesting that if anything, I underestimated the magnitude of litter
interference effects on recruitment at the sites with the greatest community
biomass.

The result that plant litter contributed strongly to the over-all inhibitory
effect of the community on the recruitment of Andropogon at sites with the
greatest plant biomass is consistent with a number of studies showing that the
accumulation of large amounts of litter can limit plant diversity in productive
systems by interfering with species colonization (e. g. Carson and Peterson 1990,
Tilman 1993). I suggest that theories designed to explain variation in plant

interactions and species diversity along gradients in community biomass and

46

associated gradients in productivity and soil fertility should take into account the
potential impact of litter on recruitment, as well as other
non-exploitative interactions that potentially occur among plants.

In addition, although this study was conducted at only ﬁve sites and over
just one growing season, the results suggest that it may be during the
recruitment phase of the life history, when seeds and seedlings are especially
susceptible to the effects of litter, that Andropogon is most sensitive to variation in
community biomass. If this is correct, an important implication is that the
inclusion of regeneration processes in plant competition studies may often lead
to different conclusions regarding how plant interactions vary along
environmental gradients compared to those based upon the responses of
established plants alone. This is of particular signiﬁcance because regeneration
processes have often been found to be more important in determining the
distribution of species in space and time than the performance of established
plants (Grubb 1977, Gross and Werner 1982). These results also suggest that the
distribution of native C4 prairie grasses such as Andropogon along old-ﬁeld
biomass gradients in southwest Michigan may be more strongly constrained by
limits to recruitment than by competition experienced during the established

phase of the life history.

CHAPTER 3
REMOVAL EXPERIMENT II: PARTITIONING THE EFFECTS OF PLANT

BIOMASS AND LITTER ON THREE NATIVE GRASSES
IN OLD-FIELD VEGETATION

47

48

INTRODUCTION

An important goal of plant ecologists is to develop models of
population and community regulation that suitably predict the dynamics and
outcomes of species interactions at any location. As a result, over the past
decade there has been a great deal of interest in describing and explaining
variation in the magnitude of plant competition, especially with respect to
variation in community biomass or standing crop associated with gradients
in soil resource availability and plant productivity (Wilson and Keddy 1986,
Wilson and Tilman 1991, 1993, Bonser and Reader 1995, Kadmon 1995,
Twolan-Strutt and Keddy 1996). There is continuing disagreement among
plant ecologists regarding the expected relationship between the magnitude of
plant competition and community biomass. Biomass-dependent theories of
plant competition (Grime 1979, Keddy 1990) predict that the magnitude of
competition will increase with community biomass. This prediction is based
on the assumption that neighboring plants are most likely to come into
contact with each other and compete at sites supporting high plant biomass
and that plant demand for both above- and below-ground resources increases
with plant productivity and community biomass. Others have argued that the
magnitude of competition should be just as strong in low productivity
habitats as in high productivity habitats, and that competition tends to change
more in quality (from below- to above-ground) than in quantity as
productivity or community biomass increases (Newman 1973, Tilman 1988,
Taylor et al. 1990). This has led to an alternative biomass-independent view of
competition which predicts that the magnitude of competition should be

unrelated to productivity or community biomass.

49

Experimental studies designed to test these contrasting views have
yielded mixed results with some studies ﬁnding an increase in competition
with community biomass as predicted by the biomass-dependent theories
(Wilson and Keddy 1986, Bonser and Reader 1995, Kadmon 1995), and others
ﬁnding no relationship at all (Fowler 1990, Wilson and Shay 1990, Wilson
and Tilman 1991, 1993). To determine the general usefulness of community
biomass as a predictor of competition across different habitats and
communities, it is clear that more studies need to be conducted that
encompass a greater number of focal species and greater range of community
biomass and productivity.

To practically measure the magnitude of competition in the ﬁeld,
many studies have used neighbor removal experiments where the
performance of target plants growing in ﬁeld plots with the surrounding
plant community removed is compared to performance in plots with the
community left intact. In almost all of these experiments where target plant
performance was found to be reduced in the intact plots, the effect was
assumed to be due to resource exploitation by living plant neighbors.
Although this assumption may hold for some types of vegetation, in many
grassland and old-ﬁeld communities, additional effects associated with
accumulated plant litter may also affect the establishment and growth of
individual plants (Goldberg and Werner 1983, Fowler 1986, Hamrick and Lee
1987, Facelli 1994). As a result, indices of competition obtained from neighbor
removal experiments in grasslands and old-field communities may often
reﬂect the combined effects of living plants and litter on target plants.

In this study I investigated variation in the effects of both living plant
neighbors and litter on three native grass species in southwest Michigan old-

ﬁelds that differ in plant community biomass. I use the term community

50

biomass to refer to the sum of above-ground living plant biomass and litter
and use this measure as an index of plant productivity. The three target
species: Andropogon gerardi (Big bluestem), Schizachyrium scoparium (Little
bluestem), and Sorghastrum nutans (Indian grass), are all C4 perennial
bunch-grasses. These species represent the dominant grasses of the tall-grass
prairies (Bazzaz and Parrish 1982) and all were important components of the
pre-settlement prairie and savanna communities of southwest Michigan
(Gotshall 1972). Despite similarity among these species in photosynthetic
pathway, growth form, relative growth rate, and ability to compete for soil
nutrients (Parrish and Bazzaz 1982, Tilman and Wedin 1991), Schizachyrium
differs somewhat from the other two species in distribution across
environmental gradients. In intact prairies, Andropogon and Sorghastrum
are distributed similarly across broad gradients of plant productivity, soil
moisture, and soil fertility (Weaver 1954, Smeins and Olson 1970), while
Schizachyrium is typically more restricted in distribution to the more xeric,
infertile sites (Weaver 1954, Bazzaz and Parrish 1982). The prevalence of
Schizachyrium in unproductive, xeric sites in prairies has been attributed to
its tolerance of moisture stress, exempliﬁed by the high water use efﬁciency
(Weaver 1954, Knapp 1985). Physiological trade-offs associated with
adaptation to low moisture conditions may limit the competitive ability of
Schizachyrium in the more mesic and productive sites. Schizachyrium also
has smaller seeds than Andropogon and Sorghastrum (Rabinowitz 1978,
Foster, unpublished data) which may limit its ability to colonize productive
sites that support large amounts of ground cover.

In the current successional landscape of southwest Michigan, all three
of these native species are largely restricted to low fertility sites and are

generally absent from more productive old-ﬁelds on fertile sites that are

51

typically dominated by introduced C3 grasses. Competition with early-
colonizing, fast-growing introduced grasses combined with limited seed
dispersal may be important in inhibiting the re-colonization of fertile sites by
native grasses after site abandonment from agriculture (Tilman and Wedin
1991, Foster and Gross, in press, Tilman 1993). Also, because ﬁre is no longer
an important factor in this landscape, litter accumulation may inhibit the re-
colonization of productive sites by native prairie species that evolved in the
presence of periodic ﬁre and which are not likely adapted for establishment in
dense litter (Tilman 1993).

In this study I conducted a neighbor removal experiment in which
seeds and transplants of the three native grasses were added to a factorial
arrangement of removal treatments within each of nine old-ﬁeld sites. The
primary goals of this experiment were: 1) to determine the degree to which
site to site variation the effects of the surrounding plant community on target
plant recruitment and growth could be explained by plant community
biomass; and 2) to partition the effects of living plant neighbors and plant
litter in order determine how each contribute to the net effect of the

surrounding plant community on target plant performance.

METHODS

Study sites

This study was conducted in nine old-ﬁeld sites at the W. K. Kellogg
Biological Station (KBS) of Michigan State University in southwest Michigan
(Kalamazoo County; 42° 24' N, 85° 24' W). Sites were selected to encompass

the entire range of community biomass common to old-ﬁelds in southwest

52

Michigan. All nine sites had been abandoned from row crop agriculture for at
least 16 years and are currently dominated by introduced perennial grasses
(Table 8; see Burbank et al. 1992). Isolated individuals of Andropogon were
found at sites 1 and 5. Schizachyrium and Sorghastrum were not found at any
of the study sites although populations of these species were located within
several kilometers of each site. Soils underlying the sites are classiﬁed as
sandy loam, but differ in the amount of top-soil lost to agricultural-related
erosion. Assignment of site numbers (1-9) was based on ranking of mean

plant community biomass (1 = lowest, 9 = highest biomass).

Experimental design and methods

In 1995 both seeds and seedling transplants of three native C4 perennial
grasses were planted into a 2 x 2 factorial arrangement of treatments
established at each site in a randomized block design. The experimental
treatments included two levels of living plant neighbors (intact and
removed) and two levels of plant litter (intact and removed). At all sites, six 3
x 3.5 m experimental blocks were established in early May 1995 (Fig. 8).

Within all sites (except sites 1 and 3), adjacent blocks were separated by buffers
zones of approximately 2 m. At sites 1 and 3, blocks were separated by greater
distances ranging from 5 - 15 m to avoid shading from several isolated shrubs.
Each block contained four parallel 0.5 x 3 m experimental treatment plots
separated by 0.5 m buffer areas. Each plot within each block contained 6
subplots, three for seed additions and three for transplants (one seed addition
and one transplant subplot per target species). The three seed addition

subplots (30 x 30 cm) were adjacent to each other at one end of the block and

53

Table 8. Vegetation characteristics of the nine study sites. Biomass values are
shown as means 1 1 SE (n = 6 for each site). Biomass values within a column
with the same letter are not signiﬁcantly different (P > 0.05; Tukey’s HSD).

 

Litter Total
Neighbor biomass community Four most abundant Relative
Site biomass (g/Inz) (g/mz) biomass (g/Inz) species biomass (%)
1 68.8 :: 15.2a 10.4 :9: 1.5a 79.1 :: 16.1a Andropogon virginicus 50.0
Rubus alleghaniensis 13.9
Danthonia spicata 9.9
Solidago nemoralis 5.7
2 109.6 2 43.721 19.6 :t 5.9a 129.1 1: 43.2ab Rubus alleghaniensis 27.8
Danthonia spicata 17.2
Centaurea maculosa 14.7
Andropogon virginicus 11.0
3 182.3 :9; 15.1b 39.1 :t 2.3b 221.4 : 14.8bc Andropogon virginicus 49.8
Danthonia spicata 9.8
Rubus alleghaniensis 6.0
Desmodium rotundifolium 4.8
4 259.6 : 27.5bc 91. 9 : 23.7b 351.6 : 50.3cd Agropyron repens 85.2
Achillea millefolium 7.3
Bromus inermis 6.4
Melilotus alba 0.1
5 348.1 :t 17.9cd 235.2 t 50.7c 583.3 :2: 54.9de Bromus inermis 58.1
Poa pratensis 10.3
Achillea millefolium 10.1
Agropyron repens 5.1
6 398.8 1 56.2cd 220.4 :1: 57.5cd 626.9 :1: 115.7de Bromus inermis 78.0
Agropyron repens 10.3
Poa pratensis 3.8
Daucus carota 3.2
7 745.2 a: 105.3de 397.6 :9; 42.4cde 1142.8 : 142.8ef Agropyron repens 71.0
Arrhenatherum elatius 28.3
Silene alba 0.1
Polygonum scandens 0.1
8 915.2 2: 55.1e 555.7 3: 54.3de 1470.9 3: 102.4f Bromus inermis 44.2
Poa pratensis 19.1
Solidago canadensis 17.4
Agropyron repens 15.4
9 1235.9 :1: 174.4e 737.3 a: 68.5e 1973.3 1 223.8f Bromus inermis 90.6
Poa pratensis 4.9
Agropyron repens 2.5
Barbarea vulgaris 0.1

 

54

Experimental Block Layout

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ag Sn Ss Sn Ag 55

T1 T1 T1 ‘
50 cm Seeds Seeds Seeds

T2 T2 T2

| l
_ 0 cm
50 cm

50 cm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 8. Experimental block layout illustrating the arrangement of the

3.0 m

3.5 m

experimental treatment plots and the transplant and seed addition subplots.

T1 and T2 refer to the position of the two transplants in each transplant

subplot.

55

were separated by buffer zones of 20 cm. The three transplant subplots (50 x 50
cm) were also adjacent to each other, but had no buffer zones between them.
Species were randomly assigned to both the seed addition and transplant
subplots within each experimental plot.

In early May, the experimental treatments were applied to the four 0.5 x
3.5 m plots located within each block. Neighbors were removed either by
clipping or mowing and by applying Glyphosate herbicide solution
(Roundup) to the re-growth a week later. Litter was removed by hand from
the litter removal treatments so that the soil was left undisturbed. To
minimize root encroachment into the two neighbor removal treatments, plot
perimeters were trenched with a ﬂat blade shovel to a depth of 20 cm twice
during the growing season. The borders between subplots were also trenched
to prevent below-ground interaction between the different target species.
Neighbor removal plots were periodically hand-weeded during the growing
season to eliminate vegetation re-growth.

Seed additions - Seeds of the three target species were collected from
local populations in the fall of 1994 and stored in an unheated garage over
winter in a metal container to expose the seeds to winter temperature
ﬂuctuations. Seeds were sown by hand into all of the 30 x 30 cm subplots at a
rate of 300/ subplot (3030/ m2) in late May 1995. The seedlings were censused
and target shoot biomass was harvested in late September 1995. Seedling
density at the time of harvest is used in this study as a measure of target plant
recruitment. All plant material was oven-dried at 60° C to constant mass and
then weighed.

Seedling transplants — Seedlings of each of the three target species were
obtained by germinating seeds in ﬂats in a heated greenhouse in mid-May

1995. Approximately 7-9 days after germination, 700 - 900 seedlings of each

56

species were transferred into individual seedling cone-containers (2.54 cm
diameter x 10.16 cm deep) containing a 3:1:1 mixture of peat moss,

commercial potting soil, and old-ﬁeld soil. After 22-25 days additional growth,
432 seedlings of each species were selected for the experiment. Initial shoot
biomass (B) of these seedlings was estimated from regressions of seedling
biomass on longest leaf length (L) obtained from subsamples of the remaining
healthy seedlings (Andropogon: B = 0.005L - 0.017, R2 = 0.71, P < 0.001, n = 30;
Schizachyrium; B = 0.008L - 0.034, R2 = 0.76, P < 0.001, n = 30; Sorghastrum: B =
0.002L + 0.016, R2 = 0.44, P < 0.001, n = 30 ).

In early June, two seedlings of each species were transplanted into their
respective transplant subplots by inserting seedling plugs into pre-formed ‘
holes made with a 2.54 cm diameter soil corer. Within each subplot, the two
transplants were separated by approximately 20 cm. All individual transplants
were watered (0.2 L / day) for twelve consecutive days following transplanting
to encourage initial establishment. At sites 5, 8 and 9 most of the transplants
in the neighbor removal treatments were heavily browsed by deer within two
days after planting. The browsed transplants were replaced with new ones and
a 1.5 m tall fence made from poultry netting was placed around the perimeter
of the neighbor removal treatment plots at all sites to prevent further deer
browsing. Fencing was not placed around the treatment plots where
neighbors were left intact because the target plants growing in these plots
were not browsed. The presence of fences had no signiﬁcant affect on light
levels at the soil surface in any of the sites. By the end of the experiment, only
16 Andropogon seedlings and 27 Schizachyrium transplants had died.
However, 66 Sorghastrum transplants died, most of which were located at
sites 1 and 2. In late September the shoots of all surviving transplants were

harvested, dried and weighed.

57

Community biomass - Above-ground biomass of neighboring plants
and litter was estimated for all sites in the seed addition subplots where both
neighbors and litter had been left intact. For six of these sites, plant neighbors
were sorted by species to document species composition (Table 8). Species
composition for the other three sites (4, 5 and 7) had been assessed in a
previous study (Foster and Gross, in press) and data from that study are given
in table 8. Plant material from all sites was dried and weighed as described

previously.

Light measurements

Photosynthetically active radiation (PAR) was measured in mid-June
above the vegetation and at the soil surface in the transplant section of each
plot with a PAR ceptometer probe (Decagon Devices Inc.) to determine how
light penetration to the soil surface varied among the sites and was affected by
the experimental treatments. In plots where litter was left intact, light
measurements were made by inserting the ceptometer beneath the litter layer.
In the neighbor removal plots, measurements were made at plot edges at
mid-day when the sun was directly overhead to minimize the impact of
transplants on light readings. Light penetration is expressed as the percentage

PAR reaching the ground surface.

Data analysis

Community biomass - Site differences in neighbor biomass, litter

biomass and community biomass (sum of neighbor and litter biomass) were

examined using one-way analysis of variance (ANOVA). This analysis was

58

performed on biomass data from the intact treatment (neighbor and litter
biomass left intact).

Light penetration and target plant performance - Percent light
penetration, seedling density and total shoot biomass (seed addition
experiment), and transplant growth (transplant experiment) were analyzed
using two-way analysis of variance separately for each site. In cases where the
interaction between neighbors and litter was signiﬁcant, contrasts among
neighbor levels were compared at each level of litter and contrasts among
litter levels were compared at each level of neighbors using unplanned
comparisons (Sokal and Rohlf 1981). Due to a lack of germination in a large
proportion of the plots at all sites in the seed addition experiment, data on
Sorghastrum density and total shoot biomass data are not reported here. In
addition, Sorghastrum transplant RGR data from sites 1 and 2 were
eliminated from all statistical analyses due to high plant mortality at these
sites. Analyses of transplant growth were performed on plot mean values.
Relative growth rate of each transplant (RGR, Chiariello et al. 1989) was

calculated as:

RGR = [1n(132) -1n(31)l/(t2 - t1)

where B2 is the transplant biomass at harvest, B1 is the estimated transplant
biomass at planting, and t2 - t1 is the number of days growth between harvest
(t2) and planting (t1). Light penetration data were arcsine-square root
transformed to reduce heteroscedasticity. Andropogon total shoot biomass
and density data were log transformed and square-root transformed
respectively to improve normality and reduce heteroscedasticity (Sokal and

Rohlf 1981). Untransformed data are presented in the ﬁgures.

59

Magnitude of community effects - The magnitude of the effects of the
surrounding plant community on target plant performance (density, total
shoot biomass, and RGR of transplants) was examined using a relative
community effect index (diffuse competition, Wilson and Keddy 1986;
competition intensity, Wilson and Tilman 1993 ). This index estimates the
degree to which target plants were inhibited or facilitated by the surrounding
community by comparing plant performance in plots cleared of all plant
material (neighbors and litter) to that in plots where all plant material was
left intact. Positive values indicate inhibition, negative values indicate
facilitation, and values of zero indicate no net effect of the surrounding
community. To assess the importance of litter in contributing to the total net
effect of the community on target performance, a separate index was
calculated for plots that were left intact and those where only litter was

removed. These indices were calculated for each experimental block as:

Net Effect = (PC - PNL)/PC

Neighbor Effect = (PC - PN)/ PC

where PC is plant performance in plots cleared of all plant material (living
neighbors and litter), PNL is performance in plots with neighbors and litter left
intact, and PM is performance in plots with only neighbors left intact (litter
removed). An additional index was calculated to measure the effect of litter

alone and was calculated as:

Litter Effect = (PC - PLit)/ PC

60

where PL“ is plant performance in plots with only litter left intact (neighbors
removed).

Two-way ANOVA and Tukey HSD tests were used to examine site and
species differences in each community effect index (total, neighbor and litter
effects). Effects on the RGR of Sorghastrum at sites 1 and 2 could not be
examined because of missing data. As a result I performed two separate two-
way AN OVA's on transplant RGR: 1) an analysis that included data from all
nine sites and two species (Sorghastrum omitted); 2) an analysis with data
from sites 1 and 2 omitted and all three species included.

I examined the degree of inter-dependence between the community
effects and community biomass using linear regression on site means (n = 9).
For Sorghastrum RGR, regressions were performed on data for sites 3-9 only.
In cases where a community effect index showed evidence of non-linear
dependence on community biomass, I log-transformed community biomass
prior to the analyses. These data are shown in the ﬁgures un-transformed
with logarithmic best-ﬁt curves presented to illustrate non-linearity. SYSTAT

statistical software (version 5.2.1; Wilkinson 1992) was used for all analyses.

RESULTS

Community biomass and species composition

Total community biomass and its two components, neighbor and litter
biomass differed signiﬁcantly among the nine sites (Table 8). Total
community biomass increased 25-fold From site 1 to site 9. Neighbor and
litter biomass increased 18- and 71-fold, respectively, from site 1 to site 9 and

were positively correlated to each other (R2 = 0.87, P < 0.0001, n = 9). The nine

61

study sites were primarily dominated by perennial grasses, though the

differed somewhat in species composition (Burbank et al. 1992; Table 8).

Light penetration

The extent to which living plant neighbors and litter affected light
reaching the soil surface varied substantially among the nine study sites (Fig.
9. At the three sites with the lowest community biomass (sites 1-3), the
removal of neighbors signiﬁcantly increased light penetration, but removal of
litter had no effect. In contrast, at sites 4-9 light penetration varied
significantly with neighbors, with litter, and with the interaction between the
two. At these sites, neighbor removal increased light penetration in the plots
where litter was removed, but had no effect where litter was left intact. At
sites 4-7, litter removal increased light penetration in both the presence and
absence of neighbors, while at sites 8 and 9, litter removal increased light
penetration only in plots where neighbors were removed. Mean light
penetration measured in the intact plots declined logrithmically with mean
community biomass (light penetration = -0.231(log community biomass) +

162, R2 = 0.86, P < 0.0001, n = 9).

Seed addition

Significance of treatment effects - In the seed addition experiment, the
number of Andropogon seedlings recruited into the plots was increased
signiﬁcantly by the removal of living plant neighbors at all nine study sites
(Fig. 10A). At the two sites with the greatest community biomass (sites 8 and

9) where there were signiﬁcant interactions between neighbors and litter, the

62

 

 

100 Nut Lns N. Lns D -Litter
I * E3 +Litter ,,
[—1 n g. n 1'
80 ,T. -F .. n I H
g {I l I
g 60
‘13
to
3:1
2
cu 40
a" I"
,E'b It
:3 20 n ..
n
0 al..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

     

a].
6;.

no so

“:3 “.3

-N +N -N +N -N +N -N +N -N +N -N +N -N +N -N +N -N +N
1 2 3 4 5 6 7 8 9

 

Site

Figure 9. Mean seasonal light penetration (mean + 1 SE) in relation to the
experimental treatments. Treatments with neighbors removed are indicated
by - N, and treatments with neighbors left intact are indicated by +N.
Treatments with litter removed are indicated by - L, and treatments with litter
left intact are indicated by shading and by +L. Letters above the bars specify the
signiﬁcance of neighbor (N) and litter (L) effects within the sites (ns = P > 0.05;
* = P < 0.05). At sites with a signiﬁcant neighbor x litter interaction, brackets
above and below the bars specify treatment mean comparisons made at each
level of the two treatment factors (ns = P > 0.05 ; * = P < 0.05).

63

Figure 10. Andropogon density (A) and total shoot biomass (B; mean + 1
SE) in relation to the experimental treatments (seed addition).

64

N“ L"

N" L‘

E] -Litter
ES +Litter

        

   

I I I I I I
\ \I‘I‘ \ \ \I\I\I\I\I\I\I\I\ \l\
IIIIIIIII IIIIII

 

 

m.”

\\\\\\\\\\\\\\\\
IIIIIIIIOIIIIII
\\\\\\\\~k\\\\\h

IIIIIIIIIIIII’IIII

 

 

 

 

  

\\\\\\\\\\\\“\\\\\\\\\\
IIIIIIIIIIIIIIIIIllllsllI

I
§\\\\\\\\\\\\\\\\\\\\ \‘

A

\\
\\
I

 

     

\\\\\\\\
IIIIIII’
\\\\\\\\L

LIB

   
    

     

 

I I I I I I I I I I I I I I I I III I I I I I I I I
\ \ \ \ \ \ \ ‘ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I

    

 

N1:

      

 

N" L“5

 

 

V\\\\\\\\\\\\\
IIIIIIIIIIIIIA

KKKLNNPLLKFLIK

 

 

 

    

NNV\\NNV\\ V“\VN‘\\§\NN
IIIIIIIIIIIIIIIIIIIIIIIII
\\\\‘\\\\\\\\\\\\\\\\\\\

N" I.“5

 

        

\\\\\\\\\

IIIIIII IIIIIIA
\\\\\\\\\\\\\\
IIIIIIIIIIIIIII

  

 

1200 _

1000..

$6 \ .05 bmmcou gavonoﬁﬁx

 

 

-N+

-N +N -N +N -N +N -N +N -N +N -N +N -N +N

-N +N

 

 

 

  

 

 

   

IIIIIIIIIIIIIIIIIIII

 

  

IIIIIIIIIIIIIIIIII.
\\\\\\\§\\\\\\\\\‘
IIIIIIIIIIIIIIIIII

 

 

            
 
 

\\§\\\\\\\\\\
IIIIIIIII’III
EKKKKKK\KI *

  

In

«N +N

-N +N

 

 

 
 

  

V\V\'\7\V\V\'\V\f\P\V\V\V\7\V\V\7\V\V\V\- *—

       

I
\\\‘\\\\\‘\\\\\
IIIIIIIIIIIIIII

bbbbbbbbbbbbbb

\\\\\\\\\\\\\\\\§

 

 

IIIIIII’IIII’IIII
\\\\\\\\\\\\\\\\“—*—

 

 

 

 

 

B
1 . .
0 0 0

A
0
w m w w

u

m

2

A E \mv $6803 305m ~53 zomoueméax

N

-N +N -N +N -N +N -N +N -N +N -N +N

-N +N

 

 

 

5
Site

Figure 10.

65

removal of neighbors significantly increased Andropogon density, both in the
absence and presence of litter. At sites 1-4, the removal of litter had no effect
on Andropogon density. At sites 5-7, litter removal significantly increased
Andropogon density. At sites 8 and 9 where litter biomass was greatest, litter
removal increased Andropogon density significantly only in plots where
neighbors were removed, accounting for the neighbor by litter interaction.

The removal of neighbors increased Andropogon shoot biomass
signiﬁcantly at all but the site with the lowest community biomass (site 1; Fig.
10B). At sites 5-9 where there were signiﬁcant interactions between neighbors
and litter, the removal of neighbors significantly increased Andropogon
shoot biomass both in the absence and presence of litter. At sites 1-4, litter
removal had no effect on Andropogon shoot biomass. At sites 5-7, litter
removal significantly increased Andropogon shoot biomass, both in the
absence and presence of neighbors. At sites 8 and 9, litter removal increased
Andropogon shoot biomass significantly only in plots where neighbors were
removed, accounting for the neighbor by litter interaction.

The treatment responses of Schizachyrium in the seed addition
experiment were similar to the responses of Andropogon (Fig. 11). The
number of Schizachyrium seedlings that recruited into the plots was
increased significantly by the removal of living plant neighbors at all sites
except site 1 (Fig. 11A). At sites 8 and 9, where there were significant
interactions between neighbors and litter, the removal of neighbors
significantly reduced Schizachyrium density, both in the absence and presence
of litter. At sites 1-4, the removal of litter had no effect on Schizachyrium
density. At sites 5-7, litter removal significantly increased Schizachyrium

recruitment. At sites 8 and 9, litter removal increased Schizachyrium density

66

Figure 11. Schizachyrium density (A) and total shoot biomass (B; mean
+ 1 SE) in relation to the experimental treatments (seed addition).

67

N" L”

N" L"

—I_.1tter
+Litter

‘
I
\ \

 

 

 

 

\ \ \ \ \ \ \ \
I I I I I I I
\ \ S \ \ \ \ \

 

 

N" L”

N” Lmi

 

 

\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
I I I "I‘I‘I I\I\I\I I\I I I I I\I I I I I I I\

IiIlIyl I I I I I I I I bbbbbbbbbbbb

 

 

All I I I I IlI
\‘III\\\\II\\\\\\\
II II IIIIIII

I II I I II I I I I I b I

 

 

 

l

\\\\\\\\\\ .\\\\ \\\\\
IIIIIIIIIII:IIIIIIIIIIIIIII
\\\\\\\\\\\\\\\\\\\\\ \\

I

  
   

\\\\\\\
IIIIIII:II:I
\\\\\\\\\ \

  

. 3.,

 

 

IIII
\\\

\
IIIII

IIIIIIIIIIIIIIIIIIIIIIIIIII
\\\\\\IIIII\\\\\\\\‘\\\\\\\\\\I\\
IIIII I IIIIIIIIIIIII I

  
 

 

N” L"5

 

 

N"3 L“5

 

1400 - A

1200 .

A

N

E\

m.

800 -

600 .

 

 

m 0 0
0
2

.05 based EESﬁSEom

-N +N -N +N -N +N -N +N -N +N -N +N

-N +N

 

 

 

 

  

\ \ \ \ \ \ \ \ \ \ \ \ \

 

 

-N +N

r’rrrtrrlltll
IIIIIFIIIIIIIIIII *—

   

  

IIIIIIIIIIIIII
\\\\\\\\s\\\\\
VVVVVEIIVVV

  

-N +N

 

 

I
\ \ \ \ \ \ \ \ \ \ \ \ \

 

 

 

-N +N

I I I I I I I I IIIIIIIDIII.**—

  
  

\ \ \ \ \ § \ \ ‘
I I\I I\I\I\I\I I I I l I I‘
\\\\\\\\\\\\\\\

   
 

-N +N

 

  
  

IIIIIIIIIIIII
\\\\\\\\\\\\‘
I I I I I I I I I I I I I

l
-N +N

 

 

A J J
\\\l\\
IIII

N" L"5
m
-N +N

 

N‘V“\
IIIIII
\\\\\\‘

I
3
I.
I
\\
I

N" L'“
ﬂ
-N +N

NML’“

-N +N

 

-N +N

 

5

W
m
NW

—
m 0

1

$8 \wv 39:03 89? :39 ESSEEEom.

 

 

 

Site

Figure 11.

68

significantly only in plots where neighbors were removed, accounting for the
neighbor by litter interaction.

The removal of neighbors increased Schizachyrium shoot biomass
significantly at all sites except sites 1 and 2 (Fig. 11B). At sites 6-9 where there
were significant interactions between neighbors and litter, the removal of
neighbors significantly increased Schizachyrium shoot biomass in both the
absence and presence of neighbors. At sites 1-4, litter removal had no effect on
Schizachyrium shoot biomass. At site 5, litter removal significantly increased
Schizachyrium shoot biomass. At site 6, litter removal significantly increased
Schizachyrium shoot biomass only in plots where neighbors were present. At
site 7, litter removal increased Schizachyrium shoot biomass in both the
absence and presence of neighbors, while at sites 8 and 9, litter removal
significantly increased Schizachyrium shoot biomass only in plots where
neighbors were removed.

Magnitude of community effects - in the seed addition experiment, the
net, neighbor and litter effects on both seedling density and shoot biomass
differed significantly among the sites, but there were no differences among
species (Andropogon and Schizachyrium) and no interaction between sites
and species (Table 9, Fig 12 and 13). For both species, the net, neighbor, and
litter effects on density and shoot biomass were all positively correlated with
total community biomass (Fig. 12 and 13). The net and litter effects on the
density of both species increased logrithmically with community biomass,
while relationships between neighbor effects on density and community
biomass were best described as linear functions. For shoot biomass of both
species, all three measures of community effect (net, neighbor, and litter

effects) increased logrithmically with community biomass.

69

Table 9. Results of two-way ANOVA for community effects on density and
total shoot biomass (seed addition).

Total Shoot

 

Density biomass

Effect/Source df (F value) (F value)
Net

Site 8 92*“ 102*“

Species 1 1.5nS 0.1nS

Site x Species 8 0.4118 0.1ns
Neighbor

Site 8 5.6“” 1.1’W

Species 1 0.1nS 0.1“8

Site x Species 8 0.4115 0.1“8
Litter

Site 8 4.3"” 42*“

Species 1 0.1nS 0.1ns

Site x Species 8 0.3118 0.3nS

 

ns P > 0.05; * P < 0.05; ** P < 0.01; *’” P < 0.001.

70

Figure 12. Relationships between the community effects on Andropogon
density (A) and total shoot biomass (B; dependent variables) and community
biomass (independent variable). Dashed line represents the no effect line.

71

Doom coma 83 gm

 
     

coca coma oooH com

Iind H M
N

.2 «SE

I Nd

I and

1 ed

 

 

coca

 

 

 

 

 

 

      

H
a.
I
a
H
a
D

fwd

ooom coma 83 com

— p - p n — 0.0
HonnwmmZIGI N
“oz-bl 1N6 m.
B
u
p.
uvd N
9.
am.
red 0..
O
1
a

t; . H

.2 «a $5 m
3
1S1

no;

 

 

72

Figure 13. Relationships between community effects on Schizachyrium
density (A) and total shoot biomass (B; dependent variables) and
community biomass (independent variable).

73

R E \wv 89:05 325888 :38.
SON coma 83 com

— n p b n p n n

.2 magma

c

lllllllllll .ll Cd

      

***mw.o “- NH

SON coma 83 com

— h P n n h — -

$.de u M
N
O

I Nd

1 «mo

1 ed

 

     

uNd

uwd

nod

 

ANS \wv 89:05 525588 EBB

OOON coma 83 com o
h

- P - b h l. I

 

 

88362 Jul
82 IOI

 

ISHQO H M
N
Izbwd H M

0A-

 

 
       

r Nd

u vd

r ed

1 wd

 

L:

page 131m

spaya .quqﬁta N pue is N

74

Transplants

Significance of treatment effects - The removal of neighbors
significantly increased the RGR of Andropogon and Schizachyrium
transplants all nine study sites (Fig. 14A and B). The removal of neighbors
increased the RGR of Sorghastrum at sites where its responses could be tested
statistically (sites 3-9; Fig. 14C). Litter removal increased the RGR of
Andropogon and Sorghastrum transplants only at site 7 in plots where
neighbors were removed, and had no effect on the RGR of Schizachyrium at
any of the sites.

Magnitude of community effects - The net effect of the plant
community on the RGR of transplants varied significantly with sites, but not
with species (Table 10). However, there was a significant site by species
interaction reﬂecting a greater net effect on Schizachyrium than on the other
species at site 6, but a weaker net effect on Schizachyrium than on the other
species at sites 7 and 8 (Figs. 15, 16, and 17). The neighbor effect on transplant
RGR varied significantly among sites, but litter effects did not vary
signiﬁcantly among sites or species (Table 10). For both Andropogon and
Schizachyrium, the net and neighbor effects on RGR increased logrithmically
with total community biomass (Fig. 15 and 16). Litter effects on Andropogon
and Schizachyrium were not correlated with community biomass. None of
the three community effects on Sorghastrum transplants were correlated with

community biomass (Fig. 17).

l1.

75

Figure 14. RGR (g-gl-d'l) of transplants (mean + 1 SE) in relation to the
experimental treatments. NT = not tested statistically.

76

A

0.08 q

N" L”

N" I.”

 

 

\I\I\I\I\I\I\I\I\I\I\I\I\I\
IIIIIIIIIIIIIA

IIIIIIIIIIIIIIIIIIIIIIIIIIIII
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
I I I I I I I I I I I I I I I I I I I I I I I I I I I I I A

 

    

  

\\\\\\\\\\\\\\\\
I I I I I I I I I I I I I I I I A

 

  

\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \

    

\ \ \ \ § \ \ \ \ § \ \ ‘ \ S \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I‘ I‘ I ‘I I I I I I I I I I I I I A
\ \ \ \ \ \ \ \ \ \ § \ \ \ \ \ \ \ \ \ \ \ \
7 7 V I 7 V ' 7.717i" 7W? 7 I f V V 7 I.’ V V P.’ III

 

Andropogon

 

 

      

I I I I I I I I I I I I A
\ \ ‘ \ \ \ \ \ \ 5 \ \
I I I I I I I I I I I I A

  

   

IIIIIIIIIIIIIIIIIIIIIIIIIIIA
\\\\\\\\\\\\\\\\\\\\\\\\\\\\
IIIIIIIIIIIIIIIIIIIIIIIIIIIA

 

\ ‘ \ \ \ \ \ \ \ § \ \ \
I I I I I I I I I I I I I A
\\\\\\\\\\\\\

\ \ \ \ \ \ \ \ \ ‘ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I I I I I I I I I I I I I I A
\ \ \ \ § \ \ \ \ ‘ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \

 

      

I I I I I I I I I I I A
\ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I A

  

 

I I I I a I I I I I I o I i I I I t I I I;
.1 s \ \ \ \ I s s \ \ \ \ I \ s s \ s \ s \
IIIIIIIIIIIIIIIIIIIII.‘

FIIIIIIIIIIIIIIIIIIII

i tu1iu‘r111w‘~“‘V<‘«t

 

 

 

         

     

\ \ I \ s I I s \ x \ \ s \ \
I I I I I I I I I I I I I I A
\ \ s \ \ \ x s \ \ \ \ \ \ \

 

411111IIIIIA:A41141JIJIII
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I :I I I I I I‘ I I\ I“ I I I I A
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ § \ \ \ \ \

 

 

 

 

\ \ \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I I I A
\ \ \ \ \ \ \ \ \ \ \ \ \ \
III/IIIIIIIIII

 

y

w \ s s \ ‘ N \ s \ \ N \ \ \ \ \ \ ~ \
I I I I I I I I I I I I I I I I » I I I .
\ \ x \ sl\ \ I\ Is \ s \ I \ \ \ \ \ \ \

 

 

       

N‘ L“5

 

I I I I I I I I I I I A
\\\\\ \\\\\\
IIIIIIIIIIIA

     

 

I I I I I I I I I I I I I\
\\\\\\\\\\I\\I\I\
I I I I I I I I I I
5 N \ \ K K N K K KlLlKlK IN

.
\A

N" L“5

 

 

 

 

0.06 a
0.02

0.
0.

 

-N +N ~N +N -N +N -N +N -N +N -N +N -N +N -N +N

~N +N

 

 

 

Schizachyrium

 

L115

\\\‘\\\§\\\\\\\\\\\\\\\\
IIIIIIIIIIIIIIIIIIIIIIIII
\\\\\\\\\\\\\\\\\\\\\\\\

 

I I I I I I I I I
\\\\\\\\
III/IIIII

IIIIIIII IIIIIIIIIIIIIII
\ \ 5 \ ‘ \ \ \ \I\ \ \ \ \ \ \ \ \ \ \ \ \ \ \
IIIIIIIIIIIIIIIIIIIIIIII

     

IIII
\\\\\\\\\\\
I I I I I I I I I I I I
\\\\

     

 

I I I I I I I I I I I I I I

    

    
  

 

\\ \\\\\\\
\I‘I I I‘l‘I‘I I I I I I I I

IIIIIIIIIIIIIIIIIIIIIIII
\\\\\\\\\\\\\\\\\\\\\\\
IIIIIIIIIIIIIIIIIIIIIIII

   
 

    

\ \ \ \ \ \ \ 5 ‘ \ \ \
I I I I I I I I I I I I
\ \ \ \ \ \ \ \ \ \ \ \

     

\ \ ‘ \ \ \ \ \ \ \ \ \ \ \ \ \ \ ‘ \ \ \ \ \ \ \
I I I I I I I I I I I I I I I I I I I I I I I I I
\ \ Q \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \

 

\ \ \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I I I
\ \ S ‘ \ \ \ \ \ \ \ \ \ \

\ \ \ \ \ \ \ \ \ ‘ \ \ \ \ \ \ \ \ \ \ \ \
IIIIIIIIIIIIIIIIIIIIIIII
\\\\\\\\\\\\\\\\\\\\\\\

 

 

 

I I I I I I I I I I I I I I
K \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I I I

 

I r I r I I I I I I 1 I r I I r I I I I [11"
\\\\\‘\‘\\\\\\‘\\\\\‘\
IIIIIIIIIIIIIIII’IIIIII

N’ L"5

 

    

    

\\\\\\\\\\‘\\\‘
I I I I I I I l I I I I I I I
\\\\\\\\\\\\\\\

     

\\\\\\\\\‘\\\\\\\
IIIIIIIIIIIIIIIII
\\\\\\\\\\\\\\\\\

N” I)“

 

 

       
      
   

I I I I I I I I I I I I I I
\ \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I I

 

I I I I‘I I I I I I I I I I I I I I
\\\\\\\\\\\\\\\\\
IIIIIIIIIIIIIIIIII

N" L”

 

.-
I‘
I‘
-l
I.-

8 2
0. 0 0 0. .
0 0 0

mucmamcmﬁ. mo MU

-N +N -N +N -N +N -N +N ~N +N -N +N -N +N -N +N

-N +N

 

 

 

 

 

C

0.08 ..

 

Sorghastrum

 

\ \ \ \ \ \ \ \ \ \ \ \ \ ‘
I I I I I I I I I I I I I I
\\\\\\\\\\\\\\

 

IIIIIIIIIIIIIIIIIEIII:IIEIIIIIIIII
\ \ \ \ \ \ \ \ \ \ \ ‘ § \ \ \ \ \ \ \ \ \ \

- \\\\\\\\\\\\\\\\\\\s\\\\\\\\\\\
'VVFVVP'7'???V'VVIVIV\VV\V\VV\VV'V'V'

 

 

 

         

 
 

I I I I I I
\ \ ‘ \ \ \ \ \ \ \ \ \ \ \ \ \

I I I I I I I I I I I I I I I I

  

\ \ \ \ \ \ \ \ \ I‘ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ ‘ \
IIIIIIIIIIIIIIIIIIIIIIIIIII
\\\\\\\\~\\\\\h§\\\\\\\\\\\

 

 

F
U“

      

\\\\\S\\\\\\\‘
I I I I I I I I I I I I I
\ \ \ \ \ \ \ \ \ \ \ \ \ \

   

I\I\I\I\I\I\I\I\I\I\I\I\I\I\I\I\I\I‘I‘I‘I\I\I\I\I\I\I\I‘I\
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

        

  

 

   

I I I I I I I I I I I I
\ \ \ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I

 

\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ ‘ ‘ ‘ \ \ \ \ \ \ \ \ \ \
I I I I I I I I I I I I I I I I I I I I I I I I I I I

‘1 I a o I I I a I I a I r 1 I t I I I I I r r a I I I I

 

 

 

11111111111
\ 5 § § \ \ \ \ ‘ \ \ \
I I I I I I I I I I I I
\ \ \ \ \ \ \ \ \ \ \ \

 

\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \

' \\\\\\\\‘\\\\\\\\\\\‘
I I I I I I I I I I I I I I I I I I I I

       

              
           

I I I I I I I I I I I I I I I
\\\\\\\\\\\\\\\
I I I I I I I I I l I I I I I

IIIIIIIIIIIIIIIIIIIIIIIII
\‘\\\\\\\\\\\\\\\\\\\\\\\\
IIIIIIIIIIIIIIIIIIIIIIIII

N" I.“5

 

I I I I I I I I I I I
\ ‘ \ \ \ \ \ \ \ \ \ \
IIIIIIIIIII

 

I I I I I I I I I I I I I I I I I I I
\ ‘ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ I‘ I‘
I I I I I I I I I I I I I I I I I I
K N K K \ § \ K \ \ NiKlK K N \ N \lhI

T

 

 

 

           

I I I I I I
\ \ \ \ \ \ \
I I I I I I

   

I I I I I I I I I I
\\\\\\\\\\
IIIIIIIIII

 

 

0

0.06 -l
0.04 -1.
002 J

o o

-N +N -N +N -N +N -N +N -N +N -N +N -N +N -N +N

-N +N

 

 

 

 

 

 

Site

Figure 14.

77

Table 10. Results of two-way ANOVA for community effects on the RGR of
transplants.

Sites 1-9, Sites 3-9,
Effect/Source Sorghastrum Sorghastrum
omitted (df, F value) included (df, P value)

 

Net
Site 8 20.0“” 6 65*“
Species 1 0.1“8 2 0.4“8
Site x Species 8 ' 2.9“ 12 3.1“”
Neighbor
Site 8 13.4’“m 6 41*“
Species 1 4.0ns 2 3.0nS
Site x Species 8 1_4ns 12 1.7“5
Litter
Site 8 0.1[‘5 6 1.4ns
Species 1 0.9“5 2 0.1nS
Site x Species 8 0.6nS 12 0.6118

 

ns P > 0.05; * P < 0.05; ’” P < 0.01; *"* P < 0.001.

Net and Neighbor effects

 

78

Andropogon

R2 = 0.76"

 

 

 

 

 

 

'0' Net
0‘1 T -0- Neighbor .
0.0 l I ' I ' I
0 500 1000 1500 2000
0.2 ~ B
o
4.. 0.1 "
u
:8
cu
H
.93. o o
:3 0.0 "‘ .
2
R = 0.06ns
-0 1 “—1 l l ' l ' I
0 500 1000 1500 2000

Total community biomass (g / m2 )

Figure 15. Relationships between the community effects on Andropogon
transplant RGR (g-g‘l'dl) and community biomass (independent variable). ns

P > 0.05, ** P < 0.001.

0.7 1

0.6 ‘

0.5 ‘

0.4 ‘

0.3 ‘

0.2 ‘

0.1 "

Net and Neighbor effects

79

Sch izachyrium

2
R = 0.74"

  

2
R = 0.80""
8

  
 

 

'0' Net
-0- Neighbor

 

 

 

0.0

0.2"

0.1‘

0.0'l

Litter effect

-0.1‘

 

I ' I ' I ' I

0 500 1000 1500 2000

2
. R = 0.01ns

 

I ‘ I ' I ' I

0 500 1000 1500 2000

Total community biomass (g/ m2)

Figure 16. Relationships between the community effects on Schizachyrium
transplant RGR (g-g'l-d'l) and community biomass (independent variable). ns
P > 0.05, *" P < 0.001, *** P < 0.0001.

 

80

 

 

 

 

 

Sorghastrum
0.60 -
3
u
m
”~55 0.55 - o
s.‘ 0. O O
s 0 ° °
.1:
.29 0.50 -
m
2 o
‘2
m 0.45 - 0 2
a ONet R = 0.13ns
Z 0 Neighbor R 2: o_o3ns
0.40 I I I ' I I 1
0 500 1000 1500 2000
0.20 ..
0.15 _
E 2
a: ' R = 0.01ns
m0w. .
H
.93
5
0.05 _
o
o
o
0.00 I I I ' l ' I
0 500 1000 1500 2000

Total community biomass (g/ m2)

Figure 17. Relationships between the community effects on Sorghastrum
transplant RGR (g-g‘l-d'l) and community biomass (independent variable). ns

P > 0.05.

81

DISCUSSION

In this experiment I found that the surrounding plant community
suppressed target plant performance at all nine study sites. For two of the
three target species (Andropogon and Schizachyrium,) the net suppressive
effect of the surrounding community on both seedling recruitment and
transplant growth increased significantly with community biomass. These
relationships were strongly non-linear, indicating that the net effects of the
community increased most rapidly with community biomass at the low end
of the gradient and then leveled off. Similar non-linear dependence of
competition on plant biomass has been observed in wetland communities
(Shipley et al. 1991), old-fields (Bonser and Reader 1995), and desert grasslands
(Kadmon 1995). Taken as a whole, these results are most consistent with
biomass-dependent view of plant competition that predict that the magnitude
of competition occurring among plants increases with community biomass,
plant productivity or soil fertility (Crime 1979, Keddy 1990). However, the
strong non-linearity of these relationships suggests that the degree to which
competition depends on community biomass or productivity may depend
upon the range of community biomass or productivity examined.

Several authors have argued that the inconsistent results obtained
from studies examining variation in competition may be related to
differences in the range of community biomass or productivity over which
competition was measured (Goldberg and Barton 1992, Bonser and Reader
1995, Kadmon 1995). Goldberg and Barton (1992) pointed out that results from
experiments conducted across natural productivity gradients most often
support the prediction that competition increases with productivity, while

experiments conducted across experimental gradients are usually more

82

consistent with the predictions of the biomass-independent view. Goldberg
and Barton (1992) argue that these different results may be due to a more
restricted range of productivity in studies using experimental gradients. In a
previous field experiment that I conducted in 1994 over a more limited range
of community biomass in these old-fields (208 - 1205 g/ m2; Foster and Gross,
in press), I found that the net community effect on Andropogon recruitment
increased with community biomass as found in the current study. However,
unlike in the current study, the net effect of the plant community on the
growth of Andropogon transplants varied little, and was unrelated to
community biomass.

To examine whether the different results obtained for transplants in
the two experiments could be attributed to differences in the range of biomass
over which effects were measured, 1 re-analyzed data from the current study,
but excluded data from sites that fell below the minimum level of
community biomass used in the previous study (sites 1 and 2 omitted). The
results of this re-analysis were largely consistent with the results from the
previous experiment: the net effects on the recruitment of both Andropogon
and Schizachyrium increased with community biomass as observed for
Andropogon in the previous study (Andropogon: R2 = 0.78, P < 0.01; and
Schizachyrium: R2 = 0.73, P < 0.05). However, the net effects on transplant
growth for all three target species were not correlated with community
biomass over the more limited range of biomass, although the net effect on
Andropogon transplants did exhibit a tendency to increase across this range
that was marginally significant (Andropogon: R2 = 0.48, P = 0.07; and
Schizachyrium: R2 = 0.32, P = 0.20; Sorghastrum: R2 = 0.30, P = 0.20).

This result suggests that target plants recruiting from seed were

sensitive to changes in community biomass across the entire range of biomass

83

examined, although these effects were most pronounced from low to
medium biomass sites than from medium to high biomass sites. In contrast,
the growth of established transplants appeared to have been sensitive to
changes in community biomass only at the low end of the biomass gradient.
This indicates that the range of biomass and productivity examined can be of
critical importance in competition studies, greatly inﬂuencing conclusions
regarding how competition is affected by productivity. In fact, the conclusion
in this study that the net effect of the plant community on Sorghastrum
transplant growth was unrelated to community biomass may reﬂect the more
restricted range of community biomass over which data was available for this
species.

Although results for the entire range of biomass are most consistent
with the biomass-dependent view of competition (Grime 1979, Keddy 1990),
in the case of competition between established plants, it may be more useful
to view both biomass-dependent and biomass-independent views as being
applicable across different ranges of biomass or productivity. Because net
effects on transplants increased signiﬁcantly only across a very limited range
of low community biomass, the biomass-independent view may be the most

relevant for all but the most extremely unproductive old-field communities.

Partitioning the effects of living neighbors and litter

In this study, living plant neighbors reduced most aspects of target
plant performance at each of the sites, indicating that resource exploitation by
living plants suppressed recruitment and growth across a wide range of plant
community biomass. Plant litter inhibited the recruitment of Andropogon

and Schizachyrium from seed at sites of medium to high community

84

biomass, but had little impact on the growth of established transplants at any
of the sites. Interestingly, the removal of litter at sites 8 and 9 which
supported the greatest neighbor and litter biomass, led to increases in
recruitment only when neighbors had also been removed, while litter
removal at sites of intermediate neighbor and litter biomass increased
recruitment both in the presence and absence of neighbors. This suggests that
litter was less of a constraint to seedling recruitment in the intact vegetation
at sites with the greatest community biomass than it was at sites of
intermediate community biomass. The negative impact of litter at sites 8 and
9 within plots where living neighbors had been removed suggests that litter
had a strong potential for limiting seedling recruitment at these sites that was
not realized in the intact plots due to the much stronger impact of living
plant neighbors. These strong suppressive effects of living neighbors were
likely related to effects on light availability. Unlike at sites of intermediate
community biomass, living plant biomass was so great at sites 8 and 9 that
neighbors reduced light at the soil surface to very low levels and to the same
degree as did litter and neighbors combined.

These results indicate that litter contributed to the net effect of the
surrounding plant community on recruitment at sites of intermediate
community biomass, but contributed little to the net effect of the plant
community on transplant growth at any of the sites. For seedling recruitment,
the separate effects of living neighbors and litter increased with community
biomass, indicating that both effects contributed to variation in net effects on
recruitment among the study sites. The observed increases in litter effects
with increased community biomass illustrates that plant litter was capable of
inhibiting seedling recruitment by itself at sites of intermediate to high

community biomass. However, significant disparity between net and

85

neighbor effects occurred only at sites of intermediate productivity, thus
indicating that the potential inhibitory impact of litter on recruitment was
not detectable in the intact vegetation at sites with the greatest community
biomass due to the strong effects of living plant neighbors on light
availability. For transplants, litter had a relatively small effect on RGR at all
sites compared to the net and neighbor effects. This, and the fact that there
was very little disparity between the net and neighbor effects for any species,
suggests that litter did not contribute significantly to the net effects of the
community on transplant growth nor variability in the net effects observed
among sites.

The strong suppressive effect of plant litter on recruitment observed in
this study at sites of medium to high community biomass are consistent with
a number of studies showing negative impacts of litter on germination,
seedling emergence, growth and survival (Werner 1975, Goldberg and
Werner 1983, Hamrick and Lee 1987, Bergelson 1990, Facelli and Pickett
1991a). Such negative effects of litter are usually observed at relatively high
levels of litter accumulation where litter can reduce light levels at the soil
surface, alter microclimate, and act as a mechanical barrier to seeds and
seedlings (Facelli and Pickett 1991a). Smaller amounts of litter has been found
to enhance plant performance in some environments by ameliorating
environmental stresses (Fowler 1986, Wilms et al. 1986). Facilitory effects of
plant litter are most common in arid environments where plants are often
exposed to extreme moisture stress. In the study, I found no strong evidence
of facilitation by litter, probably because of the relatively benign climate and

high rainfall in this region.

86

Implications for native grass distributions and species diversity

In this study, the established plant community suppressed the
recruitment of Andropogon and Schizachyrium at most of the study sites.
However, seedlings were able to successfully establish some seedlings at all
sites in the presence of vegetation, suggesting that the absence of these species
from these study sites may reﬂect dispersal limitation, in addition to any
constraints on recruitment imposed by the established plant community.
Seedling recruitment of Andropogon and Schizachyrium in the presence of
vegetation tended to be highest, and the suppressive effects of the plant
community were lowest at the most unproductive sites. This suggests that
dispersal limitation may be the most important constraint to establishment in
unproductive sites, while dispersal limitation and interference effects of
vegetation may both be important in reducing recruitment at the more
productive sites. At the three most productive sites in this study (sites 7-9),
both seedling recruitment and total shoot biomass were very low in the
presence of vegetation, indicating that the effects of the established plant
community may be sufficient to prevent colonization by these species in
highly productive communities that support large amounts of living biomass
and litter even when there is no limit to dispersal. Andropogon and
Schizachyrium differed little from each other in their ability to establish
seedlings at any of the study sites. In contrast, Sorghastrum was largely unable
to establish seedlings at any site. The inability of Sorghastrum to establish
from seed was primarily due to poor germination brought about by low seed
viability and likely does not reﬂect a genetic difference in this species ability to

establish seedlings. However, the high mortality of Sorghastrum transplants

87

at the two most unproductive sites (sites 1 and 2) may reﬂect species
differences.

Several studies suggest that strong interference effects of established
vegetation on seedling recruitment can contribute to reduced species diversity
in highly productive old-field and grassland communities by reducing rates of
species colonization (Carson and Peterson 1990, Tilman 1993). Most previous
mechanisms proposed to explain reduced diversity with increased
productivity assume that intense resource competition among living plants
(primarily for light) reduces diversity in productive habitats by increasing the
rate of species loss (Goldberg and Miller 1990, Tilman and Pacala 1993, Huston
and DeAngelis 1994). However, the number of species in a habitat is
determined by the balance between extinction and colonization (MacArthur
and Wilson 1969). The results showing that seedling recruitment of two
native grass species was inhibited by the established plant community are
consistent with the view that constraints to colonization may contribute to
low species diversity in productive old-field communities (Tilman 1993). The
results also show that the mechanism responsible for reduced colonization in
productive habitats may include the suppressive effects of accumulated plant
litter on seedling recruitment in addition to resource competition by living

plants.

CHAPTER 4

SPECIES RICHNESS IN A SUCCESSIONAL GRASSLAND: EFFECTS OF
NITROGEN ENRICHMENT AND PLANT LITTER

88

89

INTRODUCTION

The addition of limiting nutrients to herbaceous plant communities
frequently leads to increases in plant production and declines in plant species
diversity (Bakelaar and Odum 1978, Tilman 1987, Carson and Barrett 1988,
Pysek and Leps 1991, Tilman 1993). There has been a great deal of interest in
the mechanisms responsible for reduced diversity at high levels of plant
productivity and in response to nutrient addition, not only due to theoretical
interest, but also because of practical concerns regarding the consequences that
long-term eutrophication has for the conservation and restoration of native
plant communities (Berendse and Elberse 1990, Marrs et a1. 1996). A number
of prominent theories have focused on resource competition and its potential
role in determining the effects of productivity and soil fertility on species
diversity in plant communities (Grime 1973, 1979, Newman 1973, Goldberg
and Miller 1990, Tilman and Pacala 1993, Huston and DeAngelis 1994). Most
of these theories suggest that intense competition for light among living
plants causes the rate of competitive exclusion to increase at high levels of
productivity, leading to a reduction in species richness (Abrams 1995).
Although resource exploitation by living plants is undoubtedly important in
limiting the diversity of many plant communities, these theories do not
consider the potential impact of non-living plant biomass or accumulated
plant litter in contributing to reduced diversity at high levels of productivity.
For example, a number of studies suggest that the accumulation of a dense
litter layer is an important factor limiting the diversity of highly productive
grasslands and old fields through its inhibitory impact on seedling
establishment and thus species colonization (Goldberg and Werner 1983,

Carson and Peterson 1990, Tilman 1993).

90

Plant litter plays an important role in many plant communities by
directly or indirectly affecting individuals and populations (Sydes and Grime
1981, Knapp and Seastedt 1986, Carson and Peterson 1990, Facelli and Pickett
1991a, Facelli 1994). There are studies illustrating positive (facilitation) and
negative (inhibition) effects of plant litter on the performance of plants. In
general, facilitation by litter is most common in extremely unproductive or
arid environments where even small amounts of litter can ameliorate
stressful environmental conditions (e. g. low moisture levels; Fowler 1986,
Willms et al. 1986, Hamrick and Lee 1987). Inhibition by litter is most
commonly observed in highly productive, but undisturbed environments
where litter accumulation can be quite high (Haslam 1971, Werner 1975,
Goldberg and Werner 1983, Carson and Peterson 1990, Facelli and Pickett
1991a, Tilman 1993, Foster and Gross, in press). Together, these studies suggest
that litter may be an important factor inﬂuencing species diversity along
productivity and fertility gradients via changes in the direction (facilitation
versus inhibition) and magnitude of litter effects.

In this study I investigated the effects of nitrogen enrichment and litter
accumulation on species richness in a southwest Michigan successional
grassland. Prior to European settlement, the herbaceous component of the
upland prairie and savanna communities of southwest Michigan were largely
occupied by native C4 grasses and a diverse array of prairie forbs, and it is
generally recognized that these species and communities were maintained by
frequent fire (Gotshall 1972). In tallgrass prairies, fire serves to remove plant
litter, thus altering species composition and diversity (Hulbert 1969, Knapp
and Seastedt 1986, Tilman 1993). Litter accumulation has been frequently cited
as an important factor contributing to reduced yield and reduced diversity

under conditions of fire suppression in tallgrass prairie (Hulbert 1969, Knapp

91

and Seastedt 1986). In the current successional landscape of southwest
Michigan, upland herbaceous old-fields are typically dominated by introduced
C3 grasses such as Bromus inermis and Agropyron repens which are capable
of invading and dominating fields shortly after abandonment from
agriculture (Tilman and Wedin 1991a, Foster 1992). Because fire is currently
absent from this landscape, litter accumulation may be an important factor
limiting species colonization success and species richness in the most
productive ﬁelds, especially those that have become eutrophied due to past
agricultural activity (fertilization, manure application etc.). Litter
accumulation may be particularly important in inhibiting the colonization of
productive old-fields by native grassland species that evolved in the presence
of fire and which generally lack adaptations for establishment in dense litter
(Tilman 1993).

The primary goals of this study were to examine the short-term effects
of nitrogen enrichment on community biomass, species abundance, and
species richness within a moderately productive successional grassland in
southwest Michigan, and to examine the extent to which these effects could
be attributed to changes in litter biomass independently of changes in
standing crop or living biomass. To critically examine the role of litter in
contributing to changes in community structure associated with nutrient
enrichment and increased productivity, it is necessary to manipulate litter
levels in conjunction with fertilization. In this study, I report the results of a
two year field experiment in which the experimental addition of nitrogen
fertilizer was accompanied by the experimental removal and addition of plant
litter.

Another objective of this study was to examine the impact of nitrogen

enrichment and litter accumulation on the re-colonization potential of a

92

single native species. This was done by adding seed of a native C4 perennial
grass, Andropogon gerardi, (Big bluestem) to the experimental plots.
AndrOpogon was absent from the site at the time of the study, both as
established plants and as seed, thus allowing assessment of seedling
recruitment from a known seed pool. Although this once-dominant prairie
species is common in the current landscape, it is primarily restricted to low
and medium fertility sites and is largely absent from the most fertile old-
fields. The seed additions allowed me to assess whether the absence of
Andropogon from the study site, and from the more productive old-fields in
the region, may be due more to dispersal limitation or to inhibition by living

vegetation and litter.

METHODS

Study site

This study was conducted in southwest Michigan at the W. K. Kellogg
Biological Station (KBS, Kalamazoo County; 42° 24' N, 85° 24' W). The
experimental site was located within a mid-successional old—field (20-30 years
post-abandonment) that was dominated by the introduced C3 grasses:

Bromus inermis,Agropyron repens, and Phleum pratense. These species
accounted for over 80% of the total community biomass. The most common
forb species included Achillea millefolium (native), Daucus carota
(introduced), Hypericum perforatum (introduced), and Potentilla recta
(introduced). Soils underlying the field site are a well-drained, Oshtemo
sandy loam formed from glacial till and are of moderate fertility for the

region (Foster and Gross, in press).

93

Experimental treatments

In mid April 1994, six 2.5 x 2.5 m experimental blocks were established
at the study site (Fig. 18). The blocks were arranged in two parallel columns of
three blocks each that were separated by a 2m buffer strip. Blocks within each
column were separated by buffer zones of 1m. Within each block, four 1 x 1 m
treatment plots were established and placed so that there was a 0.5m buffer
between each plot (one plot at each corner of the block). Four experimental
treatments were randomly assigned to the four treatment plots within each
block. The treatments included; a control (C), litter removal (LR), nitrogen
addition (NA), and nitrogen addition + litter removal (LRNA). I later added
a litter addition treatment (LA) to examine the effects of increased litter
biomass independent of nitrogen enrichment. The goal was to add a quantity
of litter similar to the additional amount expected to accumulate in the NA
plots. To accommodate this extra treatment, one additional 1 x 1 m plot for
the LA treatment was added to each block in June 1994. These plots were
systematically positioned 0.5 m away from one side of each of the original
blocks as shown in figure 1. As a result, there was not a complete
randomization of all treatments within the blocks.

Litter was removed from the LR plots in April and November 1994 by
clipping around the perimeter of the plots and then gently lifting the litter
mat by hand. This assured that there was no disturbance to the soil surface.
Litter was removed in November to remove the current years production
that had died and fallen into the plots. In March 1995, a small amount of litter
blown into these plots over the winter was removed by hand. Nitrogen was

applied to the NA and LRNA plots as commercial ammonium nitrate

94

2.5m

-_____
n _3
ICE

 

 

 

 

 

 

 

H

 

 

 

 

 

2.5m

 

 

'2"

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_JL_I II III
_l|_| II LJLJ
<-

 

 

 

 

 

 

 

 

llm
III _,
' |1m VI
—(E-m— 2m

Figure 18. Diagrammatic representation of the experimental block and
treatment plot layout. Block numbers are indicated by roman numerals. The
litter addition (LA) plots that were added after the initiation of the
experiment are indicated by heavy lines.

95

fertilizer at a rate of 48 g N - m'2 - yrl. The pelletized fertilizer was applied
twice annually (24 g N per application), in late April and mid June of 1994 and
1995.

During the first growing season (1994), the litter addition plots were left
untreated until late November when additional litter was added. Litter was
added in November because a significant increase in litter biomass, due to
nitrogen enrichment, did not occur until the end of the first growing season.
Litter collected nearby was weighed, and then spread evenly across the litter
addition plots. The amount of litter to be added was determined by collecting
litter from a 0.1 x 1m strip in the C and NA treatment plots. The litter was
weighed and the mean difference in litter biomass was calculated. After
weighing, this litter was returned to the C and NA plots. The litter added to
the LA treatment plots was held in place by loosely stretching two strands of

bailing wire across each plot and anchoring the ends of the wire in the soil.

Seed addition

To examine the effects of nitrogen enrichment and litter on the
recruitment of Andropogon gerardi, locally collected seeds of Andropogon
were added to each experimental plot in May 1995. Within each plot, seeds
were added to a marked 30 x 30 cm quadrat at a rate of 3030 seeds/ m2 (300
seeds per quadrat). Seeds were evenly broadcast by hand from approximately
30 cm above the soil. Germination and survivorship of Andropogon
seedlings was monitored by censusing the experimental plots on a bi-weekly
basis throughout the growing season. On each of six census dates, newly

germinated seedlings were marked with toothpicks of a color specific to a

96

given census. Individual plants marked in previous censuses were accounted

for as dead or alive.

Light measurements

To determine how the experimental treatments inﬂuenced the
quantity of light reaching the soil surface, photosynthetically active radiation
(PAR) was measured in each of the plots in early August 1995. PAR
measurements were taken approximately 50 cm above the vegetation and at
the soil surface in each plot using a PAR ceptometer probe (Decagon Devices
Inc.). In plots where litter was left intact, measurements were made by
inserting the ceptometer probe beneath the litter layer. Light penetration was
expressed as the percentage of the PAR above the vegetation that reached the

soil surface.

Vegetation sampling and data analysis

All above-ground living plant biomass and litter was harvested in late
August 1995 from three 30 x 30 cm quadrats located in each plot. In addition,
all above-ground biomass of Andropogon was harvested from the separately
marked, 30 x 30 cm quadrat located in each plot. All living biomass was sorted
by species, and the number of individual forbs were counted. Numbers of
individuals of the dominant clonal grass species were not counted due to my
inability to distinguish genetic individuals. All plant material was dried at 60°
C to constant mass and weighed.

Treatment effects on plant biomass (living biomass and litter), forb

density, species richness, and Andropogon density, biomass per plant,

97

germination and survivorship were examined using a randomized block
design ANOVA (recognizing that the treatments were not completely
randomized within blocks) and Tukey HSD tests. Because no significant block
effects were found in this study, only treatment P values are reported in the
results. Biomass data were log (x+1) transformed, density and richness data
were square-root transformed, and Andropogon survivorship data were
arcsine transformed before analysis to meet the assumptions of ANOVA.
Untransformed means are presented in the figures and tables. Statistical
analyses were performed using SYSTAT Version 5.2.1 (Wilkinson 1992). Plant

species nomenclature follows Fernald (1950).

RESULTS

Community responses

Plant Biomass - Nitrogen enrichment resulted in a significant increase
in community productivity as reﬂected by an increase in living biomass in
both the presence and absence of litter (Fig. 19A), and an increase in litter
biomass (Fig. 19B). However, litter removal had no effect on living biomass,
either in the unfertilized or fertilized treatments. As intended, litter removal
significantly reduced litter biomass in both the unfertilized and fertilized
treatments. Litter addition had the intended effect of increasing litter biomass
to levels similar to that found in the fertilized treatment where litter was left
intact (NA treatment). Litter addition had no effect on living biomass (Fig.
19A).

The experimental manipulations altered the amount of light reaching

the soil surface (Fig. 20). Nitrogen enrichment significantly reduced light

98

   

 

   

B
31000 A1000
<E :1
\ 800 a 800 F=207.1"“
39 Fe
g 600 T; 600
(I)
(U
E 400 g 4
-° .9
.13
go 200 H 2
'5 g
:3 0 :5
C LR LA NA LRNA
C
N 1000 A1000 D
(a F (7” ”Wins 9
= ./
to E 300 F—32
v g — . ns
§ 3 600
E E
E .9. 400
m .o
g .o 200
H 5
0 LL. 0
C IR LA NA LRNA c LR LA NA LRNA
F
$300 20
8 250 W a = - "t
@200 5
gm 5
u: .m
5 100 g
'2: a.
e 50 m
0
‘-'~ 0
C LR LA NA LRNA C LR LA NA LRNA
Treatments Treatments

Figure 19. Treatment effects on living biomass (A), litter biomass (B), grass
bloII‘lass (C), forb biomass (D), forb density (E), and species richness (F). Means
(+ 1 SE) with different letters are significantly different (P < 0.05). See methods
fEftiOn for treatment codes. ANOVA F430 values: ns P > 0.05, “* P < 0.001,

* P < 0.0001.

99

201

F = 43.2““

15‘

10..

Light penetration (%)

 

OJ

 

C LR LA NA LRNA

Figure 20. Treatment effects on light penetration. Means (+ 1 SE) with
different letters are significantly different (P < 0.05). ANOVA F430 value: “m
P < 0.0001.

100

levels in both the presence and absence of litter. Litter removal increased light
levels significantly in the unfertilized treatments, but had no effect on light in
the fertilized treatments. Litter addition significantly reduced light to levels
observed in the fertilized treatments.

The strong positive response of living biomass to nitrogen enrichment
primarily reﬂects the response of grasses (Fig. 19C). Grass biomass increased
significantly with nitrogen enrichment due to increases in the biomass of the
dominant grass species, Bromus inermis (Table 11). None of the other grass
species were significantly affected by nitrogen enrichment. Grass biomass was
unaffected by litter removal and litter addition. In contrast, the biomass of the
subordinate forb community, which was much lower than that of grasses
overall, did not differ significantly among the treatments (Fig. 19D). For life
history groups and individual species of forbs that could be examined
statistically, there were no signiﬁcant treatment effects on biomass, despite
considerable differences in the magnitude of the means among the
treatments (Table 12).

Forb density - Unlike forb biomass, forb density responded strongly to
the experimental treatments (Fig. 19E), reﬂecting the impact of nitrogen
enrichment and litter on forb seedling establishment. Nitrogen enrichment
reduced forb density signiﬁcantly in both the presence and absence of litter.
Litter removal led to an increase in mean forb density. However, this effect
Was statistically signiﬁcant only in the fertilized treatments where litter

removal increased forb density up to levels found in the controls. Litter
addition signiﬁcantly reduced forb density to the same level as found in the
fertilized plots that contained litter (NA), indicating that the reduction in forb

Clel'ISity caused by nitrogen enrichment in these plots was largely due to

Incr eased litter biomass.

101

Table 11. Mean biomass of grass species (g/ m2) in the ﬁve experimental
treatments. ANOVA's were performed for species that were present in all
treatments. Species are listed from greatest to least in biomass. Means within
rows with different letters are signiﬁcantly different (P < 0.05, Tukey's HSD).
See methods section for treatment codes.

Biomass

 

C LR LA NA LRNA F 4, 20

 

B mm as inerm is 200.3a 215.2a 272.4a 703.3b 571.1b 25.7“” "
A gropyron repens 64.4 43.6 32.6 69.2 95.4 2.2ns
Phleum pratense 40.3 28.7 22.9 53.4 43.9 0.6ns

Poa pratensis 16.7 18.3 9.2 13.2 33.5 2.8ns
Panicum 5p. 3.1 5.2 2.2 2.6 7.9 0.1ns
Poa compressa 2.3 1.2 0.7 0.2 -

 

ns P > 0.05, W” P < 0.0001

102

.256 v m z... .86 v m 3. .36 v m * .86 A m m:

 

 

 

 

- 9m 3 - mad 3: 53:. 553.856

- 2: - :5 523:8 3.332

- mm a: 3. - 85 EN 85 3533 3.3.32

- 9m .2: 3” - 85 85 m3 statue assé

- m.m.m - cod Emu“: Ezzobﬁ

- mm as - 85 mod 22:... £30

- 3w 9o 5.: 2: - one :3 25 Ed 5...: 53:38

- me go 0mm 34. - E. :3 a: 8a 55385 sarcasm

22 3m .8 2: 3a. SN 22 32 8.2 a}: 8.8 E: 53:82? 5:33...

.3. as: a.” £3 £3 £35 23 mg 8a m2: Ea 8mm 228 25:5

2.6.2 £53 £5 £8 «3.: saw 23 £82 £85 58.8 £89. mama. 3823285

:3. 98$ 35 £2 «gm 8.5.8 22 can 8a 92: 8.2 8.8 382 Essa

- 3.: ma. 5.». - a: 3m 85 was 125
8 .E <25 <2 5 5 u 8 f <25 <2 5 5 u

5?.ch mmmﬁomm

.mmUOU EmEummb How conumm mvoﬁmﬁ mom .Aﬁmm 33x3. .mod v as Bogota bucmomﬁwmm was @832 “5.8wa 5:5
36.— 553, 282 .36ch 5 “£2 8 .83me 80b 3%: man 88QO 83258.5 2m 5 Hammond 98>» “m5 mmmowmm wan

anew ESE: «E co 3:0 36.8th v.83 m.<>OZ< .36QO show .8888 «mop: :3 m
H8 6353a 98 9:82 dug—58b BEmEinXw gm 2: 5 A~E\.o5 355% 9.8 A

N53

5 8* Es museum beam: m3
ass E s a:

103

The densities of annual, biennial, and perennial forbs all reﬂect the
impact of the experimental treatments (Table 12). There were no annual forbs
in the two treatments with elevated levels of litter (LA and NA treatments),
suggesting that the accumulation of litter caused by nitrogen enrichment may
have inhibited the establishment of annuals from the seed bank. The mean
density of biennial forbs (represented primarily by Daucus carota), was
reduced by nitrogen enrichment in the presence and absence of litter, but
these effects were not statistically signiﬁcant. Mean biennial forb density was
increased by litter removal, and decreased by litter addition. However, neither
of these effects were statistically signiﬁcant either. Biennial forb density was

signiﬁcantly lower in the two treatments with elevated levels of litter (LA
and NA) compared to the litter removal treatment that was not fertilized
(LR). Nitrogen enrichment signiﬁcantly reduced the density of perennial
forbs only when litter was present. Litter removal had no signiﬁcant affect on
the density of perennial forbs, while litter addition signiﬁcantly reduced it.
Species Richness - Species richness varied in response to the
experimental treatments in a similar manner as did forb density (Fig. 19F).
Nitrogen enrichment reduced mean species richness, although this effect was
significant only when litter was left intact. Litter removal increased mean
Species richness, but just as for forb density, this effect was signiﬁcant only in
the fertilized treatments where litter removal increased species richness up to
levels found in the controls. Litter addition signiﬁcantly reduced species
riChness relative to the control and to the same degree as observed in the
fertilized treatment where litter was left intact (NA treatment) suggesting that

the reduction in species richness in these plots caused by nitrogen enrichment

Was largely due to inhibition by litter.

104

And ropo gon recruitment

Final density and biomass per plant - Nitrogen enrichment reduced the
final density and biomass per plant of Andropogon to very low levels, both in
the presence and absence of litter (Fig. 21). In the unfertilized treatments, litter
removal signiﬁcantly increased Andropogon density and biomass per plant.
However, in the fertilized treatments, the removal of litter had no effect on
either measure. Litter addition reduced Andropogon density and biomass per
plant to the same low level as found in the fertilized plots.

Germination and survivorship - I examined Andropogon germination
and survivorship (Fig. 22) to assess the relative importance of each in
determining Andropogon recruitment (ﬁnal density). Nitrogen enrichment
significantly reduced germination and seedling survivorship in both the
presence and absence of litter. Litter removal signiﬁcantly increased
germination in both the unfertilized and fertilized treatments, but had no
affect on survivorship. Litter addition significantly reduced germination to
levels found in the fertilized plots where litter was left intact. Litter addition

reduced mean survivorship, but this effect was not statistically signiﬁcant.

105

F = 66.2""

 

Andropogon density (no. / m " 2)

 

C LR LA NA LRNA

0.08 - a

0.06 _

F = 41.9""

Andropogon bimass/ plant (g)

C C C

I I I

C LR LA NA LRNA

 

 

Treatments

I:igure 21. Treatment effects on Andropogon density (A), and biomass per
Plant (B). Means (+ 1 SE) with different letters are signiﬁcantly different (P <
0-05). ANOVA F430 values: W” P < 0.0001.

106

40-

  

30 a F = 39.9""

 

20..

10-

Andropogon germination (%)

 

OJ

C LR LA NA LRNA

80..

Andropogon seedling survival (%)

 

 

C LR LA NA LRNA

Treatment

l:"igure 22. Treatment effects on Andropogon germination (A), and
Sl-1rvivorship (B). Means (+ 1 SE) with different letters are signiﬁcantly
different (P < 0.05). ANOVA 134,20 values: m P < 0.001, *m P < 0.0001.

107

DISCUSSION

Community responses

In this study, nitrogen enrichment over two growing seasons led to an
increase in community productivity and a reduction in both forb density and
species richness. A major objective of this study was to assess the impact of
accumulated plant litter in contributing to community changes associated
with nitrogen enrichment, independent of fertilization and increased living
biomass. The results suggest that the inhibitory effects associated with
increased litter biomass on the establishment of subordinate forb seedlings
contributed strongly to the observed decline in species richness. Two lines of
evidence support this conclusion. First, the removal of litter increased forb
density and species richness in the fertilized plots up to levels found in the
controls. Second, the addition of litter reduced forb density and species
richness down to the level observed in the fertilized plots where litter was left
intact. This indicates that the increase in litter biomass associated with
nitrogen enrichment led to a reduction in forb density and species richness
independently of the increase in living biomass.

Species responses to nitrogen enrichment - The strong increase in
community productivity associated with nitrogen enrichment was due
entirely to the large biomass response of the dominant grass species, Bromus
inermis. Interestingly, none of the other grass or forb species that could be
examined statistically exhibited a signiﬁcant biomass response (either positive
or negative) to nitrogen enrichment. In contrast, the density of forbs was
reduced by nitrogen enrichment in both the presence and absence of litter, an

indication that the establishment of forb seedlings was inhibited by increased

108

living biomass. The apparent inability of most species to take advantage of
increased nitrogen availability was likely due to resource pre-emption by
Bromus. Because Bromus was the most dominant species at the site in terms
of biomass (> 80% relative biomass) and coverage (evenly distributed across
all of the experimental plots), it is likely that it was in a better position than
other species to utilize added nitrogen. This effect was likely enhanced by the
fact that fertilizer was ﬁrst applied early in the growing season when C3
grasses such as Bromus are typically most active in terms of growth (Waller
and Lewis 1979, Hetrick et al. 1990). A rapid response of Bromus to nitrogen
enrichment was visibly evident within two weeks of the first fertilizer
application. In the fertilized treatments, Bromus culms where considerably
taller and the leaves were noticeably greener than in the control, indicating
more rapid growth and higher leaf nitrogen concentrations respectively. It is
likely that the initial pre-emption of added nitrogen by Bromus, and the
resulting increase in biomass and plant height, led to the preemption of light
as well. This, in turn, could have suppressed the subordinate forb species and
further inhibited their ability to utilize the subsequent applications of
nitrogen for growth.

Species responses to litter manipulations - Surprisingly, there were no
signiﬁcant biomass responses to the litter manipulations by any of the
individual species, or by species grouped as grasses, forbs, or by life history
categories. In contrast, the litter manipulations strongly affected the density of
forbs by affecting the establishment of seedlings. Species richness responded to
the litter manipulations in a similar manner as forb density, indicating that
the decline in species richness in response to nutrient enrichment and
increased productivity in plots where litter was left intact was largely due to

the negative impact of plant litter on forb seedling recruitment.

109

Although there is no direct evidence, there are a several potential
direct and indirect mechanisms that may have contributed to the negative
effects of litter on forb density and species richness in this study. Litter can act
as a mechanical barrier to seeds and seedlings, thus directly preventing the
germination and emergence of some species due to physical impedance
(Fenner 1985, Sydes and Grime 1981, Hamrick and Lee 1987, Persson et al.
1987, Facelli and Pickett 1991a). Increases in litter biomass in this experiment
caused by fertilization and litter addition may have increased the
effectiveness of this barrier, possibly preventing many of the recently
dispersed seeds from reaching the soil surface, and allowing few seedlings to
penetrate upwards towards the light.

Plant litter can indirectly effect plant performance in a number of ways
by altering light levels and other factors that inﬂuence microclimate such as
temperature, moisture and humidity (Bazzaz 1979, Goldberg and Werner
1983, Gross 1984, Carson and Peterson 1990, Facelli and Pickett 1991a). The
negative effects of litter on forb seedling recruitment observed in this study
may partially reﬂect the elimination of germination cues caused by the
dampening effects of increased litter biomass on environmental ﬂuctuations
(Facelli and Pickett 1991a). In addition, it is likely that the moist and darkened
conditions under the litter in the fertilized treatment could have promoted
the increased activity of fungal pathogens and invertebrate herbivores,
resulting in enhanced seed and seedling mortality (Goldberg and Werner
1983, Facelli 1994). Although herbivory was not measured in this study, the
occurrence of herbivorous mollusks (slugs) were greatest in the litter addition
plots and the nitrogen addition plots where litter was left intact (unpublished

data).

110

Plant responses and light availability

In this study, the effects of living and litter biomass on light availability
likely played an important role in reducing forb density and species richness.
Light levels at the soil surface were reduced to equally low levels in the litter
addition and fertilized plots, including the fertilized plots where litter had
been removed. This indicates that both living and litter biomass were equally
effective in attenuating light. Significant reductions in light levels and forb
densities in response to nitrogen enrichment in plots where litter was
removed strongly suggest that light attenuation by living biomass contributed
to the observed reductions in forb density and species richness. Further
reduction in forb density and species richness evident in the litter addition
plots and in the fertilized plots that contained litter, suggests that in addition
to effects caused by light attenuation by living plants, additional effects
associated with litter accumulation further reduced forb density and species
richness.

It is reasonable to assume that the observed reductions in forb density
and species richness that can be attributed to increases in living plant biomass
were caused primarily through the effects of shading. However, a question
arises regarding the degree to which the negative impacts of litter addition
were caused by effects of shading versus other litter effects unrelated to light
availability (mechanical impedance, microclimatic effects etc.). These separate
effects of litter can potentially be isolated by comparing responses in the litter
addition plots to responses in the fertilized plots where litter was removed
(LRN A). The observed signiﬁcant difference in forb density and species
richness between these two treatments can be attributed to the inhibitory

effects of litter unrelated to light availability if it assumed that the separate

111

effect of living biomass on forb seedlings can be used as a reliable surrogate
measure of the isolated effect of litter biomass on forb seedlings which
occurred purely as a result of shading. This assumption is reasonable given
that living and litter biomass reduced light availability to equally low levels
in this study. With this assumption in mind, it can be inferred that the
observed reductions in forb density and richness attributable to litter addition
cannot be entirely explained by shading alone. As a result, it is likely that
multiple mechanisms involving light attenuation by both living plants and
litter, and various other direct and indirect effects of litter unrelated to light
availability all combined to reduced forb seedling recruitment and species

richness in this study.

And ropo gon recruitment

The purpose of adding Andropogon seeds to the experimental plots
was to more closely examine the effects of nitrogen enrichment and litter on
recruitment from seed. The results give some insight into the potential role
of established vegetation in affecting the colonization success of this native
species in southwest Michigan old-ﬁelds that differ in soil fertility and plant
productivity. Unlike what was observed for forb density, the removal of litter
in the unfertilized treatments led to a large signiﬁcant increase in
Andropogon density, indicating that litter represents a significant constraint
to recruitment in this ﬁeld. However, despite inhibition by litter,
Andropogon was able to successfully establish a reasonable number of
seedlings in the control plots, suggesting that the absence of this species from
this study site, and other ﬁelds of similar composition and productivity, may

be due primarily to limits on dispersal. In contrast, nitrogen enrichment and

112

litter addition reduced both Andropogon recruitment and growth (density
and biomass per plant) to near zero in this study. This results suggests that the
absence of Andropogon from the more highly productive fields in southwest
Michigan that support high standing crops may be due to inhibition of
seedling recruitment by the established vegetation in addition to any limits
on dispersal. In this study, the strong negative effect of litter addition and
nitrogen addition on Andropogon, and the lack of the positive response to
litter removal in the fertilized plots that was observed in the case of forbs,
indicates that the negative effect of nitrogen enrichment on Andropogon
recruitment in plots where litter was left intact can be suitably explained by
the separate effects of either litter or living biomass. The results also indicate
that litter reduced Andropogon recruitment primarily through inhibition of
germination and that living plants inhibited recruitment by reducing both
germination and survivorship.

The management implications of the results from these seed additions
are that in successional grasslands and old-ﬁelds of moderate productivity,
litter removal accomplished through burning or mechanical methods may be
very effective in facilitating the recruitment of native grasses from seed.
However, in highly productive communities, litter removal may be much
less effective in this regard due to intense competitive inhibition imposed by

productive shoot growth.
Productivity, diversity, and plant litter
Most theories designed to explain the interdependence of species

diversity and habitat productivity invoke the role of resource competition

and make the tacit assumption that the relevant plant interactions involved

113

are exploitative in nature. In other words, competitive exclusion occurs
through the direct consumptive effects of living roots and shoots on
resources. However, this study and others provide strong evidence that
declines in species richness with increased fertility and productivity in old-
ﬁeld and grassland communities may be closely tied to inhibitory effects
associated with accumulated plant litter. As in this study, Carson and
Peterson (1990) found that experimental additions of plant litter reduced plant
density and species richness in a Solidago dominated old-field. From this they
argued that low diversity in highly productive old-ﬁelds and reductions in
diversity occurring in response to nutrient enrichment may often be due to a
combination of both inhibition by litter and resource competition associated
with living plants. Tilman (1993) found that declines in species richness
associated with long-term nitrogen enrichment in experimental grasslands
were more strongly tied to variation in litter biomass than to variation in
living biomass. These results were largely attributed to the negative impacts
of a dense litter layer on the establishment of seedlings and the colonization
of new species in the most productive sites.

An assumption implicit in the resource-competition-based diversity
theories is that increased productivity reduces species richness because it
causes an increase in the rate of species loss (Tilman 1993). However, the
number of species within a habitat at a given time can be determined by a
balance between extinction and colonization (MacArthur and Wilson 1969).
Tilman (1993) found that extinction and colonization rates were equally
important in determining the species richness of grasslands, and argued that
earlier theories should be modiﬁed to account for the effects of productivity
on colonization. The results of this study are consistent with the conclusions

of Tilman (1993) that increased productivity can reduce species richness in

114

grassland by reducing rates of species colonization. Although I did not directly
measure species gains and loss over time in this study, it was clear that the
decline in species richness associated with nitrogen enrichment was primarily
due to the inhibition of forb seedling establishment. The strong negative
impact of nitrogen enrichment on the germination, survival, and
recruitment of Andropogon highlights the role of increased productivity in
limiting the colonization potential of a single species.

To my knowledge this is the first study utilizing litter manipulations
in conjunction with fertilization to explicitly examine the independent effects
of living plants and accumulated plant litter in contributing to community
changes associated with nutrient enrichment and increased productivity. The
results are in concordance with the frequently observed decline in species
richness in response to increased plant productivity, but suggests that the
mechanism responsible for this decline may often involve the inhibitory
effects of litter on species colonization in addition to the exploitative effects of
living plants. I suggest that theories designed to explain the interdependence
of species richness and productivity in herbaceous vegetation should take

into account the potential impacts of plant litter on community organization.

CHAPTER 5

CONCLUSIONS

115

116

Results from the removal experiments discussed in Chapters 2 and 3
showed that the surrounding plant community suppressed the recruitment
and growth of the three target species across a wide range of community
biomass. When considering the entire range of community biomass
examined (Chapter 3) the magnitude of the effects of the surrounding plant
community (net effects) on both recruitment and growth of Andropogon and
Schizachyrium increased with community biomass as predicted by the
biomass-dependent theories of plant community organization (Grime 1979,
Keddy 1990). However, these relationships were nonlinear, indicating that the
degree of inter-dependence between net community effects on target plants
and community biomass was stronger from low to medium biomass than it
was from medium to high biomass. In fact, net community effects on
established Andropogon and Schizachyrium transplants increased with
community biomass only over a very narrow range of low biomass,
indicating that the biomass-independent view of plant competition may be
more applicable, in the case of established plants, for most old-ﬁeld plant
communities in southwest Michigan. Unlike net effects on established
transplants, the net effects on seedling recruitment of Andropogon and
Schizachyrium Were sensitive to changes in community biomass across the
entire range examined, although more sensitive from low to medium
biomass than from medium to high biomass. The non-linear relationships
observed in this study between net effects and community biomass across a
broad range of biomass may explain why some studies measuring
competition over more restricted ranges of productivity often find no

relationship between competition and productivity. In fact, the conclusion in

117

this study that the net effect of the plant community on Sorghastrum
transplant growth was unrelated to community biomass may reﬂect the more
restricted range of community biomass over which effects were measured for
this species.

I suggest that it is important to include the entire range of productivity
or plant biomass of a region when examining variation in competition to test
prior predictions regarding the relationship between competition and
productivity. My results suggest that it may be especially important to include
the most infertile, and unproductive habitats in competition studies because
it is across such sites where individual plants may be most sensitive to
changes in competitor biomass.

The degree to which the effects of accumulated plant litter contributed
to the net effect of the community on target plants in this study depended on
the life history stage examined. Inhibitory effects of litter increased with
community biomass and contributed to the net effect of the community on
target plant recruitment at sites of intermediate community biomass.
However, litter had little impact on established transplants and contributed
little to net effects.

These removal experiments revealed a high degree of site to site
variability in plant community effects on the recruitment and growth of
native grasses and highlight the potential importance of accumulated plant
litter in suppressing seedling recruitment in productive old-ﬁeld
communities. The results suggest that the absence of native C4 grasses in
many southwest Michigan old-ﬁelds and their apparent restriction to
unproductive sites is related to a combination of dispersal limitation and the
inhibitory impact of both living plant neighbors and litter on colonization

success in the most productive sites that support high standing crop and litter

118

biomass. Such constraints to species colonization imposed by the established
vegetation may be important in contributing to low diversity in highly
productive communities as suggested by other studies (Carson and Peterson
1990, Tilman 1993).

The nitrogen enrichment / litter manipulation experiment (Chapter 4)
demonstrated the short term effects of nitrogen enrichment, increased
productivity and litter accumulation on species colonization and species
richness within a single old-ﬁeld site. Nitrogen enrichment led to a large
increase in the biomass of both living plants and plant litter primarily by
enhancing the productivity of the dominant grass species (Bromus inermis).
This increase in productivity resulted in a signiﬁcant decline in forb seedling
recruitment, Andropogon recruitment from added seed, and species richness.
Results of the litter manipulations suggested that these declines were due to
suppressive effects associated with both increased living biomass and plant
litter. These results are in accordance with declines in species diversity
observed in other studies, but indicate that the mechanism responsible for
reduced diversity may involve both the inhibitory effects of plant litter and
resource competition occurring among living plants.

Tilman (1993) argued that if there is empirical evidence that
productivity inﬂuences the rate at which species colonize communities, then
diversity theories will have to be modiﬁed to account for this. Tilman (1993)
demonstrated that declines in species richness with increased productivity in
Minnesota grasslands were due to changes in both colonization and
extinction rates. Furthermore, reduced colonization rates appeared to be
strongly related to increased litter biomass.

The results of the three experiments discussed in this dissertation

demonstrate the potential effects of productivity on native grass and forb

119

colonization, and on the species diversity of old-ﬁeld plant communities. I
suggest that current theories of community organization, as applied to
grassland and old-ﬁeld vegetation, should be modiﬁed to account for the
effects of productivity and community biomass on species colonization, and
the role that accumulated plant litter potentially plays in this process.

The results of these experiments have implications for management
goals which seeks to enhance or restore native plant diversity in disturbed
successional landscapes. Results from the seed addition experiments suggest
that low-impact approaches to species re-establishment may be feasible in
southwest Michigan old-fields that are of low to intermediate productivity. In
relatively infertile and low productivity old-fields, the establishment of
native grasses might be accomplished solely by mediating dispersal limitation
through the addition of seed. In old-ﬁelds of intermediate productivity, the
removal of accumulated plant litter (either through periodic prescribed burns
or mechanical methods) may be effective in combination with seed additions
to enhance the establishment success of native grasses and to promote higher

species diversity.

LIST OF REFERENCES

120

LIST OF REFERENCES

Abrams, P. A. 1995. Monotonic or unimodal diversity-productivity
gradients: what does competition theory predict? Ecology 76:2019-2027.

Al-Mufti, M. M., C. L. Sydes, S. B. Furness, J. P. Grime, and S. R. Band. 1977.
A quantitative assessment of shoot phenology and dominance in
herbaceous vegetation. Journal of Ecology 65:759-91.

Bakelaar, R. G and E. P. Odum. 1978. Community and population level
responses to fertilization in an old-ﬁeld ecosystem. Ecology 59:660-665.

Bazazz, F. A. 1979. The physiological ecology of plant succession. Annual
Review of Ecology and Systematics 10:351-371.

Bazzaz, F. A., and J. A. D. Parrish. 1982. Organization of grassland
communities. Pages 233-254 in J. R. Estes, editor. Grasses and grassland
communities. Oklahoma State University Press, Stillwater, Oklahoma,
USA.

Berendse, F. and W. T. Elberse. 1990. Competition and nutrient availability
in heathland and grassland ecosystems. Pages 93-116 in J. B. Grace and D.

Tilman, editors. Perspectives on Plant competition. Academic Press, San
Diego, California, USA.

Bergelson, J. 1990. Life after death: site pre-emption by the remains of Poa
annua. Ecology 71:2157-2165.

Bertness, M. D., and R. Callaway. 1991. Positive interactions in communities.
Trends in Ecology and Evolution 9:191-193.

Bertness, M. D., and S. D. Hacker. 1994. Physical stress and positive
associations among marsh plants. American Naturalist 144:363-372.

Bonser, S. P., and R. J. Reader. 1995. Plant competition and herbivory in
relation to vegetation biomass. Ecology 76:2176-2183.

Burbank, D., K. S. Pregitzer, and K. L. Gross. 1992. Vegetation of the W. K.
Kellogg Biological Station. Michigan State University Agricultural
Experiment Station Research Report Number 501.

121

Campbell, B. D., and J. P. Grime. 1992. An experimental test of plant strategy
theory. Ecology 73:15-29.

Carson, W. P., and G. W. Barrett. 1988. Succession in old-ﬁeld plant
communities: effects of contrasting types of nutrient enrichment. Ecology
69:984-994.

Carson, W. P., and C. J. Peterson. 1990. The role of litter in an old-ﬁeld
community: impact of litter quantity in different seasons on plant
species richness and abundance. Oecologia 85:8-13.

Chiariello, N. R., H. A. Mooney, K. Williams. 1989. Growth, carbon
allocation and cost of plant tissues. Pages 327-365 in R. W. Pearcy, J.

Ehleringer, H. A. Mooney, and P. W. Rundel, editors. Plant Physiological
Ecology. Chapman and Hall, New York, USA.

Dix, R., and F. Smeins. 1967. The prairie, meadow and marsh vegetation of
Nelson county, North Dakota. Canadian Journal of Botany. 45:21-58.

Facelli, J. M. 1994. Multiple indirect effects of plant litter affect the
establishment of woody seedlings in old ﬁelds. Ecology 75:1727-1735.

Facelli, J. M., and S. T. A. Pickett. 1991a. Plant litter: its dynamics and effects
on plant community structure. The Botanical Review 57:1-31.

Facelli, J. M., and S. T. A. Pickett. 1991b. Plant litter: light interception and
effects on an old-ﬁeld plant community. Ecology 72:1024-1031.

Fenner, M. 1985. Seed Ecology. Chapman and Hall, New York, New York,
USA.

Fernald, M. L. 1950. Gray's manual of Botany. Dioscorides Press, Portland,
Oregon, USA.

Foster, B. L. 1992. The role of land use history in structuring an old ﬁeld
plant community. Thesis. Michigan State University, East Lansing,
Michigan, USA.

Foster, B. L. and K. L. Gross. (in press). Partitioning The effects of plant
biomass and litter on Andropogon gerardi in old-ﬁeld vegetation. Ecology.

Fowler, N. L. 1986. Microsite requirements for germination and
establishment of three grass species. American Midland Naturalist
115:131-145.

122

 

. 1990. The effects of competition and environmental heterogeneity
on three coexisting grasses. Journal of Ecology 78:389-402.

Goldberg, D. E., and P. A. Werner. 1983. The effects of size of opening in

vegetation and litter cover on seedling establishment of goldenrods
(Solidago spp.). Oecologia 60:149-155.

Goldberg, D. E., and T. E. Miller. 1990. Effects of different resource additions
on species diversity in an annual plant community. Ecology 71:213- 225.

Goldberg, D. E., and A. M. Barton. 1992. Patterns and consequences of
interspeciﬁc competition in natural communities: a review of ﬁeld
experiments with plants. American Naturalist 139:771-801.

Gotshall, T. B. 1972. The vegetation of Kalamazoo County at the time of
settlement. Pages 1-21 in R. Brewer, editor. The ecology of Kalamazoo

County. Western Michigan University Press, Kalamazoo, Michigan,
USA.

Grace, J. B. 1993. The effects of habitat productivity on competition intensity.
Trends in Ecology and Evolution. 8:229-230.

Grime, J. P. 1973. Competitive exclusion in herbaceous vegetation. Nature
242:344- 347.

 

. 1979. Plant strategies and vegetation processes. J. Wiley & Sons,
Chichester, England.

Gross, K. L. 1984. Effects of seed size and growth form on seedling
establishment of six monocarpic perennial plants. Journal of Ecology
72:369-387.

Gross, K. L., and P. A. Werner. 1982. Colonizing abilities of "biennial' plant
species in relation to ground cover: implications for their distributions in
a successional sere. Ecology 63:921-931.

Grubb, P. J. 1977. The maintenance of species-richness in plant communities:
the importance of the regeneration niche. Biological Reviews 52:107-145.

Gurevitch, J. 1986. Competition and the local distribution of the grass Stipa
neomexicana. Ecology 67:46-57.

Hamrick, J. L., and J. M. Lee. 1987. Effect of soil surface topography and
litter cover on the germination, survival, and growth of musk thistle
(Carduus nutans). American Journal of Botany 74:451-457.

123

Haslam, S. M. 1971. Community regulation in Phragmites communis Trim.
1. monodominant stands. Journal of Ecology 59:65-73.

Hetrick, B. A. D., G. W. T. Wilson, and T. C. Todd. 1990. Differential
responses of C3 and C4 grasses to mycorrhizal symbiosis, phosphorus
fertilization, and soil microorganisms. Canadian Journal of Botany 68:461-
467.

Hulbert, L. C. 1969. Fire and litter effects in undisturbed bluestem prairie.
Ecology 50:874-877.

Huston, M. A. 1979. A general hypothesis of species diversity. American
Naturalist. 113281-101.

Huston, M. A., and D. L. DeAngelis. 1994. Competition and coexistence: the
effects of resource transport and supply rates. American Naturalist
144:954-977.

Kadmon, R. 1995. Plant competition along soil moisture gradients: a ﬁeld
experiment with the desert annual Stipa capensis. Journal of Ecology
83:253-262.

Keddy. P. A. 1989. Competition. Chapman and Hall, London, England.

 

. 1990. Competitive hierarchies and centrifugal organization in plant
communities. Pages 266-290 in J. B. Grace and D. Tilman, editors.
Perspectives on Plant Competition. Academic Press, San Diego,
California, USA.

Knapp, A. K. 1985. Water relations and growth of three grasses during wet
and drought years in a tallgrass prairie. Oecologia 65:35-43.

Knapp, A. K., and T. R. Seastedt. 1986. Detritus accumulation limits
productivity of tallgrass prairie. Bioscience 36:662-668.

MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography.
Princeton University Press, Princeton, New Jersey, USA.

Marrs, R. H., J. B. Grace, and L. Gough. 1996. On the relationship between
plant species diversity and biomass: a comment on a paper by Gough,
Grace and Taylor.

Miller, T. E. 1996. On quantifying the intensity of competition across
gradients. Ecology 77:978-981.

124

Newman, E. I. 1973. Competition and diversity in herbaceous vegetation.
Nature 244:310.

Parrish, J. A. D., and F. A. Bazzaz. 1982. Competitive interactions in plant
communities of different successional ages. Ecology 63:314-320.

Persson, S., N. Malmer, and B. Wallen. 1987. Leaf litter fall and soil acidity
during half a century of secondary succession in temperate deciduous
forests. Vegetatio 73:31-45.

Puerto, A., M. Rico, M. D. Matias, and J. A. Garcia. 1990. Variation is
structure and diversity in Mediterranean grasslands related to trophic
status and grazing intensity. Journal of Vegetation Science. 1:445-452.

Pysek, P., and J. Leps. 1991. Response of a weed community to nitrogen
fertilization: a multivariate analysis. Journal of Vegetation Science 2:237-
244.

Rabinowitz, D. 1978. Abundance and diaspore weight in rare and common
prairie grasses. Oecologia 37:213-219.

Shipley, B., P. A. Keddy, and L. P. Leﬂ<ovitch. 1991. Mechanisms producing
plant zonation along a water depth gradient: a comparison with the
exposure gradient. Canadian Journal of Botany 69:1420-1424.

Smeins, F. E., and D. E. Olson. 1970. Species composition and production of a
native northwestern tall grass prairie. American Midland Naturalist
84:398-410.

Sokal, R. R., and F. J. Rohlf. 1981. Biometry. W. H. Freeman, New York, New
York. USA.

Southwood, T. R. E. 1988. Tactics, strategies, and templates. Oikos 52:3-18.

Sydes, C., and J. P. Grime. 1981. Effects of tree leaf litter on herbaceous
vegetation in deciduous woodland 11: an experimental investigation.
Journal of Ecology 69:249-262.

Taylor, D. R., L. W. Aarssen, and C. Loehle. 1990. On the relationship
between r/ K selection and environmental carrying capacity: a new
habitat template for plant life history strategies. Oikos 58:239-250.

Thompson, K., and J. P. Grime. 1988. Competition reconsidered: a reply to
Tilman. Functional Ecology 2:114-116.

125

Tilman, D. 1982. Resource competition and community structure. Princeton
University Press, Princeton, New Jersey, USA.

 

. 1987. Secondary succession and the pattern of plant dominance
along experimental nitrogen gradients. Ecological Monographs 57:189-214.

 

. 1988. Plant strategies and the dynamics and structure of plant
communities. Princeton University Press, Princeton, New Jersey, USA.

 

. 1993. Species richness of experimental productivity gradients: how
important is colonization limitation? Ecology 74:2179-2191.

Tilman, D., and D. Wedin. 1991a. Plant traits and resource reduction for five
grasses growing on a nitrogen gradient. Ecology 72:685-700.

Tilman, D., and D. Wedin. 1991b. Dynamics of nitrogen competition between
successional grasses. Ecology 72:1038-1049.

Tilman, D., and S. Pacala. 1993. The maintenance of species richness in plant
communities. Pages 13-25 in R. E. Ricklefs and D. Schluter, editors. Species
diversity in ecological communities: historical and geographical
perspectives. University of Chicago Press, Chicago, Illinois, USA.

Twolan-Strutt, L., and P. A. Keddy. 1996. Above- and belowground
competition intensity in two contrasting wetland plant communities.
Ecology 77:259-270.

Waller, S. S., and J. K. Lewis. 1979. Occurrence of C3 and C4 photosynthetic

pathways in North American grasses. Journal of Range Management
32:12-28.

Weaver, J. E. 1954. North American prairie. University of Nebraska Press,
Lincoln Nebraska, USA. '

Werner, P. A. 1975. The effects of plant litter on germination in teasel,
Dipsacus sylvestris Huds. American Midland Naturalist 94:470-476.

Wilkinson, L. 1989. SYSTAT: the system for statistics. SYSTAT, Evanston,
Illinois, USA.

Willms W. D., S. Smoliak, and A. W. Baily. 1986. Herbage production
following litter removal on Alberta native grasslands. Journal of Range
Management 39:536-540.

126

Wilson, S. D., and P. A. Keddy. 1986. Measuring diffuse competition along an
environmental gradient: results from a shoreline plant community.
American Naturalist 127:862-869.

Wilson S. D., and J. M. Shay. 1990. Competition, fire, and nutrients in a
mixed-grass prairie. Ecology 71:1959-1967.

Wilson, S. D., and D. Tilman. 1991. Components of plant competition along
an experimental gradient of nitrogen availability. Ecology 72:1050-1065.

Wilson, S. D., and D. Tilman. 1993. Plant competition in relation to
disturbance, fertility and resource availability. Ecology 74:599-611.

Wilson, S. D., and D. Tilman. 1995. Competitive responses of eight old-ﬁeld
plant species in four environments. Ecology 76:1169-1180.

Zar, J. H. 1996. Biostatistical analysis. Prentice-Hall, Englewood Cliffs, New
Jersey, USA.

"Illlllllllllllllll