 

 

k.
z...

5.3.3.5.”.

.mmﬁuﬁagni an.

. $321...

 

m .1
H¥ux§.%
(.- 1 a: A

.Fémﬁbiﬁ
3%....aafn
.. n W .IVL‘WV

.: ‘
. It;
’7’...

a»... 355.1.
1.3“. I an:

. a. ‘
1.1.3” 1...... fa
,_ ,

‘ “mama‘s
43...}...
F para-a“.

2...» 9h

..
! .,
1.4.9
“A

.5... ‘ 3
.a 51‘ vhu JJIHII .

. m? .5.
, ‘ . .

fear”? 9...: .0.

a... . ...1§o.u»w ) )‘

. u: .3

.13.

3515qu s .‘ .,.. Lash 1.x!
. .. . I . .321
. x

 

(‘A‘O .‘

a o 2%

[037(000 .1! Q

This is to certify that the
thesis entitled

TEMPERATE DECIDUOUS FOREST FRAGMENTS:
EDGE EFFECTS, INVASION BY NON-NATIVE PLANTS, AND
LONG-TERM CHANGE IN MATURE FOREST STRUCTURE

presented by
CHARLOTTE MURRAY REEMTS

has been accepted towards fulfillment
of the requirements for the

MS. degree in Plant Biology

 

 

@mﬁw

Major Professor’ 3 gignature‘

1:0 J .,

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

LIBRARIES
MICHIGAN STATE UN ‘vEIESITY
EAST LASV'SING MICH 488-4-1048

-:-.-.-.----.--.—.---u-o---—.—.-.-.-.--.--o--—--—u-------.-.-.-A-.-.—.-.-o----.—--

---—‘—4...--.- -.- -.--

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

 

DATE DUE DATE DUE DATE DUE

 

I

h 0
n2
3-.
A.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 c:/ClRC/DateDua.Indd-p.15

 

TEMPERATE DECIDUOUS FOREST FRAGMENTS:
EDGE EFFECTS, INVASION BY NON-NATIVE PLANTS, AND LONG-TERM
CHANGE IN MATURE FOREST STRUCTURE
By

Charlotte Murray Reemts

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Plant Biology

2005

ABSTRACT
T EMPERATE DECIDUOUS FOREST FRAGMENTS:
EDGE EFFECTS, INVASION BY NON-NATIVE PLANTS, AND LONG-TERM
CHANGE IN MATURE FOREST STRUCTURE
By
Charlotte Murray Reemts
Fragmentation and non-native species are two of the most serious threats to
biodiversity today. The extent of edge effects and non-native invasion were quantiﬁed in
ten forest fragments (beech-maple and oak-hickory types) in southern Michigan. Aspect
signiﬁcantly inﬂuenced the extent of edge effects based on vegetation structure, but not
vegetation composition. Edge effects penetrated up to 90 m into forest interiors and were
greater in beech-maple (x=12.7 m) than in oak-hickory forests (x=5.8 m). Herb layer
structure and understory composition were most sensitive to edge effects. Non-native
Species made up 24% of sampled species richness. In beech-maple forests, most non-
natives were limited to within 20 m of the edge. Non-native richness (43 vs. 26 species),
cover (0.08% vs. 0.01%), and stem density (1.2 vs. 0.1 stems/m2) were highest in oak-
hickory forests. Soil seed bank density at ﬁve sites was higher at the edge (3.85 vs. 3.35
seeds/m2). Non-native species made up 30% of the seed bank. Most abundant non-
native species in the vegetation were absent from the sampled seed bank, suggesting that
the seed bank does not contribute to their invasiveness. While basal area, stem density,
and species richness of Tourney Woodlot, a small, old-growth beech-maple fragment, are
similar to larger old-growth forests, it is dominated by Acer saccharum rather than F agus

grandifolia. This difference could be caused by the woodlot’s small size and the

inﬂuence of edge effects on the competitive balance between the dominants.

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Peter Murphy, for his guidance and support
of this research project. Kelly Millenbah and Frank Telewski, the other members of my
graduate committee, have provided insight and direction. Frank Telewski also provided
historical information about Toumey Woodlot from the Campus Parks and Planning
archives. The Statistical Consulting Center, especially Juan Pedro Steibel, provided
invaluable help in analyzing my data. Carolyn Malmstrom kindly allowed me to use her
leaf area meter. The use of the Michigan Natural Features Inventory’s liSt of exceptional
sites helped me identify many of my research sites. The Michigan Department of Natural
Resources, the MSU Campus Natural Areas Committee, Kent County Parks, Grand
Rapids Parks, and Ken Poﬂ‘ all gave me permission to work at their sites. The MSU
Herbarium allowed me to use their facilities to conﬁrm plant identiﬁcations. Four
undergraduate students, Darin Ellair, Nicholas Daum, Elizabeth Grisham, and Maggie
Allen, assisted with ﬁeld and lab work. My research, as well as travel to present this
research, was funded by the Department of Plant Biology and the Ecology, Evolutionary
Biology, and Behavior Program. A generous grant from the Hanes Fund supported my
research and allowed me to replace missing plot markers in Tourney Woodlot. Finally, a
special thanks to my friends and family, especially my husband Adam, who provided

endless support and encouragement.

iii

TABLE OF CONTENTS

LIST OF TABLES .............................................................................. vii
LIST OF FIGURES ...................................................................................................... xii
CHAPTER I:

IMPACTS OF FOREST FRAGMENTATION AND NON-NATIVE PLANTS ON
NATIVE PLANT COMMUNITIES ............................................................................ 1
CHAPTER 2:

ANATOMY OF A FRAGMENT: EDGE EFFECTS IN BEECH-MAPLE AND
OAK-HICKORY FOREST FRAGMENTS

Introduction ............................................................................................................. 12
Methods
Site descriptions ................................................................................................. 21
Vegetation structure and composition ................................................................ 22
Leaf area index (LAI) ......................................................................................... 24
Canopy openness ................................................................................................ 25
Statistical analyses .............................................................................................. 26
Results
Vegetation structure
Depth of edge inﬂuence (DEI): Herb layer ................................................... 3O
DEI: Shrub layer ............................................................................................ 30
DEI: Sapling layer ......................................................................................... 31
DEI: Canopy layer ......................................................................................... 31
Aspect ............................................................................................................ 39
Openness ....................................................................................................... 4O
Species composition
Understory ..................................................................................................... 41
Canopy .......................................................................................................... 42
Multi-permutation response procedure (MRPP) .......................................... 43
Cluster analysis ............................................................................................. 45
Discussion
Vegetation structure
DEI: Herb layer ............................................................................................. 48
DEI: Shrub and sapling layers ....................................................................... 49
DEI: Canopy layer ......................................................................................... 49
Effects of aspect and openness ...................................................................... 51
Species composition
Understory ..................................................................................................... 52
Canopy .......................................................................................................... 54
Grouping procedures ..................................................................................... 54
Conclusions ............................................................................................................. 56

iv

CHAPTER 3:
LIFE ON THE EDGE: THE SPATIAL DISTRIBUTION OF NON-NATIVE
SPECIES IN FOREST F RAGMENT EDGES

Introduction ............................................................................................................. 59
Methods
Site descriptions ................................................................................................. 66
Vegetation sampling ........................................................................................... 66
Statistical analyses .............................................................................................. 67
mem
Herbs .................................................................................................................. 71
Shrubs and trees ................................................................................................. 72
Abundant species ................................................................................................ 73
Effects of species characteristics ........................................................................ 80
Effects of canopy openness and distance ........................................................... 81
Discussion
Herbs .................................................................................................................. 85
Shrubs and trees ................................................................................................. 86
Effects of species characteristics ........................................................................ 87
Effects of canopy openness ................................................................................ 88
Site effects .......................................................................................................... 90
Conclusions ............................................................................................................. 91
CHAPTER 4:
DRAWING ON THE BANK: NON-NATIVE SPECIES IN FOREST EDGE
SEED BANKS
Introduction ............................................................................................................. 93
Methods
Site descriptions ................................................................................................. 100
Sampling and germination ................................................................................. 100
Statistical analyses .............................................................................................. 102
Results
Seed density ........................................................................................................ 108
Species richness .................................................................................................. 108
Species distribution ............................................................................................ 113
Effects of aspect and distance ............................................................................ 113
Grouping procedures .......................................................................................... 1 16
Discussion
Seed density ........................................................................................................ 119
Species richness .................................................................................................. 120
Effects of aspect and distance ............................................................................ 121
Non-native species ............................................................................................. 123
Conclusions ............................................................................................................. 126
CHAPTER 5:

SIZE DOES MATTER: LONG-TERM CANOPY DYNAMICS IN A SMALL

OLD-GROWTH FOREST FRAGMENT

Introduction ............................................................................................................. 127
Methods
Site history and description ................................................................................ 133
Canopy tree census ............................................................................................. 136
Statistical analyses .............................................................................................. 137
Results
Species richness .................................................................................................. 140
Changes in stem density and basal area ............................................................. 140
Abundant species ................................................................................................ 143
Species distribution ............................................................................................ 147
Cluster analysis .................................................................................................. 148
Discussion
Species richness .................................................................................................. 152
Dominant species ............................................................................................... 152
Changes in stem density and basal area ............................................................. 154
Abundant species ................................................................................................ 156
Species distribution ............................................................................................ 158
Extent of edge inﬂuence ..................................................................................... 160
Conclusions ............................................................................................................. 161
CHAPTER 6:
CONCLUSIONS: THE IMPORTANCE OF EDGE EFFECTS IN FOREST
ECOLOGY ................................................................................................................... 163
APPENDIX:
2004 TOUMEY WOODLOT DATA .......................................................................... 169
LITERATURE CITED ................................................................................................ 202

vi

LIST OF TABLES

Table 1.1: Prevalence of non-native species in a variety of ecosystems in the
eastern deciduous forest realm, based on a literature survey of site ﬂoras and other
studies. --=no data available 7

Table 1.2: Comparison of characteristics of beech-maple and oak-hickory forests.
Unless otherwise noted, all values are based on trees _>_10 cm dbh. Importance
values are based on relative density and relative dominance ........................... 9

Table 2.1: Dominant canopy species for each site, as determined by basal area and
importance values (IV). Importance values were calculated as the sum of relative
dominance, relative density, and relative frequency, and are presented as the

proportion of the total (3 00). Beech-maple sites were dominated by A. saccharum

and F. grandifolia. Sites where Q. alba, Q. rubra, or Q. velutina were among the
canopy dominants, and Carya spp. were present, were classiﬁed as oak-hickory.

Data are based on 100 m2 plots located >100 m from the closest edge ...................... 19

Table 2.2: Fragment descriptions. Edges are described as permanent (p) if their
position is maintained by fences, roads, or recurring disturbance. Successing

edges (8) are moving into the adjacent old ﬁelds without interference. Adjacent

land use describes the land use within ~100m of the edge. Area and perimeter

were determined from digital aerial photos (MiGDL). oh=oak-hickory,
bm=beech-maple ......................................................................................................... 20

Table 2.3: P values (from ANOVA F-tests) for vegetation structure. Analyses

included both native and non-native species. Stem Dens.=stem density, LAI=leaf

area index, D x A= interaction between distance and aspect. Values in parentheses

are not signiﬁcant ........................................................................................................ 31

Table 2.4: Penetration of edge effects (In), determined by comparisons with the
interior (100 m) for each aspect and forest type. The analysis included both native

and non-native species. Effects are signiﬁcant at the p<0.05 level; values in
parentheses are signiﬁcant at the p<0.1 level. Only edge effects signiﬁcant at

a=0.05 were used to calculate summary statistics. Gray boxes indicate that metric
values were higher at the edge; signiﬁcant effects in white boxes indicate smaller
values at the edge. Exterior plots (ep) extend from the forest edge into the

adjacent habitat (Figure 2.2). -- = no effects detected. BA=basal area, LAI=leaf

area index .................................................................................................................... 33

Table 2.5: Cumulative R2 values from Type 1 ANOVA. R2 values represent the
proportion of variation explained by each variable and all the preceding variables.

The openness, distance, and openness x distance (0 x D) categories include
interactions with aspect. Non-signiﬁcant values (in parentheses) indicate that a

variable does not signiﬁcantly increase the explanatory power of the model.

**p<0.001, *p<0.01 .................................................................................................... 39

vii

Table 2.6: P values from ANOVA F-tests. SR=species richness, ST=shade
tolerance, D x A=interaction between distance and aspect. Values in parentheses
are not signiﬁcant ........................................................................................................ 41

Table 2.7: Edge effects (m) determined by comparisons with the interior (100 m)

for each aspect and forest type. The analysis included both native and non—native
species. Gray boxes indicate that metric values were higher at the edge;

signiﬁcant effects in white boxes indicate smaller metric values at the edge.

Effects are signiﬁcant at the p<0.05 level; values in parentheses are signiﬁcant at

the p<0.1 level. -- = no effects detected, ep=exterior ................................................ 42

Table 2.8: Multi-response permutation procedure (MRPP) results for all

understory species and for the most common understory species. Rare (<3
occurrences) and unknown species were excluded. ‘A’ is the chance-corrected
within-group agreement, a measure of effect size. Cluster analysis was only

conducted for the top ﬁve Species data ....................................................................... 43

Table 2.9: Statistical indicator species for each site. All listed species were

signiﬁcant at a=0.05. For most sites, only those species with indicator values of at
least 30% are listed. If fewer than two species reached this threshold, the species

with the next highest indicator value is listed. Non-native species are marked with

an asterisk (*) .............................................................................................................. 46

Table 2.10: Distance-plots (site-distance) included in each cluster. Listed indicator
species were signiﬁcant at a=0.05; % of perfect indication is given in parentheses.
Non-native species are marked with an asterisk (*). Site abbreviations are given in
Table 2.9 .................................... , .................................................................................. 47

Table 3.1: Non-native species found in this study. All species names are
according to Gleason and Cronquist (1991). Nativity was determined using Voss
(1972, 1985,1996). Shade tolerance categorization was based on verbal
descriptions of species occurrence in Voss (1972, 1985,1996) and Gleason and
Cronquist (1991), and on personal observation. Species with low shade tolerance
are found only in open ﬁelds, roadsides, clearings, and similar areas. Species with
mid shade tolerance are found in open woods, wood edges, trails within woods,
and in thickets. Highly shade-tolerant species are found in forest interiors and will
ﬂower under a closed canopy. Shade tolerance ranges were assigned when species
are found in a variety of shade levels. Origin: E=Europe, Ea=Eurasia, A=Asia,
Am=Americas, US=eIsewhere in US. Longevity: A=annual, B=biennial,
P=perennial, S=shrub, T=tree. Lifeforrn: h=herb, s=shrub, t=tree, g=grass. Forest
type indicates where each species was observed in this study (bm=beech-maple;
oh=oak-hickory) .......................................................................................................... 70

Table 3.2: Edge effects (m) as measured by changes in non-native species stem
density and cover. Shade tolerance categories are described in the methods. In

viii

oak-hickory forests, non-native trees were found only along northern and southern
aspects. Edge effects are Signiﬁcant at p=0.05; values in parentheses are
signiﬁcant at p=0.10. -- = no signiﬁcant effect, ep=exterior plots ............................. 72

Table 3.3: AN OVA results (p values ﬁom F-tests) for abundant non-native
species. Unless otherwise speciﬁed, data are for oak-hickory forests only. Values
in parentheses are not signiﬁcant ................................................................................ 79

Table 3.4: Edge effects (m) for abundant species. Unless otherwise speciﬁed, data

are for oak-hickory forests only. Edge effects are signiﬁcant at p=0.05; values in
parentheses are signiﬁcant at p=0.10. -- = no signiﬁcant effect, ep=exterior plots,
bm=beech-maple data, oh=oak-hickory data .............................................................. 80

Table 3.5: Non-native stem density and cover ANOVA results. P values are from
Type III F-tests for signiﬁcant effects of the model terms. Values in parentheses
are not signiﬁcant ........................................................................................................ 83

Table 3.6: Cumulative R2 values from Type 1 ANOVA. R2 values represent the
proportion of variation explained by each variable and all the preceding variables.

The openness, distance, and openness x distance (0 x D) categories include all
interactions with lifeforrn and shade tolerance (ST). Non-signiﬁcant values (in
parentheses) indicate that a variable does not signiﬁcantly increase the explanatory
power of the model. **p<0.001, *p<0.01 .................................................................. 84

Table 4.1: Species found in beech-maple and/or oak-hickory forest seedbanks.

All species names are according to Gleason and Cronquist (1991). Origin was
determined using Voss (1972, 1985,1996): n=native, A=Asia, E=Europe,
Ea=Eurasian, US=eIsewhere in US. Longevity: a=annual, b=biennial,
p=perennial, sh=shrub, t=tree. Growth form: g=grass, h=herb, r=rush, sh=shrub,
se=sedge, t=tree, v=vine. Shade tolerance categorization is based on verbal
descriptions of species occurrence in Voss (1972, 1985, 1996) and Gleason and
Cronquist (1991), and on personal observation. Species with low shade tolerance
are found only in open ﬁelds, roadsides, clearings, and similar areas. Species with
mid shade tolerance are found in open woods, wood edges, trails within woods,
and in thickets. Highly shade-tolerant species are found in forest interiors and can
ﬂower under a closed canopy. Shade tolerance ranges were assigned when species
are found in a variety of shade levels .......................................................................... 105

Table 4.2: Average seed bank density (seeds/m2) and total species richness for
each edge (0-100 m) and site (both sampled edges). bm=beech-maple; oh=oak-
hickory ........................................................................................................................ 108

Table 4.3: Index of dispersion (IOD) for the most common seed bank species.
Signiﬁcant p values indicate that seedlings are more clustered (among plots across
all sites) than expected. Species were also analyzed for clustering within sites;
sites where signiﬁcant clustering was found (a = 0.05) are indicated. HW is 3

ix

beech-maple forest; all other sites are oak-hickory forests. See Table 4.2 for site
abbreviations ...............................................................................................................

Table 4.4: P values from F tests for signiﬁcant effects of distance, aspect, and the
distance x aspect interaction (D x A) for all sites. Values for seed density were
divided into categories based on origin, shade tolerance, and lifeform (deﬁned in
methods). Values in parentheses are not signiﬁcant ..................................................

Table 4.5: Multi-response permutation procedure (MRPP) results for several
grouping variables. Within each edge, values for all samples at a distance were
averaged, creating 60 sample units. Rare (< 3 occurrences), dead, and unknown
species were excluded, leaving 44 species. ‘A’ is the chance-corrected within-
group agreement, a measure of effect size ..................................................................

Table 4.6: ‘A’ values from pairwise MRPP comparisons. Larger values indicate
greater homogeneity within groups (aspects) and greater differences between
groups. Values in parentheses were not signiﬁcantly different from zero.
**Bonferroni-adjusted p<0.001 ..................................................................................

Table 4.7: Sites, distances, and indicator species for the clusters. Percent of
perfect indication is given for each signiﬁcant indicator species (a=0.05). Non-
native species are marked with an asterisk (*) ............................................................

Table 4.8: Indicator species for sites. Only those species with signiﬁcant indicator
values (p < 0.05) are given. Non-native species are marked with an asterisk (*).
‘I'These species were not observed in vegetation under the forest canopy at a site .....

Table 5.1: Species found in Tourney Woodlot. Year of presence was not given for
the 1940-1950 data. x = species present; 0 = species present outside of sample
area or too small to be counted (2004 only) ................................................................

Table 5.2: Chi-square tests for independence of dbh class and year .........................

Table 5.3: Rate of change (per year) of basal area and stem density in Tourney
Woodlot. Total density includes all stems 2 2.5 cm dbh. Large trees are 2 12.7
cm dbh .........................................................................................................................

Table 5.4: Total basal area (mz/ha) for the most common species in Tourney. The
maximum value for each species is in bold ................................................................

Table 5.5: Tree density (trees/ha) by size classes for all years. The maximum
density in a dbh class for each species is in bold ........................................................

Table 5.6: hnportance values (IV) for all study years. hnportance values are the
sum of relative dominance and relative density, and are expressed as a percentage
of the total (200). The highest value for each species is in bold ................................

113

114

116

117

117

118

139

141

I41

143

144

145

Table 5.7: Importance values (IV) for all tree species in Tourney Woodlot in 2004.
Importance values are the sum of relative dominance, relative frequency, and
relative density, and are presented as a percentage of the total (300) .........................

Table 5.8: Index of dispersion (IOD) values for common trees in Toumey. Total
density includes all stems 2 2.5 cm dbh. Large trees are _>_ 12.7 cm dbh. Canopy
trees are not overtopped by any other tree. Values in bold were not signiﬁcantly
different ﬁ'om a random distribution. -- = too few trees to analyze ...........................

Table 5.9: Clusters based on total basal area and relativized basal area. All listed
indicator species were signiﬁcant at a=0.05 ...............................................................

Table 5.10: Comparison of the characteristics of ﬁve old-growth beech-maple
remnants. For all sites, ‘all trees’ include trees 2 2.5 cm dbh. For Hueston Woods,
large trees are 2 25 cm dbh; for Williams Woods, large trees are 2 8.9 cm dbh; for
all other sites, large trees are _>_ 10 cm dbh. Importance values are based on
relative density and relative dominance ......................................................................

Table 5.11: Edge widths in Tourney Woodlot, based on vegetation structure and
composition data from Chapter 2. Total area of the woodlot in 2004 was 11.7 ha...

xi

146

147

148

151

I60

LIST OF FIGURES

Figure 1.1: Vegetation of southern Lower Michigan, circa 1800. Beech-maple
forest (black) is common in the south of the state. Oak-hickory forest (dark gray)
is present in many smaller patches .............................................................................. 10

Figure 1.2: Land cover in southern Lower Michigan in 2000. Most of the forest,
wetlands, and grassland visible in Figure 1.1 has been converted to agriculture.
Deciduous forests, including beech-maple and oak-hickory forests, are present

only as very small fragments ...................................................................................... 1 1

Figure 2.1. Site locations. See Tables 2.1 and 2.2 for site descriptions .................... 22

Figure 2.2: Illustration of sampling grid-transect. Vegetation sampling and
hemispherical photography are shown on separate transects for clarity. The edge

was deﬁned as the average location of the outermost canopy tree trunks. When
necessary, sections of transects were shiﬁed to avoid large gaps (canopy photo
transect). Understory species were sampled in 1 m2 plots located at the corners of

25 m2 vegetation structure plots. Trees were sampled in 5x100 m belt transects ...... 23

Figure 2.3: Herb layer height, cover, and stem density by aspect in beech-maple
and oak-hickory forests. This layer included woody plants <1 m tall ....................... 34

Figure 2.4: Shrub layer height, cover, and stem density by aspect in beech-maple
and oak-hickory forests. This layer included all woody species 1-2 m in height ...... 35

Figure 2.5: Sapling layer height, cover, and stem density by aspect in beech-maple
and oak-hickory forests. This layer included all woody species >2 m tall, but <2.5
cm dbh ......................................................................................................................... 36

Figure 2.6: Canopy basal area, stem density, and leaf area index by aspect in
beech-maple and oak-hickory forests. The canopy layer includes all woody

species >2.5 cm dbh. LAI was measured only in north and south edges ................... 37
Figure 2.7: Canopy openness (%) for beech-maple and oak-hickory forests ............ 38
Figure 2.8: Understory species richness (SR) and Sorensen similarity to the

interior, and canopy tree species richness and shade tolerance (ST). For clarity,

error bars are not shown for canopy data .................................................................... 44

Figure 3.1. Non-native herb stem density (stems/m2) by shade tolerance category
(described in methods) in beech-maple and oak-hickory forests ................................ 74

Figure 3.2. Non-native herb cover (%) by shade tolerance category (described in
methods) in beech-maple and oak-hickory forests ...................................................... 75

xii

Figure 3.3. Non-native shrub and tree stem density (stems/m2) by shade tolerance
category (described in methods) in beech-maple and oak-hickory forests ................. 76

Figure 3.4. Non-native shrub and tree cover (%) by shade tolerance category
(described in methods) in beech-maple and oak-hickory forests ................................ 77

Figure 3.5. Stem density trends for three common non-native species in oak-
hickory forests. Note the changes in scale between species ...................................... 78

Figure 3.6: Shade tolerance of non-native species in oak-hickory forests. Species

with low shade tolerance are found only in open ﬁelds, roadsides, clearings, and
similar areas. Species with mid shade tolerance are found in open woods, wood

edges, trails within woods, and in thickets. Highly shade-tolerant species are

found in forest interiors and will ﬂower under a closed canopy. Shade tolerance

ranges were assigned when species are found in a variety of shade levels. Non-

native trees were found only along north and south edges ......................................... 82

Figure 4.1: Soil samples were collected at six distances from the forest edge. The
location of ﬁve soil samples at each distance was randomly determined. If no

samples fell within 10 m of one side of the sampling area, a sixth, non-random

sample was collected 5 m from the line ...................................................................... 101

Figure 4.2: Species origins for each site and across all sites. Unknowns include all
plants that could not be identiﬁed to the species level. See Table 4.2 for site
abbreviations. oh=oak-hickory, bm=beech-maple ...................................................... 110

Figure 4.3: Lifeforrns of seed bank species. Woody plants include shrubs, trees,

and vines. The monocots category includes grasses, sedges, rushes, and T ypha sp.
Unknown species for which lifeforrn could be determined are included in the
appropriate category. See Table 4.2 for site abbreviations. oh=oak-hickory,
bm=beech-maple ......................................................................................................... 1 l 1

Figure 4.4: Shade tolerance for seed bank species. ‘Low’ includes both the low

and low-mid shade tolerance categories; ‘high’ includes the mid-high and high
categories. Shade tolerance categorization is based on verbal descriptions of

species occurrence in Voss (1972, 1985, 1996) and Gleason and Cronquist (1991),

and on personal observation. Species with low shade tolerance are found only in

open ﬁelds, roadsides, clearings, and similar areas. Species with mid shade

tolerance are found in open woods, wood edges, trails within woods, and in

thickets. Highly shade-tolerant species are found in forest interiors and can ﬂower
under a closed canopy. Unknown species could not be identiﬁed to the species

level. See Table 4.2 for site abbreviations. oh=oak-hickory, bm=beech-maple ........ 112

Figure 4.5: Average seed bank species richness and diversity (H) (per sample) by

aspect for all sites. Samples were 0.5 L (5 cm deep). Bars with the same letter are
not signiﬁcantly different ............................................................................................ 1 15

xiii

Figure 4.6: Seed density (log seeds/m2) by aspect for all sites. Across all aspects,
densities at 5 and 10 m were signiﬁcantly different from interior values. Densities
in north edges were signiﬁcantly higher than those in west edges..................._ .......... 115

Figure 4.7: Species richness per sample (across all sites and aspects) with distance
ﬁ'om the edge. Samples were 0.5 L (5 cm deep). No values were signiﬁcantly
different from that at 100 m ........................................................................................ 116

Figure 4.8: Seed density by distance for lifeforrn and shade tolerance categories.
Density of perennials and woody species did not change with distance. Monocot

seed density decreased with distance (p=0.0098). For annuals, the density at 10 m

was signiﬁcantly higher than values at 100 m. Densities of mid and highly Shade-
tolerant species did not change with distance. For shade-intolerant species, the

density at 10 m was signiﬁcantly higher than values at 100 m. Soil was collected
toadepthofScm ........................................................................................................ 122

Figure 5.1: Tourney Woodlot from 1938 to 2001. All photos are the same scale

(shown in the 2001 photo). Photos for 1938, 1950, 1970, and 1981 were obtained

from the MSU Aerial Photo Archive. The 1961 photo is taken from Schneider

(1963). The 2001 photo was obtained from the Michigan Geographic Data Library
(MiGDL). The permanent pond is noted in the 1961 photo. A large blowdown

(present in 1961) is still visible in the 1970 photo. Planted conifers are pointed out

in the 2001 photo ......................................................................................................... 135

Figure 5.2: Locations of permanent plots in Tourney Woodlot. Each grid square
is 18.3 x 18.3 m (60 x 60 ft). Elevations are given in meters. From Schneider
(1963) .......................................................................................................................... 136

Figure 5.3: Stem density and basal area by species. Total stem density includes all
stems _>_ 2.5 cm dbh. Large trees are 2 12.7 cm dbh ................................................ 142

Figure 5.4: Cluster identity of each plot in Tourney Woodlot. Clustering was
based on total basal area for each species within a plot. Each plot is 18.3 x 18.3 m
(335 m2) ....................................................................................................................... 150

Figure 5.5: Cluster identity of each plot in Tourney Woodlot. Clustering was
based on basal area for each species within a plot, relativized to species maxima.
Each plot is 18.3 x 18.3 m (335 m2) ............................................................................ 150

Figure 5.6: Tree species richness in Tourney Woodlot in 2004. Darker colors
indicate higher richness ............................................................................................... 153

Figure 5.7: Basal area (mZ/ha) of F. grandifolia in Tourney Woodlot. Darker
colors indicate higher basal area ................................................................................. 154

xiv

Figure 5.8: Stem density (trees/ha) of canopy-sized A. saccharum trees in the
study area. Darker colors indicate higher basal area .................................................. 159

Figure 5.9: Core forest in Toumey Woodlot if edge effects penetrate 20 In (light
gray) and 90 m (darker gray) into the forest ............................................................... 159

XV

Chapter 1
Impacts of forest fragmentation and non—native plants on native plant communities

Anthropogenic habitat fragmentation and introduction of non-native species are
two of the greatest threats to biodiversity today (Elton 1958, Hobbs and Huenneke 1992,
Kolar and Lodge 2001). Forests across the North American continent have been greatly
altered by human activity (Harris 1984). Most forests have been logged at least once, and
those that have regrown from clearcuts are often highly fragmented (Harris 1984, Riitters
et al. 2002). Non-native diseases and insects have effectively eliminated several
dominant canopy species while non-native plants change the structure and composition of
forests (Elton 1958, Kloeppel and Abrams 1995, Collier et a1. 2002). Understanding the
interactions between forest fragmentation and non-native species is critical to the
management of existing forests as well as to forest restoration.

Fragmentation, the division of continuous habitat into smaller, more isolated
patches, affects forests through destruction of habitat, creation of edges, and isolation of
populations. In 1997, approximately 165 million hectares of forested land (representing
29% of total land area) existed in the United States (USNRCS 2000). Total forest area
has increased slightly since the early 19003: current forest cover, including plantations, is
approximately 70% of pre-settlement cover (USFS 2001). However, current forests are
more fragmented than they were historically: 62% of forest area in the continental United
States is located within 150 m of an edge (Riitters et a1. 2002).

Edge effects (abiotic and biotic changes at the forest edge) change forest structure
because light availability and wind velocities are greater, and humidity and soil moisture

lower, at the forest edge than in the interior (Murcia 1995, Burke and N01 1998,

Gehlhausen et al. 2000). Edges also concentrate nutrients and pollutants relative to both
open ﬁelds and forest interiors (Weathers et a1. 2001). These altered abiotic conditions
change vegetation structure and species composition (Palik and Murphy 1990, Matlack
1993, Murcia 1995, Gehlhausen et a1. 2000). Canopy trees increase lateral branching
towards open areas in response to increased light availability (Mourelle et a1. 2001).
Canopy and understory stem density increase (Palik and Murphy 1990, Murcia 1995,
Gehlhausen et al. 2000) and less shade-tolerant species in all vegetation layers become
more abundant along edges (Whitney and Runkle 1981, Matlack 1994, Kupfer 1996,
Burke and N01 1998, Honnay et al. 2002, MacQuarrie and Lacroix 2003). Furthermore,
edge effects may reduce recruitment of shade-tolerant species if these species require
conditions characteristic of the forest interior (Tomimatsu and Ohara 2004).

Plant population isolation, another consequence of fragmentation, affects both
dispersal and pollination. In a review on the ecology of forest herbs, Whigham (2004)
noted that many forest herbs spread primarily through clonal growth, while others are
dispersed by ants. Both these mechanisms are limited in dispersal distance, making the
colonization of ﬁ'agments unlikely (Meier et a1. 1995). Reduced pollination in fragments
has two consequences: 1) pollen limitation, leading to reduced seed production; and 2)
genetic isolation, possibly causing inbreeding depression (Aizen et a1. 2002).
Fragmentation may also increase seed predation by eliminating the predators of
granivores. In forests of the Paciﬁc Northwest, for example, Peromyscus maniculatus
(deer mouse) populations increase in response to fragmentation, causing an increase in
Trillium ovatum (Paciﬁc trillium) seed predation (Tallmon et a1. 2003). Similarly, seed

predation in New Jersey was highest along the edge (Meiners and LoGiudice 2003). In

addition to these problems, population sizes of herbs are often decreased during
fragmentation. For these reasons, isolated populations, especially those in small
fragments, are more vulnerable to extirpation (Harris 1984). The number of species a
fragment supports also depends in part on its history. Fragments that are remnants of
more extensive forests will often have more species than fragments generated from
regrowth because the latter lack residual populations of species with short dispersal
distances (Harris 1984). In either case, a fragment will support fewer species than
unfragrnented forest simply because of its decreased area.

Fragmentation has not been the only anthropogenic inﬂuence on forests: the
introduction of non-native species has also caused large changes in forest composition
and structure. Non-native (also called exotic, alien, or introduced) species are those that
were not historically present in an area and have been deliberately or accidentally
introduced through human intervention (Richardson et al. 2000). Some of these species
have become naturalized; that is, they have established self-sustaining populations
(Richardson et al. 2000). A relatively small number of naturalized species become
‘invasive’, a term whose deﬁnition has been debated. Richardson et al. (2000) called a
species invasive if it spreads more than 100 m from the parent population in less than 50
years. Many other authors, however, reserve the term ‘invasive’ for those species that
invade and cause signiﬁcant changes to natural ecosystems (e. g., Colautti and Maclsaac
2004). While the term ‘weed’ has sometimes been applied to invasive species
(Sutherland 2004), it can also apply to unwanted native species and thus should be

avoided. In this thesis, I will use the Richardson et al. (2000) deﬁnition of invasive

species. Furthermore, I will consider any species not found in the pre-settlement ﬂora, as
deﬁned by Voss (1972, 1985, 1992), to be non-native.

Much effort has been devoted to describing attributes of invasive non-native
species. Commonly cited attributes include broad habitat requirements, small seed size,
fast growth rate, short time to maturity, copious seed production, clonal growth, tolerance
of disturbance, seed dispersal by vertebrates, and invasiveness in other regions
(Rejmanek 2000, Kolar and Lodge 2001, Prinzing et a1. 2002). The characteristics of the
invaded habitat are equally important in determining non-native success. Thompson et
al. (1995) found that non-natives were ﬁrnctionally indistinguishable (based on
characteristics such as lifeforrn and phenology) from natives with expanding ranges,
indicating that similar environmental factors may be driving invasion and range
expansion. Non-native species tend to invade habitats similar to those in their native
range (Pysek 1998). On a between-habitat scale, more species-rich habitats have a higher
number of non-natives, possibly because both natives and non-natives respond similarly
to resource availability (Lonsdale 1999, Stohlgren et a1. 2002). Disturbance, especially
when anthropogenic, increases the likelihood of invasion (Forcella and Harvey 1983,
Pyle 1995, Vitousek et al. 1997).

Non-native species are ubiquitous in eastern deciduous forests, although they
often make up only a small percentage of species richness or cover (Table 1.1). A survey
of site ﬂoras and other studies in the Midwest and Northeast found that non-native
species were present at all sites, although not necessarily in all habitat types within a site.
The proportion of non-native species in forests ranged from 0-25%. Non-native

abundance (expressed as cover or stem density), an important ecological measure, is

reported less often than non-native species richness. Spyreas (2004) found that relative
non-native cover was lower in forests (9%) than in wetlands (33%), prairies (36%), and
planted grasslands (76%). In forests, non-native plants can reduce native diversity
(Collier et a1. 2002), suppress tree regeneration (Gorchov and Trisel 2003, Fagan and
Peart 2004), change forest structure (Gould and Gorchov 2000, Collier et al. 2002), and
alter forest phenology by leaﬁng out earlier and retaining leaves later than native species
(Woods 1993, Kloeppel and Abrams 1995, Anderson et al. 1996, McNab and Loﬁis
2002)

Historically, southern Michigan was covered by a mosaic of different forest types,
with beech-maple and oak-hickory communities the most common climax forests (Braun
1950; Figure 1.1). Beech-maple forests (or mesic southern forests, MNFI 1986), which
covered by far the most area, are typically found on more mesic, richer soils (Dickman
and Leefers 2003). These forests tend to have fewer canopy species than other deciduous
forests: while mixed mesophytic forests can have from 10 to 20 species, beech-maple
forests rarely have more than 14 (Braun 1950). This paucity of canopy species could be
due to the relatively young geologic age of the region (Braun 1950) or because both of
the canopy dominants (Acer saccharum, sugar maple, and F agus grandifolia, American
beech) create such a dense canopy that only seedlings of extremely shade-tolerant Species
can survive (Dickman and Leefers 2003).

Oak-hickory forests reach their northern limit in southern Michigan and were
historically represented by large but isolated fragments (Figure 1.1), primarily in dry-
mesic and xeric areas (Braun 1950, Dickman and Leefers 2003). Oak-hickory forests,

also called dry-mesic southern forests (MNFI 1986), have a fairly open canopy and

understory (Braun 1950). The ‘openness’ of oak-hickory forests is based on lower stem
density and basal area (as compared to beech-maple forests, Table 1.2), as well as the
relatively low leaf area index of the dominant Quercus spp. (oaks, Parker and Tibbs
2004). Historically, this openness, and the dominance by Quercus spp., was maintained
by frequent ﬁre (Dickman and Leefers 2003). Modern-day oak-hickory forests have
become more diverse because ﬁre suppression has allowed less ﬁre-tolerant species, such
as Carya spp. (hickories), Prunus serotina (black cherry), and Acer rubrum (red maple),
to increase in abundance (Dickman and Leefers 2003).

Many forests in Michigan today, especially those in the southern Lower Peninsula
are highly fragmented (Figure 1.2). Nearly all of the state’s forests have been destroyed
or highly altered by human activity (Ricketts 1999). This study examined the interaction
between fragmentation and non-native species, as well as long-term changes in, and
forest characteristics of, an old-growth stand using data collected between 1940 and
2004. The objectives were to compare edge effects and invasion by non-natives in two
common forest types, beech-maple and oak-hickory, and to examine the presence of non-
native species in the seedbanks of each forest type. In Chapter 2, the structural and
compositional changes to vegetation in forest edges are quantiﬁed. All study sites are
described in that chapter. Chapter 3 examines the distribution of non-native species in
forest edges. In Chapter 4, the seed bank distribution of non-natives along edges is
described. Chapter 5 details the long-term changes in an old-growth forest fragment.

Finally, conclusions are presented in Chapter 6.

 

89300.3

 

 

G3: 305 I for. w 53222 .: 3:232 8:55 “000205 0:83 .0?
$88 .3 :0 3205 I 3330 m 06:2: 0:805 0532 £5 :00Q 5238: 0308-3?
3008
3.85 .3 00 3205 I $2” M $82: 020005 0.532 we: 309 53.9008 0302
AvooNV Echc? E3 wEEoE I {on o 0030505. 0.35 03% £35 030:0 :5 .3395
Agomv Echo? :3 w:::05 I .xch N 0030505. x35 0.8m 235 03000 _35 :85 32205032802
Avoomv Echo? .23 wEEoE I o\oN _ 0030505. x35 0.8m 235 0.0000 :5 :ccBmmmn-xocE0m
25°C Echo? :3 wEEoE I go o 0030505. 0E5 0.8m £35 0.0000 :5 0.0280:
$83 Echo? :3 wEEoE I gm 2 0030::0... 0:5 0.5m 235 0.00:0 :5 0:300qu3.
0850034
333 _305 I £1. 5 333:2 3:232 833. 30:35 0:23: 3:83 0:5
03:30 6:33—
Amoomv 0:803 E3 059.0032 I <9: 2 :33:m 005.5 00.3 c 328 3:394.
$83 Echo? :3 w:::05 I Ran; m 0030505. 035 030m £35 x0000 ~35 33:53.08 5.00:2
300cm
$.9ch 3:55 23 805892. I $2 on the—03:35 38802 532305 3:35ch: @052
0858.3
$33 305 I $8 3 53322 .: 3:232 8.ch 500535 30:c.« 0:8200Q
$3: 303:5. E3 3:032 I {.ch #0 010% 302 02305 308.5 035 38E E333
$83 Echo? E3 w:::05 I :30 c 000.0050... 0E5 0.8m 0.3"— c_00..O =5 3c¢35m050
Aged Echo? :3 wan—05 I £36 2 00000505. 0.35 03% £35 0.0000 :5 boo—02-030 C0052
39E
2.83 .3 a as“: -- xi m :25: ”tome: 23.2 we: :8: “.53: 28840:
$83 Echo? :3 m:::05 I Ran; m 0030505. .535 03% £35 0.00:0 :5 5$032.30
$508 .3 H0 .305 I £5; N 0:95: 02805 0.532 3:: 309 mach 20:33: 0302
:233030 3:805
59:35 5023:3533 I :32 NN 33:22 .m 888? 50:30.5. 0308:0035
63: E03? :3 :cO— I 3mm 2. ”33:32 .0 302 3.532 3:00 :05 238.3035
3933 3:305 I .xcmm wk. 33:22 .0 302 3532 Ech3m 0338-5000m
Sam: :cwom E3 0:905 I $2 3 33:32 .0 8:08? :003m 0308:003—
._0>cc 0030mm 80:23..
033: 033: 0020050
00:80.35 -:c: .x. -:c: .x. 033.282 :3303 85 80338.5

 

23:33 03: 0:": @0630 3:8 E3 38h 0% he >033

05302: 0 :c :03: .830: 300cm 033200: E0330 05 :_ 0.803.380 he b03> 3 E 830% 03395: he 00:03>05 n: 013,—.

 

 

 

 

 

 

€88 308m :3 -- -- 8.5: 85> 285

$83 .3 s 8: -- $0 _ N z 38:: 0585 222 2:: :8: 220 823085

9.83 828m :2 -- -- .65.: 325, 05:28 035:

$ng Echo? 8:3 w5805 I femdm ow 0080805. 0.35 035 23.5 x005 :5 803 833: .885

$8 : 09.23 as :28: -- :8 M: 0:8» >52 2585 .85: 0.80 82883

803: 380 -- :3 v 50222 a me: 2850.. we:

I o\oc.m _ x8> 302 0.6085 308.5 0.05 :0 .5

€83 303% $2 -- -- :25: 25:? 05:03

08.30034

88: 33. -- $3 2 50322 d 3882 9.83 Bag: 896 .83:

$83 32:8 ea -- -- mesa 85> :28

Ava: 88:5. :8 8:052 I 8v; N 38> 302 02085 3085 035 088.2 983m
880 880% 08:8:
033: 033: 880%

00:08.85 -:o: .x. -:o: .x. 0>ua:.:o2 823004 8% 8035835

 

.8083 2 as“:

 

335 8:35:53 3328 so comnmo
A33 Eu _.m m 805 so 895+
533 Eu :6 N woo: so 383m...

2:98:83 .3 3 38.23383 388m“
333 :5 Wm N 82. co woman.»

 

egg. _ : 53: d

 

©3333: 3533; a £33:
:3: .3 3 33¢ ogqomvav 3&3 .m 5:33:85. 333
. . 88: 32:3.
Aoéw 3% ~53 U can A23: 3932 a. 33509 Agomnmvv Ae\o >C
ﬁg: 2500 335 “SEES .23: >895 a. gem umcbnrugm 3:qu 880% 2:33:35
A32: 5300 a. £853"—
.98: 33 3 5355

:3: .3 B .8225
$3: .3 3 8322::

23.32:: a?
@828: 33:53 oé

.23: .3 3 SEE
:8: 2583.— 3 33:33:.

"2333:: 3.3.?
8823: 3+

530%-:cooom
x09: «23 «no;

 

 

 

:8: .3 8:: 23>on
3 23¢ .38: .3 3 v322 ﬁrm : 3 38m .323: 3:33 3.5-8 ﬁaohwécooom
38: .33
333.3 .23: .3 3 :23; :8: 5932
.38: 322cm 3 5252 3 TN TC: 3 3:2 .38: 58 «$2 £303-20
mmoaaot 860% out 20:30
8%: 2333
:3: .3 3 .33; emoémv 3 38m .333: 353$ 3382.82 ﬁaohwécooom
£8: .33
333.3 .23: .3 3 .23; £8.39: 88:
.38: 333.5 3 525.2 m :l _m 333:: .68: 23:3 932% _ sesamea
$5835 53:2. 88m
83.: 2333
:3: .33 23; o.vm-m.2 3 3.3m .323: .3338 30.8-3.9“ :EBwécooom
£8: .33
2923 .28: .3 3 23¢ $5333-92 88: 2333: .38.: a3: N. _ m- _ .wm 53333-20

A958: 323 _ammm

 

 

mooSow

b33530

 

moousow

oEmELEBm

 

 

328588 33:28 was 53:33 33:28 no woman 23 mos—9» oo§€0-E~ A33 :8 o N 805
so 333 03.. mos—9» :3 .322“ ommEoﬁo 32:3 3328 @833-an Ea 2938-589 :0 8532835 no acmtaaaoo ”N; 053.

 

 

Legend N

- Beech-Maple
- Oak-hickory

Deciduous/Mixed Forest

7//// Coniferous forest 0 25 50 100 Kilometers
Grassland/Savannah
8888 Wetlands Map created by Charlotte Reemts using data from MiGDL

Figure 3.1: Vegetation of southern Lower Michigan, circa 1800. Beech-maple forest
(black) is common in the south of the state. Oak-hickory forest (dark gray) is present in
many smaller patches.

 

Legend
- Urban A -

 

|:| Agriculture 0 25 50 ﬂameters
L 1 I l l 1 J 1 I
- Deciduous/Mixed Forest
Other Map created by Charlotte Reemts using data from MiGDL, 2004

Figure 1.4: Land cover in southern Lower Michigan in 2000. Most of the forest,
wetlands, and grassland visible in Figure 1.1 has been converted to agriculture.
Deciduous forests, including beech-maple and oak-hickory forests, are present only as
very small fragments.

Anatomy of a fragment: Edge effects i1('t3 bzngiirzrtaple and oak-hickory forest fragments
Introduction

The detrimental effects of fragmentation on forests include outright destruction of
forest, isolation of populations, and the creation of edge effects (abiotic and biotic
changes along the forest edge). Fragmentation also increases the proximity of seed
sources for non-native species, allowing these species to invade the fragments
(Cadenasso and Pickett 2001). Because fragmentation increases the perimeterzarea ratio
of a forest ﬁagment, edge effects inﬂuence a greater proportion of contemporary forests
than they did in pre-settlement forests. Furthermore, the sharp boundaries between
forests and adjacent habitats allow edge effects to penetrate deep into forest fragments;
natural boundaries are often more diffuse, ameliorating the impact of edge effects.

Edge effects have been studied in many different forest types, ranging from
tropical rain forests (Bidham and Lawton 1999, Vasconcelos and Luizao 2004) and
tropical dry forests (Asbjomsen et a1. 2004) to temperate deciduous forests (Palik and
Murphy 1990, Burke and N01 1998) and coniferous forests (Chen et a1. 1992, Rheault et
al. 2003). In all forest types, abiotic differences between edge and interior conditions
include greater light intensity and higher wind speeds at the edge (Burke and N01 1998,
Gehlhausen et a1. 2000). These factors combine to increase air and soil temperature as
well as decrease air and soil moisture (Burke and N01 1998, Gehlhausen et a1. 2000).
More open conditions also allow greater heat loss at night, leading to greater diurnal

ﬂuctuations in temperature at the edge (Murcia 1995).

12

The depth of edge penetration depends on the orientation of the ﬁagment. In the
northern hemisphere, edges facing south or west receive more solar radiation during the
warmest part of the day than edges facing north or east; consequently, the area inﬂuenced
by edge effects is wider there. Matlack (1993) studied ten oak-chestnut ﬁ'agments in
Pennsylvania and Delaware. On average, edge effects (based on differences in light,
temperature, and humidity) extended farther into south-facing edges than into north-
facing ones. East and west edges were similar to south edges. Other authors (Ranney et
al. 1981, Brothers and Spingam 1992) have suggested that east-facing edges are more
similar to north-facing edges. Cadenasso et a1. (1997) compared a north-east edge and a
north-west facing edge in two oak-dominated ﬁ'agments in New York. Edge effects
(based on photosynthetically active radiation (PAR), maximum air temperature, relative
humidity, and soil temperature (10 cm depth)) consistently penetrated farther into the
north-west facing edge. Soil temperature at a depth of 10 cm showed the greatest edge
response (15 m), while temperature at solar noon showed the smallest (1 m). Burke and
N01 (1998), working in upland deciduous forests in Ontario, Canada, found no changes in
humidity or air temperature in east-facing edges, but light availability and soil
temperature decreased sharply beyond the immediate edge (5 m). Soil moisture and
relative humidity in the south and west edges of two mixed-mesophytic forest fragments
in Illinois differed signiﬁcantly from interior conditions as far as 40-80 m into the forest
(Gehlhausen et al. 2000). In contrast, edge effects in north and east edges for these two
factors were either not present or extended only 15 m into the forest. Edge inﬂuence on
soil moisture and humidity penetrated deeper into the forest than changes in canopy

openness.

l3

Vegetation at the forest edge responds to abiotic edge effects in two ways:
through changes in structure and through changes in species composition. Vegetation
responses are usually more extensive along the warm edges, where abiotic edge effects
penetrate deeper into the forest. Structural response to edge creation can include the
growth of shrubs and saplings in the understory, which results in increased stem density
at the edge (Ranney et al. 1981, Murcia 1995, Gehlhausen et a1. 2000). Matlack (1993)
found increased shrub cover as far as 40 m into south-, east-, and north-facing edges.
Canopy density and basal area are also inﬂuenced by edge effects. In an old-growth
forest in Michigan, canopy stem density was higher up to 10 m along a northern edge and
up to 15 m along a southern edge (Palik and Murphy 1990). Such changes in vegetation
structure at the edge may be greater than changes due to forest age. Whitney and Runkle
(1981) compared an old-growth beech-maple fragment to a nearby second-growth
fragment (~100 years). While stand age was a signiﬁcant source of variation in basal
area, plot position (edge vs. interior) was more important. Canopy structure may also be
affected if increased exposure to wind leads to higher wind throw mortality (Murcia
1995). Furthermore, canopy trees often increase lateral branching towards the edge,
enhancing light capture. In general, shade-tolerant tree species exhibit greater lateral
branching and increase foliage biomass more than shade-intolerant species. Acer
saccharum (sugar maple) and F agus grandifolia (American beech) trees at a fragment
edge reduced open ﬁeld light levels by 96%, while less shade-tolerant species like
Quercus rubra (red oak) and Populus grandidentata (bigtooth aspen) reduced light levels

by 85% (Mourelle et al. 2001).

14

Increased light capture by tree crowns and by the denser understory shields the
forest interior from edge effects. This protective effect is‘more evident along older edges
where sufﬁcient time has passed for the vegetation to respond (Brothers and Spingam
1992). Matlack (1993) showed that a closed (older) side-wall of vegetation eliminated
light gradients along two southern edges. Litter moisture and humidity, however,
continued to change signiﬁcantly with distance into the forest. Denser canopies also
reduce wind speeds. Compared to intact controls, signiﬁcantly more wind-dispersed
seeds were found inside experimentally thinned edges, indicating that intact edges
reduced wind speeds (Cadenasso and Pickett 2001).

Species composition at the forest edge differs from the interior because some
shade-intolerant species can survive there, while some interior species are excluded either
by competition or by the altered abiotic regime (Ranney et a1. 1981, Palik and Murphy
1990). All layers of vegetation—ﬁom understory herbs to canopy trees—are affected. In
the understory, old-ﬁeld and other heliophytic species may penetrate a short distance into
the forest (Burke and N01 1998, Gehlhausen et al. 2000). Changes in community
composition extended up to 20-23 m into the forest along southern edges in ancient
(>220 yrs) deciduous forest fragments in Belgium (Honnay et al. 2002). Matlack (1994)
found that all edge-favoring herb and shrub species were also found as far as 40 m (the
greatest distance studied) into oak-chestnut forests. However, north-facing edges could
not be distinguished from other edges based on species composition. In upland hardwood
forests on Prince Edward Island, the interior community composition was not reached
until more than 120 m from the edge (MacQuarrie and Lacroix 2003). The authors

attribute this unusually large edge effect to the forests’ history of disturbance.

15

The response of canopy trees is similar to that of the understory: shade-intolerant
species have higher importance values close to the edge (Whitney and Runkle 1981 ,
Kupfer 1996). This response is usually more pronounced along warm edges. Palik and
Murphy (1990), for example, found that less shade-tolerant canopy species like Q. rubra
and F raxz'nus amerz‘cana (white ash) were found only within 20 m of the southern edge of
an old-growth beech-maple forest. In contrast, A. saccharum and F. grandifolia were
found throughout the northern edge, although their importance values declined somewhat
within 10 m of the edge. Edge effects can also inﬂuence tree recruitment in gaps,
because edge species can serve as seed sources for the gaps. Heliophytic species tend to
colonize gaps that are close to an edge, while gaps in the interior favor more shade-
tolerant species (Kupfer et al. 1997).

Vegetation response to edge effects is somewhat similar to the response to gap
formation. Aﬁer gap formation, shrub and herbaceous cover increases rapidly in
response to higher light availability (Ehrenfeld 1980, Huenneke 1983, Moore and Vankat
1986). In some cases, this increase in cover comes from growth by species already
present (Ehrenfeld 1980, Moore and Vankat 1986, Goldblum 1997); in others, species
composition changes (Huenneke 1983). Compositional changes appear to increase with
gap size for both herbs and shrubs, and are driven by colonization or germination of
shade-intolerant species. Anderson and Leopold (2002) found increased herbaceous
species richness in medium and large, but not small, gaps in a species-rich conifer
swamp. Larger gaps in wetland forests were colonized by three shade-intolerant shrubs,
while small gaps were dominated by species from the surrounding forest interior

(Huenneke 1983). In a northern hardwoods forest, all experimental gaps were colonized

16

by shade-intolerant Rubus spp. within three years, but little other compositional or
structural change was observed (Collins and Pickett 1988). Even when composition does
not change, there is usually a shift in dominance. In both large and small gaps caused by
gypsy-moths, cover of a common understory shrub (Comusflorida, ﬂowering dogwood)
increased greatly (Ehrenfeld 1980).

One important difference between gaps and edges is that edges are usually
permanent. As the canopy closes over a gap (by branch extension or growth of new
canopy-size individuals), light availability declines, and vegetation structure and
composition eventually return to interior levels (Gysel 1951, Moore and Vankat 1986).
In contrast, most edges are permanent, maintaining the vegetation in an early gap
successional state (Matlack 1994). Edges are often adjacent to seed sources for shade-
intolerant, ruderal species, increasing the probability of colonization by these species
(Brothers and Spingam 1992).

Many animal species respond to edge effects. Yellow-faced honeyeaters
(Lichenostomus chrysops) have higher nest success close to a forest edge (Boulton and
Clarke 2003), while many other species, especially ground-nesting birds, face higher nest
predation close to the edge (Flaspohler et al. 2001). Nest success for hermit thrushes
(Catharus guttatus) and ovenbirds (Seiurus aurocapillus) is lower within 300 m of the
edge; decreases in nest density extend even farther (Flaspohler et al. 2001). Black-
capped chickadees (Poecile atricapillus) avoid hoarding food close to forest edge,
indicating that edges are perceived as low quality habitat (Brotons et al. 2001). Black-rat
snakes (Elaphe obsoleta obsoleta), common nest predators in temperate deciduous

forests, prefer forest edges for thermoregulatory reasons, increasing the probability of

17

contact between snakes and nesting birds along edges (Blouin-Demers and Weatherhead
2001). Van Wilgenburg et al. (2001) found that leaf-litter arthropods were less abundant
near edges. This decrease could be due to increased predation or to the changed abiotic
environment. Salarnanders avoid the warm, dry environment of forest edges (Young and
Yahner 2003).

Much of the research on edge effects has considered only a few sites in a single
forest type (Whitney and Runkle 1981, Palik and Murphy 1990, Kupfer 1996, Cadenasso
et al. 1997, Gehlhausen et al. 2000). Forest types differ in their canopy structure. Open-
canopy forests may be less inﬂuenced by edge effects, since the difference between the
edge and the interior will be smaller. I chose to compare two forests types, both common
in southern Michigan. Beech-maple (or mesic southern) forests typically have a dense
canopy and a diverse understory of spring ephemerals (MNFI 1986). Oak-hickory (or
dry-mesic southern) forests, on the other hand, have lower stem densities and sparser
canopies (Table 1.2). This study addressed the following questions:

0 How far do edge effects, measured by changes in vegetation structure and
composition, extend into beech-maple and oak-hickory forest fragments?

0 Which measures of vegetation response (i.e., structure or species composition) are
most inﬂuenced by edge effects?

0 Does depth of edge penetration depend on edge aspect?

0 What understory species are characteristic of edges and interiors in both forest
types?

0 Are differences between the edges and interiors of beech-maple fragments more

pronounced than those of oak-hickory fragments?

18

Table 2.1: Dominant canopy species for each site, as determined by basal area and
importance values (IV). Importance values were calculated as the sum of relative
dominance, relative density, and relative frequency, and are presented as the
proportion of the total (3 00). Beech-maple sites were dominated by A. saccharum and
F. grandifolia. Sites where Q. alba, Q. rubra, or Q. velutina were among the canopy
dominants, and Carya spp. were present, were classiﬁed as oak-hickory. Data are

 

 

 

 

 

 

 

 

 

 

 

 

 

based on 100 m2 plots located >100 m from the closest ed we.
Site Species Dominance IV Basal area Forest
(% BA) (%) (mz/ha) tune

Hudson Acer saccharum 43 .4 68.3 38.0 Beech-

Woodland T ilia americana 12.3 18.0 maple

(HW) Fagus grandifolia 9.3 18.9

Ionia (10) Acer saccharum 60.4 67.2 38.4 Beech-
F agus grandifolia 28.2 45.6 maple

Poff Woods Fagus grandifolia 36.2 37.1 31.7 Beech-

(PW) Prunus serotina 21.9 20.7 maple
Acer saccharum 17. 1 42.3

Tourney Acer saccharum 63 .3 79.2 35.0 Beech-

Woodlot (TO) Fagus grandifolia 28.0 37.6 maple

Aman Park Quercus rubra 25.1 31.2 49.1 Oak-

(AP) Quercus alba 15.8 22.4 hickory
Acer rubrum 14.2 39.5

Bald Mountain Quercus velutina 41.0 23.9 300' Oak-

(BM)“ Carya ovata 18.5 25.7 hickory
F raxinus pennsylvanica 15.7 24.8

Clear Lake Quercus alba 54.9 48.5 36.3 Oak-

(CL) Acer rubrum 26.2 44.3 hickory

Johnson Park Quercus rubra 29.0 25.3 46.6 * Oak-

(JP) Acer rubrum 21.6 31.8 hickory
Quercus velutina 19.0 15.2

Rose Lake Carya glabra 32.6 29.4 35.1 Oak-

(RL) Quercus alba 21.9 18.4 hickory
Quercus velutina 18.2 13.5

Seven Lakes Acer rubrum 27.1 36.2 28.5 Oak-

(SL) Carya glabra 19.3 16.3 hickory
Quercus alba 17.9 15.8

 

 

 

 

 

 

 

*Forest type for Bald Mountain was determined from transects along the edge,
because the canopy composition in the interior, which was topographically lower than
the edge, differed greatly ﬁ'om the edge canopy composition.

+Basa1 area was determined based on interior plots.

19

 

 

 

 

 

 

23me m m £838: ~E£
E .0 n4 _ c4 23me a Z 36:03:: SEES 8:585 3:38.
Ed 95 5: 22m Bo m m in: 83m :82me 23 mean— :o>om
ode 2. 3 see a m
3o 2.” E 22.: 20 e 2 as use» sea 865 £3 8.3 so:
:6 c.2 94 380 a Z 33:: :83 EB» 38>? to:
20: 2o .32 e 3
Ed 34 2. as: e m :8 358 a3. 23 nae 82%.
:.o 3% ﬁx Bow 2: m m 8:: $38.82 8:6 3:2 3:: 2:2
380 a 3 5838: :5:
“53:83
E .o n. b _ . _ 83me a m 37:02:: Easwﬁ :83:
380 a m
ood man we :3: duo: : Z :2: 25% 2:5 60$ 38 8:3 .820
33
mod msmm QB 3:: 66¢ Bo : Z 83 8:882 88m :52er 58:32 2mm
Bow 20 m 3
8o 3: 02 22.: 20 m 2 use be $an 23 as: =22
egv e5 :8: as ea: 25 Bass
833%me 82 8:258: 8088.4. owwm mowvm 8: 85 .9550 2mm

 

 

oEuEéoooana .boxoﬁiwonno @5582 2mm 8.: _.N Emma
mom .SDOSZV 88:: Eton 3&6 Set 35:58: 0.83 .5885: :5 83 .owwo 05 mo 8 o3». :23? em: :5: 05 8:58:
3: :52 808.84: .oo:o:o.t85 :55? mEom 20 808.8: 05 85 m5>o€ 8m 3 meme»: wﬁmmooosm .859:sz @538

.8 .mp8: .805.“ .3 3:532: mm .828: :65 a Ev 30:5qu we accrued: v.8 mowvm .32358: Seaman “Nd 035.

20

Methods
Site descriptions

Potentially suitable beech-maple and oak-hickory fragments were identiﬁed using
the Michigan Natural Features Inventory (MNFI) database of exemplary sites, augmented
by personal observation. Site locations are shown in Figure 2.1. To determine the forest
type of each site, the forest interior (>100 m from the edge) was sampled using randomly
located, circular 100 m2 plots. The number of plots at each site was determined by
creating a species-area curve for tree species. In each plot, the species, diameter at breast
height (dbh), and height category (understory, subcanopy, or canopy) was recorded for
each tree. Understory trees were considered to be greater than 2.5 cm dbh but less than 3
m tall. Subcanopy trees were greater than 3 m tall but were still over—topped by another
tree. Canopy trees were not over-topped by any other tree. Forests were classiﬁed
according to the species that contributed the most basal area (Table 2.1). Beech-maple
sites were dominated by A. saccharum and F. grandifolia. Sites where Quercus alba, Q.
rubra, or Q. velutina were among the canopy dominants, and Catya spp. were present,
were classiﬁed as oak-hickory. For each forest, the importance value (IV) of each
canopy species was calculated as the sum of relative dominance, relative density, and
relative frequency; IV values are presented as the percent of total (300, Mche and
Grace 2002).

I selected relatively undisturbed forest fragments with sharp edges (clear
boundaries at the forest edge) surrounded primarily by crops, pasture, old ﬁelds, or
mowed grass (Table 2.2). Most edges were permanent, but in ﬁve cases the forest edge

was invading an old ﬁeld. A total of 10 sites (4 beech-maple and 6 oak-hickory) and 16

21

edges (6 north, 2 east, 5 south, and 3 west) were sampled. Initially, ‘cool’ (north and
east) and ‘warm’ (south and west) edges were to be grouped. However, because of the

differences within these groups, each aspect was analyzed separately.

. Bald
Lakes

 
          
  
 

 

Ionia

 

 

 

 

 

 

 

Rose.

1

 

Tourney seven

p¢ff Hudson

. l

O 20 4O 80 Kilometers

ltrrlrirl

 

 

 

   

 

 

 

 

 

 

 

 

Figure 2.1. Site locations. See Tables 2.1 and 2.2 for site descriptions.
Vegetation structure and composition

Vegetation data were collected during the summers (June to August) of 2003 and
2004 in grid-transects of 5x5 m plots. Grid-transects extended from 5 m outside the
fragment edge to 100 m into the fragnent (Figure 2.2). The forest edge was deﬁned as
the average line of the outermost trunks of canopy trees; the canopy dripline was usually
located outside of this edge. The plots that extended beyond the forest edge into the
adjacent habitat will be referred to as ‘exterior plots’ or ‘ep’ (Figure 2.2). When the edge
of mowed vegetation was less than 5 m from the edge, the exterior plots extended up to
the edge of the unmowed vegetation. In one case (Seven Lakes), a stream precluded
sampling beyond 50 111 into the forest. Transect width ranged from 15 m to 25 m, with

most being 25 m wide.

22

Grid-transect locations were randomly determined, but were at least 50 m away
ﬁ'om another edge. If necessary, transect locations were shifted to avoid large canopy
gaps. If several large canopy gaps were present, sections of the transect were shifted
perpendicular to its length to avoid them (Figure 2.2). Where the edge was long enough,

more than one transect was sampled along a single edge.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 \Edge
6% 3 :3 Understory Canopy Canopy
. k Specres Gap Photos
“ng0“ C" .4. m r“: a. ,, Composition
SW9" ‘ “ “ Sampling
Sampling .. ... .. ,, (lmxlm)
(SmxSm) - ;_, .. H
r i; "L Q j
‘ Canopy
Tree "1 Gap
Sampling \\~
(5m x 100m)

 

 

 

 

 

 

 

 

Figure 2.2: Illustration of sampling grid-transect. Vegetation sampling and hemispherical
photography are shown on separate transects for clarity. The edge was deﬁned as the
average location of the outermost canopy tree trunks. When necessary, sections of
transects were shifted to avoid large gaps (canopy photo transect). Understory species
were sampled in l m2 plots located at the corners of 25 m2 vegetation structure plots.
Trees were sampled in 5x100 m belt transects.

Measurements of vegetation structure in each 25 m2 plot included the height of
the herbaceous (<~1 m) layer, the shrub layer and the sapling (<2.5 cm dbh) layer,
percent cover of each layer, and stem density of shrubs and saplings. Woody plants <1 m

tall (e.g., A. saccharum seedlings) were included in the herbaceous layer. In 1x1 m

subplots, I recorded all species present, the number of species present, and the stem

23

density of the herbaceous layer. The ﬁve most common species were ranked based on
cover. These subplots were located at the corners of 25 m2 plots at lO-m intervals
(starting at 0 m), with additional plots at 5 m (Figure 2.2). The number of subplots at
each distance from the edge ranged from 4 to 6, depending on the width of the grid-
transect.

Trees were sampled in at least two 5 x 100 m belt transects, nested within the
grid-transect (Figure 2.2). For each tree, I recorded the species, dbh, height category
(described above), and distance from the edge. Tree species were assigned a shade
tolerance rank based on verbal descriptions in Barnes and Wagner (1981): 5=high1y
shade-tolerant, 4=shade-tolerant, 3=intermediate, 2=shade-intolerant, and l=very shade-
intolerant.

Leaf area index (LAI)

In a deciduous forest, LA] (the total one-sided area of leaves above a unit of
ground area) can be measured by collecting leaf litter after leaf fall in the autumn
(Murphy et al. 1974, J onckheere et a1. 2004). While it is extremely labor-intensive, this
is the only method that measures species-speciﬁc contribution to overall LAI (Bréda
2003). Due to time restrictions, LAI was measured at only two sites, one beech-maple
forest (Tourney Woodlot) and one oak-hickory forest (Rose Lake).

Leaf litter was collected from 0.25 m2 circular plots in fall 2003. Five replicate
plots were located at each of the following distances from the edge: 5, 10, 20, 30, 40, 50,
75, and 100 m. Plots were initially arranged in a grid, but were shifted as necessary to
avoid depressions, large rocks, tree trunks, logs, and any other feature that could

inﬂuence the amount of leaf litter on the ground. The grids were located on relatively

24

level ground. Leaves were air dried and each sample was sorted by species and weighed.
A leaf areazmass relationship was established for each species by measuring the leaf area
of subsamples of leaves with an LI-3100C Area Meter (Li-Cor Environmental, Lincoln,
Nebraska) and then weighing them. These relationships were used to estimate the leaf
area of the full samples.
Canopy openness

Canopy openness (the percent of open sky seen from beneath the forest canopy)
was measured using hemispherical photos of the forest canopy. This method has become
more common with the wide-spread availability of digital cameras and software that can
easily analyze digital images (Bréda 2003). It is inexpensive and reliable, and has the
advantage of preserving a permanent record of canopy conditions (J onckheere et a1.
2004)

Photos were taken in August and September of 2004, before leaf fall, using a
Nikon CoolPix 990 with a FC-E8 ﬁsh-eye converter. Photos were taken at 10 m
increments from the edge (0 m), with additional photos at 5 m. At each distance, three
photos were evenly spaced across the grid-transect (Figure 2.2). Fifteen additional
photos were randomly located within the transect. In transects less than 25 m wide, two
photos were taken at each distance, with 10-20 randomly determined additional locations.
The camera was placed 1 m above the ground (above most understory vegetation) on a
leveled tripod and aligned with magnetic north, regardless of the orientation of the edge.
Pictures were analyzed using Gap Light Analyzer (GLA) 2.0 (Frazer et a1. 1999) to

calculate canopy openness.

25

Statistical Analyses

Tree basal area was log transformed before analysis to improve normality.
Sorensen’s index of similarity was used to compare 1 m2 plots to the forest interior (100
m plots) using presence/absence data for all 1 m2 plots at a single distance. I used PROC
MIXED in SAS (SAS Institute 1999) to test for signiﬁcant effects of distance, aspect, and
the distance x aspect interaction on the following dependent variables for each forest
type: height, cover, and stem density of the herb, shrub, and sapling layers; understory
species richness; percent similarity to the interior (Sorensen); canopy basal area and stem
density; LAI; and canopy openness. Multiple transects along an edge and multiple plots
within a transect allowed statistical analysis, even when only one edge within an aspect
was sampled. An appropriate covariance structure accounted for variation within a
sampling grid. A t—test on the least squares means (lsmeans) was used to compare
distances within aspects to the interior value (100 m). Depth of edge inﬂuence (DEI) was
deﬁned as the area within which variables differed consistently from interior values.
When both the interaction term and main effect of aspect were not signiﬁcant, DEI was
determined using a Tukey-Kramer adjustment. When appropriate, the lsmeans of
distance and aspect as main effects were also analyzed. A signiﬁcance level of a=0.05
was used.

A Type I ANOVA in SAS was used to examine the relative importance of aspect,
canopy openness, and distance (and all interactions among them) as explanatory variables
for all understory structural variables by aspect within forest type. For this analysis,
canopy openness was interpolated between sampled points using block kriging in R 1.8.1

(R Foundation for Statistical Computing 2003). The order of variables was ﬁxed as listed

26

above, because Type I ANOVA is sensitive to the order of model terms. Cumulative R2
values were calculated from the sum of squares for model terms. Because a simpler
covariance structure was used for this model, a stricter criterion for signiﬁcance was
selected (a=0.01).

The multi-response permutation procedure (MRPP) in PC-ORD (McCune and
Mefford 1999) was used to compare understory species composition. MRPP is a
nonparametric procedure that tests a hypothesis of no difference between a priori groups
(McCune and Grace 2002). A weighted mean within-group distance (delta) is calculated
for each group (this analysis used a Sorensen distance measure). These mean distances
are used to calculate a test statistic T, which is

T = (5 - ma) / 36
where m5 and 55 are the expected mean and standard deviation of delta, given the null
hypothesis of no difference. A Pearson type III distribution is used to determine a p
value. A measure of effect size (A), which is independent of sample size, describes
within-group homogeneity:

A = 1 — (6 / mg)

When all items within groups are identical, A=1; if items are distributed as expected by
chance, A=0. When A<0, there is more heterogeneity within groups than expected by
chance. In community analyses, A is usually less than 0.1 (McCune and Grace 2002).

For the MRPP analysis, understory species composition was compared using
presence/absence data for all understory species recorded in l-m2 plots. To decrease the
number of rare species, data for the six plots at each distance were combined (referred to

hereafter as ‘distance-plots’). Unknown species and those species present in fewer than

27

three distance-plots were excluded ﬁom the analysis, leaving 121 species. Groups based
on forest type, site, aspect, distance, and grouped distances (in pairs: 0 and 5, 10 and 20,
etc.) within forest types were compared. All groups had at least 15 members. The same
analysis was conducted on data for the ﬁve most common species in each plot (105
species total). Again, data for the six plots at each distance were combined, but percent
frequency was calculated instead of using presence/absence.

For the top ﬁve species data, site groups and forest-distance groups were
investigated further using indicator species analysis (Dufrene and Legendre 1997).
Indicator values are calculated based on the proportional abundance and frequency of a
species within a group:

1ij = 100 (RAkj x RF kj)
where RA is the relative abundance and RF the relative frequency for a species j in each
group k. A perfect indicator species (IV =1 00) would be highly abundant in every plot of
a group, and be present in no other groups. The percent of perfect indication (or indicator
value) is thus a statistical measure of how often a species appears in a particular
grouping. The statistical signiﬁcance of IV kj can be evaluated using a Monte Carlo
technique. Species occurrences are randomly shufﬂed among groups and IV max is
calculated for each run. The p value is the proportion of times that the shufﬂed IVmax is
equal to or exceeds the observed value. In this analysis, sampling units were reassigned
1000 times to calculate p values.

The clustering procedure in PC-ORD was used to investigate relationships among
the distance-plots for the top ﬁve species data. A ﬂexible beta (B: -0.25) linkage method

and Sorensen distance were used to cluster distance-plots. Ten cluster groups were

28

chosen, because this level maximized the number of species with signiﬁcant indicator
values and minimized the average p value for all indicator species (Dufrene and Legendre

(1997). All groups consisted of at least 12 distance-plots.

29

Results
Vegetation structure
Depth of edge inﬂuenchED: Herb layer

Distance signiﬁcantly inﬂuenced all three measures of vegetation structure
(height, cover, and stem density) in both forest types (Table 2.3). In beech-maple forests,
herb layer height, cover, and stem density were usually highest at the forest edge (Figure
2.3). The only decreases were in cover on west edges and stem density on north edges.
While most of the edge effects were small (55 m), cover in east edges and stem density in
north edges showed larger effects (35 m and 40 m, respectively, Table 2.4).

In oak-hickory forests, the herb layer was taller at the edge, but the direction of
change for cover and stem density varied with aspect (Figure 2.3). Herb cover at the
edge showed no response to edge effects on north and west edges, increased on south
edges, and decreased on east edges (Table 2.4). Herb stem density was higher on south
and west edges, and lower on north and east edges. The unusually high interior values
for cover and stem density in east edges were caused by the invasive herb V. minor.
Even when excluding east edges, however, average DEI was greater in oak-hickory (l 1.3
m) than in beech-maple forests (4.6 m, Table 2.4).

DEI: Shrub lager

Distance signiﬁcantly inﬂuenced shrub layer structure in both forest types (Table
2.3). In beech-maple forests, shrub height was lower along east and south edges, and
higher along north edges (Figure 2.4), while cover was lower along east edges (Table
2.4). In oak-hickory forests, shrubs were generally more abundant close to the edge

(Figure 2.4), but relatively few signiﬁcant edge effects were found (Table 2.4). For

30

height, the only signiﬁcant edge effect was an increase on north edges. Shrub cover and

stem density were higher on north and south edges, but those effects were limited in

extent (£10 m).

Table 2.3: P values (from ANOVA F-tests) for vegetation structure. Analyses included
both native and non-native species. Stem Dens.=stem density, LAI=leaf area index, D x
A: interaction between distance and aspect. Values in parentheses are not signiﬁcant.

 

 

 

 

 

 

 

 

Metric Beech-Maple Oak-Hickory

Distance Aspect D x A Distance Aspect D x A
Herb Height <0.001 0.009 <0.001 <0.001 (0.17) <0.001
layer Cover <0.001 0.007 <0.001 <0.001 0.029 <0.001
Stem Dens. <0.001 0.002 0.003 <0.001 <0.001 <0.001

Shrub Height <0.001 (0.48) <0.001 0.001 (0.79) 0.002
layer Cover 0.021 0.0004 0.006 0.003 (0.47) <0.001
Stem Dens. 0.010 0.0002 0.024 <0.001 (0.75) <0.001

Sapling Height 0.011 0.0001 0.0003 (0.28) .002 0.046
layer Cover 0.0003 0.0001 <0.001 <0.001 <0.001 <0.001
Stem Dens. 0.002 <0.001 <0.001 0.001 0.090 <0.001

Canopy Basal area 0.002 (0.92) 0.023 <0.001 0.010 (0.40)
Stem Dens. 0.031 (0.67) (0.051) <0.001 0.002 (0.61)

LAI (0.17) <0.001 0.002 (0.29) (0.21) (0.12)

Openness <0.001 (0.059) 0.005 <0.001 0.031 (0.18)

 

 

DEI: Saplinglaver

Distance signiﬁcantly inﬂuenced all three measures of sapling layer structure in
beech-maple forests (Table 2.3); in oak-hickory forests, sapling height was not
signiﬁcantly inﬂuenced by distance. Sapling height, stem density, and cover in beech-
maple forests were lower at the edge for north-, east-, and south-facing edges (Figure
2.5). Cover and stem density, however, were higher along west edges. Edge effects were
small in north and east edges (ep), but much larger in south and west edges (40-65 m).
The only signiﬁcant edge effect in oak-hickory forests was observed on east edges, where
cover was signiﬁcantly higher at the edge (Table 2.4).

DEI: Canopy layer

31

The main effect of distance signiﬁcantly inﬂuenced basal area, stem density, and
openness of the canopy layer in both forest types (Table 2.3). While the interaction
between distance and aspect for LAI was signiﬁcant in beech-maple forests, it was not
signiﬁcant in oak-hickory forests. In beech-maple forests, basal area and stem density
were higher at the edge (Figure 2.6), but these differences were signiﬁcant only in east
and south edges (Table 2.4). LAl was signiﬁcantly higher in north edges (5 m) and lower
in south edges (30 m, Table 2.4). Edge effects based on canopy openness extended 10-20
m into the forest (Table 2.4).

In oak-hickory forests, the increases in basal area and stem density were more
pronounced (Figure 2.6) and were almost always signiﬁcant (Table 2.4). However, edge
effects were very small (0 m), except for stem density along south edges (20 m, Table
2.4). No signiﬁcant edge effects were found for LAI. Edge effects for canopy openness
were small (5-15 m), and no signiﬁcant effect was found for east edges.

Interior (75-100 m) LAI was signiﬁcantly higher in beech-maple forests than in
oak-hickory forests (6.52i0.13 mZ/m;2 vs. 4.45i1.16 mz/mz, t-test, p=0.01). In contrast,
interior (90-100 m) canopy openness, as measured by hemispherical photography, in the
two forest types was very similar. Although beech-maple forests had signiﬁcantly higher
Openness values (2.93zt1.23% vs. 2.67i1.30%, t-test, p=0.024), the difference was small.
Due to the greater variability at the edge (0-10 m), no signiﬁcant differences were found
between forest types for LAI (6.36zt2.99 mz/m2 vs. 3.82i0.39 mz/mz, t-test, p=0.l4) or

openness (5.861161% vs. 5.31i1.69%, t-test, p=0.l6).

32

.088 £80 ::0 £8: wc80 3:: :08088 003 1:: e
88:: 3:088: 0:03 880> 88:8 02:80: 8 o: :0 820> 0: 88:09:80 3 :08:::0:0: 0:03 0:00.:0 08m:

08:88:00 0:880 @0888 88.: :0::_0x0 0:03 8:_0> 088:. 5:80: 80:0 ::0 880 8: 8:_0> 88:8 :38 3:088:

e: :0: 88:3 .858 85: 08> 02008: 0:: :o 00:08:: 0:: c: 0:: 888308 8 0w:0 8:: m:o_0 :o:0::0:0: 0m:0 .8 80:8 08.?

 

 

 

 

 

 

 

 

 

 

 

 

ﬁle 81:0 21:0 07:0 91:0 SA: 070.0 91:0 083:
m S o 3 2 n: o o 8:002
I: we :0 2: a. :0 SN 8.: 0.: 80080 088:0
0.0 E 2. m w: 0.: 3 E 0083:
m 2 -- m a: 2 2 0: cm 880:0
. -- . -- . 8 u m 2.:
0 em 0 o -- -- o -- as: 520
5 o -- o o -- o: o -- <0 888
-- -- a: -- 0:. we a s :0 -- .8: saw
-- -- :0 -- 9. 80: we :0 -- :38
-- -- -- -- -- -- :0 :0 880: :08: w8_:0m
-- 0: -- m -- :2 as: -- as: 820
-- :0 -- m -- -- 08$ :0 -- :0>oO
-- -- as :0 -- :0 2 0 30:: as: :28
2 2 .8 8 a: m o: 0:. .20: Sam
-- :2 :0 .8 -- o :0 mm 8 c :0 :25:
a: m a: 2 o :0 o :0 :0 :00: as: as:
3 m 0 z 3 m 0 2
88:80:00 2:02.88: 082
80:8

00:0 :8735 .00:0 _08nu<m .:0:00:0: 0:00.:0 o: n -- ANN 088:: 3:50: 8000.80 0:: 0:8 0w:0 :88: 0:: 88.: ::0:8 30: 08::
88cm: .0m:0 0:: :0 880> 8:080 308:8 88:: 0:83 8 80:0 88:88.: ”8:0 0:: :0 :88: 0:03 888 0:88 :08: 208:8
88.: >05 8:808 D0888 0:030:00 e: :00: 0:03 mod“: :0 2:008:80 300:0 0w:0 3:0 ._0>0_ :dv: 0:: :0 800:880

0:0 888:0:0: 8 888 :08: 36v: 0:: :0 88:8»: 0:0 80:”: 860:: 08:088.: ::0 0>:0: :8: 8:808 89:80 08:.
0:5 :88: ::0 80:00 :08 8.: A8 8: 8:08 0:: :83 8:80:88 .3 88888: A8: 800:0 0w:0 .8 8:088: Him 030:.

33

Cover (%) Height (m)

Stem density (stems/m2)

Beech-Maple Forests

 

_A .4
C) IV
0

.0
on

9.0.9.0
Guano:

 

 

100

 

 

 

 

 

 

0 T
200 -
/ /
100// /
50 <
'1:—
9/;\\t:-3: , a 4x
0 . j . . 1’

.7-

0 20 40 80

Distance from edge (m)

60

100

__—.——
O
——_'—__

_ A —

Oak-Hickory Forests

 

 

 

 

 

 

 

 

 

 

 

 

60 80
Distance from edge (m)

Nonh

East

South

West

Figure 2.3: Herb layer height, cover, and stem density by aspect in beech-maple and oak-
hickory forests. This layer included woody plants <1 m tall.

34

Beech-Maple Forests

2.0

 

Height (m)

 

 

 

 

 

 

 

 

:\°‘

8

>

O

0

$1.4.

7,1.2.

31.0.

(D

30.8 r

20.6« /\ °
0 N

“0'4 1°" M . f‘
EOZi o$\/3/“\: ‘H ﬁg
wo.o¢.°, °':°:,

0 20 40 60 80 1 00

Distance from edge (m)

 

O

——-—'_——
_ —A _

Oak-Hickory Forests

 

 

 

 

 

 

 

 

 

 

 

 

0 20 4O 60 80 100
Distance from edge (m)
North
East

South
West

Figure 2.4: Shrub layer height, cover, and stem density by aspect in beech-maple and
oak-hickory forests. This layer included all woody species 1-2 m in height.

35

Beech-Maple Forests

Oak-Hickory Forests

 

 

Height (m)

 

 

 

 

 

Cover (%)
8

 

 

 

 

0.6 ~

0.4 ~

0.2 «

Stem density (stems/m2)

 

 

0.0

 

0 20 40 60 80 1 00

Distance from edge (m)

_—_'—--n—-

___._

 

 

 

 

 

 

 

0 20 40 60 80 100
Distance from edge (m)

North
East
South
West

Figure 2.5: Sapling layer height, cover, and stem density by aspect in beech-maple and
oak-hickory forests. This layer included all woody species >2 m tall, but <2.5 cm dbh.

36

Beech-Maple Forests Oak-Hickory Forests
0.8

 

 

0.6 4

Basal area (m2/m2)

 

 

 

 

Stem density (stems/m2)

 

 

 

 

 

Leaf area index (m2/m2)

 

 

 

 

 

 

0 20 4O 60 8O 1 00 0 20 40 60 80 100
Distance from edge (m) Distance from edge (m)

——0— North

0 East
— — —v—— —- — South
— —A — West

Figure 2.6: Canopy basal area, stem density, and leaf area index by aspect in beech-
maple and oak-hickory forests. The canopy layer includes all woody species >2.5 cm
dbh. LAI was measured only in north and south edges.

37

Beech-Maple Forests Oak-Hickory Forests

 

14 - North — North

 

 

 

East

 

 

 

South . South

Canopy Openness (%)
I;

 

 

14 . West . West

 

 

 

 

 

 

0 20 40 60 80 1 00 0 20 40 60 80 1 00

Distance from edge (m) Distance from edge (m)

Figure 2.7: Canopy openness (%) for beech-maple and oak-hickory forests.

38

Table 2.5: Cumulative R2 values from Type 1 ANOVA. R2 values represent the
proportion of variation explained by each variable and all the preceding variables. The
openness, distance, and openness x distance (0 x D) categories include interactions with
aspect. Non-signiﬁcant values (in parentheses) indicate that a variable does not
signiﬁcantly increase the explanatory power of the model. **p<0.001, *p<0.01.

 

 

 

 

 

 

 

 

 

 

 

Metric Beech-maple
Aspect Openness Distance 0 x D

Herb layer Height 0.06* 0.29" (0.30) O.38**

Cover 0.08* 0.17" 0.20* 0.29"

Stern Density 0.07** 010* O.16** O.23**
Shrub Height 0.07** 0.11** (0.13) 0.15*
layer Cover 010* 0.1 3** (0.14) (0.15)

Stem Density 0.08** (0.10) 0.12* (0.13)
Sapling Height 0.08* 0.17** 020* 0.29**
layer Cover 01]" O.16** 0.34** 0.37*

Stem Density 0.04* 0.05* (0.06) O.10**

Oak-hickory
Aspect Openness Distance 0 x D

Herb layer Height (0.001) O.16** 0.19" 0.24**

Cover (0.04) 0.15** (0.16) 0.20**

Stem Density (0.03) 0.14" 0.17* 0.22“
Shrub Height (0.05) O.10** (0.10) 0.16
layer Cover (0.02) 0.04* , 0.07* 0.09*

Stem Density (0.02) 0.06" 0.09* 0.12*
Sapling Height (0.04) O.15** (0.16) 0.20**
layer Cover 0.04* 0.05* (0.06) O.10**

Stem Density (0.02) (0.03) (0.04) 0.09**

Aspect

 

The interaction between distance and aspect was always signiﬁcant for understory
variables in both forest types (Table 2.3), indicating that the effects of distance vary with
aspect. In the canopy layers, interactions were never signiﬁcant in oak-hickory forests,
but the main effect of aspect was signiﬁcant in all cases except for LAI. While aspect
had no signiﬁcant effect on canopy stem density in beech-maple forests, either the main
effects of aspect or its interaction with distance were signiﬁcant in all other cases. Aspect
usually explained very little variation in understory structure (Table 2.5). In oak-hickory

forests, a model including only aspect explained signiﬁcantly more variation than a null

39

model only for sapling cover. The effects in beech-maple forests were slightly larger
(004-01 1) and always signiﬁcant.

In beech-maple forests, northern aspects had the smallest average zone of edge
inﬂuence (7.1 m). Average edge width of east edges (9.5 m) was more similar to that of
north edges (7.1 m) than to south or west edges (17.5 and 18 m, Table 2.4). Furthermore,
the median edge width of north and east edges (0 m) was the same. Southern edges had
the widest average edge zone (17.5 m), while eastern edges had the greatest maximum
DEI (75 m).

In oak-hickory forests, average edge widths of north and east edges (5.0 and 2.5
m) were lower than widths of south and west edges (7.1 and 8.0 m). The largest DEI (80
m) was observed for herb cover and stem density in east edges. However, these results
were inﬂuenced by the occurrence of the invasive vine Vinca minor (common
periwinkle), which led to unusually high interior values for cover and stem density along
the eastern edge. Where it was present, this creeping herb completely covered the forest
ﬂoor.

enness

Openness almost always added more explanatory power to the model (Table 2.5).
The maximum increase in R2 (0.23) was larger in beech-maple forests than in oak-
hickory forests (0.16), but average increases were very similar (0.071 vs. 0.076). The
addition of distance and the openness x distance interaction usually increased model
explanation Men The one exception was for shrub cover in beech-maple forests.
Average increases in R2 for beech-maple forests were larger (0.09) than those for oak-

hickory forests (0.06), and overall R2 values were also higher in beech-maple forests

4O

(0.23 vs. 0.16). In both forest types, R2 values were highest for the herb layer and lowest

for the shrub layer.

Species composition
Understory

Understory species richness was signiﬁcantly inﬂuenced by distance (Table 2.6).
In beech-maple forests, richness was also signiﬁcantly affected by aspect; this was not
the case in oak-hickory forests. Richness in both forest types was highest at the edge
(Figure 2.8). Edge effects for beech-maple forests were smaller in north and east edges
(both 0 m) than in south and west edges (40 and 15 m, respectively). In oak-hickory
forests, richness (across all aspects) differed signiﬁcantly from interior values for 10 m.
Overall richness was higher for oak-hickory forests (5.7 vs. 1.8 species/m2) and
decreased less with distance than richness in beech-maple forests (2.1 vs. 3.3 species/m2,
Figure 2.8).

Table 2.6: P values from ANOVA F-tests. SR=species richness, ST=shade tolerance, D
x Aq'nteraction between distance and aspect. Values in parentheses are not signiﬁcant.

 

 

 

 

 

Beech-Maple Oak-Hickory
Metric Distance Aspect D x A Distance Aspect D x A
Understory SR <0.001 <0.001 <0.001 <0.001 (0.48) (0.18)
Sorensen <0.001 (0.074) (0.26) <0.001 (0.31) (0.13)
Canopy SR 0.025 (0.64) (0.50) <0.001 <0.001 (0.38)
ST <0.001 (0.29) (0.12) <0.001 (0.16) 0.005

 

 

 

 

Sorensen similarity to plots at 100 m increased signiﬁcantly with distance from
the edge (Table 2.6, Figure 2.8). While distance had highly signiﬁcant effects on
similarity in both forests, aspect did not. The maximum edge effect in both forest types

was 90 m, but average DEI was higher for oak-hickory forests (27.5 m vs. 21.0 m).

41

While Sorensen values at the edge (0-10 m) were lower in beech-maple forests (0.28 vs.
0.41), values in both forests reached ~0.5 by 30 m from the edge (Figure 2.8).
281222!

Patterns in canopy tree species richness were similar to those for the understory,
but richness was much lower. Distance signiﬁcantly inﬂuenced richness in both forests,
but aspect was signiﬁcant only for oak-hickory forests (Table 2.6). Overall richness was
slightly higher in oak-hickory forests, especially at the edge (Figure 2.8). In beech-maple
forests, richness was signiﬁcantly higher for 10 m (main effect of distance); in oak-
hickory forests, signiﬁcant edge effects were smaller (0 m, Table 2.7).

Table 2.7: Edge effects (m) determined by comparisons with the interior (100 m) for
each aspect and forest type. The analysis included both native and non-native species.

Gray boxes indicate that metric values were higher at the edge; signiﬁcant effects in
white boxes indicate smaller metric values at the edge. Effects are signiﬁcant at the

 

 

 

p<0.05 level; values in parentheses are signiﬁcant at the p<0.l level. -- = no effects
detected, ep=exterior plots
Forest Metric Aspect N E S W
Beech-maple Understory Species Richness 0 0 40 15
Sorensen’s Index 90 5 20 10
Canopy Species Richness 10a
Shade Tolerance 20a
Oak-hickory Understory Species Richness 10a
Sorensen’s Index 90 30 70 60
Canopy Species Richness (ep) 0 -- 0
Shade Tolerance (0) 0 -- 15

 

 

 

 

 

alDEI for these metrics was determined using distance as a main effect, because no
signiﬁcant edge effects were found for the distance x aspect interactions.

Canopy tree shade tolerance was higher and less variable in beech-maple forests
than in oak-hickory forests (Figure 2.8). While shade tolerance was signiﬁcantly
inﬂuenced by distance in both forest types, aspect did not inﬂuence shade tolerance
signiﬁcantly in beech-maple forests (Table 2.6). The interaction between distance and

aspect was signiﬁcant in oak-hickory forests. DEI in beech-maple forests (determined

42

from the main effect of distance) was 20 m. In oak-hickory forests, edge effects were
smaller (515 m) and more variable. ‘
Multi-response Permutation Procedure (MRPP)

MRPP effect sizes were highly signiﬁcant for all grouping variables except
distance (Table 2.8). Results were qualitatively similar for the presence/absence of all
species and for percent frequency of the most common species, but the latter data set
yielded larger effect sizes for all signiﬁcant variables. For the presence/absence data set,
composition varied most among sites (A=0.160). Using forest-distance groups offered
only a slight improvement over using forest alone (A=0.060 vs. A=0.054). For the top
ﬁve species data, effect size for site was much larger (A=0.25 8). The forest-distance
groups were again less homogenous than site groups (A=0.090). These differences were
also reﬂected in the indicator analysis data. Only one indicator species with an indicator
value of more than 20% was found for forest-distance groups. For site groups, 16
indicator species with values greater than 40% were found (Table 2.9). Three non-native
species (Alliaria petiolata, Rosa multiﬂora, and T araxacum oﬂicinale) were found to be
signiﬁcant indicator species for three oak-hickory sites (Table 2.9).

Table 2.8: Multi-response permutation procedure (MRPP) results for all understory
species and for the most common understory species. Rare (<3 occurrences) and

unknown species were excluded. ‘A’ is the chance-corrected within-group agreement, a
measure of effect size. Cluster analysis was only conducted for the top ﬁve species data.

 

 

 

 

 

All species Top 5
Grouping Number Number
Variable of groups A p of groups A p
Forest 2 0.054 <0.0001 2 0.090 <0.0001
Site 10 0.160 <0.0001 10 0.258 <0.0001
Aspect 4 0.025 <0.0001 4 0.040 <0.0001
Distance 12 -0.002 0.80 12 -0.006 0.95
Forest-distance 24 0.060 <0.0001 24 0.097 <0.0001
Clusterirg -- -- -- l 0 0.298 <0.0001

 

43

 

Beech-Maple Forests

Oak-Hickory Forests

 

Understory SR
(speCIes/mz)
A 05 on

N

O

 

 

 

Percent similarity (Sorensen)

 

 

 

 

 

 

 

 

 

 

 

1.0‘ o

‘

tr .
a)
8
.e
>.
D.
O
C
(U
0

0.0 -

SJ 0° :6 ‘9 '30 o 00:0; °
hie/NJ V i M, °° 1’
(D A a
x
a
o 21
C 0
8

1

0 . .

0 20 40 60 80 20 40 60 80 100
Distance from edge (m)

__E'___

__Q_

North
East
South
West

Distance from edge (m)

Figure 2.8: Understory species richness (SR) and Sorensen similarity to the interior, and

canopy tree species richness and shade tolerance (ST). For clarity, error bars are not
shown for canopy data.

Cluster analysis

Cluster analysis differentiated distance-plots into two large groups, one containing
predominantly oak-hickory plots and the other containing only beech-maple plots from
Hudson Woodland and Tourney (Table 2.10). Within the beech-maple group, there were
two major subgroups (clusters 9 and 10), but plots from Hudson and Toumey were
present in both groups. In the oak-hickory group, smaller clusters were formed largely by
site. Two clusters (4 and 8) contained plots from only one site each (Johnson Park and
Clear Lake), but a few (mostly edge) plots from these two sites were included in other
clusters. Two beech-maple sites (Ionia and Poft) were also included in the ‘oak-hickory’
group, as were several edge (330 m) plots from Hudson and Tourney. When most plots
from a site were clustered together, plots were often grouped by distance. Within the
APN group (cluster 3), for example, distances 10-30 m were separated from distances 40-
100 m. In cluster 4, J PE distance-plots 20-40 m were separated from plots 50-80 m,
while for JPW, 5-30 m were separated ﬁom 40-100 m. In other cases, however, divisions
were less clear. In cluster 8, CLS distance-plots 0-10, 30, and 100 m were in one group
while 20 and 40-90 m were in another.

Indicator species analysis for the 10 cluster level yielded at least one species with
an indicator value of 25% for all groups (Table 2.10). Indicator values were lower for the
two beech-maple groups (clusters 9 and 10) than for the other groups. The largest
number of signiﬁcant indicator values was found for cluster 8, where six species had
indicator values of 40% of higher. Indicator values were lowest for cluster 10, the largest

cluster: the highest value was 28%.

45

Table 2.9: Statistical indicator species for each site. All listed species were signiﬁcant at
a=0.05. For most sites, only those species with indicator values of at least 30% are listed.
If fewer than two species reached this threshold, the species with the next highest

 

 

 

 

 

 

 

 

 

 

 

indicator value is listed. Non-native s ecies are marked with an asterisk (*).
Site Forest Type Indicator species % of perfect
indication
Hudson Woodland beech-maple Acer saccharum 24
(HW)
Ionia (IO) beech-maple Lindera benzoin 42
Fagus grandifolia 38
Poff Woods (PW) beech-maple Polygonum virginianum 37
Tourney (TO) beech-maple Acer saccharum 30
Impatiens capensis 24
Aman Park (AP) oak-hickory Alliaria petiolata* 67
Bald Mountain oak-hickory Rubusﬂagellaris 42
(BM) Carya ovata 37
Carya sp. 35
Quercus alba 30
Clear Lake (CL) oak-hickory Maianthemum canadense 65
Hamamelis virginiana 62
Mitchella repens 46
Ostrya virginiana 41
Osmorhiza claytonii 38
Sassafras albidum 32
Acer rubrum 30
Johnson Park (JP) oak-hickory Smilacina racemosa 54
Polygonatum pubescens 41
Viburnum trilobum 34
Rose Lake (RL) oak-hickory Rosa multiﬂora* 49
Ulmus americana 35
Anemonella thalictroides 31
Rubus allegheniensis 31
Seven Lakes (SL) oak-hickory Taraxacum oﬁicinale* 57
Aster sp2. 57
Camus foemina 51
Aster sp. 46
Prunella vulgaris 43
Potentilla simplex 41

 

 

 

 

 

46

 

 

C. 5 8888828 3.235828%

2: e 38

8 .8 .8 .8 .2 .828

02 .8 .8 .8 .8 .8 .8 .228
o2 .8 .8 .8 .8 .8 .2 .8 .2833:
8:33:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

as $3383 82 o2 .8 .8 .8 82:3: 2
8 .8 .8 .8828 8 .8 2:28
8 .8 .8833:
8C 5.538% 82.. o2 .8 .8 .8 .8 .8 .8 .283:
A88 5:28.? 383.22.. 8 .8 .8 .8 .2 .253: m
33 323.5%?» 99:0
:8 3353.3 £3828: =e-mAU
a8 323333 83:85:38: o2 3 8-240 w
G: 38.3% $8388 289 8 .8 .8 .8 .2 .m .225
:2 v.38 82%: 28:3: 253: 2 .8 .225 222: 8
8V sage: 832.. 82-23.:
33 58:8 3.88 o2 2 8-82 8-83: 2 .8 .2m.: 8
as 88.88 secs: 8 .8 .8 .8 .8 .823:
2 9 2888883: 88:3: 2: e 8.22: 8
G2 2888 5:58.: 8 .8 .8 .8 .8 .8 .8 .8 .23.:
as 3958.: 88:8. 8 .8 .8 .8 .8 .8 .88.: 8
£3 5.2883. 82.. 220:. 2228 O2 .88.:
2: .838: 5:32 8 .8 .282 8-83: 2: s 83.2 2: 2 2.2.3. 8
$3 832885 82558: 8 .8 .2 .8 .223: o2 2 8-2:: 8.22:
:3 83838 5:8 m .32 8 .2 .8 .23.? 8-2.2 N
9.3 ”88888 2.8: :98: 2 .8 .228:
as .388... as: o 722: 22.2 _
momooqm .5329: muoWﬁoogﬁme oameucooom mac—9-02.556 ~AuOv—oiuxwo hymn—U

 

 

dd Bash E 53m 8e 3032552? 86 AL 02:83 cm 5.3 @828: Be 860
Soﬁe: me .x. ”modue 8 385%? 0.83 228% 882?: @324 .5830 :08. E woes—8: $8886-0sz 203-8385 ”EN 035.

m 03:27:02 .womoﬁcouea: cozw 2 58065

47

Discussion
Vegetation structure
DEI: Herb layer

The herb layer was most sensitive to edge effects: signiﬁcant edge effects were
found for all variables in both forest types (except for cover in north and west edges of
oak-hickory forests, Table 2.4). Herbs were usually taller and herb cover was greater at
the edge in both forest types, although overall cover was slightly higher in oak-hickory
forests (Figure 2.3). While stem density was often higher at the edge, the difference
between edge and interior (2.0% difference) is much smaller than for cover (54.5%
difference). Thus there is only a small increase in the number of plants at the forest edge,
but those plants grow much larger than their counterparts in the forest interior. The
decrease in height (0.49 m) and cover (38.3%) from edge to interior was greater in beech-
maple forests than in oak-hickory forests (0.33 m and 29.9%, Figure 2.3), matching the
greater decrease in canopy openness (Figure 2.7).

Herb layer edge effects were smaller in beech-maple forests (most 5 10 m) than in
oak-hickory forests (0-80 m). The largest effects were observed for north edges in beech-
maple forests (40 m) and east edges in oak-hickory forests (80 m). The large effects for
east edges in oak-hickory forests were due to the presence of the invasive vine V. minor,
which led to unusually high interior values for cover and stem density. When those
values are excluded, results were similar to those of other authors. Palik and Murphy
(1990) found decreases in ground layer stem density for woody species up to 5 m in north
edges and 45 min south edges. Meiners and Pickett (1999) observed that understory

cover was highest at the edge; DEI was not calculated.

48

DEI: Shrub and saplinjgﬁwers

For most aspects, shrub and sapling abundance was very low outside the forest,
peaked at 0-5 m, and then declined (Figures 2.4 and 2.5). The low values outside the
edge reﬂect anthropogenic edge maintenance (mowing, fences, etc.) rather than natural
patterns. Where edges are not maintained, shrub and sapling cover and stem density may
be higher outside the edge than inside (Wales 1972). In general, shrubs in the study sites
were taller (oak-hickory: 1.0 m; beech-maple: 0.75 m) and more dense (oak-hickory: 11.7
stems/m2; beech-maple: 5.1 stems/m2) in oak-hickory forests, while saplings were more
abundant in beech-maple forests (21.3% vs. 10.4% cover).

Fewer signiﬁcant edge effects were found for the shrub and sapling layers,
especially in oak-hickory forests, than for the herb layer (Table 2.4), partly because the
abundance of shrubs and saplings varied more than that of herbs. While many edge
effects in beech-maple forests were small (limited to the exterior plots), some large
effects were found for sapling cover and stem density in south- and west-facing edges (65
m and 40 m, respectively). All edge effects in oak-hickory forests were less than 30 m.
Matlack (1993) found larger edge effects (40 m) for shrubs in north, east, and south
edges; no effects were detected in west edges. Gehlhausen et al. (2000) found higher
sapling density close to the edge, but did not determine edge width.

DEI: Canopy

In both forest types, canopy tree basal area and stem density were always higher
at the edge (Figure 2.6), but signiﬁcant edge effects extended at most 20 m into the forest
(Table 2.4). Overall values for basal area and stem density, as well as the trends with

distance, were similar in the two forest types. Palik and Murphy (1990) also found high

49

stem densities at the edges of two beech-maple forests, with similar values for edge
effects (0-10 m for north edges, 5-15 m for south edges). Wales (1972) observed that
basal area and stem density were highest at the edge for most canopy species in a mature
oak-hickory forest. While stem density was highest at the edge for 22 sites in Ontario (all
east-facing edges), basal area was lower (Burke and No] 1998). Edge penetration at those
sites was relatively small: 20 m for stem density and 10 m for basal area.

The difference between edge and interior openness values was greater for beech-
maple forests (7.34%) than for oak-hickory forests (3.91%). Shrub and sapling cover at
the edge was lower than in beech-maple forests, allowing greater lateral light penetration
at the edge. One exception was west edges in beech-maple forests. Because one of the
transects was located close to a large gap, sapling stem density was particularly high
(Figure 2.5). Edge penetration for LAI and openness was greater in beech-maple forests
(5-30 m) than in oak-hickory forests (5—1 5 m, Table 2.4). These results were similar to
those previously reported. Matlack (1993), for example, found increases in light
availability of 10-11 m for north and east edges, and 13-35 m for south and west edges.
While Gehlhausen et al. (2000) observed a similar range (10-40 m across all aspects),
Burke and N01 (1998) found a much smaller width (5 m) for east-facing edges.

While interior LAI of beech-maple forests was signiﬁcantly higher than in oak-
hickory forests, the difference in percent Openness was very small (<0.5%). Parker and
Tibbs (2004) observed than LAI of F. grandifolia (2.1 mz/mz) is more than 10 times that
of Quercus alba (0.095 mz/mz), Q. velutina (0.14 mz/mz), and Carya glabra (0.097
mz/mz). This difference in LAI, which was reﬂected in LAI values from two sites, was

poorly detected by hemispherical photography. Photo analysis differentiates sky from

50

‘non-sky’ (e.g., leaves, branches, and trunks), but it cannot distinguish between pixels
occupied by a single leaf and those occupied by many leaves. In fact, van Gardingen et
al. (1999) found that hemispherical photography can underestimate LAI by as much as
50% in highly clumped canopies.

The differences in species LAI noted above likely explain the low LAI values
along the south edge of the beech-maple site (Tourney Woodlot). Quercus rubra and Q.
alba are completely absent from the north edge, but are relatively abundant along the
south edge (see Chapter 5 for further discussion). The increasing importance of A.
saccharum from south to north correlates with the increase in LAI from the south edge to
the forest interior (Figure 2.6).
Effects of Aspect and Openness

Aspect inﬂuenced almost all variables (except canopy stem density in beech-
maple forests) signiﬁcantly (Table 2.3). These results, which indicate that the inﬂuence
of distance varies with aspect, are corroborated by the variations in edge effects among
aspects (Table 2.4). Some authors have grouped north and east edges together as ‘cool’
edges (e. g., Brothers and Spingarn 1992). Average width of east edges in beech-maple
forests (9.5 m) was similar to that of north edges (7.1 m). However, the maximum edge
width in east edges (75 m) was much larger than in north edges (40 m) and was more
similar to the maximum in south edges (65 m). These results appear to support Matlack
(1993) in including east edges in the ‘warm’ group. In oak-hickory forests, however,
average edge widths were smaller in north and east edges (5.0 and 2.5 m) than in south

and west edges (7.1 and 8.0 m). Direction of metric response at the edge was not always

51

consistent within the warm and cool groups, but the lack of signiﬁcant edge effects for
most of the shrub and sapling variables makes evaluation of the groupings difficult.
Even though the difference in canopy openness between the two forest types was
small, partial R2 values indicate that canopy openness signiﬁcantly inﬂuenced vegetation
structure. Adding canopy openness to the statistical model increased explanatory power
signiﬁcantly in almost all cases (Table 2.5), although increases were relatively small

(average increase of 0.071 for beech-maple and 0.076 for oak-hickory forests).

Species composition

The main effect of distance signiﬁcantly inﬂuenced understory species richness
and Sorensen similarity to the interior, as well as canopy tree species richness and shade
tolerance (Table 2.6). In only three cases were aspect or the distance-aspect interaction

signiﬁcant. Thus, while aspect may inﬂuence vegetation structure, it exerts less inﬂuence

over species composition in the understory and canopy.

Understog

Understory species richness (including non-natives, which are discussed in more
detail in Chapter 3) at the edges of beech-maple forests was higher than the interior for 0-
40 m, depending on the aspect; in oak-hickory forests, edge effects were limited to 10 m
(Table 2.7). Luczaj and Sadowska (1997) found that richness for vascular plants
(understory and canopy) in a deciduous forest in Poland was higher within 3 m of the
f‘Ol'est edge than in either the forest interior or the adjacent grassland. In two mixed-
Il‘lesophytic forests in Illinois, understory species richness was signiﬁcantly higher for 15-

60 In for north, south, and west aspects; no signiﬁcant effects were detected on east edges

52

(Gehlhausen et al. 2000). The authors also found that edge penetration based on species
richness was greater than edge width for canopy openness, soil moisture, relative
humidity, and air temperature. Matlack (1994) found several ‘edge-orientated’ species in
oak-chestnut forests in Pennsylvania, but all of these species were also found up to 40 m
into the forest (the greatest distance surveyed).

The higher understory species richness in oak-hickory forests, both at the edge
and in the interior, could be due to greater light availability in this forest type. Although
only a small (<0.5%) difference in openness was detected by hemispherical photography,
LAI values of beech-maple forests exceeded those of oak-hickory forests. Furthermore,
differences in light availability may exist early and late in the growing season because A.
saccharum and F. grandifolia leaf out much earlier than many Quercus and Carya
species. In Quebec, A. saccharum leaves were fully expanded more than 100 degree-
days earlier than Q. rubra leaves; F. grandifolia was intermediate (Lechowicz 1984).
Parker and Tibbs (2004) observed that, in Maryland, F. grandifolia trees retained their
leaves for 195 days while Q. alba, Q. velutina, and C. glabra retained leaves for only

1 86, 185, and 163 days, respectively. Understory plants in beech-maple forest, therefore,
have a shorter high-light growing season.
Seasonal light availability may also inﬂuence the magnitude of change between
edge and interior. In beech-maple forests, the difference between edge and interior
Values for herb height, herb cover, understory species richness, and Sorensen similarity
Were all greater than in oak-hickory forests (Figures 2.3 and 2.8). If beech-maple
caIIOpies close earlier in the growing season, lateral light availability at the edge may be

l"11011: important for understory vegetation than in oak-hickory forests.

53

While oak-hickory forests may have greater average light availability in early
spring due to later leaf-out, spatial variability in light levels during the same time may
also be high. Acer rubrum, which leafs out earlier than Quercus spp. (Lechowicz 1984),
was an important component of the canopy in several sites (Table 2.1). Kato and
Komiyarna (2002) found that such differences in phenology led to great spatial variation
in light availability in early spring; this variability was almost eliminated after canopy
closure. Similar differences in the study sites could explain the greater variability of

species richness in oak-hickory forests than in beech-maple forests.

Canopy

The increase in canopy tree species richness at the edge was largely due to the

increased abundance of less shade-tolerant species, such as Q. rubra, U. americana, and
P. serotina (Figure 2.8). Interestingly, the edge effects based on shade tolerance were
sometimes larger than effects based on species richness (Table 2.7). This disparity may
be due to the smaller variability in shade tolerance. Whitney and Runkle (1981) also
found higher stem densities of shade-intolerant species at the edge. Because the density
of these species was greater than the decreases in density of shade-tolerant species (A.
saccharum and F. grandifolia), overall stem density was increased at the edge. Similarly,
increases in stem density of less shade-tolerant species, such as Q. rubra, F. americana,
and Tilia americana (basswood), were found in two beech-maple forests in Michigan
( P alik and Murphy 1990). In an oak-hickory forest in New Jersey, shade-intolerant

Canopy species and canopy species that often reproduce vegetatively were most common

at the edge (Wales 1972).

\GTouping procedures

54

Understory vegetation in edge plots was highly dissimilar from vegetation in
interior plots (Figure 2.8). Edge effects for Sorensen similarity to the interior were
among the largest found: 90 m for north edges in both forest types (Table 2.7). The
difference between edge and interior plots was greater for beech-maple forests (0.54)
than for oak-hickory forests (0.39). Given the greater species richness in oak-hickory
interiors, it is not surprising that they have more species in common with the forest edge
than is the case in more species-poor beech-maple interiors. While Gehlhausen et a1.
(2000) did not calculate edge effects, the trends they observed in Sorensen similarity are
similar: edge plots were least similar to interior plots, and percent similarity tended to
l evel off after 15-25 m. Clustering identiﬁed broad differences between two beech-maple
sites (Hudson and Tourney) and all other sites. These differences appear to be based less
on the presence of certain species in Hudson and Tourney (only one indicator value was
>30%) than on the absence of many other species (Table 2.10).

While the species composition of edge plots was very different from that of
interior plots, species composition varied more among sites than among distances within
a site. While MRPP effect sizes were large for groups based on sites (A=0.160-0.258),
groups based on distance across sites did not differ signiﬁcantly from each other (Table
2 -8). Clustering conﬁrmed these results: most plots from a site were concentrated within
a single cluster (Table 2.10). Furthermore, indicator species for clusters were often the
Same as indicator species for the predominant site within the cluster (Tables 2.9 and

2 - 10). However, relationships among plots within a site cluster were oﬁen based on
di Stance ﬁ'om the edge. For APN (cluster 3), plots more than 30 m from the edge were

SeIJaIated ﬁ'om plots closer to the edge. J PW (cluster 4) was divided at the same point;

55

J PE (cluster 4) was divided at 40 m. Furthermore, while most plots of a site were located
in a single cluster, edge plots were often placed in separate clusters (Table 2.10). For
example, the ﬁrst three plots (0-10 m) from CLN, BMN, and RLN are all separated from
the rest of each site’s plots. This separation conﬁrms that the understory composition of
edges differs greatly ﬁ'om the composition of forest interiors. MacQuarrie and Lacroix
(2003) used clustering within a site to determine depth of edge penetration for vegetation
composition. The resulting DEI (120 m) is among the largest reported. In contrast, using
separation of plots within a site cluster for the present study generally results in much
smaller edge zones (30-40 m). Edge effects are even smaller (0—10 m) if they are based

on the separation of edge plots into entirely different clusters.

Conclusions

Vegetation structure—understory: Estimates of edge penetration based on measures of
understory structure in both forest types varied between 0 and 75 m, depending on the
metric and aspect of interest. Herb layer structure (height, cover, and stem density) was
more sensitive to edge effects than shrub or sapling structure, especially in oak-hickory
forests. Herb height and stem density were the only metrics for which signiﬁcant effects
were found in all four aspects of both forest types. In beech-maple forests, the range of
signiﬁcant edge effects for the herb layers was smaller than for the shrub/sapling layers
(040 in VS. 0-75 m); in oak-hickory forests, herb layer effects were larger (0-30 m vs. 0-
1 0 m). Thus the extent of edge depends on the vegetation layer studied.

mum structure—canopy: The canopy was less sensitive to edge effects (max. edge:

3 0 m) than the understory (max. edge: 75 m). Edge effects for canopy basal area and

56

stem density were signiﬁcant more often in oak-hickory forests. Surprisingly, canopy
openness (as measured by hemispherical photography) for the two forest types differed
little (<0.5%). While LAI values indicated that beech-maple forests have more foliage
area than oak-hickory forests (interior LAI: 6.52i0.13 mz/m2 vs. 4.45i1.16 mz/mz),
hemispherical photography could not detect it.
Species comﬂsition: Signiﬁcant edge effects based on species composition were larger
( ave. 24.3 m) than edge effects determined from vegetation structure (9.5 m). Cluster
analysis often separated a site’s edge plots (510 m) from plots in the forest interior,
conﬁrming that there are large differences in species composition between edge and
interior plots. However, species composition varied more among sites than among
distances across sites and no ‘edge’ indicator species were found. Thus the inﬂuence of
edge effects on species composition must be evaluated on a site-by-site basis.
Aspect: Aspect signiﬁcantly inﬂuenced almost all measures of vegetation structure in
both forest types. In oak-hickory forests, average edge widths of north and east edges
(5.0 and 2.5 m) were smaller than widths of south and west edges (7.1 and 8.0 m). In
beech-maple forests, the groupings were less clear, especially for species composition.
These results cast doubt on the practice of grouping ‘warm’ and ‘cool’ edges together a
Priori; the validity of such groupings in different forest types deserves further study.
gomparison of forest types: While the direction of change in vegetation structure and
Composition within the forest edge was similar for the two forest types, the magnitude of
the change differed, especially for the herbaceous layer. The changes from edge to
irlterior in in herb height (beech-maple: 0.5; oak-hickory: 0.3 m), understory species

l"ichness (3.3 vs. 2.1 species/m2), and Sorensen index (0.5 vs. 0.4%) were all greater in

57

beech-maple forests than in oak-hickory forests. These results could be due to the greater
foliage area of beech-maple canopies or because beech-maple canopies may close earlier

than oak-hickory canopies, restricting the length of the growing season for plants in the

forest interior.

58

Life on the edge: The spatial disnibuticffrhgfgiredninative species in forest fragment edges
Introduction

Non-native invasive species have become one of the largest threats to
biodiversity. Each year, millions are spent to control these invaders in both terrestrial and
aquatic systems. These invaders, which include plants, animals, and pathogens, displace
native species, decrease species richness, and alter ecosystem processes and functions by
changing ecosystem structure, productivity, decomposition rates, ﬁre frequency, and
hydrology (V itousek et al. 1997, Collier et al. 2002, Gorchov and Trisel 2003).

Forests have long seemed to be less vulnerable to invasion than grasslands
because so many non-native plants are shade-intolerant crop weeds. Most of these
species, as well as their native equivalents, cannot survive under a closed forest canopy.
In Illinois, relative cover of non-native species was lower in forests than in wetlands,
prairies, and planted grasslands (Table 1.1, Spyreas et al. 2004). Yet an increasing
number of shade-tolerant non-natives colonize forests successfully. These plants are
predominantly woody perennials (Brothers and Spingam 1992, Webb et al. 2000) and
many were introduced for horticultural purposes (Ehrenfeld 1997, Webb et al. 2000,
Gayek and Quigley 2001).

As discussed in Chapter 2, there are large differences between the abiotic
environments of edge and forest interior. These differences allow less shade-tolerant
species—both native and non-native—to persist in the edges. Non-native invasion is
expected to be greater along south and west edges, since abiotic edges effects are greatest

there. In ancient (>220 yrs) deciduous forests in Belgium, edge communities differed

59

signiﬁcantly from interior communities, but most ‘weedy’ species penetrated only 3 m
into southern edges (Honnay et al. 2002). The pattern of invasion closely matched the
reduction in light inside the edge. In old-growth beech-maple forests in central Indiana,
both non-native species richness and frequency were higher along southern and western
edges (Brothers and Spingarn 1992). Very few of these species penetrated more than a
few meters into the forest. Gysel (1951) observed a similar pattern in beech-maple
forests of southern Michigan. In an Acer rubrum-Quercus palustris forest in New Jersey,
non-native abundance peaked within 20 m of the forest edge (Meiners and Pickett 1999).
In contrast, MacQuarrie and Lacroix (2003) documented that eight non-native species
had penetrated over 100 m (two up to 300 m, the maximum distance studied) into the
forest interior of upland hardwood forests in Prince Edward Island, Canada. While most
species had <5% cover at their maximum penetration distance, Veronica oﬂicinalis
(common gypsyweed) and Hieracium lachenalii (common hawkweed) had more than
10%. In Ontario, eight non-native species were present as far as 100 m (the maximum
distance studied) into the interiors of 22 upland deciduous forests (Burke and No] 1998).
However, with the exception of Rhamnus cathartica (common buckthom), all species
contributed <1 % cover at that distance.

For a new species to invade a forest ﬁ'agment successfully, it must progress
through several stages: dispersal into the fragment, germination, seedling survival,
growth, and successful reproduction. Wind dispersal is common among non-native
species, especially those that colonize old ﬁelds. Forest edges generally act as a barrier
to wind-dispersed seeds, but the porosity of the barrier depends on edge structure.

Cadenasso and Pickett (2001) found signiﬁcantly more wind-dispersed seeds inside

60

experimentally thinned edges than inside intact controls. Seeds also penetrated farther
into the forest through thinned edges. While wind-dispersed seeds may be partially
blocked by forest edges, seeds dispersed by animals often reach the forest interior. The
two most common non-natives (Rosa multiﬂora (multiﬂora rose) and Solanum nigrum
(black nightshade» found in the forest interiors of beech-maple forests in Indiana are
both animal-dispersed (Brothers and Spingarn 1992). Bird-dispersed seeds can reach
forest interiors (Yates et al. 2004) and deer can spread seeds of invasive Lonicera
(honeysuckle) species into forest interiors (Vellend 2002). Once a bird-dispersed species
is established in a gap, its seeds will likely spread to other gaps. Hoppes (1988) found
that bird-dispersed seeds of native species growing in forest gaps were more likely to fall
into a nearby gap than expected. Seed rain was highest at gap edges. Birds would likely
disperse seeds of non-native species in a similar pattern.

Fragmented forests are usually close to human habitation and are therefore
especially vulnerable to human—dispersed seeds. Dispersal by humans takes several
forms. Seeds can adhere to clothing, be carried in shoe treads, or even be introduced in
horse manure along riding trails. For this reason, and because they create canopy gaps,
trails and roadsides are among the ﬁrst areas to be colonized by new species. Human
activity also increases the seed sources for many horticultural species that can invade
forests (e.g., R. multiﬂora, Berberis thunbergii (Japanese barberry), and Ligustrum
vulgare (common privet)).

Edge structure can affect non-native germination and seedling survival through its
inﬂuence on the abiotic environment. A more open (oﬁen younger) edge will allow

higher light penetration. Because light can be critical to seedling survival, many species

61

have higher survival rates closer to the edge. Luken and Goessling (1995) found higher
densities of Lonicera maackii (Amur honeysuckle) seedlings along forest edges than in
the interior. Both germination and seedling survival of sown Alliaria petiolata (garlic
mustard) were higher in edge plots than in forest interior plots (Meekins and McCarthy
2001)

Soil disturbance and litter removal increase germination for many species.
Disturbance of the forest ﬂoor was correlated with increased abundance of Celastrus
orbiculatus (oriental bittersweet, McNab and Loftis 2002). High densities of intact litter
reduced emergence of C. orbiculatus by 60% (Ellsworth et al. 2004). In contrast,
Meekins and McCarthy (2001) found slightly lower A. petiolata seedling survival in litter
removal plots, possibly due to slightly higher moisture levels in the litter of control plots.
There was no difference in germination.

Population establishment depends on successful reproduction. While individuals
of some species can grow in a forest edge, they fail to ﬂower (Brothers and Spingarn
1992). In such cases, the forest acts as a population sink: population persistence depends
on dispersal into the fragment. Other species can reproduce successﬁrlly along edges
and, to a more limited extent, in the forest interior. Meekins and McCarthy (2001) found
higher adult A. petiolata survival and reproduction in edge plots than in interior plots.
Elaeagnus umbellata (autumn olive), Lonicerajaponica (Japanese honeysuckle), and R.
multiﬂora grow taller at the forest edge; R. multiﬂora and L. japonica also have greater
stem densities at the edge (Yates et a1. 2004).

Some non-natives take advantage of increased light in forest gaps. Lonicera

maackii seedlings established more successfully in gaps than in the forest interior,

62

, although establishment along edges was even greater (Luken and Goessling 1995). The
creation of gaps greatly increased the growth of C. orbiculatus (Robertson et al. 1994).
Sanford et a1. (2003) planted seedlings of three non-native species in open and understory
environments. The two shrub species, Rhamnusﬁangula (glossy buckthom) and E.
umbellata, grew faster in the open. MacQuarrie and Lacroix (2003) found more
signiﬁcant correlations between non-native abundance and canopy cover than between
non-native abundance and distance from the edge.

By creating canopy gaps and exposing mineral soil, disturbances generally
promote invasion by non-native species. Compared to other forests in the region, non-
native species were very abundant in upland hardwood forests on Prince Edward Island,
Canada (MacQuarrie and Lacroix 2003). The authors attributed the high level of
invasion to historical disturbance by human activity. Gould and Gorchov (2000) found
that densities of L. maackii in an anthropogenically disturbed forest were more than twice
those in a forest with little disturbance. However, some non-natives can successfully
establish populations even without disturbance. Acer platanoides (Norway maple)
invades undisturbed forests, although edges and the presence of large seed sources
increase the probability of establishment (Webb et al. 2000). Alliaria petiolata, B.
thunbergii, L. japonica, and Wisteria ﬂoribunda (Japanese Wisteria) also invade
undisturbed forests (Ehrenfeld 1997, Webb et al. 2000).

Phenology can play an important role in successful non-native establishment.
Non-native species that grow before spring canopy closure or after leaf fall tend to be
more successful because they are less shaded during those times. Acer platanoides is in

leaf signiﬁcantly longer than the native Acer saccharum (Kloeppel and Abrams 1995).

63

Lonicera maackii and L. tatarica (Tatarian honeysuckle) are also in leaf longer than any
native woody deciduous species (Woods 1993, Collier et al. 2002), allowing them to
grow when canopy trees are bare. Celastrus orbiculatus, an invasive vine, emerges from
winter dormancy several weeks before most trees (McNab and Loftis 2002). Alliaria
petiolata, a biennial, begins growing in late fall and early spring while native spring
ephemerals are still dormant (Anderson et al. 1996). By capturing more early- and late-
season light, some of these species can survive in the forest interior. Kloeppel and
Abrams (1995), for example, found that A. platanoides distribution was not related to
distance from the forest edge.

The phenology of non-natives, especially shrubs and trees, has signiﬁcant impacts
on native species because it changes light availability in a forest (Woods 1993). This
effect is especially pronounced in forests that have few native shrubs (Gould and
Gorchov 2000, Collier et al. 2002). Stem density and species richness of native species
were signiﬁcantly lower under canopies of A. platanoides than under native A.
saccharum or Fagus grandzfolia (American beech) canopies (Webb et al. 2000).
Similarly, species richness and stem density were reduced below L. maackii crowns
(Collier et al. 2002). Lonicera maackii also reduced survival, growth, and fecundity of
experimentally planted native species, both annuals and perennials (Gould and Gorchov
2000, Miller and Gorchov 2004). Furthermore, overall species richness, as well as stem
density of tree seedlings, were negatively correlated with residence time of L. maackii
(Collier et al. 2002). The closely related L. tatarica also decreased herb species richness
and cover (Woods 1993), but did not affect vine-like or evergreen herbs. These species

appeared able to escape the light limitation caused by increased Lonicera cover (Woods

64

1993). Rhamnusfrangula, another invasive shrub, reduced growth and survival of native
tree saplings and shifted seedling abundance to favor shade-tolerant species (Fagan and
Peart 2004). In addition to competition for light, non-natives can also change native
diversity through competition for water and nutrients, and through allelopathy (Hierro
and Callaway 2003). However, most of the studies described above did not differentiate
among these mechanisms.

In all stages of colonization and establishment, species effects are critically
important. Non-native species exhibit a wide variety of life history strategies and
adaptations. While many authors have compiled lists of characteristics that can be used
to screen new introductions, there are exceptions to every rule (Orions 1986, Rejmanek
2000, Kolar and Lodge 2001, Prinzing et al. 2002). Previous studies of non-natives in
forests have often grouped all non-native species together. While the simplicity of this
strategy is appealing, much information about individual species’ life history and habitat
requirements is lost. Few studies have considered the effects that forest type and canopy
structure have on invasion. This study addressed the following questions:

0 Which non-native species are found in beech-maple and oak-hickory forests?

o What is the spatial distribution of each non-native species with respect to distance
from the forest edge, aspect, and light availability?

0 Are growth form and shade tolerance of non-natives important in determining
their distribution?

0 Are non-native species more abundant (i.e., higher cover and stem density) in

oak-hickory forests, which have sparser canopies?

65

Methods
Site descriptions

Non-native species data were collected at the sites described in Chapter 2.
Vegetation sampling

Non-native species were deﬁned as those that were not historically present in
southern Michigan and nativity was determined using Voss (1972, 1985, 1996). The
grid—transects described in Chapter 2 were used to collect non-native species data. In the
25 m2 plots, the average height, number of stems, and percent cover of each non-native
species present was estimated. Average height was estimated visually. If there were
fewer than 20 stems of a species in a plot, individual stems were counted. For plots with
more than 20 stems, the number of stems was estimated using the following categories:
20, 30, 40, 50, 100, 200. Percent cover was also estimated visually. In the l m2 subplots,
the number of herbaceous non-native stems was recorded using the system described
above.

Non-native species were assigned a shade-tolerance category (low, mid, or high)
based on descriptions of species occurrence in Voss (1972, 1985, 1996) and Gleason and
Cronquist (1991), and on personal observation. Species with low shade tolerance are
found only in open ﬁelds, roadsides, clearings, and similar areas. Species with mid shade
tolerance are found in open woods, wood edges, trails within woods, and thickets.
Highly shade-tolerant species are found in forest interiors and will ﬂower under a closed
canopy. Shade-tolerance ranges were assigned to species found in a variety of shade
levels. For some analyses, the categories were converted into a numerical scale ranging

from 1 (low) to 5 (high).

66

Statistical Analyses

PROC MIXED in SAS (SAS Institute 1999) was used to examine the effects of
distance, lifeforrn, and shade tolerance on non-native cover and stern-density (by aspect
in each forest type). Including aspect in the model statement was not possible due to the
large size of the data set. An appropriate covariance structure was used to account for
variation within a grid-transect. Distance and aspect were both treated as categorical
variables. A t-test on the least squares means (lsmeans) was used to compare distances
within aspects to the interior value (95 m). Depth of edge inﬂuence (DEI) was deﬁned as
the area within which variables differed consistently from interior values. ‘No effect’
was found when no values differed signiﬁcantly from the interior. When the interaction
term was not signiﬁcant (a=0.05), the main effects of distance and aspect were analyzed.
For each combination of lifeforrn and shade tolerance within an aspect, DEI for both
cover and stem density was determined. In a few cases, where there were unusually high
stem densities or cover of non-native species at 95 m, the value at 90 m was used as the
‘interior’ value. Because species in the shrub and tree layers fell into only two categories
each, data are presented for only those categories. PROC MIXED was also used to
analyze the effects of distance, aspect, and the distance x aspect interaction on non-native
shade tolerance in oak-hickory forests. There were too few data to conduct a similar
analysis for beech-maple forests.

The Type I ANOVA described in Chapter 2 was used to examine the relative
importance of lifeforrn, shade tolerance, canopy openness, and distance (and all

interactions among them) as explanatory variables. Analyses were conducted for non- ,

67

native cover and stem density by aspect within forest type. The order of variables was
ﬁxed as listed above.

Height, stem density, and cover data for the most common exotic species (A.
petiolata, B. thunbergii, Lonicera spp., R. cathartica, and R. multiﬂora) were analyzed
separately using the scheme described previously. Data for the two Lonicera species
were pooled. In most cases, there were too few individuals found in beech-maple forests

for analysis; data from beech-maple forests are presented only for R. multiflora.

68

Table 3.1: Non-native species found in this study. All species names are according to
Gleason and Cronquist (1991). Nativity was determined using Voss (1972, 1985,1996).
Shade tolerance categorization was based on verbal descriptions of species occurrence in
Voss (1972, 1985,1996) and Gleason and Cronquist (1991), and on personal observation.
Species with low shade tolerance are found only in open ﬁelds, roadsides, clearings, and
similar areas. Species with mid shade tolerance are found in open woods, wood edges,
trails within woods, and thickets. Highly shade-tolerant species are found in forest
interiors and will ﬂower under a closed canopy. Shade tolerance ranges were assigned
when species are found in a variety of shade levels. Origin: E=Europe, Ea=Eurasia,
A=Asia, t.Am=tropical Americas, US=eIsewhere in US. Longevity: A=annual,
B=biennial, P=perennial, S=shrub, T=tree. Lifeform: h=herb, v=vine, s=shrub, t=tree,
g=grass. Forest type indicates where each species was observed in this study (bm=beech-
maple; oh=oak-hickory).

69

 

Table 3.1

 

 

 

 

 

 

Species Family Origin Longevity Lifeform Shade Forest
tolerance typg
Achillea millefolium Asteraceae Ea/US P h low bm, oh
Chrysanthemum Asteraceae Ea P h low bm, oh
Ieucanthemum

C irsium vulgare Asteraceae Ea B h low brn
Crepis capillaris Asteraceae Ea A/B h low oh
Dianthus armerr'a Caryophyllaceae E A/B h low oh
Duchesnea indica Rosaceae A P h low oh
Hypericum perforatum Clusiaceae Ea P h low oh
Ipomoea coccinea Convolvulaceae t.Am A v low bm
Medicago sativa Fabaceae Ea P h low oh
Phragmites australis Poaceae E/U S P g low oh
Plantago lanceolata Plantaginaceae Ea P h low bm, oh
Polygonum convolvulus Polygonaceae Ea A h low oh
Solarium nigrum Solanaceae Ea A h low oh

T nfolium arvense Fabaceae Ea A h low oh

T nfolium repens Fabaceae Ea P h low oh
Verbascum thapsus Scrophulariaceae Ea B h low bm
Vicia cracca Fabaceae Ea/U S P h low oh
Arctium minus Asteraceae Ea B h low-mid bm
Berteroa incana Brassicaceae Ea A/P h low-mid oh
Daucus carota Apiaceae Ea B h low-mid bm, oh
Hieracium caespitosum Asteraceae E P h low-mid oh
Hieraciumﬂagellare Asteraceae E P h low-mid oh
Hwochaen's radicata Asteraceae E P h low-mid oh
Plantago major Plantaginaceae Ea P h low-mid bm, oh
Potentilla recta Rosaceae Ea P h low—mid bm, oh
Silene Iatifolia Caryophyllaceae Ea B/P h low-mid bm, oh
T araxacum oﬂicinale Asteraceae Ea P h low-mid bm, oh
Tn'folium ratense Fabaceae Ea P h low-mid bm, oh
Epilobium hirsutum Onagraceae Ea P h rrrid oh
Leonurus cardiaca Lamiaceae Ea P h mid bm, oh
Lysimachia nummularia Primulaceae Ea P h mid bm, oh
Rumex crispus Polygonaceae Ea P h mid bm, oh
Rama obtusrfolius Polygonaceae Ea P h mid bm, oh
Solarium dulcamara Solanaceae Ea P h/s mid bm, oh
Ste/[aria media Caryophyllaceae Ea A h mid oh

T orilis japonica Apiaceae Ea A h mid bm
Hesperis matronalis Brassicaceae Ea B/P h mid-high oh
Urtica dioica Urticaccae E/US P h mid-high oh
Alliaria petiolata Brassicaceae Ea B h high bm, oh
Epipactis helleborine Orchidaceae E P h high oh
Vinca minor Apogmceae E P h gig oh
Berberis thunbergii Berberidaceae A S 5 mid bm, oh
Celastrus orbiculata Celastraccae A V v mid oh
Ligustrum vulgare Oleaceae Ea S 3 mid bm. oh
Rhamnus cathartica Rhamnaceae Ea S/T s/t mid bm, oh
Rhamnusfrangula Rhamnaceae Ea S s mid oh
Rosa multiﬂora Rosaceae A S s mid bin, oh
Lonicera maackii Caprifoliaceae A S s mid-high oh
Lonicera tatarica Caprifoliaceae Ea S s mid-high brn, oh
Elaeagnus umbellata Elaeagnaceae Ea T t low oh
Maclura pmfera Moraceae S US T t mid brn

 

 

70

Results

A total of 51 non-native species from 26 families were observed in this study
(Table 3.1). These included nine annuals, seven biennials, 25 herbaceous perennials,
eight shrubs, and two tree species. The majority (78%) of these species are of Eurasian
origin. Six species (11%) were found only in beech-maple forests, and 25 (49%) only in
oak-hickory forests. Only three species (A. petiolata, Epipactis helleborine, and Vinca
minor) were highly shade-tolerant, while 28 species (55%) had low or low-mid shade
tolerance.
Herbs

Non-native stem density at the edge differed from interior values only
occasionally (Table 3.2). Signiﬁcant edge effects varied between the exterior plots (ep)
and 85 m, and were generally greater for oak-hickory forests (11.3 m) than for beech-
maple forests (4.6 m). In beech-maple forests, edge effects in the herb layer were small
for all shade tolerance categories; only the outside (ep) or edge (0 m) plots were
signiﬁcantly different from interior values (Table 3.2). Even highly shade-tolerant
species were usually restricted to within 5 m of the edge (Figure 3.1). Edge effects
determined from non—native herb cover were very similar to those determined from stem
density (Table 3.2): non-native cover was restricted to the immediate edge (Figure 3.2).

In oak-hickory forests, species with low to mid shade tolerance were also
restricted to the edge, although they penetrated up to 20 m on southern edges (Table 3.2,
Figure 3.1). Highly shade-tolerant species were present throughout oak-hickory forests
along north and west edges (Figure 3.1). This trend was driven by the great abundance of

A. petiolata at one site (Aman Park). At other sites, highly shade-tolerant species were

71

restricted to within 25 m from the edge. Edge effects for cover were much smaller (ep to
15 m) than those from stem density (ep to 85 m), especially for highly shade-tolerant
herbs.

Table 3.2: Edge effects (m) as measured by changes in non-native species stem density

and cover. Shade tolerance categories are described in the methods. In oak-hickory
forests, non-native trees were found only along northern and southern aspects. Edge

 

 

 

 

 

 

 

 

effects are signiﬁcant at p=0.05; values in parentheses are signiﬁcant at p=0.10. -- = no
signiﬁcant effect, ep=exterior plots.
Lifeforrn Shade Beech-maple Oak-hickory
Tolerance N E S W N E S W
Stem Herb low -- -- -- -- ep -- 3p --
density low-mid ep 0 0 ep -- -- 20 --
mid ep 0 (O) O -- -- -- --
mid-high -- -- -- -- -- -- -- --
high 0 0 -- -- ep 25 10 85
Shrub mid -- --* l 0 -- l 5 5 0(5)* --
mid-high -- -- -- -- -- -- -- --
Tree low -- -- -- -- ep -- -- --
mid -- -- -- -- -- -- -- --
Cover Herb low -- -- ep -- -- -- ep --
low-mid ep ep ep ep -- -- ep --
mid ep 0 -- 0 -- -- -- --
mid-high -- -- -- -- -- -- -- --
high 0 -- -- -- -- -- 10 15
Shrub mid -- --* 20 -- 50 5 0* --
mid-high -- -- -- -- 10* ep -- ep
Tree low -- -- -- -- ep -- 5 --
mrd -- -- —- ep -- -- ep --

 

 

 

 

 

 

 

*Depth of edge inﬂuence was determined by comparisons to the 90 m value instead of
the 95 m value.

Shrubs and trees

In the beech-maple shrub and tree layers, signiﬁcant edge effects were found only
along south edges for shrubs (stem density and cover) and west edges for trees (cover
only, Table 3.2). Shrub edge effects (10-20 m) were larger than those for the herb layer
(ep to 0 m). Very few (<40 stems) non-native shrubs or trees were found along east and

west edges (Figure 3.3). In oak-hickory forests, shrubs and trees were common along all

72

edges (Figures 3.3 and 3.4), and signiﬁcant shrub layer edge effects were found for all
aspects (Table 3.2). Effects determined using shrub cover were somewhat larger (ep to
50 m) than those determined using stem density (0-15 m). Signiﬁcant effects in the tree
layer were limited to northern and southern edges since non-native trees were not found
in east and west edges. Though signiﬁcant, the edge effects were small (ep to 5 m) for
both cover and stem density.

Abundant species

The most common non-native herb, at least in oak-hickory forests, was A.
petiolata, which was present in 29% of oak-hickory plots. While the main effect of
distance was signiﬁcant only for stem density, the interaction term was always signiﬁcant
(Table 3.3). No signiﬁcant edge effects were found along east edges (Table 3.4), and the
species was most common along north and west edges (Figure 3.5). The great variability
in all three variables (height, stem density, and cover) along these edges made trend
detection difﬁcult.

Rosa multiﬂora was the most common non-native shrub (present in 19% of plots)
and was present at every site. In fact, it was the only non-native species common enough
in beech-maple forests to allow for statistical analysis. While distance was highly
signiﬁcant in both forest types, aspect was not signiﬁcant in oak-hickory forests (Table
3.3). Small (0-10 m) but signiﬁcant edge effects were found in north and south edges of
beech-maple forests. Larger edge effects (0-35 m) were found in oak-hickory forests

along north, east, and south edges (Figure 3.5).

73

3.5

Beech-Maple Forests

 

3.0 <
2.5 «
2.0 -
1.5 <
1.0 «
0.5 .
0.0

Stem density (stems/m2)

North

Oak-Hickory Forests

 

/'

 

3.5

vvv-vvvvv‘vvvvvvvvvvv

 

3.0 .
2.5 1
2.0 .
1.5 .
1.0 ,
0.5 .

Stern density (stems/m2)

0.0 i—O‘A-O-O-O-m—G-H-O—O—O—O—O—O—H—O—

North

\/\JN/

o—c’.

\

\

F

 

East

 

vvvvvvvvvv—vvvvvvv“"v

 

 

3.5
3.0
2.5 «
2.0 «
1.5 <
1.0 ~
0.5 <

0.0 «
3.5

Stem density (stems/m2)

South

 

 

3.0 l
2.5 .
2.0 .
1.5 .
1.0 .
0.5 .

 

Stem density (stems/m2)

0.0

West

 

 

 

 

‘s-AA--‘A--A-AAAAA!A‘A

 

 

Low
Low-mid
Mid
Mid-high
High

vvvvvvvvvvvvvvvvvvvvv

Distance from edge (m)

Figure 3.1. Non-native herb stem density (stems/m2) by shade tolerance category
(described in methods) in beech-maple and oak-hickory forests.

74

Beech-Maple Forests Oak-Hickory Forests

 

0.30
025 . North . North

 

0.20 «
0.15 .

0.10 ‘ '\

0.05. \ . f
'\ f
0.00 4):-eeeeeeeeeeeeeeecee ‘m/OHHOO\W:$:

0.30
0,25 l 1 East . East

Cover (%)

 

 

 

0.20. 1
0.15. l
0.10.
0.05.

0.00 Aesssﬁ‘eseeeeeeeeeeee ==’5"=’5‘2~:-:—::4.=:==:====-
0.30
0.25. South . South

Cover (%)

 

 

 

 

0.20 <
0.15 <
0.10 «
0.05 - 0 i

o k-AAALL-A--AA‘A‘A‘A ‘W
. vvvvvvvvvvvvvvvvvvivv

0.30
0.25 . West 1 West

Cover (%)

 

 

 

0.20 .
0.15 .
0.10 .
0.05 i \ .

Cover (%)

 

 

 

 

DC’

I vvvjvvv v v 'v""""‘-

Vvvvvvvvvvvvvvvvvvvvv

 

 

Distance from edge (m) Distance from edge (m)
———0— Low
0 Low-Mid
— — + — — Mid
— —o — — Mid-High

Figure 3.2. Non-native herb cover (%) by shade tolerance category (described in
methods) in beech-maple and oak-hickory forests.

75

1 O Beech-Maple Forests Oak-Hickory Forests

 

 

0.8 4 North 4 North

0.6 .
0.4 ~

0.2 4

Stem density (stems/m2)

 

 

 

0.0
1.0

 

 

 

0.8 4
0.6 '1
0.4 4

0.2 l 4
g v v v '1 V v v v vﬁ v v v v V v v v v

1.0

Stem density (stems/m2)

 

 

 

vvvvvvvrvvvvvvvvvvvvv

 

 

0.8 i

0.6 <

 

 

 

Stem density (stems/m2)

 

 

 

West . West

0.8 ‘

0.6 <

0.4 .

02‘ ‘
O

“AAAAA-AAAAAAA‘A‘AAA _
. VVVVVVVVVVVVVVVWWVVVV

0 20 40 60 8O 1 00 0 20 40 60 80 1 00

 

 

Stem density (stems/m2)

 

 

 

Distance from edge (m) Distance from edge (m)
o Shrub, Mid-High
— — —.— —— - Tree, Low
— —¢ — Tree, Mid

Figure 3.3. Non-native shrub and tree stem density (stems/m2) by shade tolerance
category (described in methods) in beech-maple and oak-hickory forests.

76

o 16 Beech-Mame Forests Oak-Hickory Forests

 

 

North North

0.12 4

0.08 4

Cover (%)

0.04 «

 

 

 

 

0.00 .
0.16

 

 

East
0.12 ~

0.08 .

Cover (%)

0.04 <

0.00 W

0.16

 

 

 

 

 

South
0.12 «

0.08 «

0'04 ‘ AA
0.00 a LA AAAAAAAAAAAAAA

vvvvvvvvvvvvvvvvvvvvv
0.16

Cover (%)

 

 

 

 

 

 

 

 

West West
0.12 . .

0.08 . .1

Cover (%)

0.04 . .
\ o

0.00 V'vv‘vvvvvvvvvvv’vivvvv vvvvvvvvvvvvvvvvvvvvv

0 20 40 60 80 100 0 20 40 60 80 100
Distance from edge (m) Distance from edge (m)

 

 

 

 

 

 

o Shrub, Mid-High
— — —.— — - Tree, Low
— —o -— Tree, Mid

Figure 3.4. Non-native shrub and tree cover (%) by shade tolerance category (described
in methods) in beech-maple and oak-hickory forests.

77

80

 

Alliaria petiolata

Stem density (stems/m2)

 

25

 

 

 

20-

Stem density (stems/m2)

 

 

 

Stem density (stems/m2)

 

Distance from edge (m)

Figure 3.5. Stem density trends for three common non-native species in oak-hickory
forests. Note the changes in scale between species.

78

 

 

 

-—_.-'————
__ ._¢ _—

Nonh
East

South
West

Nonh
East

South
West

Nonh
East
South
West

At least one of the two Lonicera species observed in this study (L. maackii and L.
tatarica) was found in every site and both were relatively common in oak-hickory forests
(13% of plots). In beech-maple forests, however, they were rare, with only two
individuals (found at the same site). In spite of their abundance in oak-hickory sites, only
aspect was signiﬁcant as a main effect and no signiﬁcant edge effects were found (Tables
3.3 and 3.4). Most Lonicera cover was found in one site (Rose Lake) along a north edge.
Table 3.3: ANOVA results (p values from F-tests) for abundant non-native species.

Unless otherwise speciﬁed, data are for oak-hickory forests only. Values in parentheses
are not signiﬁcant.

 

 

 

 

 

 

 

 

 

 

Species Model term Height Stem Density Cover
Alliaria petiolata Distance (0.43) 0.0002 (0.33)
Aspect 0.0025 <0.0001 0.0088
Distance x aspect 0.0015 0.0056 0.0056
Berberis thunbergii Distance 0.0059 0.036 (0.67)
Aspect (0.07) (0.57) 0.016
Distance x aspect 0.0010 (0.41) (0.86)
Lonicera spp. Distance (0.99) (0.99) (0.99)
Aspect 0.014 (0.16) 0.019
Distance x aspect (I) (1) (0.99)
Rosa multiflora Distance <0.0001 * 0.0003
(beech-maple) Aspect <0.0001 * <0.0001
Distance x aspect <0.0001 * <0.0001
Rosa multiﬂora Distance <0.0001 <0.0001 <0.0001
(oak-hickory) Aspect (0.28) (0.3 l) (0.1 7)
Distance x aspect (0.14) 0.0007 (0.11)
Rhamnus cathartica Distance (0.25) (0.99) (0.47)
Aspect 0.0053 (0.13) 0.0066
Distance x aspect (0.21) (0.99) (0.058)

 

 

*Stem density in beech-maple forests could not be analyzed, because the statistical
model did not converge.

Berberis thunbergii was another frequently encountered shrub (present in 6 sites),
but was rarely common (31 total individuals). Signiﬁcant edge effects (5-10 m) were
found for height (north and east) and cover (north, Table 3.4, Figure 3.5). Rhamnus

cathartica was encountered just as frequently as B. thunbergii but was more common (68

79

individuals). Due to its highly variable distribution, however, signiﬁcant edge effects
were found only for cover in south edges (ep).

Table 3.4: Edge effects (m) for abundant species. Unless otherwise speciﬁed, data are for
oak-hickory forests only. Edge effects are signiﬁcant at p=0.05; values in parentheses are
signiﬁcant at p=0.10. -- = no signiﬁcant effect, ep=exterior plots, bm=beech-maple data,
oh=oak-hickory data.

Species Metric N E S W
Alliaria petiolata Height -- -- -- ep (5)
Stem density 5 (10) -- -- 80
Cover -- -- 10 --
Berberis thunbergii Height 10 5 -- --
Stem density 5 -- -- --
Cover -- -- -- --
Lonicera spp. Height -- -- -- --
Stem density -- -- -- --
Cover -- -- -- --
Rosa multiﬂora (bm) Height 0 -- 10 --
Stern density * * * *
Cover -- -- 5 --
Rosa multiﬂora (oh) Height -- -- -- --
Stem density 15 (25) 5 35 --
Cover 0 -- -- --
Rhamnus cathartica Height -- -- -- --
Stem density -- -- -- --
Cover -- -- ep --
*Stern density in beech-maple forests could not be analyzed, because the statistical
model did not converge.

 

 

 

 

 

 

 

 

 

 

 

 

Eﬂects of species characteristics

In oak-hickory forests, average herb shade tolerance was generally lower along
the edge than in the interior (Figure 3.6). Signiﬁcant edge effects were found for north (0
m), east (ep), and south (ep) edges. The consistently high shade tolerance of non-native
herbs in west edges is largely due to the dominance of A. petiolata in one site (Aman
Park). The shade tolerance of shrub species differed little with distance from the edge
(Figure 3.6) because all shrub species found in this study fell into only two shade

tolerance categories (mid and mid-high). No signiﬁcant edge effects were found. Only

80

two non-native tree species (Elaeagnus umbellata and Maclura pomifera) were found.
These species occurred only along north and south edges. Tree shade tolerance increased
with distance along south edges, with a signiﬁcant edge effect up to 10 m. Along north
edges, however, there was much greater variability, and the shade tolerance actually
decreased slightly with distance into the forest (Figure 3.6). Shade tolerance could not be
analyzed in beech-maple forests, because too few non-native individuals were found.

Species characteristics (lifeforrn and shade tolerance) had signiﬁcant effects on
non-native cover and stem density in both forest types (Table 3.5). When considering
each aspect within a forest type separately, the lifeform and shade tolerance terms were
always highly signiﬁcant. However, lifeforrn and shade tolerance explained very little of
the variation in stem density or cover (Table 3.6). In beech-maple forests, R2 values were
always less than 0.04. Values in oak-hickory forests were higher, but only in the west
edges did lifeforrn and shade tolerance explain more than 15% of the variation.
Eﬂects of canopy openness and distance

The addition of canopy openness almost always increased explanatory power of
the model signiﬁcantly (Table 3.6). Canopy openness increased R2 values by 0.5 to 0.41
(signiﬁcant increases only) in beech-maple forests. Increases were larger for north and
east edges (0.33) than for south and west edges (0.09). Increases for oak-hickory forests
were more similar among aspects and were also smaller, varying between 0.1 and 0.13

(ave. 0.04).

81

 

 

 

 

 

 

 

 

 

 

 

 

Herbs
5 aaooooooooaoaaaooooa
a) A . O . .0. .
8 41 a‘ '
g 0 North
6 3. v v H 0 East
a C v'
‘8 ' v South
2 2-8 "" A West
(0 V
1.
0 T r
5
Shrubs
4< A .
a: co
0 . O . . . C
c o 0 0 0 0
g 3.5883808899¢,,¢Xv"gxz
.'°.'.
8 2~
(0
.C
o)
14
0 -
5
Trees
4.
o
8
m , 00 o '
E 3 o O':Oooo '
g2 . v v 0' o
E x" .V V .
m o
14 v . o
0 . . T . .
0 20 40 60 80 100
Distance from edge (m)

Figure 3.6: Shade tolerance of non-native species in oak-hickory forests. Species with
low shade tolerance (1) are found only in open ﬁelds, roadsides, clearings, and similar
areas. Species with mid shade tolerance (3) are found in open woods, wood edges, trails
within woods, and thickets. Highly shade-tolerant species (5) are found in forest interiors
and will ﬂower under a closed canopy. Shade tolerance ranges were assigned when
species are found in a variety of shade levels. Non-native trees were found only along

north and south edges.

82

Table 3.5: Non-native stem density and cover ANOVA results. P values are from Type
III F -tests for signiﬁcant effects of the model terms. Values in parentheses are not
signiﬁcant.

 

 

 

 

 

 

 

 

Beech-Maple
Model term N E S W
Stern Distance <0.0001 <0.0001 <0.0001 <0.0001
density Lifeform <0.0001 <0.0001 <0.0001 <0.0001
Shade tolerance (ST) <0.0001 <0.0001 <0.0001 <0.0001
Lifeform x distance <0.0001 <0.0001 <0.0001 <0.0001
Distance x ST <0.0001 <0.0001 <0.0001 <0.0001
Lifeform x ST <0.0001 <0.0001 <0.0001 <0.0001
Distance x lifeforrn x ST <0.0001 <0.0001 <0.0001 <0.0001
Cover Distance <0.0001 <0.0001 <0.0001 <0.0001
Lifeform <0.0001 <0.0001. 0.0104 <0.0001
Shade tolerance (ST) <0.0001 <0.0001 0.0002 <0.0001
Lifeform x distance <0.0001 <0.0001 <0.0001 <0.0001
Distance x ST <0.0001 <0.0001 <0.0001 <0.0001
Lifeform x ST <0.0001 <0.0001 <0.0001 <0.0001
Distance x lifeform x ST <0.0001 <0.0001 <0.0001 <0.0001
Oak-Hickory
Model term N E S W
Stern Distance (0.074) 0.0004 <0.0001 (0.23)
density Lifeform <0.0001 0.0004 <0.0001 <0.0001
Shade tolerance (ST) <0.0001 <0.0001 <0.0001 <0.0001
Lifeform x distance (0.16) <0.0001 <0.0001 (0.81)
Distance x ST <0.0001 <0.0001 <0.0001 (0.89)
Lifeform x ST <0.0001 <0.0001 <0.0001 <0.0001
Distance x lifeforrn x ST 0.0009 <0.0001 <0.0001 (0.94)
Cover Distance <0.0001 0.0058 <0.0001 (0.72)
Lifeform <0.0001 0.0020 <0.0001 <0.0001
Shade tolerance (ST) <0.0001 0.0003 <0.0001 <0.0001
Lifeform x distance <0.0001 0.0003 (0.98) (0.39)
Distance x ST <0.0001 0.0008 (0.38) (0.39)
Lifeform x ST <0.0001 <0.0001 <0.0001 <0.0001
Distance x lifeforrn x ST <0.0001 <0.0001 (0.12) (0.55)

 

 

 

While R2 values for lifeforrn, shade tolerance, and canopy openness together were
as high as 0.45 for beech-maple forests and 0.42 for oak-hickory forests, distance
increased explanation signiﬁcantly in all but one case. Signiﬁcant increases in R2 ranged
from 0.01 to 0.05 for beech-maple forests, and from 0.01 to 0.06 for oak-hickory forests.

Including the interaction between openness and distance increased model explanation

83

further, with signiﬁcant increases of 0.06-0.12 and 0.005-0.13 for beech-maple and oak-
hickory forests, respectively.

Table 3.6: Cumulative R2 values ﬁ'om Type 1 ANOVA. R2 values represent the
proportion of variation explained by each variable and all the preceding variables. The
openness, distance, and openness x distance (0 x D) categories include all interactions
with lifeforrn and shade tolerance (ST). Non-signiﬁcant values (in parentheses) indicate

that a variable does not signiﬁcantly increase the explanatory power of the model.
**p<0.001, *p<0.01.

 

 

 

 

 

 

 

Beech-maple

Variables N E S W
Cover Lifeform and ST 003’” 0.04" 0.02** 0.04**
Openness 0.34** 0.34** 0.08" O.10**

Distance 0.39" (0.34) 0.1 1** (0.1 1)
O x D 0.45** 0.41** 0.15** 0.23“
Stem Lifeform and ST 003“ 0.04“ 0.02** 0.04"
density Openness 0.34** 0.45" 0.17“ 013'”
Distance 0.38“ 0.46** 0.18** 0.14**
O x D 0.44“ 0.56** 0.28" 0.28**

Oak-hickory

N E S W
Cover Lifeform and ST 011’” 0.05** 0.15** 0.24**
Openness 0.14** 0.11" 0.17" 0.25“
Distance 0.15** 0.14** 0.19" 0.29"
O x D 0.17** 0.22** (0.19) 0.42"
Stem Lifeform and ST 012” 0.09** 0.06“ 0.39**
density Openness 0.143“ 0.14** 0.19** 0.42"

Distance 0.147* 0.20** 0.21 ** 0.43 *
O x D 0152* 0.28** 0.24** 0.58"

 

 

 

Average R2 values were slightly higher in beech-maple forests (0.22) than in oak-
hickory forests (0.21). The model for non-native abundance (both cover and stem
density) in beech-maple forests explained more of the variation in north and east edges
(0.45 and 0.49) than in south and west edges (0.22 and 0.26). In oak-hickory forests, on
the other hand, R2 values were highest for west edges (0.50); north (0.16), east (0.25),

and south edges (0.22) were more similar.

84

Discussion

Of the 216 vascular plant species recorded in the 10 sites, 51 (24%) were not
native to Michigan. These data overestimate the true proportion of non-native species: all
non-natives present in the grid-transects were recorded, whereas native species were
sampled only in 1 m2 plots. Furthermore, most native spring ephemerals were no longer
present when the sites were sampled. More non-native species (43) were found in oak-
hickory forests than in beech-maple forests (26). The proportion of non-native species
found in this study was within the high end of the range reported for other deciduous
forests in Michigan (0-25%, Table 1.1).
Herbs

Non-native herb presence in beech-maple forests was very low (12% of plots) and
80% of species were restricted to within 10 m of the forest edge (Figures 3.1 and 3.2).
Vegetation responses to edge effects were predicted to be greatest along south and west
(warm) edges since abiotic edge effects penetrate farthest into these edges. However,
herb edge response was uniformly small (ep or 0 m, Table 3.2). In contrast, Brothers and
Spingarn (1992) noted that the ﬁequency and richness of non-native species (all
lifeforrns) were higher along warm edges than along cool edges in old-growth forests in
Indiana. They too found that overall non-native presence was very low. In mixed
hardwood forests in North Carolina, Fraver (1994) found fewer non-native species in
north edges than in south edges, and edge effects based on non-native cover were greater
in south edges (60 m vs. 10 m for north edges).

With the exception of highly shade-tolerant species, herb stem density and cover

in oak-hickory forests were small and response to edge effects was often not signiﬁcant.

85

Most (40%) of the signiﬁcant edge effects were found on south edges (Table 3.2). There ’
were two highly shade-tolerant herb species: E. helleborine, an orchid, and A. petiolata,
the biennial garlic mustard. Only two E. helleborine individuals were found at one site
(Clear Lake); most of the data on highly shade-tolerant herbs therefore reﬂect patterns of
A. petiolata abundance. As might be expected of a shade-tolerant herb, A. petiolata
abundance was usually low close to the edge (ep to 0 m), but it increased rapidly within
the next 10 m (Figure 3.5). Although it is highly shade-tolerant, A. petiolata growth and
ﬂowering increases with greater light availability (Meekins and McCarthy 2001). Edge
penetration for A. petiolata was the largest for any non-native species (ep to 80 m), but
was greater for stem density than for cover (Table 3.4). In many plots, most of the A.
petiolata stems were ﬁrst-year rosettes and, although stem density was often very high,
the small rosettes contributed very little cover.
Shrubs and trees

In beech-maple forests, only two shrub species (R. multiﬂora and L. vulgare)
were observed beyond 25 m. While the R. multiﬂora individuals were all small (<1%
cover), a large L. vulgare shrub was observed at approximately 100 m (east edge, Figures
3.3 and 3.4). Although this was categorized as a mid-shade-tolerant species, the shrub
was ﬂowering and successfully producing ﬁ'uit. The presence of two gaps just outside
the transect may have provided enough light for the shrub to survive.

Shrub response to edge effects in beech-maple forests was largest (10 and 20 m
for stem density and cover, respectively) along south edges; in fact, these were the only
edges for which signiﬁcant edge effects were found. Only one signiﬁcant effect (ep),

along west edges, was found for trees. While these results support the hypothesis that

86

edge effects would be largest along warm edges, the paucity of signiﬁcant effects make
the results difﬁcult to generalize.

Non-native shrub and tree species were more frequent in oak-hickory forests
(23% of plots) than in beech-maple forests (4% of plots, Figures 3.3 and 3.4). Signiﬁcant
edge effects ranged from ep to 50 m, although most were fairly small (ep to 5 m, Table
3.2). Shrubs with mid shade tolerance, particularly R. multiﬂora, were the most
abundant. The largest edge effect was found for shrub cover (50 m, mid shade tolerance)
on north edges, which contradicted expectations. Site effects may have inﬂuenced these
results: Rose Lake, a state game area where north and south edges were sampled, was
highly invaded by R. multiﬂora.
Effects of species characteristics

The shade tolerance of non-native herb species increased with distance from the
edge, at least in oak-hickory forests (Figure 3.6). Many authors have noted that shade-
intolerant species (native and non-native) rarely penetrate far into forests, while shade-
tolerant species are less common near forest edge (Ranney et al. 1981, Whitney and
Runkle 1981, Palik and Murphy 1990, Burke and N01 1998, Meiners and Pickett 1999,
Gehlhausen et al. 2000). Similarly, this study found that only highly shade-tolerant non-
natives penetrated farther than 10 m into either forest type. Thus non-native plants
appear to be responding to environmental gradients much as native species do.

Lifeform was another important factor that inﬂuenced the density and cover of
non-native species in forests (Table 3.5). The most common non-native species in this
study were almost all shrubs or trees. Shrubs and trees have an advantage over herbs

because their greater height can allow them to escape some light limitation. Brothers and

87

Spingarn (1992) found only ﬁve non-native woody species in old-growth beech-maple
forests in Indiana, but these species were disproportionately common in the forest
interior. Of the eight non-native species present in interior forest plots in Ontario,
Canada, only R. cathartica, a small tree, had more than 6% cover (Burke and N01 1998).
The woody species found in my study may have a further advantage over most non-
native herbs: all are dispersed by birds (Brothers and Spingarn 1992, Archibold et al.
1997). While the wind-dispersed seeds of most herbaceous species do not penetrate far
into forests (Cadenasso and Pickett 2001), bird-dispersed seeds penetrate as far as their
dispersers do.
Eﬂects of canopy openness

Light availability, often cited as an important factor in determining invasion by
non-natives, may be one reason for the great difference in invasion between oak—hickory
and beech-maple forests. While Brothers and Spingarn (1992) found no relationship
between non-native abundance and light availability, they speculated that light levels
recorded throughout a greater portion of the growing season would correlate more
strongly with non-native distribution. MacQuarrie and Lacroix (2003) noted that the
abundance of non-native species was correlated with canopy cover more often than with
distance. Alliaria petiolata and L. maackii seedlings establish most successfully along
forest edges, where light availability is highest (Luken and Goessling 1995, Meekins and
McCarthy 2001).

Canopy openness (discussed in Chapter 2) in forest interior plots (90 to 100 m)
was signiﬁcantly higher in oak-hickory forests than in beech-maple forests (2.93i1.23%

vs. 2.67:1:1.30%, t-test, p=0.024). This difference, while statistically signiﬁcant, is small

88

enough that it may not have biological signiﬁcance. Furthermore, canopy openness at the
edge (0 to 10 m) did not differ signiﬁcantly, although it was slightly higher in beech-
maple forests (5.86i1.61% vs. 5.31:t1.69%, t-test, p=0.l6). However, signiﬁcant
differences in interior (90-100 m) LAI were found (beech-maple: 6.52i0.13 mz/mz; oak-
hickory: 4.45i1.l6 mz/mz), indicating that differences in foliage area do exist.
Furthermore, light availability in oak-hickory forests may be greater early in the growing
season, due to later leaf-out of canopy species. Many non-native species leaf out before
canopy trees (Anderson et al. 1996, Collier et al. 2002), and the longer shade-free
growing season in oak-hickory forests may contribute to their success. Indeed, even in
mid-summer, canopy openness was not unimportant in determining non-native
abundance. Partial R2 values for canopy openness ranged from 0.05 to 0.41 in beech-
maple forests, and ﬁom 0.01 to 0.13 in oak-hickory forests (Table 3.6). The larger R2
values for beech-maple forests may indicate that light availability is more important to
non-native species in forests with a denser canopy.

While canopy openness explains at least some of the variation in non-native
abundance, distance adds yet more information to the model. In this study, distance was
a surrogate for many factors that are inﬂuenced by edge effects, including soil moisture,
air and soil temperature, humidity, wind velocity, disturbance, and propagule availability.
These factors all play a role in determining where non-native species can invade.
Survival and reproduction of A. petiolata, for example, are higher in moist lowland plots
than in drier upland plots (Meekins and McCarthy 2001). Stapanian et al. (1998) found
more non-native species in disturbed plots than in undisturbed plots in a variety of forests

in the south- and northeast.

89

Site effects

Propagule availability is closely linked to site use, and non-native species are
often introduced along trails through natural areas. Of the four beech-maple sites, two
were university-owned natural areas and one was on private property (Table 2.2). While
there is little human traffic through any of these forests, all three are surrounded by active
crop or pasture ﬁelds. Several of the most problematic non-native species, including A.
petiolata, Lonicera spp., and R. multiﬂora, were found in both natural areas (only R.
multiﬂora was present in the private forest), indicating that propagules are reaching these
sites. The fourth beech-maple site, Ionia Recreation Area, is a state-owned park with
many hiking and horse-riding trails. While non-native species were common in the open
ﬁelds within the park (personal observation), they were almost completely absent from
the forests, even along trails. Deer were observed only in Tourney, but none of the sites
are completely fenced, so it is likely that deer are present and can act as seed dispersers in
all sites. These data suggest that beech-maple forests may be relatively resistant to
invasion by non-natives, in spite of the availability of propagules.

The six oak-hickory sites are all on publicly-owned land. The three parks (Aman
Park, Johnson Park, and Seven Lakes State Park) were observed to have higher visitation
rates than the recreation area (Bald Mountain) and the game areas (Clear Lake and Rose
Lake). Invasion of A. petiolata seemed to be related to intensity of use for the parks: the
species was most common at Aman Park, the most heavily used of the three, and
completely absent from Seven Lakes, the least visited. As a whole, however, the parks
were not more heavily invaded than the game areas. The high abundance of Lonicera

spp. at Rose Lake may be due to deer dispersal (Vellend 2002), but the absence of these

90

species at the other game area is not easily explained. The high abundance of R.
multiflora at Rose Lake is likely due to historic factors: this species was planted in ﬁelds
during the 19508 and 19608 as living fences (Ankney 1988). These examples illustrate
the importance of site history and site use in determining invasion of forest ﬁagments.
Such factors can complicate studies of these sites, because they can obscure ecological

pI'OCCSSCS.

Conclusions

Non-native presence: While 24% of species recorded in this study were not native to
Michigan, only nine species were present in more than 2% of the plots sampled.
However, six of these species (Alliaria petiolata, Berberis thunbergii, Lonicera maackii,
L. tatarica, Rhamnus cathartica, and Rosa multiflora) are known to be invasive in forest
ecosystems. More non-native species were found in oak-hickory forests than in beech-
maple forests (43 vs. 26 species).

Non-rgtiveﬁabpndance: Non-native species were more ﬁ'equent in oak-hickory forests
(38% of plots) than in beech-maple forests (11% of plots) and were also more abundant
(ave. 1.2 stems/m2 vs. 0.1 stems/m2). Edge widths estimated ﬁom non-native stem
density and cover were usually small (5 15 m), although a few larger effects were found
(25-85 m). Only species with at least mid-shade tolerance showed signiﬁcant edge
responses 2 25 m. Contrary to expectations, warm (south and west) edges were not
consistently more invaded than cool edges.

Effects of species characteristics. openness, and distance: Lifeform and shade tolerance

were almost always important in determining non-native abundance, but these variables

91

generally explained very little of the variation in abundance (ave. R2=0.09). Adding
canopy openness model increased R2 values more in beech-maple forests (0.05-0.41) than
in oak-hickory forests (0.01-0.13), possibly indicating that light availability is more
important to non-native species in forests with a denser canopy. After canopy openness
was taken into account, distance explained little variation in non-native abundance in
beech-maple forests (increased R2 by 0.07-0.11), again suggesting that light is more
limiting than other distance-correlated variables. Statistical models that included shade
tolerance, lifeform, canopy openness and distance explained less variation in non-native
abundance in oak-hickory forests (ave. R2=0.28) than in beech-maple forests (ave.
R2=0.3 5).

Site and forest type effects: Site use appeared to play an important role in determining
invasion by non-natives: sites with higher observed visitation rates were more heavily
invaded. Oak-hickory forests in this study were generally more accessible and received
more visitors than the beech-maple sites, possibly contributing to greater invasion.
Furthermore, canopy phenology could contribute to increased invasion of oak-hickory
forests because the hi gh-light growing season is longer in oak-hickory forests than in
beech-maple forests. This study also illustrated the importance of site history, especially

a history of deliberate introductions, in determining the level of invasion of a site.

92

Chapter 4
Drawing on the Bank: Non-native species in forest edge seed banks
Introduction

A seed bank consists of all viable seeds in the soil that are capable of replacing
adult plants (Baker 1989). For this reason, seed banks are important to recovery after
disturbance in a variety of ecosystems, including forests (Leckie et al. 2000). They can
also serve as reservoirs of seeds for non-native species. The number of seeds in the seed
bank represents a balance between additions from local or long-distance dispersal and
losses ﬁom germination, disease, deep burial, and predation (Pickett and McDonnell
1989). Edge effects can inﬂuence seed bank density and richness by changing the
balance between additions and losses. While some studies have examined seed-bank
edge effects in tropical forests (Saulei and Swaine 1988), few have considered these
effects in temperate deciduous forests (Buckley et a1. 1997, Landenberger and McGraw
2004)

Persistent seeds are characteristic of species that occupy habitats with temporally
unpredictable disturbance regimes (Grime 1989). Long-lived seeds allow these species to
survive long periods of unsuitable conditions. In forests, where light is limiting, many .
shade-intolerant species have long-lived seeds that germinate quickly after a gap-creating
disttu'bance (War et al. 1994). Seed banks are further supplemented by seasonal inputs
ﬁom annuals. Many of these seeds survive only until the following growing season,
causing seasonal variation in the density and richness of the seed bank (Grime 1989).

Seeds reach a forest seed bank by within-site or between-site dispersal, or they

can be remnants of an earlier successional stage. The relative importance of local and

93

long-distance dispersal depends on the isolation of a fragment as well as the dispersal
mechanism. While most seeds remain close to the parent plant, rare long-distance
dispersal events are disproportionately important for gene ﬂow, migration, and invasion
(Nathan and Muller-Landau 2000). Beatty (1991) found that the seed banks of several
different forest types in New York were more similar to seed rain than to the local
vegetation, indicating that long-distance dispersal contributes more seeds to the seed
banks than nearby adult plants do. Relatively few species (<40%) remained from earlier
successional stages. Even when seeds are successfully dispersed, they may not enter the
seed bank. McClanahan and Wolfe (1993) estimated that only 3% of bird-dispersed
seeds in an early-successional system entered the seed bank.

Seed banks in all systems differ to some extent from the aboveground vegetation,
but this difference is especially pronounced in forests (Pickett and McDonnell 1989).
Nearly 75% of vegetation species were absent ﬁom the seed bank in a Spanish Betula-
Fagus forest (Olano et al. 2002), while only 28% of aboveground species were
represented in the seed bank of a French Quercus-Carpinus forest (Decocq et al. 2004).
Tree and shrub species—common in the vegetation—were largely absent from the seed
banks of ﬁve woodlands in England, while shade-intolerant herbs were common (War et
a1. 1995). The difference between vegetation and the seed bank is due to different
strategies for survival among groups of plants. Instead of persistent seeds, most shade-
tolerant forest herbs have bulbs, corms, or roots, and many reproduce by clonal grth
(Olano et al. 2002). Some trees form seedling banks that can respond relatively quickly
to changes in light availability (F enner 1985). Seeds of these species are usually large,

providing sufﬁcient energy for seedling growth through the litter layer and for

94

establishment. Large seeds are more vulnerable to predation, but rapid germination
reduces seed loss (F enner 1985).

The degree of similarity between the seed bank and the vegetation varies with
forest age. The seed banks of early successional forests tend to be very different from the
vegetation, because many seeds of earlier seres still persist (Bossuyt et al. 2002). As
forests age, however, these seeds slowly senesce and the similarity between seed bank
and vegetation increases. For example, the seed banks of several forest types within a
single old-growth forest in Canada consist largely of native forest species that were
observed in the vegetation (Leckie et al. 2000).

Seed bank density generally decreases with succession ﬁom old ﬁeld to forest
(Bossuyt et al. 2002). As succession progresses, seed density of old-ﬁeld species
declines due to decay and lack of new inputs (Buckley et a1. 1997, Bossuyt et al. 2002).
Seed density and richness of shade-intolerant species that grow in gaps will remain fairly
constant as long as disturbance occurs ﬁequently enough for these plants to reach
reproductive maturity (Matlack and Good 1990). Oosting and Humphreys (1940)
observed the highest seed density (13181 seeds/m2) in ﬁrst-year old ﬁelds and lowest
seed densities in old-grth oak-hickory forest (1180 seeds/m2). Bossuyt et a1. (2002)
found higher seed densities in a 55-year-old F agus-Quercus forest than in 97- or 115-
year-old or ancient (>220 yrs) forests. In contrast, Landenberger and McGraw (2004)
found no consistent differences in density or richness between 6-year-old clearcuts and
nearby Acer-Quercus forests. They did, however, note a different in species
composition: clearcut seed banks were dominated by annuals, while perennials were

more common in the forest seed banks.

95

Some authors have suggested that seed density should be higher in old-growth
forests, which have a greater number of canopy gaps than secondary forests (Pickett and
McDonnell 1989, Leckie et al. 2000). Bossuyt et al. (2002) noted a slight increase in
seed density in ancient forests (>220 yrs) over younger (97-115 yrs) forests, but the
difference was not signiﬁcant. Oosting and Humphreys (1940) found a slight increase in
species richness in old-growth oak-hickory forest compared to younger forests, but no
difference in seed density.

Seed distribution within a forest can often be highly variable (Fenner 1985). For
species that lack long-distance dispersal mechanisms, seeds are clumped around adult
plants. Olano et a1. (2002) found a signiﬁcant relationship between dead stumps of Erica
arborea and seeds of that species. The authors also noted that seeds of Juncus eﬂusus
were clumped in patches of 6-8 m in diameter and suggested that these patches may be
related to the bunched growthforms of adult plants. Abiotic gradients can lead to patchy
distributions when plants preferentially grow in gaps. Matlack and Good (1990)
attributed strong clustering of shade-intolerant species in Coastal Plain forests to gap
dynamics. Bird dispersal may lead to patchy distributions of seeds beneath nest sites or
favored habitats, such as gaps (Hoppes 1988, Pickett and McDonnell 1989, Nathan and
Muller-Landau 2000).

Edge effects may also inﬂuence the spatial distribution of seeds, although few
studies have examined this inﬂuence. Forest ﬁ'agmentation increases proximity to
additional seed sources. Soderstrom (1986) found a negative correlation between seed
density and distance to a second-grth forest in an undisturbed tropical forest. This

pattern was likely caused by dispersal from the second-growth stand as well as from edge

96

species at the boundary between the stands. Buckley et al. (1997) noted a similar
decrease in seed density in several temperate forests in England. In West Virginia, seed
bank density declined signiﬁcantly across a clearcut/forest edge on west-facing forest
edges, but not across south-facing edges (Landenberger and McGraw 2004). Edge
structure may inﬂuence dispersal into a fragment: dense edges act as barriers to wind-
dispersed seeds. Compared to intact controls, signiﬁcantly more wind-dispersed seeds
were found inside experimentally thinned edges, indicating that intact edges reduced
wind speeds (Cadenasso and Pickett 2001). Seeds also penetrated farther into the forest
in thinned edges.

Numerous studies have examined the seed longevity of individual non-native
species (Baskin and Baskin 1992, Fowler and Larson 2004). The longest study,
conducted at Michigan State University, has shown that Verbascum blattaria seeds are
still viable after 120 years (Telewski and Zeevaart 2002). However, few studies have
investigated the prevalence of non-native species in forest seed banks. Leckie et a1.
(2000) found 9 (18%) non-native species in the seed bank of an old-growth deciduous
forest in Quebec, Canada. In contrast, Decocq et a1. (2004) observed no non-native or
ruderal species in the seed bank of a French forest. In greenhouse germination trials,
Hyatt (1999) determined that eight non-native species contributed almost 50% of the seed
bank of an oak-hickory forest in Pennsylvania. However, only half of those species were
observed in ﬁeld germination trials and their abundance was reduced to 20%.

Most seed bank studies determine species presence using either seed counts or
germination (Simpson et al. 1989). Seed counts require seeds to be separated from the

soil by sieving, ﬂotation, or hand-sorting (Fenner 1985). The separation and subsequent

97

identiﬁcation of seeds is labor—intensive. Small seeds are likely to be overlooked and
may be difﬁcult to identify. Seed counts must also be supplemented by tests for seed
viability using tetrazolium or germination. These tests may reveal viability rates as low
as 2% (McClanahan and Wolfe 1993).

Germination studies are less-labor intensive and measure a more ecologically
important characteristic of the seed bank: the non-dormant seed density (Fenner 1985).
Sampled soil is spread thinly onto soil in seed trays and seedlings are counted and
identiﬁed. However, some viable seed may not germinate if appropriate conditions are
not met. This method also requires considerable greenhouse or grth chamber space.
Finally, greenhouse germination may greatly overestimate the number of seeds that
would germinate after a natural disturbance. Hyatt (1999) compared germination in the
greenhouse and the ﬁeld, and found that as much as 75% of the seed bank (determined in
the greenhouse) remained ungerminated in the ﬁeld. I chose to use the germination
method for two reasons. First, a large number of samples were collected, making seed
counting impractical. Second, compared to laboratory germination tests, greenhouse
conditions are closer to those that would trigger germination of seeds in the ﬁeld.

This study investigated the inﬂuence of edge effects on forest soil seed banks in
deciduous forest fragments (including beech-maple and oak-hickory forests) in southern
Michigan. Very few studies have examined the seed banks of oak-hickory forests
(Pickett and McDonnell 1989) or seed bank edge effects in any temperate deciduous
forest type (Buckley et al. 1997, Landenberger and McGraw 2004). I addressed the

following questions:

98

What non-native species are present in the seed banks of deciduous forest
fragments in southern Michigan (including beech-maple and oak-hickory forests)?
How abundant are non-native species in forest seed banks?

Do edge effects inﬂuence seed bank density, species richness, diversity, or species
composition?

Do seed banks contribute to the invasiveness of common non-native species?

99

Methods
Site descriptions

Soil samples were collected at the following ﬁve sites (described in Chapter 2):
Aman Park, Clear Lake, Hudson Woodland, Johnson Park, and Rose Lake.
Sampling and germination

Soil samples were collected from the top 5 cm of soil in early May 2004 (Leckie
et al. 2000). Soil was removed with a hand trowel from a ~100 cm2 circle for a total
sample volume of 0.5 L. Five replicate samples were collected at each of the following
distances from the edge: 5, 10, 25, 50, 75, and 100 m. Samples were randomly located
between two parallel lines, oriented perpendicular to the edge. If no samples were
located within 10 m of either line, a sixth sample was collected 5 m ﬁ'om that line (Figure
4.1). Samples were reﬁigerated (<1 month) before further processing.

The presence of viable seeds was evaluated using germination in the greenhouse.
Soil samples were spread thinly over potting soil in 27 x 54 cm seed trays and kept moist.
The position of each tray within the greenhouse was randomly determined. Fourteen
trays with only potting soil served as controls to account for greenhouse weeds. The ﬁve
species that germinated in these trays were excluded from the analysis. Seedlings were
identiﬁed and removed as they emerged. Seedlings that could not immediately be
identiﬁed were transplanted into pots and grown until identiﬁcation was possible. When
no seedlings had emerged from a tray for at least a month, the soil was stirred and any
additional seedlings were recorded. The soil was discarded after an additional month of

no germination.

100

Canopy
Dripline

 

 

Edge
Soil Sample
Extra Sample
(5m from line)
Forest 0 O 0 CD; (75m)

E0 00 O 0 (100m)
1 i

25 m
Figure 4.1: Soil samples were collected at six distances from the forest edge. The
location of ﬁve soil samples at each distance was randomly determined. If no samples
fell within 10 m of one side of the sampling area, a sixth, non-random sample was
collected 5 m from the line.

101

Plants were identiﬁed using Voss (1972, 1985, 1992). The origin of all species
was determined using Voss (1972, 1985, 1992) and Gleason and Cronquist (1991). Some
plants, mostly grasses and sedges, could not be identiﬁed to the species level. Sedges
were divided into three groups based on leaf morphology. Several unidentiﬁed dicot
species with multiple occurrences were divided into groups based on leaf morphology
and analyzed as separate species. Seedlings that died before they could be identiﬁed
were excluded from the species-level analysis, but were included when calculating seed
density.

The sources listed above were used to assign each species a shade tolerance
category. Species with low shade tolerance are found only in open ﬁelds, roadsides,
clearings, and similar areas. Species with mid shade tolerance are found in open woods,
wood edges, trails within woods, and thickets. Highly shade-tolerant species are found in
forest interiors and ﬂower under a closed canopy. Shade tolerance ranges were assigned
when species are found in a variety of shade levels. Species were also assigned to
lifeforrn categories. These categories included annual and biennial herbs, perennial
herbs, woody species (vines, shrubs and trees), and monocots, which included grasses,
sedges, rushes, and cattails.

Statistical analyses

Species richness and Shannon diversity indices were calculated for each soil
sample. Seed density was averaged across samples at each distance and then log-
transformed to improve normality. PROC MIXED in SAS (SAS Institute 1999) was used
to test for signiﬁcant effects of distance and aspect on species richness, diversity, and

seed density. A t—test on the least squares means (lsmeans) was used to compare

102

distances within aspects to the interior value (at 100 m). Depth of edge inﬂuence (DEI)
was deﬁned as the area within which variables differed consistently from interior values.
When the interaction term was not signiﬁcant, DEI was'determined for the main effect of
distance. When appropriate, lsmeans were used to compare aspects. Species were
grouped by origin, lifeforrn, and shade tolerance (described above) and DEI within each
grouping for a category (e. g., native and non-native for origin) was calculated.

For all 13 species with more than 15 seedlings, an index of dispersion was
calculated:

IOD = 52 /xave
where 52 is the variance and x,we is the mean of seed density for a species. Because counts
are expected to follow a Poisson distribution, a species that is randomly distributed
among samples would have an IOD of one. An IOD > 1 indicates clustering, while IOD
< 1 indicates overdispersion (a regular distribution). The calculated IOD can be
compared to a Chi-square distribution with n-I degrees of ﬁeedom to ﬁnd a signiﬁcance
level. IOD values for each species were calculated across all sites and within individual
sites.

The multi-response permutation procedure (MRPP, described in Chapter 2) in
PC-ORD (McCune and Mefford 1999) was used to compare goups of plots. For this
analysis, the number of seedlings in each sample at the same distance within an edge was
averaged. Species with fewer than three occurrences (after averaging) were excluded, as
were dead seedlings and unidentiﬁable seedlings (except for those groups described
earlier). I compared groups based on forest type, site, aspect, and distance. Further

comparisons included pairs of aspects (across sites) and comparisons between the interior

103

(100 m) and all other distances. For these multiple comparisons, Bonferroni-adjusted p
values were used to determine signiﬁcance. Site groups were investigated further using
indicator species analysis (Duﬁene and Legendre 1997, described in Chapter 2).

The clustering procedure in PC-ORD was used to investigate relationships among
the plots. A ﬂexible beta (B = - 0.25) linkage method and Sorenson distance were used to
cluster the averaged distances at each edge. I chose to use seven cluster groups, because
this level of grouping maximized the number of indicator species and minimized the
average p value for all indicator species (Dufrene and Legendre (1997). All groups

consisted of at least three plots.

104

 

 

 

28 a d : m.mm 0 008822252 800.2080 0.820%

32 w a : mam m 8080.2 0.8588 Ezomzem

2mm: 2 a e 0.; m 8880a< 280.0 028.20%

32 n a e 0. 3. n 8008204... 8.205588% 280.33%

222.28 8 a e 0. _v N 08080§U 82x82 8an

28.32 a a : «.3. n 0088204 83.205830 Exagmeab

32 a a < N? m 88802 00.85 005.0225

232 a a e :3 _ 808283 828228 0.208200%

AME _— q a o. _ m N 88802 082880 8.86

28-32 a a 8m 3% v 8882?. 2052s? £808.85

3.2-28 w a 8 new m 8080A 22.5. 3.20on

28-32 00 a 0 new N 08080§O 0.28.8 880

32 8 a a o. _ N. m 888%... .9. 83¢

28-32 a a a 0.8 m 88823 08.830 038.20%

2832 a an e 2.0 v 80022882. 02.290 82%

28 am am 8 02: n 800802 3080280 828

008 2 a a 5.2: m 8882?. 8:88 eeaewtm

28-32 a a c e. _ 2 m 8882m< 282858 08228

28 2 cm 0 2.3; v 88802 28282823 233%

32 n n m wémm m 0888—88088 832: 250880;

32 a £8 : Néhm m 88820< 2.82828 8209

.. 0m .. -- odmm n 8008250 dam 88D

32 8 a 8 mm? m 8888a .1228 8> 282 80.35

32 n 8 mm mwwﬂ m 8.882% 88MB 5.820%
8:822 82 $82802 820 .20

282m 53080 88802 2280 .8820 0000 30,—. 803802 3:88.42 020055

 

 

.282 28% .20 .3283 a 2 282 08 828% 8023 @0868 803 888 08822

2085 .8080 0820 a 280 8030a 08 80:22 “882 2 282 08 880% 8822-285 3:2: 88.22 28 .8003 2:23
282 .8200 .0003 .8003 8000 2 282 08 8:822 28% 28 523 860nm .808 8:87. 28 $8820 82288 .222.“ :80 2
>80 252 08 8:822 28% 32 :23 860% 8283080 880800 :0 08 .200: 200:on 08 80820 08 609 .33 50:
80> 2 00:08:80 028% 00 8022880 .382, :0 0080 2 82880880 08822 282m .02>n> .0022 dwvomuom .8288
.288".— .2000808 80502: .202”..— .mm8wuw 882 53080 .0022 5.82020 22800008 8803.8 .3888 ”330804 .m3 2
08.2888: 88588 .oaosmnm .38}. 028% ”@8182 .28: 80> was: 85.83% as 580 .28 a 88:80
08 00820 2 820.508 08 888 860% =< .8285 008 8080.2 boo—22-0.80 .8208 288-2003 2 282 0200mm 2.0 28k

105

 

 

 

32 22 2 am We N 0088380 828025.858 88325.22

32 00 .2 8 m0 2 888022.20 2828022288080

2282.28 2 2 8 m0 2 08888< £8.38 200V

28 22 2228 8 5.0 N 800205 82.20.28. 222

22322 220 220 8 5.0 2 00822022880 8208.28.80 858.822

32 m 8 8 5.0 N 80802 58028202220220 Ezu2§o2

28-302 22 0 8 5.0 2 888.8880 588.2200 E=E2EW

28.32 m 0 8 5.0 2 88802 880.28. 8.8222ch

32 8 0 8 5.0 2 88822.20 8220223 88D

28 22 0 8m 0.N2 2 0082882 8028.80 88:00.2

28 22 0 8 0.N2 2 88832 888.28.20.22: 8.288.282

2&2 2.2 22 8 0.N2 N 888880 88.8082 88.0.2.5

- 22 - - 0.N2 n 8008028. .9. 88V

28 8 .2 8 262 2 888022.20 80.22.2028 83.80.

28 22 220 8 2 .02 n 0800832 8.80.28- 822222

28.32 220 220 8 2.02 2 08822080822 8382M 822%

28.32 00 n2 8 2 62 2 088800.20 802=8>2AQ0Q 80.80

2822.28 22 0 8 2.02 2 88288.2 822088 0.208288%!

26.32 > .2 a 2...: N 8855 .8382. 02.:

- 2.2 0 8 8.02 v 88820.8. .00 00.82200.

32 22 28 m 8.02 m 88802m< 828282 200800.

28 22.... 220 8 8.02 N 882222.800 280888 8080580.

32 2 2 8 8.02 8 0888223 88380 2.2280.

302 8 8 8 8.02 N 88888822022 8822.20.22.20; 08.220222

- - -- - 8.02 N 08008882. 20 .3082.

302 22 £8 8 8.02 2 8082085 8255.282 08.888.8U

302 22 8 8 8.02 m 88822883 828 82228.88?

32 w 0 8 2NN 2 80802 8088020828. 880.28%

2822.28 2 2 8 de m 88802 822 8.202

302 22 0 m2 de N 882820 888.8588 580.2082:

-- : : -- de v 8880.2 283082882 880

28 220 220 8 c.0N m 88832 80.88.22.283 .3852

28-302 22 2228 8 c.0N n 0800832 802M080: 8.820%

302 22 .2 8 o.0N N 0888023. 80.02.82 .2028
888022 88.2 08220002 8220. .20

888 8320 02:83 8080 028% 88 25.8- 8.2852 22288.2 828.8

 

 

80.882 2.2. 2.8.2

106

 

 

 

 

 

$3 a m c NM _ goon—o; Sci 3. 3c:
:wE-ES : a mam. ~.M _ oaooaoED 33:» 8.2.5
-- a -- -- ~.M ﬂ 883a"— dm Exacxrk
28 n a «m NM m 0802353580 8.559» #25sz
SE a a «N NM _ oncogowbom aztobmieo RES:
32 a QR ozoxu ~.M _ unoccuawﬂm Shaina 2&on
25-32 a a m NM _ 0.388% 892 uEEmBR
32 a a 0:88 NM _ unooSBm< h3==5§u =§2=§4
-- a -- -- NM _ 3888m< dm 55%me
28-32 a a a NM _ omooﬂnsm £53.33» £3.55
Bo. : a m N. M _ 8888m< 53.88». 92.3ng
25-32 a a a NM _ 08868830 5333.3 Exﬁcmetmse
ES . u a N.M _ once—€532. negation SEC
28-32 a a = NM _ 3322?. mutw=c$§e= hit
32 w a = NM _ 838m .QESMAS _SMQQEEQ
28-32 a a a NM _ ouooﬁu.m< caoxaﬂsmra EneerV
32 > a a We _ 08°85 SEA: a»:
-- a -- -- We N 33.20; .9». 50.5
28 . a = We N 08853 5:33: “Synge
.3362: . a a We N 0888M 5.3—~28 ESQ
3.3-28 u . a We ﬂ canon—Bum 3:339...» S020
28-32 a a a We N goo—233‘ .5233me EESEQSM
85.5.2 8.5.“ N Nagoomv 8% Mo
0925 £390 bingo; Ewto his“. 38 ESE 6352 358m museum

 

 

€283 3 2.3

107

Results
Seed density

A total of 1745 seedlings germinated. Seed density was highly variable among
samples (Table 4.2). Average seed densities for sampled edges (0-100 m) varied between
1677 seeds/m2 (Johnson Park west) and 15133 seeds/m2 (Hudson Woodland east). Seed
bank density was only slightly less variable when averaged across edges for each site:
values ranged from a low of 2324 seeds/m2 (Johnson Park) to a high of 8525 seeds/m2
(Hudson Woodland).
Table 4.2: Average seed bank density (seeds/m2) and total species richness for each

sampled edge (0-1 00 m) and site (both sampled edges). bm=beech-maple; oh=oak-
hickory

 

 

 

Site Edge Seed density (seeds/mz) Species richness
sampled 0-100 m Site ave. 0-100 m Site total

Aman Park 11 6452 :t 12022 4952 i 8847 34 43

(AP, oh) w 3500 :t 3501 29

Clear Lake 11 7719 i 9679 7290 :t 9007 35 46

(CL, oh) 3 6833 at 8371 28

Hudson Woodland e 15133 t 55351 8525 :1: 39135 26 35

(HW, bm) w 2129 :1: 3939 22

Johnson Park e 3344 :t 9011 2324 i 6535 28 38

(JP, oh) w 1677 :1: 1700 20

Rose Lake n 6065 d: 4250 4950 i 4080 23 37

(RL, oh) 8 3088 at 3614 28

Total 5629 3: 18581 88

 

 

 

 

 

Species richness

In all, 88 species were found (Table 4.1). Of these, 14 (16%) were not native to
Michigan (Figure 4.2). These non-native species, however, contributed 33% of the seeds
(Figure 4.2). Solanum nigrum (black nightshade) alone made up 22% of the overall seed

bank. The next two most common non-native species were Verbascum thapsus (common

108

mullein) and T araxacum oﬂicinale (common dandelion). Seven of the non-native species
were represented by only one or two seedlings. About 22% of species (20) were annuals
or biennials, 33% (29) were perennials, 19% (17) were woody shrubs, vines, or trees, and
20% (18) were monocots, a group that included grasses, sedges, and rushes (Figure 4.3).
Contribution to the seed bank by each group, however, was not closely related to the
number of species. Annuals (including S. nigrum) and biennials made up 38% of the
seed bank, perennials 13%, woody plants 10%, and monocots 29% (Figure 4.3). Of the
species that could be identiﬁed to species level, shade tolerance was generally low: 45
species (66% of seeds) had low or low-mid shade tolerance, 13 species (8% of seeds)
mid, and 14 (6% of seeds) mid-high or high (Figure 4.4).

Seed species richness for each site ranged from 35 to 46 (Table 4.2), while
richness for sampled edges (0-100 111) ranged from 20 to 35. Distribution of species
among categoﬁes (for origin, lifeforrn, and shade tolerance) was similar for all sites, but
seed density distribution vaiied greatly. Percent of non-native species was highest for
Aman Park (23%) and lowest for Clear Lake (9%, Figure 4.2). Hudson Woodland,
however, had the highest proportion of non-native seedlings (75%). The highest
proportions of annual/biennial species were found at Aman Park and Rose Lake, although
only in Hudson Woodland was more than 30% of the seed bank armual or biennial
(Figure 4.3). All sites except Johnson Park had more than 50% shade-intolerant species,

and these species contributed 48-88% of the seed density (Figure 4.4).

109

 

 

80 J - Native
Non-native
- Unknown
3
g 60 4
.C
.9
(I)
.9
8 4o —
O.
(I)
.73
o
p—
20 -
o _
AP (oh) CL (oh) JP (oh) RL (oh) HW(bm) Total
30000

 

_ Native
1: Non-native
25000 ' - Unknown

20000 -
15000 -1

10000 -

Seed density (seeds/m2)

5000 -

 

 

o-

 

AP (oh) CL (oh) JP (oh) RL(oh) HW(bm) Total

Site
Figure 4.2: Species origins for each site and across all sites. Unknowns include all plants
that could not be identiﬁed to the species level. See Table 4.2 for site abbreviations.
oh=oak-hickory, bm=beech-maple

110

 

_ Annual/Biennial

V
80 ' Perennial herbs §
- Woody \

k

\\\\\V Monocot

 

 

”’ _ Unknown
§ 60 _
.C
2
U)
.9
8 40 —
D.
1’
g
p.

20 -

o _

AP(oh) CL(oh) JP(oh) RL(oh) HW(bm) Total

30000

- Annual/Biennial
m Perennial herbs
25°00 ' - Woody

 
    
 

A m Monocot
NE - Unknown
3 20000 -

3

o

1?,

a 15000

'5

C

m

'5

'§ 10000 -

U) ‘1‘1'3'2'2‘2‘:

5000 -

........
t... '''''

 

 

 

0_ ix
AP(oh) CL (oh) JP(oh) RL(oh) HW(bm) Total

Site
Figure 4.3: Lifeforms of seed bank species. Woody plants include shrubs, trees, and
vines. The monocots category includes grasses, sedges, rushes, and T ypha sp. Unknown

species for which lifeforrn could be determined are included in the appropriate category.
See Table 4.2 for site abbreviations. oh=oak-hickory, bm=beech-maple

lll

 

_ Low
80 “ Mid
- High
[:1 Unknown

60-

40-

Total species richness

20-

 

0 _
AP (oh) CL (oh) JP (oh) RL (oh) HW (bm) Total
30000

- Low
[2:] Mid

25000 - - High I
:l Unknown

20000 -

 

15000 i

10000 1

Seed density (seeds/m2)

5000 i

oi

 

 

 

AP (oh) CL (oh) JP (oh) RL(oh) HW(bm) Total

Site
Figure 4.4: Shade tolerance for seed bank species. ‘Low’ includes both the low and low-
mid shade tolerance categories; ‘high’ includes the mid-high and high categories. Shade
tolerance categorization is based on verbal descriptions of species occurrence in Voss
(1972, 1985, 1996) and Gleason and Cronquist (1991), and on personal observation.
Species with low shade tolerance are found only in open ﬁelds, roadsides, clearings, and
similar areas. Species with mid shade tolerance are found in open woods, wood edges,
trails within woods, and thickets. Highly shade-tolerant species are found in forest
interiors and ﬂower under a closed canopy. Unknown species could not be identiﬁed to
the species level. See Table 4.2 for site abbreviations. oh=oak-hickory, bm=beech-maple

112

Species distribution

Index of dispersion (IOD) values indicate that almost all of the common species
were signiﬁcantly clustered (Table 4.3). Signiﬁcant IOD values ranged from 1.14 to
234.4, with a median of 2.54. Solanum nigrum was by far the most highly clustered
species; T. oﬁ‘icinale was only marginally clustered. Only Juncus tenuis was clustered at
all sites; most other species were clustered at three or fewer sites. Solidago altissima was
signiﬁcantly clustered only when all sites were considered together.
Table 4.3: Index of dispersion (IOD) for the most common seed bank species.
Signiﬁcant p values indicate that seedlings are more clustered (among plots across all
sites) than expected. Species were also analyzed for clustering within sites; sites where

signiﬁcant clustering was found (a = 0.05) are indicated. HW is a beech-maple forest; all
other sites are oak-hicko forests. See Table 4.2 for site abbreviations.

 

 

 

 

 

 

 

Species IOD 12 p Sites where clustered
Solanum nigrum 234.4 72444 <0.0001 CL, HW, RL

Juncus tenuis 28.6 8850 <0.0001 AP, CL, HW, JP, RL
Carex spp. 7.30 2255 <0.0001 AP, CL, HW, RL
Rhus typhina 3.49 1079.1 <0.0001 AP, CL

Verbascum thapsus 3.41 1055.1 <0.0001 AP, CL, RL

Conyza canadensis 2.92 903.4 <0.0001 CL, RL

Rubus occidentalis 2.54 784.3 <0.0001 AP, CL, JP

Rubus allegheniensis 2.24 692.3 <0.0001 AP, HW, RL
Erigeron annuus 1.72 531.3 <0.0001 AP, RL

Solidago altissima 1.28 379.4 0.004 none

Solidago canadensis 1.24 381.7 0.003 CL

T araxacum oﬁ‘icinale 1.14 352.0 0.046 JP

T ypha sp. 1.02 316.2 0.38 none

 

Eﬂects of aspect and distance

 

Aspect had highly signiﬁcant effects on species richness and diversity, as well as
on seed density (Table 4.4). The interaction between aspect and distance was never
signiﬁcant. Species richness and diversity were higher in north and south edges than in
east or west edges (Figure 4.5). Seed density in north edges was signiﬁcantly higher than

in west edges (Figure 4.6). When seed density was examined using categories based on

113

origin, shade tolerance, and lifeform, only native and perennial species showed
signiﬁcant responses to aspect (Table 4.4). For native species, seed density in north
edges was signiﬁcantly higher than in west edges (p=0.007). For perennial species,
differences were found among the following aspect pairs: east and north (p=0.005), north
and south (p=0.009), and north and west (p=0.008).

Distance did not signiﬁcantly inﬂuence species richness or diversity (Table 4.4).
However, species richness showed a decreasing trend with distance into the forest (Figure
4.7). Seed density, however, was signiﬁcantly inﬂuenced by distance: density in the 5
and 10 m plots was signiﬁcantly higher than values at the 100 m plots (Figure 4.6).
When seed density was examined using categories based on origin, shade tolerance, and
lifeforrn, signiﬁcant trends with distance were found only for native species and for
monocots (Table 4.4). For both groups, seed density at 5 m was signiﬁcantly higher than
density at 100 m (p<0.01).

Table 4.4: P values from F tests for signiﬁcant effects of distance, aspect, and the
distance x aspect interaction (D x A) for all sites. Values for seed density were divided

into categories based on origin, shade tolerance, and lifeforrn (deﬁned in methods).
Values in parentheses are not signiﬁcant.

 

 

 

 

Metric Distance Aspect D x A
Species richness (0.089) 0.0005 (0.29)
H (0.63) <0.0001 (0.069)
Seed density 0.015 0.049 (0.93)
Native 0.004 0.036 (0.57)
Non-native (0.16) (0.26) (0.14)

 

Low shade tolerance (0.086) (0.33) (0.20)
Mid shade tolerance (0.81) (0.58) (0.056)
High shade tolerance (0.63) (0.31) (0.70)

 

 

 

Annual (0.1 1) (0.32) (0.17)
Perennial (0.44) 0.010 (0.88)
Woody (0.60) (0.84) (0.86)
Monocot 0.010 (0.1 l) (0.54)

 

114

A
.3
C

 

 

 

 

 

 

 

8
gm €0.8-
if $0.6
.32. g
8 £204.
9- ‘o
"’1. 8 j
§ (00.2
(I)

0. 0.0l

 

N E S W
Aspect Aspect
Figure 4.5: Average seed bank species richness and diversity (H) (per sample) by aspect

for all sites. Samples were 0.5 L (5 cm deep). Bars with the same letter are not
signiﬁcantly different.

4.6

44 _ —0— East

 

4.2 -
4.0 -
3.8 -
3.6 ~
3.4 <
3.2 -

Seed density (log seeds/m2)

3.0 -

2.8 | | | I I l
0 20 40 60 80 1 00

Distance from edge (m)

 

 

 

 

Figure 4.6: Seed density (log seeds/m2) by aspect for all sites. Across all aspects,
densities at 5 and 10 m were signiﬁcantly different from interior values. Densities in
north edges were signiﬁcantly higher than those in west edges.

115

4.0

 

3.5 -

3.0 -

2.5 -

Species richness per sample

2.0 .

 

 

1.5

 

0 20 40 60 80 1 00
Distance from edge (m)
Figure 4.7: Species richness per sample (across all sites and aspects) with distance from

the edge. Samples were 0.5 L (5 cm deep). No values were signiﬁcantly different from
that at 100 m.

Grouping procedures

MRPP indicated that most grouping variables, with the exception of distance,
were signiﬁcant, although the effect size was sometimes small (Table 4.5). Sites and
aspects both showed greater effect sizes than forest type. Pairwise comparisons among
aspects showed that only east and west aspects did not differ signiﬁcantly (Table 4.6).
Comparisons for all other aspects were highly signiﬁcant, and effect sizes varied between
0.028 and 0.057.
Table 4.5: Multi-response permutation procedure (MRPP) results for several grouping
variables. Within each edge, values for all samples at a distance were averaged, creating
60 sample units. Rare (<3 occurrences), dead, and unknown species were excluded,

leaving 44 species. ‘A’ is the chance-corrected within-group agreement, a measure of
effect size.

 

 

 

 

Grouping variable Number A p

of groups
Forest type 2 0.020 <0.001
Site 5 0.087 <0.001
Aspect 4 0.054 <0.001
Distance 6 0.002 0.38
Clustering 7 0.176 <0.001

 

116

 

 

 

o/

“/

/

 

 

Table 4.6: ‘A’ values ﬁ'om pairwise MRPP comparisons. Larger values indicate greater
homogeneity within groups (aspects) and greater differences between groups. Values in

 

 

0arentheses were not signiﬁcantly different ﬁ'om zero. ** Bonferroni-adjusted p<0.001
E S W
N 0.054“ 0.035" 0.028"
E -- 0.057** (0.016)
S -- 0.044"

 

 

 

 

Table 4.7: Sites, distances, and indicator species for the clusters. Percent of perfect
indication is given for each signiﬁcant indicator species (a=0.05). Non-native species are
marked with an asterisk (*).

 

 

 

 

 

Group Sites and distances included Signiﬁcant indicators

number (percent of perfect indication)

1 APN-5, 50 CLN-lOO Verbascum thapsus“ (3 8)
APW-S, 10, 25, 75, 100 JPE-25 Rubus allegheniensis (35)
HWW-75 RLN-25, 50

2 APN-10, 25 HWW-S Juncus tenuis (90)

(edge) CLN-S JPE-5 Panicum capillare (36)
HWE-S RLN-5 Verbascum thapsus“ (19)

3 APN-75 RLN-75 Glyceria striata (29)
CLN-25, 50, 75 RLS-50, 100
CLS-75, 100

4 APN-100 RLN-100 Phytolacca americana (31)
APW-SO RLS-25, 75

JPE-50, 75, 100
JPW-5, 10, 25, 50, 75

 

 

 

5 CLN-l 0 RLS-l 0 Carex spl. (74)
CLS-5, 10, 25, 50 Conyza canadensis (58)
RLN-10 Rubus allegheniensis (17)
6 HWE-l 0 JPW-100 Vitis aestivalis (43)
JPE-1 0
7 HWE-25, 50, 75, 100 Sambucus canadensis (26)

 

 

 

 

HWW-lO, 25, 50, 100

 

Clustering did not produce groups based solely on site, edge, or distance (Table
4.7). Although most plots ﬁ'om Hudson Woodland were in the same group, other sites
were divided among many groups. All the sampling units contained in Group 2 were
located within 25 m of a forest edge, but many more ‘edge’ plots were placed in other
groups. There were 16 signiﬁcant indicator species, but indicator values were generally

low. Only four species (Carex sp1., J. tenuis, S. nigrum, and V. thapsus) had values

117

above 50%. The cluster group had the highest effect size (A=0.176, Table 4.5),
indicating that these groups are more homogenous than groups created using other
variables. Indicator species for sites differed from those for the clusters (Table 4.8). Five
site indicator species were not observed in the forest understory vegetation.

Table 4.8: Indicator species for sites. Only those species with signiﬁcant indicator values

(p<0.05) are given. Non-native species are marked with an asterisk (*). 'l'These species
were not observed in vegetation under the forest canopy at a site.

 

 

 

 

 

 

 

Site Indicator Species Indicator Value
Aman Park Euthamia graminifolia'l‘ 37
Rubus allegheniensis 27
Clear Lake Carex spl. 68
Conyza canadensis 41
Scinus atrovirens‘l' 33
Duchesnea indica *‘l’ 29
Carex sp2. 23
Hudson Woodland Solanum nigrum*'l 80
Sambucus canadensis 28
Johnson Park Rhus glabra 33
Vitis aestivalis 28
Rose Lake Geum canadensis 47
Erigeron annuus‘l' 30
Aster pilosus 26

 

 

118

Discussion
Seed density

Seed density was highly variable among plots, edges, and sites (Table 4.2). The
average seed density per site (5629 seeds/m2, median 2000 seeds/m2) is comparable to
values reported in previous studies. F enner (1985) reported that seed densities for
temperate deciduous forests ranged ﬁom 1000 to 10000 seeds/m2. Leckie et a1. (2000)
found a median of 1218 seeds/m2 in an old-growth forest (more than eight forest types) in
Quebec, Canada. Olano et al. (2002) observed a total seed density of 7057 seeds/m2 in
Betula-Fagus forest in Spain. Higher densities (8296 and 12426 seeds/m2) were
observed in a French Quercus-Carpinus forest (Decocq et al. 2004) and Belgian Fagus-
Quercus forest (Bossuyt et al. 2002). War et al. (1994) found a similarly wide range
(from 3240 to 112480 seeds/m2) in ﬁve woodlands (dominated by Quercus spp., Picea
spp., Larix spp., or Prunus avium) in England.

The variability in seed bank density appears to be caused by the patchy
distribution of the more common species (Table 4.3). The species with the highest index
of dispersion (IOD) were Solanum nigrum, Juncus tenuis, and Carex spp. Olano et al.
(2002) speculated that limited long-distance dispersal mechanisms led to patches of seeds
around present or former adults for two common species. While this hypothesis was
supported by the high concentration of seeds around shrub stumps, they did not ﬁnd any
Juncus eﬁ'usus adults. It seems unlikely that inefficient local dispersal is causing
patchiness of J. tennis in my sites, because no adults were observed anywhere under the
canopy, including large gaps. Furthermore, plants were observed outside the forest

(along an unpaved road) only at Hudson Woodland. Seeds must therefore have a long-

119

distance dispersal mechanism. Wind dispersal is a possibility, because the seeds are very
small. The distribution of S. nigrum is more easily explained: seeds will be concentrated
by the animals that eat its berries and will be deposited in places that these animals favor
(Brothers and Spingarn 1992).

Most authors report seed density as the number of seeds per surface area of soil,
regardless of the volume of the sample. One review even used this method of reporting
results as a criterion for inclusion in a meta-analysis (Bossuyt and Hermy 2001). Soil is
usually collected to a depth of 4-10 cm (Matlack and Good 1990, Hyatt 1999, Leckie et
a1. 2000, Olano et al. 2002, Landenberger and McGraw 2004), but can be collected to as
deep as 20 cm (Bossuyt et al. 2002, Decocq et a1. 2004). Reported sample volume varies
between 0.2-2.0 L (Matlack and Good 1990, Hyatt 1999, Leckie et al. 2000, Bossuyt et
al. 2002, Olano et al. 2002, Decocq et a1. 2004, Landenberger and McGraw 2004). While
most seeds are concentrated in the upper 5 cm of soil, seeds can be present as deep as 20
cm (Bossuyt et al. 2002). Because the ratio of surface area to volume varies greatly
between studies, reporting density simply as seeds/m2 may under- or overestimate density
per volume.

Species richness

The observed seed bank species richness for the study sites (35 to 46, Table 4.2)
was on the high end of the range reported by other studies. Species richness in four forest
types in New York ranged from 8 to 11 (Beatty 1991). While only 19 species germinated
ﬁom the soil of an oak-hickory forest in Pennsylvania (Hyatt 1999), Olano et al. (2002)
found at least 28 taxa in a Spanish forest. The same number of species were found in a

managed forest in France (Decocq et al. 2004). Seed bank richness in different areas of

120

ﬁve woodlands in England varied between 6 and 39 (War et al. 1994). Landenberger and
McGraw (2004) observed 44 species in the seed bank of a mesophytic forest in West
Virginia. Seed bank species richness was slightly higher in an old-growth forest in
Quebec, where 49 taxa were identiﬁed (Leckie et a1. 2000). Most of these studies
focused on sampling forest interiors; the high species richness in the present study could
be due to inputs from adjacent habitats. Inputs from old ﬁelds, for example, would add to
the true forest seed bank, inﬂating the number of species.

The large proportion of shade-intolerant (46-51%) and annual/biennial seed bank
species (21-32%) found in this study is not unusual (Pickett and McDonnell 1989, War et
al. 1993). Bossuyt et al. (2002) found that about 80% of seed bank species were
characteristic of forest edges, while up to 10% of species were characteristic of disturbed
environments. In oak-pine forests in the New Jersey pine barrens, species typically found
in early successional, disturbed sites made up 50% of total species richness (Matlack and
Good 1990). Many other studies have reported that such species are ‘common’ or
‘frequently present’ (War et al. 1994, Decocq et a1. 2004, Landenberger and McGraw
2004). Proportion of species is not, however, directly related to proportion of seeds: in at
least some of the sites, the proportion of shade—intolerant seedlings was smaller than the
proportion of species (Figure 4.4).

Eﬂects of aspect and distance

Surprisingly, aspect was more important than distance in determining species
richness and diversity (Table 4.4). While species richness declined with distance into the
forest (Figure 4.7), no plots differed signiﬁcantly from the interior. MRPP analysis

conﬁrmed that, while species composition differed signiﬁcantly among forest types, sites,

121

and aspects, there were no differences with distance across all sites (Table 4.5). No
consistent patterns emerged for cool (north and east) and warm (south and west) edges:

north and south aspects had higher richness and diversity than east and west aspects
(Figure 4.5).

14000
12000 1
10000 -
8000 ~
6000 -
4000 «

2000. _
0. . wliﬁ?

 

—o— annuals
vo . monocots

Seed density (seeds/m2)

 

 

 

 

14000
12000-
10000«
8000 l
6000 -
4000 .
2000 <

 

Seed density (seeds/m2)

 

 

 

 

0 20 40 60 80 1 00

Distance from edge (m)

Figure 4.8: Seed density by distance for lifeforrn and shade tolerance categories. Density
of perennials and woody species did not change with distance. Monocot seed density
decreased with distance (p=0.0098). For annuals, the density at 10 m was signiﬁcantly
higher than values at 100 m. Densities of mid and highly shade-tolerant species did not
change with distance. For shade-intolerant species, the density at 10 m was signiﬁcantly
higher than values at 100 m. Soil was collected to a depth of 5 cm.

Seed density showed greater responses to distance than richness or diversity did.
Soil collected within 5-10 m of the forest edge had signiﬁcantly more seeds than soil

collected 100 m from the edge (Figure 4.6). Landenberger and McGraw (2004) found a

122

similar decrease in density in forests adjacent to recent (~6 years) clearcuts. When
density examined by category (origin, lifeforrn, and shade tolerance), distance was rarely
signiﬁcant (Table 4.4), possibly due to the smaller sample sizes. Monocot seed density
decreased with distance (p=0.0098) and native density at 5 m was signiﬁcantly higher
than at 100 m. While distance did not have a signiﬁcant effect on annuals as a group
(p=0.11), the value at 10 m was signiﬁcantly different from that at 100 m (Figure 4.8).
Distance was not signiﬁcant for any shade tolerance category, but seed density of shade-
intolerant species decreased with distance into the forest (Figure 4.8).

These results indicate that increased seed density at the edge is likely the result of
increased seed deposition from adjacent habitats: many of the annual, shade-intolerant
species are characteristic of open ﬁelds. These species include Conyza canadensis,
Erigeron annuus, Euthamia graminifolia, Solidago spp., and Verbascum thapsus. While
most of the monocot species could not be identiﬁed to the species level, very few grasses
and sedges, and no rushes, were observed in the study sites; most of these seeds likely
came from adjacent areas. Furthermore, T ypha sp., a wetland species, was relatively
common in the seed bank, even at sites with no adult populations in sight. Clustering,
which often separated edge and interior plots within a site, conﬁrmed these differences in
species composition. Buckley et al. (1997) also noted that seeds of shade-intolerant herbs
and grasses were more common at the edges of woodland sites. Landenberger and
McGraw (2004) observed that the densities of two common early-successional species
(Erichtites hieraciifolia and Lobelia inﬂata) declined rapidly beyond 10 m into the forest.

Non—native species

123

Few studies have reported the number or proportion of non-native species in
forest seed banks. Decocq et al. (2004) found no non-native species in a 4000 ha old-
growth Quercus-Carpinus forest in France, while Hyatt (1999) found eight species in an
old-growth oak-forest and adjacent blowdown in Pennsylvania. The highest reported
number of non-natives was for a 1000-ha old-growth forest (covering a variety of
deciduous and mixed forest types) in Quebec, where Leckie et al. (2000) observed nine
(18%) non-native species. This study found 14 non-native species (16%), seven of which
were represented by only one or two seedlings (Table 4.1). Aman Park had the highest
proportion of non-native species (23 %), followed by Rose Lake with 18% (Figure 4.2).
Most of the non-native species observed in these two sites were shade-intolerant,
annual/biennial species characteristic of the adjacent, unmanaged old-ﬁelds. Examples of
species common to both sites include Duchesnea indica, T araxacum oﬂicinale, and V.
thapsus. Additional species at Aman Park included Crepis tectorum, Hypericum
perforatum, Leontodon autumnalis, Rumex obtusifolius, and Senecio vulgaris. While
Hudson Woodland had a very low number of non-native seed bank species, it had the
highest proportion of non-native germinants. This difference was due largely to >300 S.
nigrum seedlings that germinated from one plot.

Of the 39 non-native species observed in the above-ground vegetation of these
ﬁve sites, only seven were found in the seed bank. None of the ﬁve most abundant non-
native species (Alliaria petiolata, Berberis thunbergii, Lonicera spp., Rhamnus
cathartica, and Rosa multiflora) were found in the seed bank. The lack of the shrub
species is not surprising: shrubs are generally scarce in forest seed banks (War et al.

1993). Lonicerajaponica, an invasive relative of the species observed at my sites, does

124

not form a persistent seed bank (Fowler and Larson 2004). Furthermore, all of the shrub
species observed in this study are bird-dispersed and have fairly large seeds, which
generally do not persist in the seed bank. Archibold et a1. (1997) observed that most
seeds of R. cathartica germinated within 50 days of deposition.

More surprising was the absence of A. petiolata ﬁ'om the seed bank. In the
vegetation, A. petiolata was abundant at Aman Park (97% of plots), common at Rose
Lake (31% of plots), and present at Hudson Woodland and Johnson Park. Previous
studies have reported that A. petiolata forms a persistent seed bank, with seeds
germinating for 3—4 years aﬁer sowing (Cavers et al. 1979, Baskin and Baskin 1992).
Anderson et al. (1996) noted that 10% of total germination occurred in the second spring
after sowing, the limit of their observation time.

Alliaria petiolata seeds require cold stratiﬁcation for germination (Baskin and
Baskin 1992). This requirement had clearly been met, as I observed many small rosettes
at the time of soil sampling. It is possible that most of the viable seeds had already
germinated: peak germination was in February in Illinois (Anderson et al. 1996) and
March-April in Ontario (Cavers et al. 1979). However, Anderson et al. (1996) observed
germination from ﬁeld-collected soil as late as mid June. Another possibility is that
appropriate germination conditions were not met. While soil was kept moist, it may not
have been wet enough for A. petiolata, which requires very moist conditions for
germination (Meekins and McCarthy 2001). Greenhouse conditions may also have been
too warm to trigger germination.

On the other hand, these results may indicate that seeds of A. petiolata in the

sampled forests do not persist for more than a year or two. Removal of adult plants may

125

largely eliminate populations from a site if plants were removed before seed set. While
continual re-introduction will be difﬁcult to control, management of A. petiolata is

possible.

Conclusions

Seed density: Seed density varied greatly among sites, ranging ﬁ'om 1677i1700 to
15133i55351 seeds/m2. Density was highly variable, likely due to the clumped
distribution of the most common species.

Species richness: Of the 88 seed bank species observed in this study, shade-intolerant
species made up 51% of species and 66% of seeds, while annual species contributed 23%
of species and 38% of seeds. Overall, less than half of the seed bank species were
observed in the forest vegetation, not even in large canopy gaps.

Effects of aspect and distance: Seed banks of forest edges are greatly inﬂuenced by '
adjacent areas: species characteristic of old-ﬁeld habitats were abundant in the seed bank,
especially close to the edge. This increase in density indicates that adjacent habitats are
important sources of seeds to forest edge seed banks. Clustering conﬁrmed the
differences in species composition between edge and interior samples. Seed bank species
richness declined with distance into the forest, but this decline was not statistically
signiﬁcant.

Non-magic species: The 14 (16%) non-native species contributed only 30% of the forest
seed bank, and none of the most common invasive species were present in the seed bank.
These results indicate that, at least for the species observed at these sites, the seed bank

does not contribute to their invasiveness.

126

Chapter 5
Size does matter: Long-terrn canopy dynamics in a small old-growth forest fragment

Introduction

Most studies of long-term canopy-tree dynamics in eastern deciduous forests have
focused on relatively large ﬁ'agments (V ankat et al. 1975, Fore et al. 1997), and little is
known about the inﬂuence of edge effects on long-term dynamics in small forest
ﬁagments. Furthermore, relatively few studies have examined old-growth forests (Abrell
and Jackson 1977, Runkle 1982, Parker et al. 1985). Because so many small forest
ﬁ'agments exist, knowledge of long-term dynamics is critical for understanding their
conservation value. By studying these dynamics in old-growth forests, the impacts of
fragmentation and edge effects can be separated from within-forest disturbances.

‘Old-growth’ eastern forests have been deﬁned in several ways. In a general
sense, old-growth forests are those that have never been clearcut (Runkle 1982, Davis
1996). Because historic information is not available for every stand, some authors have
used vegetation characteristics to identify old-growth stands. Parker (1989), for example,
deﬁned old-growth forests as having canopy trees older than 150 years and an all-aged,
multilayered canopy. Runkle (2000) noted that, in old-growth forests, enough time has
passed since the last major disturbance that under- and overstory species composition
have converged; thus species composition and stand structure are ‘in equilibrium’. This
deﬁnition excludes forests with extensive natural disturbance; Davis (1996) includes such
forests in her deﬁnition. Held and Winstead (1975) reported that, in most climax stands
of mesic forests, basal area of trees 2 10 cm dbh is about 30 mZ/ha. Other vegetation

characteristics sometimes used to describe old-growth forests include the presence of

127

fallen logs and snags, pit-and-mound microtopography, undisturbed soils, abundant
herbs, lichens, and fungi, and tree species characteristic of late successional stages
(Leverett 1996).

While old-growth forests are assumed to be in dynamic equilibrium, they are not
static. Runkle (2000) studied a number of old-growth forests in the southern
Appalachians, Ohio, and Pennsylvania. Over 15 years (from 1976 to 1991), density
decreased by 0-0.52% per year and basal area increased by 0.03-0.45% per year (basal
area in one site decreased by 0.22% per year). Species composition also changed
slightly: F agus grandifolia (American beech) importance decreased in the Appalachians
and in Ohio, and increased in Pennsylvania. Abrell and Jackson (1977) found that,
between 1965 and 1975, stand-wide basal area (for trees 2 10 cm dbh) of an old-growth
beech-maple fragment increased 2.4% (from 27.46 to 28.11 mz/ha) and stand density
decreased from 204.8 to 198.5 trees/ha. No change in species composition was found.

In some cases, forest change appears to be directional. Parker et a1. (1985)
studied an oak forest in Indiana, the 20.6-ha Davis-Purdue Forest. Between 1926 and
1976, stem density increased by 93.9%, while basal area increased by 30.8%. In 1926,
Quercus rubra (red oak), F raxinus americana (white ash), Q. alba (white oak), and
Ulmus americana (American elm) were the most important species. In 1976, the top four
species were Q. alba, U. americana, Acer saccharum (sugar maple), and F. americana.
The authors noted that late-seral, shade-tolerant species, such as A. saccharum and
Aesculus gIabra (Ohio buckeye), appeared to be replacing the dominant Quercus spp.
Schmelz et al. (1974) observed similar changes in Donaldson’s Woods, a 32.4-ha mixed

woods forest in Lawrence County, Indiana. Between 1954 and 1974, the importance

128

value (based on relative dominance and relative density) of the dominant Q. alba
decreased from 36.9% to 32.2%. F agus grandzfolia and A. saccharum both increased in
importance, with A. saccharum showing a larger increase (ﬁom 11.9% to 16.9%) than F.
grandifolia (15.4% to 16.6%). The authors noted that a minor climatic shift to cooler and
moister conditions could be the cause of the Quercus decrease. Stand density decreased
by 0.25% per year from 292.9 trees/ha to 277.1 trees/ha, while basal area increased by
0.45% per year from 28.0 mZ/ha to 30.4 mZ/ha.

Most non-anthropogenic disturbance in old-growth forests is limited both
spatially and temporally. Common mechanisms of disturbance include relatively minor
damage from wind- and ice storms, as well as occasional ﬁres (Runkle 1996). In a large
sample of mesic forests in North Carolina, Tennessee, Ohio, Pennsylvania, and New
York, Runkle (1982) found that gaps occupied 3.2-24.2% of land area, with lower values
in the north and higher values in the south. New gaps were formed and old gaps closed at
an average rate of 1% of total land area per year. Similarly, the rate of canopy turnover
in old-growth, hemlock-dominated forests in New York (determined from tree-ring
widths) was 3.1-4.5% (Ziegler 2004). Tree fall rates in an old-growth beech-maple forest
in Michigan (Warren Woods) ranged from 0.15 trees/ha per year to 1.64 trees/ha per year
(Poulson and Platt 1996).

Less than 1% of presettlement old-growth forests remain in the central hardwood
region, which extends from the Appalachian Mountains west to the Iowa prairies and
north into Michigan (Parker 1989). Michigan has 82,151 ha of old-growth forest, but
most of this forest occurs in the Upper Peninsula (Davis 1996); in southern lower

Michigan, only 474 ha (in 22 stands) remain (Parker 1989).

129

One of the best known beech-maple remnants in Michigan is Warren Woods,
located in Berrien County, Michigan. Cain (1935) surveyed a ~16 ha section of the
forest; the entire forest, including areas of second-growth, is ~83 ha. Cain found 18 tree
and shrub species with dbh of at least 2.5 cm; 12 were classiﬁed as canopy species.

F agus grandzfolia made up more than 50% of the total basal area; A. saccharum was
second in importance. Basal area for the portion of the forest sampled was 51.2 mz/ha. A
more recent survey of the same area found that F. grandifolia dominance had increased
to 68.2%; A. saccharum was still second, with 20% dominance (Donnelly and Murphy
1987). Combined importance of the two species was more than 85%. The latter study
recorded only nine tree species in the sample plots; an additional nine trees were listed as
present in low numbers. F urtherrnore, based on data derived from surveyors’ notes,
Donnelly and Murphy (1987) concluded that the species composition of Warren Woods
was very similar to that of pre-settlement forests.

Another well-studied old-growth ﬁagment is the 67-ha Hueston Woods in Preble
County, Ohio, which is also dominated by F. grandifolia (dominance in 1977 = 57.5%)
and A. saccharum (dominance in 1977 = 24.2%, Runkle 2000). In 1991, total forest
basal area was 35.6 mz/ha, a slight decline from the 1977 value (36.7 mz/ha, Runkle
2000). During that same period, density of trees 2 25 cm dbh declined 7% from 171 to
159 trees/ha. A 26.3-ha old-growth forest in the North Chagrin Reservation (near
Cleveland, Ohio), later named the AR Williams Woods, was characterized by Williams
in 1936. While F. grandifolia and A. saccharum were not uniformly distributed

throughout the woods, they made up 51% and 26.5% of the forest (importance values),

130

respectively (Williams 1936). The ages of fallen trees (determined ﬁ'om tree rings)
ranged from 150-250 years for F. grandifolia and 115-190 years for A. saccharum.

A fourth stand, Hoot Woods, is a 25.9 ha old-growth, beech-maple forest in Owen
County, Indiana. This forest, like the other stands, is dominated by F. grandifolia (Abrell
and Jackson 1977). Esten (1932) surveyed a 0.40 ha portion of a beech-maple forest at
Turkey Run State Park in Parke County, Indiana. Acer saccharum density of trees 2 5.1
cm dbh (160 trees/ha) was slightly higher than F. grandifolia density (148 trees/ha), but
dominance (based on crown area) of F. grandifolia was twice than of A. saccharum.

Cain (1935) noted that, in Warren Woods, A. saccharum dominated the smaller
size classes. Other authors (Esten 1932, Braun 1950, Abrell and Jackson 1977) noticed
similar patterns in other beech-maple remnants and have speculated that A. saccharum
may be replacing F. grandifolia in these forests. In Hoot Woods, F. grandifolia declined
in importance (46.0% to 42.7%) between 1965 and 1975 (Abrell and Jackson 1977).
Meanwhile, A. saccharum importance increased by 10% (to 31.8%). The authors
hypothesized that these changes were part of multidecadal wave-form periodicities in
abundance of these two species. In Hueston Woods, Vankat et al. (1975) found that F.
grandifolia was successfully replacing itself. Since that study, however, F. grandifolia
dominance and stem density (for trees 2 25 cm dbh) have declined by more than 15%
(Runkle 2000). Meanwhile, dominance and stem density of similarly-sized A. saccharum
trees have increased by approximately 15%. Runkle (2000) suggested that patterns of F.
grandifolia decline could be due to an unspeciﬁed environmental change that shifted the

competitive balance between the two species.

131

Poulson and Platt (1996) conducted a more detailed study of replacement patterns
of these two species. In Warren Woods between 1933 and 1980, A. saccharum
dominated the smallest size classes while F. grandifolia dominated the largest classes.
F agus grandifolia outcompeted A. saccharum in the understory, while the situation was
reversed in gaps. The authors noted that the rate of gap formation increased almost
tenfold between 1949 and 1994 (from 0.16 trees/ha per year to 1.64 trees/ha per year).
Under the higher rate of gap formation, they predicted that A. saccharum would
eventually become dominant in the stand.

Little is known about the effects of forest size on long-term canopy tree dynamics.
Most of the studies described above avoided sampling near forest edges because, as
discussed in Chapter 2, edge effects can inﬂuence species composition. For example,
Kupfer et a1. (1997) showed that edge species were more common in gaps close to the
edge than they were in forest-interior gaps. In a small ﬁagment, the area of both edge
and gap inﬂuence would be proportionately larger than in larger forests. Consequently,
species composition and abundance may differ from a larger fragment of the same forest
type. This study examined long-term (> 60 years) canopy dynamics in a small (11.7 ha)
old-growth beech-maple forest and addressed the following questions:

0 Is species composition in this fragment changing?

o Are stand basal area and stem density in a steady-state equilibrium?

0 Are stand characteristics, including species composition, stem density, and basal
area similar to those of larger old-growth forests?

o How much of this small fragment is inﬂuenced by edge effects, and is there any

core forest present?

132

Methods
Site history and description

Tourney Woodlot is an 11.7-ha old-growth maple-beech fragment in lngham
County, Michigan. The bulk of the woodlot (6.07 ha) was acquired in 1939 by Michigan
State University (MSU) from the Frank Bennett estate (Schneider 1966). The woodlot
had been owned by the Bennett family since 1852. The family had maintained the
woodlot in its undisturbed state, not allowing logging or grazing. The only
anthropogenic disturbance was the occasional removal of dead trees for ﬁrewood. The
woodlot likely had not been burned for at least 200 years (Schneider 1966).

After Tourney became university property, it was set aside as a research reserve.
In 1941-1942, two evergreen windbreaks were planted to ameliorate edge effects
(Schneider 1963). Five species (a total of 802 trees) were used for the plantings: Pinus
strobus (white pine), Abies concolor (white ﬁr), Picea mariana (black spruce), P. abies
(Norway spruce), and Thuja occidentalis (white cedar). Two staggered rows (6.7 m
wide) were planted along the north edge of the woodlot. The western windbreak was 3-5
rows (13.4 m) wide. Most of the western windbreak had died by 1960, due to
competition ﬁom regenerating hardwoods (Schneider 1963). Sections of the northern
windbreak are still in existence (Figure 5.1), but no regeneration was observed.

The woodlot has increased in size since it was originally acquired by the
university (Figure 5.1). A block of forest has been allowed to regenerate along the west

edge, and the southeast comer has also been reforested, at least partly through planting of

evergreens (1961 photo in Figure 5.1).

133

Figure 5.1: Tourney Woodlot from 1938 to 2001. All photos are the same scale (shown
in the 2001 photo). Photos for 1938, 1950, 1970, and 1981 were obtained from the MSU
Aerial Photo Archive. The 1961 photo is taken from Schneider (1963). The 2001 photo
was obtained from the Michigan Geographic Data Library (MiGDL). The permanent
pond is noted in the 1961 photo. A large blowdown (present in 1961) is still visible in
the 1970 photo. Planted conifers are pointed out in the 2001 photo.

134

 

 

’1]?
r.
,3.)
.1...
.,..
. w—
... .—
‘ 5'3
0

 

1961

Currently, Tourney is bordered on the north and south by active pastures, on the
east by a road, and on the west by active crop ﬁelds. A small, semi-permanent pond is
located in the east of the woodlot (Figure 5.1). This pond has standing water for most of
the year, except sometimes in late summer. Dutch elm disease was discovered in the
woodlot in 1962 (Schneider 1963). While the presence of emerald ash borer (Agriclus
planipennis) in Tourney has not been conﬁrmed, the beetle is known to occur in Ingharn
County. Tourney was added to the National Parks Service Register of Natural

Landmarks in 1976.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A B C D E F G H I J K L M N O P Q R S
1 \ rel . 97
2 f if
. v x \ , l i: ﬁrst"
4 W275 A W 2533‘ 361?? ”.5630 L1
5 / /V //J//36l is,"
. ”fad/(N1 it N
7 ( / / i l l
a “j . k‘i/ / l l
. J1)
10 276\

 

 

 

 

11

 

 

 

 

 

 

 

Leer...) Map of
”‘— °°"‘°"*"* Toumey Woodlot

W POND

Figure 5.2: Locations of permanent plots in Tourney Woodlot. Each grid square is 18.3
x 18.3 m (60 x 60 ft). Elevations are given in meters. From Schneider (1963).

Canopy tree census

In 1940, 163 permanent plots (18.3 x 18.3 m, 60 x 60 ft) were installed in Tourney

(Figure 5.2). Trees in these plots were censused every 10 years from 1940 to 1970.

136

Summary data ﬁom 1940-1960 were published in a doctoral dissertation (Schneider
1963) and a research bulletin (Schneider 1966). Data from 1970 were obtained ﬁ'om an
MSU Forestry Department archive.

In the winter of 2003 -2004, the permanent plot markers were relocated and trees
in each plot were censused. The dbh and height category (understory, subcanopy, and
canopy) of all trees 2 2.5 cm dbh were recorded. Understory trees were greater than 2.5
cm dbh but less than 3 m tall. Subcanopy trees were greater than 3 m tall but were still
overtopped by another tree. Canopy trees were not overtopped by any other tree. Trees
that could not be identiﬁed were marked and identiﬁed in the spring.

Statistical Analyses

Summary data from 1940-1970 were presented as number of trees/acre in l-inch
dbh classes. These data were converted to trees/ha and also to basal area (in m2)/ha for
comparison to the 2004 data. Relative dominance and relative density were calculated
for all years. Because plot-by-plot data were not available for the earlier data sets,
relative frequency was calculated only for the 2004 data. Importance values were
calculated as the sum of relative density and relative dominance, and are presented as a
percent of the total (200); for 2004, a second importance value that included relative
frequency was also calculated (McCune and Grace 2002). For eight of the most common
species (Table 5.2), the FREQ procedure in SAS (SAS Institute 1999) was used to test
whether the total stem density and the density of large trees (dbh 2 12.7 cm) varied by
year. For each of those species, I also tested whether the number of stems in a dbh class
varied by year. These species were selected because detailed information (1940-1970)

was available in Schneider (1963) and an MSU Forestry Department archive.

137

An index of dispersion (IOD, described in Chapter 4) was calculated for total
stem density of the nine most abundant tree species (all species with 2 40 individuals).
Because the majority of stems included in this analysis were very small, IOD values were
also calculated for large trees (2 10 cm dbh, seven species with Z 10 individuals), and
canopy trees (ﬁve species with > 10 individuals).

The clustering procedure in PC-ORD (McCune and Mefford 1999) was used to
investigate relationships among plots. I used the Sorensen distance measure and a
ﬂexible beta ([3: -0.25) linkage method. All species present in fewer than four plots were
excluded ﬁ'om the analysis. The basal area of each species in a plot was used to create a
dendrogram. Clustering is sensitive to species abundance and plots can cluster based on
species abundance. For this reason, data were relativized to each Species’ maximum
basal area and a second dendrogram was created. This dendrogram reﬂects species
composition more than species abundance. In both cases, indicator species analysis
(Dufrene and Legendre 1997, described in Chapter 2) was used to choose the optimal

number of clusters.

138

Table 5.1: Species found in Tourney Woodlot. Year of presence was not given for the
1940/ 1950 data. x = species present; 0 = species present outside of sample area or too
small to be counted (2004 only).

 

Year recorded
Species 1940/ 1960 1970 2004
1950

 

X

Acer saccharum

Catpinus caroliniana

Catya cordiformis

Fagus grandifolia

F raxinus americana

F raxinus quadrangulata

Ostrya virginiana

Prunus serotina

Quercus alba

Quercus rubra

T ilia americana

Ulmus americana

Ulmus rubra

Zanthoxylum americanum

Acer nigrum

Acer rubrum

F raxinus nigra

Platanus occidentalis

Salix sp.

Acer saccharinum

Amelanchier arborea

Celtis occidentalis

“Apple”

Acer negundo

Carya tomentosa

Cornusﬂorida

Crataegus spp.

Hamamelis virginiana

Lindera benzoin

Populus tremuloides

Sambucus sp.

Carya glabra

Carya ovata

Juglans nigra x
Number of species 22 19 23 17 (21*)

* For 2004, species richness in parentheses includes species that were observed but not

counted.

>< ><><><O><

XXXXXXXNXXXXXX
><

XXXXXXXXXXXXXXXXXXX
><O><><><><

XXXXXXXXXXXXXXXXXXXXX

XNXXXXXXX
>40

XXO

 

 

 

 

 

139

Results
Species richness

In 2004, 17 tree species were found in the study area (Table 5.1). Overall species
richness was similar to data from 1940-1970. Three new species (Carya glabra (pignut
hickory), C. ovata (shagbark hickory), and Juglans nigra (black walnut)) were recorded.
Four additional species (Carpinus caroliniana (American hombeam), Cornusflorida
(ﬂowering dogwood), Sambucus spp. (elderberry), and Zanthoxylum americanum
(prickly ash)) are still present in the woodlot but were either located outside the sampling
area or were too small to be recorded.

Changes in stem density and basal area

Total stem density (for all stems 2 2.5 cm dbh) varied dramatically over the 60
year study period (Figure 5.3), ranging from a low of 981.7 trees/ha in 2004 to a high of
1823.6 trees/ha in 1970. Total stem density varied signiﬁcantly among years (x2=2167,
p<0.001). For the most common species, the number of stems in a dbh class also varied
signiﬁcantly (Table 5.2). For most years, the rate of increase in total stem density was
low (Table 5.3), ranging between 1.84% and 3.1% per year. Between 1970 and 2004,
however, total stem density decreased by 15.81% per year. Over the entire study period
(1940 to 2004), total stem density increased on average by 0.13% per year.

Unlike total stem density, stem density of large trees (2 12.7 cm dbh) remained
relatively stable (Figure 5.3). However, differences among years were signiﬁcant
(fr—207.8, p<0.001). Large tree density was lowest in 2004 (213.9 trees/ha) and highest
in 1960 (270.1 trees/ha). Changes in large tree density were smaller than those for total

density and were generally negative (Table 5.3). Large tree density decreased by 0.61 -

140

3.16% in most years, except between 1950 and 1960, when it increased by 0.83% per
year. Over the 65 year study period, large tree density decreased by 0.30% per year.

Table 5.2: Chi-square tests for independence of size class and year.

 

 

 

 

 

 

Species x2 p

Acer saccharum 1 016 <0.0001
F agus grandifolia 105.8 <0.0001
T ilia americana 38.6 0.030
Ulmus americana 51.3 <0.0001
Prunus serotina 39.1 0.026

F raxinus americana 400.1 <0.0001
Ulmus rubra 161.8 <0.0001
Quercus rubra 150.3 <0.0001

 

Basal area also varied less than total stem density (Figure 5.3); total basal area
ranged from 27.6 mz/ha (1970) to 35.4 mz/ha (1950, Table 5.4). On an annual basis,
basal area increased by 0.28-1.46% in most years, except between 1960 and 1970, when
it decreased by 1.98% per year (Table 5.3). Over the study period, basal area increased
by 0.24%.

Table 5.3: Rate of change (per year) of stem density and basal area in Toumey Woodlot.
Total density includes all stems 2 2.5 cm dbh. Large trees are 2 12.7 cm dbh.

 

 

 

 

Time Stem density Stem density Basal area
(total) (large trees) (total)
trees/ha % trees/ha % mz/ha %

1940-1950 16.68 1.84 -1.61 -0.61 0.08 0.28

1950—1960 ' 32.69 3.05 2.08 0.83 0.44 1.46

1960-1970 43.41 3.10 -3.42 -1.27 -0.68 -1.98

1970-2004 -25.08 -15.81 -0.65 -3.16 0.18 0.65

Total (1940-2004) 1 .17 0.13 -0.80 -0.30 0.07 0.24

 

 

 

 

 

141

2000

 

 

 

’5
£
3 1500.
2
$
.é‘ 2004
g 1000-
o
'O
E
2
3 500.
N
‘6
.—

o.

2000

 

(A)
O
O

 

M
0'!
O

N
O
O

_-
O
0

Stem density for large trees (stems/ha)
U: at
O O

O

 

Basal area (m2lha)
N w -h
0 O O

_\
O

 

 

 

 

1940 1950 1960 1970 2000

Year

Acer saccharum
Fagus grandifolia
Tilia americana
Ulmus americana
Fraxinus americana
Prunus serotina
Quercus rubra
Ulmus rubra

Other species

\
\
3“

 

Acer saccharum
Fagus grandifolia
Tilia americana
Ulmus americana
Fraxinus americana
Prunus serotina
Quercus rubra
Ulmus rubra

Other species

I .
2

Acer saccharum
Fagus grandifolia
Tilia americana
Ulmus americana
Fraxinus americana
Prunus serotina
Quercus rubra
— Ulmus rubra

- Other species

'0.
K
a

Figure 5.3: Stem density and basal area by species. Total stem density includes all stems

2 2.5 cm dbh. Large trees are 2 12.7 cm dbh.

142

Table 5.4: Total basal area (mz/ha) for the most common species in Toumey. The
maximum value for each species is in bold.

 

 

 

Total basal area (mZ/ha)
1940 1950 1960 1970 2004

Acer saccharum 16.45 16.74 19.29 16.77 22.52
F agus grandifolia 5.83 5.92 7.36 6.21 6.53
T ilia americana 1.93 1.92 1.70 1.04 1.05
Ulmus americana 1.88 2.00 1.83 0.21 0.17
Prunus serotina 0.24 0.22 0.21 0.51 0.10
F raxinus americana 0.63 0.62 0.72 0.80 0.77
Ulmus rubra 0.85 0.99 1.30 0.28 0.12
Quercus rubra 0.58 0.87 1.13 1.35 2.05
0sttya virginiana 0.22 0.22 0.20 0.22 0.09
Other species 0.54 0.48 0.64 0.16 0.28

Total 29.15 29.98 34.37 27.56 33.66

 

 

 

 

Abundant species

Acer saccharum was the most abundant species in Tourney, with more than 700
trees/ha in all years (Table 5.5). Density was highest in 1970 due to the large number of
trees in the smallest size class. Stem density for the two largest classes increased from
1940 to 2004. Basal area for A. saccharum was similar in most years, with the highest
values in 1960 (19.3 mz/ha) and 2004 (22.5 mZ/ha, Table 5.4). Importance values for this
species increased from 62.08% in 1940 to 74.01% in 2004 (Table 5.6). The 2004
importance value that included relative frequency was substantially lower (61.0%, Table
5.7).

F agus grandifolia was the second most common species in the study area, with
density exceeding 100 trees/ha in all years except 2004 (Table 5.5). For most size
classes, density was highest between 1940 and 1960; density for the largest trees
increased from 1940 to 2004. F agus grandifolia basal area peaked in 1960 (7.37 mz/ha,
Table 5.4). Importance values declined slightly, from 17.06% in 1940 to 14.07% in

2004. Including relative frequency increased the 2004 importance value to 16.43%.

143

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N. ~ N 03. md? d.Nv o.wN w.N~ aﬁN d.N. N6 of m6 mdm Two 0.3 wdm .89
dd dd dd dd dd d.N Nu— n; hd Nd dd Nd bd dd Nd m.md
dd Nd dd dd dd Nd NJ d4 d._ bd dd Nd 54 m; 5d wdm
dd Nd dd dd dd m; md d.— m.— N._ Nd Nd md d.N N.N _.wm
Nd md Nd vd Nd dd nd Nd bd wd hd dd 54 d._ N._ WmN
vd _N d.N N.v d.m ﬁn hd d4 md d4 _._ N.N Nah Na— 54 FN—
dd— d.NN 02 Nd N.: m.m Wm d; Nd Nd M: méN mdv N.NN dd Tm
md_ 93. de N.wN M.N~ MA e.m— d.N nd dd 5m .vNN m. 3V m.me N.NN nN
ugh; 9920 and»: 38.03% and»: «.355
N.N. m.Nd— 1am 0.x d.N. ddN bﬁN d6 of Wm m.N~ d.N w.w— N.NN N.mN :30.
dd Nd md Nd Nd dd md Nd Nd Nd dd Nd m.N N.N m4 mdg
Nd NA md 5d bd dd md Nd md Nd dd nd NA d.N d.N wdw
Nd Nd NA N— d; Nd rd Nd Nd Nd dd Nd b; d.N dd ﬁwm
a; Nd d._ NM— h; vd md dd Nd Nd bd dd d; w; d.N v.mN
N.N hm N; d4 N._ .vd NA Nd Nd Nd w; md N.N N.N Nam N.N—
hd N.VN v.2 m; N._ d6 N6. Na— d._ An— Nd Nd v.5 Nd dd _.m
bd Ndh MKM N.N b; N: ddN N.N m; N.N Nam Nd N.N Nam vé mN
geotmﬁe 5552K SEES. SEEK uneetmse 2:25
N.: dd— WMN NdN 9mm 06w 5d: ddﬁm n.mN_ d.wN_ mddh mﬁwnn ﬁdwd mth d.mdh _So~
m4 Nd m._ mg mg dd N.m d.N N.N 54 d6“ vd —.d \ld N.N. m.md
dd d.N N.N NA NA m6 v.5 vd v6 Nd —.mN dd“ 0.2 Q: v.2 wdw
dd N._ Na— m.m Nd. _d vd Wm— dd~ YNH _ﬁN ddN d.—m —._m \ldN fwm
m._ b; N._ n; N.N m6 N.N. Y: in Vd~ _.mN ﬁdm down m. :u mdv YWN
_4 mN N.N m4 EN vd m.m_ wd— WM: md— Ewm mNc d.Nb mNd 5:. N.N—
v.N N6 v.0 QM oN— QMN YMWN wdm aim v.5. v.mNN 0th eddm EMS Nag mm
N.N Nd N.N N.N. vd MEN mdv mdw Wm». ddm dd; ﬁddd WV; foam 9me wN
VddN dbd— ddd— dmd— dVd— VOdN due dog dmdd dve VddN chm; doe dma 9d— 320
Imm—

e=~§§Ee 32k

 

 

BENCEEM Sack

 

EEBn‘UUBu. kmbvx

.23 E m_ 86on :08 cod $20 and a E 362% 85::me 2E. 38» :m Sm momma—o one. .3 Q5805 bacon 8:. “Wm 032.

 

T ilia americana total density and basal area declined monotonically over the
study period (Tables 5.4 and 5.5). Density in most dbh classes was highest in 1940;
density of the largest trees has remained steady. This species. decreased in importance by
50% between 1940 and 2004 (5.18% to 2.13%, Table 5.6). The inclusion of relative
ﬁ'equency increased the 2004 importance value to 3.02% (Table 5.7). Ulmus americana
total density declined from 1940 to 1970, but was higher again in 2004 (Table 5.5).
However, density in the three largest dbh classes declined over the study period; in 2004,
no trees larger than 38 cm dbh were found. Ulmus americana basal area was lowest in
2004 (Table 5.4). The importance value for this species declined tenfold between 1940
(4.61%) and 1970 (0.43%, Table 5.6). The 2004 importance value increased with the
inclusion of relative frequency (from 0.89% to 3.80%, Table 5.7).
Table 5.6: Importance values (IV) for all study years. Importance values are the sum of

relative dominance and relative density, and are expressed as a percentage of the total
(200). The highest value for each species is in bold.

 

 

 

 

1940 1950 1960 1970 2004
Acer saccharum 62.08 61.76 63 .38 66.61 74.01
Fagus grandifolia 17.06 15.72 15.49 14.28 14.07
T ilia americana 5.18 4.42 3 .31 2.40 2.13
Ulmus americana 4.61 4.37 3.33 0.43 0.89
Prunus serotina 0.73 0.55 0.48 1.68 1.21
Fraxinus americana 1.52 1.44 2.98 4.25 1.52
Ulmus rubra 3.64 5.93 5.26 2.04 0.56
Quercus rubra 1.21 1.69 1.93 3.15 3.70
Ostrya virginiana 1.96 2.32 2.01 2.42 1.24
Other species 2.02 1.79 1.84 2.69 0.70

 

 

Prunus serotina density was low (< 6 trees/ha) ﬁ'om 1940 to 1960 (Table 5.5).
Density in 1970 (27.7 trees/ha) was more than ﬁve times the 1960 density (4.9 trees/ha);
density in 2004 declined only slightly (20.9 trees/ha). While density in all size classes
increased between 1960 and 1970, the greatest difference was in the smallest class. Basal

area was highest in 1970 (Table 5.4). Importance values for this species varied, but were

145

lower in 1940-1960 (0.48-0.73%) than in 1970 and 2004 (1.68% and 1.21%, Table 5.6).
Density of F. americana increased sixfold between 1950 and 1960, and almost doubled
again by 1970 (Table 5.5). The increase was limited mostly to the two smallest size
classes. Density in 2004 was similar to the 1940-1950 values. Basal area was highest in
1970 (Table 5.4). Importance values increased ﬁ'om 1940 to 1970, and then decreased in
2004 (Table 5.6). The inclusion of relative frequency increased the 2004 importance
value ﬁom 1.52% to 2.28% (Table 5.7).

Table 5.7: Importance values (IV) for all tree species in Tourney Woodlot in 2004.

Importance values are the sum of relative dominance, relative frequency, and relative
density, and are presented as a percentage of the total (300).

 

 

 

 

Species IV (%)
Acer saccharum 60.99
F agus grandifolia 16.43
Ostrya virginiana 3.89
Ulmus americana 3.80
Quercus rubra 3.76
T ilia americana 3.02
F raxinus americana 2.28
Carya cordiformis 2.03
Ulmus rubra 1.83
Hamamelis virginiana 0.71
Celtis occidentalis 0.46
Carya ovata 0.23
Juglans nigra 0.18
Prunus serotina 0.15
Crataegus spp. 0.08
Acer nigrum 0.07
Carya glabra 0.07

 

Ulmus rubra density increased from 1940 to 1960 and then declined to its lowest
level in 2004 (7.5 trees/ha, Table 5.5). Density in the smallest size class peaked in 1950,
but density in several size other classes was highest in 1960. Basal area peaked in 1960
and then declined to 0.12 mz/ha in 2004 (Table 5.4). Importance values were highest in

1950 and 1960 (5.93% and 5.26%, respectively), and reached their lowest level in 2004

146

(0.56%, Table 5.6). Quercus rubra density increased ﬁ'om 1940 to 1970 and declined
somewhat in 2004 (Table 5.5). The density of the smallest trees was highest in 1970;
density of the largest size class was highest in 2004. Basal area increased monotonically
from 1940 to 2004 (Table 5.4). Importance values also increased ﬁom 1940 to 2004,
from 1.21% to 3.70% (Table 5.6). Ostrya virginiana density increased from 1940 to
1970 and then decreased dramatically in 2004 (Table 5.5). Peak density for most size
classes was reached in 1970 (Table 5.5), but basal area remained fairly constant from
1940 to 1970 (Table 5.4). Ostrya virginiana importance values varied between 1.24% in
2004 and 2.42% in 1970, but showed no distinct pattern (Table 5.6). Despite its low
density and basal area, this species was ranked third by importance value (IV=3.89%) in
2004 due to its high relative frequency (Table 5.7).

Species distribution

Table 5.8: Index of dispersion (IOD) values for common trees in Toumey. Total density

includes all stems 2 2.5 cm dbh. Large trees are 2 12.7 cm dbh. Canopy trees are not
overtopped by any other tree. Values in bold were not signiﬁcantly different ﬁ‘om a

 

 

 

 

random distribution. -- = too few trees to analyze.

Total stem density Large trees Canopy trees
Species IOD p IOD p IOD p
Acer saccharum 4.78 <0.001 1.05 0.30 1.27 0.01
F agus grandifolia 4.48 <0.001 2.06 <0.001 1.57 <0.001
Prunus serotina 6.11 <0.001 -- -- -- --
Ulmus rubra 1.73 <0.001 1.24 0.022 -- --
Ostrya virginiana 3.79 <0.001 -- -- -- --
T ilia americana 2.28 <0.001 1.51 <0.001 1.13 0.13
Quercus rubra 11.2 <0.001 5.45 <0.001 3.39 <0.001
Ulmus americana 6.65 <0.001 4.72 <0.001 -- --
Fraxinus americana 4.78 <0.001 4.41 <0.001 4.02 <0.001

 

 

 

 

Index of dispersion (IOD) analysis for total stem density (Table 5.8) indicated that
all species analyzed were signiﬁcantly clustered. The highest IOD value was found for

Q. rubra; U. rubra had the lowest value. When IOD was calculated for large trees (2

147

12.7 cm dbh), A. saccharum was found to be randomly distributed through the forest.
Quercus rubra and U. rubra still had the highest and lowest values, respectively. Finally,
IOD was calculated using only canopy-sized trees. In contrast to the large tree results, A.
saccharum trees were clustered. However, T. americana was not. F raxinus americana
canopy trees had the highest IOD; A. saccharum was the lowest.

Cluster analysis

Table 5.9: Clusters based on total basal area and relativized basal area. All listed
indicator species were signiﬁcant at o=0.05.

 

 

 

 

 

 

 

 

 

Total basal area
Indicator
Cluster # Plots Indicator species Values (%)
1 82 Acer saccharum 39
2 15 Quercus rubra 90
3 32 Fagus grandifolia 71
4 23 Acer saccharum 21
5 11 Ulmus americana 30
Ulmus rubra 23
Relativized basal area
Indicator
Cluster # Plots Indicator species Values (%)
1 34 Acer saccharum 23
F agus grandifolia 29
2 34 T ilia americana 51
Quercus rubra 45
Ulmus americana 37
Prunus serotina 35
Ostrya virginiana 31
3 32 F agus grandifolia 62
4 51 Acer saccharum 36
5 12 Acer nigrum 41

 

 

 

 

Clustering by total species basal area produced six clusters. However, because
one of these clusters contained only two members, it was combined the most closely
related cluster (5). Clustering by relativized basal area produced ﬁve clusters. Roughly
50% of plots were in different clusters on the two dendrograms. Indicator values based

on total basal area were higher than values based on relativized basal area (Table 5.9).

148

For the total basal area data, plots with a high abundance of A. saccharum (cluster 1)
were separated from those with lower abundance (cluster 4). Plots in cluster 3,
characterized by high abundance of F. grandifolia, were located mostly in the
northwestern section of the woodlot (Figure 5.4). Plots with high basal area of Q. rubra
(cluster 2) were generally located in the south. Plots from clusters 4 and 5 were scattered
throughout the woodlot. For the relativized data set, plots with high abundance of either
F. grandifolia or A. saccharum were separated (clusters 3 and 4). A third cluster (1)
contained plots with intermediate abundances of both. The F. grandifolia plots (cluster
3) were still located mostly in the western section of the woodlot (Figure 5.5). Plots from
cluster 2, which were characterized by high abundance of several less shade-tolerant
species, were located around the edges of the sampling area. Plots with high relative

abundance of A. nigrum (cluster 5) were located throughout the woodlot.

149

 

Figure 5.4: Cluster identity of each plot in Tourney Woodlot. Clustering was based on
total basal area for each species within a plot. Each plot is 18.3 x 18.3 m (335 m2).

 

Figure 5.5: Cluster identity of each plot in Tourney Woodlot. Clustering was based on
basal area for each species within a plot, relativized to species maxima. Each plot is 18.3
x 18.3 m (335 m2).

150

Seconded 033—8 0305 mos—m.» occur—88m 305 552:» @6on 8: wood 85% 2F...

 

 

 

 

 

 

 

 

 

Ea: eemxoea
as :22 $21.35:
:8: e5 boson

aEa Essa one: see baa 63: 282:3 88$ 20:; 38: sec assom
38-9.2 $2.32 So _ ﬁe as _ -32 86% 25
meeoow 03.6.50? 030.506 093.55 >~ wzwnk E QMGNSU

so: ska... tea smm sew Sassoon sees

as so. as; see sew assesses seem
”mos—g countenan—

2 NN N_ moo: owuﬂ

a E moo: =e
. 58:33.. 362$

38. 32 em: 32 as: owes.

a. :3 3%: as: __e
”dag—E85 36:3 88m
em“ :a 9mm in ESE 8:. _eam
EwEomE £865 0E0 020 53:22 c2283
e. : an mom 9% 5.8 $5 85

$83 8683
headoh £503 doom $503 max—“Ema mdooa acumen: £503 :053

 

 

.oocmcwcou gum—2 Ea based 03:22 no woman 8m 82? oo§:OQE_ A3“. Eu 2 M 2m moo: 0me—
.mo:m uoﬁo =w com End :8 a.” N Be moot. 092 £503 252:3 com ”and Bo N N 08 moo: owe: £503 c8853 8m ﬁnd Eu
m.N N 89: “6205 .80: =n. .85 =« com 95:88 oEsEéooon 530%-20 3% do moﬁtogomamao 05 mo nomtmmﬁoo ddm 03mg.

151

Discussion
Species richness

Tourney Woodlot has relatively high tree species richness compared to other
beech-maple forests. Between 1940 and 2004, sampled species richness varied between
17 and 23 (Table 5.1). Braun (1950) reported the lowest canopy species richness for
beech-maple forests: 3 and 14 (average 9.5). In contrast, Cain (1935) found 15 species
(stems of dbh 2 2.5 cm) in Warren Woods, Michigan; 12 of these species were classiﬁed
as canopy species (Table 5.10). On the other hand, Hoot Woods, a 25.9 ha old-growth
fragment in Indiana, had 22 species of trees (dbh Z 10 cm, Abrell and Jackson 1977).
Toumey’s high tree species richness may be due to its small size. While the dense
canopy of beech-maple forests may suppress many less shade-tolerant species, increased
light at the forest edge allows some of these species to persist (Whitney and Runkle 1981,
Kupfer 1996). The high edgezarea ratio of Tourney creates a large area of edge inﬂuence
in which less shade—tolerant species may persist. Indeed, species richness in Tourney is
somewhat higher in the south of the woodlot (Figure 5.6). Furthermore, the area around
the pond appears to be suitable for species characteristic of wetter habitats (such as Acer
rubrum and A. saccharinum), further increasing species richness.
Dominant species

While richness may be high in Toumey, the woodlot is typical in the high
dominance of A. saccharum and F. grandifolia. During the study period, A. saccharum
and F. grandifolia made up 64-92% of the total basal area (Table 5.3). Combined
dominance of these two species was lowest in 1950 and highest in 1970. According to

Braun (1950), these two species make up ~80% of the basal area in beech-maple forests.

152

A study of 36 forests in and around Indiana by Schmelz and Lindsey (197 0) identiﬁed 16
beech-maple sites; combined importance values of A. saccharum and F. grandifolia in
these sites ranged from 48-84%. In most sites, F. grandifolia has higher importance
values than A. saccharum (Table 5.10). A survey of an unnamed woodlot in Ohio found
that, of the 32.7 m2 of basal area in the l-ha sample plot, F. grandifolia contributed 66%,
while A. saccharum contributed only 18% (Gilbert and Riemenschneider 1980). In Russ
Forest, Michigan, F. grandifolia made up 48% of the canopy, while A. saccharum made
up 23% (Braun 1950). In Tourney, on the other hand, A. saccharum (IV=61%) is much

more abundant than F. grandifolia (IV=16%, Table 5.7).

 

Figure 5.6: Tree species richness in Tourney Woodlot in 2004. Darker colors indicate
higher richness.

This preponderance of A. saccharum could be due to the woodlot’s small size.
Poulson and Platt (1996) found that A. saccharum seedlings grow faster than F.
grandifolia seedlings in gaps; the opposite is true under a closed canopy. In a small

woodlot like Tourney, a single gap occupies a much larger proportion of total area than it

153

does in a larger forest like Warren Woods. Acer saccharum trees in Tourney are
therefore more likely to be near a gap than they are in Warren Woods. Furthermore,
increased light availability at the edge would also favor A. saccharum over F.
grandifolia. These speculations are supported by the concentration of F. grandifolia in
the northwest portion of the sample area (Figure 5.7). This region is the most shielded

from edge effects and from the increased light availability around the pond.

“ 20

   

Figure 5.7: Basal area (mZ/ha) of F. grandifolia in Tourney Woodlot. Darker colors
indicate higher basal area.

Changes in stem density and basal area

Stem density varied signiﬁcantly among years both for the whole forest and for
individual species (Tables 5.2 and 5.5). While total stem density reached its peak in
1970, stem density for large trees (2 12.7 cm dbh) was highest in 1960 (Figure 5.3).
Thus the great increase in stem density in 1970 was due only to changes in the number of
smaller stems. By 2004, total stem density had returned to the 1940-1950 values; large

tree density was lower. Stem density in Tourney was comparable to values in other sites

154

(Table 5.10). Stem density of large trees was slightly lower in Hueston Woods (159
trees/ha, Runkle 2000) and Hoot Woods (199 trees/ha, Abrell and Jackson 1977) than in
Tourney (214 trees/ha); density was higher in Williams Woods (243 trees/ha, Williams
193 6).

During the 64-year study period, total basal area in Tourney varied between 27.6
and 35.4 mz/ha (Figure 5.3). In Hoot Woods, an old-growth beech-maple forest in
Indiana, total basal area was 28.1 mZ/ha. In 1991, basal area in Hueston Woods was 35.6
mZ/ha (Runkle 2000). Total basal area for Warren Woods was much higher (51.2 mZ/ha,
Cain 1935). Held and Winstead (1975) reported that, in most climax stands of mesic
forests, basal area of trees 2 10 cm dbh is about 30 mZ/ha. In 2004, the equivalent basal
area for Tourney was 32.3 mz/ha.

In mature ecosystems, individuals are expected to be larger and density to be
lower (Odum 1969). Basal area in Tourney has increased by 0.24% per year over the
study period (Table 5.3). While overall stem density has ﬂuctuated (changing by
-15.81% to 3.10% per year), density of large trees has decreased by 0.30% per year.
These directional changes contradicted the expectation that Tourney was in dynamic
equilibrium. However, the increase in size of the woodlot may have allowed trees around
the edge of the study area (formerly the edge of the woodlot) to increase in size. The
reduction of edge effects in the study area would also have led to a decrease in stem
density. Similar changes in basal area and stem density have been found in other old-
growth forests. In Hoot Woods, stem density decreased 3% and basal area increased
2.4% over 10 years (Abrell and Jackson 1977). Runkle (2000) found that some old-

growth forests from Pennsylvania to the southern Appalachians also showed a trend

155

towards fewer, larger stems over 14 years. Basal area changed by -0.22% to 0.45% per
year, while stem density decreased by -0.52% to 0% per year.
Abundant species

Because A. saccharum is the most abundant species in the woodlot, patterns in its
abundance had a large inﬂuence over forest-wide patterns. In 1970, A. saccharum
density reached its highest levels in the smallest size class, almost equaling total density
in 1960 (Table 5.5). However, densities in the next two size classes had reached peak
density in 1960. While the number of mid-sized trees (25.4-38 cm dbh) appears to be
declining, the density of the two largest size classes has been increasing. Basal area has
also increased by 36% since 1940, to 22.52 mZ/ha (Table 5.4). This species has increased
substantially in importance since 1940 (62.08% to 74.01%, Table 5.6). Already the most
important species in Tourney, A. saccharum may yet be increasing in importance.

There has been little directional change in F. grandifolia abundance. Stem
density in most size classes reached peak abundance between 1940 and 1960, but the
number of very large trees (2 63.5 cm dbh) has tripled since 1940. Basal area peaked in
1960, but 2004 values (6.53 mz/ha) were still higher than 1940 values (5.83 mZ/ha, Table
5.4). While importance values have decreased since 1940, the change has been small
(17.06% to 14.07%, Table 5.6). If A. saccharum is indeed increasing in abundance, it is
not doing so at the expense of F. grandifolia.

Two species (U. americana and T. americana) have shown steady declines in
stem density and basal area. Ulmus americana density declined 90% between 1940 and
1970 (Table 5.5), while basal area decreased by more than 90% between 1950 and 2004

(Table 5.4). Importance values declined from 4.61% to 0.89% (Table 5.6). This decline

156

was due to the loss of large trees killed by Dutch elm disease, which was present in the
woodlot by 1962 (Schneider 1963). The increase in stem density of smaller trees in 2004
is not due to resprouting from stumps: none of the observed U. americana saplings were
attached to stumps. It is interesting to note that, at least for the two smallest size classes,
U. americana abundance in 2004 was similar to levels of the 1950s.

T ilia americana density and basal area were highest in 1940 and, by 2004, both
had declined by about 50% (Tables 5.4 and 5.5). Importance values for this species
declined from 5.18% in 1940 to 2.13% in 2004 (Table 5.6). This decline may be due to
the decrease in edge effects: as forest regenerated around the east and west edges of the
study area, light availability would have declined. Unfortunately, the locations of the
pre-2004 trees are not known, preventing further examination of this hypothesis.

Five species (A. saccharum, P. serotina, F. americana, Q. rubra, and 0.
virginiana) reached peak density in 1970. Stem density in the smallest size class roughly
doubled for A. saccharum, F. americana, and 0. virginiana, and increased eight- to
tenfold for Q. rubra and P. serotina. This great increase in small stems, especially for the
less shade-tolerant species, may indicate an increase in light availability between 1960
and 1970. Changes in light availability may be caused by wind- or ice storms that create
large canopy gaps. Schneider (1963) noted two large blow-downs in a 1961 aerial photo
of Tourney. One of these is still visible in the 1970 aerial photo (Figure 5.1). The death
of large elm trees (Table 5.5) may also have contributed to increased light availability.
Surprisingly, T. americana, another fairly shade-intolerant species, showed no increase in

stem density or basal area in 1970 (Tables 5.4 and 5.5). If the change in stem density was

157

caused by gaps, these gaps could have been localized in the south of the woodlot, where
the other species are common.
Species distribution

Index of dispersion (IOD) analyses indicated that stems were clustered for all nine
species analyzed (Table 5.8). It was not surprising that total stem density showed non-
random spatial patterns: edge and gap effects cause higher densities of small trees. Gap
effects can also lead to clustering of shade-intolerant trees of any size. Williamson
(1975) found that less shade-tolerant species (Liriodendron tulipfera (tuliptree), Fraxinus
pennsylvanica (green ash), and F. americana) were clustered and that the size of the
clusters matched the size of canopy gaps in Hoot Woods. Other than A. saccharum and
F. grandifolia, most species listed in Table 5.8 were more common along the southern
edge of the woodlot. A clumped distribution was conﬁrmed by high IOD values for both
large trees and canopy trees. Interestingly, T. americana canopy trees were not clustered;
large trees were. As mentioned previously, F. grandifolia was more frequent in the
northwest portion of the woodlot, leading to a clustered distribution. The most surprising
result from the IOD analysis was that A. saccharum canopy trees were clustered. While
A. saccharum was present in every plot, canopy trees of this species were infrequent in
the northwest of the woodlot (Figure 5.8). It is possible that competition from F.

grandifolia reduced the size of A. saccharum trees in this portion of the woodlot.

158

   

 

 

_l.

Figure 5.8: Stern density (trees/ha) of canopy-sized A. saccharum trees in the study area.
Darker colors indicate higher basal area.

3:5: ,
“Tia .
h.-...J

o‘ 40 80 160 Meters

 

Figure 5.9: Core forest in Tourney Woodlot if edge effects penetrate 20 m (light gray)
and 90 m (darker gray) into the forest.

159

Clustering plots based on species abundance (total basal area) and species
composition (relativized basal area) resulted in similar groupings. Plots in the south of
the woodlot, characterized by the presence of less shade-tolerant species, tended to
cluster together (Table 5.9). Plots with high F. grandifolia basal area were separated
ﬁom plots with high A. saccharum basal area. While many of the F. grandifolia plots
were located in the northwest of the study area, A. saccharum plots were distributed more
evenly throughout the woodlot (Figure 5.4), as might be expected for such a dominant
species.

Extent of edge inﬂuence

Table 5.11: Edge widths in Tourney Woodlot, based on vegetation structure and
composition data from Chapter 2. Total area of the woodlot in 2004 was 11.7 ha.

 

 

 

 

Basis for width Aspect Edge width Edge area Interior area Core forest
(m) (ha) tha) (%)

structure (ave.) NE 10 2.3 9.4 80%
SW 20

structure (max.) NE 70 8.3 3.3 28%
SW 60

composition all 20 3.0 8.6 74%

(ave.)

composition NE 90 8.0 3.4 23%

(max.) SW 40

overall all 90 10.3 1.3 1 1%

maximum

 

 

 

 

 

The extent of edge inﬂuence can be calculated based on edge effects determined
from vegetation structure and composition (Chapter 2). If the average depth of edge
penetration (DEI) for structure is used, 80% of Tourney is core forest (Table 5.11). Using
the maximum DEI instead reduces the amount of core forest to only 28% of the woodlot.
Similarly, edge widths based on average and maximum composition DEI values reduce
core forest area to 8.6 and 3.4 ha, respectively. If edge effects penetrate as far as 90 m

(the largest DEI found), only 11% of Tourney is core forest (Figure 5.9).

160

These calculations included second-growth portions of the woodlot. If only the .
old-growth portion were considered, core forest areas would be much smaller. While
most of the east and west edges of the old-growth forest are now protected by younger
forest (Figure 5.1), the north and south edges are still exposed. Matlack (1994) noted that
forest edges are representative of earlier stages in forest succession. In Tourney, edge
effects are evident in the dominance of A. saccharum, as well as in changes in species
richness and composition, especially along the southern edge of the woodlot. Even
though portions of this site have never been clearcut, the forest’s edges are not

representative of old-growth beech-maple forest.

Conclusions

meies richness: Species richness of all stems (21 species) is higher than in the much
larger Warren Woods, possibly due to increased light availability at the forest edges.
These edge effects may allow relatively shade-intolerant species like Zanthoxylum
americanum and Crataegus spp. to persist. Unlike other sites, Tourney is dominated by
A. saccharum rather than F. grandifolia. This difference may be caused by the stand’s
small size: A. saccharum grows faster in gaps than F. grandifolia and will likely be the
better competitor close to edges as well.

Changes in species composition, stem density aﬂl basal area: Eight species recorded in
the original survey are no longer present in the study area. These species were mostly
shade-intolerant, and included Salix sp., Platanus occidentalis, and Populus tremuloides.
Three new species were recorded in 2004. Between 1940 and 2004, basal area increased
at an average rate of 0.07 mz/ha/year, while stem density of large trees (2 12.7 cm dbh)

decreased by 0.80 trees/ha/year. Total stem density (stems 2 2.5 cm dbh: 981.7

161

stems/ha), density of large trees (stems 2 10 cm dbh: 213.9 trees/ha), and basal area (33.7
mZ/ha) are comparable to values for other forests.

Extent of edge inﬂuence: Because Toumey Woodlot is small, the proportion of core
forest is greatly reduced by edge effects. Effects“. based on average values for vegetation
structure and composition reduce effective core forest by 20%; effects based on the
maximum observed values reduce core forest by as much as 89%. Consequently, this

1 1.7-ha fragment may only have as little as 1.3 ha of interior forest remaining. While
there are many similarities between Tourney and larger beech-maple forests, the
dominance of A. saccharum and differences in species composition along the south edge

show that forest size matters.

162

Chapter 6
Conclusions: The importance of edge effects in forest ecology

Edge effects are an important concern in forest ecology and management: 62% of
forest area in the continental United States is located within 150 m of an edge (Riitters et
al. 2002). The extent to which forest structure and species composition along the edge
differ from the forest interior is related to several factors, including forest type, aspect,
and vegetation layer. The present study of beech-maple and oak-hickory forest types
examined the response to edge effects by vegetation structure and composition, non-
native species, and the forest soil seed bank, as well as how edge effects inﬂuence long-
terrn canopy tree dynamics.
Vegetation structure and composition

Estimates of edge penetration based on measures of vegetation structure in both
forest types varied between 0 and 75 m and, on average, were greater in beech-maple
forests (12.7 m) than in oak-hickory forests (5.8 m). Height and stem density of herbs
were the only metrics for which signiﬁcant edge effects were found in all four aspects of
both forest types, suggesting that the herb layer is the most sensitive to edge effects.

Species composition showed a greater response to edge effects than vegetation
structure: signiﬁcant edge effects based on species composition were larger (ave. 24.3 m)
than edge effects determined from vegetation structure (9.5 m). Maximum edge
penetration based on understory species composition (all plants <2.5 cm dbh) was 90 m
in both forest types. While Sorenson similarity between the edge and interior was low

(beech-maple: 0.28; oak-hickory: 0.41), cluster analysis revealed that differences among

163

sites were even larger. No ‘edge’ indicator species were found, indicating that changes in
composition must be evaluated on a site-by-site basis.
Non-native species

While 51 (24%) of the vegetation species ”sampled were not native to Michigan,
only 12 were found in more than 1% of the plots. Six of these 12 (Alliaria petiolata,
Berberis thunbergii, Lonicera maackii, L. tatarica, Rhamnus cathartica, and Rosa
multiflora) are known to be invasive in forest ecosystems. More non-native species were
found in oak-hickory forests (43) than in beech-maple forests (26); non-native plants
were also more abundant in oak-hickory forests (ave. 1.2 stems/m2 vs. 0.1 stems/m2).

Lifeform and shade tolerance, two factors rarely considered in other studies,
explained a signiﬁcant but small (ave. R2=0.09) proportion of the variation in non-native
abundance. Canopy openness, on the other hand, increased R2 values by 0.21 in beech-
maple forests and only by 0.04 in oak—hickory forests. The greater increase in
explanation in beech-maple forests may indicate that light availability is more important
to non-natives in forest types, such as beech-maple, with a higher leaf area index.
Soil seed bank

A total of 88 species were germinated from soil collected in beech-maple and
oak-hickory forests. Shade-intolerant species made up 51% of species and 66% of seeds
in the seed bank, while annual species contributed 23% of species and 38% of seeds.
Species richness declined with distance into the forest, but this decline was not
statistically signiﬁcant. Seed density varied greatly among sites, ranging from

1677i1700 to 151332t55351 seeds/m2. Across all sites, seed density was signiﬁcantly

164

higher at the edge. The prevalence of shade-intolerant, annual species suggests that this
increase is due to inputs from adjacent habitats.

The 14 non-native species (16%) observed in the seed bank contributed 30% of
the seeds. Only one of the 12 most common non-natives (Rumex obtusifolius) present in
the vegetation was observed in the seed bank. While it was surprising that Alliaria
petiolata was not observed in the seed bank, soil samples may have been collected after
most spring germination occurred. If this were true, this invasive could be eradicated
from a site (barring re-introductions) if all plants were removed before seed dispersal.
Inﬂuence of aspect

Because abiotic edge effects penetrate farther into a fragment along ‘warm’ edges
(south and west), it was expected that biotic edge effects would also be larger along these
edges. Average edge responses in vegetation structure were indeed smaller along north
and east edges (8.3 and 5.8 m) than along south and west edges (17.7 and 15.7 m).
However, for species composition, the largest edge effects (90 m) were found for
northern edges. Furthermore, edge effects based on non-native stem density was smaller
in north edges (1.4 m outside the forest) than in east, south or west edges (6.0, 5.0, and
26.7 m). While seed bank density was higher in north edges (3.7 seeds/m2) than in west
edges (3.] seeds/m2), seed bank diversity was highest in north and south edges (H’=0.8
vs. 0.5 for east and west). These results cast doubt on the practice of grouping ‘warm’
and ‘cool’ edges together a priori; the validity of such groupings in different forest types
deserve further study.

Inﬂuence of forest type

165

Edge effects based on vegetation structure penetrated farther into beech-maple
forests (12.7 m) than into oak-hickory forests (5.8 m). In contrast, species composition
showed greater responses in oak-hickory forests (27.5 m) than in beech-maple forests (21
m). Differences from edge to interior in herb height (beech-maple: 0.5; oak-hickory: 0.3
m), understory species richness (3.3 vs. 2.1 species/m2), and Sorensen index (0.5 vs.
0.4%) were all greater in beech-maple forests than in oak-hickory forests. Non-native
species were more frequently encountered in oak-hickory forests (38% of plots) than in
beech-maple forests (11% of plots).

These differences could be due to the greater foliage area in beech-maple forests,
where interior LAI was 6.52i0.13 mz/m2 (compared to 4.45:L~1.16 mZ/m2 for oak-hickory
forests). Hemispherical photography, used to calculate canopy openness, was relatively
insensitive to these differences in LAI: although interior openness was signiﬁcantly
higher in beech-maple forests (2.93i1.23% vs. 2.67i1.30%), this difference was small.
CanOpy phenology may also contribute to these differences. Because Acer saccharum
and F agus grandifolia leaf out earlier and retain their leaves longer than Quercus spp.
and Carya spp., the high-light growing season for the understory in beech-maple forests
is shorter than in oak-hickory forests.

Finally, the oak-hickory forests in this study were all located in parks and game
areas and received more visitors than the beech-maple sites (mostly university natural
areas and private property). Furthermore, in at least one oak-hickory site, non-native
species were deliberately introduced (into ﬁelds), greatly inﬂating their presence there.

However, even the beech-maple site located in a state park was less invaded by non-

166

native species than the oak-hickory sites. These results suggest that beech-maple forests
may be more resistant to invasion than oak-hickory forests.
Long-term canopy tree dynamics

Edge effects can inﬂuence long-term canopy tree dynamics in small forest
fragments. Tourney Woodlot, an 11.7-ha old-growth maple-beech forest, is similar to
other, larger old-growth forests in total stem density (stems 2 2.5 cm dbh: 981.7
stems/ha), density of large trees (stems 2 10 cm dbh: 213.9 trees/ha), and basal area (33.7
mZ/ha). Species richness of stems 2 2.5 cm dbh (17-21 species) is only slightly higher
than that of other forests. However, unlike other old-growth forests in the region,
Toumey is dominated by A. saccharum (70% of basal area) rather than F. grandifolia.
This difference may be caused by the stand’s small size: in gaps, Acer saccharum grows
faster than F. grandifolia, and will likely be the better competitor close to edges as well.

The amount of core forest in Tourney is greatly reduced by edge effects. Effects
based on average values for vegetation structure and composition reduce core forest by
20%; effects based on the maximum observed values reduce core forest by as much as
89%. Consequently, this forest has only 1.3 to 9.4 ha of interior forest remaining. These
calculations included second-growth portions of the woodlot, which shield some of the
old-growth from edge effects. Those sections of old-growth at the forest edge are not
truly representative of old-growth beech-maple forest, even though they have never been
logged.

In conclusion, the effects of forest fragmentation extend beyond habitat
destruction. Because edge effects inﬂuence forest structure, species composition, and

non-native invasion, they must be taken into account when assessing the conservation

l67

value of forest fragments. As the example of Tourney Woodlot shows, the effective
interior area of a forest may be much smaller than it appears. However, even small
woodlots provide some beneﬁt as habitat for forest ﬂora and fauna that are less sensitive

to edge effects, and can serve as population reservoirs for future forest restoration.

168

APPENDIX

2004 TOUMEY WOODLOT DATA

169

Data below are organized by plot (Figure 5.2), species (abbreviations listed below), dbh (cm), height

category, and triangle. Height categories (understory (u), subcanopy (s), and canopy (c)) are deﬁned in

Chapter 2. Triangles were created by connecting plot comers and are labelled north (N), east (E), south

(S), or west (W). Species abbreviations: AceN=Acer nigrum, AceS=Acer saccharum, CarC=Carya

cordiformis, CarG=Carya glabra, CarO=Carya ovata,lCelO=CeItis occidentalis, CraS=Crataegus spp.,

FagG=Fagus grandifolia, FraA=Fraxinus americana, HamV=Hamamelis virginiana, JugN=Juglas nigra,
OstV=Ostrya virginiana, PruS=Prunus serotina, QueR=Quercus rubra, TilA=TiIia americana,
UlmA= Ulmus americana, UlmR= Ulmus rubra

 

 

a1
al
a1
a1
a1
a1
a1
a1
al
a1
a1
31
a1
a1
al
a1
a1
31
a1
a1
a1
a1
a1
a1
a1
al
al
a1
a1
a1
a1
a1
a1
a10
alO
a10
a10
310
310
310
a10
alO
a10
a10
a10
a10
a10
a10

AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
OstV
PruS

PruS

35.7
5.2
43.2
4.6
8.0
12.1
2.7
3.5
9.0
3.6
2.6
2.7
3.4
3.5
13.0
4.0
4.2
6.7
7.4
8.5
7.7
3.1
4.3
4.5
4.6
2.8
5.9
4.9
7.5
4.3
66.9
3.0
76.1
3.2
5.5
9.5
6.9
7.1
4.5
27.6
4.2
5.5
5.0
3.2
5.4
30.7
2.6
9.2

S

mmwmmmmmmmmmmmwmmmmmwommowommmmmmmommmmmmommmom

mmmmmmmmmmmmmmmgggggggﬁgmwmmmmmmmzzzzzzzzmmmmmmm

 

 

 

a10 PruS 5.6 s E
a10 QueR 50.3 c E
a10 AceS 21.6 s N
a10 AceS 2.8 s N
310 AceS 12.5 s N
a10 AceS 5.0 s N
a10 AceS 5.0 s N
a10 AceS 6.6 s N
a10 AceS 10.5 s N
a10 AceS 3.5 s N
a10 PruS 4.5 u N
a10 PruS 79.5 s N
a10 PruS 3.8 s N
a10 AceS 18.7 s S
a10 AceS 3.8 s S
a10 AceS 2.6 s S
a10 AceS 2.6 s 8
a10 AceS 3.2 s S
a10 AceS 3.1 s S
310 AceS 3.0 s S
a10 PruS 6.9 s S
a10 PruS 2.9 s S
a10 PruS 6.6 u S
a10 PruS 4.0 s 8
a10 PruS 7.6 s S
a10 PruS 3.6 s S
a10 PruS 5.1 s S
a10 PruS 4.2 s S
310 PruS 6.1 s S
a10 QueR 6.5 c S
a10 AceS 27.8 s W
310 AceS 6.7 s W
a10 AceS 3.7 s W
a10 AceS 3.6 s W
a10 AceS 8.9 s W
310 AceS 5.1 s W
a10 AceS 17.7 s W
a10 AceS 2.9 s W
310 AceS 2.7 e W
310 PruS 3.6 s W
a10 PruS 63.1 n W
a10 PruS 2.5 u W
a2 AceS 2.5 s E
a2 AceS 3.3 s E
a2 AceS 10.3 s E
a2 AceS 7.1 s E
a2 AceS 69.9 s E
a2 FagG 3.1 s E

 

 

 

 

 

1N

ﬁ$$$$ﬁ$$ﬁEﬂﬂﬂﬂﬂﬂﬂﬁﬂﬂﬂﬁﬁﬂﬂbﬂﬂﬁﬁﬁﬁﬂﬁﬂbﬂﬂﬂﬁﬁﬂﬂﬂﬂﬁﬂﬂ

FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
FagG
PruS

PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

4.4
38.5
3.1
16.4
28.9
10.8
7.5
4.6
5.6
7.6
13.4
8.5
8.1
6.9
94.6
8.0
2.7
2.8
4.5
11.5
3.4
7.5
36.1
6.4
6.2
52.8
6.7
6.3
4.2
3.8
7.9
2.9
48.4
3.4
57.1
3.3
6.2
6.5
4.7
2.6
2.6
10.2
3.7
2.9
3.6
2.6
2.8
4.1

mmmommmmmmmmommwmmmmomomwmwomoommmmgmmmommmmmmmm

mmmmmmmmmmmgggggggggggmwmwmmwwwwwmzzzzzzzzzzzzzm

 

 

 

 

a3 AceS 2.5 s E
a3 AceS 7.3 s E
33 AceS 41.3 s E
a3 AceS 39.3 s E
a3 AceS 2.9 s N
a3 AceS 3.8 s N
a3 AceS 3.3 s N
a3 AceS 2.7 s N
a3 FagG 4.9 s N
a3 FagG 61.2 s N
a3 FagG 2.7 s N
a3 AceS 17.8 s N
a3 AceS 3.4 s S
a3 AceS 5.2 s S
a3 AceS 17.2 s S
a3 FagG 2.7 c S
a3 FagG 3 .4 s S
a3 FagG 2.7 s 8
a3 AceS 3.1 s W
a3 AceS 5.5 e W
a3 AceS 4.2 s W
a3 AceS 5.2 s W
a3 AceS 4.3 s W
a3 AceS 5.8 s W
a3 AceS 2.9 e W
83 FagG 50.6 e W
84 AceS 2.7 s E
a4 AceS 3.1 s E
34 AceS 3.0 s E
a4 AceS 4.4 s E
34 AceS 2.5 s E
84 AceS 2.6 s E
a4 AceS 3.0 s E
34 FagG 2.9 s E
a4 AceS 4.7 s N
34 AceS 3.8 s N
a4 AceS 3.9 s N
a4 AceS 4.0 s N
a4 AceS 4.2 s N
a4 AceS 3.5 s N
34 AceS 2.9 s N
34 FagG 2.7 s N
a4 FagG 4.0 s N
a4 FagG 3.7 c N
34 AceS 2.7 s S
a4 A068 3.6 s S
a4 AceS 2.8 s S
84 AceS 2.6 s 8
a4 AceS 3.8 s S
a4 AceS 49.8 s S
a4 AceS 67.0 s 8
a4 FagG 10.7 s S
34 FagG 34.2 s S
a4 FagG 28.9 s S
34 FagG 16.0 s S
a4 FagG 28.7 s S

 

 

 

 

a4 AceS 5.2 s W a7 AceS 67.3 s N
34 AceS 2.6 s W 87 AceS 8.6 s N
a4 AceS 2.7 s W a7 AceS 6.0 s N
34 AceS 3.7 s W a7 AceS 4.5 s S
34 AceS 4.0 s W a7 AceS 2.9 s S
34 AceS 2.9 s W a7 AceS 29.9 s S
a4 AceS 2.6 s W a7 FagG 15.2 c 8
a5 AceS 3.1 c E a7 FagG 28.6 c S
a5 AceS 2.6 s E a7 PruS 5.2 s S
a5 AceS 2.6 s E a7 AceS 4.7 s W
a5 AceS 2.7 c E a7 AceS 4.6 s W
a5 FagG 3.8 s E 37 AceS 22.5 s W
a5 FagG 3.5 s E 37 AceS 6.4 s W
a5 FagG 3.0 s E a7 AceS 7.6 s W
a5 AceS 2.5 s N a7 AceS 20.0 s W
a5 AceS 3.2 s N a7 AceS 7.3 s W
a5 AceS 2.8 s N 87 FagG 7.9 s W
a5 AceS 3.9 s N 37 FagG 2.9 e W
35 AceS 2.8 s N a7 FagG 20.9 u W
a5 AceS 2.7 s N 37 FagG 3.4 e W
a5 AceS 3.7 s N a8 AceS 36.5 s E
35 FagG 2.7 c N a8 AceS 53.5 n E
a5 FagG 2.6 s N a8 AceS 12.1 s E
a5 AceS 2.6 s S a8 AceS 7.4 s E
a5 AceS 2.5 s 8 a8 AceS 5.3 s E
a5 AceS 4.9 s 8 a8 AceS 25.4 s E
a5 AceS 4.5 c S 38 AceS 67.3 s E
a5 AceS 63. l s W 38 AceS 57.3 s E
a5 AceS 55.3 s W a8 AceS 54.0 s E
a5 AceS 13.6 s W a8 AceS 50.6 s E
a5 AceS 10.0 s W a8 AceS 4.8 s E
a5 F agG 9.1 s W a8 AceS 3.1 s E
a5 FagG 40.2 s W a8 AceS 2.5 s N
a6 AceS 36.9 s E 38 AceS 2.7 s N
36 AceS 2.5 s E a8 AceS 2.6 s N
36 AceS 3. 1 s E a8 AceS 2.7 s N
36 AceS 5.7 c E a8 AceS 2.8 c N
a6 AceS 5.7 c E a8 AceS 3.0 s N
a6 AceS 8.5 c E a8 AceS 3.5 s N
a6 AceS 7.7 c E 88 AceS 3.5 s N
a6 FagG 10.6 s E a8 AceS 4.4 s N
a6 AceS 77.0 s N a8 AceS 4.7 s N
a6 AceS 2.5 s N a8 AceS 3.8 s N
a6 FagG 27.7 c N a8 AceS 30.9 s S
a6 AceS 5. l c S a8 AceS 3. 1 c S
a6 F agG 4.2 s 8 a8 AceS 3.6 s S
a6 FagG 16.8 s S a8 AceS 35.0 s S
a6 AceS 15.4 s W a8 AceS 2.6 s S
a6 AceS 10.5 s W a8 AceS 3.8 s S
a6 AceS 13.7 s W a8 AceS 4.2 s S
36 AceS 65.9 s W a8 AceS 2.5 s S
a7 AceS 45.1 s E a8 AceS 18.3 s S
a7 AceS 41.7 s E a8 AceS 7.8 s S
a7 AceS 21.6 s E 88 AceS 10.9 s S
37 AceS 6.3 s E a8 AceS 9.3 s S
a7 FagG 64.8 c E a8 PruS 4.2 u S

 

 

 

 

171

 

 

 

 

a8
a8
a8
a8
a8
88

a8
a8
a8
a8
a8
a8
a8
a8
88
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9

PruS
PruS
PruS
AceS
AceS
AceS
FagG
FagG

F agG
PruS

PruS" '

PruS

PruS

PruS

PruS

PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

M6
m
40
42
u
m
m
65
35
5.4
n2
m
m
95
35
3.8
u.o
26
m
3.7
mo
34
67
74
M.7
5.3
m.7
28
m
7A
98
4.7
5.5
”7
2.8
”.0
M
M
M
n.1
59
5.4
u.5
26
32
9.6
m.7
yd
H5
my
B7
34
88
5.8
3.1
88

U:mmmmmmmmmmmmmmmommmmmmmmwmmmmmmnwmmmmmmmmmmmmmoooommmcw

mmmmmmmmmmmmmmmmzz222222222:zzzmmmmmmmmmmgggggggggggggmmm

 

 

 

 

a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
a9
b1
bl
b1
b1
b1
b1
b1
b1
b1
bl
b1
b1
bl
b1
b1
bl
b1
b1
b1
bl
bl
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10

AceS
FagG
HamV
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

PruS

PruS

AceS
AceS
AceS

58.1
5.8
3.2
58.0
3.6
7.5
8.0
2.5
5.7
7. 4
26.0
27.1
2. 9
11.1
14.0
3.0
33.9
2. 9
11.8
64.0
78. 2
47.9
2.7
3.9
4.4
3.0
2. 7
10. O
8.4
3.7
3.1
58.5
10. 7
5. 8
17.4
8.6
8.7
6.8
9. 3
10.4
61. 2
5. 9
10.1
7.2
8.5
6.8
6. 2
79. 6
3.9
5.3
5.8
7. 6
18. 4
3.3
4.6
5.1

mmmcmmmmmmmmmmmmmmmmmmmmmmmmmomomommmmomomommmmmmmmmmmom

wwwzzzzzzzzzzzzzmmmmmgagegwmwmmmzzzzzzmmmmgeggggggmwmwww

 

 

 

 

172

b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b10
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2
b2

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
PruS
PruS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG

3.0
5.3
6.0
11.0
8.4
5. 2
12. 6
8.4
9.3
9.7
5.4
3.3
2.5
5.1
2.5
3. 9
24. 7
2.8
7.9
7.0
16. 4
7.5
9.7
9.7
9. 7
17.5
16.4
77.1
2.8
3.9
4.7
2.7
3.4
4.7
4.1
3.1
4. O
43. 5
4.6
3. 0
42. 7
2.9
3.2
5.6
5.1
3.4
3.5
2.7
3.0
11.2
2.6
2.7
44.8
5. 6
56.6
2.5

moommmmmmmmmmwoommwmmmmmmmmmmmwmmmmmcmmmmmommmmmmmmmmmow

2222222zzzzzzzzmmmmmmmmmmmmmmmmmmmmggggggggggmmmmmmmmmmm

 

 

 

 

 

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
F agG
PruS

AceS
AceS
AceS
AceS
AceS

2.7
6.1
3.2
3.6
2.5
3.1
19.1
4.8
4.2
2.8
2.5
6.0
2.5
2.7
3.7
2.6
4.4
61.6
5.3
8.2
4.7
3.2
4.5
11.8
3.5
7.0
3.7
2.8
3.5
3.7
2.8
3.2
2.8
2.9
3.7
6.5
2.8
3.2
29.0
7.5
7.2
3.0
2.8
31.5
7.9
7.6
2.6
2.5
2.5
3.2
37.6
3.6
3.7
3.1
72.7
3.9

mmmmmmggggggggmmmmwmmmmzzzzzzzzzzzzzzzzmmmgssggggzggmmww

 

 

 

E$18832ZIEEEIIIIIIIIEEEIIIXIIIIIZIEIE

AceS
AceS
AceS
FagG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
FagG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
FagG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
AceS
AceS
AceS
FagG
AceS

3.9
30.0
90.1
6.0
7.6
5.1
10.6
2.7
3.5
19.1
5.3
62.2
3.1
3.7
17.3
31.7
6.1
9.2
13.0
23.7
26.3
5.6
92.1
15.2
7.3
2.7
3.0
3.8
4.2
18.1
4.5
5.4
28.9
11.9
2.6
3.7
6.3

5.9
7.9
4.9
5.4
5.7
4.2
6.8
5.8
2.8
2.7
6.1
73.5
3.3
6.6
3.9
4.6
2.5
22.1

mmmmmmomcmmmmcmommammmmmmvat/20mmmmmmmmmmommmmmmmmommgmmwarm

gmmmmzzzzzzzzzmmmmmgggggggmmmmmmmmmmmmzzzzzzzzzzzzmmmmmm

 

 

 

b5
b5
b6
b6
b6
b6
b6
b6
b6
b6
b6
b6
b6

b6
b6
b7
b7
b7
b7
b7
b7
b7
b7
b7
b7
b7
b7
b7
b7
b8
b8
b8
b8
b8
b8
b8
b8
b8
b8
b8
b8
b8
b8
b8

b8
b8
b8
b8
b8
b8
b8
b8
b8
b8

AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
F agG
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

2.5
2.6
14.4
69.6 ,
5.1
7.8
14.1
5.6
5.2
2.9
3.0
4.1
3.4
3.1
3.0
3.6
40.3
2.7
4.1
3.7
2.9
2.5
3.1
4.0
2.8
3.5
36.1
3.0
2.7
3.5
2.5
36.0

3.2

6.3
3.1
3.0
3.0
6.0
4.4
3.5
20.5
2.7
3.2
29.2
19.8
3.7
4.6
80.8
70.1
45.3
39.0
5.2
2.5
6.2
3.7
4.6

tnmmmmmmmmmmmmooommmmmmmmmmocmommmommwommommmmmowmmmcwowm

saunasbbabnanooooooooooooogggmmmmaaooooogzzzzzzzzzmmmmgs

 

 

173

 

 

 

b8 AceS 55.8 s 11
b8 AceS 3.7 s 11
b8 FagG 5.4 c 11
b8 AceS 4. 3 s s
b8 AceS 11.4 u s
b8 AceS 58.2 s 3
b8 AceS 13.1 s 3
b8 AceS 12.5 s s
b8 AceS 3.5 s 3
b8 AceS 2. 7 s s
b8 AceS 11.3 s 3
b8 AceS 3.5 s s
b8 AceS 3.8 s 8
b8 AceS 7.2 s s
b8 AceS 3.8 c s
b8 FagG 4 .4 s 3
b8 FagG 75. 8 s 5
b8 AceS 3. 6 s w
b8 AceS 23. 3 s w
b8 AceS 3.9 s w
b8 AceS 3.4 s w
b8 AceS 1 1.1 s w
b8 AceS 50.3 s w
b8 AceS 5.5 s w
b8 AceS 2.5 s w
b8 AceS 4.4 s w
b9 AceS 6. 3 s E
b9 AceS 55.1 s E
b9 AceS 7. 9 c E
b9 AceS 28. 8 c E
b9 AceS 4.9 s E
b9 AceS 6.0 s E
b9 AceS 2. 8 s E
b9 AceS 83. 5 s E
b9 AceS 4.6 s E
b9 AceS 74.0 s E
b9 AceS 8.4 s E
b9 AceS 3.5 s E
b9 AceS 7 .3 c E
b9 AceS 2.9 c 11
b9 AceS 3. 0 s 11
b9 AceS 16.2 s 11
b9 AceS 2.6 s n
b9 AceS 3.1 s 11
b9 AceS 2.7 s n
b9 AceS 2.6 s 11
b9 AceS 6.3 s N
b9 AceS 8.0 s N
b9 AceS 8.2 s N
b9 AceS 5. 8 s N
b9 AceS 17. 4 c N
b9 FagG 55.1 s N
b9 FagG 51.4 c N
b9 FagG 11.2 c N
b9 FagG 8. 6 s N
b9 FagG 70. 8 s N

 

 

 

 

 

b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
b9
cl
c l
c 1
c 1
Cl
C 1
c 1
c1
0 1
c 1
cl
c 1
c1
c1
0 l
c 1
c 1
c 1
c 1

FagG
FagG
FagG
FagG
FagG
HamV
OstV
PruS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

14.3
7.3
4.3
2.7
7. 0
61.2
5.7
7.5
4.3
2.5
6.0
2.8
3.3
3.2
2.5
5. 5
63. 5
3.1
8.5
3.6
7.7
5. 6
10. 8
11.1
13.1
13.3
14.0
14.5
2. 7
24. 8
12.1
2. 7
12. 9
6.3
8.7
7. 0
14. 0
76.8
10.2
11. 5
6. 3
19. 4
12.1
14.2
13. 8
5.1
13.2
5. 6
10. 0
7.2
3.3
4.2
5. 5
11.6
6.1
3.8

mocha:ommomwmmooommmmmmmwmmmmmmmmmmmmmmmmmmmmmmcmmmcgmwmmm

gsggwwwwmzzzzzzzzmmgggggggeggggggggsasswmmmmmmmmzzzzzzzz

 

 

 

174

CID
c10
c10
c10
CIO
010
010
010
c10
c10
c10
c10
c10
c10
c10
c10
c10
C10
010
CID
c10
c10
c10
c10
c10
c10
cll
cll
cll
c11
cll
cll
cll
c1]
01]
cll
cll
CH
c1]
c1]
cll
c11
cll
cll
cll
cll
cll
cll
c1]
c1]
c1]
cll
CH
CH
cll
cll

AceS
AceS
AceS
AceS
AceS
OstV
OstV
OstV
OstV
OstV
OstV
OstV
PruS

PruS

PruS

AceS
AceS
AceS
AceS
QueR
AceS
AceS
AceS
PruS

PruS

TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
QueR
QueR
QueR
QueR
QueR
QueR
QueR
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FraA

4.1
3.8
3.1

' 4.2

3.1
5.4
7.6
8.4
8.7
5. 5
71. 2
36.1
3. 2
16. 6
3.1
49. 3
3.4
3.3
2.6
2.7
2.7
2.6
2.9
2.6
5.5
3. 2
46. 6
4.5
4.3
3.5
7.4
2.7
3.5
2.9
6.1
3. 5
14. 2
3.5
3.0
4.4
2.9
3.1
32. 6
2.5
3.5
2.9
3. 0
47.1
2.6
2. 6
14.9
3.3
3.0
2.7
3.2
4.0

zzlzzzzzzmmmmmmmmmmmmmmmmmmmmmmggggﬁgmmmmmzzzzzzzmmmmmmmm

 

 

 

 

cll
cll
cll
cll
cll
cll
cll
cll
cll
cll
c11
cll
cll
c11
cll
c11
cll
cll
cll
cll
cll
c11
cll
cll
cll
c11
cll
CH
CH
cll
CH
CH
cll
cll
cll
cll
CH
CH
cll
cll
cll
c2

c2

c2

c2

c2

c2

c2

02

02

c2

c2

c2

c2

c2

c2

F raA

OstV
QueR
QueR
TilA

AceS
AceS
AceS
AceS
AceS
PruS

QueR
QueR
QueR
QueR
QueR
QueR

UlmA
UlmA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
CarO
QueR
QueR
QueR

2.8
5.5
3.1
6.9
4.0
3.6
3.5
3. 0
13. 7
2. 6
10. 7
26.0
2.5
2. 5
76. 3
53.1
86.1
4.0
3.2
2.6
2. 6
58.3
52.6
2.7
3.8
3.5
2.9
5.3
3.9
3.9
2.8
7.7
4. 3
57.0
6.8
3.7
6.1
11. O
3.2
7. 8
67. 9
4.2
2.7
4.9
2.7
5. 8
57.2
2.9
5.4
3.7
4.2
3.5
2.6
5.4
4.1
6.0

mmmmmmwmmmmwmmmmmmmomommwmmmmmwmmmmmmmmomommocmmcmmmmmmo

 

 

 

 

c2
c2
c2
c2
02
c2
c2
c2
c2
c2
c2
c2
c2
c2
02
c2
c2
c2
02
c2
c2
c2
c2
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3
c3

AceS
AceS
AceS
AceS
AceS
PruS

PruS

PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS
FagG
FagG
FagG
FagG
FagG
FagG

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
F agG
UlmR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS

5.5
2.9
2.7
6. 3
63.7
4.6
5. 4
11. 7
2. 8
33.0
11.8
24.3
4.2
2.9
3.9
4. 6
37. 8
24.8
2.7
3.0
5.4
6. 7
11.2
2. 7
17. 9
8.1
3.3
4.5
2. 8
62. 7

8.6 .

7.1
6.4
6. 5
31. 8
17.1
10. 6
3. O
49.9
24.2
6.3
7.8
5.1
10. 1
6. 6
52. 3
3.2
2.5
2.9
4.3
3.4
5.3
7.0
2. 9
l4. 4
14.0

mmu:mmma:mm5:35:32:at:aunt-3000000ooooooagggggggggggggggmmmmmmmm

 

 

 

 

175

3888883388388388

$833

04

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

16.0

33.6
40. 3
3. 2
16.3
13.1
11.5
7.1
8. 4
l3. 3
11.6
11.3
24.1
81
31.2
35.9
4.3
2. 5
21. 5
10.7
13.0
3.6
3.5
2. 5
10.1
9.1
5.5
4.9

13:33:35DDDﬁﬂbﬂﬂbbﬁﬂﬂ5590G000000000€£ﬁggggggggfggggggmmmmmm

 

 

 

 

 

c4 AceS 13.6 s n c5 FagG 5. 0 s w
c4 AceS 38.4 s 11 c6 FagG 59.0 s e
c4 AceS 49.0 s 11 c6 AceS 6.4 u n
04 AceS 2. 9 s 11 c6 AceS 4. 3 s 11
c4 AceS 14. 3 s 11 c6 AceS 66. 7 s 11
c4 AceS 5.2 s 11 c6 AceS 8. 4 s 11
c4 AceS 4.2 s s c6 AceS 36.7 s 11
c4 AceS 5.3 s s c6 AceS 9.3 c 11
c4 AceS 2.6 s 3 c6 FagG 2.8 c 8
c4 AceS 4.5 s 3 c6 FagG 5.3 c s
04 AceS 7.6 s 8 c6 AceS 4.7 s w
c4 AceS 6.2 u 8 c6 AceS 2.6 c w
c4 AceS 5.7 s 8 c7 AceS 2.7 s e
c4 AceS 4.6 s w c7 AceS 4.6 s e
c4 AceS 4.2 s W C? AceS 6.6 s e
c4 AceS 2.6 s w c7 AceS 4.1 s e
c4 AceS 4.6 s w c7 AceS 19. 8 c n
c4 AceS 2.9 s w c7 FagG 6.0 c 11
c4 AceS 4.1 s w c7 AceS 4. 7 s 3
c4 AceS 5.1 s w c7 AceS 10. 2 c 8
c4 AceS 15.5 s w c7 AceS 4.3 u 8
c4 AceS 3.0 s w c7 FagG 9.8 s s
c4 AceS 5.1 s w c7 AceS 4.1 s w
c4 AceS 2.9 s w c7 AceS 8.4 s w
c4 AceS 7.1 s w c7 AceS 6.3 s w
c5 AceS 2.8 s e c7 FagG 5.6 s w
c5 AceS 6. 4 s e c7 FagG 6.9 c w
c5 AceS 15. 9 s e c8 AceS 8.3 s e
05 AceS 9.3 s e 08 AceS 9.2 s e
c5 AceS 5.5 s e c8 AceS 5. 8 s e
c5 AceS 7.8 s e c8 AceS 80. 2 s e
05 AceS 7.4 s e 08 AceS 7.8 c e
c5 AceS 6.1 s e c8 AceS 9. 3 s e
c5 FagG 10. 8 s e c8 AceS 11. 6 s e
c5 FagG 22.8 s e c8 AceS 3.3 s 11
c5 AceS 42.7 s 11 c8 AceS 5. 9 s n
c5 AceS 10.0 n 11 c8 AceS 10. 2 s 11
c5 AceS 3.7 s n 08 AceS 33.9 s n
c5 AceS 3.6 s 11 c8 AceS 45. 3 s 11
c5 AceS 6.5 s 11 c8 AceS 9. 6 s n
c5 FagG 6.6 s n c8 AceS 66.7 s s
c5 FagG 9.5 s 11 c8 AceS 2. 7 s 5
c5 FagG 3.0 s n c8 AceS 13. 0 s 3
c5 FagG 2.5 s 11 c8 AceS 7.0 s s
c5 TilA 2.9 c 11 c8 AceS 5.6 s s
c5 AceS 5.6 s s c8 AceS 2. 5 s s
c5 AceS 4.0 s 3 c8 AceS 25. 3 s s
c5 AceS 2.6 s s c8 AceS 4.6 s s
c5 FagG 5.7 c s c8 AceS 9. 3 s s
05 FagG 4.4 s 3 c8 AceS 27.1 s 8
c5 FagG 7.0 s s c8 AceS 6.0 s s
c5 FagG 2. 5 s 8 c8 AceS 2.8 s s
c5 AceS 10. 5 s w c8 AceS 2.9 s w
c5 AceS 9.9 s w c8 AceS 2.6 s w
05 AceS 4. 0 s w c8 AceS 2.8 s w
CS 389 19. 3 s w c8 AceS 3.6 s w

 

 

 

 

 

 

 

176

c8
c8
c8
c8
c8
c8
08
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9
c9

'c9

c9
c9
c9
c9
c9
c9
c9
c9
c9

AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
OstV
OstV
OstV
OstV
QueR
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS "

AceS
AceS
AceS
AceS
AceS

19.0
7.6
8. 8
29. o
3.7
3.9
7.1
4.8
4.5
3.4
4.3
7.6
4. 2
45.0
4.5
2.7
6. 9
1o. 7
2.8
5. 5
12. 3
9. 8
10. 6
70.0
4. 4
48. 8
10.1
35.0
2.9
6.8
6.0
5.5
3. o
10. 6
3.7
3. 3
14.8
43.6
3.5
5. 7
33.7
37.7
10.0
13.1
12.7
60.8
6.1
5. 2
51.9
7.2
6.9
7.2
5. o
43.6
38.2
5.5

mmmwmmmmmmmmmnwmc:mmmmmmommmmmmommmmmmwmmwmmmmoommommmmmm

gggaggﬁ£8ggismtnmmwmmmmmmmmmanpanaanbnbaooooaoooooggggggg

 

 

 

 

 

c9
c9
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d1
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10
d10

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
F agG
AceS
AceS
AceS
FagG
FagG
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
AceS

6.4
9. 6
71.7
74.3
7.0
9.4
4. 2
54. 9
9.2
9. 5
65. 8
4.0
2. 9
46.7
36.7
2.7
5.9
8.9
5.7
5.6
3.2
5.5
6.5
4.8
3.1
2.7
3.0
2.9
3.7
2.6
2.5
5.2
4.3
8.8
6.1
2. 5
10. 6
12.9
10. 8
3. 8
13.2
4. 5
l6. 3
6. 7
29. 4
2. 6
11.3
2. 5
57.7
2. 5
62. 7
2.7
3.5
2. 7
37.3
7.2

 

 

gmmmmmmmmmzzzzzzzzzz222mmmmmmmmmmmmmggmmmmmzzzmmmmmmmmss

d10
d10
d10
d10
(111
(111
dll
dll
d1]
d1]
dll
dll
dll
(111
d1]
d1]
(111
(111
(111
<11]
(111
(111
dll
(111
(111
(111
dll
(111
(111
(11]
d1]
dll
(111
(111
dll
dll
dll
dll
dll
(111
(111
(111
dll
dll
(111
(111
(111
d1]
(111
(111
dll
dll
dll
dll
dll
d11

AceS
AceS
AceS
AceS
FraA

AceS
AceS
AceS
AceS
AceS
AceS
QueR
AceS
AceS
AceS
AceS
QueR
AceS
AceS
AceS
AceS
AceS
AceS
QueR
AceS
AceS
AceS
AceS
QueR
AceS
AceS
AceS
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

TiIA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

4.1
9. 8
35. 7
6.2
2.9
54.2
12.5
2.7
2.7
4. 5
39. 8
3.5
3.1
3.1
2.9
4.2
4.1
3. 6
56.5
5.6
5. 8
78.1
6.9
5.9
5.0
2. 6
62.8
6.4
2.6
2.9
3.4
3.7
3.8
5.8
5.3
6.7
8.9
2.6
3. 0
10. O
3.8
4.2
4.2
3.7
3.3
2.7
5.5
2.7
3. 5
10.1
5.4
5.7
4.3
6.0
2. 6
18. 3

mmmmmmwmmwmmmmwmmmmmmmmmwwmommmwoommwmcomwmmommmmmmomwmm

gigssggsgﬁggmmmmmwwwmmmmmmrowanabut:anoooaoaoaoooaoonaéggg

 

 

 

8858383838383858585838383858:838383&83838383&8385&E383%EBBEEBBBBESSBESESESBSB

 

 

1W

dll
dll
d11
dll

CLO-n.0-
p—op—ot—op—I
pan—spay...

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG

AceS
AceS
AceS
AceS
AceS

74.6
7.5
4. 1
10. 3
3 .4
l3. 4
5.3
2. 8
13. 3
11.4
9.9
4.5
5.5
5.4
6.3
7.5
2.7
5.8
3.5
7.9
6.7
6.6
7.6
6.9
4.3
2.8
3.0
5.0
4. 5
70. 5
41.0
2.5
4.5
4.1
6. 8
79. 5
6.7
8.8
5.7
7. 8
54. 0
5.6
6.9
3.5
4. 8
70. 2
2.7
7.0
2.8
8.6
2.6
67.7
3.1
8. 9
34. 4
25.0

mart/3mmmmmmommmmmmmmommmmmwmmmmmwmommmmmmmmommmmmmmmmmmmmm

mmmmmseen:beenbananaooooonoooégmmmmmzzzzzmmmmmmmsitseﬁas

 

 

 

 

88888888888

88888

888888888888

d4

AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
AceS
FagG
F agG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
F agG

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

3.0
4.7
26.7
2.5
7.7
3.3
2.7
10.0
71.7
4.5
6.0
3.9
5.2
5.8
3.6
5.7
44.5
76.1
63.5
7.2
4.7
6.6
3.0
33.7
9.3
7.3
4.2
6.2
5.7
25.7
3.1
2.9
2.8
47.5
3.1
2.8
2.6
32.2
16.2
14.8
8.2
17.5
8.7
10.8
3.4
8.0
3.7
3.3
2.8
2.8
4.4
48.0
2.8
2.6
7.9
8.1

sssﬁs888$ésmmmmmmwmwmwwpeenneeeeeuseneeoaoooonogggggmmmm

 

 

 

888888888888888888888888888&8888888888888888888

mar—ac.
O‘O‘QQ

AceS
AceS
AceS
F agG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
F agG
FagG
FagG
TilA

AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
FagG
FagG
F agG
FagG
FagG
FagG
FagG
AceS

AceS
AceS

47.3
8.3
7.8
6.1
69.7
3.8
2.7
2.5
3.3
3.7
2.7
3.4
4.2
3.1
3.3
2.5
39.9
3.0
4.7
5.6
2.6
2.9
3.1
7.9
2.5
55.6
8.5
28.4
3.2
5.0
16.7
2.5
13.9
14.9
11.6
12.8
71.1
9.3
4.1
2.6
11.0
3.8
55.4
67.2
10.6
4.1
46.3
43.8
6.4
4.1
25.6
29.3
2.9
3.0
4.2
2.6

0000fssgsgggsggssmmmﬂﬂbbbbﬁbbb§§DGOGOOOGGOOGOOOOOGO€€ﬁgs

 

 

 

 

 

178

d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d6
d7
d7
d7
d7
d7
d7
d7
d7
d7
d7
d7
d7
d7
d8
d8
d8
d8
d8
d8
d8
d8
d8
d8
d8

d8.

d8
d8
d8
d8
d8
d8
d8

FagG
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
F agG
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
F agG
TilA

TilA

TilA

TilA

TilA

TilA

AceS
AceS
F agG
FagG
AceS
AceS
AceS
AceS

4.3
3.2
3.9
3.8
2.9
3.8
2.9
59.9
62.5
50.3
53.6
3.4
3.9
8.2
4.0
5.5
3.4
3.3
2.8
6.0
8.9
3.3
5.0
5.9
7.2
8.2
8.8
6.1
6.6
5.1
2.7
48.6
3.1
35.0
3.4
6.0
61.9
3.4
3.1
3.0
3.1
4.3
45.2
45.4
4.6
6.0
4.6
3.6
4.8
6.3
4.0
6.1
54.5
3.6
6.3
3.4

nmmmmmmomommmmm03mmmmommmo0:02:00oneommwmmmocmmomommwmcwm:m

smmmbbunoaooooooooogﬁgmmmpaunoooggggsggammmmaunnpunnunoo

 

 

 

 

d8
d8
d8
d8
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
d9
e 1
e 1
e1
e1
e 1
e 1
e1
el
e1
e1
e 1
e1
e1
e1
e1
e 1
e 1
e1
e 1
e 1
e1
e 1
e1
e1

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
TilA

TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

4. 3
37.0
45.2
60.7
37.3
47.2
5.3
3. 7
11. 5
19.7
11.6
9.7
9. 3
14. 9
20.7
10.7
8.6
7.4
5. 2
13.2
61.8
3.0
3. 2
70.1
5.1
46.0
2.6
6.8
4.3
2.6
3.8
7. 0
ll. 6
9.9
2.6
2. 9
10.4
2.6
4.7
4. 3
16. 2
9.7
2.5
4. 9
16.5
3.9
4. O
11.3
7. 9
65.2
7.4
3.2
2.8
5. 9
12.4
2.5

mmmmmmmmmmmmmmmmonmmmmwmommmmmmmmmmmmmmmmmmmmmommmmmomom

mmmmmmzzzzzzzzzmmmmmmmmmssgas:swwmmwwmm=b=b555=00““°£s£€

 

 

 

 

e1
e1
e1
e1
e1
e1
e1

e1
e1
e1
e1
e1
e1

e1

e1

e1

e1

e1

e1

e1

e1

e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e10
e11
e11
e11
e11
e11
e11

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
TilA

TilA

TilA

F agG
AceS
FagG
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
TilA

AceS
AceS
AceS
AceS
AceS
AceS

42.4
3.8
3.6
3.0
4.1
5.5
3.3
4.8
5.5
8. 8 i
53.0
29.1
4.4
5. 3
11. 8
14.9
15.5
37.7
41.1
42.8
53.5
8. 2
11.0
11.1
57.3
8.2
8.2
4. 0
65. 9
41. 6
2. 7
39.4
40.0
3.3
4.7
3.7
4.1
4.7
3.5
4.4
3.1
35.2
68.6
3.8
6. 7
53.7
41.6
18.5
54. 4
4. 0
55.2
2. 6
52.4
4.9
9.5
4.7

mmmmmmommwmmmwmmmmmmmomwmmwmommmmmmmmmmmmmmmmmommmmommom

mmmmmmse2222seeeewwwwzzzzzzzzzmmm22424822222eeswwwmwwwww

 

 

 

 

179

e11
e11
e11
e11
ell
e11
ell
ell
e11
e11
e11
e11
e11
e11
e11
e11
e11
ell
e11
ell
ell
e11
e11
e11
ell
e11
e11
ell
e11
ell
e11
e11
ell
e11
e11
e11
e11
ell
e11
e11
ell
e11
e11
e11
e11
e11
e11
e11
ell
e11
e11
e11
e11
ell
e11
e11

CarC
PruS

AceS
AceS
AceS
AceS
AceS
CraS

OstV
OstV
OstV
OstV
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
PruS

PruS

QueR
QueR
QueR
QueR
QueR
QueR
QueR
QueR
QueR
QueR

AceS
AceS

9.0
57.4
2.6
4.1
61.2
2.9
3.4
2.7
2.7
3. 4
12.6
5.0
2.7
2.9
6.1
7.1
3. 0
46.1
12.8
33. 7
3. 7
49.0
55.5
61.5
4.6
5.5
4.3
3.5
4. 4
61.3
9.6
9.1
2.5
6.8
4.6
6.5
8.3
3 .4
10. 4
3.7
6.5
2. 7
26.1
22. 4
3. 7
49. 7
6.3
5.6
4.9
4.8
3.3
4. 6
73. 8
4.6
5.7
5.8

mwmtnVar/amenmmwmmmmmmmmmmmmm0mmmmmmmmwmmmwmmmmmmwmmmoomwmmmm

gammmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmzzzzzzzzzzzmmmmm

 

 

 

 

e1 1
e1 1
e1 1
e1 1
e1 1
e1 1
e1 1
e1 1
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
62
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
e2
62
e2
e2
e2
e3
e3
e3
e3
e3
e3
e3
e3
e3
e3

AceS
AceS
FraA

OstV

QueR
QueR
QueR
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

21.5
2.7
12.0
5.0
23.3
4.7
7.9
6.3
6.1
9.2
2.5
2.5
4.6
6.2
7.4
3.5
6.7
3.2
7.0
77.3
5.7
8.8
3.7
2.5
65.6
6.5
5.7
3.7
4.5
5.1
39.4
7.7
6.0
47.5
35.7
4.5
4.6
6.5
9.4
7.0
6.4
7.8
67.6
56.8
11.3
3.5
7.4
3.5
4.2
2.8
53.2
36.6
3.3
6.0
3.4
3.7

mmmmmmmmmmommmmmmmmmnmmmmmmmmmmmmmommommmmmmmmmmmmmmmomm

mmmmmmmmmmgggggggggmmmwwmwmzzzzzzzzzzzzmmmmmmmmmséeéﬁéﬁé

 

 

8888888888888888

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

2.7
3.0
22.7
2.6
2.7
3.2
39.1
2.5
2.6
102.4
74.6
6.2
66.7
54.2
3.6
17.8
60.3
26.7
6.1
2.7
6.0
5.5
7.2
3.4
4.9
55.4
4.6
7.4
4.3
7.8
5.6
4.2
10.4
5.3
4.4
6.7
58.9
3.6
4.5
4.7
3.6
5.4
4.4
4.9
4.7
5.8
4.9
4.6
3.3
4.3
6.1
6.5
30.5
5.6
5.6
6.6

mmmmmmmommmag:mccmmoommmmmmmmmommmmmcommmmommmmmmmmmmmmmm

anabbannnoooooooooégégégggéggmmmmmmmmmmmmmmmzzzzzzzzzzmm

 

 

 

 

 

1%

88888888888888888

FagG
AceS
AceS
AceS
FagG
F agG
F agG
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
PruS

AceS
AceS
AceS

AceS ~

AceS
AceS
F agG
FagG
FagG
AceS
AceS
AceS
AceS
FagG
FagG
FagG
PruS

PruS

PruS

PruS

PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

62.8
11.2
4.3
7.4
5.3
3.9
6.8
5.1
5.4
25.6
5.7
61.1
54.1
4.3
4.4
6.6
14.8
7.6
3.2
4.5
64.4
5.0
9.4
5.5
2.9
4.3
2.6
2.8
10.6

4.8
7.5
7.7
13.5
13.6
45.1
17.1
3.3
8.0
2.5
3.0
2.8
28.8
3.3
4.0
5.3
2.5
3.1
2.5
4.3
7.3
4.0
3.3
3.3
33.7
4.0

cm:mmm:mmmczmmmmoommcmmo:mccgmmmmmgcgggmommmmmmmmmomwmmo

€€€€€€€€ﬁ“mm“"’V’V’V’V’mmmbﬂﬂﬂbbﬂbboonoooooaggggfgggﬁgmmmwmwa

 

 

 

 

e5
e5
e5
e6

e6
e6

e6

e6
e6

e6

e6
e6

e6

e6

e6
e6
e6
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e7
e8
e8
e8
e8
e8
e8
e8
e8
e8
e8

AceS
AceS
AceS

AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
FagG
PruS

PruS

PruS

PruS

AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
FagG
PruS

PruS

AceS
AceS

AceS
AceS
AceS
AceS
AceS
FagG
PruS

AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
FagG
FagG
AceS
AceS

AceS
AceS

5.7
4.8
3.8
2.6
4.4
15.2
2.5
15.5
4.1
74.9
6.8
6.8
4.9
59.5
13.9
3.7
3.0
3.5
3.8
3.9
4.7
5.0
4.0
4.7
2.6
5.9
4.4
10.3
6.8
61.2
6.1
3.1
3.2
4.4
3.4
2.7
54.4
2.7
3.0
22.8
4.0
3.3
3.9
30.2
2.6
3.2
4.8
6.1
4.0
3.2
5.5
2.7
6.5
47.1
3.1
5.2

omommommmoogmg:mmomc03mg:garment/acne:var/1mmmcomOGGSmocomcccccmmmm

ggwmmeeneoggsggmmmseer:be:ooogggggmwmwwmeeeeeeooooooonggg

 

 

 

2222222222222222222222222222222222222

AceS
FagG
FagG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
F agG
AceS
AceS
AceS
AceS
TilA

TilA

TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

3.9
4.1
3.1
3.4
2.9
2.6
4.2
2.5
4.3
3.1
2.5
2.6
75.1
2.8
4.0
3.1
5.3
3.2
3.3
2.5
3.7
2.6
4.8
3.3
2.9
3.0
59.7
35.8
2.5
3.2
4.9
3.3
2.8
4.0
18.7
6.4
28.1
44.6
43.8
4.7
5.5
3.7
3.6
18.3
8.7
3.0
36.4
5.3
23.0
29.9
61.3
20.4
3.6
9.1
3.5
4.9

éééémmmmwmwwmwmmzzzzzzzzmmmmmmmmmmmmmsssagsgmwweeeeonoog

 

 

 

 

 

18]

22222222222222

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
OstV
OstV
AceS
AceS
AceS
AceS
AceS
AceS
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
QueR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS

3.0
2.5
3.3
2.6
2.7
61.5
3.5
7.3
3.1
30.6
74.3
5.2
21.9
4.1
5.6
5.3
7.6
9.5
14.1
7.8
11.8
9.1
7.0
11.0
5.4
5.1
4.4
6.9
4.4
3.1
13.6
3.9
4.5
3.5
5.2
9.1
7.9
7.9
5.3
6.8
7.1
7.5
3.9
45.6
3.2
3.5
48.1
5.4
6.7
2.6
3.6
5.2
4.7
2.5
2.7
3.0

nmmmmmoommmmmmomommmmmmmmmoowm05mmmommmmmmmmmomwmmmmmmmm
ZZZzmmmmmmmgggggmmmmmmmmmmmmmzzzzzzzmmmmmmgggggggggggﬁﬁg

 

 

 

 

22232222322322232322

AceS
OstV
TilA
TilA
AceS
AceS
AceS
AceS
AceS
OstV
OstV
OstV
OstV
OstV
OstV
OstV
OstV

QueR

QueR
QueR

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
OstV
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceN
FagG
AceS
AceS
AceS
AceS

33
3.3
47.7
3.5
2.5
4.1
2.9
3.1
3.0
2.5
2.9
11.5
2.8
7.0
12.8
4.1
5.6
4.6
7.4
4.3
9.5
8.4
2.7
2.5
14.7
9.1
8.1
10.5
7.3
12.4
11.8
6.3
17.2
4.3
6.0
12.2
5.8
8.8
3.4
8.4
64.6
5.7
5.5
3.8
12.8
14.3
6.3
6.5
4.9
5.6
86.0
13.3
7.8
2.9
8.7
12.3

mmmmomommmmmmmmmmmmmmmmmmmmommmmmmommmommmwmmmmmmwmmommm

ZZZmemmmmmmmmmmmmmmggggggggggggémmmmmmmmmmmmmmmmmmmzzzz

 

 

 

 

f2 AceS 2.5 s N
B AceS 2.5 s N
f2 AceS 3.2 s N
f2 AceS 10.2 s N .
f2 AceS 7.2 s N
12 AceS 3.6 s N
12 AceS 76.4 s N
12 A068 1 1.9 s S
12 AceS 2.5 c S
12 AceS 77.2 s S
12 AceS 2.9 s S
12 AceS 3.2 s S
12 AceS 36.0 s S
12 AceS 2.8 s S
12 AceS 4.7 s S
f2 AceS 21 .3 c W
12 AceS 8.5 s W
12 AceS 62.5 c W
12 AceS 3.9 s W
f2 AceS 2.8 s W
12 AceS 3.4 s W
12 AceS 4.9 s W
12 AceS 3.9 s W
12 AceS 2.5 s W
B AceS 3.4 s W
f2 A068 2.5 s W
12 AceS 4.2 s W
B AceS 20.9 s W
12 AceS 17.4 s W
B AceS 21.5 s c
B AceS 22.0 s e
B AceS 5.7 s c
B AceS 4.0 s c
B A068 31 .2 c e
B AceS 14.1 s c
B AceS 7.0 s n
B AceS 7.0 s n
B AceS 5.4 s n
B AceS 2.5 s n
B AceS 5. 1 s n
B AceS 7 .5 s n
B AceS 4.8 c n
B AceS 2.9 s n
B AceS 4.1 s s
B AceS 6.9 s s
B AceS 10.1 c s
B AceS 3.1 s s
B AceS 5.9 s s
B AceS 50.4 s s
B AceS 7.2 s s
B AceS 5.0 s s
B AceS 9.5 s w
B AceS 46.1 s w
B AceS 57.2 s w
B AceS 17.4 s w
B FagG 56.1 c w

 

 

 

 

182

f4 AceS 4.3 s E
f4 AceS 50.3 s E
f4 AceS 28.2 s E
f4 AceN 2.8 c E
f4 AceS 3.1 s E
f4 AceS 38.4 s E
f4 AceS l 1.3 s N
f4 AceS 2.5 s N
f4 AceS 5.2 s N
f4 AceS 3.5 s N
f4 AceS 7.4 s N
14 AceS 2.6 u N
f4 AceS 4.0 s N
f4 AceS 4.7 c N
f4 AceS 3.6 u N
f4 AceS 5. 1 s N
f4 AceS 2.9 s S
f4 AceS 2.8 s S
f4 AceS 3.4 s S
f4 AceS 8.4 s S
f4 AceS 3.2 s S
f4 AceS 6.4 s S
f4 A068 5.2 s S
f4 A068 5.8 s S
f4 AceS 9. 1 s S
f4 AceS 2.7 s S
f4 FagG 4.2 c S
f4 AceS 3.3 s W
14 AceS 3.3 s W
f4 AceS 6. 1 s W
f4 AceS 6.2 s W
f4 AceS 3.5 s W
f4 AceS 64.5 s W
f4 AceS 7.9 s W
f4 AceS 7.2 s W
f4 A068 3. 1 c W
f4 AceS 4.0 s W
f4 AceS 7.5 s W
15 AceS 10.8 11 6
f5 A068 8.6 s c
15 AceS 7. l u 6
f5 AceS 78.9 n 3
f5 AceS 60.4 n 6
f5 AceS 30.7 c e
15 AceS 42.4 11 e
5 AceS 2.5 s e
f5 AceS 5. 1 s e
15 AceS 3.0 s 6
f5 AceS 4. 1 s 6
f5 AceS 2.6 u 11
f5 AceS 4.0 u n
15 AceS 3.6 u 11
f5 AceS 4. 1 u 11
f5 AceS 4.4 s n _
f5 AceS 5. 1 s 11
f5 F339 3.3 c n

 

 

 

 

388333333333383383388838383888833838388338

3883

8338883388

F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
FagG
FagG
AceS
AceS
AceS
AceS
F agG
F agG
FagG
FagG
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

4.9
3.7
4.4
10.4
3.8
3.7
5.1
2.7
3.2
2.5
3.9
4.0
4.4
2.5
3.3
6.5
5.0
3.9
6.2
5.4
5.0
6.0
8.1
7.1
2.7
3.2
2.8
2.9
5.3
3.0
10.6
4.4
4.4
5.1
9.7
4.9
9.0
3.7
4.0
3.2
6.6
4.2
10.6
6.3
51.2
29.6
43.0
44.3
13.7
2.6
4.5
4.3
5.5
4.6
3.2
2.6

cccccccmcacwmamoocccscnmwoccms:1:11:01:mmwowzcccmmccwmmwmwccm

mmmwmmwmmmwbbbbonnaooooggggggggsgggggmwmmwwwmmmmmmmmmmmp

 

 

 

 

3131313{33833333388833333358585858333333321333121223333888888838883

FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
A068
A068
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
TilA
AceS
AceS

10.4
6.2
2.5
2.7
2.5
2.6
3.8
92.7
9.6
4.3
34.9
5.0
24.9
13.2
24.1
3.2
3.1
2.7
47.6
5.9
5.9
2.8
21.3
4.2
5.0
4.2
13.9
85.4
8.2
9.8
3.7
4.0
2.5
5.5
9.8
7.2
4.3
6.4
5.0
4.2
6.6
3.3
6.9
2.5
3.4
6.7
4.9
3.1
3.5
2.9
7.4
18.8
2.5
5.8
2.5
4.6

mmbbbbbﬂﬂggﬁgmmwmabbbbooooggsgmmwmmunponooasssggsgggsgram

 

 

 

 

183

f9 AceS 4.1 s s
19 QueR 4.4 c s
19 AceS 3.3 c w
B AceS 36.9 s w
f9 AceS 31.9 c w
19 AceS 1 1.5 c w
g1 AceS 2.7 c E
g1 AceS 3.8 s E
g1 FagG 3.0 s E
g1 FagG 4.3 s E
g] FagG 4.6 s E
g1 FagG 3.8 s E
g1 FagG 3.2 s E
g1 FagG 2.7 s E
gl FagG 2.9 s E
g] FagG 4.6 s E
gl AceS 3.9 c N
g1 AceS 4.4 c N
g1 AceS 3.4 s N
g1 AceS 2.8 s N
g1 AceS 3.8 s N
g1 AceS 3.3 s N
g1 AceS 2.9 s N
g1 AceS 2.9 s N
g1 AceS 2.9 s N
g1 AceS 2.6 s N
g1 AceS 4.0 s N
g1 AceS 2.8 s S
g1 AceS 24.9 s S
g1 A068 9.3 s S
g1 AceS 3.3 c S
g1 AceS 4.1 s S
g1 AceS 4.7 s S
g] AceS 2.5 s S
g1 AceS 2.9 s S
g1 AceS 2.7 s S
g1 AceS 2.5 s S
g] AceS 2.5 s S
g1 AceS 3.2 s S
g1 AceS 2.9 s S
g1 AceS 9.4 s W
g1 AceS 2.5 s W
g1 AceS 2.7 s W
g1 AceS 2.8 s W
g1 AceS 65.2 s W
g1 AceS 2.7 s W
g1 AceS 9.2 s W
g1 AceS 6.7 c W
g1 AceS 2.6 s W
g1 AceS 73.5 s W
g1 AceS 66.4 s W
g1 AceS 28.4 s W
g1 AceS 29.9 s W
g1 AceS 3.5 s W
g1 FraA 52.7 c W
_glO A068 2.5 s E

 

 

 

£10 ACC
1 £10 Ace
. glO ACC
1 gIO Fag‘
i 210 111:”
' £10 Ulm

210 Um

no r
O
n
(D

"Q "Q "Q "Q "Q "0 "SJ-4 .

oooooooooo
o

37

9.

 

1

1

1

1

I

1
:21 AceS
g1 Ost\
g1 AceS
g1 AceS
.310 AceS
:210 AceS
€10 AceS
€10 051V
€11 AceS
g” AceS
EH AceS
811 AceS
3H AceS
311 Ulm.‘
g1] Aces
g” FagG
g1] FagG
g1] OstV
g1] QUCR
g1] AceS
g1] AceS
g1] AceS
g1] AceS
g1] ACeS
g1] AceS
3” AceS.
g1] ACES.
g1] ACeS
g1] ACeS
g1] ACeS
g1] ACeS
£1] ‘ CeS
g1] ACCS
1311 05W
1:11 QUCR
€11 QueR
EH AceS
811 ‘
2“ ACeS
\ 1 .ACCS

['3
._J

4/1)
F)
O
{/7

 

 

 

glO AceS 33.8 s E
g10 AceS 3.2 s E
g10 AceS 2.7 s E
g10 FagG 2.7 s E
g10 TilA 2.5 s E
310 UlmR 1 1.9 s E
g10 UlmR 55.6 s E
g10 AceN 4.1 c N
g10 AceS 3.6 s N
g10 AceS 2.9 s N
g10 AceS 3.5 s N
g10 AceS 4.1 s N
g10 AceS 3.1 s N
g10 AceS 2.7 s N
g10 AceS 3.2 c N
g10 AceS 2.7 s N
g10 OstV 4.9 s N
g10 AceS 10.3 s W
g10 AceS 4.0 s W
g10 AceS 63.2 s W
g10 AceS 14.9 s W
g10 AceS 6.0 s W
g10 OstV 2.9 s W
gl 1 AceS 2.5 s E
gll AceS 3.4 c E
gl 1 AceS 3.2 s E
31 1 AceS 4.3 s E
gl 1 AceS 3.5 s E
gl 1 UlmA 3.4 s E
gl 1 AceS 2.5 s N
gl 1 FagG 17.1 s N
gl 1 FagG 6.4 c N
gl 1 OstV 12.8 s N
gl 1 QueR 2.5 c N
gl 1 AceS 5.6 s S
gl 1 AceS 13.3 s S
gl 1 A068 4.9 c S
gl 1 AceS l 1.3 s S
gl 1 AceS 2.5 s S
gl 1 AceS 4.6 s S
gl 1 AceS 7.1 u S
gl 1 AceS 36.5 s S
gl 1 AceS 3.5 s S
gl 1 AceS 9.4 s S
gl 1 AceS 45.7 s S
gl 1 A068 7.1 s S
gl 1 AceS 12.6 s S
gl 1 AceS 32.1 s S
gl 1 OstV 4.6 s S
gl 1 QueR 31.0 c S
gl 1 QueR 3.8 c S
gl 1 AceS 2.7 c W
gl 1 AceS 5.5 s W
gl 1 AceS 15.2 s W
gl 1 AceS 13.9 s W
_gl 1 AceS 6.1 s W

 

 

 

 

388%EﬁﬁﬁﬁﬁﬁﬁﬁﬂﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁR%%%%%ﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁﬁg

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

89.3
2.5
4.3
13.3
26.3
25.7
3.3
7.0
5.8
37.9
22.1
5.1
2.5
4.7
5.0
6.0
4.9
6.4
45.9
44.9
45.8
2.5
2.6
2.6
2.6
2.6
2.6
2.7
2.8

' 2.8
2.8
3.0
3.0
3.1
3.2
3.2
3.3
3.3
3.4
3.5
3.6
3.7
4.0
4.1
4.2
4.2
4.3
4.4
4.5
4.5
4.7
4.7
4.8
4.9
4.9
5.7

mmmmmmwmmmmmmmwmmmommmmmmmmmommmmmmmmmmmmmmmmmwmmommmmmo

gwwwmwuasnooooooﬁgggggggﬁggmmmmmmmmzzzzzzzzzzzmmmmmmmmmg

 

ﬁﬁﬂﬁﬁﬁ

 

 

 

184

A068
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

5.7
5.8
6.0
6.4
6.4
6.5
6.7
6.7
9.5
14.2
19.1
48.6
49.5
16.4
21.1
23.6
32.4
2.6
3.3
68.0
37.8
7.7
2.8
2.6
4.0
3.6
3.1
4.5
3.5
61.3
40.4
3.1
2.8
5.8
4.2
3.4
6.3
4.0
5.6
7.4
2.9
3.9
10.6
8.1
2.6
2.7
7.2
3.6
3.3
4.1
2.7
3.9
4.9
2.6
4.2
7.1

mccmcgmmmmccmmcmmmmwmmmmmmmommmowmccmmmmmOOmwmmamwommcnmw

annoooaooooggggﬁgs5gmmwmmmmmmmmpnunaunanaooooonoooggggag

 

 

 

 

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

5.7
4.5
5. 7
57.9
4.8
9.2
9.8
2.9
3.1
2.9
3.4
3.8
3.2
3. 0
29. 2
3. 2
31.8
52.2
16.4
5.4
3.4
6.0
5.9
2.8
5.4
2.5
2.7
4.2
5.1
3. 7
11.1
73.0
5. 4
12. 0
3. 6
12. 7
3. 7
65.5
3.5
4. 3
53.6
25.0
19.4
5. 5
24. 2
2. 7
22. 4
9.3
4. 3
11.1
64. 8
48. 4
58.4
16.3
15.7
9.1

=1:cscccmocczcwwcwmwocccccmccwwm===:=.wm==owo==anamoccmmcc

bnbaannnanonaosssssssassesssssssssswwwwwwwmwma==========

 

 

 

g6 AceS 3.1 u n
g6 AceS 11. 5 s n
g6 FagG 12.7 c n
g6 FagG 3.9 u n
g6 FagG 5.9 u n
g6 FagG 4.4 c n
g6 FagG 3.4 u n
g6 FagG 2.6 u n
g6 A068 3.5 u s
g6 AceS 2.7 u s
g6 A'ceS 4.2 c s
g6 AceS 2.7 s s
g6 AceS 3.0 s s
g6 AceS 2.7 u w
g6 AceS 7. 6 u . w
g6 AceS 12. 7 c w
g6 AceS 12.0 s w
g6 A068 9.9 s w
g6 FagG 7.0 c w
g6 FagG 64.3 11 w
g7 AceS 24.5 11 c
g7 AceS 12.5 s c
g7 AceS 17.6 s e
g7 AceS 22.6 c e
g7 AceS 5. 3 u c
g7 FagG 12. 4 s e
g7 AceS 10.9 n n
g7 AceS 7.5 s n
g7 FagG 8. 2 s n
g7 AceS 37.0 s s
g7 AceS 2. 7 s s
g7 AceS 10. 6 s s
g7 AceS 3.2 s s
g7 AceS 3.3 u s
g7 FagG 2.5 s s
g7 AceS 8.7 c w
g7 AceS 4.2 s w
g7 AceS 7.5 s w
g7 AceS 3.7 s w
g7 FagG 8. 9 c w
g8 AceS 23. 3 s e
g8 A068 7.3 s e
g8 AceS 6. 3 s e
g8 A068 13. O s e
g8 FagG 4.9 s e
g8 AceS 2.8 s n
g8 FagG 5.2 u n
g8 FagG 2.7 s n
g8 FagG 5. 2 c n
g8 FagG 38. 8 u n
g8 FagG 2.9 s n
g8 FagG 4.6 s n
g8 FagG 5.0 s n
g8 AceS 5. 3 s s
g8 AceS 17. 9 s s
18 AceS 3.8 s s

 

 

 

 

185

I82:°12°8°8°8°8°$°®°®°8°~8°®°®°®°®°®°®°~8$°®°~8§$°®°~8°8“%°%%°®°8°%°%°®°~2>

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG

QueR
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG

FagG
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
OstV
QueR
QueR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS

8.6
5. o
24.7
14. 1
4. 5
10.1
13.0
7.1
4.2
8. 7
16. 7
4. 3
12. 4
4.0
9.0
5.0
4. 3
15. 7
7.3
4.4
4.9
4. 3
86. 8
2.9
4.8
2.6
2.5
3.4
3.7
2.9
2. 5
13. 3
2.9
4.3
3.0
3.5
3. 5
12. 5
4.9
3.4
3.3
2.5
2. 5
19.4
3.5
3. 4
18. 7
2. 5
13. 7
9. 2
14.1
12.7
18.2
3.0
3.5
3.7

mmmm0mmmmmmmnommmomma:wommmmmwmmmommmmmommmmmmmmomwmmmmmm

mmwmanbubnnaooooooooooooooaoooanonoggggggggggggmmmm'mwmmm

 

 

 

 

 

EE%%%%%%Q%%%%%

hl

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
QueR
QueR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG

FagG
AceS
AceS
AceS
AceS
AceS
A068
A068
AceS

3.1
2.9
2.7
3.7

1 1.1
7.3
6.6
10.8
29.5
25.0
2.7
36.9
71.5
24.1
3.1
3.9
3.6
4.3
3.1
4.6
2.5
4.1
4.1
3.8
3.3
2.5
2.6
2.5
93.1
63.6
6.9
6.0
9.3
79.8
14.6
14.3
83.3
2.8
20.0
54.9
8.5
4.7
4.2
40.5
46.8
26.5
5.0
23.3
5.3
20.8
5.3
38.9
8.4
3.0
20.8

mmmmmmomommommm:mmommmmommmmmmmommmomwwwmmmmnowmmmmmmwmm

mmmmmmmmggggggsgsmmmmmmmmmmmmzzzzzzmmmmmmmmmssagsgwwwwwm

 

 

h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
h10
hll
hll
hll
hll
hll
hll
h11
hll
hll
hll
hll
hll
hll
hll
hll
hll
hll
hll
hll
hll
hll
hll
hll

AceS
QueR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
CarG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
TilA

AceS
AceS
AceS
AceS
AceS
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS

2.5
5.1
3.9
3.7
2.6
2.9
22.1
62.7
20.0
2.9
2.5
3.3
2.9
3.9
40.9
33.6
5.0
3.8
5.0
5.4
4.8
4.5
20.1
3.5
4.7
5.6
5.3
3.5
3.0
2.5
42.9
2.8
3.8
4.0
3.0
5.8
3.1
5.8
5.0
3.4
3.1
4.1
2.7
2.9
2.9
2.6
3.3
2.6
2.8
3.0
2.7
3.0
3.0
3.8
86.6
8.6

mocorn0mmmmmmmmmmmwmmmmmmmtnmmmommmmmmmmmmmmmmmmommommomom

zmmmmmmmmrun-1mmmmmmmmmmmmggggggmmwmmmmwwmmmmmmmmmzzzzzz2mm

 

 

 

EEEEEBSEESEBEEEEEEEEBEEEBBEB

5.0
4.6
3.0
38.9
7.1
7.7
4.8
6.8
5.5
5.7
3.9
5.5
3.0
5.4
4.2
3.2
17.7
2.5
2.6
5.4
9.5
3.7
7.1
4.3
4.1
4.7
8.2
3.1
6.5
4.2
4.9
4.3

- 9.3

3.5
61.3
3.0
7.5
7.1
7.7
71.0
3.0
61.3
73.5
54.8
7.0
60.2
13.7
2.5
16.3
12.3
23.8
4.1
4.4
4.5
5.5
3.8

mmmmmmmoommmmmmmmmmwmmmmwmmm«amt/3mg:mmmmmmommmmmmmmmmmmmmmm

mmmzzzzzzmmmmmmmmmmmmmmmmmmmggggggmwwwmmwmwwwwwwwzz22222

 

 

 

1%

 

 

 

EEEEEEEEEIEEEEEESEEEBEEBEEEBEBEEEEBEBBEBEE65383553653555

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
F agG
FagG
AceS

AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG

FagG
AceS

AceS
AceS
FagG
FagG
AceS
AceS
AceS

56.6
5.2
3.1
7.5
2.6
6.5
4. 4
23. 3
7.5
3.1
2.9
4.0
5.3
3.3
8.8
2.8
5.7
7. 6
24. 2
2.8
3.0
2. 9
18. 5
3.4
3. 2
51.3
2.6
4.6
4.3
3.0
6. 7
78. 3
14. 7
6. 6
12. 0
14.9
6.0
3.1
6.5
3.9
5.4
9. 8
14. 9
14. 8
9. 9
51.2
10.6
7.2
5.3
7.1
6.7
8. 0
l6. 7
4.5
8.1
5.5

mmmmcmmmmommmmmmmoommmommmmmmmmmmmmmwmmmwmmmmmmmmmmmmmmm

anbnaonoooogggsggmmwmmbnanﬂbbba50000000000000aﬁgégggmmmm

 

 

 

85533535353333533335333EEEEEEEEEEEEEEEEEEEEEEEE

AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
F agG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
TilA

AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
TilA

AceS
AceS

11.5
19.0
14. 2
2. 8
18. 8
3.4
3.2
6.9
3.4
4.3
4.1
13. 4
6. 6
15.3
6. 8
11. O
10. 0
64.1
35. 4
4. 9
45. 2
3. 7
40.5
79.0
9. 4
12. 3
11.4
16.3
2.5
4.4
3.4
4.1
2.5
2.5
2. 5
10. 3
3.2
5.1
7.5
3.3
3.0
3.0
3.0
4.1
5.1
3.5
5. 3
19.1
2. 5
21. 3
12.1
15.6
3.1
2.6
2.6
9.2

accommmwmmcccocccar/2;:m:canwcccccocmommmcmmmmommwmm:magnum

 

 

 

h6
h6
h6
h6
h6
h6
h6
h6
h6
h6
h6
h6
h6
h6
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h7
h8
h8
h8
h8
h8
h8
h8
h8
h8
h8
h8
h8
h8
h8
h8

AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

FagG

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

2. 9
73.0
56.2
2.9
4. 5
14.5
2.9
3.0
4.8
2. 9
25. 7
54.9
2.6
3.4
5. 3
13. 1
52.7
43.4
4.7
9. 3
68. o
25.5
11.4
18.0
12.9
12. 6
2. 6
10. 8
22.1
3.5_
3.3
3.9
4.7
7. 2
82.1
4.5
6.1
8.8
6.0
7. 5
23.2
3.8
8.6
2.6
4.6
9.0
3.1
2.7
2.7
5.9
9. 9
6o 0
4.9
4. 8
11. 1
8.3

mmommmmwmmmmmmmmcn:mane:cccgmmmmmmmgmcmmgcmmgcmmommomgccgm

abbbnbaoooaooangﬁggagggggmmmwmmwnbbonnnwnoggggmwwmabﬂbbb

 

 

 

187

 

 

 

EﬁgggggggggggEEEEEEE%EEEEEEEE%ESEEEEE

FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
AceS
AceS
AceS
AceS

11.6
5.0
7.5
2.9
4.8
46.7
56.8
2.9
7.0
2.5
2.5
3.2
3.5
5.0
4.5
12.8
3.8
50.0
42.6
8.4
7.8
4.7
4.1
6.4
2.5
4.6
5.0
4.4
4.4
5.8
6.4
23.7
4.5
22.7
5.6
5.3
8.2
8.6
74.1
28.7
10.8
4.8
46.2
4.8
4.3
5.2
2.5
4.5
3.2
3.6
10.8
3.9
4.0
18.1
3.5
3.0

mmmmmmmmmmmmmmmwmmmmmmmmmmomwmmmmmmmmmmmmwwmmmmmmwmmwmmo

3553000000000000000000000000000000000€£€€ségmmmwmwmmmmmb

 

'56EEEEEEEE$EEEEEEEEEEEEEE

 

 

AceS
AceS
AceS
AceS
AceS ‘
AceS
QueR
QueR
QueR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

3.5
2.5
3.0
3.9
2.5
54.1
2.9
3.1
2.9
2.6
7.8
2.6
3.3
14.8
2.5
4.7
60.5
7.4
2.9
6.0
6.0
3.8
9.0
7.8
10.5
11.6
8.9
4.3
7.9
42.0
9.5
4.4
7.0
7.0
54.6
2.8
53.8
10.5
82.2
9.4
5.3
6.7
38.9
55.1
4.3
6.3
4.4
6.8
7.5
5.7
4.1
5.4
20.5
6.2
49.5
3.4

mmmmowmmmmmmmcoomcmmmmcmmmmmommowmmmmmmmmwowmmoooommmmmm

gmmwwmmwzzzzzzzzzzzzmmmmmmmmmmmssass258244842ssnbbaauban

 

 

 

 

 

188

CSCJCJCJCSCJ'CJCSCJQ

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
AceS
FagG
AceS
AceS

34.2
3.0
5.9
2.9
2.9
46.7
29.8
5.6
4.0
3.1
3.6
22.3
3.6
71.5
2.6
4.1
13.6
25.9
3.2
2.6
3.7
5.9
6.3
4.3
3.0
46.9
4.7
3.8
6.2
5.1
2.5
3.4
5.8
4.5
4.9
3.5
5.0
6.1
2.8
32.0
7.8
3.2
8.0
63.6
8.2
7.4
51.7
55.2
4.4
5.0
5.9
4.7
6.0
4.4
6.1
5.8

mmmmmmmcmommowmmmmmmmmmommmmcoommmmmmmmmmmmmmgmmommmnmmm
w”D=°==99seeseeaaseeewmwwwwwwwwzzzzzzzzzzmmmmmmmm2224242

 

 

 

 

'CSCJ'CS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS
PruS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
UlmA
AceS
AceS
AceS
AceS
AceS
A068

7.3
7.3
3.7
4.1
3.5
4.3
6. 9
48. 6
3. 2
61.4
43.7
5.5
5.4
2.7
2. 7
11.9
78.9
7.9
8. 5
58. 2
9.1
8. 8
56. 6
8.5
5.9
2.6
7. 6
15.3
7. 6
17. 9
2.5
3.2
3. 5
53.5
3.2
3.1
3.0
7.3
6. 5
70. 2
2.7
4.4
6. 8
22.2
3. 4
18. 8
4.3
3.8
3.8
2.6
73.5
3.9
5.4
3.1
8.3
2.5

cccccnccccccocmmmowmmomcmowmmmmmocmmwmammmmmmommmommmmmm

 

 

 

anannnnognooaasﬁgsgmmmnnbanoooooooooooaaooggggggggggsggg

i5
15
15
15
15
i5
15
15
15
15
15
15
15
15
15
15
15
16
16
16
16
16
16
16
16
16
i6
16
i6
16
16
16
16
16
16
16
16
16
16
i6
i6
16
i6
i6
i7
17
i7
17
17
17
17
17
i7
17
17
i7

AceS
AceS
AceS
FagG
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

2.5
2.8
2.5
2.5
3.0
6.9
9.5
5. 9
25.8
64.9
8.3
2.5
2.8
6. 9
11.2
2.5
2.6
5. 4
61. 5
7.9
2.7
7.5
2. 5
43.5
3.9
3.0
4. 7
23.5
23.1
5.2
8. 7
11.6
3.8
3. 0
46. 3
10.0
4.6
7.7
5. 7
27.1
4.4
3. 8
17. 2
35.7
4.4
3.6
3.7
7. 5
36. 4
3.7
9.8
2. 9
54. 6
9.3
3.0
2.9

mcccmomccmmmcocmc:mm:ovacant:gmccmccccocccomgcmcmm:owmocwcc

wmwmanannoaog5ggmmwcnmwmagapauaaaoooooaogsgggsmmmmmﬂbﬂbbﬂ

 

 

 

 

 

189

17
17
17
17
17
i7
i7
17
18
18
18
i8
18
18
18
18
18
18
18
18
i8
18
18
18
18
18
18
18
18
i8
18
18
18
19
i9
19
19
19
19
19
19
19
i9
19
19
i9
i9
19
19
i9
19
19
19
19
19
19

AceS
FagG
AceS
AceS
AceS
FagG
FagG
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

70.8
41.2
42.2
2. 7
38.3
4.9
4.2
5.5
3.3
6.6
2. 6
49.4
4. 5
40.6
7.2
4.6
4.1
4.4
2.9
4.9
2.7
7.8
3.8
5.2
4.4
5.3
4.1
3.9
3. 8
16. 7
4. 7
20. 5
4.1
3.1
53.1
3.8
7.8
4.6
3.3
2.6
6.0
3. 2
58.8
3.6
8.1
7.1
6.2
3. 6
17. 5
34.6
4.6
3.5
2.5
8.2
3.1
4.5

mmenu:mmmcmmommmmwmenu:mmmomvaommmcomommmmmwmwmmmmmnoomocmcom

m:pannnpanuooooooonaoaoggggsgmmmmmmmmbbnoofbnﬂoooosﬁgaﬁgww

 

 

 

 

19 AceS 8.9 s s
19 AceS 7.1 s s
19 AceS 3.2 s s
19 AceS 2.6 s s
19 AceS 2.6 s s
19 AceS 3.3 s s
19 AceS 4.8 s s
19 AceS 3.6 s . s-
19 AceS 3. 1 s s
19 AceS 3.0 s s
19 UlmA 3.3 s s
19 UlmR 4.2 c w
19 A668 2.9 s w
19 AceS 3.8 s w
19 A068 2.7 s w
19 AceS 3.7 s w
19 AceS 5.9 s w
19 AceS 2.8 u w
19 AceS 4.7 s w
19 A068 3.5 s w
19 AceS 8.4 s w
19 AceS 24.8 s w
19 AceS 29.2 s w
j l AceS 13 .6 s E
j l AceS 55.7 s E
j 1 AceS 34.6 s E
j 1 AceS 20.5 s E
j 1 FagG 5.4 s E
j 1 AceS 5.1 s N
j 1 AceS 3 .4 s N
j l AceS 4.9 s N
j 1 AceS 5.7 s N
j 1 AceS 4.5 s N
j 1 AceS 3. 1 c N
j 1 AceS 9. 1 u N
j 1 FagG 5.1 s N
j 1 TilA 60.5 c N
j 1 AceS 5.0 s S
j l AceS 8. 1 s S
j 1 AceS 5.0 s S
j 1 A068 5.7 s S
j 1 AceS 8.9 c S
j 1 AceS 4.4 s S
j l AceS 9.6 s S
j l FagG 15.8 s S
j l AceS 4.3 s W
j 1 AceS 6.7 s W
j 1 AceS 5.4 c W
j 1 AceS 45.0 s W
j l A068 7. 1 s W
j 1 AceS 4.3 s W
j 1 FagG 14.6 s W
j 1 FagG 3.1 s W
j 1 FagG 19.5 s W
j 1 OstV 4.0 s W
j2 AceS 7 .9 c E

 

 

 

j 2 A065 3.9 s E
j2 AceS 7.7 s E
j2 A068 3.5 c E
j2 AceS 6.3 c E
j 2 FagG 10.1 c E
j2 AceS 16.2 s N
j2 AceS 7.0 s N
j2 AceS 7.6 s N
j2 AceS 10.6 s N
j2 AceS 19.0 s N
j2 AceS 3.1 s N
j2 AceS 2.7 s N
j2 FagG 3.3 c N
j2 FagG 5.3 s N
j 2 FagG 6.3 s N
j2 AceS 4.0 s S
j2 AceS 3.3 s S
j 2 AceS 13.1 s S
j2 FagG 13.2 s S
j2 FagG 8.0 s S
j2 FagG 4.5 c S
j2 F agG 2.7 s S
j2 AceS 1 1.9 s W
j2 AceS 9.1 s W
j2 A068 3.4 s W
j2 FagG 2.7 s W
j2 AceS 3.1 s W
j2 A068 3.9 s W
j2 F agG 9.1 s W
j2 PruS 3.5 n W
j3 AceS 6.0 s c
j3 AceS 6.3 s e
j3 AceS 2.7 c e
j3 AceS 15.3 s c
j 3 AceS 10.8 s n
j3 AceS 2.9 s n
j3 AceS 2.7 s n
j3 AceS 28.0 s n
j3 AceS 5.8 s n
j3 FagG 3.0 s n
j3 FagG 3.0 s n
j3 FagG 2.5 c n
j3 FagG 2.6 s n
j3 FagG 2.6 s n
j3 FagG 4.6 s n
j3 AceS 23.1 s s
j3 AceS 4.9 c s
j3 AceS 6.5 s s
j3 AceS 44.0 s s
j3 AceS 4.7 c s
j3 AceS 2.8 s w
j3 AceS 73.8 n w
j3 FagG 5.1 s w
j3 FagG 4.9 s w
j3 F agG 6.4 c w
j 3 FagG 4.1 u w

 

 

 

 

 

190

j 3 FagG 7. 1 s w
j3 FagG 2.5 s w
j4 A068 5.3 s c
j4 AceS 73.8 s e
j4 AceS 3.7 c e
j4 AceS 8.4 s c
j4 AceS ' 9.4 s e
j4 AceS 10.9 n e
j4 AceS 20.5 s e
j4 AceS 23.8 s e
j4 AceS 35.7 s e
j4 AceS 57.9 s e
j4 AceS 58.8 s n
j4 AceS 21.8 c n
j4 AceS 66.4 s n
j4 AceS 38.7 s n
j4 AceS 10.9 s n
j4 PruS 18.2 n n
j4 AceS 8.8 s s
j4 AceS 12.3 s s
j4 AceS 3.3 s s
j4 AceS 28.9 s s
j4 AceS 6.9 s s
j4 AceS 64. 1 s s
j4 AceS 9.4 s s
j4 AceS 22.7 s s
j4 AceS 9.9 s w
j4 AceS 16.7 s w
j4 AceS 39.4 s w
j4 AceS 2.5 s w
j4 AceS f 4.7 s w
j4 AceS 4. 1 s w
34 AceS 3.9 s w
j4 AceS 64.2 c w
j4 AceS 4.3 s w
j4 AceS 4.7 s w
j4 AceS 3.4 s w
j4 AceS 4.7 s w
j5 AceS 3.0 c e
j5 AceS 4.2 u e
j 5 AceS 2.5 u e
j5 F agG 3.0 c e
j5 FagG 5.5 s e
j5 AceS 3.6 c n
j5 AceS 2.9 u n
j5 AceS 3 .8 u n
j5 AceS 2.7 u n
j5 A088 4.9 u n
j 5 AceS 2.5 u n
j5 AceS 2.7 u n
j5 AceS 4.0 u n
j 5 AceS 2.8 u n
j 5 AceS 14.7 n n
j5 AceS 6.4 u 11
3'5 AceS 6.3 u n
j5 Fagg 4.3 c n

 

 

 

 

j5 AceS 3.2 c s
j5 AceS 49.3 11 s
j5 AceS 3.1 u s
j5 AceS 4.0 u s
j5 AceS 3.9 u w
j5 AceS 3.6 u w
j5 AceS 32.7 s w
j5 A068 2.7 u w-
j5 AceS 2.6 c w
j5 A068 3.2 u w
j5 AceS 2.9 s w
j6 AceS 5.5 c c
j6 AceS 2.9 u e
j6 AceS 4.1 u e
j6 AceS 6.2 u e
j6 AceS 6.7 c c
j6 AceS 5.7 c e
j6 AceS 12.8 n n
j6 AceS 4.4 s n
j6 A068 5.3 s n
j6 FagG 4.8 c n
j6 AceS 9.1 c s
j6 AceS 27.3 11 s
j6 AceS 60.3 11 s
j6 AceS 2.5 c s
j6 FagG 9.5 u s
j6 FagG 76.1 11 s
j6 FagG 61.1 1.1 s
j6 AceS 61.2 n w
j6 AceS 3.1 u w
j6 AceS 2.6 c w
j6 AceS 3.5 s w
j6 AceS 4.8 u w
j6 AceS 4.1 s w
j6 A068 3.0 u w
j6 FagG 3.1 u w
j6 QueR 5.1 c w
j7 AceS 6.1 u c
j7 AceS 3.0 s c
j7 AceS 18.9 s e
j7 AceS 66.5 c e
j7 AceS 5.2 u e
j7 AceS 3.4 u n
j7 AceS 2.5 c n
j7 AceS 8.2 s n
j7 AceS 4.2 u n
j7 FagG 6.5 u n
j7 FagG 2.9 u n
j 7 FagG 13.1 11 n
j7 FagG 2.5 s n
j7 AceS 3.7 s s
j7 AceS 2.9 u s
j7 AceS 91.4 11 s
j7 AceS 4.3 s s
j7 AceS 3.5 c w
j7 AceS 4.4 c w

 

 

 

j8 UlmR 4.6 s c
j8 AceS 4.7 u e
j8 AceS 5.4 c c
j8 AceS 4.9 s e
j8 AceS 28.4 s e
j8 A038 2.7 s c
j8 AceS 4.5 s e
j8 OstV 2.5 s c
j8 AceS 3.5 s n
j8 A068 6.5 c n
j8 AceS 4.3 s n
j8 AceS 2.5 s n
j8 OstV 5.4 s n
j8 AceS 3.7 u s
j8 AceS 2.8 c s
j8 AceS 44.7 s s
j8 AceS 6.0 u s
j8 AceS 2.9 s s
j8 AceS 3.7 s s
j8 FagG 10.8 s s
j8 OstV 1 1.1 s s
j8 AceS 16.1 s w
j8 AceS 16.3 s w
j8 AceS 19.0 c w
j8 AceS 21.6 s w
j8 AceS 24.3 s w
j8 AceS 28.8 s w
j8 AceS 2.9 c w
j8 AceS 3.1 s w
j8 AceS 3.2 s w
j8 AceS 3.2 s w
j8 OstV 3.3 s w
j9 A068 3.3 c e
j9 AceS 3.4 s e
j9 AceS 3.5 s c
j9 AceS 3.6 s e
j9 AceS 4.6 s e
j9 PruS 72.3 11 c
j9 PruS 3.9 u e
j9 PruS 2.5 s e
j9 PruS 3.3 s c
j9 UlmR 3.5 s e
j9 AceS 5.7 s n
j9 AceS 6.0 c n
j9 AceS 4.5 s n
j9 AceS 2.8 s n
j9 AceS 5.3 c n
j9 AceS 3.0 c n
j9 AceS 5.2 s n
j9 AceS 3.3 s n
j9 AceS 2.9 s s
j9 AceS 3.2 s s
j9 AceS 5.0 s s
j9 AceS 3.2 s s
j9 AceS 2.5 s s
19 AceS 6.5 s s

 

 

 

 

 

191

65555651356561.3515;

AceS
AceS
AceS
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
QueR
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
F agG
FagG
FagG
FagG
FagG
AceS
AceS
AceS
AceS

10.8
4.0
3.5
4.0
3.9
5.9
3.3
5.9
6.5
5.0
3.0
6.1
5.1
3.5
3.1
3.3
19.1
5.6
74.4
48.2
6.4
3.1
12.0
20.1
4.2
21.7
7.0
27.7
37.5
19.8
18.4
9.7
24.4
7.8
46.2
9.6
24.1
16.2
21.8
3.5
23.1
12.4
11.9
8.8
9.5
3.9
6.5
3.3
5.4
9.0
3.4
4.6
2.8
35.6
2.8
2.5

moommmmmommmmmommom000mmmammmmmommmmmmomoommwmmmmmmommmm

mwwmzzzzzmmmmmmggagsegggwwmwmwzzzzzmmmmmmsassess:2299999

 

 

 

 

 

 

 

 

k2 A068 5.1 s W
k2 AceS 4.7 s W
k2 AceS 4.2 s W
k2 AceS 3.2 s W
k3 AceS 19.6 s c
k3 AceS 4.5 s 1:
k3 AceS 3.0 s 6
k3 AceS 3 .0 s 6
k3 AceS 3.5 c 6
k3 AceS 18.9 s 11
k3 AceS 2.9 s 11
k3 AceS 4.4 s 11
k3 F agG 3.5 c 11
k3 FagG 22.8 s 11
k3 FagG 3.0 s 11
k3 F agG 4. l s n
k3 A668 5.5 s s
k3 AceS 9.8 s 8
k3 AceS 7.6 s s
k3 AceS 3.2 s s
k3 A068 2.6 s s
k3 AceS 2.6 c 8
k3 FagG 4. l s 8
k3 AceS 4.2 s w
k3 A068 5.9 s w
k3 FagG 3.7 s w
k4 A068 6.0 s 6
k4 AceS 3.6 s 6
k4 AceS 7.2 s 6
k4 AceS 31 .9 s e
k4 AceS 5. 1 s 6
k4 AceS 5.3 s 6
k4 FagG 4. 1 s e
k4 F agG 5.7 s 6
k4 F agG 4.2 u 6
k4 FagG 7.2 s e
k4 A068 2.5 u 11
k4 AceS 6.8 s 11
k4 AceS 4.6 s 11
k4 AceS 3.0 s n
k4 A068 2.8 c 11
k4 AceS 3.5 s 11
k4 AceS 4.9 s 11
k4 FagG 4.0 c 11
k4 AceS 2.8 s 8
k4 AceS 3.0 s 8
k4 AceS 4.0 s s
k4 AceS 1 1 .9 s 8
k4 AceS 18.5 s w
k4 AceS 16.0 s w
k4 AceS 12. 1 s w
k4 AceS 53.4 s w
k4 AceS 1 12.8 s w
k4 AceS 5.9 s w
k4 AceS 1 1.0 s w
k4 Egg 12.7 s w

EGGGGGGGEGGGGGEGGGGGGGGK

FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
F agG
FagG
AceS
AceS
AceS
AceS
F agG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
F agG
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS

5.0
4.9.
4.4
6.3
5.6
3.0
4.8
3.3
4.7
3.4
2.6
3.0
18.5
2.8
4.2
6.6
4.4
6.8
3.4
3.0
3.9
5.5
3.0
2.7
5.3
6.0
4.6
5.5
2.6
2.9
3.9
3.1
2.9
2.7
7.2
2.6
2.9
2.7
2.6
3.7
33.0
11.2
62.8
3.9
57.5
6.4
55.2
4.5
71.1
3.0
5.4
39.8
4.9
12.4
4.0
3.8

mmmmomcommmmmmmmwooommmommmwommnocmowmwonmcomm:omccmccg:

gstow?»DannaoaoogggégmmmzzzzzzmmmmggammmmpagagpnnnooaaoaaS

 

 

 

 

192

OstV

AceS

AceS

2.7
5.7
5.1
2.5
2.8
2.6
3.5
4.8
4.0
5.4
3.1
4.5
2.6
3.5
4.3
3.6

4.9
3.3
4.3
4.5
3.2
4.0
5.3
3.6
2.7
3.2
16.6
41.0
4.2
2.7
55.8
7.0
3.4
4.0
4.7
6.2
2.7
6.7
7.0
5.0
9.0
7.2
3.0
39.7
2.7
2.5
6.2
9.8
3.0
2.7
3.1
2.6
2.5
9.1
5.9

wwwmoommmmocmmmmmmmmmmmmm:mmmmooommmmmmoommmommommmccwoo

mmmmmmmmggggagsggwwmwmmwwmmnuanabooaoggggggwmmmuuaooooag

 

 

 

 

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

5.7
7.0
13.6
7.7
6.0
8.1
10.2
3.2
17.5
35.8
3.2
60.8
45.7
54.1
43.2
3.4
4.3
6.6
36.6
23.8
7.6
5.1
2.7
4.2
5.9
3.1
3.0
4.3
6.0
3.7
3.0
4.3
3.2
3.1
3.1
6.7
2.8
3.1
3.3
3.3
5.2
6.6
3.4
26.4
4.9
3.4
63.4
4.2
6.3
37.5
18.8
2.8
2.5
3.2
4.3
2.8

mommmmmmmomommmmmmommmmmm:w0mmmmoomomoomwmmomwmmoommmomm

but::35:31:10ooogsgaggmmmzzzzzmmmmmgggggggmmmmmmmmmmmzzzzzzzz

 

 

 

 

13 AceS 4.6 s n
13 AceS 5.7 u n
13 AceS 3 .5 s n
13 AceS 2.7 s n
13 AceS 3.6 s s
13 AceS 2.5 s s
13 AceS 3 .0 s s
13 AceS 2.9 s s
13 AceS 5. l s s
13 AceS 2.6 s s
13 AceS 3. 1 s s
13 AceS 3.5 u s
13 AceS 57.2 11 w
13 AceS 2.8 s w
13 AceS 3.2 s w
13 A068 2.7 s w
13 AceS 2.6 s w
13 AceS 2.5 s w
13 AceS 4.8 s w
13 AceS 3.6 s w
13 A068 3.2 s w
14 AceS 3.5 s e
14 AceS 5.0 s c
14 AceS 4.2 s e
14 AceS 39.4 s e
14 AceS 15.8 s e
14 AceS 14.0 s e
14 AceS 14.8 s e
14 A068 53 .2 s e
14 AceS 15.5 s c
14 AceS 16.4 s n
14 AceS 5. 1 s n
14 AceS 4.7 s n
14 FagG 4.6 s n
14 F agG 3.3 c n
14 AceS 5.5 s s
14 AceS 5.0 s s
14 AceS 44.4 s s
14 A068 4.8 s s
14 AceS 15.7 s s
14 AceS 1 1.1 s s
14 AceS 9.8 s s
14 FagG 82.0 s s
14 AceS 3. l u w
14 AceS 22. 1 s w
14 A068 2.9 s w
14 AceS 2.8 s w
14 AceS 2.5 s w
14 A068 2.6 s w
14 FagG 2.8 s w
14 FagG 3 .6 c w
14 FagG 4.5 s w
15 AceS 13 .6 s e
15 A068 19.9 c E
15 FagG 26.3 s E
15 FagG 20. 1 s E

 

 

 

 

193

15
15
15
15
15
15
15
15
15
15
15
15
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
17
17
17
17
17
17
17
17
17
17
17
17
18
18
18
18
18
18
18
18
18
18
18
18

FagG
AceS
AceS
A068
A068
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

FagG

AceS
PruS
AceS
AceS
AceS
AceS
CarC
CarC
OstV
OstV
AceS

2.5
3.0
5.0
3.1
2.7
4.0
11.0
17.3
2.6
2.7
6.4
4.3
48.6
6.6
3.7
4.0
2.5
3.9
5.3
76.9
5.0
50.9
51.9
3.5
11.6
49.9
5.3
3.2
2.8
41.2
3.2
17.3
3.5
4.5
4.9
3.0
3.9
58.2
5.4
4.2
3.7
25.8
2.9
13.8
7.5
4.0
7.0
7.5
8.5
8.0
8.2
8.8
2.7
3.1
3.1
2.7

mmmmmonooa0:0:mec:omowoomchoc:mmcoocomcoocmccgccmanommmonmmm

m=:_.nb:s=:=ooog gmam:13:15:00ggggmmmmmzzzzzzzmmmmggmmmmmzzzzm

 

 

 

 

18
18
l8
18
18
18
18
18
18
18
18
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
l9
l9
l9
l9
19
19
19
l9
19
19
m1
m1
m1
m1
m1
m1
m1
m1
m1
m1
m1
m1
ml
ml
m1
m1
m1

m1

AceS
AceS
AceS
AceS
PruS
AceS
AceS
AceS
AceS
CarC
TilA
AceS
AceS
AceS
AceS
AceS

mm
nu
Hm
nm

AceS
AceS
PruS
OstV

AceS
AceS
AceS
FraA
FraA

QueR

AceS
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
A068
A068
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

4.7
2.8
44.9
4.1
4.2
2.5
73.1
4.0
4.1
4.0
3.5
35.2
52.7
2.6
7.5
5.6
3.5
5.9
3.0
3.1
65.7
5.8
18.4
2.7
63.5
4.6
3.4
4.6
3.2
2.5
7.0
8.6
6.0
2.8
3.1
7.4
19.8
2.8
4.0
3.3
3.8
2.8
3.0
5.0
6.8
2.8
2.8
2.6
6.2
3.0
4.0
3.0
5.6
2.6
2.7
8.2

mmmmomowwmmmmmmoom0mmcanmmmmwmmcoommomOCmmmmmggcwo:gonzo

mmmmmzzzzzzzzzzmmmmgggmmmmwmwmzzzzmmmmmmmmmmmsssssswmwww

 

 

 

8888288888882

EEEQEEEBEEEEEEEEEEEEEEEEEEEEEEEEEEEEEREEEEE

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
A068
A068
UlmR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV

AceS
AceS

3.3
30.9
3.0
3.0
6.7
18.6
6.1
2.7
2.7
50.1
33.2
80.0
3.1
2.5
6.3
7.2
15.2
6.4
7.2
3.5
6.8
4.6
3.9
3.7
2.7
5.5
17.8
2.7
53.8
2.7
3.0
2.5
56.3
20.8
66.3
56.7
4.3
8.2
13.5
3.2
51.4
5.6
18.1
8.4
8.7
44.1
11.2
4.3
2.5
2.6
3.2
3.1
2.9
3.0
3.2
2.9

Ommmmmommomma:mommmmmmmmmwmmmmmmmmmmwmmmmmmmmmmcommommmom

”weﬁéééé€2mmmmmmmmmzzzzzzzzzzmmmmmmmmmmmmmmggggggggmmmmm

 

 

8%888888888888888888888888888888888888888EEEEEEEEEEBEEEE

 

 

 

1%

AceS
AceS
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS
A068
A068
A065
AceS
AceS
AceS
AceS
AceS

AceS

34.2
3.8
3.4
2.9
4.9
3.1
57.6
52.3
18.7
4.0
3.5
4.9
3.0
8.1
3.7
7.1
11.7
4.3
2.9
7.2
24.2
11.4
4.5
17.5
105.7
17.9
33.2
3.7
3.2
3.5
4.3
3.9
3.6
21.4
3.3
3.6
7.2
13.7
6.7
3.5
5.3
4.7
6.7
3.6
5.0
2.9
6.3
6.4
65.7
9.6
2.9
12.6
15.4
13.9
6.4
2.7

ccococccmacco:mmcmcmmmgczs:o:mmwomaccmcocc:commmmucoommmo

°éééééé£2éémmmmmmmzzzzzmmmmmmsssasgegwmwwwmmzzzzzzzzzmmm

 

 

 

 

888838838888883838888833888838838

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
FagG
FagG
TilA

AceS
AceS
AceS
FagG

FagG
UlmR
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS

5.6
5.1
6. 4
20. 5
8.4
3. 9
38. 7
17.0
61.4
3.0
7. 5
30. O
2.9
6.9
7.6
6.0
9. 7
19. O
3.0
2. 8
17. 4
4.8
7.0
2. 8
11. 5
8.1
3. 5
14. 3
20.7
10.8
12.8
6.7
6.8
3.0
7.2
8.5
5.3
8. 2
29. 6
5.9
3. 6
14. 6
4.2
5.9
9.1
5.7
6.1
5.1
3.5
3.1
6.5
3.8
44.6
2.6
5.6
6.9

onozocmcmgmwcggmggcccmmomommmgcwmmccmmcoommmcggcmmmmmmmc

sgsmmmubana50000000000003gggagmmmmmmmmmppappnanoooaoaono

 

 

 

AceS
AceS
AceS
AceS
AceS
FraA
AceS
AceS
AceS
AceS
AceS
TilA
AceS
AceS
AceS
AceS
FraA
AceS
AceS
AceS
AceS
TilA
TilA
TilA
TilA
CarC
CarC
AceS
AceS
TilA
AceS
AceS
AceS
OstV
UlmA
CarC
AceS
CarC
AceS
AceS
AceS

‘ AceS

AceS
AceS
CarC
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

5.7
8. 2
l9. 5
21.5
12.8
2.7
4.2
3. 2
33. 7
60.5
3.4
2. 7
85. 9
41.7
5.9
3.0
4.1
4.3
2.5
6.3
4.7
7.0
4.1
4.5
3.0
4. 3
15. 7
7.2
5.5
9.9
9. 6
32.1
5.4
6.2
2.6
6.3
4. 3
54.3
3.7
4.6
9. 6
14. 0
15.8
29.6
9.5
7.3
4.1
3.1
11.4
25.5
29.2
27.5
28.9
34.5
39.2
66.5

mmmmommmmmmmoomomovzoommommomcnoomoomowmmowcommmmmmoommmmm

zzmmmmmmmmggggggggmwmzzzzzmmmmmggggggggmmmmmzzzzzzmmmmmm

 

 

 

 

 

1%

n1 AceS 17.0 s N
n1 AceS 65. 7 s N
n1 AceS 4 .0 s N
'nl AceS 45.5 s N
n1 AceS 4. 3 s N
n1 AceS 32.9 s N
n] PruS 4.4 s N
n1 TilA 4. 2 s N
n] AceS 47. 2 s S
n1 AceS 2.6 s S
n1 AceS 3.1 s S
n1 AceS 3. 9 c S
n1 AceS 18. 9 s S
n1 AceS 3.7 s S
n1 AceS 5.4 s S
n1 AceS 6.4 s S
n1 AceS 5.4 s S
n1 AceS 7 .7 c W
n1 AceS 2.6 c W
n1 AceS 5.2 s W
n1 A068 3.9 s W
n1 AceS 4.9 s W
n1 AceS 5.5 s W
n1 OstV 4.3 s W
n2 AceS 3.2 s E
n2 AceS 4. 2 c E
n2 AceS 19. 7 c E
n2 AceS 18.7 s E
n2 AceS 2.8 s E
n2 AceS 3.5 s E
n2 FagG 2.5 s E
n2 FagG 2.5 s E
n2 FagG 2.6 s E
n2 FagG 2.6 s E
n2 AceS 2.6 s N
n2 AceS 2.6 s N
n2 AceS 2.6 c N
n2 AceS 2.7 s N
n2 AceS 2.7 s N
n2 AceS 2.9 s N
n2 AceS 2.9 c N
n2 AceS 3.1 s N
n2 FagG 3.1 s N
n2 OstV 3.1 s N
n2 AceS 3.1 s S
n2 FagG 3.1 s S
n2 FagG 3.2 s S
n2 FagG 3.2 s S
n2 FagG 3.4 c S
n2 FagG 3.4 s S
n2 FagG . 3.4 s S
n2 FagG 3.7 s S
n2 FagG 3.7 s S
n2 FagG 3.7 s S
n2 FagG 3.8 s S
n2 FagG 4.0 s S

 

 

 

 

333iiiiiB88633%338%BaBﬁBBBBBEE38Baﬂﬂaaaﬂaﬂaaaaﬂaaﬂﬂﬂﬂﬂﬂﬁ

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
A068

4.1
4.1
4.1
4.4
4.8
5.1
5.8
5.9
6. 5
10.1
17.7
3.7
9.4
8. 5
52.4
20.5
11.8
9.8
8.8
4. 2
56. 3
2. 6
23.1
5. 7
11.0
18.9
17. 8
6. 2
12. 2
2.5
3.8
3.7
2.5
9.1
3.7
6. 3
78.1
3.4
4.0
6.7
3.1
3.4
5.0
3.6
5.7
6.6
3.4
4.1
4.3
5.1
10. 5
38.6
10.4
2.9
3.0
6.7

EmmmC-‘V’I:“€3661:6OCCI:(06666::‘0‘001396'0217’62606836‘06mOmCSCV’OmV‘mV’mV"1’

zzzmmmmmggggggggggggggmmwwwwmmmmzzzzzzzzzzzzmmmmmggggéég

 

 

 

333333383333333333aaaﬁaaaﬂaaaaﬂaiiiiiiiiiiiiiiiiiiiiiiii

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceN
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS
AceS

2.6
2.5
4.0
2.8
40. 5
88. 4
2.8
3. 5
20. 3
2. 8
14. 8
2.6
2.6
2.7
2.7
2.5
8.3
8.6
7 7
70. 8
2. 8
58. 0
2.5
3.7
2.7
2.7
2.7
2.8
2.6
7.1
3.7
5. 4
63.3
2. 9
19. 6
68.6
26. 3
5. 7
69. 5
7.2
5. 2
37. 8
2.5
4.1
7.4
2.8
2. 6
28.1
5.4
2.5
5.3
9. 9
21.3
3.0
3.0
3.2

mccmoccccccocococccc00:00::czar:t:mwcwowccwcnmccccmcccwcco

£2€mmmmmmmwwmmmzzzzzzzzmmmmmmmmmggggggggggggmmmmmmmmmzzz

 

 

 

 

 

196

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
CelO
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

AceS -

AceS
A038
A068
AceS
AceS
A065
A668
A068
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

3.2
2.9 '
4.6
2.7
2.5
4.0
4.3
7.1
47. 2
2.8
3.5
3.0
4.2
5. 4
27. 7
4.5
4. 0
102.6
63.8
21.8
32. 0
2. 6
37. 9
3.0
2.8
2.7
3. 9
69. 9
30.6
3.0
2.9
2.7
3.4
2.9
3.1
3.2
7.0
2.6
4.0
3.1
3.3
3. 6
51.4
8. 6
31.0
2.7
8.1
11.1
3.6
60.9
6. 9
28. 2
23.0
39.7
3.3
4.3

cnmommmmmmommmmmm0mmmmmmommmwmmmmmmmmmmmmommmmmmmmmmmcccm

zzzz'mmmmmggggggggmmmmmmmmmmmmmmmmzzzzzzzzzmmmmmmmmmmgggﬁ

 

 

 

 

n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n7
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
n8
01
01
01
01

A068
AceS
AceS
AceS
CarC
TilA
TilA

AceS
AceS
AceS
AceS

2.9
2.7
3.5
7.0
29.4
9.4
4.4
3.6
3.1
4.7
3.3
5.2
2.5
3.3
2.7
2.8
4.6
3.5
3.0
5.8
3.3
3.2
3.0
13.4
2.5
12.5
4.5
4.6
2.8
6.7
5.7
3.7
3.2
4.0
3.6
4.1
3.3
2.5
2.9
2.7
11.4
42.7
2.9
3.5
2.9
2.6
2.7
3.5
2.7
2.5
2.9
2.6
2.7
13.0
2.9
60.0

monomomoommmommmmommmoommnmommmmmmmmmoomwcnmmmmmmmgmmmmmm

mmmmggsgg2g2gmmmmmmmmzzzzzzmmmmmmmmmmggggggwwmwwzzzzzzzz

 

 

 

 

01
01

01
01
01

01
01
01
01
01
01
01
01
o 1
01
01
01
01
01
01
01
o 1
o 1
01
01
01
01
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02
02

OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
FagG
F agG
FagG

AceS
AceS
OstV
OstV
OstV
OstV
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

4.8
4.2
3.4
2.8
4.1
16.6
3.6
70.3
3.3
6.7
13.4
58.9
61.0
6.2
4.3
5.8
8.1
66.7
5.3
8.2
4.3
3.1
53.5
29.1
43.8
15.2
20.6
11.5
5.9
6.3
4.8
3.5
19.4
3.8
3.2
53.0
6.6
4.8
5.1
8.7
5.8
53.2
6.5
5.7
5.3
6.9
54.7
8.1
5.6
4.6
47.1
5.5
6.3
5.2
2.6
45.0

mmmmmmmmmmmmwmmmmmmommmmmmmomornmwooommmmmmmmmmmowmmmmmwm
222eeeewwwwwwzzzzzzzmmmmmmmaseeesseeesemmwwwwmzzzzzzzzzm

 

 

 

888883338833333833883833

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
A068
A068
AceS
AceS

11.7
2.6
4.5
5.2
7.0
5.8
9.4
7.0
3.1
4.0
3.1
2.6
3.1
2.7
2.5
2.8
3.3
3.2
2.5
58.7
3.1
4.2
56.5
60.5
4.1
5.1
64.0
3.1
2.5
2.5
3.2
3.1

4.8 -

4.2
3.0
3.3
3.5
2.5
3.3
5.7
3.1
2.9
3.4
5.2
5.5
82.1
50.9
3.5
7.0
10.3
8.3
7.6
7.2
2.7
3.1
5.1

1:131:mwcncc:coca:ccocmm:cocsccaccccccscccccccccoccccocoommm
gaggmmmmwmzzzzzzzmmmmmmmggggasgmmmmmmzzzzzzzzzmmmmmmmggs

 

 

197

 

 

 

05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
05
06

06
06
06
06
06

06
06
06
06
06
06

06
06

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
AceS
AceS
FagG
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG

FagG
AceS

AceS
AceS
FagG
FagG
F agG
FagG
FagG
AceS
AceS
FagG
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
FagG

AceS

6. 4
23. 0
2.6
4.1
5. 6
16. 7
14.5
10.4
13.3
3.7
4.8
4.5
5.8
6.9
7.2
3.5
6. 5
10. 7
48. 5
3. 5
49.6
6.9
8.7
4.3
2. 8
50. 9
6.5
2.9
3. 7
39.1
2. 9
39. 4
50. 9
3. 3
50. 0
30.4
57.0
47.1
2.9
2.7
4.4
3.8
9.6
2.8
2.8
3.4
8.6
3.9
7.4
5.8
5. 4
30.1
5.2
3. 7
11.5
4.1

mmomommmmmmmmmmmmmmmo:coccccccgcm:cgccccccogmmccmcocamgg

mzzzzzzzzzmmmmmmmmmmmggsg$2ﬁgmmmmmmmmmmzzzzzzzmmmmmmmmgg

 

 

 

06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
06
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07

AceS
AceS
AceN
AceS
AceS
AceS
AceS
AceS
AceS
FagG
AceS
AceS
AceS
UlmR

AceS
C610
C610
AceS
AceS
AccS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

UlmA
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
UlmA

3.2
4. 7
59.0
2. 9
82.8
4.8
6. 7
10. 2
9.1
6.1
11.2
3. 3
77.5
4.5
5.0
8.1
5.6
6.2
2.5
2.7
3.4
3.5
3.8
3.8
4.7
5.0
5.3
5.4
5.7
5.8
6.0
6.1
6.3
6.3
6.4
8.0
8.7
8. 7
10. 2
22.3
30.9
3.6
5.6
6.6
6.7
7.2
7.7
9. 1
72. 7
2.7
9. 8
38. 4
5.1
6.3
3.1
3. O

mmmmmmmmmmmoommmmmmmmmmmmcnmmmmmmcmmomommmmmmnmmmmmmmmomm

zzzzzzzzzzzzzzzmmmmmmmmmmmmmeggggggmmmmmwmmmmmwmmwmmmmmm

 

 

 

 

 

198

07 UlmR 3.8 s N
07 AceS 4.2 s N
07 AceS 2.8 s N
07 AceS 5.5 s N
07 A068 2.7 s N
07 AceS 2.9 s N
07 UlmA 3.0 s S
07 UlmA 5.0 s S
07 UlmR 4.7 s S
07 FraA 2.6 s S
07 FraA 5.1 c S
07 AceS 4. O s S
07 A068 58.3 s S
07 AceS 52.0 s S
07 AceS 52.4 s S
07 AceS 3. 6 s S
07 AceS 49.0 s S
07 A068 4.0 s S
07 AceS 9. 0 s S
07 AceS 27. 3 s S
07 UlmA 33.7 s W
07 AceS 29.6 s W
07 AceS 25.2 s W
07 AceS 7. 8 s W
p1 AceS 10. 5 c N
pl AceS 48.3 s N
p1 AceS 34.0 s N
p1 AceS 39.2 s N
p1 AceS 2.8 c N
p1 AceS 44.1 s N
p1 AceS 48. 4 .c S
p1 AceS 15.0 s S
p1 AceS 8. 8 s S
p1 AceS 37. 7 s 8
p1 OstV 2. 7 s S
p1 OstV 10. 5 s S
pl AceS 3.3 c W
p1 AceS 6. 7 s W
p1 AceS 30. 9 s W
pl AceS 6.2 s W
p1 AceS 5.0 c W
pl AceS 7.3 s W
p1 FagG 5 .4 s W
p1 FagG 31. 5 c W
p1 OstV 4.1 s W
pl PruS 3. 8 n W
p1 PruS 12. 5 s W
p1 TilA 5. 2 s W
p2 AceS 10. 0 s E
p2 AceS 5. 9 s E
p2 AceS 11 .5 s E
p2 AceS 8.4 s E
p2 AceS 2.5 s E
p2 AceS 5.9 s E
p2 AceS 5.2 s E
p2 AceS 8.7 c E

 

 

 

 

p2 AceS 5.9 s E
p2 AceS 4.1 s E
p2 OstV 5.2 s E
p2 OstV 2.9 s E
p2 AceS 1 1.2 s N
p2 AceS 54.0 s N
p2 AceS 3.8 s N
p2 AceS 18.4 c N
p2 OstV 3.0 s N
p2 OstV 4.9 s N
p2 OstV 7.5 s N
p2 AceS 48.9 s S
p2 AceS 6.7 s S
p2 A668 4.4 s S
p2 AceS 6.6 s 8
p2 AceS 9.0 s S
p2 OstV 5.0 s S
p2 AceS 5.8 c W
p2 AceS 2.5 s W
p2 AceS 24. 1 c W
p2 A068 5. 1 c W
p2 AceS 3.5 s W
p2 AceS 3.6 s W
p2 AceS 3.7 c W
p2 OstV 3.5 s W
p2 OstV 3.0 s W
p3 AceS 3.4 s E
p3 AceS 32. 1 s E
p3 AceS 34.7 s E
p3 AceS 4.1 s E
p3 AceS 47.8 s E
p3 AceS 4.0 s E
p3 AceS 4.0 s E
p3 AceS 3.9 c E
p3 AceS 2.7 s E
p3 OstV 3.7 s E
p3 AceS 3. 1 s N
p3 AceS 38.3 s N
p3 AceS 7.5 c N
p3 A068 3.5 s N
p3 AceS 6.4 s N
p3 AceS 7.7 s N
p3 AceS 4.0 s N
p3 AceS 3.9 s N
p3 OstV 2.9 s N
p3 OstV 53.7 s N
p3 OstV 63.8 s N
p3 AceS 4.9 s S
p3 AceS 2.9 c 8
p3 AceS 3.4 s S
p3 AceS 2.8 s S
p3 AceS 67.0 s S
p3 AceS 39.8 s S
p3 AceS 7.3 c S
p3 AceS 3.9 c S
p3 AceS 1 5 .6 s W

 

 

 

AceS
AceS
AceS
AceS
AceS
AceS
AceS
FagG
OstV
AceS
AceS
AceS
AceS
AceS
AceS
OstV
AceS
AceS
AceS
FagG
OstV
AceN
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceN
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
Ach
AceS
F raA
AceS
AceS
AceS
AceS
AceS

24.3
2.5
25.8
2.8
2.5
2.7
2.7
2.7
4.2
12.0
7.0
2.8
4.1
29.5
38.9
10.5
20.1
23.6
34.6
39.4
10.5
10.5
14.7
8.5
5.0
14.9
4.0
10.1
34.7
8.6
4.2
3.0
2.7
24.8
29.0
3.1
3.9
3.1
5.5
4.6
3.5
2.5
3.2
3.6
5.2
5.5
2.7
3.6
2.7
2.6
2.8
5.4
3.4
3.8
3.5
4.6

mmmmmmgggggggmmwzzzzzmmmgégéémmmmwmzZZZmemmmmm€€€€€€€€€

 

 

 

 

 

199

p6 AceS 3.6 s N
p6 AceS 19.3 s N
p6 A068 4.4 c N
p6 AceS 4.4 s N
p6 AceS 8.0 s N
p6 AceS 4.4 s N
p6 FraA 4.6 s S
p6 FraA 4.8 s S
p6 FraA 3.2 s S
p6 AceS 3.2 c S
p6 AceS 25.7 c S
p6 AceS 2.5 s S
p6 AceS 4.0 s S
p6 AceS 3.0 s S
p6 AceS 3.5 s W
p6 AceS 2.9 s W
p6 A068 3.7 s W
p6 AceS 20.5 s W
p6 AceS 3.7 s W
p6 AceS 8.8 s W
p6 AceS 3.3 s W
p6 AceS 2.7 s W
p6 AceS 4.7 s W
p6 AceS 3.5 s W
p6 AceS 1 1.9 s W
p6 AceS 3.9 s W
p6 AceS 2.9 s W
p6 AceS 2.6 s W
p6 AceS 2.6 s W
ql AceS 3.0 s E
ql AceS 39.9 c E
ql AceS 7.1 c E
ql AceS 3.4 s E
ql AceS 3.9 s E
ql AceS 2.7 s E
ql AceS 19.2 s N
ql AccS 2.7 s N
ql AceS 3.1 s N
ql AceS 7.6 s N
ql AceS 3.4 s N
ql A068 3. l s S
ql A068 2.9 s S
ql A068 3.6 s S
ql AceS 13.0 c S
ql A068 3.0 s W
ql AceS 4.1 s W
ql AceS 3.3 s W
ql AceS 3. l s W
ql AceS 4.9 s W
ql AceS 9.5 s W
ql AceS 44.7 s W
ql AceS 4.0 s W
ql AceS 3.0 c W
q] FagG 44.3 s W
q2 AceS 6.9 c E
q2 AceS 3.0 s E

 

 

 

 

q2 AceS 2.9 s E
q2 AceS 4.2 s E
q2 AceS 5.3 s E
q2 OstV 2.8 s E
q2 AceS 4. l c N
q2 AceS 3.1 s N
q2 AceS 3.3 s N
q2 AceS 9.1 s N
q2 AceS 61.8 s N
q2 AceS 53.4 s N
q2 AceS 6.7 s N
q2 AceS 5.2 s S
q2 AceS 2.7 s S
q2 AceS 3.9 s S
q2 AceS 2.5 c S
q2 AceS 10.3 c S
q2 AceS 3.0 s S
q2 AceS 2.6 s S
q2 AceS 4.5 s S
q2 A068 3.4 s S
q2 AceS 3.4 s S
q2 AceS 9.2 s S
q2 AceS 16.2 s S
q2 OstV 4.8 s S
q2 AceS 5.2 s W
q2 AceS 72.6 s W
q2 AceS 10.0 s W
q2 AceS 17.6 s W
q2 A068 9. 1 c W
q2 AceS 16.1 s W
q2 AceS 1 1.4 s W
q2 OstV 6.5 s W
q3 A068 7.4 s E
q3 AceS 4.3 s E
q3 AceS 10.1 s E
q3 AceS 5.9 s E
q3 AceS 97.0 s E
q3 A068 5 1 .2 s E
q3 AceS 7.6 c E
q3 AceS 2.9 s E
q3 AceS 6.2 c E
q3 AceS 7.3 s E
q3 A068 4. 1 s E
q3 AceS 1 1.8 s E
q3 AceS 47.0 s E
q3 AceS 55.2 s E
q3 AceS 2.8 s E
q3 AceS 5.9 s E
q3 AceS 3.7 s E
q3 AceS 3.0 s E
q3 AceS 2.8 s E
q3 AceS 3.0 s E
q3 A068 5.2 s E
q3 AceS 3.4 s E
q3 AceS 2.5 s E
(13 AceS 3.6 s E

 

 

 

AceS
AceS
AceS
AceS
AceS
AceS
FraA
OstV
PruS
TilA

AceS
AceS
AceS
AceS
AceS
TilA

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
PruS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
A068
A068
AceS
AceS
AceS
AceS
AceS
AceS
AceS

22.1
21.9
4.0
3.6
36.9
43.8
6.2
36.5
7.0
10.8
8.1
8.1
3.0
49.1
61.6
6.5
62.5
5.9
2.9
3.3
6.2
3.1
4.6
7.4
3.8
3.1
4.5
5.6
7.6
4.3
8.0
7.8
80.7
6.6
5.1
3.7
6.8
5.5
4.4
5.2
6.9
9.1
11.6
2.9
2.7
4.8
4.5
4.7
53.3
4.0
5.9
5.6
3.8
2.6
2.6
54.8

22zzzzzzzzzzmmmmmmmggwmwwwmmzzzzzzzzzmmoaanonmmmmmmmmmmm

 

 

 

11
11
11
1'1
11
11
11
11
11
1'1
11
11
1'1
11
12
12
12
12
12
12
12
12
12
12
1'2
12
12

12
12
12
12

12
12
12
12

12
12
12
12
13
13
13
13
13
13
14
14
14
1'4
1'4
14
r4
r4

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
OstV
OstV
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
FraA
TilA

TilA

TilA

FraA
AceS
AceS
AceS
AceS
AceS
AceS

3.3
2.7
32.6
4.0
2.6
70.4
2.9
16.4
52.9
11.1
12.8
4.7
8.7
4.4
17.8
3.2
2.7
19.6
43.2
43.7
3.7
4.0
4.7
3.4
2.6
5.3
2.5
3.1
2.6
3.4
3.6
47.0
3.5
4.4
2.6
4.5
2.8
5.0
8.0
2.6
32.9
2.5
5.6
9.1
2.6
3.4
5.3
3.2
5.0
3.6
4.3
2.5
4.6
4.4
9.1
58.0

mmmmowmommmmonmmmmmmommmmmmmmmwmmmmommmmmmmgmmmmmmmmmmmm

mmmmmmmmnnas:agggsgggmzzzzzzzzmmmmmmmmmmmmgaggmmmmmmmmmm

 

 

 

2W

 

 

 

AceS

AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS

mmmmmmmwmmcowmmmmwoommomwmmwwmwot/300mmmmooommmmmwcoommmwo

zzzzzzzzzzzzmmmmmmmmmmms£282828288228ssssssnbnnannbmmmmm

 

 

AceS
AceS
UlmA
AceS
AceS
AceS
AceS
OstV
AceS
AceS
AceS
AceS
AceS
AceS
FagG
OstV
OstV
OstV

FraA
FraA
AceS
AceS
AceS
AceS
AceS
AceS
AceS
AceS
TilA

AceS
AceS
AceS
AceS
AceS
AceS
OstV

monmmmmmoommmmmmmoommowownommmmmmowonommomwmmmwowmwmomww

mmmssssssewwwwwmwwmzzzzzzzwwmmmmmmmmmsssszéssseswwwwmzzz

 

AceS
AceS
AceS
AceS
AceS
PruS
UlmA
AceS
AceS
AceS
AceS
AceS
AceS

mmmwwmowmmnoowommm'1:mmmmov:1nInoooommmmmwmmmwwmmmommmwwmm

ggggggggﬁgggﬁggmmmwmmmmmwwmmmmmzzzzzzzzzzzzzzzmmmmmmm

 

 

 

 

Literature Cited

Abrell, DB, and M.T. Jackson. 1977. A decade of change in an old-growth beech-
maple forest in Indiana. The American Midland Naturalist 98: 22-32.

Aikrnan, J .M., and A.W. Smelser. 1938. The structure and environment of forest
communities in central Iowa. Ecology 19: 141-150.

Aizen, M.A., L. Ashworth, and L. Galetto. 2002. Reproductive success in fragmented
habitats: do compatibility systems and pollination specialization matter? Journal of
Vegetation Science 13: 885-892.

Anderson, KL, and DJ. Leopold. 2002. The role of canopy gaps in maintaining
vascular plant diversity at a forested wetland in New York State. Journal of the Torrey
Botanical Society 129: 238-250.

Anderson, R.C., T.C. Kelley, and SS. Dhillion. 1996. Aspects of the ecology of an
invasive plant, garlic mustard (Alliaria petiolata), in central Illinois. Restoration Ecology
4: 1 8 1-1 91 .

Ankney, R.M. 1988. History of the Rose Lake Wildlife Research Center. Wildlife
Division Report No. 3054, Michigan Department of Natural Resources.

Archibold, O.W., D. Brooks, and L. Delanoy. 1997. An investigation of the invasive
shrub European buckthom, Rhamnus cathartica L., near Saskatoon, Saskatchewan.
Canadian F ield-Naturalist 1 1 I : 617-621 .

Asbjornsen, H., K.A. Vogt, and MS. Ashton. 2004. Synergistic responses of oak, pine
and shrub seedlings to edge environments and drought in a fragmented tropical highland
oak forest, Oaxaca, Mexico. Forest Ecology and Management 192: 313-334.

Baker, H.G. 1989. Some aspects of the natural history of seed banks. In Leck, M.A.,
V.T. Parker, and R.L. Simpson. Ecology of Soil Seed Banks. Academic Press, Inc., New
York.

Barnes, B.V., and W.H. Wagner. 1981. Michigan Trees. The University of Michigan
Press, Ann Arbor, Michigan.

Baskin, J .M., and CC. Baskin. 1992. Seed germination biology of the weedy biennial
Alliaria petiolata. Natural Areas Journal 12: 191-197.

Beach, J .H., and W.D. Stevens. 1980. A study of Baker Woodlot. 11. Description of
vegetation. The Michigan Botanist 19: 3-13.

Beaman, J .H. 1970a. A botanical inventory of Sanford Natural Area. I. The
environment. The Michigan Botanist 9: 116-139.

202

Beaman, J .H. 1970b. A botanical inventory of Sanford Natural Area. 11. Checklist of
vascular plants. Michigan Botanist 9: 147-164.

Beatty, S.W. 1991. Colonization dynamics in a mosaic landscape: the buried seed pool.
Journal of Biogeography 18: 553-563.

Blouin-Demers, G., and P.J. Weatherhead. 2001. Habitat use by black rat snakes
(Elaphe obsoleta obsoleta) in fragmented forests. Ecology 82: 2882-2896.

Bossuyt, B., and M. Hermey. 2001. Inﬂuence of land use history on seed banks in
European temperate forest ecosystems: a review. Ecography 24: 225-238.

Bossuyt, B., M. Heyn, and M. Hermy. 2002. Seed bank and vegetation composition of
forest stands of varying age in central Belgium: consequences for regeneration of ancient
forest vegetation. Plant Ecology 162: 33-48.

Boulton, R.L., and M.F. Clarke. 2003. Do yellow-faced honeyeater (Lichenostomus
chrysops) nests experience higher predation at forest edges? Wildlife Research 30: 119-
125.

Burton, A.J., K.S. Pregitzer, and DD. Reed. 1991. Leaf-area and foliar biomass
relationships in northern hardwoods forests located along an 800 km acid deposition
gradient. Forest Science 3 7: 1041-1059.

Braun, EL. 1950. Deciduous Forests of Eastern North America. Hafner Publishing
Company, New York, NY.

Bréda, N.J.J. 2003. Ground-based measurements of leaf area index: a review of
methods, instruments and current controversies. Journal of Experimental Botany 54:
2403-2417.

Brothers, TS, and A. Spingarn. 1992. Forest fragmentation and alien plant invasion of
central Indiana old-grth forests. Conservation Biology 6: 91-100.

Brotons, L., A. Desrochers, and Y. Turcotte. 2001. Food hoarding behaviour of black-
capped chickadees (Poecile atricapillus) in relation to forest edges. Oikos 95: 511-519.

Buckley, G.P., R. Howell, and MA. Anderson. 1997. Vegetation succession following
ride edge management in lowland plantations and woods. 2. The seed bank resource.
Biological Conservation 82 : 305-3 1 6.

Burke, D.M., and E. N01. 1998. Edge and ﬁ'agment size effects on the vegetation of
deciduous forests in Ontario, Canada. Natural Areas Journal 18: 45-53.

203

Cavers, P.B., M.I. Heagy, and RF Kokron. 1979. The biology of Canadian weeds. 35.
Alliaria petiolata (M. Bieb) Cavara and Grande. Canadian Journal of Plant Science 59:
217-229.

Cain, SA. 1935. Studies on virgin hardwood forest: III. Warren’s Woods, a beech-
maple climax forest in Berrien County, Michigan. Ecology 16: 500-513.

Chason, J .W., D.D. Baldocchi, and MA. Huston. 1991. A comparison of direct and
indirect methods for estimating forest canopy leaf area. Agricultural and Forest
Meteorology 5 7: 107-128.

Chen, J ., J .F . Franklin, and T.A. Spies. 1992. Vegetation responses to edge
environments in old-grth Douglas-ﬁr forests. Ecological Applications 2: 387-396.

Cadenasso, M.L., and S.T.A. Pickett. 2001. Effect of edge structure on the ﬂux of
species into forest interiors. Conservation Biology 15: 91-97.

Cadenasso, M.L., M.M. Traynor, and S.T.A. Pickett. 1997. Functional location of forest
edges: gradients of multiple physical factors. Canadian Journal of Forest Research 27:
774-782.

Cobbe, TJ. 1943. Variations in the Cabin Run Forest, a climax area in southwestern
Ohio. American Midland Naturalist 29: 89-105.

Colautti, R.I., and H]. Maclsaac. 2004. A neutral terminology to deﬁne ‘1nvasive’
species. Diversity and Distributions 10: 135-141 .

Collier, M.H., J .L. Vankat, and MR. Hughes. 2002. Diminished plant richness and
abundance below Lonicera maackii, an invasive shrub. American Midland Naturalist
14 7: 60-71 .

Collins, B.S., and S.T.A. Pickett. 1988. Demographic responses of herb layer species to
experimental canopy gaps in a northern hardwoods forest. Journal of Ecology 76: 43 7-
450.

Crow, GE. 1969. Species of vascular plants of Pennﬁeld Bog, Calhoun County,
Michigan. Michigan Botanist 8: 131-136.

Davis, M.B., ed. 1996. East Old-Growth Forests: Prospects for rediscovery and
recovery. Island Press, Washington, DC.

Decocq, G., B. Valentin, B. Toussaint, F. Hendoux, R. Saguez, and J. Bardat. 2004. Soil

seed bank composition and diversity in a managed temperate deciduous forest.
Biodiversity and Conservation 13: 2485-2509.

204

Dickman, D.I., and L.A. Leefers. 2003. The Forests of Michigan. University of
Michigan Press, Ann Arbor, MI.

Didharn, R.K., and J .H. Lawton. 1999. Edge structure determines the magnitude of
changes in microclimate and vegetation structure in tropical forest fragments. Biotropica
31: 17-30. -

Donnelly, G.T., and PG. Murphy. 1987. Warren Woods as forest primeval: a
comparison of forest composition with presettlement beech-sugar maple forests of
Berrien County, Michigan. Michigan Botanist 26: 17-24.

Dufrene, M, and P. Legendre. 1997. Species assemblages and indicator species: the
need for a ﬂexible asymmetrical approach. Ecological Monographs 67: 354-366.

Ehrenfeld, J .R. 1980. Understory response to canopy gaps of varying size in a mature
oak forest. Bulletin of the Torrey Botanical Club 107: 29-41.

Ehrenfeld, J .R. 1997. Invasion of deciduous forest preserves in the New York
metropolitan region by Japanese Barberry (Berberis thunbergii DC). Journal of the
Torrey Botanical Society 124: 210-215.

Ellsworth, D.S., and PB. Reich. 1993. Canopy structure and vertical patterns of
photosynthesis and related leaf traits in a deciduous forest. 0ecologia 96: 169-178.

Ellsworth, J .W., R.A. Harrington, and J .H. Fownes. 2004. Seedling emergence, growth,
and allocation of Oriental bittersweet: effects of seed input, seed bank, and forest ﬂoor
litter. Forest Ecology and Management 190: 255-264.

Elton, OS. 1958. The ecology of invasions by animals and plants. Methuen, London.

Esten, M.M. 1932. A statistical study of a beech-maple association at Turkey Run State
Park, Parke County, Indiana. Butler University Botanical Studies 2: 183-201.

Pagan, M.B., and DR. Peart. 2004. Impact of the invasive shrub glossy buckthom
(Rhamnusfrangula L.) on juvenile recruitment by canopy trees. Forest Ecology and
Management 194:95-1 07.

Fassnacht, K.S., and S.T. Gower. 1997. Interrelationships among the edaphic and stand
characteristics, leaf area index, and aboveground net primary production of upland forest

ecosystems in north central Wisconsin. Canadian Journal of Forest Research 2 7: 1058-
1 067.

Fenner, M. 1985. Ecology of Seed Banks. In Fenner, M. 1985. Seed Ecology.
Chapman and Hall, New York.

205

Feist, M.A., L.R. Phillippe, D.T. Buserneyer, and J .E. Ebinger. 2004. Vegetation survey
of Dean Hills Nature preserve, Fayette County, Illinois. Castanea 69: 52-66.

Flaspohler, D.J., S. A. Temple, and RN. Rosenﬁeld. 2001. Species-speciﬁc edge effects
on nest success and breeding bird density in a forested landscape. Ecological
Applications 11: 32-46.

Flemming, C.A., and BE. Wofford. 2004. The vascular ﬂora of Fall Creek Falls State
Park, Van Buren and Bledsoe Counties, Tennessee. Castanea 69: 164-184.

F orcella, F ., and 8.]. Harvey. 1983. Eurasian weed infestation in western Montana in
relation to vegetation and disturbance. Madrono 30: 102-109.

Fore, S.A., J .L. Vankat, and R.L. Schaefer. 1997. Temporal variation in the woody
understory of an old-growth F agus-A cer forest and implications for overstory
recruitment. Journal of Vegetation Science 8: 607-614.

Fowler, SP, and KC. Larson. 2004. Seed germination and seedling recruitment of
Japanese honeysuckle in a Central Arkansas natural area. Natural Areas Journal 24 : 49-
53.

Fralish, 1.8., F .B. Crooks, J .L. Chambers, and RM. Harty. 1991. Comparison of
presettlement, second-growth, and old-growth forest on six site types in the Illinois
Shawnee Hills. American Midland Naturalist 125: 294-309.

Fraver, S. 1994. Vegetation responses along edge-to-interior gradients in the mixed
hardwood forests of the Roanoke River basin, North Carolina. Conservation Biology 8:
822-832.

Frazer, G.W., Canham, CD, and Lertzrnan, K.P. 1999. Gap Light Analyzer (GLA),
Version 2.0: Imaging software to extract canopy structure and gap light transmission
indices from true-colour ﬁsheye photographs, user’s manual and program documentation.
Simon Fraser University, Burnaby, British Columbia, and the Institute of Ecosystem
Studies, Millbrook, New York.

Gayek, A., and M.F. Quigley. 2001. Does topography affect the colonization of
Lonicera maackii and Ligustrum vulgare in a forested glen in southwestern Ohio? Ohio
Journal of Science 101: 95-100.

Gehlhausen, S.M., M.W. Schwartz, and CK Augspurger. 2000. Vegetation and
microclimatic edge effects in two mixed-mesophytic forest fragments. Plant Ecology
14 7: 2 l -3 5.

Gilbert, G.E., and V.L. Riemenschneider. 1980. Vegetative structure of an essentially

undisturbed beech-maple ecosystem in central Ohio. Ohio Journal of Science 80: 129-
133.

206

Gleason, H.A., and A. Cronquist. 1991. Manual of Vascular Plants of Northeastern
United States and Adjacent Canada. New York Botanical Garden, Bronx, NY.

Goldblum, D. 1997. The effects of treefall gaps on understory vegetation in New York
State. Journal of Vegetation Science 8: 125-132.

Gould, A.M.A., and D.L. Gorchov. 2000. Effects of the exotic invasive shrub Lonicera
maackii on the survival and fecundity of three species of native annuals. American
Midland Naturalist 144: 36-50.

Gorchov, D.L., and DE. Trisel. 2003. Competitive effects of the invasive shrub,
Lonicera maackii (Rupr.) Herder (Caprifoliaceae), on the growth and survival of native
tree seedlings. Plant Ecology 166: 13-24.

Grime, J .P. 1989. Seed banks in ecological perspective. In Leek, M.A., V.T. Parker, and
R.L. Simpson, eds. Ecology of Soil Seed Banks. Academic Press, Inc., New York.

Gysel, L.W. 1951. Borders and openings of beech-maple woodlands in southern
Michigan. Journal of Forestry 49: 13-19.

Harris, L.D. 1984. The Fragmented Forest: Island Biogeography Theory and the
Preservation of Biodiversity. University of Chicago Press, Chicago.

Held, M.B., and J .E. Winstead. 1975. Basal area and climax status in mesic forest
systems. Annals of Botany 39: 1147-1148.

Hierro, J .L., and R.M. Callaway. 2003. Allelopathy and exotic plant invasion. Plant
and Soil 256: 29-39.

Hobbs, R.I., and LP. Huenneke. 1992. Disturbance, diversity, and invasion:
implications for conservation. Conservation Biology 6: 324-337.

Honnay, O., K. Verheyen, and M. Herrny. 2002. Permeability of ancient forest edges for
weedy plant species invasion. Forest Ecology and Management 161: 109-122.

Hoppes, W.G. 1988. Seedfall pattern of several species of bird-dispersed plants in an
Illinois woodland. Ecology 69: 320-329.

Huenneke, LP. 1983. Understory response to gaps caused by the death of Ulmus
americana in central New York. Bulletin of the Torrey Botanical Club 110: 170-175.

Hutchison, B.A., D.R. Matt, R.T. McMillen, L.J. Gross, S.J. Tajchman, and J .M.

Norman. 1986. The architecture of a deciduous forest canopy in eastern Tennessee,
USA. Journal of Ecology 74: 635-646.

207

Hyatt, L.A. 1999. Differences between seed bank composition and ﬁeld recruitment in a
temperate zone deciduous forest. American Midland Naturalist 142: 31-38.

Jonckheere, 1., S. Fleck, K. Nackaerts, et a1. 2004. Review of methods for in situ leaf
area index determination - Part I. Theories, sensors and hemispherical photography.
Agricultural and Forest Meteorology 121: 19-35.

Kato, S., and A. Komiyama. 2002. Spatial and seasonal heterogeneity in understory
light conditions caused by differential leaf ﬂushing of deciduous overstory trees.
Ecological Research 17: 687-693.

Kloeppel, RD, and MD. Abrams. 1995. Ecophysiological attributes of native Acer

saccharum and exotic Acer platanoides in urban oak forests in Pennsylvania, USA. Tree
Physiology 15: 739-746. 1

Kolar, C.S., and D.M. Lodge. 2001. Progress in invasion biology: predicting invaders.
Trends in Ecology and Evolution 16: 199-204.

Kron, K.A., and BS. Walters. 1986. A study of Red Cedar Natural Area. 11. Checklist
of vascular plants. Michigan Botanist 25: 35-44.

Kupfer, J. 1996. Patterns and determinants of edge vegetation of a midwestem forest
preserve. Physical Geography 1 7: 62-76.

Kupfer, J .A., G.P. Malanson, and J .R. Runkle. 1997. Factors inﬂuencing species
composition in canopy gaps: the importance of edge proximity in Hueston Woods, Ohio.
Professional Geographer I99 7: 165-178.

Landenberger, RE, and J .B. McGraw. 2004. Seed-bank characteristics in mixed-
mesophytic forest clearcuts and edges: does “edge effect” extend to the seed bank?
Canadian Journal of Botany 82 : 992-1000.

Lechowicz, M.J. 1984. Why do temperate deciduous trees leaf out at different times?
Adaptation and ecology of forest communities. The American Naturalist 124: 821-842.

Leckie, S., M. Vellend, G. Bell, M.J. Waterway, and M.J. Lechowicz. 2000. The seed
bank in an old-growth, temperature deciduous forest. Canadian Journal of Botany 78:
1 81-192.

Leverett, R. 1996. Deﬁnitions and history. In: Davis, M.B., ed. 1996. East Old-
Growth Forests: Prospects for rediscovery and recovery. Island Press, Washington, DC.

Lonsdale, W.M. 1999. Global patterns of plant invasions and the concept of invasibility.
Ecology 80: 1522-1536.

208

Luken, J .O., and N. Goessling. 1995. Seedling distribution and potential persistence of
the exotic shrub Lonicera maackii in fragmented forests. American Midland Naturalist
133: 124-130.

Luczaj, L., and B. Sadowska. 1997. Edge effect in different groups of organisms:
vascular plant, bryophyte and fungi species richness across a forest-grassland border.
F olia Geobotanica and Phytotaxinomica 32: 343-353.

MacQuarrie, K., and C. Lacroix. 2003. The upland hardwood component of Prince
Edward Island’s remnant Acadian forest: determination of depth of edge and patterns of
exotic plant invasion. Canadian Journal of Botany 81: 11 13-1 128.

Matlack, GR. 1993. Microenvironment variation within and among forest edge sites in
the eastern United States. Biological Conservation 66: 185-194.

Matlack, GR. 1994. Vegetation dynamics of the forest edge—trends in space and
successional time. Journal of Ecology 82: 113-123.

Matlack, GR, and RE. Good. 1990. Spatial heterogeneity in the soil seed bank of a
mature Coastal Plain forest. Bulletin of the Torrey Botanical Club 11 7: 143-152.

McClanahan, TR, and R.W. Wolfe. 1993. Accelerating forest succession in a
fragmented landscape: the role of birds and perches. Conservation Biology 7: 279-288.

McCune, B., and J .B. Grace. 2002. Analysis of Ecological Communities. MjM
Soﬁware, Gleneden Beach, Oregon.

McCune, B. and M. J. Mefford. 1999. PC-ORD. Multivariate Analysis of Ecological
Data. Version 4.33. MjM Software, Gleneden Beach, Oregon.

McNab, W.H., and D.L. Loﬁis. 2002. Probability of occurrence and habitat features for
oriental bittersweet in an oak forest in the southern Appalachian mountains, USA. Forest
Ecology and Management 155: 45-54.

Meekins, J .F., and BC. McCarthy. 2001. Effect of environmental variation on the
invasive success of a nonindigenous forest herb. Ecological Applications 11: 1336-1348.

Meier, A.J., S.P. Bratton, and DC. Duffy. 1995. Possible ecological mechanisms for
loss of vemal-herb diversity in logged eastern deciduous forests. Ecological Applications
5: 935-946.

Meiners, S.J., and K. LoGiudice. 2003. Temporal consistency in the spatial pattern of
seed predation across a forest — old ﬁeld edge. Plant Ecology 168: 45-55.

Meiners, S.J., and S.T.A. Pickett. 1999. Changes in community and population
responses across a forest-ﬁeld gradient. Ecography 22: 261-267.

209

Michigan Geographic Data Library (MiGDL). http://www.mcg1.state.mi.us/mgd1/

Michigan Natural Features Inventory (MNFI). 1986. Michigan natural community types.
Available online: http://web4.msue.msu.edu/mnﬁ/data/natural_community_types.pdf

Miller, KB, and D.L. Gorchov. 2004. The invasive shrub, Lonicera maackii, reduces
grth and fecundity of perennial forest herbs. 0ecologia 139: 359-375.

Mitchell, R.S., and GO Tucker. 1994. Flora of an unusually diverse virgin and old-
growth forest area in the southern Adirondacks of New York. Bulletin of the Torrey
Botanical Club 121: 76-83.

Monk, C.D., G.I. Child, and SA. Nicholson. 1969. Species diversity of a stratiﬁed oak-
hickory community. Ecology 50: 468-470.

Moore, M.R., and J .L. Vankat. 1986. Responses of the herb layer to the gap dynamics of
a mature beech-maple forest. American Midland Naturalist 115: 336-347.

Mourelle, C., M. Kellman, and L. Kwon. 2001. Light occlusion at forest edges: an
analysis of tree architectural characteristics. Forest Ecology and Management 154: 179-
192.

Murcia, C. 1995. Edge effects in fragmented forests: implications for conservation.
TREE 10: 58-62.

Murphy, P.G., R.R. Sharitz, and A.J. Murphy. 1974. Leaf-litter production in the aspen
and maple-birch forest types and the contribution by individual tree species. In Rudolph,
TD. 1974. The Enterprise, Wisconsin, Radiation Forest: Preirradiation Ecological
Studies. Technical Information Center, United States Atomic Energy Commission.

Nathan, R., and HG Muller-Landau. 2000. spatial patterns of seed dispersal, their
determinants and consequences for recruitment. Trends in Ecology and Evolution 15:
278-285.

Odurn, E.P. 1969. The strategy of ecosystem development. Science 164: 262-270.
Olano, J .M., I. Caballero, N.A. Laslcurain, J. Loidi, and A. Escudero. 2002. Seed bank
spatial pattern in a temperate secondary forest. Journal of Vegetation Science 13: 775-

784.

Oosting, H], and ME. Humphreys. 1940. Buried viable seeds in a successional series
of old ﬁeld and forest soils. Bulletin of the Torrey Botanical Club 6 7: 253-273.

Orians, G. H. 1986. Site characteristics favoring invasions. In H. A. Mooney and J. A.
Drake, eds. Ecology of Biological Invasions of North America and Hawaii. Springer-
Verlag, New York.

210

Palik, B.J., and PG. Murphy. 1990. Disturbance versus edge effects in sugar
maple/beech forest fragments. Forest Ecology and Management 32: 187-202.

Parker, GR. 1989. Old-growth forests of the central hardwood region. Natural Areas
Journal 9: 5-11.

Parker, G.R., D.J. Leopold, and J .K. Eichenberger. 1985. Tree dynamics in an old-
growth, deciduous forest. Forest Ecology and Management I I : 31-57.

Parker, G.G., and DJ. Tibbs. 2004. Structural phenology of the leaf community in the
canopy of a Liriodendron tulipifera L. forest in Maryland, USA. Forest Science 50: 387-
397.

Petty, R.O., and AA. Lindsey. 1961. Hoot Woods, a remnant of virgin timber, Owen
County, Indiana. Proceedings of the Indiana Academy of Science 71: 320-326.

Pickett, S.T.A., and M.J. McDonnell. 1989. Seed bank dynamics in temperate deciduous
forest. In Leck, M.A., V.T. Parker, and R.L. Simpson. Ecology of Soil Seed Banks.
Academic Press, Inc., New York.

Poulson, T.L., and W.J. Platt. 1996. Replacement patterns of beech and sugar maple in
Warren Woods, Michigan. Ecology 77: 1234-1253.

Prinzing, A., W. Durka, S. Klotz, and R. Brand]. 2002. Which species become aliens?
Evolutionary Ecology Research 4: 385-405.

Pyle, LL. 1995. Effects of disturbance on herbaceous exotic plant species on the
ﬂoodplain of the Potomac River. American Midland Naturalist 134: 244-253.

Pysek, P. 1998. Is there a taxonomic pattern to plant invasions? Oikos 82: 282-294.

Ranney, J .W., M.C. Bruner, and J .B. Levenson. 1981. The important of edge in the
structure and dynamics of forest islands. In Burgess, R.L., and D.M. Sharpe, eds. 1981.
Forest Island Dynamics in Man-Dominated Landscapes. Springer-Verlag, New York.

Read, RH. 1975. Vascular Plants of Pictured Rocks National Lakeshore. Michigan
Botanical Club, Special Publication No. 3, University of Michigan, Ann Arbor, MI.

Rejmanek, M. 2000. Invasive plants: approaches and predictions. Austral Ecology 25:
497-506.

Rheault, H., P. Drapeau, Y. Bergeron, and RA. Esseen. 2003. Edge effects on epiphytic

lichens in managed black spruce forests of eastern North America. Canadian Journal of
Forest Research 33: 23-32.

211

Richardson, D.M., P. Pysek, M. Rejmanek, M.G. Barbour, et al. 2000. Naturalization
and invasion of alien plants: concepts and deﬁnitions. Diversity and Distributions 6: 93-
107.

Ricketts, T.I-1.. 1999. Terrestrial ecoregions of North American: a conservation
assessment. Island Press, Washington, DC.

Riitters, K.H., J.D. Wickham, R.V. O’Neill, et al. 2002. Fragmentation of continental
United States forests. Ecosystems 5: 815-822.

Robertson, D.J., M.C. Robertson, and T. Tague. 1994. Colonization dynamics of four
exotic plants in a northern Piedmont natural area. Bulletin of the Torrey Botanical Club
121: 107-118.

Runkle, J .R. 1982. Patterns of disturbance in some old-grth mesic forests of eastern
North America. Ecology 63: 1533-1546.

Runkle, J .R. 1996. Central mesophytic forests. In: Davis, M.B., ed. 1996. East Old-
Growth Forests: Prospects for rediscovery and recovery. Island Press, Washington, DC

Runkle, J .R. 2000. Canopy tree turnover in old-growth mesic forests of eastern North
America. Ecology 81: 554-567.

Sanford, N.L., R.A. Harrington, and J .H. Fownes. 2003. Survival and growth of native
and alien woody seedlings in open and understory environments. Forest Ecology and
Management 183: 377-385.

SAS Institute. 1999. SAS User’s Guide. SAS Institute, Cary, NC.

Saulei, S.M., and MD. Swaine. 1988. Rain forest seed dynamics during succession at
Gogol, Papua New Guinea. Journal of Ecology 76: 1133-1152.

Schmelz, D.V., J .D. Barton, and AA. Lindsey. 1974. Donaldson’s Woods: two decades
of change. Proceedings of the Indiana Academy of Science 84: 234-243.

Schmelz, D.V., and AA. Lindsey. 1970. Relationships among forest types in Indiana.
Ecology 51 : 620-629.

Schneider, G. 1963. A 20-year ecological investigation in a relatively undisturbed sugar
maple-beech stand in southern Michigan. Doctoral dissertation, Michigan State
University.

Schneider, G. 1966. A 20-year investigation in a sugar maple-beech stand in southern

Michigan. Research Bulletin 15, Michigan State University Agricultural Experiment
Station.

212

Shotola, S.J., G.T. Weaver, P.A. Robertson, and WC. Ashby. 1992. Sugar maple
invasion of an old-grth oak-hickory forest in southwestern Illinois. American Midland
Naturalist 127: 125-138.

Simpson, R.L., M.A. Leck, and V.T. Parker. 1989. Seed banks: general concepts and
methodological issues. In Leck, M.A., V.T. Parker, and R.L. Simpson. 1989. Ecology
of Soil Seed Banks. Academic Press, Inc., New York.

Soderstrom, E. 1986. Effect of secondary forest on the soil seed bank of primary forest.
In F.E. Putz. Tropical Biology: an ecological approach. Organization of Tropical
Studies Publication OTS 86-1, Durham, NC.

Spyreas, G., J. Ellis, C. Carroll, and B. Molano-Flores. 2004. Non-native plant

commonness and dominance in the forests, wetlands, and grasslands of Illinois, USA.
Natural Areas Journal 24 : 290-299.

Steparrian, M.A., S.D. Sundberg, G.A. Baumgardner, and A. Liston. 1998. Alien plant
species composition and associations with anthropogenic disturbance in North American
forests. Plant Ecology 139: 49-62.

Stevens, W.D., and J .H. Beach. 1980. A study of Baker Woodlot III. Checklist of
vascular plants. Michigan Botanist 19: 51-69.

Stohlgren, T.J., G.W. Chong, L.D. Schell, K.A. Rimar, Y. Otsuki, M. Lee, M.A.
Kalkhan, and CA. Villa. 2002. Assessing vulnerability to invasion by nonnative plant
species at multiple spatial scales. Environmental Management 29: 566-577. ‘

Sutherland, S. 2004. What makes a weed a weed: life history traits of native and exotic
plants in the USA. 0ecologia 141: 24-39.

Tallmon, D.A., E.S. Jules, N.J. Radke, and LS. Mills. 2003. Of mice and men and
Trillium: cascading effects of forest fragmentation. Ecological Applications 13: 1193-
1203.

Telewski, F.W., and J.A.D. Zeevaart. 2002. The 120-yr period for D1. Beal’s seed
viability experiment. American Journal of Botany 89: 1285-1288.

Thompson, K., J .G. Hodgson, and T.C.G. Rich. 1995. Native and alien invasive plants:
more of the same? Ecography 18: 390-402.

Thompson, R.L., and CA. Fleming. 2004. Vascular ﬂora and plant communitiesof the
John B. Stephenson Memorial Forest State Nature Preserve (Anglin falls ravine),
Rockcastle County, Kentucky. Castanea 69: 125-138.

Tomirnatsu, H., and M. Ohara. 2004. Edge effects on recruitment of Trillium
camschatcense in small forest fragments. Biological Conservation 1 I 7: 509-519.

'213

Trernblay, NO, and GR Larocque. 2001. Seasonal dynamics of understory vegetation
in four eastern Canadian forest types. International Journal of Plant Sciences 1 62: 271-
286.

USDA Forest Service (USFS). 2001. 2000 RPA assessment of forest and range lands.
US Department of Agriculture, Forest Service, Washington, DC.

USDA Natural Resources Conservation Service (USNRCS). 2000. Summary Report
1997 National Resources Inventory. US Department of Agriculture, Natural Resources
Conservation Services, Washington, DC.

van Gardingen, P.R., G.E. Jackson, S. Hemandez-Daumas, G. Russell, and L. Sharp.
1999. Leaf area index estimates obtained for clumped canopies using hemispherical
photography. Agricultural and Forest Meteorology 94: 243 -257.

Vankat, J .L., W.H. Blackwell, and WE. Hopkins. 1975. The dynamics of Hueston
Woods and a review of the question of the successional status of the southern beech-
maple forest. Castanea 40: 290-311.

Van Wilgenburg, S.L., D.F. Mazerolle, and K.A. Hobson. 2001. Patterns of arthropod
abundance, vegetation, and microclimate at boreal forest edge and interior in two
landscapes: Implications for forest birds. Ecoscience 8: 454-461.

Vasconcelos, H.L., and F .J . Luizao. 2004. Litter production and litter nutrient
concentrations in a fragmented Amazonian landscape. Ecological Applications 14: 884-
892.

Vellend, M. 2002. A pest and an invader: White-tailed deer (Odocoileus virginianus
Zimm.) as a seed dispersal agent for honeysuckle shrubs (Lonicera L.). Natural Areas
Journal 22: 230-234.

Vitousek, P.M., C.M. D’Antonio, L.L. Loope, M. Rejmanek, and R. Westbrooks. 1997.
Introduced species: a signiﬁcant component of human-caused global change. New
Zealand Journal of Ecology 21: 1-16.

Voss, E.G. 1972. Michigan Flora. Part I: Gymnosperms and Monocots. Cranbrook
Institute of Science, Bloomington Hills, MI.

Voss, E.G. 1985. Michigan Flora. Part II: Dicots (Saururaceae — Cornaceae).
Cranbrook Institute of Science, Bloomington Hills, MI.

Voss, E.G. 1992. Michigan Flora. Part III: Dicots (Pyrolaceae — Compositaceae).
Cranbrook Institute of Science, Bloomington Hills, MI.

Wales, BA. 1972. Vegetation analysis of north and south edges in a mature oak-hickory
forest. Ecological Monographs 42: 451-471.

214

War, S.J., M. Kent, and K. Thompson. 1994. Seed bank composition and variability in
ﬁve woodlands in south-west England. Journal of Biogeography 21 : 151-168.

War, S.J., K. Thomson, and M. Kent. 1993. Seed banks as a neglected area of
biogeographic research: a review of literature and sampling techniques. Progress in
Physical Geography 1 7: 329-347.

Weathers, K.C., M.L. Cadenasso, and S.T.A. Pickett. 2001. Forest edges as nutrient and
pollutant concentrators: potential synergisms between fragmentation, forest canopies, and
the atmosphere. Conservation Biology 15: 1506-1514.

Webb, S.L., M. Dwyer, C.K. Kaunzinger, and PH. Wyckoff. 2000. The myth of the
resilient forest: case study of the invasive Norway maple (Acer platanoides). Rhodora
102: 332-354.

Whigham, DP. 2004. Ecology of woodland herbs in temperate deciduous forests.
Annual Review of Ecology, Evolution and Systematics 35: 583-621.

Whitney, G.G., and J .R. Runkle. 1981. Edge versus age effects in the development of a
beech-maple forest. Oikos 3 7: 377-381.

Williams, AB. 1936. The composition and dynamics of a beech-maple climax
community. Ecological Monographs 6: 317-408.

Williamson, GB. 1975. Pattern and seral composition in an old-grth beech-maple
forest. Ecology 56: 727-731.

Woods, K.D. 1993. Effects of invasion by Lonicera tatarica L. on herbs and tree
seedlings in 4 New-England Forests. American Midland Naturalist 130: 62-74.

Yates, E.D., D.F. Levia Jr., and CL. Williams. 2004. Recruitment of three non-native
invasive plants into a fragmented forest in southern Illinois. Forest Ecology and
Management 190: 119-130.

Young, G.I., and RH. Yahner. 2003. Distribution of, and microhabitat use by, woodland
salarnanders along forest-farmland edges. Canadian F ield-Naturalist 1 I 7: 19-24.

Ziegler, 8.8. 2004. Composition, structure, and disturbance history of old-growth and
second-growth forests in Adirondack Park, New York. Physical Geography 25 : 152-169.

215