IN FO R M A TIO N TO U SE R S
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.

U n ive rsity Microfilm s In t e r n a ti o n a l
A Bell & Howell In fo rm atio n C o m p a n y
3 0 0 N o rth Z e e b R o a d . A n n Arbor, Ml 4 8 1 0 6 - 1 3 4 6 U SA
313/761-4700
800/521-0600

O rder N u m b er 9314733

T h e h isto r ica l d ev elo p m en t o f b ig to o th a sp en -d o m in a ted forests
in n o rth ern low er M ich igan
Palik, Brian Josef, Ph.D.
Michigan State University, 1992

UMI

300 N. Zeeb Rd.
Ann Arbor, MI 48106

THE HISTORICAL DEVELOPMENT OF BIGTOOTH ASPEN-DO M I N A T E D
FORESTS IN NORTHERN LOWER M I C HIGAN
By
Brian Josef Palik

A DISSERTATION
Submitted to
M ichigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Forestry

1992

ABSTRACT
THE H I S T O R I C A L DEVELOPMENT OF BIGTOOTH A S P E N-DOMINATED
FORESTS IN NORTHERN LOWER MICHIGAN
By
Brian Josef Palik

Presettlement
development
dominated
this

forest

were

examined

landscapes

study

in

included:

composition;

popular

and
two

post-settlement
bigtooth

Michigan.

examining

ii)

assessing

for

influence

of

the

developmental

the

theory

aspen-

Objectives

regimes on presettlement

and

even-aged

within

northern
i)

prevailing disturbance
forest

composition

and current

applicability

in

bigtooth

of

aspen

forests.
Land

survey

presettlement

records

forest

were

used

composition.

to

Vegetation

reconstruct
sampling

on

replicate plots w i t h i n each landscape was used to determine
current
turn

composition

of

the

in

century

forests

logging.

that
Stem

established
analysis

following

was

used

to

reconstruct age distributions and growth histories of trees
in the current forest.
Presettlement

forest composition of the two

differed substantially.
sensitive
landscape

eastern
2 was

Landscape 1 was dominated by

hem l o c k

dominated

pine and white pine.

landscapes

by

and

American

fire-dependent

beech,
red pine,

firewhile
jack

Compositional differences may have been

related to differences in presettlement fire frequency.
current

overstories

of

both

landscapes

are

dominated

The
by

bigtooth

aspen,

convergence

red

was

oak,

and

attributed

red

to

maple.

similar

Compositional
post-settlement

disturbance regimes.
Tree age distributions in the current forests reflected
rapid

initial

cohort

establishment,

understory reinitiation.
forests,

bigtooth

turn

Within
were

aspen,

taller

populations

often

heights
no

was

an

highly

influence

determined
variation

on

within

environment

maple,

red

oak

and

with

to bigtooth

height

species

a

red

some

species,

tolerant

maple,

was

Age

p a t t e r n s . Growth

populations.

rate

patterns

to

heights
reaching

differences

Differences

contributed

species.

stem

individuals

aspen.

stratification

likely

and

a mid-tolerant species, which

red

variable,

equivalent

intolerant

than

of

exclusion,

Within the initial cohorts of both

typically taller than red oak,
in

stem

differences
and

in

height

had

height

competitive
growth

rate

variation within red oak and red maple populations.
The
plots,

duration
both

of

within

Understory

reinitiation

releases

in

establishment

stem

exclusion

and

between

often

overstory
occurred

resource availability,

two
w ith

to

an

on

all

landscapes.
radial

suggesting

response

possibly

similar

the

coincided

stems,
in

was

growth

that
increase

new
in

following a wave of natural

thinning within the o v e r s t o r y .
The
forests

results
follow

a

indicate
pattern

of

that

bigtooth

development

aspen-dominated

characteristic

many forests types that initiate after ma j o r disturbance.

of

This dissertation is dedicated to Richard Tracy of Stiles
Elementary, John Clark of Ithaca High School, and Professor
Richard Bowker of Alma College, for their dedication to
t eaching and learning.

iv

ACKNOWLEDGEMENTS

This study was every bit a collaborative effort between
myself and my advisor,
for his patent

Dr.

Kurt S.

and persistent

quality graduate program.
Peter

G.

members

Murphy,
of

my

Pregitzer.

efforts to provide me with

I also thank

Donald

I.

graduate

I thank him

Dickmann,

committee,

Drs.

and

for

Carl

Thomas

advise

a

W.

Ramm,

R.

Crow,

during

the

course of my program.
I

thank

the

organizations

for

presented herein.

following
their

additional

contributions

David Cleland,

Terri,

Dr.

Burton

Biological Station,
Brooks,
Lynn

Jessie

Hanninen,

Palik,

Leslie

Dr.

Barnes,
J.

B.

Chiang-Fung,
Elizabeth
Jaeger,

The

David

research

U n i v ersity

Hartson,
Swee

Lorraine
May

Collins,

Dr.

of Michigan

Michael Mueting,

Jacqmain,

Leslie L o e f f l e r , Joh n Vigneron,

the

Robert Vande Kopple,

Hart,

William

to

and

the H u r o n - Manistee National

Forest-Harrisville Ranger District,
James

people

Tang,

Ronald

Elaina
Grimm,
Josef

Hendrick,

The Department of Forestry,

Michigan State University.
I thank Andrew Burton,
my colleagues

in ecology,

and

necessary

laughs

Phu

Nguyen

assistance,

for

over

Ronald Hendrick,

and Zhijun Liu,

for providing helpful
the

providing

last
a

four

v

years.

listening

and countless free lunches.

discussions

ear,

I than

Dr.

computer

Finally,
Joanna

and

and Josef,

Elizabeth,

most
my

importantly,

sisters

Autu m n

I
and

thank
Dawn,

my

parents

and my

wife

for always believing what I did was important.

TABLE OF CONTENTS

LIST OF T A B L E S .................................................. 12
LIST OF F I G U R E S ................................................. 14
Chapter
I. INTRODUCTION
Development of forests following major d i s t u r b a n c e . ... 1
The development of Great Lakes aspen dominated
f o r e s t s . ..........
6
Study objectives and d e s i g n .............................. 8
O b j e c t i v e s ................................................8
Study l o c a t i o n s .......................................... 9
Vegetation s a m p l i n g .................................... 10
Destructive sampl i n g ................................... 10
Age d e t e r m i n a t i o n ................
11
Radial growth a n a l y s i s ...........
12
Presettlement forest composition and disturbance
r e g i m e s ...................................
12
Overview of c h a p t e r s . . ................................... 13
B i b l i o g r a p h y .............................................. 16
II. THE RELATIVE INFLUENCE OF ESTABLISHMENT TIME AND HEIGHT
GROWTH RATES ON SPECIES V E R TICAL STRATIFICATION DURING
SECONDARY FOREST SUCCESSION
A b s t r a c t ................................................... 22
I n t r o d u c t i o n ...............................................23
M e t h o d s ..........
25
Stand d e scription and vegetation s a m p l i n g ........... 25
Stem s a m p l i n g ........................................... 26
Stem a n a l y s i s ........................................... 27
Data a n a l y s i s . .................................
27
R e s u l t s .................................................... 29
Forest composition and s t r u c t u r e ..................... 29
Patterns of Establishment and Vertical
S t r a t i f i c a t i o n ........................................31
Intraspecific height growth r a t e s .................... 40
Interspecific height growth r a t e s .................... 44
D i s c u s s i o n ................................................. 49
Summary and c o n c l u s i o n s .................................. 56
B i b l i o g r a p h y ...............................................58

vii

III. A COMPARISON OF PRESETTLEMENT AND PRESENT-DAY FORESTS
ON TWO BIGTOOTH ASPEN-DOMINATED LANDSCAPES IN NORTHERN
LOWER MICHIGAN
A b s t r a c t ................................................... 61
I n t r o d u c t i o n ...............................................63
65
M e t h o d s ...............................
Study l o c a t i o n s ...................
65
Presettlement forest composition and disturbance
r e g i m e s ....................
.............. 68
Present-day vegetation s a m p l i n g ..........
70
R e s u l t s .................................................... 71
Presettlement forest c o m p o s i t i o n ..................... 71
Presettlement disturbance r e g i m e s .................... 71
Present-day overstory c o m p o s i t i o n .................... 73
Regeneration. .......................................... 76
D i s c u s s i o n ..................................
76
76
Presettlement f o r e s t s . .....
Present-day f o r e s t s . . .............
... 79
13rib1 3_ocjr ajpbiy.........................................84
IV.

PATTERNS OF EVEN-AGED FOREST DEVELOPMENT WITHIN TWO
BIGTOOTH ASPEN-DOMINATED LANDSCAPES IN NORTHERN LOWER
MICHIGAN
Al>)S 1 31*aCt . . . . . . . . as. . . . . . s . . . . . . . . . . . . . a.

iroiuof i o n . . .... ......................
....................
Study l o c a t i o n s
MOfllOClS . . . . . . . . . . . . . . . . . . . .a oa . a . . . . . .oo

Plot selection and vegetation sampling
Destructive s a m p l i n g ....... ...... .....
Age d e t e r m i n a t i o n ....... ............
Radial and height growth analysis.
White pine seed-source c h a r a c t e r i s t i c s . .
R e s u l t s ........ ...................... .
Forest composition and s t r u c t u r e . .
Population age s t r u c t u r e s ......
Bigtooth aspen heights and height growth rates.
Patterns of radial growth and understory
r e i n i t i a t i o n .....................................
White pine seed-source c h a r a c t e r i s t i c s .........
D l S C U S S l O n . e o . o . . o . a e e . a. . . . . . . . a . . . . . . . . . . . . . . . .

Summary and conclusions.
.............. .........
Bibl i o g r a p h y
........ .........................

.90
.93
.97
.98
.98
.99
100
102
104
106
106
106
119
120
131
133
141
144

V. THE VERTICAL DEVELOPMENT OF EARLY SUCCESSIONAL FORESTS
IN NORTHERN LOWER MICHIGAN
A b s t r a c t .................
149
Introduction.
........................................ 153
Study l o c a t i o n s ..............................
157
M e t h o d s ................
158
Stand and plot s e l e c t i o n ........
158
Stem s a m p l i n g .......................................... 159

viii

V.

(con'd).
Stem a n a l y s i s .......................................... 160
Data a n a l y s i s .......................................... 161
R e s u l t s ................................................... 164
R e g e n e r a t i v e - m o d e s.....................................164
Population and height s t r u c t u r e s .................... 166
Population height v a r i a t i o n ............................ 172
Individual height growth t r a j e c t o r i e s ............ ..183
D i s c u s s i o n ................................................ 189
Population height s t r u c t u r e s .................
189
A model for the development of d o m i n a n c e ........... 193
B i b l i o g r a p h y ...........................
200

VI. THE AGE AND HEIGHT STRUCTURE OF RED MAPLE (A cer rubrum
L . ) POPULATIONS IN NORTHERN LOWER M I C H I G A N BIGTOOTH
ASPEN (Populus qrandidenta Michx.) FORESTS
A b s t r a c t ..........................................
207
I n t r o d u c t i o n.....................
210
Study location and original forest c o m p o s i t i o n ...... 212
M e t h o d s ................................................... 213
Stand s e l e c t i o n ...............................
213
Vegetation s a m p l i n g ................................... 214
Stem s a m p l i n g ........
214
Stem a n a l y s i s ....................................... . .215
Data a n a l y s i s
...........................
216
R e s u l t s ....... ......................................... . .218
Forest c o m p o s i t i o n.....................................218
223
Red maple age s t r u c t u r e .......................
Red maple height s t r u c t u r e ...........................230
Red maple height growth p a t t e r n s ..............
233
Temporal patterns of overstory height i n c r e m e n t . ..239
D i s c u s s i o n ......
246
Origin of the red maple sprout c o h o r t ................ 246
Red maple age s t r u c t u r e.............................. 248
Understory reinitiation and overstory
d e v e l o p m e n t ..........................
249
Red maple height s t r u c t u r e ........................... 251
Landscape p a t t e r n s .................................... 254
Summary and c o n c l u s i o n s ........
255
B i b l i o g r a p h y ..............................................258

ix

LIST OF TABLES

Table 2.1.

Forest composition and structure by stratum...30

Table 2.2.
Mean time of establishment (number of years
after stand initiation) for three later successional tree
species and analysis of variance for species differences
in establishment t i m e .......................................3 6
Table 2.3.
Repeated measures analysis of variance for age
group differences in height-growth rates (m/yr) within
two later successional tree species. . .....
43
T able
2.4.
Repeated measures
analysis
of
species differences in height-growth rates

variance
for
(m/yr)...... 48

Table 2.5.
Summary information for second order polynomial
regressions of stem height as a function of stem age for
three
later
successional
tree
species
in
a Populus
.............. 50
qra n d i d e n t a t a - P . tremuloides forest.
Table
3.1.
Mean
climatic
variables
for
the
districts
con t aining the UMBS and Huron study a r e a s ............... 67
Table 3.2.
Percentage of bearing and line trees in the
pre settlement
forests
of
the
UMBS
and
Huron
study
4 21*e cis ..... . . . . . . . . . . . . . . . . . . . . . a . . . . . . . . . . . . . . . . . . .....72
Table 3.3.
Present-day forest composition and structure of
the UMBS and Huron study a r e a s ............................ 74
Table 4.1.
Present-day forest composition and structure
of
the H u r o n and UMBS study a r e a s ........................... 107
Table 4.2.
plots

Structural attributes
...........

of

the

Huron

and

UMBS
115

Table 5.1.
Regeneration methods of Ouercus rubra and Acer
rubrum genets in the a) Huron and b) UMBS f o r e s t s
165
Table 5.2. Repeated measure analysis of v a r iance and species
contrasts
for mean heights of Populus a r a n d i d e n t a t a .
Ouercus rubra and Acer rubrum over time in the Huron and
UMBS f o r e s t s ......
170

x

Table 5.3.
Numbers and density of Ouercus rubra and Acer
rubrum genets by height class in the a) Huron and b) UMBS
f o r e s t s ......................................................177
Table
5.4.
Repeated
measure
analysis
of
variance
and
specific contrasts for heights of Populus grandidentata
and dominant Ouercus rubra and Acer rubrum over time in
the Huron and UMBS f o r e s t s ............
181
Table 5.5.
Height growth patterns of Populus g r a n d i d e n t a t a r
Ouercus rubra and Ace r rubrum in the a) Huron and b) UMBS
f o r e s t s ........................................
185
Table 5.6.
Proportion of suppressed or released subdominant
Ouercus rubra and Acer rubrum in the a) Huron and b) UMBS
forests
that
had
height
growth
rates
equal
to
or
exceeding Populus grandidentata prior to a suppression or
subsequent to a release. ............
188
Table 6.1.
Overstory composition and structure of five
bigtooth
aspen-dominated
stands
within
the
HuronManistee National f o r e s t
............................ 219
Table
6.2. Age and site index of bigtooth aspen in five
stands within the Huron-Manistee National
Forest and
analysis of variance for comparisons among s t a n d s
220
Table
6.3. Sapling and seedling densities by species in
five bigtooth aspen-dominated stands within the HuronManistee National Forest.
......
221
Table 6.4.
Mean age and height of red maple sprout-origin
cohorts in five bigtooth aspen-dominated stands within
the
Huron-Manistee
National
Forest
and
analysis
of
variance for comparisons among s t a n d s ................... 226
Table 6.5.
M ean age and height of bigtooth aspen and
sprout-origin red maple on 20 plots
within the HuronManistee National
Forest
.......................... .227
Table 6.6.
Mean age of red maple
seedlings less than 1.5 m
t a l l , stratified by
0.5 m
height classes,
in
five
bigtooth aspen-dominated stands within the Huron-Manistee
National Forest.
................
229
Table 6.7.
a) Repeated measure analysis of variance and b)
specific contrasts for height increment across time in
bigtooth aspen and sprout-origin red m a p l e ............. 245

LIST OF FIGURES

Figure 2.1.
Vertical growth reconstructions of individual
Populus g r a n d i d e n t a t a . Ouercus r u b r a . Fraxinus americana
and Acer rubrum over a 42 year period in four plots. Each
line represents an individual s t e m ........................33
Figure 2.2.
The number of new individuals establishing as a
function
of
stand
age
for three
later
successional
species
in
a
Populus
q r andidentata- P .
tremuloides
f o r e s t .....................
35
Figure 2.3.
Vertical stratification patterns among four
tree
species
over
a
42
year
period
in
a
Populus
qrandid e n t a t a -P. tremuloides forest.
Heights are mean
values (+/- standard error) of four plots at each a g e . . 39
Figure 2.4.
Vertical growth rates (m/yr) fox* two age groups
over four height intervals within two later successional
tree species.
Old individuals established w i thin 1-21
years
after
stand
initiation.
YOUNG
individuals
established within 22-42 years after stand initiation.
Values are back-transformed means of inverse-square root
transformed growth rates
(m/yr), with 95% confidence
intervals. Sample size (number of plots) = 3 for Q. rubra
(O L D )
and
4
for
all
other
age
group-species
c o m b i n a t i o n s ................................
42
Figure 2.5.
Vertical growth rates (m/yr) for three later
successional tree species over eight h e ight intervals,
pooled across age groups. Values are back-transformed
means of inverse-square root transformed growth rates
(m/yr), with
95%
confidence
intervals.
Sample
size
(number of plots) = 3 for A. rubrum and 4 for Q. rubra
and F. a m e r i c a n a ............................................ 47

xii

Figure 2.6.
Second order polynomial regressions of stem
height
as
a
function
of
stem
age
for three
later
successional
tree
species
(A-C)
in
a
Populus
qra n d i d e n t a t a - P . tremuloides forest.
Dashed lines are
95% confidence bands for predicted values. Best-fit lines
for
all
three
species
are
compared
in
D.
Symbols
represent
number
of
observation
as
follows:
Ouercus
r u b r a . 1-15 years=3-30 observations (per s y m b o l ) , 16-29
years=l-15
observations;
Fraxinus
americana.
1-15
years=8-30 observations, 16-29 years=l-14 observations;
Acer r u b r u m , 1-15 years=2-15 observations, 16-30 years=l5 observation. ............................................. 52
Figure 4.1.
Age-height distributions for all sampled stems
on the 20 Huron plots.
Individuals > 1.5 m tall were
completely sampled on each plot. Individuals < 1.5 m tall
were
subsampled. Note that
each
plot contained
an
additional 15-35 bigtooth aspen ramets of similar age
as
the sampled stem.
Red oak and black oak are combined in
the f i g u r e .................
110
Figure 4.2.
Age-height distributions for all sampled stems
on the 16 UMBS plots.
Individuals > 1.5 m tall were
completely sampled on each plot. Individuals < 1 . 5 m tall
were
subsampled. Note that
each
plot contained
an
additional 15-35 bigtooth aspen ramets of similar age as
the sampled stem.
The red oak in the upper right corner
of plot U15 was actually a individual from plot U16 that
predated
the
remaining
individuals
on
the
plot
by
approximately 20 y e a r s .........
113
Figure 4.3.
Five-year height increment curves for the
sampled bigtooth aspen in each forest.
Values are means
± se.Sample size for
the Huron forest is 20. Sample
size for the UMBS forest is 16
(14 at 5-10 y e a r s ) ..... 122
Figure 4.4.
Ring-width indices for the Huron plots. Each
chronology is the mean for the sample size indicated in
the upper right corner of each graph.
The doted line at
a ring-width index of 1.0 is the standardized mean for
the entire chronology (see text for derivation of ringwid t h indices) .
The dashed line in each graph is the
cumulative
establishment
distribution
for
advanced
regeneration (stems > 1.5 m t a l l ) . The beginning of each
establishment distribution marks the start of understory
reinitiation (at least two stems establishing every five
year for a minimum of 20 y e a r s ; this is indicated by an U
for plots without advanced r e g e n e r a t i o n ) . The C in each
graph
marks
the
start
of
continuous
understory
establishment (at least one stem establishing every year,
or two every two years, for a minimum 10 y e a r s ) ....... 124

Figure 4.5.
Ring-width indices for the UMBS plots. Each
chronology is the mean for the sample size indicated in
the upper right corner of each graph.
The doted line at
a ring-width index of 1.0 is the standardized mean for
the entire chronology (see text for derivation of ringwidth indices) .
The dashed line in each graph is the
cumulative
establishment
distribution
for
advanced
regeneration (stems > 1.5 m tall).
The beginning of each
establishment distribution marks the start of understory
reinitiation (at least two stems establishing every five
year for a minimum of 20 years) .
The C in each graph
marks the start of continuous understory establishment
(at least one stem establishing every year, or two every
two years, for a min i m u m 10 y e a r s ) ................... ...129
Figure 5.1. Heights of Populus g r a n d i d e n t a t a . Ouercus r u b r a .
and Acer rubrum at 10 year stand age intervals in the
Huron forest (a) and the UMBS forest (b) . Huron values
are means
(±SE)
of 19 observations.
UMBS values are
backtransformed means (+95% confidence intervals) of 16
square-root transformed observations.
Q.
rubra and A.
rubrum values are the means of all individuals pooled by
169
species on each p l o t .................
Figure
5.2.
Height
growth
reconstructions
of
Populus
g r a n d i d e n t a t a . Ouercus rubra,
and Acer rubrum within
randomly selected plots from the Huron forest (a) and the
UMBS
forest
(b). Each line represents an individual
174
s t e m ............
Figure 5.3.
Heights of Populus grandidentata compared to
dominant Ouercus rubra or Acer rubrum at 10 year stand
age intervals in the Huron forest (a,b) and the UMBS
forest (c, d) .
Huron values are backtransformed means
(+95% confidence intervals)
of 16 log (y+1)
transformed
observations for Q. rubra, and 10 log (y+1)
transformed
observations for A. r u b r u m .UMBS values are
means (±SE)
of 13 and 3 observations
for Q. rubra and
A. r u b r u m .
r e s p e c t i v e l y ................................................179

xiv

Figure
5.4.
A
model
for
the
development
of
vertical
stratification among tree species differing in understory
tolerance and relative successional status in an evenaged forest, (a) Immediately following a stand initiating
disturbance,
initial heights of an intolerant,
earlysuccessional species (clear crown) and a relatively moretolerant, later successional species (filled crown) may
be
similar.
The
intolerant
species
is
numerically
dominant because of superior dispersal ability.
(b) With
the onset of crown differentiation, some individuals of
both species lapse into suppression.
A limited number of
individuals of the tolerant species may have early height
growth rates similar to the intolerant species, probably
because of locally reduced competitive pressure from the
intolerant species.
The initially low number of stems of
the tolerant species reduces the probability that many
will occur in competitively favorable neighborhoods, (c)
In the mature forest, the intolerant species is typically
vertically
dominant
over
the
more-tolerant
species
although
some
stems
of
the
tolerant
species
have
maintained growth rates similar to that of the intolerant
species.
The population-level height growth rate of the
intolerant species will be g reater than that of the
tolerant species because all suppressed stems of the
former have died, leaving only fast-growing s t e m s , while
slower
growing
stems
of
the
t o l erant
species
have
survived
..........
196
Figure
6.1.
The
number
of
new
red
maple
individuals
establishing as a function of stand age in five bigtooth
aspen-dominated
forests
within
the
Huron-Manistee
National Forest.
........
225
Figure 6.2.
Height growth reconstructions of individual
bigtooth aspen and sprout-origin red maple within one
randomly selected plot from each of five bigtooth aspen
dominated-stands
w i thin
the
H uron-Manistee
National
Forest.
Each line represents an individual s t e m ...... 232
Figure 6.3.
The relationship between stem age (years) and
total height
(meters)
for individual
seed-origin red
maple in five bigtooth aspen-dominated stands w i thin the
Huron-Manistee National Forest.
One randomly selected
plot from each of the five stands is shown.
In each
plot,
height
classes
above
the
dotted
line
were
completely sampled.
Height classes below the dotted line
were randomly s u b s a m p l e d . . . ...............................235

xv

Figure 6.4.
The relationship between height
increment
(meters) over the five-year period immediately prior to
sampling and initial height (meters) at the beginning of
this period for individual seedling-origin red maple in
five bigtooth aspen-dominated stands w i t h i n the HuronManistee National Forest.
One randomly selected plot
from each of the five stands is s h o w n ................... 238
Figure 6.5.
The relationship between stem age (years) and
height growth rate (m/yr) over the first 1.5 m for the
fastest
growing
seedling-origin
red
maple
in
five
bigtooth aspen-dominated stands within the Huron-Manistee
National Forest.
For each stand, the fastest growing
individuals from consecutive five-year stem-age classes
(i.e. 5-10 years old, 10-15 years old, etc.) with i n each
of four plots were pooled for a n a l y s i s .................. 241
Figure 6.6.
Height increments (m) over consecutive fiveyear stand-age classes, beginning 10 years after stand
initiation,
for bigtooth aspen and sprout-origin red
maple.
Values
are
back-transformed
means
(+
95%
confidence intervals) of 20, 18 and 19 log-transformed
observations for bigtooth aspen, the fastest growing red
maple on each plot and the median growing red maple on
each plot, respectively. In each graph, the bott o m point
of the inverted triangle indicates the y ear after stand
initiation (mean of five stands) by w h i c h 10% of the
seed-origin red maple > 1.5 m tall had established. The
single open triangle at the right end of each graph is
the mean (+ 95% confidence intervals) height increment of
the fastest growing seedling-origin red maple over a
stand age of 65-70 y e a r s .................................. 244

xv i

Chapter 1

INTRODUCTION

Development of forests following major disturbance
Studies
that

in various

large-scale,

common

1973;
Peet

stand-replacing

features

Mutch 1973;

of

many

1976;

Romme

Veblen

and

natural

forested

Heinselman 1973;

Sprugel
1981;

forest ecosystems have demonstrated

landscapes

Kilgore
and

Knight

1973;

Ashton

1981;

Oliver et a l . 1985; Whitney 1986).

disturbances
(Habeck

analogy

to

large-scale

1978;

O l iver

1981;

Loucks

1984;

While these works

focus

Canham

and

anthropogenic

such as clearcutting and fire.
literature

on

stand

Day

1972;

Bormann

Oliver
and

1978,

Likens

Indeed,

development

anthropogenic disturbance

1983;

1986;

Larson

and Hodges

Guldin

following

Stubblefield

1979;

1986;

1988;

Lorimer 1990,

and
Peet

Wierman

Lorimer
and

large-scale

.1985;

and O l iver

and

O l iver

1981;
1978;
1979;

1981; Carleton 1982;
Harcombe

Christensen

Foster 1988;

1986;

1987;

Scheiner et al.

Kelty

Clatterbuck

1988;

Hix and

1991; Oliver and Larson 1990; Deal et al 1991;

Segura and Snook 1992).
suggest

disturbances,

there is an abundant

Christensen and Peet 1981; Oliver 1980,
Hibbs

there is a

(Cayford 1957; Mar q u i s 1967,

1980;

and

Rowe and Scotter

on forest development after natural disturbances,
clear

are

that

characteristics

certain
are

Results from many of these studies
structural
common

1

to

and
a

wide

developmental
range

of

2

forest
1980,

i)

types

originating

after

major

disturbance

(Oliver

1981; Peet 1981; Peet and Christensen 1987):

The

overstories

of

forests

disturbance are often even-aged.

initiating

after

major

An initial cohort of trees

establishes within a relatively short period of time,

often

spanning less than 20% of the pathological rotation age for
the dominant tree species in the

forest (Smith 1986; Oliver

1981; Oliver and Larson 1990) .

ii)
within

In

mixed-species

initial

forests,

overstory

the

cohorts

crowns
are

of

individuals

often

stratified,

following an order paralleling relative understory tolerance
and

successional

status,

successional

species

are

successional

species

(Oliver

1978;

Bormann

and

Likens

i.e.,

taller

than

1978;

1979;

intolerant,

early

m o r e - t o l e r a n t , latter

Stubblefield
Hibbs 1983;

and

Oliver

Kelty

1986;

Larson 1986; Hix and Lorimer 1990; Oliver and Larson 1990).

iii)

Af t e r

during which
establish

initial

cohort e s t a b l i s h m e n t , a period

few or no new individuals

or

survive.

This

period

of

of any tree
low

or the thinning phase of stand development

Christensen

1987),

dens ity-dependent
individuals
trees.

within

since

it

ischaracterized by

mortality
the

initial,

of

species

establishment

and/or high mortality has been termed stem exclusion
1981),

ensues

suppressed

post-disturbance

(Oliver

(Peet and

intense
overstory
cohort

of

3

iv)

Following

species

stem

begin

understory.

to

exclusion,
establish

This

reinitiation
development

event

and

individuals

and

survive

has

(Oliver 1981),
(Peet

new

been

of

in

tolerant

the

termed

forest

understory

or the transition phase of stand

Christensen

1987),

since

the

cohorts

that begin establishing at this time promote the transition
to an uneven-age structure in the forest.

There
and

is

little

repeatability

information

of

these

concerning

the

developmental

features.

studies that have documented the structural
outlined

above

replicate

plots

and Richardson
or

have

have

been based

(Oliver

1978;

1985;

Kelty

data

from

pooled

on

Oliver

1986;

replicate

analysis

(Bloomberg

Oliver

Oliver

a l . 1985;

a l . 1989;

Segura

inappropriate
repeatability

and
for

of

et

Snook

one

a l . 1985;

or

samples

into

few

Roberts

and Nowacki
a

1992),
single

Stubblefield

Harcombe

assessing

Most

characteristics
from

1950;

1992).

developmental

et

Abrams

forest-level
1978;

results

generality

1986;

and

Sharik

et

These

methodologies

are

the

variability

and

patterns

landscapes characterized by homogeneous

within

and

among

site conditions and

disturbance h i s t o r i e s . The need for assessing the generality
of developmental

patterns

some studies have
originating

after

is

demonstrated by

the

fact

that

found no stem exclusion period in forests
ma j o r

disturbance

1982; Roberts and Richardson 1985),

(Peet

1981;

Carleton

while others have

found

no stratification among species within the initial cohort of

4

trees

(Guldin and Lorimer 1985).
In

addition

developmental
concerning
events.

patterns,

the

from

Nesbitt

and Marks

is

that

also

i.e.,

Marks

1975;

silvicultural

studies

competitive

interactions
intolerant

physiological

in

relatively

lack

have

height

(Loach

Bormann

and

1991).

may

species

e n v i r o n m e n t s , the

species

tolerant

species

1978;

Wierman

Smith

1986;

have

been

(Cayford

and

Oliver

Oliver,

of

consensus
structural

grow t h

be

may

equally

important.

them

outgrow

equal

rates

1957;

Stubblefield

and

to

or

H o w e v e r , in n o n ­

to

Clatterbuck

Canham

m o rphological

growth

Guldin

and

interspecific

found

1979;

Drury

1979;

that

allow

height

197 0;

grow

On the other hand,

possess

that

rates

species

Likens

demonstrated

characteristics

tolerant

of

various

more-tolerant species when in competition.
competitive

variability

intolerant

species

1985; Kozlowski et al.

Relatively

a

control

differences

more-tolerant

1973;

regarding

overstory stratification is thought to

inherent

than

there

mechanisms

species,

faster

uncertainty

For example,

result
among

to

those

and

of
of

and

Lorimer

Burkhardt

and Larson 1990; D e a l , Oliver and Bormann 1991;

1990;

some
less-

Oliver
1985;
Oliver

S h a insky and

Radosevich 1992).
There is also little consensus regarding the m e c h anisms
that

control

even-aged
development

stem

exclusion

forests.
suggest

of competitive

and understory

Resource-based
that stem

pressure

r e i n i tiation

theories

exclusion

from overstory

occurs
stems

of
as

in

forest
a

(Oliver

result
1981;

5

Christensen and
1987) .

Given

Peet
this

1981;

Peet

1981;

assumption,

it

Peet and

follows

Christensen

that

understory

reinitiation should be triggered by an increase in light or
soil

resource

development.

availability

mature overstory
1981;

the

course

factors

including,

individuals

Peet

1981;

canopy gap

(Sprugel

Peet

and

1976;

Christensen and

Christensen

1979,

Oliver

et

a l . 1985;

Oliver

reductions in rates of root growth

1987),

concentrated

Fried

et

al

wave

1988;

resourse-based

of

natural

Oliver

and

theories,

(Bormann and
Larson

individuals

thinning

Larson

some

and

crown

1990),

(Oliver and Larson 1990),

and mortality of suppressed overstory
a

stand

formation by

differentiation within the even-aged overstory
Likens

of

The cause of this increase has been attributed

to a variety of

Peet

during

following

(Harcombe

1990).

studies

In

1986;

contrast

suggests

to

that

understory reinitiation may simply reflect the reproductive
maturation

of

successional

a

trees

post-disturbance
(Day

1972;

cohort

Carleton

1982;

Richardson 1985; Sharik et al 1989; Sakai et al.
1990).
This

Prior to this time,

vie w

understory

implies
is

establishment
growth

may

increase.

be

not
and

that
so

minimal

later

Roberts

and

1985; Sakai

seeds are simply not available.

resource
severe

survival

of

of

until

limitation
enough
tolerant

resource

that

in

the
it

species,

forest

prevents
however

availability

does

6

The development of Great Lakes Aspen-dominated forests
Forest

dominated

cxrandidentata

Michx.)

M i c h x .),

alone

hectares

in

1990).

or

the

and

in

bigtooth

trembling

aspen

combination,

Great

Lakes

aspen
(P.

occupy

region

(Populus
tremuloides

over

(Einspahr

5

million

and

Wyckoff

Despite aspen's areal e x t e n t , and its importance as

wildlife habitat
and a source of
Blyth

by

and

Smith

(Brelich
fiber

et

a l . 1972; Gullion

(Hughes and Brodie

1988),

relatively

1977,

1972;

little

is

1984)

Keays

1972;

known

about

patterns of stand development in the covertype.
Aspen's

areal

extent

in

the

Great

Lakes

region

directly attributable to post-settlement changes
fire

regimes.

Rapid

fueled wildfires
favored aspen
to

establish

clonal
Cramer

growth

combinations
A i t . , P.
are

No

bare

to

studies

forest
disturbance

mineral

conifers

soil

have

by

aspen

composition and
history

and development.

(Pinus

repeated
20th

on

spread

192 3;

slashedcenturies

many

(Kilburn

the

of

L. ,

present-day

the

1957;

influence

structure

readily

to settlement,

strobus

or
aspen

ability

Benninghoff

Tsucra canadensis

dominated

assessed

and

Prior

Lam. , and

have

occupied

in natural

sites because of the taxon's

after fire (Gleason

banksiana

and

late 19th and early

Whitney 1987).
of

believed

currently

in the

on many
on

1963;

deforestation

P.

via
and

various
resinosa

(L.)

Carr.)

drier

sites

W hitney

of

is

1987).

presettlement

post-settlement
forest

structure

7

An

even-aged

reasonable
taxon

structure

assumption,

occupies

among

given

soil

the

space

aspen

rapidity

ramets
with

following

a

is

which

a
the

disturbance

(Stoeckeler and Macon 1956; Buell and Buell 1959; Zahner and
Crawford

1965;

Kemperman

and

Barnes
Barnes

1966,

1976;

1969;

Graham

Scheiner et

et

al.

a l . 1988).

1963;

However,

little is known about the age structure of associated,

later

successional

many

stems

of

origin,

species.

latter

successional

havi n g

wildfires

Limited

resprouted

evidence
hardwoods

from

suggests
are

root

of

that

vegetative-

systems

surviving

(Graham et a l . 1963; Roberts and Richardson 1985).

Such forests would be even-aged in a strict sense,

however,

the generality of this developmental pattern is not k n o w n .
Limited studies of overstory canopy structure
forests

suggests

successional

that

species

aspen

(Cayford

does

stratify

1957;

Graham

et

in these

over

latter

al.

1963),

however no information on species age structures were given
in these

particular

contrasting

aspen

studies.

site

Results

classes

from

suggest

that

a

study

stem

in

two

exclusion

occurs on productive aspen sites, but not on less-productive
sites

(Roberts

and

Richardson

1985).

The

differences

were

attributed to the ability of rapidly growing overstory trees
to

preempt

tree

vigor

preempt
assess

resources
and

density

resources.
the

on

good

sites,

are

while

insufficient

There have been no

g e n e rality

of

aspen or any forest-type.

these

on
to

studies

developmental

poor

sites,

completely
designed

to

patterns

in

8
STUDY OBJECTIVES AND DESIGN
Objectives
The current
uncertainties

study was

associated

designed to address
with

stand

some of the

development

patterns

following major disturbance and aspen forest development

in

particular.

The general objectives of the study were as follows:
i)

to

composition
origin,

assess
and

the

post-settlement

development,

bigtooth

influence

original

disturbance

and potential

aspen-dominated

of

forests

forest

history

on

the

successional pathways of

within

several

landscapes

in northern Lower Michigan?
i i ) to
bigtooth

determine

if

species

aspen-dominated

age

forests

distributions

reflect

an

within

even-aged

pattern of development characterized by rapid initial cohort
e s t a b l i s h m e n t , stem exclusion,
iii)
aspen

to determine

over

overstory
forests

later
cohort

and,

if

and understory reinitiation;

if height stratification of bigtooth

successional
of

so,

trees
to

was

assess

tree

species

within

characteristic
the

influence

of

of

the
these

time

of

establishment and height growth rates on the development of
species stratification;
iv)

to

gain

insight

into

the

possible

changing resource and seed availability on the
stem

exclusion

forests ?

and

understory

reinitiation

influence

of

induction of
in

even-aged

9

v)

to

developmental

quantify

patterns

spatial

w i thin

and

variability

bet w e e n

of

bigto o t h

forest

aspen-

dominated landscapes characterized by relatively homogeneous
physical site conditions and disturbance histories.

Study locations
The majority of the research reported on in this study
was

conducted

bigtooth

in

replicate

aspen-dominated

stands

located

landscapes

in

within

northern

two
Lower

Michigan.

Site 1 was a 15 k m 2 landscape located within the

University

of

northwestern

Michigan

portion

of

Biological
Lower

Station

M i c higan

in

the

(Cheboygan

extreme
County).

Site 2 was a 18 k m 2 landscape located w i t h i n the Harrisville
Ranger District
in

(Oscoda County)

northeastern

indicated

that

Lower

the

of the Hu r o n National Forest

Michigan.

current

Initial

overstories

of

observations

both

landscapes

were dominated by bigtooth aspen, with lesser amounts of red
oak

(Ouercus rubra L . ) and red maple

(A c e r rubrum L . ).

Both

landscapes were characterized by similar surficial geology,
soil

type,

and

postsettlement

disturbance

histories

(see

chapter 3 for d e t a i l s ) .
In addition to the two primary research sites,
study

was

located
This

conducted

in

pilot

within

north-central
project

was

a

single

Lower

aspen-dominated

Michigan

designed

a pilot

to

(Clare

assess

stand

County).

vegetation

sampling schemes and to refine destructive sampling and stem
analysis techniques,

before initiating the pri m a r y study.

10

Vegetation sampling
Arboreal

vegetation

was

sampled

in

located circular plots within each stand.
from 250 m 2 to 475 m 2 .
(at 1.37 m)
saplings

of all

In each plot,

density of tree seedlings
1-m2

frames.

Seedling

species and diameter
(dbh > 10 cm)

were tallied.

(dbh < 2.5 cm)

frames

were

plot

boundary.

individuals

The

were tallied in 12

spaced

at

regenerative-origin

(sprout or seedling)

and

Species and

3-m

along two perpendicular diameters of each plot,
the

randomly

Plot size ranged

overstory individuals

(2.5 cm < dbh < 10 cm)

four

intervals

beginning at
of

overstory

was assessed when possible.

Destructive sampling
On each plot,
all

individuals

one randomly selected bigtooth aspen,

>

1.5

destructively

sampled

establishment,

rates

growth

of

selected

aspen,

only

the

stemmed genets.

m

tall

to

of

determine

stems.

tallest

For

were

For bigtooth aspen,

codominant

crown

dominant

or

codominant)

was

internal

each plot.

total

Sampling

decay)

of

radial

bigtooth

from

multi­

one individual from the

randomly

intensity

tha n

sampled

were

times

rates of

other

classes

and

species,

heights,

and

species

ramets

or

with

tree

of height growth,

dominant

ramets

other

and

(most

ramets

selected

destructively

were

(rejecting
sampled

for bigtooth aspen was

on

much

lower than for other species because little height variation
was

observed

and

little

among

age

dominant-codominant

variation

was

expected.

ramets
On

w i thin
each

plots

plot,

a

11

subsample

of

individuals

<

1.5

m

tall

was

sampled

stratifying on species and 0.5 m height classes.

by

Stems < 1.5

m tall were sampled primarily to d etermine the age range of
these stems on each plot, particularly their m a x i m u m ages.
Stems were
on

size)

were

and

cut

felled at

total

at

those

additional

section

individuals

base

continued

to

level

was

or g r ound

were

greater

(dbh >

Additionally,
the

heights

ground

(generally

0.5 m

determined.

from
than

all
m
m

(depending

Stem

larger

2.5

removed at 1.4

10 cm)

level

sections

individuals

in height).
from

An

all overstory

for use in radial g r owth analysis.

all stems were marked at 0. 25-m intervals from
2

to

m

and

at 1-m

the

1-m

multiple

dominant leader.

intervals

thereafter.

closest

to

the

Marking

end

of

the

Stem sections were cut at each measurement

interval up to a 3 cm diameter top. All d e structive sampling
was completed during summer and fall of 1990.

Age determination
Small stems of deciduous tree species
<

2.5

m

tall)

and

the

terminal

leaders

(generally those

of

b i g tooth

aspen

were aged by counting the number of terminal bud scale scars
preceding each

height

species

pine

branch
were

(white
whorls.

sanded

examination.

to

and red

Basal
a
Ages

interval.

stem

smooth

pine)

Small stems
were

sections
surface and

of all

sections

of

aged

of

coniferous

by

larger

counting

individuals

wetted

to aid

were

determined

counting ring number on at least two radii

ring
by

(typically a long

12

and short

axis

microscope.

on

oblique-shaped stems)

Terminal bud scale counts,

under a dissecting
branch whorl counts,

and ages of basal stem sections were used to determine time
of

establishment

establishment
plot.

for

all

histories

The additional

cumulative

height

sampled
of

stems

surviving

and

to

reconstruct

individuals

in

each

section ages were used to reconstruct

growth

curves

and

periodic

height

increment curves for each sampled stem.

Radial growth analysis
Annual
selected
1.4

m

radii

stem

widths

ring

widths

measured

(rejecting decayed

sections

wer e

were

of

measured

dissecting microscope

overstory
to

the

or

along

injured

stems

randomly

areas)

(dbh

nearest

one

>10

0.1

cm).

mm

and ocular micrometer.

on

the
Ring

using

a

Radial growth

measurements were used to detect periods of suppression and
release

p o s sibly

indicative

of

changing

resource

availability over the course of stand development.

Presettlement forest composition and disturbance regimes
Pre-settlement
two
Land

primary

study

Office

Lorimer

(1977,

forests
areas

(GLO)
1980),

and

were

survey

disturbance

regimes

reconstructed

using

records

and Whitney

following

(1986).

of

the

General

methods

of

13

OVERVIEW OF CHAPTERS
Chapter 2.
THE RELATIVE INFLUENCE OF ESTABLISHMENT TIME A N D
GROWTH RATES
ON
SPECIES VERTICAL STRATIFICATION
SECONDARY FOREST SUCCESSION

HEIGHT
DURING

In Chapter 2, the results from the pilot study on aspen
forest development are presented and discussed. The research
was conducted in a single aspen-dominated stand occurring on
a site

formerly occupied by a northern hardwood

Originally,

the

project

was

designed

to

assess

community.
and

refine

the stem analysis methodologies that were to be u sed in the
primary project.
to

be

The

interesting

results

in

from this

themselves

pilot

since

the

study

proved

developmental

pattern for this single stand was found to differ from those
previously reported for other stands originating after major
disturbance,

and from those of the primary study.

Chapter 3.
A COMPARISON OF PRESETTLEMENT AND PRESENT-DAY FORESTS
TWO BIGTOOTH ASPEN-DOMINATED LANDSCAPES IN NORTHERN
MICHIGAN
In

Chapter

composition
landscapes
addressed
relating

of

the

are
in

3,

presettlement
two

compared.

this

chapter

presettlement

primary

and

bigtooth

Specific
include:

forest

present-day

landscape;

and
and

present-day

composition

(H i )

forest

assessing

that

reconstructing
to

natural disturbance regime of each landscape;
presettlement

forest

aspen-dominated

objectives
(i.)

ON
LOWER

the

and

potential

(ii) comparing

composition
the

are

of

influence

each
of

postsettlement disturbance history on the development of the

14

current

forests

and

the

potential

successional

pathways

of

the two landscapes.

Chapter 4.
PATTERNS OF E V EN-AGED FOREST DEVELOPMENT WITHIN TWO BIGTOOTH
ASPEN- D O M I N A T E D LANDSCAPES IN NORTHERN LOWER MICHIGAN
Chapter

4

reports

on

the

generality

of

establishment

patterns with i n and between the two primary study landscapes
and

provides

controlling
forests

insight

stem

into

exclusion

initiating

the

and

after

possible

understory

major

mechanisms

reinitiation

disturbance.

objectives addressed in this chapter include

in

Specific

(i) quantifying

the va r i a b i l i t y of establishment patterns with i n and between
the

two

study

landscapes

and

(ii)

exploring

relationships

betwee n the timing of understory reinitiation and changes in
resource

and

seed

availability

during

the

course

of

stand

development.

Chapter 5.
THE V E R T I C A L DEVELOPMENT
NORTHE R N LOWER M I C HIGAN
Chapter

5

OF

examines

EARLY

height

SUCCESSIONAL

FORESTS

structures

of

populations with i n the post-disturbance cohort trees
two pri m a r y study landscapes.
in this

chapter

include:

(i)

IN

tree
in the

Specific objectives addressed
determining

if species

height

stratification paralleling understory tolerance and relative
successional
(ii)

status

assessing

species

are

characteristic

the degree

populations;

(i i i )

of

of variability
determining

if

these

forests;

of heights
age

within

differences,

15

regenerative-mode
individuals

(sprout

competitive

or

environment

population height structures;
to which
from

height structure

that

expected

differences

in

seedling),

in the

under

height

and

a

had

(iv)

growth

rates

changes

any

in

an

influence

on

assessing the degree

forests
model

or

examined

based

on

among

deviated
inherent

species

of

differing understory tolerance and successional status.

Chapter 6.
THE AGE AND HEIGHT STRUCTURE OF RED MAPLE (Ace r rubrum L. )
POPULATIONS
IN
NORTHERN
LOWER
MICHIGAN
BIGTOOTH
ASPEN
(Populus qrandidenta Michx.) FORESTS
Chapter 6 presents an autecological examination of red
maple in the bigtooth aspen forests of the Huron study area.
Specific

objectives

addressed

in

this

chapter

include:

i)

quantifying the compositional importance of red maple in the
forests

of

the

study

height

structures

area;

iii)

of

a r e a ; i i ) characterizing
red

examining

estab l i s h m e n t , height,

maple

populations

relationships

the

in

between

and height growth rates

age

the

and

study

time

of

in overstory

and understory red maple p o p u l a t i o n s ; and i v ) characterizing
the degree of variability in age and height structure among
red maple populations within a local landscape.

BIBLIOGRAPHY
Abrams, M. D. and G. J. Nowacki. 1992.
Historical variation
in fire, oak recruitment, and post - l o g g i n g accelerated
succession in central Pennsylvania.
Bull. Torrey Bot.
Club 119: 19-28.
Barnes, B. V.
1966.
The clonal
aspens.
Ecology 47: 439-447.

growth habit

of

American

________.
1969.
Natural variation and delin e a t i o n of clones
of Populus tremuloides and P. grandidentata in northern
Lower Michigan.
Silvae. Genet. 18: 130-142.
Benninghoff,
W.
S.
and
K.
J.
Cramer.
1963.
Phytosociological
analysis
of
aspen
communities
on
three site classes for Populus grandidentata in western
Cheboygan C o u n t y , MI.
V egetatio 11: 253-264.
Bloomberg, W.
157-161.

J.

1950.

Fire and spruce.

For.

C h r o n . 26:

Blyth, J. E. and W. B. Smith.
1988.
Pulpwood p r o d uction in
the north-central region by county, 1986. U . S .D.A. For.
S e r . Res. B u l l . N C - 1 0 6 .
B o r m a n n , F. H. and G. E. Likens.
1979.
Pattern and process
in a forested ecosystem.
S p r i n g e r - V e r l a g , New York.
Brelich,
J . D . , J.
L.
Cook
and R.
I . Blouch.
1972.
Management for deer.
In Aspen: symposium proceedings.
U.S.D.A. For. Ser. Gen. Tech. Rep. N C - 1 . pp. 120-125.
B u e l l , M. F.
prairie.

and H. F. B u e l l . 1959. A s p e n invasion
B u l l . Torrey Bot. Club 86: 264-265.

of

Canham, C. D. and 0. L. L o u c k s . 1984. Catastrophic windthrow
in the presettlement forests of Wisconsin.
Ecology 65:
803-809.
C a n h a m , C. D. and P. L. M a r k s . 1985.
The response of woody
plants to disturbance: patterns of establishment and
growth.
In The ecology of natural disturbance and
patch dynamics.
Edited by S. T. A. Pickett and P. S.
White. Academic Press, Orlando,
pp. 197-216.
Carleton,
T.
J.
1982.
The
p attern
of
invasion
and
establishment of Picea mariana (Mill.) BSP. into the
subcanopy layers of Pinus banksiana Lamb, dominated
stands.
Can. J. For. Res. 12: 973-984.

16

17
Cayford, J. H.
1957.
Influence of the aspen overstory on
white spruce growth in Saskatchewan.
Canada Department
of Northern Affairs and National Resources Forestry
Branch, For. Res. Div. Tech. Note 58.
Christensen, N. L. and R. K. P e e t . 1981.
Secondary forest
succession on the North Carolina Piedmont. In Forest
succession: concepts and application.
Edited by D. C.
W e s t . , H. H. S h u g a r t , and D. B. Botkin.
SpringerVerlag, New York.
pp. 230-245.
Clatterbuck, W. K. and J. D. Hodges.
1988.
Development of
cherrybark
oak
and
sweetgum
in
mixed,
even-aged
bottomland stands in central Mississippi, U.S.A. Can.
J. For. Res. 18: 12-18.
Day,

R. J.
1972.
Stand s t r u c t u r e , succession, and use of
southern Alberta's Rocky Mountain forest.
Ecology 53:
473-478.

D e a l , R.
L . , C.
D.
Oliver
Reconstruction
of
mixed
coastal southeast Alaska.
654.
D r u r y , W. H. and
Ar nold Arbor.

and
B.
T.
Bormann.
1991.
hemlock-spruce
stands
in
Can. J. For. R e s . 21: 643-

I . C. T. N i s b e t .
54: 331-368.

1973.

Einspahr, D. W. and G. W. W y c k o f f . 1990.
aspen:
timber
supply,
utilization,
North. J. A p p l . For. 7: 168-171.

Succession.

J.

North American
and
research.

Foster,
D.
R.
1988.
Disturbance
history,
community
organization and vegetation dynamics of the old-growth
Pisgah Forest, south-western New Hampshire, U.S.A. J.
E c o l . 76: 105-134.
Fried,
J. S., J. C. Tappeiner and D. E. Hibbs.
1988.
Bigleaf maple seedling establishment and early growth
in Douglas-fir forests.
Can. J. For. Res. 18: 12261233.
Gleason, H. A.
Michigan.

1923.
Botanical observations in
J. New York Bot. Club 24: 273-283.

northern

Graham, S. A. , R. P. Harrison, Jr. and C. E. W e s t e l l , Jr.
1963.
Aspens: phoenix trees of the Great Lakes region.
University of Michigan Press, Ann A r b o r .
Guldin, J. M. and C. G. L o r i m e r . 1985. Crown differentiation
in even-aged northern hardwood forests of the Great
Lakes region, U.S.A.
For. Ecol. Manage. 10: 65-86.

18
Gullion, G. W.
1977.
Maintenance of the aspen ecosystem as
a primary wildlife habitat. In Proceeding of the Xlllth
International Congress of Game Biologist.
Edited bv T.
J. Peterle.
Wildlife Management Institute, Washington
D.C. 538 p.
________ .1984.
Managing northern forest for wildlife.
Ruffed Grouse Society.

The

Habeck, J. R. and R. W. M u t c h . 1973.
Fire-dependent forests
in the northern Rocky Mountains.
Q u a t . Res. 3: 408424.
Harcombe, P. A.
1986.
Stand development in a 130-year-old
spruce-hemlock forest based on age structure and 50
years of mortality data.
For. E c o l . Manage. 14: 41-58.
Heinselman, M. L.
1973.
Fire in the virgin
Boundary Waters Canoe Area, Minnesota.
329-382.

forests of the
Quat. Res. 3:

Hibbs, D. E.
1983.
Forty years of forest succession
central New England.
Ecology 64: 1394-1401.
Hix,

in

D. M . , and C. G. L o r i m e r . 1990. Growth-competition
relationships
in
young
hardwood
stands
on
two
contrasting sites in southern Wisconsin.
For. S c i . 36:
1032-1049.

________.
1991.
Early stand development on two former oak
sites of southwestern Wisconsin. For. E c o l . Manage. 42:
169-193.
H u g e s , J. M. and J. D. Brodie.
1972. Selected economic
aspects
of
management.
In
in
Aspen:
symposium
proceeding.
U.S.D.A. For. Ser. Gen. Tech. Rep. N C - 1 .
pp. 27-34.
Keays, E. A.
1972.
The resource and its potential in North
America.
In Aspen: symposium proceedings. U.S.D.A.
For. Ser. Gen. Tech. Rep. N C - 1 . pp. 4-9.
Kelty, M. J.
1986.
Development patterns in two hemlockhardwood stands in southern New England.
Can. J. For.
Res. 16: 885-891.
K e m p e r m a n , J. A. and B. V. B a r n e s . 1976. Clone
American a s p e n s . Can. J. Bot. 54: 2603-2607.

size

in

Kilburn, P. D.
1957.
Historical development and structure
of the aspen, jack pine, and oak vegetation types on
sandy
soils
in
northern
lower
Michigan.
Ph. D.
dissertation, University of Michigan, A n n Arbor. 267 p.

19
Kilgore, B. M.
1973. The ecological role of fire in Sierran
conifer f o r e s t s . Quat. R e s . 3: 496-513.
Kozlowski, T. T . , P. J. Kramer, and S. G. Pallardy. 1991.
The Physiological Ecology of Woody Plants. Academic
Press, San Diego.
Larson, B. C.
1986.
Development and growth of even-aged
stands of Douglas-fir and grand fir.
Can. J. For. Res.
16: 367-372.
Loach, K.
Growth
shade.

1970.
Shade tolerance in tree seedlings I I .
analysis
of plants
raised
under artificial
New P h y t o l . 69: 273-286.

L o r i m e r , C. G.
1977.
The presettlement forest and natural
disturbance cycle of northeastern Maine. Ecology 58:
139-148.
________ . 1980. The use of land survey records in estimating
presettlement fire frequency.
In Proceedings of the
fire history workshop.
Edited by M. A. Stokes and J.
H.
Dieterich.
20-24 October 1980,
Tucson,
Arizona.
U.S.D.A. For. Ser. Gen. Tech. R e p . , R M - 8 1 . pp. 57-62.
M a r k s , P. L.
1975. On the relation between extension growth
and successional status of deciduous trees of the
northeastern United States. B u l l . Torrey Bot. Club 102:
172-177.
Marquis, D. A.
1967.
Clearcutting in northern h a r d w o o d s :
results after 30 years.
U.S.D.A. For. Ser. Res. Pap.
NE-85.
________ .
1981.
Removal
or retention of unmerchantable
saplings in Allegheny hardwoods: effect on regeneration
after c l e a r c u t t i n g . J. For. 79: 280-283.
Oliver, C. D.
1978.
The development of northern red oak in
m i x e d stands in central New England.
B u l l . Yale Univ.
Sch. For. Environ.
Stud. : 91. Yale University,
New
Haven, Connecticut.
________ .
1980.
Even-aged
development
stands.
J. For. 78: 201-203.

of

mixed-species

________ . 1981. Forest development in North America following
ma j o r disturbances.
For. E c o l . Manage. 3: 153-168.
Oliver,
C.
D . , A.
B.
Adams and R.
J.
Zasoski.
1985.
Disturbance
patterns
and
forest
development
in
a
recently deglaciated valley in the northwestern Cascade
Range of Washington, U.S.A.
Can. J. For. Res. 15: 221232.

20

Oliver, C. D. , W. K. Clatterbuck,
and E. C. Burkhardt.
1990. Spacing and stratification patterns of cherrybark
oak and American sycamore in mixed, even-aged stands in
the southeastern United States.
For. Ecol. Manage. 31:
67-79.
Oliver, C. D. and B. C. Larson.
McGraw H i l l , New York.
Peet, R. K.
Range.
Peet,

1990.

Forest stand dynamics.

1981.
Forest vegetation of the Colorado Front
Vegetatio 45: 3-75.

R. K. and
tree death.

N. L. Christensen. 1987.
BioSci. 37: 586-595.

Competition

and

Roberts, M. R. and C. J. Richardson.
1985.
Forty-one years
of population change and community succession in aspen
forests on four soil types in northern Lower Michigan,
U.S.A.
Can. J. Bot. 63: 1641-1651.
Romme,
W.R. and D. H. Knight.
1981. Fire frequency and
subalpine
forest
succession
along
a
topographic
gradient in Wyoming.
Ecology 62: 319-326.
Rowe,

J . S. and G. W. S c o t t e r .
1973.
forest.
Quat. Res. 3: 444-464.

Sakai,
A.K. 1990.
Sexual
northern Lower Michigan.

Fire

in the boreal

reproduction of red maple in
Am. M i d i . Nat. 123: 309-318.

Sakai,
A. K . , M.
R. Roberts
and
C.
J. J o l l s .
1985.
Successional
changes
in a mature
aspen
forest
in
northern Lower Michigan: 1974-1981. Am. M i d i . Nat. 113:
271-282.
S c h e i n e r , S. M. , T. L. Sharik, M. R. Roberts and R. Vande
Kopple.
1988.
Tree
density
and
modes
of
tree
recruitment in a Michigan pine-hardwood forest after
clear-cutting and burning.
Can. F i e l d - N a t . 102: 634638.
Segura, G.
and L.
C. Snook.
1992.
Stand dynamics
regeneration patterns of a pinyon pine forest in
central Mexico.
For. E c o l . Manage. 47: 175-194.

and
east

S h a i n s k y , L. J. and S. R. Radosevich.
1992.
Mechanisms of
competition
between
Douglas-fir
and
red
alder
seedlings.
Ecology 73: 30-45.
Sharik,
T.
L. , R.
H.
Ford
and
M.
L.
Davis.
1989.
Repeatabi1 ity of invasion of eastern white pine on dry
sites in northern Lower Michigan.
Am. M i d i . Nat. 122:
133-141.

21

Smith, D. M.
1986. The practice of silviculture.
and Sons, New York.

John Wiley

Sprugel, D. G.
1976.
Dynamic structure of wave-generated
Abies balsaemea
forests in the northeastern United
States.
J. Ecol. 64: 889-911.
________.
1991.
Disturbance, equilibrium, and environmental
variability: what is "natural” vegetation in a changing
environment?
Biol. Cons. 58: 1-18.
Stoeckeler, J. H. and J. Macon. 1956.
Regeneration of aspen
cutover areas in northern Wisconsin.
J . F o r . : 54: 1316.
Stubblefield, G. W. and C. D. Oliver.
1978. Silvicultural
implications
of
the
reconstruction
of
mixed
alder/conifer s t a n d s . In Utilization and management of
red alder.
U.S.D.A. For. Ser. Gen.
Tech. R e p . : P N W - 7 0 .
pp. 307-320
Veblen,
T.
T.
and
D.
H. Ashton.
influences on the vegetation of
Chile.
V egetatio 36: 149-167.

1978. Catastrophic
the Valdivian Andes,

Whitney, G. G.
1986.
Relation of Michigan's presettlement
pine
forests to substrate and disturbance history.
Ecology 67: 1548-1559.
________.
1987.
An ecological history of the Great
forest of Michigan.
J. E c o l . 75: 667-684.

Lakes

________.
1990.
The history and status of the hemlockhardwood forests of the Allegheny Plateau. J. E c o l . 78:
443-458.
Wierman, C. A. and C. D. Oliver. 1979.
Crown stratification
by species in even-aged mixed stands of Douglas-firw estern hemlock.
Can. J. For. Res. 9: 1-9.
________.
1990.
The history and status of the hemlockh ardwood forests of the Allegheny Plateau. J. E c o l . 78:
443-458.
Z a h n e r , R. and
N. A. Crawford.
1965.
The clonal concept
in aspen site relations.
In Forest-soil r e l a t i o n s h i ps.
Edited by C. T. Youngberg. Oregon State University
Press, Corvallis,
pp. 230-243.

Chapter 2
THE RELATIVE INFLUENCE OF E S T ABLISHMENT TIME A N D HEIGHT
GROWTH RATES ON SPECIES V E R TICAL STRATIFICATION DURING
SECONDARY FOREST SUCCESSION

ABSTRACT

The

vertical

growth

histories

rubra L . , Fraxinus americana

L.

of

individual

Ouercus

rub rum L.

growing

and A cer

in a 42-year-old Populus grandidentata Michx.-P.

tremuloides

Michx.-dominated

using

analysis.

forest

Species

were

established

r e constructed

contempor a n e o u s l y

stem

early

in

the sere but temporally separated periods of p e a k individual
establishment occurred among species,
of

Q.

rubra

americana
species

established

and
by

42

A.

rubrum.
years

prior
The

to

such that the majority
the

vertical

paralleled

m a j ority

of

F.

stratification

of

establishment

patterns.

Vertical growth rates were similar among species and between
different-aged

individuals

within

species.

This

suggests

that species vertical stratification of species by 42 years
after stand

initiation was primarily a function differences

in species establishment patterns.

22

23
INTRODUCTION

Tree

species

succession

has

establishment

often

been

during

secondary

characterized

by

forest

an

initial

floristic model in which the majority of individuals for all
overstory

tree

species

that

will

occur

assumed to be present at the time of,
initiation

(Egler 1954;

Oliver

and

Larson

1984;

Sakai

relationships

among

stratification,
differences

al.

in

function of

1985) .

result

height

such

that

dominance

growing,

short-lived
species

to

(Drury

is assumed
of

Nisbet

serai

Species
to be

slow

1973;

a

component

relatively

relatively

and

1978;

vertical

1983).

from

stand

interspecific

and mortality

shifts

species

in

from

are

Dickmann

model,

(Hibbs

secondary succession

species,

Oliver

manifest

rates

sere

and

this

primarily

growth

1973;

Abrams

Under
as

growth

differential

longer-lived

1983;

species,

must

replacement during

Hibbs

the

or soon after,

Drury and Nisbet

1990;

et

in

fast

growing,
Peet

and

Christensen 1980; Hibbs 1983; Tilman 1988).
This scenario becomes less straightforward if there has
been

temporal

within
that

a

separation

species

the

majority

established

at

floristics

(sensu

initially
translate

level

into

of

within

times.

This

1954)

since

Egler

Delayed

individual
resource

individuals

of

floristics

individuals

differential

early-establishing

establishment

initial

different

present.

in

is
all

individuals,

framework,
each
not

achieve

species

true

species

establishment

acquisition
and

such

such

maintain

relay
were
may
that
size

24

superiority
1987).

(Harper

This

has

1977,

been

pp.

247-248;

demonstrated

herbaceous monocultures

by

Huston
several

(Black and Wilkinson

and

Smith

workers

1963;

Ross

in
and

Harper 1972; Adul-Faith and Bazzaz 1979; Naylor 1980; Howell
1981)

and within naturally established populations of trees

in early successional
1986).

Within

forests

tree

(Sakai and Sulak 1985;

populations,

early

Larson

establishing

individuals may gain size superiority by virtue of enhanced
resource capture and either increased vertical growth rates,
relative

to

rates

individuals,

given

or

individuals

contemporaneous

suppression

(Oliver

and

of

establishment

later

Larson

1990).

of

establishing
Alternatively,

vertical dominance may be a function of increased residence
time,

i.e.

early establishing individuals have simply had a

longer time to grow.
Given

temporal

individuals,
profile

species

will

individuals

be

separation
level

each

would

not

vertical

on

Thus,

sufficiently

account

stratification.

An

s u c c e s s , as
influenced

reflected
by

understanding
succession.

in

the

in

the

the

of

vertical

majority

established,

as

well

of
as

an initial floristics model
for

observed

examination

vertical

establishment
of

when

species

rates of vertical g r o w t h .

establishment

relationships

dependent

within

in

growth

time,

mechanisms

of

of

individual

pattern,

would

driving

patterns

and

further

secondary

as
our

forest

25

Stem
growth

analysis

was

histories

used

of

to

reconstruct

individuals

from

the

vertical

three

later

successional tree species in a pioneer Populus grandidentata
Michx.-P.
to:

1)

tremuloides Michx.

describe

vertical

species

growth

within

rates

species

and

forest.

Study objectives were

establishment
between

among

p a t t e r n s ; 2)

different

species;

aged

and

compare

individuals

3)

compare

and

contrast the influence of time of establishment and vertical
growth

rates

on

individual

and

species

position

in

the

vertical profile.

METHODS
Stand Description and Vegetation Sampling
This

study

was

grandidentatacentral

P.

Lower

Soil

Alfic Haplorthod
42

year

determined
s p p . ) on
Remnant

of the

aspen
basal

a

formerly

old

trees

(SW

one

developed

1/4,

stand

was

of

grazed

old

fire

on

hectare

located

study area

ages

and

a

forest

U.S.A.

and has

old

from

in

tremuloides

Michigan,

Clare County) .

The

conducted

SEC

north-

T19N,

is classified
a morainal

several

charred

in

3,

initiated

field

Populus

in

stumps

as an

landform.
1947

overstory

subsequent

R3W,

(as

Populus
to

fire.

suggest

the

original forest contained both Pinus strobus L. and northern
hardwood species,
Arboreal
circular

plots

each plot,

in particular Acer saccharum Marsh.
v e g e tation

distributed

was

sampled

randomly

species and diameter

in

within

four
the

250

stand.

m2
In

(at 1.37 m) of all overstory

26

(> 10 cm dbh)

and sapling

were determined.

(2.5 cm > dbh < 10 cm)

Tree seedlings

individuals

(< 2.5 cm dbh) were sampled

in 12 one m 2 frequency frames spaced at 3 m intervals along
four opposing radii of each plot.
Stem Sampling
Stratified
individual
rubrum
using

random

L.

for

stem

reconstruction

analysis.

follows.

In

species

height
Some

class
height

which

used

to

vertical

growth

individuals

appeared

to

histories

were

be

of

were

plot,

one

randomly
from

classes

each
did

to

five

selected
1

m

not

of each species in each plot.
americana and 26 A.

to

Populus

determine

dominants.
soil.
(10

Stems

cm)

to

from

height

the

contain

any

continued
dominant

to

inte r v a l .

the

leader.

individuals

age

stems

and

thereafter.

were

and

at

1 m

1 m

.25 m

>

10

r u b r a . 58 F.

were

in height)

was

pattern
cm

intervals

above

to

the

removed
at

one

sampled
of

the

mineral

from the

thereafter.

closest

sections
2 m

at

intervals

multiple

Stem

(usually

at

felled

individual

Additionally,

growth

m

individuals.

from each plot
the

within

0.5-1.0

class

A total of 33 Q.

grandidentata

were marked

2 m

from

rather

individuals

rubrum were sampled.

stand
All

all

seed

Selection proceeded to the height of the tallest

overstory

select

Selection was stratified by height class

each

and

of

Sampled

genets

than sprout origin.

each

was

Ouercus rubra L . , Fraxinus americana L. and Acer

single-stemmed

as

sampling

base

Marking

end
from

of

the

large

each measurement

27
Stem Analysis
Small
terminal

individuals
leaders

counting

the

of

number

larger
of

each height i n t e r v a l .
were

sanded

to

examination.
section
used

a

and

2

m

in

bud

height) ,
were

scale

and

aged

scars

by

preceding

Stem sections from larger individuals

smooth

surface

dissecting

reconstruct

establishment
sampled

a

<

individuals,

terminal

Ring number was

under

to

(usually

the

years

individual.

and

aid

ring

counted on two radii of

each

microscope.

vertical
to

wetted

Vertical

Section

growth

reach

a

of

trajectories

of

each

of

all

Section ages were

also used to determine vertical growth rates
(e.g.

were

(time

height)

individuals were plotted together by plot.

one meter height interval

ages

history

given

growth

to

(m/yr)

for each

growth rate over 0-1 m,

1-2

m, e t c . ) of all sampled trees.

Data Analysis
Mean

times

of

stand initiation)
and A.
data

met

(number

were compared among Q.

rubrum using

sets

establishment

one-way ANOVA.

assumptions

of

All

of

years

after

r u b r a . F. americana
establishment

normality

of

residuals

time
and

homogeneity of v a r i a n c e s .

Tukey's test was used to separate

individual species means.

Mea n species heights,

age of
(data
and

42 years,
failed

wer e

to meet

homogeneity

of

Under the main test,
=.1

level.

compared using
assumptions

variances

of

even

at a stand

a Kruskal-Wallis
normality
w ith

of

test

residuals

transformation).

mea n species heights differed at the P

M a n n - Whitney

U-tests

were

used

to

locate

28

differences
used

as

between

individual

replicates

for

species

comparisons

means.

of

Plots

establishment

were
times

and heights among species.
Vertical

growth

rates

were

compared

establishing

individuals

(22-42-year-old

establishing

individuals

(1-21-year-old

species.
species

Vertical
by

growth

pooling

rates

individuals

between

stems)
stems)

were

and

the

later

within

compared

across

early

two

among
age

one

meter

height

Analysis section.
of

height

plots

as

described

in

For intraspecific comparisons,

intervals

maxim u m height

interval

used

in

each

analysis

shared by both age groups

the

for
Stem

the number

depended

on at

all

groups.

Comparisons were made using the growth rates determined
each

each

on

the

least three

(0-1 m for A. r u b r u m . 0-1 m through 3-4 m for Q. rubra

and F. a m e r i c a n a ) .

For the interspecific comparison,

rates for the first eight height intervals

growth

(0-1 m through 7-

8 m) were used.
A
the
A.

t-test

was

0-1 m height
rubrum.

u sed

Q.

rubra

using

split-plot

compare

rates

and

F.

were

accepted

a m e r i c a n a . and

repeated

Gurevitch

for both

and

and

compared

measures

interval was the repeated measure.
was

mean

i n t e r v a l , of early

Growth

within

to

1986)

later

rates,

among

over

establishing

between

ANOVA,

age

all

groups

species,

where

height

The sphericity assumption

intraspecific

Chester

growth

but

comparisons
was

rejected

(P >

.25;

for

the

interspecific comparison

(P < .001) . For the latter analysis

adjusted

used

F

values

were

to

test

hypotheses

regarding

29

height

interval

growth

rates

contrasts

and

height

(Moser

et

al.

used

to

separate

were

across height intervals.
all

growth

rate

transformed

to

interval
1990).

species

effects

Orthogonal

trends

in

Data

assumptions

were
of

growth

rates

inverse-square

normality

on

polynomial

Plots were used as replicates

analyses.

meet

by

of

for
root

residuals

and homogeneity of variances.
The relationships between stem age and stem height were
examined
individual
each

by

fitting

age-height

species.

least-squares
curves,

Regression

data so that the natural
stem height,
and

vigor

the

pooled

lines

were

relationship

often sigmoid

of

regression
across
fit

t r e e ; Husch

et

plots,

to

b etween

or hyperbolic
al.

lines

to

within

untransformed
stem age

and

(depending on

age

1972)

would

not

be

obscured.

RESULTS
Forest Composition and Structure
The

forest

overstory

grandidentata

and

characterized

by

P.

tremuloides.

species

hardwood association

was

typical

in the

Curtis 1959; Table 2.1).

dominated
while
of

upper Lake

Ouercus rubra,

by

regeneration

the

northern

States

Fraxinus

Additionally

these

species

was

mesic

(Braun

and Acer rubrum were the only later successional
the o v e r s t o r y .

Populus

1950;

americana
species

had the

in

highest

Table 2.1.

Forest composition and structure by stratum.

Overstory

Sapling

Seedling

Density;
Basal Area
(stems/ha);
(m /ha)

Density
(stems/ha)

Density
stems/ha

Species

Pooulus arandidentata
P. tremuloides
Ouercus rubra
Fraxinus americana
Acer rubrum
Prunus serotina
Tilia americana
Ostrva vircriniana
Facms arandifolia
Acer saccharum
Carva cordiformis

390a (173); 10.4
310 (234);
7.3
90 (45) ; 1.2
30 (19) ; 0.5
30 (19) ; 0.3

(4.5)
(5.1)
(0.7)
(0.4)
(0.2)

80
30
260
320
210
470
60
30
20
10

(32)
(19)
(90)
(91)
(145)
(330)
(39)
(10)
(20)
(10)

400
60
552
9950
8260
2230
6000
150
130
1650
10

(361)
(39)
(199)
(4343)
(1118)
(1005)
(2868)
(100)
(105)
(425)
(10)

Total

850

(3.7)

1490

(555)

29,391

(4516)

aValues are means
sapling,

(150);

19.7

(+/“ standard error)

of four replicates.

2.5 cm < dbh < 10 cm; seedling, dbh < 2.5 cm.

Overstory dbh > 10cm;

31

understory

density

the exception

among

of Prunus

latter species were

later

successional

serotina Ehrh.

found primarily

species,

Individuals

in one plot,

with

of the

hence

its

importance was not considered representative of the stand in
general

and

for simplicity sake

it was not

included

in the

vertical growth reconstruct i o n .

Patterns of Establishment and Vertical Stratification
Individual
illustrated

in

age-height
Figure

reconstructions

2.1.

The

by

four Pooulus

plot

are

grandidentata

that w ere sampled achieved vertical dominance both by virtue
of

early

establishment

establishment

and

patterns

for

rapid

early

growth

r u b r a . F.

Q.

rates.

americana

The

and

A.

rubrum were characterized by temporally separated periods of
peak individual e s t a b l i s h m e n t .
patterns

among

frequency

species

Differences in establishment

are better

distributions

of

illustrated

individual

function of stand age

(Figure 2.2).

were

in

present

surviving
surviving

Q.

early
rubra

F.

establishment

stand

americana
(number

of

and
years

as

a

While all three species

prior
A.

examining

establishment

history,

established

by

the

to

the

rubrum.

after

stand

majority

of

majority

of

Mean

time

of

initiation)

for

Q. rubra was significantly earlier than for F. americana and
A.

rubrum

distributions
(Figure
not

2.2).

(P

<

of

.05;

the

latter

Further,

significantly

Table

mean

different

two

2.2).
species

time
for

The

of

these

were

establishment
very

establishment
species

(P

similar
of
>

was
.05;

32

Figure 2.1. Vertical growth reconstructions of individual
Populus
grandidentata.
Ouercus
rubra.
Fraxinus
americana and A c e r rubrum over a 42 year p e riod in
four plots. Each line represents an individual stem.

Populus grandidentata

PLOT 1

— Fraxinua americana

HEIGHT (M )

Acer rubrum
Quercua rubra

24
22
20
18
16
14

F ig u r e

20

PLOT 2

12
10

8
6
4
2
0
0

4

'

0

........... .

4

8

... . ... . i ' i ■ i ' i i i i i i i ■ i» * | i i i i ■ i i i

12

16

20

24

28

32

36

40

44

co

2
4T
22
-

-

HEIGHT (M)

20

PLOT3

to

22
20
18
16
14

PLOT 4

12

10

8
6
4
2
0
8
YEARS AFTER STAND ESTABLISHMENT

12

16

20

24

28

32

36

YEARS AFTER STAND ESTABLISHMENT

40

44

34

Figure 2.2.
The number
as
a
function
of
successional
species
tremuloides forest.

of new individuals establishing
stand
age
for
three
later
in a Populus qrandidentata- P .

35
Figure 2.2

Quercus rubra

5

10

15

20

25

30

35

40

24

zt

a

21

>
o

1B

z
z
u_
o
s
m
3
3

Z

Fraxinua americana

15
12

8
6
3

0

I

5

10

15

20

25

30

35

40

15

20

25

30

35

40

24
Acer rubrum

21
18
15
12

9
8
3
0

5

10

YEARS AFTER STAND ESTABLISHMENT

36

Table 2.2.
Mean time of establishment (number
of years after stand initiation) for three
later successional tree species and analysis
of variance for species differences in
establishment time.

Species

Ouercus rubra
Fraxinus americana
Acer rubrum

Years

S.E.

19.2a
25.2b
25.9b

(0.8)
(1.3)
(1.8)

na

4
4
4

Analysis of Variance

Source of
Variation

Species
Error

df

2
9

SS

104.803
66.030

P > F

0.010

aSample size (number of p l o t s ) . Note: within
the years column, means followed by the same
letter were not significantly different at
P = .05.
SS, Sums of Squares.

37

Table

2.2).

The right

skewed establishment

americana and A. rubrum

(Figure 2.2)

profiles

for

F.

could be the result of

recruitment from seed produced by a limited number of early
establishing individuals
by 20-35 years

that

reached

(see Figure 2.1),

reproductive

maturity

thus providing a local seed

source for additional establishment.
By

year

americana
pattern

42,

and

vertical

A.

seen

stratification

rubrum

in

approximated

establ ishment

of

the

Q.

r u b r a . F.

stratification

times.

Most

£>.

rubra

individuals were taller than most F. americana and A. rubrum
individuals

(Figure

the

two

latter

2.1).

species

Occasional

individuals

plots 2-4)

and A.

established

earlier
achieved

positions

the

heights

show
of

rubrum

individuals
in

The

for

considerably

F.

americana

was

and

profile

typically

maintained

than

the

more

(see

(see Figure 2.1,

than

individuals

overlap.

Figure

plots

of

2.1,

1,3

and 4)

observed.

These

taller

"a v e r a g e ”

vertical

species

level

positions.
Mean
over
2.3.

the

species-level
42

year

Populus

stand

changes

in

history

are

grandidentata

vertical

was

stratification

illustrated
taller

than

species throughout the period of examination.
was taller than both

F.

americana

and remained so through y ear 42.
rubra

over

the

other

two

and A.

in
all

at

other

Ouercus rubra

rubrum by year 32

The greater height

species

Figure

year

42

was

of Q.
only

marginally significant

(Us= 1 6 , n=4 and 4, P <. 1 and U s= 14,

n= 4 and 4,

for comparison with F. americana and

.1< P <.2,

38

Figure 2.3.
Vertical stratification patterns among four
tree species over a 42 year period in a Pooulus
qrandidentata- P . tremuloides forest. Heights are mean
values (+/“ standard error) of four plots at each age.

25

20

15

-

Populus grandidentata
E S S Qu ercus rubra
I N l Fr a xi nus a m e r i c a n a
Acer r u b r u m

-

I

X
O

LU
X

10

I

-

I
5

I

-

0-

2

7

12

17

22

27

32

37

YEARS AFTER STAND ESTABLISHMENT

42

40

A.

rubrum

respectively).

Mean

heights

of

species were

not significantly different

4,

The height

P >.2).

influence
between

of
27

and

individuals
years,

newly

profile

for A.

established

32

years

reduced

the

(Us=

rubrum

stand

height

latter

11,

individuals

after

mean

the

n=

4 and

reflects
in

the

the

the
sere

initiation.

of

two

These

species

at

32

relative to 27 years.

Intraspecific Height Growth Rates
Within

each

individuals
initiation

species,

establishing
and

those

vertical

within

followed

by

1-21

95%

For A.

confidence

(0.08-0.17)

not significant
for

£>.

intervals

examined,

each

species,

were

not

both

species,

2.3,

.001

<.005;

americana.

are

overall

there

growth

a

(P >

0-1

was

2.3).

The

americana

quadratic

significant
This

is

of

m/yr

Growth rates

the

the
.5;

four

height

2.4.

Within

age

groups

two
Table

linear

2.3).

For

increase
(P <

in

.005

r e s p e c t i v e l y ; Table

term

for

m)

(0.07-0.15)

Figure

significant

rubra and F.

2.4).

effect

Table

different

was

in
of

for

.3) .

across

rates

after

The difference was

P >

illustrated

stand

years

from lower to upper height intervals

for Q.

Figure

interval

F.

significantly

growth rates
and

and

df= 3,4,

of

r u b r u m . growth rates

r e s p e ctively.

(t= -1.030,

rubra

intervals

m/yr,

after

22-42

the older and younger age groups were 0.10
and 0.12

rates

years

establishing within

stand initiation were similar.
(means

growth

F.

a reflection

with i n

the

americana
of

the

height
only

(P

non-linear

41

Figure 2.4.
Vertical growth rates (m/yr) for two age
groups over four height intervals within two later
successional tree species. Old individuals established
within
1-21
years
after
stand
initiation.
YOUNG
individuals established within 22-42 years after stand
initiation.
Values
are
back-transformed
means
of
inverse-square root transformed growth rates (m/yr),
with 95% confidence intervals. Sample size (number of
plots) = 3 for £. rubra (O L D ) and 4 for all other age
group-species combinations.

42
Figure 2.4

( M /Y R )

0. 6 -

HEIGHT

GROWTH

0 .8 -

RATE

I

A. Q u e r c u s r u b r a

0.4-

0 .2 ©•■« OLD
GHD YOUNG

0.0

(M /Y R )

06

HEIGHT

GROWTH

0.8 -

RATE

0-1

1-2

2-3

3-4

B. F r a x i n u s a m e r i c a n a

.

-

0.4-

0 .2 ®--® OLD
YOUNG

0.0
0-1

1-2

2-3

HEIGHT INTERVAL (M )

3-4

43

Table 2.3.
Repeated measures analysis of variance for age
group differences in height-growth rates (m/yr) within two
later successional tree species.

Species
Ouercus rubra

Fraxinus

americana

Source of
Variation

df

Age group
1
Plot
5
Height interval
3
Linear
1
Quadratic
1
Height interval
3
x Age group
Error
15

SS, Sums of Squares.

SS

0. 003
0.165
3.386
2.566
0.536
0.012
3.259

P > F

0.779
0.012
< 0.005
> 0.100
0.997

df

SS

1
6
3
1
1
3

0.056
0.542
9.021
8.481
0.481
0. 083

18

0.793

P > :

0.460
0.001
< 0.001
< 0.005
0.608

44

increase in growth rate between 2-3 m and 3-4 m seen within
the older age group of this species
the

interaction

between

height

(Figure 2.4b).

interval

not significant for either species

and

age

Overall,
group

was

(P > .5; Table 2.3).

Interspecific Height Growth Rates
Height

growth

rates

individuals

within

each

intervals.

This

height

stratification at

of

early

species

and

were

suggests

late

establishing

similar

that

over

species

42 years was not a result

of

shared

vertical

significant

changes in growth rates between early and late establishing
individuals

within

individuals within

a

species,

species were

thus,

pooled

different

for

aged

a comparison

of

growth rates among s p e c i e s .
The

pattern

of

species at 42 years
function

of

vertical

stratification

(Figures 2.1 and 2.3)

establishment

time

interspecific vertical growth rates.
of

peak

could

be

establishment
coupled

growth r a t e s ; ii)
relative

to

F.

with:

scenarios

i)

species
similar

americana

and A.

interacting

found

in

this

interspecific

r u b r u m ; or

result

vertical stratification profile,

in

with

The separation in time

iii)

the

observed

study

vertical

for Q.

in the latter two species.

could

among

was necessarily a

h i gher vertical growth rates

vertical growth rates
three

among

seen

rubra

increased
While all
species

the influence of the second

possibility would tend to accentuate the pattern,
third would tend to ameliorate it.

while the

45

Growth rates
across

the

Figure

2.5.

for Q.

eight

r u b r a . F.

height

Overall

intervals

growth

different among species
significant
upper
The

linear

height

quadratic

2.4).

rates

(P >

.05;

increase

intervals
term

americana and A.

in

(P <

was

also

examined,

are

were

significantly

Table

not
2.4).

growth

.001;

rubrum.

rate

Table

significant

shown

in

There was a
from

2.4,
(P

lower

Figure
<

to

2.5).

.01;

Table

This is a reflection of the decrease in growth rates

over 6-8 m seen in both £>. rubra and A.

rubrum

(Figure 2.5).

This decrease did not occur in F. a m e r i c a n a . w h i c h accounts
for

the

marginally

significant

interaction term (Pad j .
term

was

also

height

< •04to

significant

interval

by

.10; Table 2.4).
(P

< .01;

Table

species

The cubic
2.4)

but

interpretation of this is difficult given the high amount of
variability

in

growth

rates across mid -

intervals within all species

to

upper

height

(Figure 2.5).

The similarity of vertical growth rates among the three
species
stem

can

be

by

age-stern height

curves for all
too

seen

few

separate

examining

data.

For

this

lines

analyses,

fit

to

age-height

individuals were pooled across plots because

individuals

were

sampled

r e g r e s s i o n s .Regressions

maximum age,

regression

within each species,

within
were

each

plot

to

restricted

fit

to

a

with at least four height

observations.
For
function
(r2=

all

three

provided

.659,

species

the

.822 and

best

.910

a

second

interpretable

for Q.

r u b r a . F.

order
fit

polynomial
to

the

americana

data

and A.

46

Figure 2.5.
Vertical growth rates (m/yr) for three later
successional tree species over eight height intervals,
pooled across age groups. Values are back-transformed
means of inverse-square root transformed growth rates
(m/yr), wit h 95% confidence intervals. Sample size
(number of plots) = 3 for A. rubrum and 4 for Q. rubra
and F. americana.

Figure

(M /Y R )

0. 6-

HEIGHT

GROWTH

0.8 -

RATE

# - - # Acer r u b r u m
EHE] F r a x i n u s a m e r i c a n a
▲ - A Quercus rubra

0 4
.

-

- 4?

0.2

0.0
0-1

1-2

2-3

3-4

4-5

5-6

HEIGHT INTERVAL (M)

6-7

7-8

48

Table 2.4.
Repeated measures analysis of variance for
species differences in height-growth rates (m/yr).

A d i . P > Fa
Source of
variation

Species
Plot
Height interval
Linear
Quadratic
Cubic
Height interval
x Species
Error

df

SS

P > F

2
8
7
1
1
1
14

0.382
0.468
17.257
9.072
6.281
0.699
1.748

0. 092

56

3 .601

0. 001
< 0.001
< 0.010
< 0. 010
0.041

G-G

H-F

0.001

0.001

0.104

0.041

aA d j . P > F are probabilities associated w i t h the
Greenhouse-Geisser (G-G) and Huynh-Feldt (H-F)
adjusted F-tests.
SS, Sums of Squares.

49

rubrum

respectively,

2.5).

Individual age-height plots and regression lines

the

three

species

P

are

comparison

of

age-height

relationships

similar.

by

shown

lines

increased

year period.
of

42

(Figure
the

the

summary

in

(Figure
among

statistics

Figure

2.6d)

2.6.

in

A

indicates

the three

species

Table
for

visual

that
were

the
very

A 10-15 year period of relatively slow growth was

followed

age

<.01;

growth

over

the

remainder

The greater mean height of Q.

y e a r s , compared

2.3),

was

age-height

to

achieved

F.

despite

relationship,

relative

to

the

42

rubra at a stand

americana
lower

of

and

A.

rubrum

predictability
the

other

of

species

(Figure 2. 6a-c, Table 2.5) and lower g r owth rates over 6-8 m
(Figure 2.5).
DISCUSSION
The

influence

establishment
rates,
has

on

life

requirements,

promoting

been

of

species

incorporated
Slatyer

into

1980;

history
shade

traits,

tolerance

replacement
several

Huston

including
and

during

succession

successional

and

Smith

growth

1987;

models

(Noble

and

1988) .

Under these m o d e l s , combinations of attributes that

convey competitive success change over a sere.
Populus

grandidentata.

requirement

(or

by

prolific

virtue

of

a

resp r o u t i n g ) ,

Tilman

For example,

mineral
shade

seed

bed

intolerance,

and a high vertical growth rate, achieves early successional
importance

soon

Conversely,

for a species such as Acer s a c c h a r u m . an organic

seed

bed

after

requirement,

a

shade

stand-initiating

tolerance

and

disturbance.

a

relatively

50

Model:

HEIGHT

Species

ll
do
o
+

Table 2.5.
Summary information for second order polynomial
regressions of stem height as a function of stem age for
three later successional tree species in a Populus
g r a n d i d e n t a t a - P . tremuloides forest.

B X * AGE + b 2 * AGE 2

Coefficients

(SEE)

r2

SEE

Ouercus rubra

B0 :
Bl:
B2 :

.4038
.0341
.0095

(.2609)
(.0443)
(.0016)

.659

(P< .001)

1.465

Fraxinus americana

B 0 : .4513
Bl*. -.0238
B 2 : .0127

(.1276)
(.0231)
(.0008)

.822

(P<

001)

0.895

Acer rubrum

Bq ’ -.0390
.0287
Bt:
B 2 : .0109

(.2034)
(.0321)
(.0010)

.910

(P<

001)

0.768

51

Figure 2.6.
Second order polynomial regressions of stem
height as a function of stem age for three later
successional
tree
species
(A-C)
in
a
Populus
a r a n d i d e n t a t a -P.
tremuloides
forest.
Dashed
lines
are 95% confidence bands for predicted values. Bestfit lines for all three species are compared in D.
Symbols represent number of observation as follows:
Ouercus r u b r a . 1-15 years=3-30 observations
(per
s y m b o l ) , 16-29
years=l-15
observations;
Fraxinus
americana.
1-15
years=8-30
observations,
16-29
years=l-14 observations; Acer r u b r u m . 1-15 years=215 observations, 16-3 0 years=l-5 observation.

A. Q u e r c u s r u b r a

HEIGHT

(M)

12

TOTAL

HiQ
d
H

14

14

B. F r a x i n u s a m e r i c a n a

(D

to

10

CTi

8

* '*

*

6
■

m

4

»* «

)*-*■« * »

» * * * * ftotjr# ft ft 9

_ - -£

m m Kjfi

2

0- *«
■i'

-2

-2

15

10

20

25

30

(M)

15

10

20

25

30

35

to

C. A c e r r u b r u m

12

HEIGHT

5

14

14

TOTAL

0

35

D. All S p e c i e s

10

8
6
4
*

m

2

0

—

-2

2
0

5

10

15

20

TOTAL AGE ( Y R S )

25

30

35

A. r u b r u m

— F. am e ric an a
•••• Q- rubro

0

5

10

15

20

TOTAL AGE ( Y R S )

25

30

35

53

slower

vertical

growth

rate

leads

to

later

successional

impor t a n c e .
The

underlying

reasons

for

changes

in

relative

importance among species over time is not so easily resolved
when

considering

history
species

species

characte r i s t i c s .
examined

americana

and

successional

in

with

a

The

three

this

Acer

study,

rubrum.

status

(climax

and 6.0 respectively;

Ouercus

are

1959)

suite

later

rubra.
in

density)

subtle
example,
in

in

differences
Hibbs

vertical

England

and

(1983)

differences

life

he ight

differences

in

height

differences

in

relative

profile,

history

and Oliver

forests

in

6.0,

of
8.0,

and understory tolerance

vertical

relationships

hardwood

Fraxinus

Differences in species importance,

the
in

life

terms

numbers=

both in terms of traditional horizontal measures
and

of

successional

similar

adaptation

Curtis

(Spurr and Barnes 1980).

similar

resulted

growth

in

rates.

from

For

New
small

Interspecific

wo u l d

positions

on

changes

central

p rimarily

rates

species

depend

found that

species

growth

may

characteristics.

(1978)

among

(basal area

account
in

a

for

vertical

profile given contemporaneous establishment of species.
Alternatively,

relative

importance

may

be

less

influenced by species characteristics

and more dependent on

differences

In

in

establishment

time.

the

present

study

interspecific height growth rates of Ouercus r u b r a . Fraxinus
americana and Acer rubrum were very similar through much of
the

42

year

period

examined

(Figure

2.5).

While

species

54

establishment
periods

of

occurred

was

contemporaneous,

peak

individual

(Figure

2.2).

temporally

establishment

The

majority

separated

among

of

species

Ouercus

rubra

individuals established prior to the m a j ority of individuals
from

the

other

the vertical
influenced

two

species.

profile by

by

Species

42 years

differences

in

relationships

(Figure

2.3)

establishment

within

were

largely

times.

Ouercus

rubra achieved a vertically dominant position in the forest,
relative
most

to

F.

americana

individuals

of

and

the

A.

species

r u b r u m . primarily
had

hence had a longer period to grow,

b een

there

because

longer

rather than as

and

a result

of higher height growth rates.
The

ability

importance

to

an

species

examined

greater

heights

individuals

of

early

establishment

individual
in

of

this
the

that

study.

several

growth

size advantage
time

superiority,

alone.
through

suppression
enhancing

rates
that

This

convey
among

is

americana

prior

to

and

In
over
early

of

the
the

beyond that

concluded

enhanced

resource

later

advantage
present

shared

gained
study,

height

emergence

that

has

the

intervals
not

yet

additional
competitive

led

led

additional

similar
among

of

individuals

individuals,

through

rubrum

bulk

acquisition,

establishing

the

Ross and Harper

conveyed by

They

three

from
A.

the

(Figure 2.1).

vertical

the

evident

found that early establishing herbaceous

gained a

time.

shared

F.

established

individuals for each species
(1972)

was

to

to

a

thus
growth

vertical
species

to

growth

suggests

competitive

55

advantage
relative

through

enhanced

species

resource

success

in

the

acquisition.
vertical

Rather,

profile

was

p rimarily a function of length of time available for growth.
Of

course

and

will

all

three

continue

later
to

successional

be,

influenced

species
by

the

have

been,

competitive

effects of the early establishing P o p u l u s .
It

is

clear

from

the

g rowth accelerated as
through

time.

results

the

rate

and

species,

Bazzaz

(1991)

have

height

shown,

increasing the volume of soil

available to an individual can have a positive
growth rate.

of

individuals of all three species grew

McConnaughay

that for some annual

that

influence on

They suggest that changes in root morphology,

in response to differences in available space, m a y alter the
ability of an individual to acquire resources.
the current
be

study suggest

directly

competitive
from

related

understory

increased

resources,

to

that tree height
occupation

environments.

acquisition

or both,

of

of

Results from

growth may

physical

Whether

below-ground

in response to changes

space

this

or

also
in

results

above-ground

in morphology

is

not known.
Early
advantage

establishment
both

w i thin

may

and

yet

among

convey

a

species.

As

matures and the overstory continues to develop,
environment experienced by smaller,

competitive
the

stand

the resource

younger individuals will

be very different from that currently experienced by larger,
older
reduced

individuals.
growth

rates

Resource
in

many

limitation
of

the

will

smaller

result

in

individuals

56

(Oliver and Stephens 1977; Hibbs 1983; Sakai and Sulak 1985)
and high rates of mortality within these populations

(Oliver

and Larson 1990).
SUMMARY AND CONCLUSIONS
Rates
americana

of

height

and

Acer

trees

became

influenced

Ouercus

in the

time

The
of

rubra.

secondary

For all species,

larger.

by

for

rubrum

were very similar.
as

growth

sere examined

growth rates

increase

with

result of

establishment.

differences

in time

increased

size was
The

stratification apparent 42 years after stand
primarily a

Fraxinus

not

vertical

initiation was

of

establishment

among individuals and species.
The results of this study may prove to be applicable to
secondary

forest

xeric sites
resource
reduced
levels
the

successions

in the Lake States.

limitation,
propagule

of

relative

site

was

other

to more
can

stocking. For
low

highly

disturbed

In these situations,

availability

overstory

study

on

productive

lead

to

(19.7 m 2/ha)

either

sites,

initially

example,

to

or
low

basal area

relative

or

of

Populus

grandidentata dominated stands on relatively less disturbed,
more

productive

m 2/ha

at

stocking

65-70
may

establishment.
disturbed
Waring

sites

recognized

years;

in

in

turn

northern
Chapter

favor

an

Lower

3).

in

Oliver

western

extended

et

forests

North

al.
of

1985)

eastern

Michigan

Reduced

This has been observed

forests

1979;

in

(33-47

overstory

period

of

stem

for xeric and highly

America
but
North

(Franklin

is

less

America

and

widely
where

57

individuals

in

similar

aged

stands

have

been

found

establish with i n shorter periods of time than documented
the present study

to
in

(Oliver 1978; Hibbs 1983).

The generality of patterns and mechanisms of secondary
forest

succession

could

expanding

the

individual

establishment

succession,

be

examination
and

resolved
of

the

growth,

to the landscape scale.

to

some

degree

interaction
and

its

Of course,

by

between

influence

on

this is more

than a trivial task when using the stem analysis technique.

BIBLIOGRAPHY
Abrams,
M.
D.
and
D.
I.
DicJcmann. 1984.
Floristic
composition before and after prescribed fire on a jack
pine
clearcut
site
in
northern lower
Michigan.
Canadian Journal of Forest R e s earch 14: 746-749 .
Adul-Faith, H. A. and F. A. B a z z a z .
1979.
The biology of
Ambrosia trifida L.
I I . Germination, emergence, growth
and s u r v i v a l . The New Phytologist 83: 817-827.
Black, J. N. and G. N. Wilkinson.
1963.
The role of time
of emergence in determining the growth of individual
plants in swards of subterranean clover
(Trifolium
subterranean L . ) .
Australian Journal of Agricultural
Research 14; 628-638.
B r a u n , E. L.
America.

1950.
Deciduous forests
The Free Press, New York.

of

Curtis, J. T.
1959.
The vegetation of
University of Wisconsin P r e s s , Madison.

eastern

North

Wisconsin.

The

D r u r y , W. H. and I . C. T. N i s b e t . 1973. Succession.
Journal of the Arnold A rboretum 54: 331-368.

The

E g l e r , F. E.
1954.
Vegetation science c o n c e p t s . 1. Initial
floristic composition-a factor in old-field vegetation
development.
Vegetatio 4: 412-417.
Franklin, J. F. and R. H. Waring. 1979. Distinctive features
of the northwestern coniferous
forest:
Development,
structure,
and
function,
in
R.
H.
Waring
(ed.) ,
Forests: Fresh Perspectives from Ecos y s t e m Analysis,
Proceedings of the 40th Annual
Biology Colloquium,
Oregon State University Press.
Gurevitch, J. and S. T. Chester,
repeated measures experiments.
H a r p e r , J. L.1977.
Press.

Jr.
1986. Analysis
Ecology 67: 251-255.

Population biology of p l a n t s .

of

Academic

Hibbs,
D. E.
1983.
Forty years of forest succession
central New England.
Ecology 64: 1394-1401.

in

H o w e l l , N.
1981.
The effect of seed size and relative
emergence time on fitness in a natural population of
Impatiens
capensis Meerb.
( B a l saminaceae). American
Midland Naturalist 105: 312-320.
Husch, B . , C. I . Miller and T. W. Beers.
1972.
mensuration.
John Wiley and Sons, New York.

58

Forest

59
Huston,
M.
and T.
Smith.
1987.
Plant succession:
history and competition.
The American Naturalist
168-198.

life
130:

Larson, B. C.
1986.
Development and growth of even-aged
stands of Douglas-fir and grand fir.
Canadian Journal
of Forest Research 16: 367-372.
M c C o n n a u g h a y , K. D. M. and F. A. B a z z a z . 1991.
space a soil resource?
Ecology 72: 94-103.

Is physical

Moser,
E. B. , A. M.
Saxton and S. R.
P e z e s h k i . 1990.
Repeated measures analysis of variance: application to
tree research.
Canadian Journal of Forest Research 20:
524-535.
Naylor, R. E. L.
1980.
Effects of seed size and emergence
time on subsequent growth of perennial r y e g r a s s . The
N e w Phytologist 84: 313-318.
Noble, I . R. and R. O. S l a t y e r . 1980. The use of vital
attributes to predict successional changes in plant
communities
subject
to
recurrent
disturbances.
V ege t a t i o 43: 5-21.
Oliver, C. D. and E. P. Stephens.
1977.
Reconstruction of
a mixed-species forest in central New England.
Ecology
58: 562-572.
Oliver, C. D.
1978.
The development of northern red oak in
m i x e d stands in central New England. School of Forestry
and Environmental Studies Bulletin Number 91.
Yale
University, New Haven, Connecticut, USA.
Oliver,
C.
D . , A.
B.
Adams and R.
J . Zasoski.
1985.
disturbance
patterns
and
forest
development
in
a
recently deglaciated valley in the northwestern Cascade
Range of Washington, U.S.A.
Canadian Journal of Forest
R e s earch 15:221-232.
Oliver, C. D. and B. C. Larson.
McGraw Hill.
P e e t , R. K. and N. L.
population process.

1990.

Forest stand dynamics.

Christensen.
1980. Succession:
Vegetatio 43: 131-140.

a

Ross, M. A. and J. L. Harper. 1972. Occupation of biological
space during seedling e s t a b l i s h m e n t . Journal of Ecology
68: 77-88.
Sakai,
A.
K . , M.
R.
Roberts
and
C.
Successional
changes
in a mature
northern Lower Michigan: 1974-1981.
Naturalist 113: 271-282.

L.
Jolls.
1985.
aspen
forest
in
American Midland

60

Sakai,
A.
K.
and J.
H.
Sulak.
1985.
Four decades
of
secondary succession in two lowland permanent plots in
northern Lower Michigan.
American Midland Naturalist
113: 146-157.
Spurr, S. H. and B. V. Barnes. 1980.
edition.
John Wi l e y and Sons.
Tilman, D.
1988.
Plant strategies
structure of plant communities.
Press, Princeton.

Forest ecology.

Third

and the dynamics and
Princeton University

Chapter 3
A COMPARISON OF PRESETTLEMENT AND PRESENT-DAY FORESTS ON TWO
BIGTOOTH ASPEN-DOMINATED LANDSCAPES IN NORTHERN LOWER
MICHIGAN

ABSTRACT

Forest

composition

influenced

by

physical

disturbance regimes.
the

natural

have

led

site

a

local

landscape

characteristics

and

is

prevailing

In many areas of eastern North America,

disturbance

presettlement
activities

within

regimes

that

influenced

forest composition have been altered by human

associated
to

with

substantial

settlement.

changes

in

These

alterations

composition

and

the

development of successional pathways m a r kedly different from
the presettlement conditions.
and

present-day

forest

composition

dominated

landscapes

in

examined.

Objectives

were

presettlement
disturbance

forest
regime

presettlement
landscape;

and

and

disturbance
forests and

In this s t u d y , presettlement
of

northern
to:

(i)

composition
of

each

present-day

two

bigtooth

Lower

M i c higan

reconstruct

to

the

and

potential

landscape;
forest

aspen-

(i_i)

c o mposition

were
relate

natural
compare
of

each

Ciii) assess the influence of p ostsettlement

history

on

the

the potential

development

of

the

current

successional

pathways

of the

composition,

reconstructed

two

landscapes.
Presettlement
General

Land

between

the

Office
two

forest
survey

landscapes.

r e c o r d s , differed
Landscape

61

1 was

using

substantially
dominated

by

62

fire-sensitive

eastern

hemlock

and

American

beech,

while

landscape 2 was dominated by fire-dependent red pine,
pine and jack pine.
related

to

climates.
are

Compositional differences m a y have been

differences

to differences

in presettlement

in physical

The

dominated

present-day

overstories

bigtooth

aspen,

species that were

of minor

importance

of

the

study

areas.

attributed

to

the

similar

red

sources

of

advanced

both

oak

and

Compositional
influence

regeneration

hemlock,

frequency,

of

in the

disturbance history on each landscape.
eliminated

fire

beech

and

landscapes
red

the

Individuals

convergence

of

disturbance

postsettlement
availability
differences

and

postsettlement

man y

pines

landscape.

and

remnant

in

species

the

two

development

changes

in

the

seed

the

red oak
survived
in

the

source

led to marked

understories

of

Changes in seed rain and fire exclusion

forest

are
types

presettlement conditions.
induced

forests have

recruitment

landscapes
of

vegetatively
in

seed

favored

latter species

Differences

in the present-day

the two landscapes.
in

proliferated

was

Logging and wildfires

and

of the

maple;

presettlement

development of forests dominated by bigtooth aspen,
and red maple.

or

site characteristics or regional

by

forests

white

apparently
markedly

The results

in disturbance

regimes

leading

different

to
from

the
the

illustrate how human
can

long-lasting effects on forest composition.

have

i m p o r t a nt,

63
INTRODUCTION

Forest
influenced by
local

edaphic

landscape

1983; Grimm,
and

composition

1976;

1988).

Vebl e n

and

In

1981;

abundance

are

across

Pregitzer et

addition

to

composition

regimes

Ashton,

1986).

species

a

a l ..

1986; Host et a i ., 1987; Roberts

vegetation

disturbance

Whitney,

Knight,

1984; Whitney,

characteristics,

tree

and physiographic v ariation

(Romiue and

Christensen,

natural

and

physical

is

(Heinselman,

1978;

Runkle,

influenced
1973;

1982;

These relationships are

site
by

Sprugel,

Grimm,

1984;

interrelated since

natural disturbance regimes are often influenced by physical
site characteristics
(Van W a g n e r , 1970;
1983; Grimm,

and vegetation composition and pattern

Heinselman,

1984; Whitney,

presettlement

pathways

have

instance,
upper
and

been

fire

Great

remnant

forest
altered

proliferation

Lakes

region

seed

sources

S p r u g e l , 1976;

the disturbance regimes that
composition
by

human

after

in

the

activities.

advanced

dominant

the

presettlement

by

landscapes

1973; Whitney,

Conversely,

eastern

invasion

in

For
the

regeneration

coniferous

species.

(Ouercus s p p . )

(Populus s p p . )-dominated forests where few existed

Heinselman,
1990).

successional

deforestation

eliminated
of

and

This favored the development of extensive oak
and aspen

Foster,

1986).

In eastern North America,
defined

1973;

and
mesic,

(Kilburn,

1987; Crow,

1957,

1960b;

1988; Nowacki et a l . ,

fire exclusion in natural oak forests of
midwestern

United

shade-tolerant

States

hardwood

has

allowed

species

(Dix,

64

1957;

Monk,

1978;

Lorimer,

al.,

1961;

Buell

1984;

et

al. , 1966;

McCune

and

Anderson

Cottam,

1985;

and

Adams,

Fralish

et

1991).
Successional predictability at a scale that emphasizes

the

influences

natural

of

local

disturbance

physical

regimes

is

site

characteristics

arguably a

tenable

and

concept.

H o w e v e r , when natural disturbance regimes have been altered
by activities associated with settlement,
in

community

occur.

composition

successional

The magnitude of this change,

a disturbed,
level

and

postsettlement

an ensuing change
d e velopment

and the potential

may

be

highly

for

forested landscape to recover a

of community composition resembling

condition,

will

dependent

on

its presettlement

the

degree

to

which

successional potential has been influenced by human activity
(for examples see Glitzenstein et a l . , 1990; Fralish et a l . ,
1991).
In

this

composition

study

of

two

the

presettlement

forested

landscapes

and
in

present-day

north e r n

Lower

Michigan were examined.

Initial observations indicated that

the

both

current

bigtooth
amounts

forests

aspen
of

rubrum L . ).

red

of

(Populus
oak

Both

surficial geology,

landscapes

arandidentata

(Ouercus

rubra

landscapes were

are

M i c h x . ),

L . ) and

red

dominated
w ith

lesser

maple

characterized by

by

(Acer

similar

soil type and postsettlement disturbance

histories. The study locations were located within different
climatic and physiogragphic settings.
to:

(i)

reconstruct

and

relate

Study objectives were
presettlement

forest

65

compositions to the potential natural disturbance regimes of
the

two

landscapes;

(ii)

compare

presettlement

forest

composition to the present-day forest of each landscape; and
(jlii)

assess

histor y

on

the

the

influence

development

of
of

postsettlement
the

current

disturbance

forests

and

the

potential successional pathways of the two l a n d s c a p e s .

METHODS
Study locations
Study area 1 was a 15 k m 2 landscape located within the
U niversity

of

Michigan

Biological

Station

(UMBS)

extreme northwestern portion of Lower Michigan
40

N,

longitude

84°

40'

W) .

Study

area

2

in

the

(latitude 45°
was

a

18

km2

landscape located within the Harrisville R a n g e r District of
the

Huron-Manistee

National

Forest

(Huron)

in

Lower Michi g a n

(latitude 44° 15' to 45° 00' N,

15' to 84° 45'

W).

northeastern
longitude 83°

The distance between centers of the two

study areas was approximately 110 km.
Surficial geology of both study areas consisted of deep
outwash

sands

Padley,

1989).

areas

were

Lapin,

US DA

State

SCS

as

of

of Michigan

a

Entic

Cheboygan

University,

The study areas were
districts

till

deposits

(Cooper,

1981;

Soils throughout the majority of both study

classified

1990;

Michig a n

overlaying

regional

Haplorthods
County

Forestry

soil

(Cooper,

1981;

survey,

1991;

Department

data

located within different
landscape

ecosystem

file).

climatic

classification

(UMBS= Presque Isle district; Huro n = Highplains;

66

Albert

et

several

a l . , 1986).
climatic

These

districts

potential evapotranspiration and average temperatures

(Table

3.1).

annual

districts

have

p r e c i p i t a t i o n , M a y -September
precipitation to potential
3.1).

notably,

in

sums,

two

most

somewhat
heat

The

variables,

differ

identical

total

p r e c i p i t a t i o n , and

evapotranspiration

In g e n e r a l , climate does

not

differ

July-August

ratios

(Table

greatly between

the two study a r e a s .
The

physiographic

settings

substantially different.

of

the

landscapes

were

The UMBS study area was located on

a 3 km wide strip of land bordered on the north and south by
two large kettle lakes.

The study area was bordered on the

west by a massive north-south oriented morainal ridge

(Lapin

1990), with soils generally finer textured than those
in the

study

survey).

area

by

farmland.

conifer-hardwood
(Kilburn,

east

plain)

L a m b . ; Simard

landform

forests

1957).

immediately

eastern

(USDA

SCS,

Cheboygan

County

soil

The eastern portion of the study area was bordered

predominantly

outwash

prop e r

found

Huron

a

90

d ominated
and Blank,

portions

of

the

characterized

to

occupied

The
of

Prior

jack

by

The
area

sandy-clay

of

area

outwash

1982).
study

m uch

study

km2
by

settlement,

this
was

system

pine

graded
soils

Lake

banksiana

southern

into

area

located

(Mack

(Pinus

northern,

lowland

and

a morainal

(Padley, 1989).

Northern hardwood forests currently occupy muc h of the mesic
portions

of

observations).

the

moraine

(Padley,

1989;

personal

Both study areas were internally homogeneous

Table 3.1.
areas.

Mean climatic variables for the districts containing the UMBS and Huron study

District
Variable

Presque Isle (UMBS)

Growing season length (days)
April-October heat sum
(°C-days, base 7.2°C)
Heat sum prior to last spring freeze (°C-days)
Ratio of night heat sum to
total heat sum (%)
May-September potential
evapotranspiration (mm)
July-August precipitation to potential
evapotranspiration ratio (%)
Total annual precipitation (mm)
May-September precipitation (mm)
Annual average temperature (°C)
May-September average temperature (°C)
Annual extreme minimum temperature (°C)

Data taken from Table 2 of Albert et al.

1986.

Highplains

120
2020

115
2140

240
24

300
23

470

490

70

69

770
400
6.2
15.9
-29

770
400
6.7
16.9
-28

(Huron)

68

wit h

respect

(bigtooth

to

physiographic,

aspen-dominated)

regions

deforested
centuries
and

in

late

(Gleason,

1923;

Cramer,

1963;

compositional
(Lapin,

1990;

Forestry Department data f i l e ) .

containing

the

and

characteristics

Michigan State University,
The

edaphic

the

study

nineteenth
Kilburn,

Whitney,

to

1957,

1987).

locations
early

1960a;

were

twentieth
Benninghoff

Frequent

slash-fueled

wildfires swept over these landscapes in the years following
logging.

Fires

eliminated

advanced

regeneration

and

most

remnant overstory

individuals of the dominant presettlement

species

1957,

(Kilburn,

1 9 6 0 a ; Whitney,

1987).

Presettlement forest composition and disturbance regimes
Pre-settlement
reconstructed
records.

using

The

and Whi t n e y
and

(1986,

diameter

corners
used

to

points)
measu r e

discussed

and

of

were
of

1987).

bearing

quarter

m ark

in

recorded.
importance

two

study

Office

surveys

areas

(GLO)

for

were

survey

reconstructing

well

as

criticisms

detail

by

Bourdo

of

(1954,

the
1956)

During these s u r v e y s , the species
trees

section

section

the
Land

of GLO

forests, as

are

of

General

use

presettlement
technique,

forests

points)

lines

used
and

between

Within
was

(trees

each

line

mar k

for

section

trees

corners

study

derived

to

and

area,
each

a

(trees
quarter

relative

species

by

dividing the number of times a species was used as a bearing
or

line

bearing

tree
and

(citation frequency)
line

trees

examined.

by

the

A total

total
of

number
79

and

of
111

69

bearing

and

line trees,

line

1.6

km), were

=

landscapes,
upland

all

were
or

11 and

examined

respectively.

forests

included

from

for

Only

examined.

portions

section lines

the

UMBS

section
The

of

19

and

lines

lines

sections

Huron

traversing

examined
26

(one

and

at

UMBS

31-34

of

Township 37N, Range 3W. Huron lines included all or portions
of sections 2-4 and 9-11 of T25N,
T26N,

R4E.

The physiographic

R4E and sections 33-35 of
settings

of

the

two

study

areas precluded larger sample sizes.

were

Presettlement

disturbance

reconstructed

using

methods

of

Surveyors

Lorimer

often

disturbance,
line
the
This

survey

(1977,

recorded

each

records

1 9 80a),

any

of

and

evidence

study

of

a

Typically,

section

the

W hitney

indicative

line

surveyors

affected

by

information was used to determine the

area

following

such as major windfalls or burnt land,

descriptions.
length

GLO

regimes

also

the

the

(1986).
of

past

in their
recorded

disturbance.

total

length

of

section lines and percentage of each study area affected by
windfall
was

or

fire.

annually

percentage

affected

following Whitney
by 30 and 15.

The

(1986),

by

of

each

disturbance

study
was

area

that

estimated,

by dividing each total percentage

These values represent high and low estimates

for the number of years that evidence of a past disturbance
would

remain

disturbance

visible
recurrence

to

a

surveyor.

interval

(Lorimer,

Finally,

the

1977)

was

estimated by dividing 100 by the percentage of area annually
disturbed by a particular type of disturbance.

Additionally,

70

township plat maps

(constructed by the surveyors)

of the two

study areas were examined for any evidence of disturbance.

Present-day vegetation sampling
Four and five stands were randomly selected from larger
data pools

(Lapin,

Department

data

file)

respectively.
to

similar

Barnes

et

1990; Michigan State University,

All

for

stands,

ecological
al. ,

characteristics

1982) ,
and

ground

Selected

stands m e t
size

of

the

1

ha;

since initiation; and

and

Huron

classification

based

State University,

UMBS

study

within each study area,

land

Michig a n

minimum

the

on

flora

units

areas,

belonged

(following

surficial

geology,

composition

(Lapin,

Forestry

Department

following additional
(2)

Forestry

free

from

1990;
file).

criteria:

obvious

(3) 0-5% slopes.

data

soil

(1)

disturbance

The distance between

stands within each study area ranged from 1-4.5 km.
Arboreal vegetation was sampled
randomly located within each stand.
UMBS and

272

required

to meet

plot,

species

individuals
cm)
(dbh

m 2 at Huron.
an

and

(dbh

tallied.

<

cm)

2.5

diameter
cm)

were

of

(at

and

Species

a

in

1.37

m)

of
(2.5

density
12

size

related

saplings

and

tallied

Plot size was 475 m 2 at

larger plot

objective

> 10

were

A

in four circular plots

1-m2

of

at UMBS

study.
all

In

was
each

overstory

cm < dbh <
tree

10

seedlings

f r a m e s . Seedling

frames were spaced at 3-m intervals along two perpendicular
diameters of each plot,
origin

of

overstory

beginning at the plot b o u n d a r y . The

individuals

(sprout

or

seedling)

was

71

assessed w h e n possible.

RESULTS
Presettlement forest composition
Presettlement forest composition of the two study areas
is summarized in Table 3.2.
in

decreasing

frequency,

order

species

of bearing

by eastern hemlock

American beech
strobus

The UMBS

L.)

forest was dominated,

and

line

tree citation

(Tsuga canadensis

(L.) Carr.),

(Fagus grandifolia E h r h .), white pine

and

accounted

red

pine

for

87%

(Pinus
of

all

resinosa
b earing

A i t . ).

and

white

trees.

pine

Aspen,

overstory

accounted for

78%

of

red

red

oak were

maple and

species

proportional

occurring

similarity

richness and importance
the

lowest

relative

at

index

all

both

that

presettlement

landscapes.

line

minor

areas.

integrated

A

species

(proportional similarity= the sum of
measure

of

importance

b etween

samples over all s p e c i e s ; Brower and Z a r , 1977)
compare

and

only

study

trees.

jack pine

bearing
the

These

line

The Huron study area was p i n e - d o m i n a t e d ? red pine,
and

(Pinus

overstory

Proportional

composition

similarity

bearing and line tree citation)

(based

two

was used to
of

the

two

on

relative

was 33%.

Presettlement disturbance regimes
The line descriptions and the township plat map for the
UMBS study area contained no evidence of past fire. Two line
descriptions

referred

to

large

windfalls

("timber

much

72

Table 3.2.
Presettlement forest composition of the UMBS and
Huron study areas.

UMBS
Species

Acer rubrum
Acer saccharum
Betula papyrifera
Facius qrandifolia
Pinus banksiana
Pinus resinosa
Pinus strobus
Pinus sp.—
Populus sp.—
Ouercus alba
Ouercus rubra—
Tsuaa canadensis

Huron

Percentage of bearing and line trees

3.8
1.3
3.8
21.5
--3.8
11.4
6.3
1.3
--2.5
44.3

(3) —
(1)
(3)
(17)
(3)
(9)
(5)
(1)
(2)
(35)

5.4
----3.6
17.1
45.1
15.3
--7.2
3 .6
2.7

(6)

(4)
(19)
(50)
(17)
(8)
(4)
(3)

^Values in parentheses are numbers of individuals.
— The lack of P. banksiana citations at UMBS suggests that
these individuals were P. strobus and P. r e s i n o s a .
— Site characteristics suggest these were mostly Populus
grandidentata.
dMay include some Ouercus velutina (Lam.) and Ouercus
ellipsoidialis (E. J. H i l l ) .

73

fallen")
Thus,

no

but the extent of the disturbance was not recorded.
estimate

of disturbance

recurrence

interval

could

be calculated.
The line descriptions for the Huron study area did
not

indicate

recent

evidence

possibly

13.3%

the

of

thicket"
indicated

or

burns.

They

indicating

section

lines

"oak t h i c k e t " .

that

they

contained,

older
fell

Some

included

however,

burns.

within

many,

small

Approximately

"aspen

references

some

and

to these

pine
areas

i n d i v i d u a l s . The

township plat maps corroborated the locations and extent of
these

areas.

Assuming

these

within the older pine matrix,
the

presettlement

study area was

fire

114-227

areas

recovering

burns

which they may not have been,

recurrence
years.

were

interval

for

the

Huron

There were no references

to

large windfalls within the Huron line descriptions.

Present-day overstory composition
Both the

UMBS

bigtooth

aspen

species

was

successional
locations

and Huron overstories

(Table
74%

at

species

included

3.3).

Relative

UMBS

and

of minor
red

maple

basal

73%

at

importance
and

red

similarity of present-day overstories

were dominated
area

Huron.
occurring

oak.

(m2/ha)

and density

(stems/ha)

h igher in the Huron forest,

this
Later

at

both

Proportional

for the two locations

(based on species relative basal areas) was 90%.
area

for

by

Mean basal

were both substantially

than at UMBS

(Table 3.3).

Table 3.3.
Present-day forest composition and structure of the UMBS and Huron
study areas.

UMBS

Canopy—
Species

m 2/ha

stems/ha

Sapling—
stems/ha

Seedling—
stems/ha

Pooulus arandidentata
Ouercus rubra
Acer rubrum
Betula oaovrifera
Pinus resinosa
Pinus strobus
Facrus arandifolia
Abies balsamea

21.3
2.5
2.4
0.9
0.9
0.6
0.02

(1.2)—
(0.6)
(0.4)
(0.3)
(0.4)
(0.2)
(0.02)

560
67
93
46
18
45
3

(41)
(20)
(23)
(12)
(10)
(15)
(3)

4
82
335
54
44
642
17
3

(4)
(23)
(69)
(13)
(20)
(127)
(7)
(3)

573
4740
64687
—
—
1562
104

(225)
(635)
(8805)
—
—
(569)
(71)

Totals

28.6

(1.5)

890

(44)

1180

(112)

71666

(8777)

Table 3.3

(cont'd).

Huron
Canopy a
Species

m 2 /ha

Populus arandidentata
Ouercus rubra
Acer rubrum
Ouercus velutina
Ouercus alba
Fraxinus americana
Betula papvrifera
Prunus serotina

30.3
5.5
4.9
0.5
0.3
0.06
0.05
0.01

(2.1) —
(1.0)
(0.6)
(0.2)
(0.2)
(0.06)
(0.05)
(0.01)

Totals

41.6

(2.0)

stems/ha

Sapling—
stems/ha

Seedlingstems/ha

638
166
282
14
10
4
2
2

(48)
(29)
(28)
(7)
(5)
(4)
(2)
(2)

13 (6)
644 (89)
------4 (4)
---26 (9)

3625 (669)
64833 (6522)
---458 (260)
167 (97)
42 (42)
14583 (3813)

1117

(48)

686

83708

(96)

(8205)

^Dbh > 10 cm; — 2.5 cm < dbh < 10 cm; ^dbh < 2.5 cm.
— Values are the means (+ SE) of 16 observations at UMBS and 20 observations at Huron.

76

Most red oak and
origin.
genets

red maple were

apparently of

sprout-

Individual ramets were often part of multi-stemmed
that

had

Additionally,

branched

many

at

ramets

or

below

were

the

growing

soil
in

arrangement or had crescent shaped stem bases,

surface.

a

circular

both of which

are indicative of development around remnant s t u m p s .
Regeneration
White pine
3.3).
layer

Red
and

dominated the

maple

was

sapling

moderately

dominated

the

seedling

included

red

oak,

components

layer

abundant
layer

American

at UMBS
in

the

(Table
beech,

(Betula papyrifera M a r s h . ) and red pine.

understory

(Prunus

serotina Ehrh.)

components.

The

3.3).

Minor

paper

birch

(Table 3.3).

and red oak were minor

sapling and

seedling

layers

the two locations had proportional similarity indices
on

mean

relative

Mean total

densities)

of

sapling density was

30%

sapling

The vast majority

of Huron saplings and seedlings were red maple
Black cherry

(Table

and

84%,

substantially

of

(based

respectively.
high e r

in the

UMBS forest.
DISCUSSION
Presettlement forests
The
forests

compositions
were

of

markedly

the

UMBS

different,

soil type and surficial geology.
pine

community

The study

area

on

the

occurred

Huron

and

despite

presettlement

similarities

in

The occurrence of a mixed-

landscape

on sites

Huron

is

that were

not

surprising.

edaphically

and

physiographically similar to the sandy uplands described by

77

Whitne y

(1986).

The forests of these sites were also mixed-

pine communities.
maintained

by

Pine dominance

periodic

hardwood vegetation,

fires

thinned

is believed

that

the

to h ave

eliminated

overstory,

mineral seedbed for new pine establishment

been

competing

and

provided

a

(Van Wagn e r 1970;

Heinselman 1973; W hitney 1986).
The

Huron

landscape

setting

conducive

located

adjacent

presettlement

to

was

fire.

to

the

large-fire

(Simard and Blank,

located
The

in

Mack

study

area

(4,000+

ha)

a

Lake
on

physiographic
outwash

the

plain,

east,

interval

of

had

35

a

years

1982). Fires spreading up from the plain,

with the prevailing westerly winds, may have predisposed the
study area itself to fire, but at a lower frequency than the
plain

proper.

estimate

of

landscape
(1986)
he

This

the

examined.

however,

be

and

pine

years)

Huron

fire

supported

interval
was

for

similar

1973;

that

may

Whitney,

by

the

to

recurrence

be

1987,

interval

Huron

forests
should,

it was based

types

indicative
1990;

the

Whitney's

for the mixed-pine

of particular vegetation

"thickets")

(Heinselman,

(aspen,
of

on

oak,

past

fire

et

al. ,

Nowacki

rather than the presence of obvious burns themselves.
presettlement

consistent

with

hemlock-pine
The

y e a r s ) , which

is

interpreted with caution since

occurrence

A

recurrence

(129-258
The

the

1990),

fire

(114-227

estimate

interpretation

Kilburn's

and

importance

hemlock-beech-pine
(1957;

hemlock-beech

of

hemlock

and

forest

1960b)

UMBS

is

reconstruction

of

forests

for

beech,

both

the

at

same

area.

fire-sensitive

78

species

(Barnes

presettlement

and

fire

Wagner,

1981),

frequency

at

suggests

UMBS

may

that

have

the

b een

low,

despite edaphic and physiographic characteristics that have
b een

associated

(Whitney,

with

1986).

supported

by

fire

elsewhere

in

northern

Michigan

An interpretation of low fire frequency is

the

lack

of

evidence

for

fire

in

the

presettlement line descriptions.
The UMBS
setting

study area was located within a physiographic

that

may

have

presettlement fire.

reduced

area

would

have

fires

from

these

directions,

fire

originating

Heinselman

(1973)

vegetation

was

the

west

often

of

acted

as

barriers

of

the

(1984)

associated

the

of

spread

hemlock
study

and

area

of

beech

proper.

found that fire-sensitive

with

the

landscapes prone to

the

to

protecting

outside

and Grimm

water bodies within
to

likelihood

The lakes to the north and south of the

study

from

the

study area would

leeward

fire.
have

sides

The

of

moraine

acted

as

an

additional barrier to fires moving east, with the prevailing
westerly winds.

Much

was

a

covered

by

dominated northern

of the moraine,

sugar

maple

hardwood

prior

(Acer

forest

to

settlement,

saccharum

(Kilburn,

M a r s h . )-

1957).

Mesic

hardwood forests are believed to be quite resistant to
(Vogl,

1967;

Canham

1987)

and

thereby

I n t e r e s t i n g l y , the
large

burnt

moraine

area

outside

and

Loucks
act as

township
abutting
of

1984;

natural fire

plat
the

Grimm

map

at

northwest

the study

area.

1984;

fire

Whitney

barriers.

UMBS

did show

terminus
Finally,

of

a

the

fires

79

originating

in

the

lowland

conifer-hardwood

east of the study area were probably
estimated that the fire recurrence
type

was

3000-6000

years;

the

forest

to

the

rare.

Whitney

(1986)

interval

for this

forest

longest

of

any

northern

Michig a n forest type he examined.
The

presettlement

compositional

differences

between

study areas may have been related to differing natural
regimes,

however,

there

interacting,

explanations.

have

related

been

characteristics
(for

example,

till,

or

Climate

the

may

statistics
somewhat

sand

apparent
particle

have

(Table 3.1)
cooler

Eastern

hemlock

extent,

are

(Godman

and

addition

in

of

been

possibly

distinctions

may

physical

site

in

than

landform

and

size

distribution,

f iner-textured
influential.

soil

type

depth

soil

Modern

to

bands).
climatic

suggest that the UMBS study area was

and

often

wetter

than

particular,
associated

Lancaster,

to

differences

frequency

also

alternative,

Compositional

to

less

are

fire

1990;

influencing

the

and
with

Tubbs

Huron

study

area.

American

beech

to

a

humid

climate

cool,

and

composition

Houston,

directly,

some

1990).
a

In

cooler,

wetter climate at UMBS could also have contributed to lower
fire frequency.

Present-day forests
Early
of

the

s u c c e s s i o n a l , postsettlement

two

differences

study
in

areas

was

presettlement

forest

similar,
forest

composition

despite

composition.

marked
This

80

convergence was

a function

proliferation;

it

postsettlement

disturbance

recurrent
pines,

of both

occurred

wildfire).

because

These

hemlock and beech.

and

red

maple)

availability

of

stumps
oak

able

and

1990).

rare,

in

(Kilburn,

followed

eliminated

in
by
most

landscapes

(bigtooth aspen,

best

mineral

able

seedbeds

to

exploit

(bigtooth

red
the

aspen;

or propagate vegetatively from roots or

survive

frequent

Gates,

fire

1930;

(bigtooth

Hutnick

and

aspen,

red

Y a w n e y , 1961;

The resulting forest type was u n r e p r e s e n t e d ,
the

1957,

There

(logging

disturbances

those

abundant

red maple;

Sander,
or

to

similarities

and

The species that did proliferate

were

Graham et a l . , 1963)

of

regimes

in these early successional
oak

species elimination

presettlement

1960b; Whitney,

were

differences

landscapes
1986,
in

of

the

region

1987) .
composition

of

tree

regeneration between the two l a n d s c a p e s ; p rimarily a lack of
pine in the sapling layer of the Huron stands.

At least one

or two large remnant overstory white pines were observed in
all UMBS

stands

found

the

in

graphically
source
Seed

to

(although not

Huron

stands

illustrates
species

source

in any plots) , but

(formerly

p i n e - d o m i n a t e d ) . This

the

importance

recovery

following

elimination

may

also

none were

be

of

a

remnant

major

disturbance.

responsible

failure of hemlock to recover at U M B S . Additionally,
recovery may
by

be

white-tail

Curtis,

1959;

inhibited because
deer

(Odocoileus

Anderson

and

of

preferential

vircrinianus

Loucks,

1979;

seed

for

the

hemlock
browsing

Zimmerman;
Frelich

and

81

Lorimer,

1985)

substrates
Coffman,

or a lack of coarse,

(Oosting

and

Hess,

1978; Hix and Barnes,

The

present-day

purely speculative.

The

1956;

Olson

et

al.,

1959

;

1984).

forests

structurally different.

moist woody germination

of

the

two

possible

landscapes

reasons

were

for this

Lower overstory density and basal

are
area

at UMBS may be related to the intensity of past disturbance
or

to

site

differences.

UMBS forest,
turn,

to

Lower

overstory

stocking

relative to the Huron forest,

increased

understory

in

the

may have led,

establishment

and

in

greater

recruitment to sapling size c l a s s e s .
The loss of bigtooth aspen from the overstories of both
study areas
the

limit

is imminent,

of

Great Lakes

the

region

species

ramets

pathological

are

approaching

rotation

age

in

the

(70-80 y e a r s ; Graham et al. , 1963) . New

overstory recruitment
species

since most

of bigtooth aspen,

(Laidly, 1990),

is dependent

on

a shade-intolerant
complete

overstory

removal through either silvicultural manipulation or natural
stand-replacing
now

fire

(Graham et a l . , 1963).

uncharacteristic

suppression

(Whitney,

of

the

1987).

region
In

The

latter

because

contrast,

red

is

of

fire

oak

will

remain an overstory component of both forests after loss of
bigtooth

aspen

simply

because

(3 0 0 + y e a r s ; Barnes and Wagner,

it

and

Y a w n e y , 1961)

and

a

long-lived

species

1981).

Bigtooth aspen mortality will
of numerous canopy gaps.

is

result

in the

Shade-tolerant red maple

mid-tolerant

white

pine

formation
(Hutnick

(Wendel

and

82

Smith,

1990)

recruitment
Hibbs,

both

in

have

canopy

1982) .

potential

gaps

Increased

species could occur,
characterized

the

by

(Gerrard,

for

1969;

overstory

release

Lorimer,

recruitment

to

successional

1980b;

for

moderate

trends

or

in

sized

toward

combination

canopy

trends

have

several

landscapes
been

areas

both

gaps.

increasing

red

with

pine,

white

signal the development of forest types unrepresented
presettlement

and

given a natural disturbance regime now

small

i m p o r t a n c e , alone,

potential

of

suggested

of the

the

upper

for

Lake

States.

oak-dominated

eastern United

States

The

maple
may

in the
Similar

forests

(Lorimer,

in

1984;

Host, et a l . 1987; Abrams and N o w a c k i , 1992).
The cultural histories of the UMBS and Huron landscapes
may

have

long-lasting

effects

communities.

Continued

seed

on

sources

composition

will

presettlement
may

postsettlement
by

differential

study

remain

the

area

markedly
A

reduced

presettlement

forest
the

their

composition

availability

ability

of

respective

forest

fire exclusion and the lack of pine

Huron

condition.

reflect

influenced

the

on

various

of

assures

that

different
fire

from

the

frequency at UMBS

condition.
has
seed

forest

However,

been

heavily

sources

presettlement

and

species

the
to

recover following fire.
The results of this study illustrate the
the

link

between

disturbance

and

importance of

composition

in

plant

c o m m u n i t i e s ; a relationship that is now widely recognized by
ecologist.

The

results

also

illustrate

how

historical

83

circumstances
pathways.

can

influence

Ecologists

have

composition
recognized

and

for

some

historical factors m a y influence composition
Cain,

1947;

Whittaker,

Auclair and
Recently,

Cottam,

the

reemphasized
Pickett,
forests
how

1971;

importance
and

1989).
of

pre-

The

Egler,

Oliver,

1981;

current

and postsettlement
historical

Curtis,

1959;

influences

and

that
1926;

Heinselman,

examined
study,

time

(Gleason,

1954;

of historical

critically

human-associated

important,

1953;

successional

(for

1981).

has

been

example

see

others

landscapes,
circumstances

comparing
illustrate
can

long-lasting effects on forest composition.

have

BIBLIOGRAPHY
Abrams, M. D. and G. J. N o w a c k i . 1992.
Historical variation
in fire, oak recruitment, and post-logging accelerated
succession in central Pennsylvania.
B u l l . Torrev B o t .
C l u b , (in p r e s s ) .
Auclair, A. N. and G. Cottam. 1971. Dynamics of black cherry
(Prunus
serotina
E r h r .)
in southern Wisconsin
oak
forests.
E c o l . M o n p g r ., 41: 153-177.
Albert,
D. A . , S.
R.
Denton and B. V.
Barnes.
1986.
Regional landscape ecosystems of Michigan.
School of
Natural Resources, University of Michigan, Ann Arbor.
32 p.
Anderson, R. C. and D. E. A d a m s . 1978.
Species replacement
patterns in central Illinois white oak forests, p. 284301. In: P. E. Pope (ed.).
Proceedings of the central
hardwood forest conference.
Purdue University,
West
Lafayette, Indiana.
A n d e r s o n , R. C . and O . L. L o u c k s . 1979. White-tail deer
fOdocoileus v i r g i n i a n u s ) influence on structure and
composition of Tsuga canadensis f o r e s t s . J. A p p I .E c o l .,
16: 855-861.
B a r n e s , B. V. and W. H. Wagner Jr.
1981. Michigan t r e e s .
University of Michigan Press, Ann Arbor.
383 p.
Barnes, B. V . , Pregitzer, K. S., Spies, T. A . , and Spooner,
V. H.
1982.
Ecological forest site classification.
J. F o r . . 80: 493-498.
Benninghof f ,
W.
S.
and
K.
J.
Cramer.
1963.
Phytosociological
analysis
of aspen
communities
on
three site classes for Populus grandidentata in western
Cheboygan County, M I . V e g e t a t i o . 11; 253-264.
B o u r d o , E. A. , Jr.
1954.
A validation of methods used in
analyzing original forest cover.
Ph.D. dissertation,
University of Michigan, Ann Arbor.
207 p.
________ .
1956.
A review of the General Land Office survey
and of
its use
in quantitative studies of
former
forests. E c o l o g y . 37: 754-768.
Brower, J. E. and J. H. Z a r .
1977. Field and laboratory
methods for general ecology.
W. C. Brown Publishers,
D u b u q u e , Iowa.
226 p.
B u e l l , M. F. , A. N. Langford, D. W. Davidson and L. F.
O h m a n n . 1966.
The upland forest continuum in northern
New Jersey. E c o l o g y . 47: 416-432.

84

85

Cain,

S. A.
1947.
Characteristics of natural areas and
factors in their development.
E c o l . M o n o a r ., 17: 185200.

Canham, C. D. and O. L. Loucks. 1984. Catastrophic windthrow
in the presettlement forests of Wisconsin. E c o l o g y . 65:
803-809.
Coffman, M. S.
1978. Eastern hemlock germination influenced
by light,
germination media,
and moisture content.
M i c h . B o t . . 17: 99-103.
Cooper, A. W.
1981. Above-ground
biomass accumulation and
net primary production during the first 70 years of
succession in Populus grandidentata stands on poor
sites in northern lower Michigan, p. 339-360. In: D. C.
West, H. H. S h u g a r t , and D. B. Botkin, (eds.) . Forest
succession: concepts and application.
Springer-Verlag,
New York.
Crow,

T. R.
1988. Reproductive
mode and mechanisms for
self-replacement of northern red oak (Ouercus r u b r a )- a
review. F o r . S c i ., 34: 19-40.

Curtis, J. T. 1959. The vegetation of Wisconsin.
of Wisconsin Press, Madison.
657 p.
Dix,

R.
L.
1957. Sugar
maple
forest
Washington, D. C.
E c o l o g y . 38: 663-665.

University

succession

at

E g l e r , F. E. 1954.
Vegetation science concepts.
I . Initial
floristic composition, a factor in old-field vegetation
development.
V e g e t a t i o . 4: 412-417.
Foster, D. R.
1983.
The history and pattern of fire in the
boreal forest of southeastern labrador.
C a n . J . B o t ..
61: 2459-2471.
Fralish, J . S., F. B. Crooks, J . L. Chambers, and F. M.
Harty.
1991.
Comparison of p r e s e t t l e m e n t , secondgrowth and old-growth forest on six site types in the
Illinois Shawnee Hills.
A m . M i d i . N a t . . 125: 294-309.
Frelich,
L.
E.
and
C.
G. Lorimer.
1985. Current
and
predicted long-term effects of deer browsing in hemlock
forests in Michigan, U.S.A.
B i o l ■ C o n s ., 34: 99-120.
Gates, F. C.
Michigan.

1930.
Aspen
Bot G a z ., 90:

association innorthern Lower
233-259.

Gerrard, D. G.
1969.
Competition quotient: a new measure
of the competition affecting individual forest trees.
M i c h . State U n i v . A g . Exp. S t a . R e s . B u l l .. 20.
32 p.

86

Gleason, H. A.
Michigan.

1923.
Botanical observations in northern
J. New York B o t . C l u b . 24: 273-283.

________ .
1926.
association.

The individualistic concept of the plant
B u l l . Torrev B o t . C l u b . 53: 212-229.

Glitzenstein, J. S., C. D. Canhaxn, M. J. McDonnell and D. R.
Streng.
1990.
Effects of environment and land-use
history on upland forests of the Cary Arboretum, Hudson
Valley, New York.
B u l l . Torrey B o t . C l u b . 117: 106122.
G o d m a n , R. M. and K. Lancaster. 1990.
Tsuaa canadensis (L.)
Carr. (Eastern h e m l o c k ) , p . 604-612.
I n : R. M. Burns
and B. H. Honkala,
(tech. c o o r .) . Silvics of North
America, Vol 1.
U. S .D. A. For. S e r . Ag. Hand. , 654.
675 p.
Graham, S. A. , R. P. Harrison, Jr. and C. E . W e s t e l l , Jr.
1963.
A s p e n s : phoenix trees of the Great Lakes region.
University of Michigan Press, Ann Arbor.
272 p.
Grimm, E. C.
1984.
Fire and other factors controlling the
Big Woods vegetation of Minnesota in the mid-nineteenth
c e n t u r y . E c o l . M o n o a r . . 54: 291-311.
Heinselman, M. L.
1973.
Fire in the v i rgin forests of the
Boundary Waters Canoe Area, Minnesota. O u a t . R e s . . 3:
329-382.
________.
1981.
Fire and succession in the conifer forests
of northern North America, p. 374-405.
In: D. C. West,
H.
H.
S h u g a r t , and
D.
B.
botkin,
(eds.).
Forest
succession.
S p r i n g e r - V e r l a g , New York.
Hibbs, D. E.
forest.

1982.
White pine in the transition hardwood
C a n . J. B o t . . 60: 2046-2053.

Hix,

D. M. and B. V. Barnes. 1984.
Effects of clear-cutting
on
the vegetation and soil of an eastern hemlock
dominated ecosystem, western Upper Michigan. C a n . J.
F o r . R e s ., 14: 914-923.

Host,

G. E . , K. S. P r e g i t z e r , C. W. Ramm, J. B. Hart and D.
T. Cleland.
1987. Landform-mediated differences in
successional pathways among upland forest ecosystems in
northwestern Lower Michigan.
F o r . S c i . , 33: 445-457.

Hutnick,
R.
J.
and
H.
W.
Yawney.
1961.
Silvical
characteristics of red maple (Ace r r u b r u m ) . U.S.D.A.
F o r . S e r . S t a . P a p ., 142.
18 p.

87

Kilburn, P. D.
1957.
Historical development and structure
of the
aspen, jack pine, and oak vegetation types on
sandy
soils in northern
lower
Michigan.
Ph.D.
dissertation, University of Michigan, Ann Arbor. 267 p.
________.
1960a.
Effect of settlement on the vegetation of
the University of Michigan
Biological Station.
Pap.
M i c h . A c a d . S c i . Arts L e t t ., 45: 77-88.
________.
1960b.
Effects of logging and fire on xerophytic
forests in northern Michigan.
B u l l . Torrey B o t . C l u b .
87: 402-405.
L a i d l y , P. R.
1990.
Populus qrandidentata Michx. (Bigtooth
a s p e n ) , p. 544-550. In: R. M. Burnes and B. H. Honkala,
(tech.
c o o r .).
Silvics
of
North
America,
Vol
2.
U . S .D.A. F o r . S e r . Aq. H a n d . . 654.
877 p.
Lapin, M.
1990. The landscape
ecosystem groups of the
U n i v ersity
of
Michigan
Biological
Station:
classification,
mapping,
and analysis of ecological
diversity.
M. S. thesis, University of Michigan, Ann
Arbor.
178 p.
Lorimer, C. G.
1977.
The presettlement forest and natural
d isturbance cycle of northeastern Maine.
E c o l o g y . 58:
139-148.
________.
1980a.
The
use
of
land
survey
records
in
estimating presettlement fire frequency, p. 57-62.
In:
M. A. Stokes and J . H. Dieterich, (eds.).
Proceedings
of the fire history workshop,
20-24 October
1980,
Tucson, Arizona.
U . S .D.A. F o r . S e r . G e n . T e c h . R e p . ,
RM-81.
________ . 1980b.
Age structure and disturbance history of a
southern Appalachian virgin forest.
E c o l o g y . 61: 11691184.
________.
1984.
Development of the red maple understory in
n ortheastern oak f o r e s t s . F o r . S c i ., 30: 3-22.
M c C u n e , B. and G. Cottam.
1985.
The successional status of
a southern Wisconsin oak woods. E c o l o g y . 66: 1270-1278.
Monk,

C.
D.
1961.
Past
and
present
influences
on
reproduction
in the W illiam
L.
Hutcheson Memorial
Forest, New Jersey.
B u l l . Torrey B o t . C l u b . 88: 167175.

N o w a c k i , G. J. , M. D. Abrams and C. G. Lorimer.
1990.
Composition, structure, and historical development of
northern red oak stands along an edaphic gradient in
north-central Wisconsin.
F o r . S c i . , 36: 276-292.

88

Olson,
J.
S.,
F.
W.
Stearns and H.
N i e n s t a e d t . 1959.
Eastern
hemlock
seeds
and
seedlings:
responses to
photoperiod and temperature.
C o n n . A g r i c . E x p . Stn.
Bull., 620.
Oliver, C. D.
following
153-168.

1981.
Forest development in North America
m a j o r disturbance.
F o r . E c o l . Manage., 3:

Oosting, H. J. and D. W. Hess.
1956. Microclimate and a
relic stand of Tsuga canadensis in the lower Piedmont
of North Carolina.
E c o l o g y . 37: 28-39.
Padley, E. A.
1989.
Associations among glacial landforms,
soils, and vegetation in northeastern Lower Michigan.
Ph. D. d i s s e r t a t i o n , Michigan State University,
East
Lansing.
279 p.
Pickett, S. T. A.
1989.
Space-for-time substitution as an
alternative to long-term studies, p. 110-156.
In: G.
Likens, (ed.). Long-term studies in ecology: approaches
and a l t e r n a t i v e s . Springer-Verlag, N e w York.
P r e g i t z e r , K. S., B. V. Barnes and G. D. Lemme.
1983.
Relationship of topography to soils and vegetation in
an Upper Michigan ecosystem.
Soil S c i . S o c . A m . J. ,
47: 117-123.
R o b e r t s , M. R. and N. L. Christensen.
1988.
Vegetation
variation among mesic successional forest stands in
northern lower Michigan.
C a n . J . B o t ., 66: 1080-1090.
Romme, W. R. and D. H. Knight.
1981. Fire frequency and
subalpine
forest
succession
along
a
topographic
gradient in W y o m i n g . E c o l o g y . 62: 319-326.
Runkle, J. R.
1982. Patterns of disturbance in some oldgrowth
mesic
forests
of
eastern
North
America.
E c o l o g y . 63: 1533-1546.
Sander, I . L.
1990. Ouercus rubra L. (Northern red oak) ,
p. 727-733.
In: R. M. Burnes and B. H. Honkala, (tech.
c o o r . ).
Silvics of North America, Vol 2. U . S .D.A. F o r .
S e r . Ag. H a n d . . 654.
877 p.
Simard, A. J. and R. W. Blank. 1982. Fire history of
Michigan jack pine forest.
M i c h . A c a d . . 15: 59-71.

a

S p r u g e l , D. G.
1976.
Dynamic structure of wave-generated
Abies balsaemea
forests
in the northeastern United
States.
J. Ecol., 64: 889-911.

89
Tubbs, C. H. and D. R. Houston. 1990. Faqus qrandifolia
Ehrh. (American beech), p . 325-332.
In: R. M. Burns and
B. H. Honkala, (tech. c o o r .). Silvics of North America,
Vol 2.
U.S.D.A. For. Ser. Ag. Hand., 654.
877 p.
V an Wagner, C. E.
1970.
Fire and red pine.
Timbers Fire E c o l . C o n f .. 10: 211-219.

Proc. A n n . Tall

Veblen,
T.
T.
and
D.
H.
Ashton.
1978.
Catastrophic
influences on the vegetation of the Valdivian Andes,
Chile.
V e q e t a t i o . 36: 149-167.
Vogl,

R. J.
Wisconsin.
47-96.

1967.
Controlled burning for wildlife in
P r o c . Tall Timbers Fire Ecology C o n f . . 6:

Wendel. G. W. and H. C. Smith.
199 0. Pinus strobus L.
(Eastern white pine), p. 476-488.
In: R. M. Burns and
B. H. Honkala, (tech. c o o r . ). Silvics of North America,
Vol 1.
U.S.D.A. For. Ser. Ag. Hand., 654.
675 p.
Whitney, G. G.
1986.
Relation of Michigan's presettlement
pine forests to substrate and disturbance history.
E c o l o g y . 67: 1548-1559.
________ .
1987. An ecological history of the Great
forest of Michigan.
J. Ecol., 75: 667-684.

Lakes

________ .
1990.
The history and status of the hemlockhardwood forests of the Allegheny Plateau. J. Ecol.,
78: 443-458.
Whittaker, R. H.
1953.
A consideration of climax theory:
the climax as a population and pattern.
Ecol. Monoqr..
23: 41-78.

Chapter 4
PATTERNS OF EVEN-AGED FOREST DEVELOPMENT WITHIN TWO BIGTOOTH
ASPEN (Populus arandidentata M i c h x . )-DOMINATED LANDSCAPES IN
NORTHERN LOWER MICHIGAN, USA

ABSTRACT

Forest
purported
An

to

initial

(stand

development
follow

cohort

a

of

following

fairly
trees

initi a t i o n ) , new

extended

period of

individuals
classes

in the

The current

the

was

landscapes

at the

(Huron

histories.

The

within

and

objectives

precluded

two

an
new

new

age

the generality

controlling

bigtooth

the

and

in even-aged

of

gain
stem

forests.

aspen-dominated

similar

compositions,
for

for

finally,

landscape-level

UMBS) having

overstory

rapidly

(understory r e i n i t i a t i o n ) .

possible mechanisms

conducted

c h a rac t e r i s t i c s ,

relatively

is

is

pattern.

to establish, creating

exclusion and understory reinitiation
Research

temporal

designed to assess

pattern

disturbance

exclusion),

forest understory

study was

into

establishes

(stem

begin

this developmental
insight

repeatable

establishment

time

again

major

physical

and

study

site

disturbance
included

i)

describing forest development patterns within each landscape
and

assessing

within

and

the

between

repeatability
the

two

of

development

landscapes, and

ii.)

patterns
exploring

relationships between the timing of understory reinitiation
and

overstory

availability.

growth
Stem

characteristics
analysis

90

was

used

and
to

seed-source
reconstruct

91

establishment

and

growth

histories

of

sampled

stems

on

replicate plots within each landscape.
The age distributions of surviving individuals
forests
rapid

reflected
initial

a developmental

cohort

establi s h m e n t ,

understory reinitiation.
was

similar

There

both

were,

forest

stem

characterized
exclusion,

by
and

The duration of development stages

within

however,

with

pattern

in both

a

and
small

substantially

between
number

reduced

the
of

two

plots

stem

landscapes.
in

the

exclusion

UMBS

lengths,

relative to the remaining plots in this forest. Vari a t i o n in
the timing of understory reinitiation in the UMBS forest was
related to characteristics of remnant seed trees.
exclusion

period

plots

close

in

was

shorter,

proximity

or

to

nearly

several

The stem

non-existent,

large

seed

on

trees.

Substantial increases in establishment frequency on all UMBS
plots,
the

as

well

Huron

radial

as

forest,

growth

all

new

were

associated

releases

suggest

that

the

forests

examined

in

timing
was

understory

of

establishment

with

overstory

by

to

s t e m s . These

understory

influenced

moderate

an

within

results

reinitiation
interaction

large

in

the

between

seed availability and changing resource levels in the forest
understory.

A limited amount of new establishment occurred

relatively

early

in

experienced

abundant

forest
early

development

seed

rain,

yet

on
the

plots
bulk

that
of

new

establishment occurred after an increase in radial growth of
overstory

stems.

The

radial

growth

releases

presumably

reflected an increase in resource availability in the forest

92

understory,

possibly

occurring

in

response

to

natural

thinning w i t h i n the b i g tooth aspen-dominated overstory.
general
study

patterns
were

forest

strikingly

landscapes.
address

of

Similar

the

similar

Additionally,
unequivocally

of

studies

the

influence

of

should

seed-source

be

and

understory establishment patterns.

resource

this

between

at

to
the

forest-types.
needed

controlling

in even-aged

given

in

needed

patterns

are

mechanisms

and

are

additional

experimental

attention

within

studies

exclusion and understory reinitiation
Particular

described

development

within

identify

both

descriptive

repeatability

landscape-level

development

The

to

the

to
stem

forests.
combined

availability

on

93
INTRODUCTION

Even-aged
attention
Bormann

forest development has

over
and

the

last

Likens

several

1979;

received

decades

Oliver

considerable

(for

1980,

example,

1981;

see

Peet

and

Christensen 1987).

This interest can be attributed in part

to

of

the

prevalence

forest
scale

an

even-aged

l a n d s c a p e s , but also
natural

forests

in

Oliver

disturbances

many

1981).

thought

to

(Oliver

1981;

to the

regions

The

have
(see

a

Peet

fairly
1981;

First,

of

established

period

becomes

of

time

(stand

managed

of

class

Lorimer

even-aged

others

large-

single-age
in

repeatable

Christensen 1987).
trees

examples

and

in

realization that

created

development

follow

structure

198 0;

forests

temporal

cited

in

is

pattern
Peet

and

an initial post-disturbance cohort
during

initiation

a

sensu

relatively
Oliver

short

1981).

A

second period follows during which few or no new individuals
become established

(stem e x c l u s i o n ) . Finally,

individuals of

overstory tree species again begin to establish s u c c e s f u l l y ,
creating

new

age

classes

in

the

forest

understory

(understory r e i n i t i a t i o n ) .
Many

silvicultural

studies

of

even-aged

forest

development have explored growth patterns within the initial
post-disturbance
1967,

1981;

W ierman
1985;
Hix

and

Oliver
Oliver

Kelty 1986;
and

cohort

Lorimer

of

1978;
1979;

trees

Stubblefield
Hibbs

Larson 1986;
1990,

(Cayford

1991;

1983;

1957;

and

Oliver

Guldin

and

Marquis
1978;
Lorimer

Clatterbuck and Hodges
Deal

et

al.

1991).

1988;
Less

94

attention

has

understanding,
exclusion

been
from

and

given

a

to

mechanistic

understory

characterizing
perspective,

reinitiation

phases

and

the

of

stem

even-aged

forest development. While one or both of these developmental
stages

have

types

been

documented

(Bloomberg

1950;

Christensen

and

Roberts

Richardson

1989;

and

Abrams

studies
stand

Peet

been

characterized
species

1981;

Peet

1985;

to

patterns

by

similar

compositions,

wide

1981;

that

disturbance
exclusion

some

have

period

1986;

a t . 1985;

Sharik

et

al.

1992),

few

repeatability

and

among

site

disturbance

of

landscapes

c h a r a c t e r i s t i c s,
histories

(for

a

The need for assessing

patterns

is demonstrated by

initiating

been

1981;

et

the

forests

(Peet

Oliver

1978;

assess

physical

actually

Oliver

Snook

the generality of developmental
fact

forest-

and

notable exception see Sprugel 1976).

the

of

Segura

within

and

variety

197 6;

Harcombe

1992;

designed

development

a

Sprugel

and Nowacki

have

in

found

after

to

Carleton

lack

1982;

major
a

stem

Roberts

and

regarding

the

Richardson 1985; Chapter 1).
There

is

mechanisms

that

reinitiation
that
local

stem

in

also

a

lack

control

stem

even-aged

exclusion

seed-sources

may

of

exclusion

forests.

simply

early

consensus

in

Some

and

understory

studies

suggests

reflect

an

initial

lack

forest

development,

of

while

understory reinitiation reflects the reproductive maturation
of

a

post-disturbance

cohort

of

trees

1982; Roberts and Richardson 1985;

(Day

1972;

Carleton

Sharik et aJL 1989;

Sakai

95

et

al.

1985;

Sakai

1990).

In

theories of forest development
Peet

1981;

Peet

1981;

Peet

contrast,

resource-based

(Oliver 1981; Christensen and

and

Christensen

1987)

suggest

that stem exclusion begins after the initial cohort of trees
attains

full

site

occupancy,

thereby

establishment through competition,
local

seed-source.

begin

only

when

mechanisms have
in

resource

Understory
resources

availability,

reinitiation

late

Christensen and
1987),

in

to account

including

Peet

even-aged
1981;

canopy

1981;

to

Various
increase

gap

formation

of mature

overstory

(Sprugel

Peet and

1976;

Christensen

crown differentiation within the even-aged overstory

(Bormann

and

Likens

1979,

Oliver

et

al.

1985;

Larson 1990) , reductions in rates of root growth
Larson

thought

for this

development

Peet

is

less-limiting.

following density-independent mortality
individuals

successful

despite the presence of a

become

been proposed

limiting

1990),

individuals
thinning

Oliver

and

(Oliver and

and density-dependent m o rtality of suppressed
following

a

concentrated

wave

of

natural

(Harcombe 1986; Fried et al 1988; O l i v e r and Larson

1990) .
This paper

reports on

a mensurative

study designed

to

assess the generality of developmental patterns in even-aged
forests
possible

at

the

landscape-level

mechanisms

controlling

understory reinitiation.
bigtooth

aspen

and

(Populus

gain

stem

insight

into

exclusion

and

The study was conducted w i t h i n two
arandidentata

landscapes in northern Lower Michigan,

USA.

M i c h x . )-dominated
Both landscapes

96

are

representative

of

aspen-dominated

forests

that

occupy

over 5 million hectares in the Great Lakes region

(bigtooth

and

combined,

trembling

Einspahr

and

reconstruct
all

patterns

The

establishment
in

patterns,

within

tremuloides

1990).

species

development

(P.

Wyckoff
the

tree

aspen

and

initial

history

these
and

objective

of

the

describe

the

two

was

individuals

forests,

assess

between

Michx.)

from

forest

repeatability

landscapes.

to

A

of

second

objective was to explore relationships between the timing of
understory reinitiation and overstory growth characteristics
and seed availability.

A working hypothesis associated with

this

that

objective

controlled

by

associated with
overstory.
that
onset

i)

was
an

increase

natural

Testable

understory

of

understory

understory
in

reinitiation

mortality

in

reinitiation

resource

thinning within

corollaries

mature,
would

be

reinitiation

to

the

this

would

associated

availability
bigtooth

aspen

hypothesis

were

begin

overstory

prior

resource

availability

survivorship

in

understory

reinitiation

characteristics

the

of

as

understory,

remnant

size) when they were present.

a

and

with

would
seed

trees

factor
iii)

be

to

dominants,

radial growth in surviving overstory individuals,
increased

is

the

the
ii)

increased
suggesting

influencing
timing

unrelated
(proximity,

of
to

number,

97

STUDY LOCATIONS
Resea r c h
National

was

Forest

conducted
(Huron)

with i n

in

the

northeastern

Huron-Manistee
Lower

Michigan

(latitude 44° 15' to 45° 00' N, longitude 83° 15' to 84° 45'
W)

and the University of Michigan Biological Station

in

the

extreme

northwestern

(latitude 45° 40 N,
are

separated

soils,

and

by

the

portion

of

Lower

approximately

physiographic

Prior to settlement,
(Pinus

L a m . ) and white pine

150

and

km.

Surficial

climatic

beech

geology,

settings

of

the

(Chapter 3).

the Huron landscape was dominated

resenosa A i t . ) , jack pine

(P.

banksiana

(P. strobus L . ). The UMBS landscape was

dominated by eastern hemlock
A merican

Michigan

longitude 84° 40' W ) . The two landscapes

two study areas have been described previously

by red pine

(UMBS)

(Fagus

(Tsucra canadensis

qrandifolia

E h r h .),

(L.) C a r r . ) ,

white

pine,

and

red pine. Minor species common to both presettlement forests
included bigtooth aspen,

red maple

oak

A more detailed description of the

(Ouercus rubra L . ).

(Acer rubrum L . ), and red

presettlement forests of both areas can be found in Chapter
3.
The regions containing the study areas were deforested
in

the

late

(Gleason,

1923;

nineteenth
Kilburn,

C r a m e r , 1963; Whitney,
over

these

landscapes

to
1957,

early

twentieth

1960a,b;

centuries

Benninghoff

and

1987). Slashed-fueled wildfires swept
in

the

years

following

logging,

eliminating advanced reproduction and most remnant overstory
individuals

of the dominant presettlement

species

(Kilburn,

98

1957,

1960ab;

promoted
rapidly

Whitney,

vigorous

1966),

clonal

which

strictly even-aged
study

area

did

area

root

led,

not

systems

in

forests

turn,

any

fires

aspen

(Graham et al.

to

the

remnant

species.

red pine were

of

the

from
1963;

development

(Graham et a l . 1963).

contain

dominant presettlement
pine and

Additionally,

vegetative-regeneration

spreading

Barnes

1987).

The Huron

individuals

In contrast,

scattered throughout

of

of

the

remnant white
the UMBS

study

(Chapter 3).
METHODS

Plot selection and vegetation sampling
Vegetation
within

five

UMBS.

was

sampled

stands

Stands

in the

within

on

Huron

each

land

a l . , 1982;

Pregitzer and Barnes

following

from

obvious

slopes.

from

located

used

1-4.5

within

forest
at

UMBS

i)

disturbance

The distance

ranged

Huron

classification

criteria:

Four

475

so

that

m 2 at
a

and

four

size

of

similar

to

analyses.
m)

of

all

each

was

number

of

randomly

m2

plot

in

the

size

was

overstory

2.5

cm dbh

were

stems
(see

in all statistical

The species and diameter at breast height
>

0-5%

landscape

were
272

larger

Plots were used as replicates

stems

et

free

i_ii)

might be sampled as in the higher density Huron forest
results).

at

similar

Barnes

and

plots

size
A

stands

1 h a ; ii)

within

circular

UMBS.

located

Selected stands met

initiation;

Plot

plots

(following

1984).

since

of

belonged

units

m inimum

stand.

and

forest

between stands

km.

each

series

landscape

ecological

the

a

recorded

on

(dbh=l.4

each

plot.

99

Seedling
plots,

(dbh< 2.5 cm)

spaced

numbers were recorded in twelve 1 m 2

equidistantly

at

3

m

intervals

along

four

opposing radii of each plot, beginning at the p lot boundary.
Seedlings were assigned to one of two h e ight classes;

< 1.5

m and > 1.5 m.

Destructive sampling
On each plot,
all

individuals

one randomly selected b i g tooth aspen,

>

1.5

m

tall

of

additional

tree

species,

were destructively sampled to determine total heights,
of establishment,
selected s t e m s .
the tallest
For

and rates

crown

codominant)

was

one

randomly

decay)

Sampling

intensity

observed

and

growth

from m u l t i - s t e m m e d

individual
(most

selected

bigtooth

because

from

ramets

the
wer e

(rejecting

destructively

for

species

among

sampled

classes

internal

other

were

aspen,

codominant

for

and h e i g h t

times

For species other than b i g t o o t h aspen,

ramets

bigtooth

of radial

sampled

aspen

little

dominant-codominant

was

ramets

little age variation was expected.

of

only

genets.

domin a n t

or

dominant

or

ramets

on

each

m uch

height

and

with
plot.

lower

than

variation

was

within

plots

and

The age dis t r i b u t i o n of

sampled bigtooth aspen substantiated the

latter expectation

(see r e s u l t s ) .
On each plot,
was

sampled

classes.
attempt

by

Stems
to

a subsample

stratifying
<

1.5 m

determine

the

on

of

individuals

species

tall were
age

range

and

sampled
of

< 1.5 m tall
0.5

m

height

p r imarily

these

stems

on

in an
each

100

plot, particularly their maximum ages.
Stems were
on

size)

were

and

cut

total

at

(generally
additional
individuals

those

greater

or ground

were

level

section was

level

determined.

from

all

than

m

1.4 m

(depending

Stem

larger

2.5

removed at

(dbh > 10 cm)

in

sections

individuals
height).

from all

An

overstory

for use in radial growth analysis.

all bigtooth aspen stems were marked at 0.25-m

from

thereafter.

0.5 m

heights

ground

Additionally,
intervals

felled at

the

base

Marking

to

2

m

and

continued to the

to the end of the dominant

leader.

at

1-m

intervals

1-m multiple

closest

Stem sections were cut

at each measurement interval up to a 3 cm diameter top.

All

destructive sampling was completed during summer and fall of
1990.

A ge determination
Small stems of deciduous tree species
<

2.5

m

tall)

and

the

terminal

leaders

(generally those

of

bigtooth

aspen

were aged by counting the number of terminal bud scale scars
preceding each
species

(white

b r anch whorls.
counts
stem

was

height
pine

and

red

by

dissecting

pine)

counting

stems

were

of

aged

rings

on

basal

(typically the most difficult

stem to age by these methods)
a

Small

coniferous

by

counting

The accuracy of bud scale and branch whorl

checked

sections

interval.

microscope.

and

0.25

portion of

m

the

on a subsample of stems under

Age

determinations

for

the

two

methods seldom differed by more than ± 2 years and then only

101

on individuals > 2 cm basal diameter.
Basal
to

a

stem

smooth

sections

surface

of

and

larger

wetted

individuals were

to

aid

ring

sanded

examination.

Ages of all sections were determined by counting ring number
on

at

least

two

oblique-shaped
oldest

age

radii
stems)

under

each

section

of

differed.

(typically

Precision

of

periodically recounting
of

15-25

than

sections

1 year.

paper

birch

was

20% of

(Betula

counting
the stem

papvriferia

if

was

made

axis

the

accurate

Marsh.)

counts

from

differed

on
The

assessed

sections

seldom

rings

short

microscope.

recorded

Recounts

annual

and

dissecting

ring

each.

Faint

a

a long

by
lots

by more

ageing

impossible.

of
All

genets of this species were apparently of vegetative-origin
(all multi-stemmed)

and were

the sampled bigtooth aspen.
important

component

on

balsam

fir

sample plots

(Abies

Terminal
ages of basal

bud

of

balsamea

scale

the

white ash

(see results)

equivalent

age

as

F u r t h e r , paper birch was not an

any

r e s u l t s ) . American beech,
and

probably of

sample

plots

(see

(Fraxinus americana L . )

L . ) were

also

rare

in

the

and were excluded from analysis.
c o u n t s , branch

whorl

counts,

and

stem sections were used to determine time

of

establishment for all sampled stems. Time of plot initiation
(years

before

individuals
plot

sampling)

that

development.

was

established
Time

defined as the plot age

of

based
within

on

the

the

understory

(years after plot

mean

first

age
10

of

years

reinitiation
initiation)

all
of
was

after

which at least two new individuals established every 5 years

102

for at least a 20 year period.
was

defined

as

Unequivocal
rates

of

For this study,

survival

inferences

natality

or

to

cannot

the

be

mortality

time

made
that

of

about
led

establishment

to

sampling.

the

changing

observed

age

distr i b u t i o n s .

Radial and height growth analysis
Annual

ring

selected radii
1.4

m

stem

widths

widths

measured

(rejecting decayed

sections

were

were

of

measured

or

overstory
to

the

along

injured

stems
nearest

dimensionless

randomly

areas)

(dbh

>10

0.1

mm

dissecting microscope and ocular micrometer.
a standardized,

one

on

the

cm) . Ring
using

a

For each stem,

index of radial grow t h

(ring-

width index) was derived by dividing the m e a n ring-width for
the entire series

into the actual

Ring-width

were

series

dendroclimatological
no

a

priori

not

studies

reason

to

ring-width

detrended,
(Fritts
remove

as

is

1976),
age

at each
often

year.

done

in

since there was

effects

from

the

chronologies.

Mean chronologies were developed by averaging

the ring-width

indices of several

The

ring-width

standardization

individuals on each plot.
prevented

fast-growing

individuals from dominating the mean chronologies
a l . 1991).
attempt
growth

(Veblen et

Individual tree chronologies were averaged

to

dampen

patterns,

suppressions

or

any

non-synchronous

while
releases

retaining
that

might

changes
any
reflect

in

in an
radial

synchronous
plot-level

changes in resource availability during the course of forest

103

development.
several

The

mea n

red oak and

chronologies

were

developed

red maple on each plot.

Subtle

using

changes

in resource availability within the matrix of bigtooth aspen
ramets would be reflected best in the radial growth patterns
of

red

oak

lived,

and

red

relatively

aspen,

and

are

radial growth
1990) .

more

known

since

these

species

understory-tolerant

to

respond

to

than

release

with

number

of

varied

from

individuals

used

2

depended

to

9

and

their

forest

bigtooth

form
on

ii)

i)

the

total

and

<

indices

heights

13

m

from the r e c o r d ) , and

(individuals

in

the

UMBS

<

15 m

forest

from sampled bigtooth

chronologies.

rapidly,
species

the

their age

in

were

the

Huron

typically

suppressed so they were excluded from this a n a l y s i s ) .
width

the

(the goal was to extend chronologies as far b a c k as

possible w ithout dropping individuals
iii)

long-

increased

to

number of genets of these species on a plot,
at 1.4 m

are

(Laidly 1990; Sander 1990; Walters and Yawney

The

chronologies

maple,

These

ramets

aspen were
always

Ring-

not use d

grew

in

relatively

as might be expected for stems of a ver y intolerant
(Laidly 1990)

that were in dominant-codominant crown

positions at maturity.
The additional
aspen were
section
increment

prepared

ages

curves

interpolation
necessary

were

of

since

stem sections
and

aged

used
for

stem
the

to
each

height

stem

as

from the sampled bigtooth
described

reconstruct
stem.
at

sampling

a

In
given

method

previously.
5-year
some
plot

The

height

instances
age

controlled

was
stem

104

height

but

not

height plots

age.

In

for the

these

cases,

individuals

the

shapes

involved were

of

age-

examined

to

insure accurate determination of heights at the 5-year plotages of interest.

W h i t e pine seed-source characteristics
The distance to and diameter of all remnant white pine
seed trees were m e a sured for each UMBS plot.
were

easily

Potential

seen

remnant

after
seed

leaf-fall

trees

were

of

Remnant trees

bigtooth

restricted

to

aspen.
a

130

m

radius around the center of each plot. This was the distance
to

a

that

sole remnant
c ontained

Large white
individuals
bore)
tree
limit

had
was

a

tree

(within

small

amount

pine could
(from

a

extensive

not be

of

internal

as

one

based

on

two

>

m

white

aged

subsample

defined

was

a 500

stem rot.
cm

lines

pine

for a plot

regeneration.

accurately because most

examined

45

radius)

dbh.

of

with

an

increment

Instead,

a remnant

The

45

evidence.

publis h e d age-diameter data for white pine

cm

diameter

First,

non­

(UMBS data

file)

indicated that trees < 45 cm at the time of sampling in the
current study would have predated bigtooth aspen by no more
than 25 years.

These

individuals would have been too young

to have reached reproductive maturity by the time of forest
initiation.
study

Second,

indicated

sampling of white

that

the

maximum

pine

in the

diameter

of

current
stems

establishing at or soon after stand initiation was < 20 cm.
Only

one

stand

(four

plots)

contained

any apparent

remnant

105

individuals with diameters greater than 20 cm but less than
45 cm.
Multiple regression was used to develop an exploratory
model

to

source
for

examine the

relationship between

characteristics

this

species.

and

The

of

included

i)

tree,

i i ) mean

distance

total

number

distance

to

to

potential

all

these

several,

variables
large

probability
greater
area

of

seed

was

was

potential

seed

seed
early

rain

used

that

trees

trees,

not

as

a

crude

Larger

measure

diameter

of

trees

as

size

trees.
because

necessary

timing
may

Ranked
of

for

data

low

residual

coefficients

of

1981)

in

close

have

mean

not

were

a

of

had

measure

a

however,

multiple

in

sizes

plots.

basal

For

to

greater

because
Mea n

of

reproductive

necessarily

used

iii)

proximity

development.

reproduction

sample

standardized

iv)

the
seed

trees,

of

basal

potential

larger trees more seeds,

than smaller diameter trees,
the

seed

and

in

potential

establishment

forest

only

reproductive o u t p u t , i.e.

plots

pine

in

used

The assumption associated

would

white

early

closest

seed-

reinitiation

variables

the

area of all potential seed t r e e s .
with

understory

independent

model

of

time

white pine

but also

maturation.

have

been

older

they proba b l y reached
earlier

the
and

than

regression
poor

all

determination

smaller
analyses

structure

analyses,
(Sokal

of

adjusted
and

Rohlf

are r e p o r t e d . There were too few plots wit h associated

remnant red pine to conduct a similar regression analysis.

106
RESULTS
Forest composition and structure
The present-day forest composition and structure of the
Huron

and

Bigtooth

UMBS
aspen

forests.

study
was

Minor

areas

the

dominant

overstory

included red oak,

are

summarized
overstory

species

found

in

Table

species
in

red maple and paper birch.

b oth

4.1.

in

both

forests

Red maple was

virtually the only species regenerating in the Huron forest.
White pine and red maple were both abundant as regeneration
in the UMBS forest.
and

density

than

at

higher

of

the

UMBS.
in

the

Both overstory
Huron

Sapling
UMBS

forest
(2.5

forest,

<

(dbh > 10 cm) basal area

were
dbh

while

<

substantially
10

cm)

seedling

higher

density

densities

was
were

similar.

Population age structures
The

age-height

distributions

for

all

sampled

individuals on the Huron and UMBS plots are shown in Figures
4.1 and 4.2.

Recall that only one bigtooth aspen ramet was

sampled per plot.

Each plot

contained

bigtooth aspen ramets of similar age.
an

initial

period

cohort

of

trees

an additional
In the Huron

established

within

a

15-35

forest,
10-year

(Figure 4.1), beginning 65-70 years prior to sampling

(see plot ages in Table 4.2).

All plots were characterized

by an obvious period of stem exclusion during which

few or

no surviving individuals of any species established

(Figure

Table 4.1.
Present-day forest composition and structure of the Huron and UMBS
study areas.

Huron
Canopy a
Species

m /ha

Bigtooth aspen
Red oak
Red maple
Black oak
White oak
White ash
Paper birch
Black cherry

30.3
5.5
4.9
0.5
0.3
0.06
0.05
0.01

(2.1)—
(1.0)
(0.6)
(0.2)
(0.2)
(0.06)
(0.05)
(0.01)

Totals

41.6

(2.0)

stems/ha

Saplingstems/ha

Seedling—
stems/ha

_ —__

638
166
282
14
10
4
2
2

(48)
(29)
(28)
(7)
(5)
(4)
(2)
(2)

13 (6)
644 (89)
------4 (4)
---26 (9)

3625 (669)
64833 (6522)
---458 (260)
167 (97)
42 (42)
14583 (3813)

1117

(48)

686

83708

(96)

(8205)

Table 4.1 (cont'd).

UMBS

Canopy—
Species

m^/ha

21.3
2.5
2.4
0.9
0.9
0.6
0.02

Totals

28.6

(1.2)—
(0.6)
(0.4)
(0.3)
(0.4)
(0.2)
(0.02)

560 (41)
67 (20)
93 (23)
46 (12)
18 (10)
45 (15)
3 (3)
mmmm’ ”

(1.5)

890

“ *'

(44)

Sapling—
stems/ha

Seedling—
stems/ha

4
82
335
54
44
642
17
3

(4)
(23)
(69)
(13)
(20)
(127)
(7)
(3)

573
4740
64687
—
1562
104

(225)
(635)
(8805)
—
—
(569)
(71)

1180

(112)

71666

(8777)

—

— Dbh > 10 cm; — 2.5 cm < dbh < 10 cm; ^dbh < 2.5 cm.
Values are the means (+ SE) of 20 observations in the Huron forest and 16 at UMBS.

108

Bigtooth aspen
Red oak
Red maple
Paper birch
Red pine
White pine
American beech
Balsam fir

stems/ha

109

Figure 4.1.
Age-height distributions for all sampled
stems on the 20 Huron plots.
Individuals > 1.5 m tall
were completely sampled on each plot. Individuals <
1.5 m
tall were subsampled.
Note that
each plot
contained an additional 15-35 bigtooth aspen ramets of
similar age
as the sampled stem.
Red oak and b l a c k
oak are combined in the figure.

no
Figure 4.1

30

to

25 4

1

A

o

20 4

•

Bigtooth a s p e n

O

Red m ap le

▲

Red oak

T B
l
a
c
kc
h
e
r
r
y

15 -oQ^O

:7> A

♦

o

10

Wh it e o a k

5-

H1

0

H2

25
20

15-?3
10

HEIGHT

(M)

5

H3

0

H4
y o -.

25 A

20 c
154

A

°s

STEM

10

54

o
°

H5

0

H6

--,
--,
--,
--,
--n
-I--,

25 4

20 4k®
1544,
104
54

0

H7

I48

H9

H 10

25
20

0)

154
10

54
0

o

10

20

30

70

~n
10

20

TIME O F E S T A B L I S H M E N T
( Y E A R S AFTER PLOT INITIATION)

70

Ill
Figure 4.1 (cont'd).

25

~

o«

Of' -I

]Q-_

H 12

25-:

io:
H 13

o o,

H 14

25

I
—
X

20 jpr

CD
X

10-i

2
X
H 15

.
qaaffiflpggfco

H16

g

25 -

15:
10:

H 16

H 17

25 :
20-

-o
1 0 "

H 19

o

10

H20
20

30

40

50

70

10

20

30

TIME OF E S T A B L I S H M E N T
( Y E A R S AFTER P L O T INITIATION)

+0

50

50

70

112

Figure 4.2.
Age-height distributions
for all
sampled
stems on the 16 UMBS plots.
Individuals > 1.5 m tall
were completely sampled on each plot. Individuals <
1.5 m
tall were
subsampled.
Note that
each plot
contained an additional 15-35 bigtooth aspen ramets of
similar age as the sampled stem.
The red oak in the
upper
right
corner
of
plot
U15
was
actually
a
individual from plot U16 that predated the remaining
individuals on the plot by approximately 20 years.

113
Figure 4.2

B i g t o o t h asp e n
20 • *

Red m a p l e
Red oak
White pine

15 -

Red pme
1 0

-

U1

U2

U3

U4

U5

U6

U7

U8

201510

-

20

-

10

-

STEM

HEIGHT

(M)

GD

20 J
151 0

-

20-

10

-

U9

0

U 1° ......
10

20

30

40

50

60

70

80

10

20

30

r

40

TIME OF E S T A B L I S H M E N T
( Y E A R S AFTER PLOT INITIATION)

'-" I

50

60

70

80

114
Figure 4.2

(cont'd).

25
20
15
10

(M)

5
25

STEM

HEIGHT

20
15
10

5
25
20
15
10

5

0
0

10

20

30

40

50

60

70

30

10

20

30

40

TIME O F E S T A B L I S H M E N T
( Y E A R S A F TE R PLOT INITIATION)

50

60

70

80

115

Table 4.2.
plots.

Structural attributes of the Huron and UMBS

Huron

Plot
#

Overstory—
m 2/ha

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

54.3
44.8
32.6
40.9
32.1
34.7
31.1
35.9
36.3
36.1
39.3
44.3
46.5
49.4
38.3
47.3
65.0
50.1
35.0
38.9

(83)*
(72)
(71)
(93)
(53)
(59)
(81)
(81)
(85)
(64)
(72)
(66)
(80)
(78)
(60)
(80)
(80)
(58)
(41)
(88)

Plot Age (yrs)
mean— (max)

66
67
67
64
66
67
67
65
66
66
66
66
65
66
65
67
66
66
65
64

(70)
(70)
(68)
(65)
(69)
(69)
(70)
(66)
(67)
(66)
(70)
(70)
(67)
(67)
(68)
(69)
(68)
(67)
(70)
(66)

Understory— (stems/ha)
> 1.5 m ht
0-1.5 m
ht

735
74
515
2720
478
37
1838
919
4154
4411
3272
993
662
993
0
551
257
2316
2757
2794

128785
46887
93210
72941
37941
52537
56703
42941
66666
90833
46666
129166
107499
111666
175833
67500
66274
104534
87107
48995

116

Table 4.2

(cont'd).

UMBS

Plot
#

Overstory—
m 2/ha

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

21.6
30.3
26.3
37.7
27.0
28.2
17.7
20.6
32.7
29.5
25.7
26.5
30.8
33.8
29.4
39.8

(75)
(88)
(74)
(76)
(63)
(84)
(87)
(69)
(62)
(89)
(74)
(68)
(78)
(77)
(65)
(71)

Plot Age (yrs)
mean— (max)

71
69
72
72
71
72
71
71
76
75
81
78
67
70
71
71

(75)
(72)
(73)
(75)
(74)
(73)
(73)
(75)
(79)
(82)
(83)
(81)
(71)
(73)
(74)
(73)

Understory— (stems/ha)
> 1.5 m ht
0-1.5 m
ht

716
1221
1010
1116
1558
632
1179
526
1979
1179
1642
1726
2231
2968
1200
1326

79187
40854
35833
54188
44166
34167
100833
22500
68354
79166
65833
159166
89187
99166
112500
58333

•^Sterns > 10 cm dbh.
— Stems establishing subsequent to the start of understory
reinitiation.
— Mean age of all stems establishing within 10 years of plot
initiation.
— Plot total (relative basal area of bigtooth a s p e n ) .

117

4.1).

M ean

time

of

(+1 sd) time to understory reinitiation from the

plot

synchronous
is

timing

reflected

event

(11%) .

1.5

tall

m

initiation

in

of

the

Note

older

was

34.5

understory
low

that

than

reinitiation

The

on

in

instance was

only one
oldest

1 year

level

stems

variation

understory

(plot

near

all

of

the

for

years.

coefficient

t a l l , and then only by
densities

(3.8)

plots

for

a stem

<

> 1.5

m

stem

H 7 , Figure 4.1).

establishing

this

Plot-

subsequent

to

understory reinitiation are shown in Table 4.2.
In

the

UMBS

forest,

most

overstory

established within a 10-year period
71-83 years prior to sampling

(Figure 4.2),

(see plot ages

Most plots were characterized by an
exclusion,

however

period was

less-obvious

(±1

sd)

time

years.

to

in

or

unequal variances:
for this

event

coefficient
no

of
stems

at UMBS

(Ul,

1.5

m

in

plot

increase was

stem

was

this

4.2).
24.9

Mean
(12.0)

forest occurred

forest

(t-test

for

establishment
initiation
used to

high,

as

(48%) .
tall

reflected

Note

were

that

older

in the
in

the

than the

high
UMBS

oldest

tall.

In some Huron plots,
increase

of

U13-U16)

(Figure

reinitiation

in the Huron

was

vari a t i o n
<

in Table 4.2).

t = - 3 .05, d f = 1 7 , P = 0 . 007), but variability

understory stem > 1.5 m

after

plots

Understory reinitiation in the UMBS
than

beginning

obvious period

non-existent

u n d e rstory

significantly earlier

forest

several

individuals

and most UMBS plots,
frequency

(Figure

define

b eginning

4.2).

a more

there was an

The

30-40

timing

conservative

of

years
this

definition

118

of

understory

establishment
least

one

individuals

reinitiation.

was

defined

as

Continuous

the

number

individual

established

established

every

of

years

every

2 years,

understory
until

year,

for

at

or

least

at
two

a

10-

year period.

Determination of the timing of this event was

not biased by

incomplete subsampling of stems <

1.5 m tall

since the max i m u m ages of these individuals never e x c e e d e d ,
and,
1.5

in
m

fact,

tall

rarely

that

approached,

typically

understory establishment
exclusively by

stems

the

defined

age

the

of

individuals

start

of

>

continuous

(note that this period was defined

<

1.5 m

tall

on plots H 2 , H6 and H15

because of a lack of stems > 1. 5 m t a l l ; Table 4.2).
In the Huron forest, mean time to continuous understory
e s t a b l i s h m e n t , from
(C.V.=

9%) .

The

plot

m ean

initiation,
(+

se)

was

36.8

difference

(3.5)

between

years

time

to

understory reinitiation and continuous establishment of -2.3
(0.5)

years

was

low,

but significantly

(one-tailed Wilcoxon's
P< 0.005).
the UMBS
se)

Mea n

reinitiation
years

was

Wilcoxon's

and

rank difference test:

33.8

(7.1)

between

years

time

continuous

significantly
rank

T s=

from
36,

zero
n=20,

(+1 sd) time to continuous establishment in

forest was

difference

different

to

initial

establishment

different

difference

(cv=21%) .

test:

from

T s=

1,

of

zero

The mean

(+

understory
-8.8

(2.1)

(one-tailed

n = 1 6 , P<

0.005).

Non-normal residual distribution and heterogeneous variances
prevented

statistical

comparison

of

time

to

continuous

understory establishment between the Huron and UMBS forests.

119

Failure

to

presence
U15) .

meet

of

a

these

assumptions

single

In this plot,

outlier

was

among

attributed

the

UMBS

U15

(+1 sd)

from

of 35.1

the

plots

the
(plot

continuous establishment occurred much

earlier than in the remaining 15 UMBS plots
a mean

to

(4.6)

analysis,

years) .

variances

(13 years versus

After excluding plot

were

homogeneous

(but

residuals were still distributed no n - n o r m a l l y ) , and the mean
time

to

continuous

significantly

understory

different

between

establishment

the Huron

(Wilcoxon two-sample test:

Ug= 181,

Also,

timing

variability

establishment

in

in

the

the

remaining

was

and UMBS

n= 20 and 15,

of
15

at

forests

P >0.20).

continuous
plots

not

understory

UMBS

was

low

(C.V . =13%).

Bigtooth aspen
reinitiation
In

the

sampled

heights

Huron

the

mean

aspen

at

the

bigtooth

continuous

the

time

of

(17) % and
five

10.9

UMBS

reinitiation
percentages

63

at

with
(Ul,

times

establishment were 3.6

67

height

of

the

understory

of

16.5

(5) % of total

(13) %,

mean

initial

and

(1.3) m and 16

heights

(3.8)

m.
at

heights
m

and

39

r e s p e c t i v e l y . For the

relatively

U13-16)

(1.8)

analogous

forest were 8.4

m and 50

plots
times

understory

of

was

resp e c t i v e l y . The

in the UMBS
(2.9)

sd)

time

establishment

(6) % and

sampling,

and percentages

(+1

during

(2.1) m, while mean height at the time

understory

These heights were

growth

forest,

reinitiation was 15.5
of

and height

early
(+1

sd)

understory
heights

continuous
(7)% and 8.6

and

understory
(4) m and

120

38

(17)%,

UMBS

respectively.

forest

understory
bigtooth

ranged

The sampled bigtooth aspen

from

15-35

reinitiation.

aspen were,

In

years
both

at most,

35

old

at the

forests,

years

old

in the
time

the

at

of

sampled

the

time

of

continuous understory e s t a b l i s h m e n t .
Height
aspen

are

Huron

increment

shown

forest

increment,
large

in

Figure

was

20-25

one

occurred
pattern

time

after

of aspen

to

the

start

of

coincided

and

a

continuous
wit h

forest

shows

increase,

gradual

height
a

understory

establishment

height
an

increment

initial

early

40-45 years after

decrease

thereafter.

establishment

secondary peak

the

followed

Both

understory

in

in

initiation,

The

bigtooth

trend

culmination

understory

the

O v e r a l l , height

overall

decline.

in the UMBS

initiation,

sampled

this period.

decline followed by a secondary
plot

The

forest

continuous

subsequent

the

early

after
and

for

4.3.

of

years

decline

reinitiation

over

typically

in height

growth rates of bigtooth aspen

The

increment.
at UMBS were

low and only approached those of Huron bigtooth aspen early
in forest d e v e l o p m e n t .

Patterns of radial growth and understory reinitiation
Ri n g -width
in Figure 4.4.
H20

because

inadvertently
The

m ost

chronologies

for the Huron plots

are

shown

The chronologies are missing for plots H17the

stems

d e stroyed

obvious

and

sections
prior

to

consistent

from

these

measuring
feature

plots

radial
of

the

were

growth.
Huron

121

Figure 4.3. Five-year height increment curves for the
sampled bigtooth aspen in each forest.
Values are
means + se.
Sample size for the Huron forest is 20.
Sample size for the UMBS forest is 16 (14 at 5-10
years).

122
Figure 4.3

3.5

HURON
3.0

2.5

INCREMENT

(M )

2.0

1.5

1.0

0.5

HEIGHT

5-10

15-20

25-30

35-40

45-50

55-60

65-70

3.5

FIVE-YEAR

UMBS
3.0

2.5

2.0

1.5

1.0

0.5

i

5-10

i i ii i
15-20

25-30

i

35-40

i i

45-50

i

i i r

55-60

YEARS AFTER PLOT INITIATION

65-70

123

Figure 4.4.
Ring-width indices for the Huron plots. Each
chronology is the mean for the sample size indicated
in the upper right corner of each graph. The doted
line at a ring-width index of 1.0 is the standardized
mean
for
the
entire
chronology
(see
text
for
derivation of ring-width i n d i c e s ) . The dashed line in
each
graph
is
the
cumulative
establishment
distribution for advanced regeneration (stems > 1.5 m
tall).
The
beginning
of
each
establishment
distribution
marks
the
start
of
understory
reinitiation (at least two stems establishing every
five yea r for a minimum of 20 years; this is indicated
by an U for plots without advanced regeneration) . The
C
in
each
graph
marks
the
start
of
continuous
understory
establishment
(at
least
one
stem
establishing every year, or two every two years, for a
m i n imum 10 y e a r s ) .

CUMULATIVE PROPORTION OF ADVANCED REGENERATION
o

i

~

n

6

o

m

6

6

o

m

d

o

6

i

n

6

o

6

i

6

n

6

o

m

6

o

6

i

n

6

i

d

n

6

o

i

d

n

6

o

d

i

n

6

o

6

K)

O)

n

o

o

o
_o

o

F ig u r e

_o

m

o

in

o

o

o

o

in

o

o

o

o

X3QNI H1QIM-0NIH

o

o

in

o

o

o

m

o

6

PLOT

00

AFTER

o
CN

YEARS

_o

INITIATION

_o

CUMULATIVE PROPORTION OF ADVANCED REGENERATION
m
pv
o

o
o
—*

m

o
m

CM
0

0

0
0
0

m
0

0

m
0

m

CM
O

m

O
O
O

0

O

0

in

o

0

O
_o

(£>

,o

m

_o

‘o

,o

O
_o

in

<

oJ
—

CM
rH

C
L

(£>

cc

L
l
J
I
—
L
_

<
(c o n t'd ).

in
z> o _o

F ig u re

4.4

o

o
o

o

X3QNI HiaiM-ONiy

o

in

o

if)

<

L
d

126

chronologies was a marked increase

in mean ring-width

that

plot

occurred

4.4).

years

At their peak,

90% and

lasted 5 to

growth

rate

growth

that

obvious,

years

30-35

after

growth

a

increases

years

plot

5

plots,
of

years.

in m ean

after

plot

(Figure
averaged

the

initial

reduced

features

initiation.

and typically

increases

period

consistent

increases averaged 60-65%,
10 years)

rate

approximately

fairly

initiation

In most

followed

included

occurred

45-50

10 years.

lasted

but

after

these growth

increase

chronologies
that

15-25

radial

Other

less-

the

Huron

of

ring-width

initiation
At

index

their

indices

and

again

peak,

these

lasted a short period of time

followed a short period

of

(5-

reduced

(lasting 5 years or l e s s ) .

Timing of understory reinitiation on each Huron plot is
shown relative to the radial growth patterns
The

dashed

line

in

each

establishment distribution

graph

is

for advanced

in Figure 4.4.

the

cumulative

regeneration

(stems

> 1.5 m tall that established subsequent to the beginning of
understory r e i n i t i a t i o n ) .
establishment
understory

curve

on

three

regeneration

(plots

indicated

an

continuous

marks

reinitiation

reinitiation

by

U

The beginning of each cumulative

began.
plots

H2,
on

understory

the

The

with

H6,

the

year

(plot

timing
little

and

H15;

graph.

Also,

establishment,

age)

of
or

that

understory
no

Figure
the

regardless

advanced
4.4)
timing
of

is
of
stem

height class,

is indicated by a C on each g r a p h .

Typically,

understory

reinitiation

understory

and

continuous

127

establishment both began within a 5-year period

immediately

following the moderate radial growth increase that occurred
30-35

years

after

plot

exception was plot H 6 .
(U) began almost
increase,

and

initiation

(Figure

4.4).

The

one

On this plot understory reinitiation

10 years

after the

continuous

30-35 year growth

understory

rate

establishment

(C)

occurred concurrently with a second moderate growth increase
beginning 50 years after plot initiation.
cumulative

establishment

The shapes of the

distributions

for

advanced

regeneration are worth examining because they represent the
establishment

patterns

of

individuals

that

are

p r e s umably

most likely to recruit to dominant canopy positions.
p l o t s , the
high

slopes

initially,

understory

of

or

the

they

reinitiation

establishment

increased
in

distributions

soon

concert

On most

after

with

the

were

start

increases

in

of
the

ring-width indices.
A

consistent

feature of most UMBS

chronologies was

an

increase in mea n ring-width index approximately 35-40 years
after plot initiation
releases
years.

averaged

for

the

15-20

of time

probably

least

years

chronologies

(plots U 1 - U 2 ) , or were
(plots

show

U5,

15

marked

The

lack

reflects

of

this

actual

feature
growth

in

slowly

U7-U8,

U14,

decreases

in

growth 15-20 years after plot initiation
U16).

for at

these

Radial growth rates prior to release were either low

increasing
of

At their maximum,

around 83% and persisted

for extended periods

Some

(Figure 4.5).

U16) .
radial

(plots U 6 , U 8 - U 1 2 ,
other

patterns

of

chronologies
the

sampled

128

Figure 4.5.
Ring-width indices for the UMBS plots. Each
chronology is the mean for the sample size indicated
in the upper right corner of each graph. The doted
line at a ring-width index of 1.0 is the standardized
mean
for
the
entire
chronology
(see
text
for
derivation of ring-width i n d i c e s ) . The dashed line in
each
graph
is
the
cumulative
establishment
distribution for advanced regeneration (stems > 1.5 m
tall).
The
beginning
of
each
establishment
distribution
marks
the
start
of
understory
reinitiation (at least two stems establishing every
five year for a m inimum of 20 years) .
The C in each
graph
marks
the
start
of
continuous
understory
establishment (at least one stem establishing every
year, or two every two years, for a minimum 10 y e a r s ) .

CUMULATIVE PROPORTION OF ADVANCED REGENERATION
t
r^

8
—
*

0

n

Q

i

0

n

i

f
0

n
N

o
o

r

0

i

v
0

n

u

o

0

T

i

C
0

n

N

Q
O
0

i

n
v

r
0

i

o

0

n

i

c
0

n

N

o

o

0

in

i

r
0

n
^

i

O

i

6

L

D

0

O
c

Q
s

o

0

i
0

r

n

^

o

m

0

i

t

0

n
N

o
O
0

co

CM

CO

_o

PLOT

_o

_o

AFTER

_o

YEARS

00

CM

_o

_o

Figure

o
K)
O

O

m

o

m

o

o

m

o

o

X 3 Q N I H 1 Q IA A -0 N I3

INITIATION

_o

o

o

o

o

130

i.0-■

-

0.50

0.5 -

-

0.25

0.0

0.00
N= 4

U13

U 14

N=2
-

0.75

1.0

-

0.50

-

0.25

- ■

0.5 -

0.0

0.00

i

U15

N=3

U 16

/

N=2

1.0

- •

0.5 -

-

0.75

- -

0.50

-

0.25

/

0.0

0.00
20

50

40

50

80

30

YEARS AFTER PLOT INITIATION

40

50

60

80

REGENERATION

1. 5 -

OF ADVANCED

1. 5 -

PROPORTION

RING-WIDTH

(cont'd).

CUMULATIVE

INDEX

Figure 4.5

131

trees

(plots Ul,

U 1 5 ) , while in others it probably reflects

the short lengths of the chronologies
In

the

associated

UMBS

with

chronologies

continuous

understory

concurrently

with

increase that

noted

understory

consistently

ring-width

(Figure

plots,

A

any

or

obvious

(Figure

soon

after

35-40

In

most

the start

150%

the

began

growth

forest

trend was

understory

of a

of

plots

rate

initiation

plot U 1 5 . As

establishment

this plot began 13 years after plot initiation,
coincide with

not

contrast,

radial

after

this

continuous

on

was

feature

4.5).

the

years

exception to

previously,

reinitiation

establishment

occurred

4.5).

(plots U 4 , U 1 3 - U 1 4 ) .

ring-width

on

but this did
increase

that

began 10 years after plot initiation and persisted for over
15

years

(Figure

chronology

for

4.5).

this

plot

Unfortunately,
did

not

the

begin

ring-width

early

enough

to

adequately assess growth patterns much before this release.
On

mos t

plots,

large

increases

cumulative

establishment

regeneration

often

understory
indices

the

slopes

distributions

coincided

establishment

in

with

and

the

the

the

for

advanced

of

continuous

in

ring-width

start

increases

of

(Figure 4.5).

White pine seed-source characteristics
White

pine was

the

dominant

tree

on most of the UMBS plots

(Figure 4.2).

included

U7

maple,

plots

and

U2-U4

plot U14

and
which

had

which
large

species

were

regenerating

Notable exceptions
dominated

amounts

by

of both

red

white

132

pine and red maple.

On most plots,

first

begin

species

to

white pine was also the

establishing

in

u n d e r s t o r y , including plots U7

and U 1 4 .

pine

opportunity

seed

trees

relationship

provided

between

an

seed-source

The

the

forest

remnant white

to

examine

characteristics

the

and

the

timing of understory reinitiation for this species.
Differences
accounted

for

in
a

white
large

pine

seed-source

and

significant

characteristics
proportion

of

variation in timing of understory establishment among plots
(r— adj.= 0.82;

see= 1.748; P= 0.0004) . This analysis excluded

plots

U3

U2

and

since,

by

definition,

initial

understory

reinitiation of white pine had not occurred on these plots.
Differences

in

seed-source

low and only marginally

characteristics

accounted

for

a

significant proportion of variation

in timing of continuous understory establishment among plots
(r"adi= 0.53 ; see= 2.468 ; P= 0.052).

Note that this analysis

excluded not only plots U2 and U 3 , but also plots U4 and U 7 ,
since continuous understory establishment of white pine,
definition,
for

had not yet occurred on these plots.

initial

excluding
results
included

understory

plots

w ere

U4

and

reinitiation
U7,

marginally

plots

U4

and

as

well

improved
U7

was

over

(r~adj.=

The model

reanalyzed

as

plots
the

by

after

U2-U3.

results

0.87;

see=

that

timing

The
that

1.313;

P=

0.0008) .
The

regression

pine understory
new stem

results

suggest

reinitiation,

establishment was

and

hence the

prevented

for

of

white

length of time

this species,

may

133

have been influenced by remnant seed-source characteristics.
Initial

establishment

proximity
with

to

occurred

several

large

less-favorable

contrast,

the

seed

on

of

the

white

pine

plots

in

close

relative

to

plots

characteristics.

of

timing

on

trees,

seed-source

influence

characteristics
establishment

earlier

remnant

of

seed-source

continuous

appeared

to

In

be

understory

muc h

weaker,

suggesting additional controls on this e v e n t .

DISCUSSION
Many of the studies that have documented b oth the stem
exclusion
forest

and

understory

development

have

reinitiation
been

one or few replicate plots
Richardson

1985;

data

replicate

from

analysis
Sharik

et

al.

variability

site
the

and

and

important

(1987)

into

among

landscapes

a

stem

and

of

step

exclusion

forest-types.

For

Roberts and
have

pooled

forest-level

1992).

by

understory
Peet

the

patterns

homogeneous

Characterizing

patterns
the

These

assessing

developmental

histories.

example,

from

1985? Harcombe 1986;

for

identifying
and

or

Snook

developmental
in

even-aged

results

single

characterized

disturbance
of

on

1992)

inappropriate

and

first

Segura

repeatability

repeatability

various

Nowacki

samples

are

of

(Oliver et a l . 1985;

and

1989;

conditions

control

either

(Bloomberg 1950; Oliver et al.

methodologies

within

Abrams

based

phases

may

be

mechanisms
reinitiation
and

an
that
in

Christensen

argue that understory reinitiation in most even-aged

134

forests is triggered by an increase in resource availability
following density-independent mortality of mature
individuals

late

dominated

in

by

deterioration

even-aged

bigtooth

can

forests

is

dominants

it

In

complete

rapidly

(<

10

forests
overstory

years;

Graham

et

If understory reinitiation in these

triggered

then

aspen,

progress

a l . 1963; Laidly 1990).

development.

overstory

by

mortality

should

begin

of

senescent

synchronously

canopy

across

space

and it should also occur late in even-aged development.
In the current s t u d y , the population age structures of
surviving

individuals

reflected

synchronous

cohort

trees

of

establishment

on

plots

developmental

established

was

within

curtailed

within
for

the

Huron

patterns.
a

10-year

landscape

An

initial

period,

approximately

35

new

years,

after which time individuals again began to establish in the
forest u n d e r s t o r y .
within

the

UMBS

In contrast,

landscape

were

the developmental
more

variable.

patterns

An

initial

cohort of trees established within a 10-year period,

but the

length of a stem exclusion period was highly variable
years).
UMBS

(0-35

Synchronous development was a characteristic of the

forest

only

under

a

more

restrictive

definition

of

continuous understory es t a b l i s h m e n t .
In both

forests,

in establishment
aspen were

in

the

at most

Great

increases

frequency began when the dominant bigtooth
67%

(and often much less)
age

understory reinitiation and

of their

total

heights

at

maturity

and no more than half of their maximum

Lakes

region

(Graham

et

al.

1963) . The

135

mechanism

t r i g gering

understory

reinitiation

in

these

forests was not increased resource availability occurring in
response

to

overstory

development,
Huron

despite

forest.

resource

The

the

may

synchrony

results

availability

develo p m e n t

deterioration

do

of

suggest

occurring

have

late

the

even-aged

development
that

an

relatively

influenced

in

increase

early

timing

in

of

in

the
in

forest

understory

reinitiation.
U n d e rstory
associated
overstory

with

most

radial

in

resource

with i n

plots,

the

the

typica l l y began

was

growth

s t e m s ; wit h both

increases
thinning

reinitiation

soon

releases

in

events occurring
availability

bigtooth

b ulk

hypothesized

of

aspen
new

following

understory

be

surviving

in response

overstory.

after moderate to

to

to

natural

Indeed,

on

establishment

large

radial

growth

releases that occurred 30-35 years after forest initiation.
The

releases

thinning,

may

not

however,

have

concentrated

mortality,

subsequent

and

releases

gr owth

characteristic
forests
(1963.

increases

of

in

surviving

developmental

p.

90-94)
occur

development.
releases

the

response
of

stems

are

believed

of

most

For example,

40

natural

availability,

years

of

to

be

even-aged

Graham et a l .

two or three waves

first

to

density-dependent

resource

features

report that

within

in

waves

(Oliver and Larson 1990).

thinni n g

subtle

occurred

of natural

aspen

forest

Short periods of reduced growth and subsequent

w i thin
changes

the
in

aspen

overstory

resource

w ere

attributed

to

availability

occurring

in

136

response

to

the

mortality

waves.

As

in

the releases reported by Graham et al.
magnitude
occur

and

shorter

following

a

duration

major

the

than

those

canopy-opening

the

majority

of

new

occurred

in

response

to

increased

following

mortality

crown
have
It

growth
been

has

stand

the

is

to

reached

development,

1983,

that

establishment

by

default,

individuals
canopy

closure

asymptotically

early

and

thereafter

decreases

availability

then,

complete

canopy

(for

1984).

light

stems

typically

disturbance

overstory

maintain

hypothesized

that

understory

suppressed

remaining

inadequate

been

forests

of

of

study,

(1963) were of lesser

examples of the latter see Lorimer 1980,
If

current

in

closure.

in
the

must

even-aged
course

(Zeide

of

1987,

1991), because of the combined influence of natural thinning
and

a

reduced

individuals.
sampled

rate

The

bigtooth

of

crown

height
aspen

increment
ramets

expansion rates may have been
understory reinitiation
height

increment

understory

reinitiation.

with

continuous

did

low at times

increased

understory

height growth rates were

UMBS,

the

that

just

bigtooth
at

crown

aspen

to

height

coincident

however,

low in this

forest,

prior

a time

establishment,

few

coincident with

In the Huron

slightly

always

surviving

for

suggest

substantially
At

in

patterns

(Figure 4.3).

dropped

increment actually

expansion

overall

forest,

except

for early in forest development.
Low
combined

height
with

growth

the

low

rates

of

overstory

bigtooth
basal

area

aspen
of

at

this

UMBS,
forest

137

(Table 4.1),
primary
are

may indicate that light limitation was not the

factor leading to stem exclusion.

characteristic

of

and nutrient-poor
and

Richardson

Chapter
may

in the Great
Scheiner

stem

exclusion

that

canopy

2) .

stem

closure

The

UMBS

the

can

despite

plots

periods

results

exclusion,

establishment,
forests,

or
be

low

that

et al.

leads

on

of

the

current

to

Laidly

age

basal

these

continuous

current

study

reduced

characteristic

overstory

had

(Roberts
1990;

forests

pr e s u m a b l y

substantially
a

dry-mesic

region

1988;

areas

because
resource

(Roberts and Richardson 1985;

basal

demonstrate
frequency

feature

areas.

distributions

of stem exclusion were

lowest

Lakes

period,

availability in the understory
Chapter

forests

Previous studies have shown that

a

incomplete

sites

1985;

2).

lack

aspen-dominated

Low basal

of

In

these

fact,

suggesting

of

the

definite

also those that had some of

areas

in

the

forest

(U7-U8, Ull-

U 1 2 ; Table 4.2).
If

light

changes

in

is

the

not limiting

ability

of

in the

root

resources during the course of
more

important

factor

Changes

in

rates

important

factor

aspen-dominated

forest

systems

to

growth may

influencing
forests

pr e e m p t

soil

forest d e velopment may be

influencing understory

of root

understory,

be

understory

given

the

a

establishment.
a

particularly

establishment

taxon's

ability

in
to

saturate soil space rapidly following fire through growth of
clonal

root

Crawford

systems

1965;

(Stoeckeler

Barnes

1966,

and Macon
1969;

1956;

Graham

et

Zahner
al.

and

1963;

138

Kemperman

and

research

Barnes

suggests

deteriorates

that

example

gradual

a

Scheiner
clone's

progressively

most ramets having
(for

1976;

32

during

et

al.

1988).

inter-ramet
forest

Limited

root

system

development,

with

independent root systems by an early age
years;

deterioration

Ruark

may

push

and

Bockheim

1987).

This

soil

resource

availability

above some minimum threshold required for new establishment
in the forest u n d e r s t o r y .
In

the

between
growth

current

understory
in

the

study,

the

reinitiation

overstory

suggests

repeatable
and
that

association

releases

in

resource

limitation

was an important factor leading to stem exclusion.
the
that

patterns
changes

forest

of
in

establishment
growth

pine

establishment

availability

may

have

also

establishment.
of

releases

initial

pine

seed

development

new un d e rstory

growth

white

white
in

the

establishment
release.

The

establishment

overstory,
prior

the

in

soon

while
to

suggest

course

of

timing

of

initial

after

on

any

timing

significantly

seed-source c h a r a c t e r i s t i c s .

UMBS

the

influenced

occurred

variation

was

during

However,

On some UMBS plots,

pine

began

at

radial

radial

other

plots

obvious
of

radial

initial

related

to

white

remnant

Initial establishment occurred

earlier on plots in close proximity to several large remnant
white p ine seed t r e e s , sometimes prior to any radial growth
release.

This

relationship

suggests

that

the

timing

of

white pine understory reinitiation may have depended on the
matura t i o n

or

proximity

of

a

local

seed-source.

A

similar

139

argument has been used to account for differences in initial
establishment times

of white pine

dominated forests at UMBS

in other bigtooth

(Sharik et al.

1989).

aspen-

However,

in

the current s t u d y , there was a correspondence between large
increases

in

subsequent
radial

the

to

growth

Additionally,
white

frequency

understory
rates

of

white

pine

reinitiation,

overstory

red

establishment,

and
maple

increases
and

red

in
oak.

the relationship between timing of continuous

pine

establishment

characteristics was weak.
establishment
initiation,

of

found

remnant

seed-source

A similar increase in white pine

frequency,

was

and

beginning

in

the

73

46

years

year-old

after

forest

bigtooth

aspen-

dominated forest examined by Sharik et a l . (1989;

see their

Figure 1).
These results suggest that timing of new establishment
of

white

between

pine

may

changing

have

resource

and seed availability.
pine

may

seed

have

input

development,

been

levels

on

plots

remnant

however

in

the

by

an

interaction

forest

understory

Limited early establishment of white

occurred
from

influenced

that

trees

the bulk

received

during

of new

substantial

early

forest

establishment may

not

have occurred until resource levels in the forest understory
became

favorable

mortality
overstory.

wave
Most

for

survival,

within
forests

of safe establishment

the

possibly

bigtooth

probably contain

sites

the peak of stem exclusion.

following

a

aspen-dominated
a limited

(sensu Harper 1977)

number

even during

Saturation of an area with seed

140

would

insure

alternative
white

that

all

of

explanation

pine

is

that

these

for

the

sites

are

occupied.

establishment

increases

in

An

patterns

resource

of

availability

actually hastened the reproductive ma t u r a t i o n of seed trees
that

had

either

disturbance

or

survived

pre-disturbance
maturation,

established
as

forest.

acting

alone

immediately

advanced

following

regenerat ion

A

near-synchronous

or

in

combination

from

the

seed-source
with

improved

resource levels in the forest u n d e r s t o r y , could have led to
the

large

increases

in

white

pine

establishment

frequency

that occurred on most p l o t s .
The

establishment

dominated
suggest

understories

that

an

patterns

on

plots

(predominantly

interaction

between

with

red

hardwood-

maple)

resource

also

availability

and seed-source maturation may have influenced the timing of
understory reinitiation.

In both forests,

the stem exclusion

periods for red maple always lasted 30-40 years.
in

new

has

red

been

maple

establishment

attributed

reproductive

maturation

sprout-origin trees
a l . 1985;
maple

Sakai

Scheiner

low
of

(Roberts

1990).

(and red oak)

and at UMBS

to

in

seed
a

aspen-dominated
availability

post-disturbance

and Richardson

Indeed,

in the Huron

observations
forest

mature

forests

prior

to

cohort

of

1985;

Sakai

on mature

(Chapters

et
red

3 and

6)

(Roberts and Richardson 1985; Sakai et a l . 1985;

et

al.

1988 ;

Chapter

3)

do

indicate

overstory individuals w ere of sprout-origin.
a

This delay

red

maple

seed-source

had

existed

that

many

Presumably,
early

in

if
the

141

development
relatively
1991)

the

have

forest,

established

white

understory

UMBS

understory-tolerant

would

tolerant

of

pine

reinitiation

individuals

species
at

(Wendel
on

some

(Walters

least
and

plots

as

and

early

Smith

w ith

of

Yawney

as

less-

1990) .

red

this

While

maple-dominated

understories may have been dependent on maturation of local
s e e d - s o u r c e s , new
typically

began

o v e r s t o r y ; the

understory

following

establishment

a

radial

repeatability

of

growth

of this

this

release

pattern

As with white pine

the

understory

red

maple

suggests

and
that

between

new

radial

growth

increases

in

releases

resource

in

suggests

it was more than coincidence.
correspondence

species
the
that

at U M B S ,

establishment
in

the

of

overstory

availability

led

to

decrea s e d mortality of new s t e m s . Reproductive maturation of
red

maple

seed-sources

and

the

increases

in

resource

availability that led to new establishment may have occurred
at similar times simply by chance,
that

increases

in

resource

or again,

availability

it is possible

actually

hastened

the maturation of local see d - s o u r c e s .

SUMMARY AND CONCLUSIONS
Quantification
forest

development

of

variability

patterns

is

an

and

repeatability

essential

first

step

in
to

developing a comprehensive mechanistic understanding of stem
exclusion and understory reinitiation

in even-aged

forests.

The current study provides this first step for a forest-type
that

has

been

understudied

from

a

stand

development

142

perspective.
indicate

The results from the Huron and UMBS landscapes

that bigtooth

aspen-dominated

forests

do

follow

a

patter n of development believed to be characteristic of most
forests
1981,

that

initiate

Peet 1981,

cohort

were

distributions
landscapes.
understory

early

scheme
example,

disturbance

While

the

to

exclusion,

repeatable

these

forest,

forests,
bulk

of

establish
there was

understory

used

stem

of

began

within each

major

Peet and Christensen 1987).

establishment,

reinitiation

of

following

at

this

understory

features of

near

variation

age

and

among

in the

forest

synchronous

some variation
at

the

within

regeneration

reinitiation

allowed

Rapid initial
and

both

(Oliver

in the

UMBS.
to

be

The

times
timing

sampling

detected.

For

if research at UMBS had been restricted to the one

stand containing plots U 1 3 - U 1 6 , there would have been little
reason

to

believe

that

a true

stem

exclusion

period was

a

characteristic feature of bigtooth aspen forest development.
The

results

also

indicate

that understory

reinitiation

not triggered by mortality of dominant overstory stems
in

even-aged

reinitiation
betwee n

development.
may

have

increased

been

The

factors

timing of

influenced

resource

response to natural thinning,
sources.

The

by

interaction
possibly

and maturation of local

controlling

late

understory

an

availability,

was

stem exclusion

in

seedand

understory reinitiation may be even more complex than recent
overviews
1990;

of

Peet

forest
and

development

Christensen

suggest

1987),

since

(Oliver
these

and

Larson

treatments

143

have largely

ignored the potential

on new establishment.

influence of seed-source

These results suggests that there

is

still mor e to be learned about even-aged forest development,
particularly
stem

with

exclusion

detected

in

experimental

respect

and

this

these mechanisms.

the

understory

study

studies

to

will

that

are

mechanisms

reinitiation.

be

useful

needed

to

for

that
The

control
patterns

designing

correctly

the

identify

BIBLIOGRAPHY
Abrams, M. D . , and N o w a c k i , G. J. 1992. Historical variation
in fire, oak recruitment, and post-logging accelerated
succession in central Pennsylvania.
Bull. Torrey Bot.
Club 119: 19-28.
Barnes, B. V.
1966.
The clonal
a s p e n s . Ecology 47: 439-447.

growth

habit

of American

________.
1969.
Natural variation and delineation of clones
of Populus tremuloides and P. grandidentata in northern
Lower Michigan.
Silvae. Genet. 18: 130-142.
Barnes, B. V . , Pregitzer, K. S., Spies, T. A . , and Spooner,
V. H. .
1982.
Ecological forest site classification.
J. For. 80: 493-498.
Benninghof f ,
W.
S. ,
and
Cramer,
K.
J.
1963.
Phytosociological
analysis
of
aspen
communities
on
three site classes for Populus grandidentata in western
Cheboygan County, MI.
Vegetatio 11: 253-264.
Bloomberg, W.
157-161.

J.

1950.

Fire and s p r u c e .

For.

C h r o n . 26:

B o r m a n n , F . H . , and L i k e n s , G . E . 1979.
Pattern and process
in a forested ecosystem.
Springer-Verlag, New York.
Carleton,
T.
J.
1982.
The
pattern
of
invasion
and
establishment of Picea mariana (Mill.) BSP. into the
subcanopy layers of Pinus banksiana Lamb, dominated
s t a n d s . Can. J . For. Res. 12: 973-984.
Cayford, J. H.
1957.
Influence of the aspen overstory on
white spruce growth in S a s k a t c h e w a n . Canada Department
of Northern Affairs and National Resources Forestry
Branch, For. Res. Div. Tech. Note 58.
Christensen, N. L . , and Peet, R. K.
1981.
Secondary forest
succession on the North Carolina Piedmont.
In Forest
succession: concepts and application.
Edited by D. C.
W e s t . , H. H. S h u g a r t , and D. B.
Botkin. SpringerVerlag, New York.
pp. 230-245.
Clatterbuck, W. K . , and H o d g e s , J. D.
1988.
Development of
cherrybark
oak
and
sweetgum
in
mixed,
even-aged
bottomland stands in central M i s s i s s i p p i , U.S.A. Can.
J. For. Res. 18: 12-18.
Day,

R. j.
1972.
Stand s t r u c t u r e , succession, and use of
southern Alberta's Rocky Mountain forest.
Ecology 53:
473-478.

144

145
Deal,

R. L . , Oliver,
C. D . ,
Reconstruction
of
mixed
coastal southeast Alaska.
654.

and Bormann,
B. T.
1991.
hemlo c k - s p r u c e
stands
in
Can. J. For. Res. 21: 643-

E i n s p a h r , D. W . , and W y c k o f f , G. W.
1990.
aspen:
timber
supply,
utilization,
N o r t h . J. A p p l . For. 7: 168-171.

North American
and
research.

Fried, J. S., Tappeiner, J. C . , II, and Hibbs, D. E. 1988.
Bigleaf maple seedling establishment and early growth
in Douglas-fir forests.
Can. J. For. Res. 18: 12261233.
Fritts, H. C.
London.
Gleason, H. A.
Michigan.

1976. Tree rings and climate.

Academic Press,

1923.
Botanical observations in
J. New York Bot. Club 24: 273-283.

northern

G r a h a m , S. A., Harrison, R. P., J r . , and W e s t e l l , C. E . , Jr.
1963.
A s p e n s : phoenix trees of the Great Lakes region.
University of Michigan Press, Ann Arbor.
Guldin,
J.
M. ,
and
Lorimer,
C.
G.
1985.
Crown
differentiation in even-aged northern h a r d w o o d forests
of the Great Lakes region, U.S.A.
For. E c o l . Manage.
10: 65-86.
Harcombe, P. A.
1986.
Stand development in a 130-year-old
spruce-hemlock forest based on age structure and 50
years of mortality data.
For. E c o l . Manage. 14: 41-58.
Harper, J. L.
1977.
Press, London.

Population biology of plants.

Academic

Hibbs, D. E.
1983.
Forty years of forest succession
central New England.
Ecology 64: 1394-1401.
Hix,

in

D. M. , and Lorimer, C. G. 1990. G r owth-competition
relationships
in
young
hardwood
stands
on
two
contrasting sites in southern Wisconsin.
For. S c i . 36:
1032-1049.

________.
1991.
Early stand d e velopment on two former oak
sites of southwestern Wisconsin. For. E c o l . Manage. 42:
169-193.
Kelty, M. J.
1986.
Development patterns in
hardwood stands in southern New England.
Res. 16: 885-891.
K e m p e r m a n , J. A . , and
American aspens.

two hemlockCan. J. For.

Barnes, B. V.
1976.
Clone
Can. J. Bot. 54: 2603-2607.

size

in

146
Kilburn, P. D.
1957.
Historical development and structure
of the aspen, jack pine, and oak vegetation types on
sandy
soils
in
northern
Lower
Michigan.
Ph.D.
dissertation, University of Michigan, Ann Arbor.
________.
1960a
Effect of settlement on the vegetation of
the University of Michigan Biological Station. Pap.
Mich. Acad. S c i ., Arts Lett. 45: 77-88.
________.
1960b
Effects of logging and fire on xerophytic
forests in northern Michigan.
B u l l . Torrey Bot. club
87: 402-405.
L a i d l y , P. R.
1990.
Populus grandidentata Michx. (Bigtooth
a s p e n ) , p. 544-550.
In Silvics of North America, Vol
2.
Edited by R. M. Burns and B. H. Honkala. U.S.D.A.
For. S e r . Ag. H a n d . , 654.
pp. 544-550.
Larson, B. C.
1986.
Development and growth of even-aged
stands of Douglas-fir and grand fir.
Can. J. For. Res.
16: 367-372.
Lorimer, C. G.
1980.
Age structure and disturbance history
of a southern Appalachian virgin forest. Ecology 61:
1169-1184.
________.
1983.
Eighty-year development of northern red oak
after partial cutting in a mixed-species Wisconsin
forest.
For. S c i . 29: 371-383.
________. 1984. Methodological considerations in the analysis
of forest disturbance history.
Can. J. For. Res. 15:
200-213.
Marquis, D. A.
1967.
Clearcutting in northern h a r d w o o d s :
results after 30 years.
U.S.D.A. For. Ser. Res. Pap.
NE-85.
________.
1981.
Removal
or retention of unmerchantable
saplings in Allegheny h a r d w o o d s : effect on regeneration
after clearcutting. J. For. 79: 280-283.
Oliver, C. D.
1978.
The development of northern red oak in
mixed stands in central New England.
B u l l . Yale Univ.
Sch. For. Environ.
Stud. : 91. Yale University,
New
H a v e n , Connecticut.
________.
1980.
Even-aged development
s t a n d s . J. For. 78: 201-203.

of

mixed-species

. 1981. Forest development in North America following
major disturbances.
For. E c o l . M a n a g e . 3: 153-168.

147
Oliver,
C. D . , Adams,
A.
B, and Zasoski,
R. j.
1985.
Disturbance
patterns
and
forest
development
in
a
recently deglaciated valley in the northwestern Cascade
Range of Washington, U.S.A.
Can. J. For. Res. 15: 221232.
Oliver,
C.
D . , and
Larson,
B.
C.
dynamics.
McGr a w H i l l , New York.
Peet, R. K.
Range.
Peet,

1990.

Forest

stand

1981.
Forest vegetation of the Colorado Front
Vegetatio 45: 3-75.

R. K. , and Christensen, N . L.
1987.
tree death.
BioSci. 37: 586-595.

Competition and

Pregitzer. K. S., and Barnes, B. V.
1984. Classification
and
comparison
of
upland
hardwood
and
conifer
ecosystems
of
the
Cyrus H.
McCormick
Experimental
Forest, upper Michigan.
Can. J. For. Res. 14: 362-375.
Roberts, M. R . , and Richardson, C. J.
1985. Forty-one years
of population change and community succession in aspen
forests on four soil types in northern Lower Michigan,
U.S.A.
Can. J. Bot. 63: 1641-1651.
R u a r k , G. A . , and Bockheim, J. G. 1987. Below-ground biomass
of 10-, 20-, and 32-year-old Populus tremuloides in
Wisconsin.
Pedobiologia 30: 207-217.
S a k a i , A. K.
northern

1990.
Sexual reproduction of red maple in
Lower Michigan. Am. M i d i . Nat. 123: 309-318.

S a k a i , A.
K . , R o b e r t s , M. R,
and J o l l s , C. J.
1985,
Successional
changes
in a mature
aspen
forest
in
northern Lower Michigan: 1974-1981. Am. M i d i . Nat. 113:
271-282.
Sander, I . L.
1990.
Ouercus rubra L. (Northern red oak) .
In Silvics of North America, Vol 2.
Edited by R. M.
Burns and B. H. Honkala.
U.S.D.A. For. Ser. Ag. H a n d . ,
654.
pp. 727-733.
Scheiner, S.
M . , Sharik, T. L, Roberts, M. R . , and Vande
Kopple, R.
1988.
Tree density and modes of tree
recruitment in a Michigan pine-hardwood forest after
clear-cutting and burning.
Can. F i e l d - N a t . 102: 634638.
Segura, G . , and Snook, L. C.
1992. Stand dynamics and
regeneration patterns of a pinyon pine forest in east
central Mexico.
For. E c o l . Manage. 47: 175-194.

148

Sharik,
T.
L . , Ford,
R.
H . , and
Davis,
M.
L.
1989.
Re p eatability of invasion of eastern white pine on dry
sites in northern Lower Michigan.
Am. M i d i . Nat. 122:
133-141.
S o k a l , R. R . , and R o h l f , F. J.
1981.
H. Freeman and Company, New York.

Biometry,

2nd Ed.

W.

S p r u g e l , D. G. 1976.
Dynamic structure of wave-generated
Abies balsaemea
forests
in the northeastern United
States.
J. E c o l . 64: 889-911.
Stoeckeler, J. H . , and Macon, J. 1956. Regeneration of aspen
cutover areas in northern Wisconsin.
J. F o r . : 54: 1316.
Stubblefield, G. W . , and Oliver, C. D.
1978.
Silvicultural
implications
of
the
reconstruction
of
mixed
alder/conifer stands.
In Utilization and management of
red alder.
U.S.D.A. For. Ser. Gen. Tech. R e p . : PNW-7 0 .
pp.
307-320
Veblen , T. T. , Hadley, K. S , Reid, M. S ,
and R e b e r t u s , A.
J.
1991. The response of subalpine forests to spruce
beetle outbreak in Colorado.
Ecology 72: 213-231.
Walters, R. S., and Y a w n e y , H. W.
1990. Ace r rubrum L. (Red
m a p l e ) . In Silvics of North America, Vol 2.
Edited by
R. M. Burns and B. H. Honkala.
U.S.D.A. For. Ser. Ag.
H a n d . , 654.
pp. 60-69.
W e n d e l . G. W. , and Smith, H. C.
1990. Pinus strobus L.
(Eastern white p i n e ) , p. 476-488.
In Silvics of North
America, VolTS
Edited by R. M. Burns and B. H. Honkala.
U.S.D.A. For.
Ser. Ag. H a n d ., 654.
pp. 476-488.
Whitney, G. G.
1987.
An ecological history of the
Lakes forest of Michigan.
J. E c o l . 75: 667-684.

Great

Wierman, C. A . , and Oliver, C. D. 1979. Crown stratification
by species in even-aged mixed stands of Douglas-firwes t e r n hemlock.
Can. J. For. Res. 9: 1-9.
Z a h n e r , R. , and Crawford, N. A.
1965.
The clonal concept
in aspen site relations.
In Forest-soil re l a t i o n s h i p s.
Edited by C. T. Youngberg.
Oregon State University
Press, Corvallis.
pp. 230-243.
Zeide, B.
1987.
Analysis of the 3/2
thinning.
For. S c i . 33: 517-537.

power

law

________.
1991.
Self-thinning and stand density.
37: 517-523.

of

self­

For. Sci.

CHAPTER 5
THE VERTICAL DEVELOPMENT OF EARLY SUCCESSIONAL FORESTS IN
NORTHERN LOWER MICHIGAN
ABSTRACT

(1)
of

Differences
tree

species

relative

that

successional

greater

juvenile

species and thus,
an

in height growth

early

in

status.

height

are

characteristic

understory

tolerance

Less-tolerant

growth

rates

than

in a mixed-species forest,

competitive

disturbance.

differ

rates

advantage

Species-level

after

height

a

and

species

have

more-tolerant
the former gain

stand-initiating

stratification

is

the

predicted outcome under this differential growth rate model.

(2)

Silvicultural

differential
development
species

growth

differing

an

model

Stratification

understory

is

even-aged

rates

relatively

in

status

relatively tolerant
growth

rate

suggests

that
of

a

species-level

even-aged

forest

is idealistic since height stratification among

successional
within

evidence

not

equaling
may

tolerant

always

forest.

species

fail

may

those
to

species

tolerance

In

consistently
fact,

of

competitively

a few years earlier than less-tolerant species,

height

of

a

uninhibited,
establish

are present

or are of vegetative-origin.

149

of

species.

individuals

released during the course of forest development,

as advanced regeneration,

expressed

and

less-tolerant
if

relative

individuals

achieve heights

develop
are

and

150

(2)

The establishment and height growth histories of three

tree species differing in understory tolerance and relative
successional
northern

status

Lower

included:

i)

paralleling

were

Michigan,

determining
understory

these f o r e s t s ; ii)

examined
U.S.A.
if

species

tolerance

changes

en v i r o n m e n t s , or

regenerative

s t r u c t u r e s ; and
structure

expected

under

of

height

was

a

forests
the

in

study

stratification

characteristic

of

assessing the degree of height variation

distributions,

height

two

Objectives

within species p o p u l a t i o n s ; jLii)

height

within

in

a

in

iv)

determining

if species age

individual

competitive

origin

assessing

these

influenced
the

forests

species-level

population

degree

deviated

differential

to

which

from

that

height

growth

rate model.
(4.)

In both forests,

Populus g r a n d i d e n t a t a . an intolerant,

early successional species was typically taller than Ouercus
rubra.

a

species,

mid-tolerant,
which

tolerant,
Ouercus

in

later
and

variable,

Acer

with

turn

relatively
was

taller

later
than

successional

species.

populations,

heights

some

individuals

successional

Acer

rubrum.

However,
were

equaling

within

often

the

a

highly

heights

of

P o p u l u s . Age differences had no influence on height patterns
either

among

mortality
since

within

contributed

many

sampling,

or

had survived.

to

suppressed

where

species

as man y

the

populations.

development

Populus

had

died

slowering growing

of
by

Differential
stratification
the

Ouercus

time

of

and Acer

151

(5) Growth rate differences among surviving stems determined
species stratification patterns and height variation within
populations.

Populus

A c e r , however,
had

height

some

growth

typically

individuals

rates,

outgrew
of

including

the

both
latter

juvenile

Ouercus
two

and

species

growth

rates,

equvilent to P o p u l u s .

Most Ouercus and Acer appeared to be

of

as

vegetative-origin,

individuals
height

of

growth

former

two

rates

were

likely

thus

rates

Populus.

allowing

pressure within
in

to

height

greater

of

growth

than
to

variation

of Populus

ramets.

potential

individuals

Stochastic

ability

Populus

species,

some

the matrix

competitive

contributed

all

the

seedlings,
of

were

in

ramets,

individual

For

juvenile

that

of

match

true

growth

competitive
or variation

genets,

likely

rate variation within

Ouercus

and Acer p o p u l a t i o n s .

(6)

The root systems of Populus proliferated following the

fires

that

initiated

the

forests

examined,

saturating stands w ith rapidly growing ramets.

quickly

In contrast,

new Ouercus and Acer ramets were restricted to the locations
of remnant s t u m p s . Populus numerically dominated the forests
because a large proportion of growing space was occupied by
its root

systems

and

r a m e t s , relative to Ouercus

and A c e r .

Vertical domination by Populus was apparently the result of
competitive

inhibition

A c e r , not because of
in

juvenile

height

of

most,

inherent,
growth

but

not

a l l , Ouercus

and

species-specific differences
rates.

This

suggests

that

152

numerical
limit

dominance

height

growth

by Populus was
rate

of

most

sufficient
Ouercus

in itself

and Acer

and lead to species-level vertical stratification.

to

sprouts

153

INTRODUCTION
P hysiologically-

or

in height growth rates,

morphologically-based

differences

particularly juvenile growth rates,

are characteristic of tree species that differ in understory
tolerance

and

relative

Drury & Nesbitt
Likens
1988,

1979,

1973;

successional

Marks

p p . 115-128;

1975;

Canham

status

Bazzaz

(Loach

1979;

& Marks

1985;

Bormann
Tilman

1990; K o z l o w s k i , Kramer & Pallardy 1991, pp.

Species-specific

growth

expressed

in

even

availability;
disturbance

for

example,

(Marks

species

1975;

1977)

over

forest

development

height

growth

with

&

thought

increased

following

Bormann

are

a

have

an

more-tolerant
because

rates,

and

of

initial

superiority quickly

thus

to

be

resource

1979,

p.

advantage

early

inherently

would

1985,

stand-initiating

Likens

species

and

100-102).

their Figure 1). Under this view,

should

Harper

(vertical)

differences

environments

Canham & Marks 1985,
tolerant

rate

1970;

in

less(sensu

even-aged

faster

achieve

125;

juvenile

competitive

following a stand

initiating

disturbance.
Indeed,
accepted

species

rankings

height

of

stratification

understory

tolerance

paralleling
and

relative

successional status is a common characteristic of many evenaged forests

(Marquis 1967,

1981;

Oliver

1978;

Stubblefield

& Oliver 1978; Bormann & Likens 1979; Wierman & Oliver 1979;
Hibbs

1983;

O'Hara

Clatterbuck & Hodges

1986;
1988;

Oliver & Larson 1990).

Kelty

1986;

Foster 1988;

However,

Larson

1986;

Hix & Lorimer 1990;

silvicultural

studies have

154

shown a population-level,
be

idealistic

expressed

within

range

within

quite

wide,

height

since

stratification

an

even-aged

populations
with

growth

some

rates

tolerant species
Oliver

differential g r owth rate model to

1980;

of

often

stand.

In

equal

(Cayford 1957;

inconsistently

fact,

more-tolerant

individuals

that

Guldin

is

species

attaining

or

exceed

Lorimer

may

fail

1985;

height
can

heights

those

Stubblefield

&

the

be
and

of

less-

& Oliver

1978;

Kelty

1986;

Hix

&

Lorimer 1990).
Stratification
spacing

between

tolerance
1978;

is

tolerant

less

Radosevich

&

of

an

(e.g.

faster

species

comparable to

may

be

1990;

or

juvenile
maintain

less-tolerant

understory

the

the

catch

of
up

tolerant

regeneration
1986;

species

(Marquis

Foster 1988;

is

1981;

Deal,

Oliver
Smith

competitive

individual
course

Bormann

&

&

1985;

of

a

stand

1991).

Facing

initially

slower

individuals

of

growth r a t e s . Individuals

of

early

species

if

to

height

growth

relative

the former initially exceed that of the latter;
when the

relative

Oliver & Larson

when

individuals

able to

in

Lorimer

during

see D e a l , Oliver

species may

species with
tolerant

&

suppressed

improves

resource-limitation,

growing

1992)

whe n

Stubblefield

Guldin

initially

species

development

1957;

Clatterbuck & Burkhardt

Shainsky

environment

1979;

develop

differing

(Cayford

Oliver

&

1986; Oliver,
1990;

individuals

large

Wierman

to

initially
Guldin

&

rates

heights

of

for example,

present
Lorimer

Oliver & Bormann 1991),

as

advanced

1985;

Kelty

or when it

155

establishes
species

several

(Newton,

years

earlier

than

El Hassan & Zavitkovski

be

considered

even-aged

(Smith

less-tolerant

1968;

& Oliver 1978; Chapter 2). By definition,
still

a

Stubblefield

such forests would

1986) . Regenerative-

origin may also influence height relationships since sprouts
of hardwood
over

species

seedlings

of

often have
the

same

an

initial

species

growth

(Jensen

1968; Oliver 1978; Beck & Hooper 1986).

Thus,

may

a

be

less

likely

predominantly
potential
specific

of

for
in

to

develop

sprout-origin

compensation

juvenile

mechanisms outlined above,
within

a

particular

regenerative-origin,
studies

have

assessed

stratification

forest

comprised
Given

differences

growth

Wilson

in

rates,

the

species-

through

the

the expression of stratification

forest
or

1943;

individuals.

of

height

in

advantage

may

depend

population

the

influence

on

age

stand

structure,

distributions.

of

these

Few

compensating

mechanisms on the development of vertical structure in evenaged forests.
This
three

tree

relative

study

examined

species

height

differing

successional

grandidentata

the

Michx.

status

(bigtooth

northern Lower Michigan.

in

and

age

understory
in

two

structures
tolerance

mature

aspen)-dominated

of
and

Populus

forests

in

These forests are typically viewed

as even-aged, having initiated following region-wide logging
and

repetitive

twentieth

slashed-fueled

century

(Kilburn

wildfires

1957,

1960;

at

the

Graham,

turn

Harrison

of
&

Westell 1963; Roberts & Richardson 1985). The fires promoted

156

vigorous

vegetative-regeneration

spreading

clonal

root

observations

indicated

Acer

L.

rub rum

species

of

of

suggests

&

after

forest

intolerant,

tolerant,

later

Objectives
species

(red oak)

additional

and

and
tree

in

studied.
A cer

from stumps

the

Limited

in

the

Harrison

The

&

three

successional
later

mature

study

Acer

in

species

(Spurr

&

(i)

1963;

to

mid-

Ouercus.
Barnes

to

1980).

determining

(Ai)

in

s t a t u s ; from

paralleling

for e s t s ;

the

differ

Populus.

included:

these

from

Westell

successional

stratification

occurred

remnant

and relative successional

early

height

tolerance

fire

successional

for

initial

importance

Ouercus

1985).

relatively

tolerant,

only

forests

(Graham,

Richardson

understory tolerance
very

the

rapidly

1966).

rubra L.

numerical

many

from

forests may also be of vegetative-origin,

having regenerated
predisturbance

were

particular

that

Populus

(Barnes

that Ouercus

maple)

the

Populus-dominated

Roberts

systems

consistent

overstories
evidence

(red

of

if

understory

assessing

the

degree of variability in heights within species p o p u l a t i o n s ;
(i i i ) determining
changes

in

if age differences,

an

individuals

influence

on

population

assessing

the

degree

to

regenerat i v e - m o d e , or

competitive
height
which

environment

structures;

vertical

had

and

structure

any
(iv)

in

the

forests examined was accounted for by a developmental model
emphasizing species-specific differences
growth rates.

in juvenile

height

157
STUDY LOCATIONS
R e s earch
National

was

Forest

conducted
(Huron)

within

in

the

Huron-Manistee

northeastern

Lower

Michigan

(latitude 44° 15' to 45° 00' N, longitude 33° 15' to 84° 45'
W)
in

and the University of Michigan Biological Station
the

extreme

(latitude
areas

45°

are

geology

40

of

both

sands

Padley,

1989) .

were

Lapin,

N,

portion

longitude

separated

outwash

areas

northwestern

by

study

84°

of

40'

overlaying

110

consisted

till

Michigan

W) . The

approximately
areas

Lower

two

km.

study

Surficial

primarily

deposits

(UMBS)

of

(Cooper,

deep
1981;

Soils throughout the majority of both study

classified

as

Entic

Haplorthods

1990; Soil Conservation Service,

University,

Forestry

descriptions

of

the

Department

physiographic

1981;

1991; Michigan State

data
and

(Cooper,

file).

climatic

Detailed

settings

of

the two study areas are found in Chapter 3.
Overstory
similar

composition

(Chapter

3) .

Mean

of

the

two

study

relative density

areas

was

of Populus was

approximately 60% at both locations. Mean relative densities
of Ouercus
Huron

and Acer were

forest,

Additional

and

overstory

some of the UMBS
strobus L.

8%

15% and 25%,
and

species

plots

10%,
that

included

respectively,

respectively,
occurred

Pinus

at

in the
UMBS.

infrequently

in

resinosa A i t . , Pinus

and Betula papyrifera Marsh.

Mean

(+ SE)

total

basal area and total density were both substantially higher
in

the

Huron

respe c t i v e l y ) ,

forest
than

(41.6+2.0

m 2/ha

at

(28.6+1.5

UMBS

and

1117+48
m 2/ha

stems/ha,

and

890+44

158

stems/ha,

respectively).
METHODS

Stand and plot selection
P o p u l u s . O u e r c u s . and
examined
UMBS.

within

Stands

locations

five

were

Department
belonged

1990?

data

to

(following

stands

selected

(Lapin,

similar
et

at
from

population

structures

Huron

four

and

larger

data

All

ecological

land

1

ha;

(ii)

initiation;

a l . , 1982;

Pregitzer

free

and

at

from

(i i i )

0-5%

a

&

at

both

Forestry

study

area

classification

Selected stands met the following criteria:
of

pools

stands within

were

stands

Michigan State University,

file).

Barnes

Acer

Barnes

units
1984).

(i) minimum size

obvious

disturbance

slopes. The

distance

since

between

stands w i t h i n each study area ranged from 1-4.5 km.
Four
stand.

circular

Plot size

larger plot

plots

were

randomly

located within

each

was 272 m 2 at Huron and 475 m 2 at UMBS.

size

was used at UMBS

so that a

A

similar number

of Ouercus and Acer stems might be sampled as in the higher
density Huron forest.

The only criterion for plot selection

was that,

at least one overstory

Ouercus

if possible,
and

criterion,
Acer.

Acer

plots

Initial

genet
were

occur

never

locations

in

each

rejected
for

two

plot.

because
of

(dbh > 10 cm)

the

Given
of

36

a

this

lack

plots

of

were

rejected because of a lack of O u e r c u s .

The location of one

of

additional

these

selected
than

plots
plot

stands,

was
did

were

retained
not
u sed

since

contain
as

any

an

Ouercus.

replicates

in

randomly

Plots,
all

rather

subsequent

159

statistical
since

analyses.

sampling

Replicating

was

conducted

by

physical

c h a r acteristics,

Stands

were

relatively

simply

subjective

plots

within

characterized
site

on

was

forests

homogeneous
and

units

justified
landscapes

composition,

disturbance
defined

by

history.
harvesting

patterns or road locations within each landscape and not on
any

obvious

criterion

which

may

have

influenced

overstory

development.

Stem sampling
On each plot,

selected Populus and all Ouercus and Acer

at least 1.5 m tall were destructively sampled as part of a
related study of stand development.

This

paper

reports

on

the development of the i nitial, post-disturbance cohorts of
Populus.

Ouercus and

Acer at

each

study

youngest

Ouercus and

Acer in

these

location.

cohorts

The

established

several decades prior to the next oldest individuals of the
species

(see Chapter 4).

For
sampled

Ouercus and
from

individual
dominant

from

or

decaying

A c e r , only

multi-stemmed
the

dominant

codominant)

was

the tallest

genets.
crown

class

randomly

ramets) and destructively

For

ramets
Populus.

(all

ramets

selected

sampled

on

were
one
were

(rejecting
each

plot.

Sampling intensity for Populus was much lower than for other
species because little height variation was observed within
a plot

and little age variation among

Both total

heights of

additional

ramets was

expected.

P o p u l u s . estimated with

a

160

clinometer,

and

ages

these expectations
All
heights

sampled
were

of the

were

determined.

thereafter.

felled

Stumps

at

and

0.5

stems

m

Marking continued

to the

leader.

and

were

from the base to 2 m and at

to the end of the dominant
at ground

P o p u l u s . substantiated

(see R e s u l t s ) .
stems

0.25-m intervals

sampled

marked

at

1 ra intervals

1-m multiple

closest

Stem sections were cut

level and at each measurement

interval up to a 3

cm diameter top. The regenerative-mode of all genets
or seedling)

total

(sprout

was assessed when possible.

Stem analysis
The terminal

leader of each stem was aged in the

field

by counting the number of terminal bud scale scars preceding
each height i n t e r v a l .

Stem sections were removed to the lab

where they were sanded to a smooth surface and wetted to aid
ring

examination.

number
axis

on
on

at

Ages

least

two

obiique-shaped

were

determined

by

radii

(typically

a

stems)

of

each

dissecting microscope.

The

oldest age

the

was

recorded.

counts

differed,

counting
long

and

section

of each

ring
short

under

section,

Precision

of

a
if

ring

counting was assessed by periodically recounting 20% of the
stem

sections

seldom

from

differed

by

lots

of

more

than

used to determine time
the height
Stem

growth

heights

at

15-25
one

year.

of establishment

trajectories
10

sections

year

stand

of

each

age

each.

Section

Recounts
ages

were

and to

reconstruct

sampled

individual.

intervals

were

derived

161

from the

growth trajectories

(see Data

analysis b e l o w ) . in

some instances interpolation of stem height at a given stand
age was

required since

the stem

stem height but not age.
of

age-height

examined to

plots

sampling method

controlled

When this was required,

the shapes

for

the

insure accurate

individuals

determination

involved

of

were

height at the

10-year stand-age of interest.
The
were

age-height

examined

might

for

reflect

environment
following
growth

for

of

in

ability.

procedure

analysis

examined

evidence

changes

or

a

trajectories

an

similar

Ouercus
or

and

Acer

release

that

in d i v i d u a l 's
episodes

to

one

1984).

slope

each

suppression

These

(Lorimer

obvious

of

were

suggested

Height

changes

competitive
defined

for

radial

trajectories

that

persisted

were

for

at

least 15 years and resulted in height growth rates that were
minima l l y 50%

above or below growth

rates

for the previous

15 yea r period.

The 15 year minimum was used to filter out

short periods

increased or reduced growth

related
status,

to

of

factors

such

as

other

climate

than

change

a
or

change
insect

that might be
in

competitive

defoliation.

The

50% mi n i m u m growth rate change was used to filter out slower
declines

in height

increment

that

occur

as

a

stem matures

among

species

(Oliver & Larson 1990; Zeide 1991).

Data analysis
Total
using

heights

were

random-order

compared
crossover,

split-block

over

time

repeated

162

measurement analysis

of variance.

were considered blocks,
"mensurational

plots

species were considered random-order

t r e a tments” within

repeated measure.

For these analyses,

blocks,

Wit h this design,

and time was

the

the overall effects of

species can be analyzed as a randomized b l o c k design,

since

no nuisance trends are associated wit h a random sequence of
"treatments” within a block
Variances
within

a

(Gill 1978).

in height over time were often heterogeneous

species

(F-max

test,

P<

0.2).

Additionally,

the

correlation structure of most data sets were not homogeneous
across

time

Gurevitch

(sphericity

& Chester

assumption

1986).

In most

rejected
cases,

homogeneous among species at specific times.
sets,

log

or

homogenize

square-root

variances

required).

In

some

transformations

over

time

instances

(and

these

at

P=

0.25;

variances

were

For some data
were

within

used
time

to
when

transformations

were

effective at homogenizing variance over time but resulted in
increased

heterogene ity

among

variances

within

time

and

lessened the normality of residuals. When this was the case,
untransformed
adjustments

data

were

analyzed

using

conservative

for heterogeneous variance-covariance

as described

in the next two p a r a g r a p h s .

structure

All of the final

data sets used met the assumption of normality of residuals
(Shapiro-Wilke test,
When
over time,
x

species

P> 0.1).

variance-covariance
F-tests

structure

was

heterogeneous

for the overall effects of time and time

interactions

were

evaluated

using

the

highly

163

conservative

Greenhouse-Giesser

1978) .

test

This

reduces

from (b-1), (b-1)(r-1)(t-1)
number of time periods,
the number

the

degrees
degrees

to 1,

of

freedom

freedom

(r-1)(t-1), where b is the

For time x

degrees

of freedom

are

1 ) (t-1)

to (t-1),

(r-1)(t-1).

freedom

was used

evaluate

reduced

effect when time x species

(Gill

for time

r is the number of blocks,

of treatments.

to

of

species

from

and t is

interaction,

(b-1)(t-1),

Box's adjusted

overall

(b-1)(r-

degrees

F-tests

for

of

species

interaction was not significant,

but variances among species were heterogeneous

(Gill 1978).

Individual species m e a n s at specific time periods were
separated using
priori

Bonferroni

hypotheses

understory

contrasts

regarding

tolerance

height

differences

conducted to determine whe t h e r
taller

than

(ii)

Ouercus

time.

These

height
not

Ouercus
was

contrasts

patterns

in

imply

that

to

influencing

the

contrast

adjusted,

according

to

within

variance-covariance
c a s e s ) . For

and

of
means
Gill

a

specific

(tolerant)
to

may
Acer
and

have

statistical

were

at

allow

been

was

time
the

and
same

ranking
to

of

Populus.

consistently

throughout

the

degrees

freedom

were

variances

were

(1978),

were

Test

a
on

(intolerant)

A c e r , relative

specific

matrices
all

Ace r

at

test

based

species.

(i.) Populus

designed

Ouercus

growth

of

(most

wer e

Ouercus

Variances

heterogeneous

than

to

patterns

among

(mid-tolerant)

taller

(two-sided)

times
not

of

when

(some
un i f o r m

tests, a

forests.

cases)
across

type-one

probability of 0.05 was considered significant.

or
time

error

164
RESULTS
Regenerative-modes
Regenerative-origin of Ouercus and Acer was assessed by
assigning

genets

to

vegetative-origin

one

of

following

(multi-stemmed

probability of vegetative-origin
highly
which

convoluted
may

be

s t u m p s ) ; and
uniform,
of

or

crescent

indicative

genets);

stem

Ace r

development

(i i i ) seed-origin

genets

in

both

(ii)

bases,

high

(single-stemmed

forests

both

around

approximately circular stem b a s e s ) .

the

were

The

apparent

seed-origin

the smallest height classes.
were

found

tallest

in

a

A cer

wide

sampled

genets

vegetative-origin,
stemmed .

however

with

Ninety percent
apparently

were

of

(Table

restricted

to

Apparent sprout-origin genets

range
in

genets

of

remnant

vegetative-origin and 90% of these were multi-stemmed
5.1).

(i)

(single-stemmed genets with
shaped

of

categories:

of

height

this

study

not

all

c l a s s e s . All

of

the

appeared

be

of

of

these

to
were

multi­

Eighty-four percent of the Ouercus genets

in the

Huron forest appeared to be of vegetative origin; two-thirds
of these w ere multi-stemmed
of

the

only

Ouercus

one-third

genets
of

(Table 5.1).

appeared

these

were

to

be

of

At UMBS,

only 66%

sprout-origin

multi-stemmed

(Table

and

5.1).

Both apparent sprout-origin and seed-origin Ouercus occurred
in a range of height classes,
individuals,

but

apparent

including some of the tallest

sprout-origin

found in the smallest height classes.
high

early

radial

growth

rates,

genets

were

not

Most Ouercus did have

relative

to

later

years,

165

Table 5.1.
Regeneration methods of Ouercus rubra and Acer
rubrum genets in the a) Huron and b) UMBS forests.

Regeneration method
Vegetative—

Likely vegetative—

Seedling—

n—

a) Huron
Q. rubra
A. rubrum

0.57—
0.83

0.27
0.06

0.16
0.11

75
92

0.33
0.11

35
69

b) UMBS
Q. rubra
A. rubrum

0.22
0.78

0.44
0.11

^ Multi-stemmed genets.
-Single-stemmed genets with highly convoluted stem bases.
^Single-stemmed genets with uniform, circular stem bases.
— Number of genets.
— Proportion of total genets sampled.

4

166

including

those

that

appeared

to

be

of

seed-origin

(unpublished d a t a ) . Rapid early radial growth suggests that
even apparent seed-origin Ouercus may have b een vegetativesprouts

(Sander 1990) .

Population age and heig h t structures
All

sampled

resprouted,

stems

within

a

in the Huron
10

approximately 70 years

year

forest established,

period

prior to

following

sampling.

20 sampled Populus differed by no more
the

UMBS

forest,

Within stands,
within

ages

the

age

ranged

from

4 to

plot

ranged

The ages of the

than

from

four years.

74

to

containing

a

distributions
15 years,

Ouercus

for

with

the

youngest

growing A c e r .
older

than

that

was

they

Bormann

distributions
species
plot)

appeared
as

65.4+0.2,

was

they

Their

likely

similar

if

all

Also

species
of

approximately

20

either

had

In both study
the

dropped

slowest

minimal

(Huron,

since

do

on
mean

UMBS,

n=16)

near

(Larson

plot

age

ages

among

species

on a

Populus= 6 5 .9+0.2,

n=19;

location

rings

sometimes

individuals pooled by

A cer= 6 5 .8+0.2.

at

years.

exception

always

influence

Ouercus- 7 3 .8+1.2. A c e r = 7 1.0+1.1,
different

were

suppressed stems

1965).

(years+ SE;
were

individuals

In

These individuals may have been several years

their stem bases,
1956;

82

all

the

years older than the oldest of its associates.
areas,

disturbance

Populus ages differed by 4 to 10 years.

stands,

combined
one

stand

or

Ouercus=

Populus= 7 3 .3+1.1.

and not significantly

(Friedman's

randomized

block

167

method: Huron,
power

to

P= 0.368; UMBS,

detect

alpha=0.05)

a

mean

was high

Mean

age

P= 0.051).

For both forests,

difference

of

±2

years

(at

(Huron= 0.99; UMBS= 0.82).

heights

of

Populus.

Ouercus

and

Acer

(all

individuals pooled by species on a p l o t ) , over 10 year stand
age

intervals,

were

are

extended

shared

by

to

all

Huron forest,
0.0001;

time

70

in

the

These

maximum

increase

heights

that

locations.

temporal

differed

among

different at all stand ages

among

species

Table

x

time

5.2),

increase

indicating

were

similar

among

effect was significant
Populus was
(Fig.

was

that

taller
not

Populus was taller

than

A c e r . At

significant

temporal

patterns

s p e c i e s . The

taller than Ouercus at almost

all

UMBS,

(P=

0.096;

in

height

overall

(P< 0.0001; Table 5.2).

species

Within time,
time periods

5.1), however the differences were significant only at

stand ages of 60 and 70 years

(P< 0.05; Table 5.2).

was significantly taller than Acer at all
(P<

except

The differences followed the

in turn was

interaction

of

significantly

(P< 0.05-0.01; Table 5.2)

(P> 0.05; Table 5.2).

then Ouercus which

patterns

s p e c i e s . Within

were

predicted order based on shade tolerance;

species

interval

(P<

indicating

study

year

species x time interaction was significant

height

both

10

comparisons

the

5.2),

within

5.1.

In

periods,

10 years

Figure

years,

stands

Table

cumulative

shown

0.01;

Table

5.2).

The

relative

Ouercus

10 year intervals

height

relationships

among species remained unchanged at the time of sampling in
the older UMBS stands

(data not s h o w n ) .

168

Figure 5.1.
Heights of Populus a r a n d i d e n t a t a . Ouercus
r u b r a . and Acer rubrum at 10 year stand age intervals
in the Huron forest (a) and the UMBS forest (b). Huron
v alues are means (+SE) of 19 observations. UMBS values
are backtransformed means (+95% confidence intervals)
of 16 square-root transformed observations. Q. rubra
and A. rubrum values are the means of all individuals
pooled by species on each plot.

169
Figure 5.1

CUMULATIVE

STEM

HEIGHT

(M )

30
25

-

20

-

15

-

10

-

Populus
A -A Quercus
Acer

sr'

5

-

HURON
0

10

20

30

40

50

60

70

80

30
25

-

20

-

15

-

10

-

5

-

UMBS
0

10

20

30

40

50

60

Y E A R S A F T E R P L O T IN ITIATIO N

70

80

Table 5.2.
Repeated measure analysis of variance and species contrasts for mean heights
of Populus a r a ndidentata. Ouercus rubra— and Acer rubrum— over time in the Huron and
UMBS forests.

Huron

Source

18
2
36
6
108
12
216

(27)
(1)
(2)
(36)

SS

641.437
1329.998
504.199
16656.831
203.734
357.171
207.089

P > F—

0.0001

(0.0001)

0.0001

(0.0001)

0.0001

(0.0001)

df

15
2
30
6
90
12
180

SS

8.171
44.317
(26)
13.161
297.292
(1)
3.985
0.924
(2)
(30)
5.424

P > F

0. 0001

(0.0001)

0.0001

(0.0001)

0.004

(0.096)

170

Plot
Species
Error (Spp.)
Time
Time x Plot
Time x Spp.
Error (Time)

df£

UMBS—

Table 5.2 (cont'd).

Huron
P. crrandidentata vs.
0. rubra

Stand age
(years)

1.250
3.986
2.788
2.711
3.931
4.876
5.406

df—

36
36
36
36
36
36
36

**
*
*
**
**
**

0. rubra v s .
A. rubrum

*b
1.647
2.767
4.660
4.140
3.294
3.231
2.809

df

36
36
36
36
36
36
36

P. crrandidentata vs.
0. rubra

tb

*
**
**
**
**
*

0.778
1.596
1.491
1.416
1.960
2.635
2.424

df

27
30
30
30
30
23 *
21 *

^All individuals pooled by species on each plot.
— UMBS analysis done on square-root transformed data.
— Degrees of freedom (df) and probabilities in parentheses are associated
(Species)
or Greenhouse-Giesser (Time and Time x Species) conservative
— tb~v a lues are for Bonferroni contrasts; df values may vary depending on
structure within and across time.
Critical value= +tfc, a/2, 2, df; * significant at P= 0.05, ** significant

rubra vs.
2- :
A. rubrum

tfc>
4.195
4.788
4.994
4.585
5.797
5.471
5.663

with Box's
F-tests.
variance
at P= 0.01.

df

30
30
30
30
30
30
30

**
**
**
**
**
**
**

171

10
20
30
40
50
60
70

t^~

UMBS

172
Population height variation
The examination of species mean heights over time
5.1)
and

suggests
Ace r

plots,

in

the

guite

examining
plots

Huron

individual

often

In

stratification
and

heights

variable.

Populus

UMBS

for

This

time-height

of

the

both

Ouercus

f o r e s t s . However,

within

latter

two

variability

can

reconstructions

the

Huron

forest,

respectively.
and

5.8

since

the

At UMBS,

(2.4)

some

additional
Ages

for

mean

(+1

of

m.

the

These

seen
of

plot-level

ramets

ranges
came

by
the

in Figure 5.2.

(3.2) m and 7.3

height

were

height
(0.4) m,

are

from

(3.4)

conservative
genets

having

slower growing but similar aged ramets.

(mean years+ SE)

height ranges

be

the analogous ranges were 5.4

shorter

shorter,

SD)

species

several

(one randomly selected from each stand)

ranges of Ouercus and Acer were 8.7

m

over

(Fig.

of individuals at the extremes of the

on a plot were

similar

for all

species-study

area combinations

(Huron O u e r c u s . t a l l = 6 5 .5+ 0.3 vs.

short=

65.6+0.5,

n=16;

Huron

64.8+0.6,

n=19 ;

UMBS

70.0+1.2,
72.1+

n=

3.9,

15;

n=8) .

Acer.
Acer.

UMBS
In

t a l l = 6 6 .2+0.4

vs.

short=

t a l l = 7 2 .9+1.6

vs.

short=

Ouercus.

all

cases,

t a l l = 7 2 .5+1.5
the mean

vs.

short=

differences

were

not significantly different from zero

(paired t-tests: Huron

Ouercus.

0.099;

P=

0.837;

0.156;

Wilcoxon's

detect

a m ean

Huron
test:

UMBS

difference

high

(0.80 or greater)

UMBS

(0.50).

Acer.

of

P=

O u e r c u s . P>
+2

for all

years

UMBS

Acer.

P=

0.10).

Power

to

(at alpha=

comparisons,

0.05)

except Acer

was
at

173

Figure 5.2.
Height growth
reconstructions of
Populus
q r a n d i d e n t a t a . Ouercus r u b r a . and Ace r rubrum within
randomly selected plots from the Huron forest (a) and
the
UMBS
forest
(b) . Each
line
represents
an
individual stem.

174
Figure 5.2

(a)

50
25-

20

P o p u lu s
Q u e rc u s
A ce r

2520-

15-

15-

10-

P lo t 1

CUMULATIVE

STEM

HEIGHT

(M )

0

P lo t 7

5 10 15 20 25 50 55 40 45 50 55 60 65 70 75

0

50

50

25-

25-

20-

20-

5 10 15 20 25 50 55 40 45 50 55 60 65 70 75

1510-

5-

P lo t 9
0

5 10 15 20 25 50 55 40 45 50 55 60 65 70 75

P lo t 14
5 10 15 20 25 50 55 40 45 50 55 60 65 70 75

YEARS AFTER PLOT INITIATION

2520-

10-

P lo t 19
0

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

YEARS AFTER PLOT INITIATION

175
Figure 5.2

(b)

25

25

P o p u lu s
20

20-

Q u e rcu s
A cer

15

(M)

5

HEIGHT

10

0

10-

P lo t 2

P lo t 6

0 5 10 15 20 25 JO 35 40 45 50 55 60 65 70 75 SO 85

STEM

25

25

CUMULATIVE

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

20

20 -

15

10

10-

5

5-

P lo t 12

P lo t 14

0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

YEARS AFTER PLOT INITIATION

YEARS AFTER PLOT INITIATION

176

There

was

less

within plots.
at

the

time

variability

In the Huron
of

sampling

analysis was 24.5

in

the

forest,

of

the

the mea n

20

difference

0.773) .

In

additional

not

UMBS

sampled Populus was
the

used

was 25.2

significant

forest,

21.7

Populus

height
in

stem

the

(0.3)

was

m,

also

(1.2) ra.

(paired

mean

had

at

least

The

t-test,

16

while the m ean height

of

(0.5)

of

m.

one O u e r c u s . and

Again,

the

(P= 0.970).
some

had

A c e r , that approached the height of Populus at 70 years
examples see Fig.
to

attain

5.2).

dominant

The potential

canopy

p=

the

21.7

height

paired heights were not significantly different
Many plots

Populus

(0.5) m. The mean height of the additional

was

the

of

(+ SE)

Populus

P o p u l u s . estimated with a clinometer,
mean

heights

an

(for

for Ouercus and Acer

positions

was

assessed

by

tabulating the number of individuals on each plot that were
within 2 m of
sampling

(or exceeded)

(Table

5.3).

Populus in height at the time of

In

the

Huron

forest,

37%

of

the

Ouercus ramets had total heights that approached or exceeded
the heights of P o p u l u s . compared to 13% for A c e r .

At U M B S ,

53% of the Ouercus ramets approached or exceeded the heights
of the P o p u l u s . while

only 6% of the Ace r ramets were this

tall.
Similarity

in

height

growth

rates

of

Populus

and

dominant Ouercus and Ace r

(the tallest stem from each of the

plots

individuals

that

assessed
species

by
pairs

contributed
comparing
at

10

mean

year

to

cumulative

stand

age

Table

5.3)

heights

intervals

was

between

(Fig.

5.3).

177

Table 5.3.
Numbers and density of Ouercus rubra and Acer
rubrum genets by height class in the a) Huron and b) UMBS
forests.

Species
0. rubra

A. rubrum

a) Huron

n—

#/ha—

n

#/ha

1. Dominant height class—
a) equals or exceeds
P. grandidentata
b) 0.5-1 m shorter than
P. grandidentata
c) 1-2 m shorter than
P. grandidentata

28
13

52
24

(10)
(9)

12
6

22
12

5

10

(4)

3

6 (3)

10

18

(6)

3

6 (3)

2. Subdominant height class—

47

84

(9)

80

149

Total cohort

75

136

(20)

92

171 (21)

b) UMBS

n

#/ha

n

#/ha

1. Dominant height class
a) equals or exceeds
P. grandidentata
b) 0.5-1 m shorter than
P. grandidentata
c) 1-2 m shorter than
P. grandidentata

19
10

25
13

4
3

5 (3)
4 (3)

2. Subdominant height class

16

25

(6)

65

87

(13)

Total cohort

35

50

(9)

69

92

(14)

(6)
(5)

4

5 (2)

5

7 (3)

(6)
(5)

(15)

----

—
1

1 (1)

^Total number of genets sampled, pooled across plots.
— Values are means (±SE) of 20 and 16 plots in the Huron and
UMBS forests, respectively.
— Dominant Q. rubra and A. rubrum were those that exceeded,
equaled (+0.5 m) or approached (2.0-0.5 m shorter) the
height of P. grandidentata on the same plot.
— Subdominant Q. rubra and A. rubrum were more than 2 m
shorter than P. grandidentata on the same plot.

178

Figure 5.3.
Heights of Populus grandidentata compared to
dominant Ouercus rubra or Acer rubrum at 10 year stand
age intervals in the Hu r o n forest (a,b) and the UMBS
forest (c, d ) . Huron values are backtransformed means
(+95%
confidence
intervals)
of
16
log
(y+1)
transformed observations for Q. r u b r a . and 10 log
(y+1) transformed observations for A. r u b r u m . UMBS
values are means (±SE) of 13 and 3 observations for Q.
rubra and A. r u b r u m . respectively.

179

Figure 5.3

30
P opulus
Q uercus

20

HURON

30-:
P opulus
Acer

i—

x

o
LjJ

X
HURON

in

30

LU 25
>
20

A~A

P opulus
Q uercus

i 15
UMBS

30
# • • • P opulus
Acer

20

UMBS

0

10

20

30

40

50

60

70

YEARS AFTER PLOT INITIATION

80

180

There were

only three plots wit h dominant Acer at UMBS,

few

included

to

total

be

height

of

was only 0.5 m

in

these

these
less

analyses,

individuals at

than Populus

however,

the time

from the

the

of

to

mean

sampling

same plots

(Fig.

5.3) .
Time

x

species

effects

were

Populus-Ouercus comparisons
Table 5.4),
forest
the

Table

cumulative

species pairs.

5.4),

height

Table

comparisons

the

5.4),

forest
were

heights

within

only

20

at

Table

each

Populus=
Populus=

0.87;

P = 0 .935;

the shapes

similar

for

the

between

30

years

0.05)

power

high
Overall

A c e r -Populus
5.4),

(Fig.

although
For

significantly
initiation

65.9+0.2
66.1+0.2

vs.
vs.

detect

species
in

height
example,

different
(P<

(0.29)

above

at

both

comparison

5.3).

plot

for

(at alpha= 0.05) was moderately low

the

UMBS:

in heights

mean

(years+ SE)

to

was

0.94).

were

after

P = 0 .302;

a i m

of

of

Power to detect a 1-m mean

Table

minor

time-periods

(Huron:

a similarity

UMBS=

(P=0.006;

However,

Mean ages
in

alpha=

clearly

and

5.4)

difference

(at

significant

differences

were

comparisons

indicating

(Huron=

was

Huron

both

Overall species effects were non-significant

difference

effect

UMBS:

indicating that

each 10 year plot age i n t e r v a l .
height

P=0.086;

trajectories

Populus-Ouercus

P = 0 .550;

(Huron:

in

and in the Populus-A cer comparison in the Huron

(P=0.077;

in both

non-significant

0.05;
height

of Populus and the other species
comparisons
Quercus=

Acer=

were

65.4+0.3,

66.5+0.5,

n=10;

similar

(Huron

n=16;
UMBS

Huron

Populus=

Table 5.4.
Repeated measure analysis of variance and specific contrasts for heights of
Populus qrandidentata and dominant Ouercus rubra and Acer rubrum over time in the Huron
and UMBS forests.

P. qrandidentata vs. Q.

rubra

Huron—

Source

dfk

Plot
Species
Error (Spp.)
Time
Time x Plot
Time x Spp.
Error (Time)

15
1
15
6 (1)
90
6 (1)
90 (15)

SS

1.794
0.025
0.332
116.584
0.626
0.143
0.481

UMBS

P > F—

df

0.302
0.0001

(0.0001)

0.0005

(0.086)

12
1
12
6 (1)
72
6 (1)
72 (12)

SS

347.430
0.982
31.303
6890.447
102.151
1.287
51.518

P > F

0.550
0.0001
0.935

(0.0001)

Table 5.4 (cont'd).

P. arandidentata vs . A . rubrum

P. qrandidentata vs. A. rubrum

Huron

Source

df

SS

Plot
Species
Error (Spp.)
Time
Time x Plot
Time x Spp.
Error (Time)

9
1
9
6
54
6
54

0.850
0.288
0.164
70.130
0.453
0.087
0.387

(7)
(1)
(1)
(9)

Huron

P > F

0.003
0.0001
0.077

Stand age

(0.006)
(0.0001)

10
20
30
40
50
60
70

(yrs)

(contrasts)

tb~

1.269
3.050
2.950
1.678
1.807
0.824
0.829

df—

9
9 *
8 *
8
6
9
9

— Huron analyses done on natural log (y+1) transformed data.
— Degrees of freedom (df) and probabilities in parentheses are associated with Box's
(Species) and Greenhouse-Giesser (Time and Time x Species) conservative F-tests.
— tb-values are for Bonferroni contrasts; df values may vary depending on variance
structure within and across time.
Critical value= +tb a / 2 , 2 , df; * significant at P=
0.05, ** significant at P= 0.01.
—

183

72.0+1.2 vs. Ouercus= 7 3.0+0.8.
differences

were

not

n = 1 3 ) . In all cases,

significantly

(paired t - t e s t : Huron Populus
Populus
vs.

vs.

A c e r . P=

0.628;

O u e r c u s . P> 0.10).

detect

a

+2

year

vs.

different

the mean

from

O u e r c u s . P= 0.289;

Wilcoxon's

test:

UMBS

For all three comparisons,

mean

age

difference

(at

zero
Huron

Populus
power to

alpha=0.05)

was

0.80 or greater.

The age difference between Populus and the

small

dominant

number

of

(Populus= 7 3 .3+4.5

similarities,
between

that

at

with
plot

juvenile

between s p e c i e s .
were not

a

at

UMBS

was

A c e r = 7 1.0+4.5,

combined

species

indicates

vs.

Acer

the
age

height

also

n=3).

The

similarity

of

10

growth

of

years
rates

minimal
age

heights

(Fig.
were

5.3),
similar

In other w o r d s , dominant Ouercus and Acer

initially shorter and then

simply caught up

later

in stand development.

Individual height growth trajectories
Each

sampled

assigned

to

patterns

of

in

an

one

Populus.
of

individual

individuals

category

period

greatly

included
increased

stems

height

included

that

growth.

categories

competitive

suppressed
of

four

Ouercus.

growth

stems
growth.

had

least

The

Acer

that
and

constant

that
The
one

ramet

described
reflected

environment

reduced
at

and

or
had

their

changes

ability.
at

period

The

least

released

category

was

one

category

of

included

greatly
stems

showing no periods of major suppression or release in height
growth

(although actual growth rates were not constant over

184

the entire life of the individual). Finally,
category

included

trajectories
release.
seen

individuals

indicated

at

whose

least

one

the combination

cumulative

major

suppression

Examples of suppressed and released

in

Figure

5.2

(for

example,

shortest Ouercus and Fig.

5.2b,

Fig.

stems

5.2a,

Plot 12,

height
and

can be

Plot

9,

the

the third-shortest

Acer).
None of the Populus
any

evidence

for

a

ramets

in the Huron

suppression

Three of the 16 Populus ramets

(18%)

show evidence for a suppression
SD)

growth

began

at

rate
stem

reduction
ages

of

for

15,

or

Populus

had

been

release

(Table 5.3).
these

16

stems

and

47

growing

same

stands

(for

example,

suppressions

were

therefore

5.5).

forest

was

73

years.

(+ 1

(9) %

The

and

growth

relative to suppressed

to

showed

fast

non-suppressed
see

Fig.

more

of

evidence

Populus

5. 2b,
an

growth rates than an actual inhibition.
UMBS

forest did

T h e m ean

exceptionally

compared

the

5.5).

(see b e l o w ) , in that the suppressed

suppression,

in

showed

(Table

in the UMBS

patterns of these stems were unique,
Ouercus and Acer at UMBS

forest

prior

to

from

the

6).

The

Plot

equalization

Three

for

a

(18%)

of

Populus

release

(Table

The mean growth rate increase for these stems was 177

(108)% and occurred between stem ages of 20 and 40 y e a r s .
The
locations
release

majority
showed
(Table

O u e r c u s . had

of
no

5.5).

growth

dominant
evidence
The
rate

Ouercus
for

few

and

either

that

reductions

a

were
of

A cer

at

both

suppression

or

suppressed,

all

approximately

60%

Table 5.5.
Height growth patterns of Populus q r a ndidentata. Ouercus rubra and
Acer rubrum in the a) Huron and b) UMBS forests.

Height growth pattern—
Constant

a) Huron

P. qrandidentata

Suppressed

n—

20

Released

Combination

Proportion—

1.00

---

---

0.04
0.06

0.04
0.06

--0.10

--0.06

185

---

2* rubra
Dominants—
Subdominants—

28
47

0.71
0.28

0.21
0.60

A. rubrum
Dominants
Subdominants

12
80

1.00
0.49

--0.35

Table 5.5 (cont'd).

b) UMBS

P. qrandidentata

16

0.75

0.06

0.06

0.12

19
16

0.95
0.50

0.05
0.44

--0. 06

--———

4
65

0.75
0.61

0.25
0.11

--0.11

0. rubra
Dominants
Subdominants

A . rubrum
Dominants
Subdominants

--0.17

— The constant category includes ramets with no evidence for major suppression
or release; the suppressed category includes ramets with evidence for one
major suppression; the released category includes ramets with evidence for one
major release; the combination category includes ramets with evidence for one
or more major suppressions and releases.
— Total number in catagory.
^Proportion of total (n).
— Dominant Q. rubra and A. rubrum were those that exceeded, equaled (±0.5 m)
or approached (0.5-2.0 m shorter) the height of P. qrandidentata from the
same plot; subdominant Q. rubra and A. rubrum were those more than 2 m
shorter than P. qrandidentata from the same plot.

187

that began at a stem age of 40 years or later. Only three of
the

dominant

increase

individuals

were

in growth rates

and occurred early,

released

after release

(Table

5.5).

The

averaged about

120%

around a stem age of 15 y e a r s .

Between 40 and 72% of the subdominant Ouercus and Acer
showed

some

5.5).

evidence

Most

reductions

of

these

averaged

ages of

30-50 years.

for the

released

200%,

for

large

were

growth

rate

suppressions.

around

70%

and

changes

The

occurred

The mean height growth

s t e m s , most

of which

were

(Table

growth

rate

between

stem

rate

increase

A c e r , was

over

but was highly variable. Most releases occurred around

a stem age of 30 years.
The
Ouercus

growth
and

rate

Acer

potential

from

the

of

suppressed

subdominant

and

height

released

classes

was

assessed by comparing growth rates of these stems to that of
Populus from the same plot,

over the same stem ages prior to

suppression

to

or

subsequent

release.

Growth

rates

considered equivalent

if they were within +0.03 m/yr

heights

<

differing

A pproximately

50%

by
(15

Huron forest and 30%
equivalent

to

(3-4 stems)

and Acer
can

be

prior

(< 16%)

seen

in

initial
Figure

a

of

stem

age

suppressed

of

65

(total
years) .

Ouercus

in

the

at UMBS were growing at rates
to

suppression

of suppressed Acer
at rates

(Table 5.6).

having

at

stems)

were growing

to suppression

m

(2 stems)

Populus

small percentage

2

were

(Table

5.6).

A

in both forests

similar to

Populus

prior

Examples of suppressed Ouercus

growth rates
5.2.

All

of

equivalent
the

to

Populus

relatively

few

188

Table 5.6.
Proportion of suppressed or released
subdominant— Ouercus rubra and Acer rubrum in the a) Huron
and b) UMBS forests that had height growth rates equal— to
or exceeding Populus qrandidentata prior to a suppression
or subsequent to a release.

Species
0. rubra

a) Huron
Suppressed
Released

b) UMBS
Suppressed
Released

n—

31
6

n

7
1

A . rubrum
*

proportion—

0.48
1.00

proportion

0.29
1.00

n

33
15

n

19
14

proportion

0.12
0.40

proportion

0.16
0.43

— Includes all stems more than 2 m shorter than P.
qrandidentata from the same plot.
— Equivalent growth rate= +0.03 m/yr.
— Total number in the catagory (note: total may exceed sums
of the suppressed+combination or released+combination
catagories from Table 5.5 if a stem in the combination
catagory had multiple suppressions or releases.
— Proportion of Q. rubra and A. rubrum in the catagory with
growth rates equivalent to or exceeding P. qrandidenta over
the same stem ages prior to a suppression or subsequent to
a release.

189

released Ouercus in both forests
UMBS,

respectively)

equivalent
released
rates

Populus

(Table

Acer

in both

forests

suggest that

to

Populus

competitive

subdominant

Ouercus

development

growth

height

to

equivalent

stand

had

rates

prior

after release,

and

and,
to

(6 and 1 stems at Huron and
growth
5.6).
(6

stems)

or

changed

some

had

release

growth
results

abilities

of many

over

or

of

These

the

instances,

suppression,

40%

height

release.

environments

in

after

Approximately

after

Acer

rates

course

initial

later

of

height

growth

rate

equaled height growth rates of P o p u l u s .

DISCUSSION
Population height structures
In

both

populations

of

of

Populus

successional
populations

the

species
of

species,

species,

least after the

differences

on

rubra,

of Acer

development.

rubrum,

height

differences

having

structures.

These

a

which

first

in turn

little

or

no

results

contrast

later

the

importance

of

mean

than

relatively
than

successional

decades

of

stand

from

species-specific

with

populations

influence
w ith

earlier study of Populus forest development
demonstrated

early

were taller

several

rates,

study,

taller

mid-tolerant,

resulted

growth

this

intolerant,

average,

a tolerant,

Stratification
in

in

q r a n d i d e n t a t a . an

populations
at

examined

were,

Ouercus

later successional

forests

age

on

those

age

height
from

an

(Chapter 2) that

differences

species to the development of height stratification.

among

190

Al t hough
mortality

not

among

species

stratification.
immediately
>24,000

quantified

Populus

following

ramets/ha

Scheiner

et

suppression

al

in

also

ramet
major

1988).

dominant-codominant

forests.

In contrast,

is

to

most

ramets

height
be

high

(for

example,

and

burning;

survive

Thus,

ramets

to

know to

clearcutting

Unable

1990),

differential

contributed

disturbance

onset of crown differentiation.
only

study,

density

following

(Laidly

this

die

even

mild

following

the

in the current s t u d y ,

were

found

in

the

mature

many slower growing Ouercus and Acer

were able to survive varying degrees of suppression,

thereby

lowering heights

for these

species

surviving

Populus

and height

growth rates

relative to P o p u l u s .
Within

plots,

the

heights

ramets were often similar.
was not
Acer

an attribute

populations.

of

the

In contrast,

of surviving

Heights

a predictable height

stems within Ouercus

often

ranged

variation was clearly not related to age.
of

individuals

similar.

differing

Obviously,

by

height

up

to

15

growth

individuals

varied

widely

within

populations,

so

so

some

growth

rates,

much

including

Populus.

Individual

that

shorter Ace r

man y

experienced
normally

that

height

develops

juvenile

height
and

growth
as

a

growth
Ouercus

ages

and

the

In fact, the ages
m

in

height

rates

of

similar-aged

Ouercus

rates,

had

height

&

to

indicated

in both

beyond

(Oliver

Acer

equivalent

trajectories
genets

were

and

individuals

suppression,

stem

widely

and

forests

that

which

Larson

1990;

191

Zeide

1991) .

Some

of

these

stems

also

had

juvenile

grow

rates comparable to P o p u l u s .
While

these

differences

in

stratification

results

demonstrate

height

growth

of

surviving

that

rates

Populus

species-specific

typically

led

over Ouercus

to

and A c e r ,

they also imply that lower juvenile height growth rates were
not

inherent

Rather,
Acer

the

may

typically

have

reduced

of

the

reduced

resulted

from

competitive

Inhibitory
as

attributes

latter
growth

ability

by

the

extreme

species

rates

competitive
of

influences were spatial

indicated

two

of

per

se.

Ouercus

and

inhibition

individual

and/or
genets.

and temporally variable,

variation

in

heights

within

Ouercus and Acer populations.
Early
Ouercus

inhibition

and

variation
density

in

Acer

competitive
by

both

w i thin

Ross

&

and

Harper

1985;
Hix

1990).
may

have

consistent

probably

environments.

its

affect

on

feature

because
For

crown

of

of

spatial

example,
exposure,

local
root

in differential plant growth performance

among

1972;

Oliver 1979; Weiner

1989;

a

or even direct physical abrasion of soft terminal

leaders can result

Pacala

not

populations,

variation,

crowding,

was

&

Harper
1984;

Kelty

1986;

Lorimer

1990;

In the current
been

in

stand development.

species

populations

1977;

Oliver

(Cayford
1978;

Wierman

Guldin & Lorimer 1985;
Goldberg
Oliver,

1987;

van

Clatterbuck

uncontested

sites

&

Silander &

der
&

Meijden

Burkhardt

s t u d y , the tallest Ouercus

relatively

1957;

and Acer

throughout

The fact that most dominant Ouercus and

192

Acer were

not

initially

suppressed,

at rates comparable to Populus,
Reduced competitive
of

regeneration,

individuals

Oliver 1978;
and

ability m a y

since

However,

and Ace r

were

genets

larger

seedling-origin,

no

An

since

seed trees,
is

likely

had

species

initiated,

if they

existed,

at

frequency

few

the

and

seed

time

extent

Wilson

1968;

lead

to

height

growth

of

stem

s p r o u t - o r i g i n . Although
morphologies

remnant

suggesting

individuals
acted as

were

found

seed sources

No cutting has occurred in any of the

they

that

aged

intermix of seedling-

in any of the stands that might have

stands

similar

in the current study, most Ouercus

obvious

(also true for A c e r ) .

sprout-origin

1943;

could

apparently

Ouercus

grew

function of mode

out-grow

(Jensen

individuals

rate variation.

be a

species,

Beck & Hooper 1986).

sprout-origin

some

within

individuals

rather always

suggests this was the case.

characteristically

seedling-origin

but

remnant

Ouercus

and

Acer

should have b een apparent.

bearing

the
of

so

trees

stands

existed

for

initiated,

post-logging

(Kilburn 1957,

1960; Pyne 1982, pp.

potential

for

long-distance

ruled out

(Darley-Hill & Johnson 1981).

fires

199-218).

dispersal

of

either

given

in

the

the

region

Of course,

acorns

It

cannot

the
be

Even if all Ouercus and Acer were of s p r o u t - o r i g i n , the
age and size of the originating stumps could have influenced
early ramet growth rates and therefore competitive ability.
For example,
with

stump

in O u e r c u s . sprout height growth rate increases
diameter

(Sander

1971)

and

size

of

advanced

193

regeneration

(Sander

1972).

In

A c e r , younger

stumps

higher sprout height growth rates than older stumps
&

Blum

1967).

Additionally,

individual growth
parent
1975;
Acer

stump

vigor

Stroempl
genets

developed

rates
or

and this
size

1983).

that

from

ramet

In

always

grew

&

current

slower

physiologically

(Solomon

may

affect

in turn may be related to

(Solomon
the

number

have

Blum

1967;

Johnson

s t u d y , Ouercus

than

Populus

inferior

root

may

and
have

systems

and

were therefore at an initial competitive d i s a d v a n t a g e .
The shorter mea n height of P o p u l u s . Ouercus and Acer at
U M B S , relative to the Huron forest at comparable stand ages,
does

undoubtedly

site

quality

H o w e v e r , the
to

the

Huron

on

reflect
height

influence

growth

low number
forest,

the

rate

of

(Hix

differences
Lorimer

&

of dominant A cer at UMBS,

may

not.

The

Huron

plots

most similar to those at U M B S , in terms

of total

basal area and density,

index

d a t a ) , contained
the

apparently

and Populus site

in

1990).

relative
that

were

overstory

(unpublished

fast - g r o w i n g , dominant A c e r , while some of

more

productive

Huron

plots

did

not.

These

results suggests that inherently slower growth rates of Acer
populations

on

the

poorer

UMBS

sites

were

not

solely

responsible for the low number of dominants in this forest.

A model for the development of dominance
Several
identified
important

analyses

of

differential
factor

leading

secondary
growth
to

forest

rates

initial

among
height

succession
species

have
as

an

stratification

194

and early vertical dominance by a particular species

(Drury

& Nesbitt

Tilman

1985,

1973;

1988,

1990).

mixed-species
juvenile

&

Likens

These

growth

stand

development

successional

based

rates,

hierarchies.

1979;

treatments

stratification

height

competitive
for

Bormann

Hibbs
present

on

a

inherently

which,

in

The

implication

is

that

species must possess

1983;

a

model

different

turn,
of

of

initiate

such

a

model

successful

early

life-history traits which

allow it to numerically dominate physical space, but it must
also have

inherently

fast juvenile growth if it is to gain

an early height advantage over less-abundant species.
model

appears

found

in

Ouercus

the

which

sufficient

study,

grow

the

developmental

namely,

in height

growth.

understory

numerical

m echanism

stratification,

not have

to

with

the

at

patterns

potential

rates

for

equivalent

to

These results suggest an alternative developmental

in

height

current

and Acer

Populus.
model

inconsistent

Such a

in

itself

through the
Under

tolerance

inherently

such

and

dominance
to

inhibition
a

Populus

promote
of Ouercus

m o d e l , species

relative

different

by

successional

juvenile

height

is

a

vertical
and

Acer

differing
status
growth

in

need
rates

for stratification to develop.
This model
In

the

forests

is illustrated stylistically in Figure 5.4.
examined,

Populus

was

able

to

gain

initial

numerical dominance because of the its propensity to rapidly
saturate

soil

space

with

numerous

vegetative

propagules.

Clonal roots of Populus are able to extend greater than 30 m

195

Figure 5.4.
A model
for the development of vertical
stratification
among
tree
species
differing
in
understory tolerance and relative successional status
in an even-aged forest,
(a) Immediately following a
stand initiating disturbance, initial heights of an
intolerant, early-successional species (clear crown)
and a relatively more-tolerant,
later successional
species (filled crown) may be similar. The intolerant
species is numerically dominant because of superior
dispersal
ability.
(b)
Wit h the onset of
crown
differentiation,
some
individuals
of
both
species
lapse
into
suppression.
A
limited
number
of
individuals of the tolerant species may have early
height growth rates similar to the intolerant species,
probably
because
of
locally
reduced
competitive
pressure from the intolerant species. The initially
low number of stems of the tolerant species reduces
the probability that many will occur in competitively
favorable neighborhoods, (c) In the mature forest, the
intolerant species is typically vertically dominant
over the more-tolerant species although some stems of
the tolerant
species have maintained growth
rates
similar to that
of the
intolerant
species.
The
population-level height growth rate of the intolerant
species will be greater than that of the tolerant
species because all suppressed stems of the former
have died,
leaving only
fast-growing stems,
while
slower growing stems of the tolerant species have
survived.

196

197

from

an

originating

disturbance,
saturate

(Buell

& Buell

1959).

resulting

1956;

(up to
in

Zahner

&

1 ha

high

in Michigan)

density

Crawford

1965;

with sucker-

stands

(Stoeckeler

Barnes

1969;

Harrison & Westell 1963; Scheiner et a l . 1988).
Ouercus

and

(Fowells

Acer

1965),

resprout

of

originating

genets.

the

Ouercus

these

occupied

nor

Ace r

the

species
Many

of

presettlement

originally

from

restricting

propagules

from

Following

the root system from a single clone can rapidly

large areas

sprouts,
Macon

genet

stumps

or

to

the

these

of

may

conifer-dominated

these

sites

were

abundant

collars

vegetative

locations

genets

of

be

the

remnant

forests

(Chapter
in

Graham,

In contrast,
root

distribution

&

that

3).

Neither

these

forests.

C o n s e q u e n t l y , the number of vegetatively derived individuals
in the current

forests may have been a function of Ouercus

and Ac e r presettlement abundance.
The
that

large

this

starting

species

capital

would

of

Populus

continue

to

ramets

insured

dominate

most

competitive neighborhoods in the mature forest,

despite high

dens ity-dependent

development.

Variation

in

mortality

inherent

height

early

in

growth

forest

rates

among

surviving

ramets would have been minimal since most individuals within
a given area

(for example the plot sizes used)

the same clone.

were part of

The interspecific competitive pressure faced

by Ouercus and Acer genets within the Populus clonal matrix
would

have

because

of

b een

intense,

stochastic

but

spatially

variation

in

variable,

Populus

ramet

probably
or

root

198

density

within

individual

competitive

Differences in the physiological
m ay

also

have

influenced

state of the parent genets

height

growth

individual Ouercus and A c e r .

In any case,

remnant

genets

Ouercus

disturbance

and

Acer

forests,

and

the

neighborhoods.

in

the

early,

of

post­

distribution

of

insured that few stems

of these species would be vertically dominant
forests.

of

the low number

restricted

ramets originating from these genets,

potential

in the mature

Under this m o d e l , vertical stratification develops

because

of

species-specific

differences

in

the

ability to

achieve maximal potential height growth rates in the face of
competitive pressure,

not because of inherent differences in

juvenile growth rates that initiate competitive hierarchies.
This model of forest development may be characteristic
only

of

forests

regenerating

vegetative m e a n s .
growth

rates

species

1970)

sprout - v i g o r .
neighborhoods
rates

of

In such forests,

characteristic

(Loach

predominantly

may

of

inherently lower height

seedlings

sometimes

be

of

then

less-tolerant

be

able

species.

to

more-tolerant

compensated

Sprout-origin stems occurring
would

through

match

However,

as

for

by

in uncontested
height

growth

shown

in the

current s t u d y , most individuals of tolerant species, despite
being

of

degree,

sprout - o r i g i n ,
which

in

turn

experience

suppression

leads

species-level

to

to

some
height

stratification.
A
current

remarkable
study

is

feature
that

the

of

the

forests

apparent

examined

inhibitory

in

ability

the
of

199

Populus has persisted for over 70 years.

Not only were the

majority of Ouercus and Ace r in both study areas suppressed
to varying
for

any

degrees but

species

(see

canopy positions.
and

can

clearcuts.
for

a

be,

no

individuals

Chapter

4)

have

recruited

cohorts

into

examined

experimentally

of

dominance

in

in

young

Populus

this model would account
a

forest

type

(Populus

qrandidentata and Populus tremuloides Michx.

combined)

covers

northern

nearly

upper

Of course the proposed model needs to be,

Should it prove valid,

pattern

from you n g e r

five

million

hectares

Lakes region of the United States
1986; Spencer et aJL. 1988).

in

the

(Jakes 1980;

that
Great

Smith & Hahn

BIBLIOGRAPHY
Barnes, B. V.
(1966).
The clonal growth habit of American
aspens.
E c o l o g y . 47, 439-447.
________ .
(1969).
Natural variation and delineation of
clones of Populus tremuloides and P. qrandidentata in
northern Lower Michigan, Silvae Genetica, 18, 130-142.
Barnes, B. V., Pregitzer, K. S., Spies, T. A. & Spooner, V.
H.
(1982).
Ecological forest site classification.
Journal of F o r e s t r y . 80, 493-498.
Bazzaz, F. A.
(1979).
The physiological ecology of plant
succession.
Annual Review of Ecology and S v s t e m a t i c s .
10, 351-371.
Beck,

D. E.
& Hooper,
R. M.
(1986).
Development of a
southern Appalachian hardwood stand after c l e a r c u t t i n g.
Southern Journal of Applied F o r e s t r y . 10, 168-172.

Bormann, F. H.
white pine
277.

(1965).
Changes in the growth pattern of
undergoing suppression.
E c o l o g y . 46, 269-

Bormann, F. H. & Likens, G. E.
(1979).
Pattern and process
in a forested e c o s y s t e m . Springer-Verlag, New York.
B u e l l , M. F.
prairie.
264-265.

& B u e l l , H. F.
Bulletin of the

(1959) . Aspen invasion of
Torrev Botanical C l u b . 86,

Canham, C. D. & Marks, P. L.
(1985).
The response of woody
plants to disturbance: patterns of establishment and
growth.
The ecology of natural disturbance and patch
dynamics (Ed. by S. T. A. Pickett & P. S. White) , pp.
197-216.
Academic Press, Orlando.
Cayford, J. H.
(1957).
Influence of the aspen overstory on
white spruce growth in S a s k a t c h e w a n . Canada Department
of Northern Affairs and National Resources Forestry
Branch, Forest Research Division Technical N o t e . 58.
Clatterbuck, W. K. & Hodges, J. D.
(1988).
Development of
cherrybark
oak
and
sweetgum
in
mixed,
even-aged
bottomland
stands
in
central
Mississippi,
U.S.A.
Canadian Journal of Forest R e s e a r c h . 18, 12-18.

200

201

Cooper, A. W.
(1981). A b o v e -ground biomass accumulation and
net primary production during the first 70 years of
succession in Populus qrandidentata stands on poor
sites in northern lower Michigan. Forest s u c c e s s i o n ;
concepts and application (Ed. by D. C. West, H. H.
S h u g a r t , & D. B. B o t k i n ) , pp. 339-360. Springer-Verlag,
New York.
Da r l e y - H i l l , S. & Johnson, W. C.
(1981). Acorn dispersal by
the blue jay (Cvanocitta c r i s t a t a ) . Oecologia, 50, 231237.
D e a l , R.
L . , Oliver,
C.
D.
Reconstruction
of
mixed
coastal southeast Alaska.
R e s e a r c h . 21, 643-654.

& B o r m a n n , B.
T.
(1991).
hemlock-spruce
stands
in
Canadian Journal of Forest

D r u r y , W. H . & N i s b e t , I . C. T.
(1973) . Succession.
Journal of the Arnold A r b o r e t u m . 54, 331-368.

The

Foster,
D.
R.
(1988).
Disturbance
history,
community
organization and vegetation dynamics of the old-growth
Pisgah
Forest,
south-western
New Hampshire,
U.S.A.
Journal of E c o l o g y . 76, 105-134.
Fowells, H. A.
(1965).
Silvics of forest trees of the
United States.
U . S .D.A. Agricultural H a n d b o o k . 271.
G i l l , J. L.
(1978).
Design and analysis of experiments in
the animal and medical s c i e n c e s .
Volumes 1 & 2. Iowa
State University Press, Ames.
Goldberg, D. E.
(1987). Neighborhood competition in an Old­
field plant community.
E c o l o g y . 68, 1211-1223.
Graham, S. A . , Harrison, R. P., Jr. & W e s t e l l , C. E . , Jr.
(1963).
Aspens: phoenix trees of the Great Lakes
r e g i o n . University of Michigan Press, Ann Arbor.
Guldin, J. M. & Lorimer, C. G. (1985). Crown differentiation
in even-aged northern hardwood forests of the Great
Lakes region, U.S.A.
Forest Ecology and M a n a g e m e n t .
10, 65-86.
Gurevitch, J. & Chester, S. T . , Jr.
(1986) . Analysis of
repeated measures e x p e r i m e n t s . E c o l o g y . 67, 251-255.
Harper,
J.
L.
(1977).
Population
Academic Press, L o n d o n .

biology

of

plants.

Hibbs, D. E.
(1983).
Forty years of forest succession
central New England.
E c o l o g y . 64, 1394-1401.

in

202

Hix,

D. M. & Lorimer,
C. G.
(1990).
Growth-competition
relationships
in
young
h a r dwood
stands
on
two
contrasting
sites
in
southern
Wisconsin.
Forest
S c i e n c e . 36, 1032-1049.

Jakes, J. L.
(1980).
Minnesota forest statistics, 1977.
U . S .D.A. Forest Service Resource B u l l e t i n . N C - 5 3 .
Jensen, V. S.
(1943).
Suggestions for the management of
northern hardwood stands in the northeast.
Journal of
F o r e s t r y . 41, 180-185.
Johnson, P. S.
(1975).
Growth and structural development
of red oak sprout clumps.
Forest S c i e n c e . 21, 413-418.
K e l t y , M. J.
(1986).
Development patterns in two hemlockhardwood
stands
in southern
New
England.
Canadian
Journal of Forest R e s e a r c h . 16, 885-891.
Kilburn, P. D.
(1957). Historical development and structure
of the a s p e n , iack p i n e , and oak vegetation types on
sandy
soils
in
northern
Lower
Michigan.
Ph. D.
dissertation, University of Michigan, Ann Arbor.
________ .
(1960).
Effect of settlement on the vegetation of
the University of Michigan Biological Station. Papers
of the Michigan Academy of Science, Arts and L e t t e r s .
45, 77-88.
Kozlowski, T. T. , Kramer, P. J. & P a l l a r d y , S. G. (1991) .
The Physiological Ecology of Woody P l a n t s . Academic
Press, San Diego.
Laidly, P. R. (1990). Populus grandidentata Michx. (Bigtooth
a s p e n ) , p. 544-550.
Silvics of North A m e r i c a . Vol 2.
(Ed. by R. M. Burns & B. H. Honkala) , pp. 544-550.
U . S . D . A . Forest Service Agricultural H a n d b o o k . 654.
Lapin, M.
(1990).
The landscape ecosystem groups of the
University
of
Michigan
Biological
Station:
c l a s s i f i c a t i o n . m a p p i n g . and analysis of ecological
diversity.
M. S. thesis, University of Michigan, Ann
Arbor.
Larson,
P.
R.
(1956).
Discontinuous
growth
suppressed slash pine.
Tropical W o o d s . 104,

rings
80-99.

in

Larson, B. C.
(1986).
Development and growth of even-aged
stands of Douglas-fir and grand fir.
Canadian Journal
of Forest R e s e a r c h . 16, 367-372.
Loach, K.
(1970).
Shade tolerance in tree seedlings I I .
Growth
analysis
of plants
raised
under
artificial
shade.
New P h y t o l o g i s t . 69, 273-286.

203

Lorimer, C. G.
(1984). Methodological considerations in the
analysis
of
forest
disturbance
history.
Canadian
Journal of Forest R e s e a r c h . 15, 200-213.
Marks, P. L.
(1975).
On the relation between extension
g r owth and successional status of deciduous trees of
the northeastern United States.
Bulletin of the Torrev
Botanical C l u b . 102, 172-177.
Marquis, D. A.
(1967).
Clearcutting in northern h a r d w o o d s :
results
after
30
years.
U.S.D.A.
Forest
Service
Research P a p e r . N E - 8 5 .
________ .
(1981) .
Removal or retention of unmerchantable
saplings in Allegheny h a r d w o o d s : effect on regeneration
after c l e a r c u t t i n g . Journal of F o r e s t r y . 79, 280-283.
Newton, M . , El Hassan, B. A. & Z a v i t k o v s k i , J . (1968).
Role
of red alder in western Oregon
forest succession.
Biology of alder (Ed. by J. M. T r a p p e , J. F. Franklin,
R. F. Tarrant, & G. H. Hassen) , pp. 73-84. U.S.D.A.
Forest
Service
Pacific Northwest
Forest
and Range
Experiment Station, Pullman, Washington, U.S.A.
O 'H a r a , K. L.
(1986) .
Developmental patterns of residual
oaks
and
oak
and yellow-poplar
regeneration
after
release in upland hardwood s t a n d s . Southern Journal of
Applied F o r e s t r y . 10, 244-248.
Oliver, C. D.
(1978).
The development of northern red oak
in
mixed
stands
in
central
New
England.
Yale
U n i v ersity School of Forestry and Environmental Studies
B u l l e t i n . 91.
Yale University, New Haven, Connecticut.
________ .
(1980) .
Even-aged development of mixed-species
stands.
Journal of F o r e s t r y . 78, 201-203.
Oliver, C. D. & Larson, B. C.
M c G r a w H i l l , New York.

(1990).

Forest stand d y n a m i c s .

Oliver,
C.
D . , Clatterbuck,
W.
K.
& Burkhardt,
E.
C.
(1990).
Spacing
and
stratification
patterns
of
cherrybark oak and American sycamore in mixed, evenaged stands in the southeastern United States. Forest
Ecology and M a n a g e m e n t . 31, 67-79.
P a d l e y , E. A.
(1989). Associations among glacial l a n d f o r ms.
s o i l s , and vegetation in northeastern Lower M i c h i g a n .
Ph.D.
dissertation,
Michigan State University,
East
Lansing.

204

Pregitzer. K. S. & Barnes, B. V.
(1984). Classification and
comparison of upland hardwood and conifer ecosystems of
the
Cyrus H.
M cCormick Experimental
Forest,
upper
Michigan.
Canadian Journal of Forest R e s e a r c h .
14,
362-375.
Pyne,

S. J.
(1982).
Fire in A m e r i c a .
Press, Princeton, New Jersey.

Princeton University

R o b e r t s , M. R. & Richardson, C. J.
(1985).
Forty-one years
of population change and community succession in aspen
forests on four soil types in northern Lower Michigan,
U.S.A.
Canadian Journal of B o t a n y . 63, 1641-1651.
Ross, M. A. & Harper, J. L. (1972) . Occupation of biological
space
during
seedling
establishment.
Journal
of
E c o l o g y . 68, 77-88.
Sander, I . L.
(1971).
Height growth of new oak sprouts
depends on size of advance r e p r o d u c t i o n . Journal
of
F o r e s t r y . 69, 809-811.
________ .
(1972) .
Size of oak r e p r o d u c t i o n : key to growth
following
harvest
cutting.
U . S . D . A . Forest
Service
Research P a p e r . N C - 7 9 .
________ .
(1990).
Ouercus rubra L. (Northern red o a k ) .
Silvics of North A m e r i c a . Vol. II. (Tech. Coordinators
R. M. Burns & B. H. H o n k a l a ) , pp. 727-733.
USDA Forest
Service Agricultural H a n d b o o k . 654.
S c h e i n e r , S. M. , Sharik, T. L. , Roberts, M. R. , & Vande
Kopple, R.
(1988).
Tree density and modes of tree
recruitment in a Michigan pine-hardwood forest after
c lear-cutting
and
burning.
The
Canadian
FieldN a t u r a l i s t . 102, 634-638.
S h a i n s k y , L. J. & Radosevich, S. R.
(1992).
Mechanisms of
competition
between
Douglas-fir
and
red
alder
seedlings.
E c o l o g y . 73, 30-45.
S i l a n d e r , J.
A.
& Pacala,
S.
W.
predictors of plant performance.
263.
Smith, D. M.
(1986).
The
Wiley & Sons, New York.

practice

(1985).
Neighborhood
O e c o l o g i a . 66, 256-

of

s i l v i c u l t u r e . John

Smith,
W.
B.
St Hahn,
J.
T.
(1986) . Michigan's
statistics, 1987: an inventory u p d a t e . U.S.D.A.
Service Technical R e p o r t . N C - 1 1 2 .

forest
Forest

Soil Conservation Service.
(1991). Soil survey of Cheboygan
County, Michigan.
U.S.D.A. Soil Conservation S e r v i c e .

205

Solomon. D. S. & Blum, B. M.
four
northern
hardwoods.
Research P a p e r . NE-59.

(1967) . Stump sprouting of
U.S.D.A.
Forest
Service

Spencer, J. S. Jr., Smith, W. B . , Hahn, J. T. & Raile, G. K.
(1988).
Wisconsin's fourth forest inventory,
1983.
U.S.D.A. Forest Service Resource B u l l e t i n . N C - 1 0 7 .
Spurr, S. H. & Barnes, B. V. (1980).
Forest e c o l o g y .
edition.
J ohn Wiley and Sons, New York.

Third

S t o e c k e l e r , J. H. & Macon, J. (1956).
Regeneration of aspen
cutover
areas
in
northern
Wisconsin.
Journal
of
F o r e s t r y . 54, 13-16.
S t r o e m p l , G.
(1983).
Thinning clumps of northern hardwood
stump sprouts to produce high quality timber. Ontario
Ministry
of
Natural
Resourses.
Forest
Research
Information P a p e r . 104.
Stubblefield, G. W. & Oliver, C. D.
(1978). Silvicultural
implications
of
the
reconstruction
of
mixed
alder/conifer stands.
Utilization and management of
red a l d e r , pp. 307-320. U . S . D . A . Forest Service General
Technical R e p o r t . P N W - 7 0 .
Tilman,
D.
(1985).
The
resource
ratio
hypothesis
succession.
American N a t u r a l i s t . 125, 827-852.
________.
(1988).
Plant strategies
structure of plant c o m m u n i t i e s .
Press, Princeton, New Jersey.

of

and the dynamics and
Princeton University

________.
(1990).
Constraints
and
tradeoffs:
toward
a
predictive
theory
of
competition
and
succession.
O i k o s . 58, 3-15.
van der Meijden, E.
(1989).
Mechanisms in plant population
control.
Toward a more exact ecology (Ed. by P. J.
Grubb
& J.
B.
W h i t t a k e r ) , pp.
163-181.
Blackwell
Scientific Publications, Oxford.
Weiner, J.
(1984).
Neighborhood interference amongst Pinus
rigida i n d i v i d u a l s . Journal of E c o l o g y . 72, 183-195.
Wierman, C. A. & Oliver, C. D.
(1979). Crown stratification
by species in even-aged mixed stands of Douglas-fir western hemlock.
Canadian Journal of Forest R e s e a r c h .
9, 1-9.
Wilson, B. F.
(1968).
Red maple stump s p r o u t s : development
the first year.
Harvard Forest P a p e r . 18.

206

Zahner, R. & Crawford, N. A.
(1965).
The clonal concept in
aspen site relations.
Forest-soil relationships (Ed.
by
C.
T.
Youngberg) , pp.
230-243.
Oregon
State
University Press, Corvallis.
Zeide, B.
(1991).
Self-thinning and stand density.
S c i e n c e . 37, 517-523.

Forest

Chapter 6
THE AGE A N D HEIGHT STRUCTURE OF RED MAPLE (Acer rubrum L . )
POPULATIONS IN NORTHERN MICHIGAN BIGTOOTH ASPEN (Populus
qrandidenta Michx.) FORESTS

ABSTRACT

Red maple
later

(Acer

rubrum L.)

successional

tree

is often the most

species

abundant

recruiting

in

the

understories of aspen and oak-dominated forests on dry-mesic
sites
that

in

eastern North

this

canopy

species

positions

increasing

on

of

(Populus

importance

study

red

height

Given
of

examined

maple

populations

growth

the

red

USA.

Stem

suggests

to

dominant

potential

maple

in

the

age

for

these

in

a

and

height
aspen

landscape

analysis

overstory

are

bigtooth

was

and

in

used

between establishment times,

rates

for

examinations

M i c h x . )-dominated

Michigan,

examine relationships
and

recruiting

sites.

grandidentata
Lower

of

population-level

This

structures

Limited evidence

capable

these

detailed

warranted.

northern

is

overstory

forests,

America.

to

heights,

understory

red

maple from 20 replicate plots in five stands located within
a 18 k m 2 area.
Red

maple

was

forests of the study

a

minor

area.

overstory

component

The understories

were overwhelmingly dominated by red maple.

of all

contained

established

mostly

concurrently

sprout-origin
with

bigtooth

207

the

stands

The populations

were composed of two clearly defined age c o h o r t s .
cohort

in

The first

individuals
aspen

within

that
a

10

208

year

period,

sampling.
was

beginning

Mean

not

age

70

of

significantly

years

the

prior

to

sprout-origin

different

among

the

red

stands,

time

maple

of

cohort

nor

did

it

differ from the mean age of bigtooth aspen.
Mean height
not

of the

significantly

stand,

different

sprout-origin cohort was

among

stands.

height growth rates of these

variable.
stem

red maple

Within

each

individuals were highly

The variability was not related to differences in

age.

Recent

height

growth

increment

of

the

sprout-

origin stems was weakly related to position of an individual
in

the

overstory,

suggesting

competitively

suppressed

by

sprout-origin

red

in

maple

dominant bigtooth aspen

that

most

taller
all

red

bigtooth

stands

in height growth

did

maple

were

aspen.

Some

approach

rate.

the

These were

likely stems that were never competitively suppressed.
The

second

red maple

cohort

c ontained

seedling-origin

individuals that began establishing 30-35 years after stand
initiation,
increment
maple.

immediately
in

dominant

after

culmination

overstory

bigtooth

of

aspen

height
and

red

This suggests that increasing resource availability,

as a result of declining overstory vigor and canopy closure,
may be a factor triggering understory reinitiation in these
even-aged
red

maple

forests.
were

In g e n e r a l , heights

more

dependent

sprout-origin indiv i d u a l s .
similar
still

aged

highly

individuals
variable.

on

However,
within

The

stem

seedling-origin

age,

compared

to

height growth rates for

the

tallest

of

seedling

cohort

individuals

were

generally

209

had the greatest rates of recent height
were

at

a

competitive

environment,
There

but

was,

growth

in

rate

origin

these

fact,

over

a

in

were

not

trend

of

time

individuals,

availability

advantage

the

for

the

again

increment,

within

always

the

increasing
fastest

was

understory

oldest

stems.

initial

height

growing

seedling-

that

resource

suggesting

understory

the

and thus

increasing

over

the

dominance

and

course of stand development.
Red

maple's

overwhelming

understory

ability to reach the dominant canopy positions in the stands
examined

suggests

importance
including
and

the

on

a

potential

dry-mesic

shade
ability

sites.

tolerance,
to

for
Life

vigorous

respond

increasing

with

history

overstory
attributes,

resprouting
increased

potential

growth

upon

release, may foster the development and maintenance of a red
maple-dominated cover type in the Great Lakes region.

210
INTRODUCTION

Post-settlement
led

to

significant

changes

natural

alterations

structure

of

many

America.

On

dry-mesic

region,

in

forested

in

the

landscapes

sites

in

the

fire

(Whitney

1987).

led to

the development

spp.)

and

aspen

tremuloides

of

eastern

northern

strobus

L. , P.

dominant

in

Heinselman

resinosa

the

1973;

forests
and

presettlement
Whitney

1987;

North

Great

Lakes

circumstances

even-aged oak

grandidentata

Ait.

and

favored species

historical

of extensive

M i c h x . )-dominated

have

soil and resprout readily

These

(Populus

regimes

composition

deforestation and repeated wildfire

able to establish on bare mineral
after

fire

Michx.

where

P.

and

pines

banksiana

landscape
Crow

(Ouercus

(Pinus

Lam.)

(Kilburn

1988;

P.

Nowacki

were
1957;

et

al.

1990; Chapter 3).
In many of these second-growth forests,

red maple

(Acer

rubrum L . ) is often one the most abundant later successional
tree species recruiting

in the understory

1952;

Cramer

Benninghoff

and

1963;

(Brown and Curtis

Graham

et

al.

1963;

Roberts and Richardson 1985; Host et a l . 1987; Sharik et a l .
1989;

Heinen

abundance

and

of

Sharik

red

1990;

maple

has

Nowacki
also

et
been

a l . 1990).
noted

High

in

the

understories of forests historically dominated by oak in the
southern

Great

Lakes

region

(Gysel

and

Arend

1953;

Larsen

1953; Rudolph and Lemmien 1976; Peet and Loucks 1977; Pastor
et

a l . 1982;

United States

Dodge
(Hibbs

and

Harman

1983;

1985)

and

the

northeastern

Lorimer 1984; Abrams and Nowacki

211

1992).

The development of red maple understories in all of

these forests is apparently in response to
(Lorimer

1984;

1992).

Barring

Dodge

and

major

Harman

1985;

disturbance

fire suppression

Abrams

capable

and

of

Nowacki

redirecting

successional p a t h w a y s , for example extensive cutting,
insect

or

disease

outbreaks,

red

maple

may

increasingly important overstory component

fire,

become

in these

an

forests

(Lorimer 1984).
Given the potential for increasing overstory importance
of red maple in eastern forests on dry-mesic sites,
ecological
those of

examinations

Lorimer

(1984)

of

these

populations,

and Sakai

(1990a,b)

detailed
following

are warranted.

This is particularly important in aspen-dominated landscapes
of the northern Great

Lakes

region,

since

in man y of

these

forests the current aspen overstory is approaching the limit
of

the

taxon's

pathological

Graham et a l . 1963)

rotation

(70-80

years

old;

and thus its successional replacement is

imminent.
This study examined populations of red maple
even-aged

bigtooth

dominated

forests

aspen
in

(Populus

northern

in mature

arandidentata

Lower Michigan,

M i c h x . )-

USA.

Initial

observations indicated that red maple was virtually the only
tree species recruiting in the understory of these
Specific

objectives

the compositional
the

study

structures

the

study

included:

importance of red maple

area;
of

of

the

2)

characteriz ing
red

maple

the

1)

forests.

quantifying

in the forests of
age

populations;

and
3)

height

examining

212

relationships

between

height growth rate
and 4)

time

of

establishment,

height,

and

in overstory and understory populations;

characterizing the degree of variability

in age

and

height structure among red maple populations w i t h i n a local
landscape.

STUDY LOCATION AND ORIGINAL FOREST COMPOSITION
Research
dominated
the

conducted

landscape

Huron-Manistee

Michigan,
15'

was

USA

on

the

within

a

Harrisville

National

Forest

bigtooth

Ranger

in

W) .

were pine-dominated

The presettlement
(Chapter 3).

District

northeastern

(latitude 44° 15' to 45° 00' N,

to 84° 45'

aspenof

Lower

longitude 83°

forests of this area

Red pine,

jack pine,

and

white pine accounted for 77.5% of all b earing and line trees
in

the

original

Aspen,

red

beech

(Fagus

General

maple,

Land

Office

oak

(Ouercus

white

alba

grandifolia E h r h . ) and red

L . ) were minor overstory components
accounted

survey

for

only

5.4%

of

oak

of

the

L.),

bearing

and

line

American

(Ouercus

(Chapter 3).

area.

rubra

Red maple

trees

in

the

presettlement forests of the study area.
Virtually all of northern Lower Michigan,
study area,

was deforested

twentieth centuries
wildfires

swept

following

logging

(Whitney 1987).

the

coniferous

cut-over

(Pyne

regeneration and most
dominant

in the

1982).

remnant
species

including the

late nineteenth to

early

Frequent slashed-fueled

landscapes
Fires

the

eliminated

overstory
(Kilburn

in

advanced

individuals

1957,

1960;

years

of

the

Whitney

213

1987)
(P.

and

promoted

grandidentata

vigorous
and

P.

vegetative-sprouting

tremuloides

spreading clonal root systems

Michx.)

of

from

aspen

rapidly

(Barnes 1966).

METHODS
Stand Selection
Five

stands

classification

unit,

characteristics
Barnes et al.
1 9 8 7 )

from

and

1 9 8 2

based

same

on

ground

ecological

surficial

flora

composition

; Pregitzer and Barnes

restricted

1 9 8 4

(Padley
sandy,

1 9 8 9 )

mixed,

Important
stands

deep
.

outwash

Soils
frigid

herbaceous

included:

anoustifolium
canadense

Ait.;

Lam.;

following additional
from obvious
and 4)

were

Haplorthods

species

(L.)

acerifolium

L. ;

Orvzopsis

asperifolia

( 1 9 6 3 )

.

criteria:

Fern.
The
1)

disturbance

s e r i e s ).
in

L. ;
Michx.

1

ha

initiation;

1 - 4 . 5

km.

and

follows
the

in size;
3)

relative basal area of bigtooth aspen >

The distance bet w e e n stands ranged from

all

Carex

selected met

minimum of
since

as

Maianthemum

Nomenclature

stands

deposits

K u h n ; Vaccinium

procumbens

f .)

till

(Rubicon

Gaultheria

(Michx.

Stands were

classified

understory

aquilinum

Viburnum

Gleason and Cronquist

slopes;

; Host et al.

(Michigan State

overlaying

stands

woody

Pteridium

Amelanchier arborea

free

all

Entic
and

D e s f .,

pensylvanica

of

sands

soil

(following

Forestry Department research f i l e ) .
to

land

geology,

, were selected from a larger data pool

University,

2)

the

0-5%
65%.

214
Vegetation Sampling
Arboreal vegetation was sampled in four 272 m 2 circular
plots placed by random compass bearing and direction in each
stand.

Species

individuals
cm)

and

(dbh

diameter

> 10

were tallied

cm)

(at

and

1.37

m)

saplings

in each plot.

of

(2.5

Species

all

overstory

cm < dbh < 10

and number of tree

seedlings

(dbh < 2.5 cm) were tallied in twelve 1 m 2 frames

spaced

3 m

plot,
maple

at

intervals

along

four

opposing

beginning at the plot boundary.
individuals

(sprout

or

radii

of

each

The origin of all red

seedling)

was

assessed

when

possible.

Stem Sampling
In

each

plot,

all

red

maple

>

1.5

m

in

height

were

destructively sampled to determine year of establishment and
subsequent

height

growth

patterns.

Only

the

ramets were sampled from multi-stemmed genets.
a

subsample

randomly

of

5-15

selected

Additionally,
selected

from

one

All

the

base

stems were

to

dominant

individuals
(usually

<

height

those

2

in

aspen

individuals)

In each plot,
height
and

and

were

examined.

was

randomly

sampled

in

felled at groung level and total

m and

at

1

m intervals

the 1 ra multiple closest to

the
>

m

classes

bigtooth

leader. Stem sections
and

1.5

live

Stems were marked at 0.25 m intervals

Marking continued to
the

m

decaying

height was recorded.
from

0.5

overstory

(rejecting

each plot.

individuals

tallest

basal
3

m

ends
in

were

of

some

height) at

cut
small
each

thereafter.
the end of
from

large

individuals
measurement

215

interval.

Stem Analysis
Small

individuals

(usually those

< 3 m

in height)

and

terminal leaders of larger individuals were aged by counting
the number of terminal bud scale scars preceding each height
interval.

The

individuals was
0.25 m
of

stem

subsample

stems

determinations
+

diameter.
to

a

for

2 years

by

scale

counting

using

a

and

then

only

surface

on

small

scale

portion

scars) on

microscope.

seldom

and

a

Age

differed by more

individuals

>

2

cm

basal

from larger individuals were sanded

and

wetted

to

short axis on oblique-shaped stems)
dissecting microscope.

on

on the basal

dissecting

aid

Ring number was counted on two radii

differed,

counts

most difficult

bud

the two methods

Stem sections

smooth

bud

(typically the

to age

of

of

checked by counting rings

stem sections

the

than

accuracy

ring

examination.

(typically a long and

of each section under a

The greater of the two ages,

if they

was recorded.

Section ages were used to reconstruct the height growth
history

(time

height)

of

meters)

for

sampling

was

increments
beginning

of establishment

to

reach

each sampled individual. Height increment
the

five

year

determined

over
10

and years

for

consecutive

years

period
each

immediately
sampled

five-year

after stand

initiation,

were

(in

prior

ramet.

stand-age

a given

to

Height

intervals,
determined

for all bigtooth aspen and selected overstory red maple.

In

216

some instances interpolation of stem height at a given stand
age was

required since

stem height but not
of

age-height

examined to

the stem

age.

plots

sampling method

When this was

for

the

insure accurate

controlled

required,

individuals

determination of

the

shape

involved

were

height

at

the

five-year stand-age of interest.

Data Analysis
The number of years to understory reinitiation and red
maple

and

bigtooth

among

stands

aspen

using

ages

one-way

and

heights

analysis

of

were

compared

variance.

All

sets met the assumption of normality of residuals
Wi l k test
some

did

for
not

small
meet

test,

P <

significant

for

(Fm ax

variances

no

sample

the

0 . 2 5 )

any

size,

assumption

P >
of

Since the overall

of

the

conservative

sets where the overall

(Shapiro-

sets

corrections

, but

1 9 7 8 )

homogeneous

.

data

Gill

0 . 1 ;

data

variances

F-tests were not

w ith
were

heterogeneous

used.

F-tests were significant,

For

data

individual

stand means were separated using T u k e y 's t e s t s .
Paired t-tests were used to compare several
of

matched

including

red

maple

ages,

and

bigtooth

heights,

growth

aspen
rates,

on

attributes
each

and

plot,
height

in c r e m e n t s . All data sets met the assumption of normality of
residual differences
Temporal
and

trends

selected

consecutive

(Shapiro-Wilk test,
in height

overstory

five-year

red

increment
maple

stand-age

P >
of

were

intervals

0 . 1 ) .

bigtooth

aspen

examined

over

using

complete

217

block repeated measure analysis of variance, where each stem
was

a

block

and

time

period

was

Height increments were transformed,
assumption
over

of

normality

heterogeneous

transformation
sets,

P<

degrees
This

of

adjustment

time

Because

variance-covariance

matrices

the

highly
was

(p-1)(r-1)
and

conservative

used

reduces

periods

1988).

to

test

degrees

of

periods

performed

of

concern

even

after

r

is

the

(or

Bonferroni

period

freedom

for

effects.

the

F-test

where p is the number

number

of

replicates

(Gill

increment between specific

combinations

using

Greenhouse-Giesser

time

to 1 and r-1,

Contrasts of mean height

time

measure.

as I n ( y + l ) , to the meet

residuals.

freedom

from p-1 and

repeated

(sphericity assumption rejected for all data

0.25)
of

of

the

of

time

t-tests,

heterogeneous variance-covariance structure

periods)

were

modified

for

(Gill 1988).

Correlation analysis was used to examine relationships
between

stem

age,

height

and

height

incr e m e n t . Spearman's

rank correlation was used because most data
meet

the

assumption

of

bivariate

sets

normality

failed

required

to
for

parametric correlation analysis. For some of these analyses,
stems

were

pooled

justified

since

height

height

stems

or

rather

type-one

than

error

significant.

across
the

plots in

each

"mensurational

increment
plots.

were
For all

probability

of

stand. Pooling

treatments"

attributes

of

of

was
age,

individual

statistical analyses,
0.05

was

a

considered

218
RESULTS
Forest Composition
The

overstories

bigtooth aspen

of

(Table

all

6.1).

stands

were

Individual

dominated

ramets differed

by
in

age by no more than three years both within and among stands
(65-68

years).

among stands
aspen

(50

stands

1,

Mean

age

(P> 0.9,

was

Table

not

6.2).

year base)

was

3,

(P> 0.05;

4 and 5

not

significantly
Site

index

different

for bigtooth

significantly different
Table

6.2).

The site

among
index

for stand 2 was significantly less than all other stands

(P<

0.05; Table 6.2).
Later successional
overstories
(Table
some

of

6.1).

stands

all

species of lesser importance

stands

Additional
included

included
minor

black

red

maple

overstory

oak

and

species

(Ouercus

in the
red

found

velutina

oak
in

L a m .),

white oak, paper birch

(Betula papvrifera M a r s h . ), white ash

(Fraxinus

) and

americana

E h r h . ; Table

L .

6.1).

Red

black

maple

cherry

(Prunus

contributed

7-18%

serotina
of

total

overstory basal area in the five stands.
The

majority

were red maple
91-100%

and

of

saplings

(Table 6.3).
62-90%

Additional species,

for

and

seedlings

in

all

stands

Relative densities ranged from
the

two

strata,

respectively.

including red oak, white oak, black oak,

black cherry, white ash,

and paper birch, were of only minor

importance in the understory of some stands

(Table 6.3).

Table 6.1. Overstory composition— and structure of five bigtooth aspen-dominated stands
within the Huron-Manistee National Forest.

Stand:

1

2

Species

34.5
5.3
3.2

Total basal area:
Total density (#/ha):

43.2
1158

(4.7)^
(1.2)
(1.1)

23.0
3.3
7.2
-

-

0.2

(2.6)
(0.7)
(2.5)

(0.2)

0.1

27.7
4.8
5.8
0.7

(m2/ha)

(1.8)
(1.0)
(1.7)
(0.6)

(0.1)

-

(4.6)
(88)

— Values are the means

33.5
1398

(1.0)
(55)

34.1
8.2
2.5

38.9
996

(2.0)
(58)

(+ SE) of four plots.

(3.7)
(1.1)
(1.2)
-

-

-

5

4

Basal Area

Bigtooth aspen
Red maple
Red oak
Black oak
White oak
Paper birch
White ash
Black cherry

— Dbh > 10 cm.

3

0.2
0.2
0.1

(0.2)
(0.3)
(0.1)

45.3
1011

(2.4)
(89)

32.4
3.2
8.9
1.6
1.1

(7.7)
(0.7)
(2.9)
(0.7)
(0.7)
-

47.1
1021

(6.7)
(112)

220

Table 6.2.
Age and site index of bigtooth aspen in five
stands with i n the Huron-Manistee National Forest and
analysis of variance for comparisons among stands.

Age

Stand

1
2
3
4
5

66
66
66
66
66

(years)

Site Index

21.1
17.0
21.2
23.6
22.2

(0.4)—
(0.4)
(0.5)
(0.6)
(0.6)

(m at 50 years)

(0.3)a
(0.5)b
(0 .6) a
(0.9)a
(1.0)a

^Values are the means (+ SE) of four observations,
Values followed by the same letter were not significantly
different at P=.05.

Age
Source

Stands
Error

df—

4
15

SSe

0.700
16.500

^df= degrees of freedom,

Site Index
P > F

0.96

SS

97.823
28.795

SS= sum of squares.

P > F

0.0001

Table 6.3.
Sapling and seedling densities by species in five bigtooth aspen-dominated
stands within the Huron-Manistee National Forest.

Sapling-

Stand
2

1

3

4

5

Stems/ha

Species
496

(149)—
-

377 (118)
28 (18)
-

“

Sum:

496

(149)

405

(106)

1241
19
46
19

(149)
(19)
(28)
(19)

368

1325

(173)

396

(98)
(28)

735 (126)
19 (19)
19 (19)

(106)

773

-

28
“

(154)

221

Red maple
Red oak
Black cherry
White ash

Table 6.3 (cont'd).

Seedling-

Stand
1

2

3

4

5

Stems/m2

Species

7.4
0.4
0.2

Sum:

(1.3)
(0.2)
(0.1)

4.2
0.5

(0.6)
(0.2)
-

8.1
0.6

(1.8)
(0.1)
-

-

-

-

-

-

-

0.7

(0.3)

0.1

(0.1)

0.3
0.02

(0.2)
(0.02)

8.8

(1.7)

4.8

(0.4)

9.0

(1.7)

^2.5 cm < dbh < 10 cm.
— Values are the means (+ SE) of four plots.
^Db h < 2.5 cm.

7.2
(1.9)
(0.04)
0.2
0.2
(0.02)
0.02 (0.02)
0.02 (0.02)
4.1
(0.4)
0.06 (0.04)

11.8

(2.3)

5.4
0.2

(0.8)
(0.1)
-

1.9

(0.7)

7.5

(1.2)

222

Red maple
Red oak
White oak
Black oak
Paper birch
Black cherry
White ash

223
Red Maple Age Structure
The

age

distributions

for

red

reflect two clearly defined cohorts
in the

first cohort

of stand development

this

(Fig.

established within
and ranged

the time of sampling.
in the

maple

1.5

6.1).

the

in age

>

m

tall

Individuals

first

ten years

from 60-70 years

at

Ninety percent of the genets sampled

cohort were

apparently of s p r o u t - o r i g i n .

Most

ramets were from m u l ti-stemmed genets that had branched from
stumps

or

growing
stem

root

in

a

bases,

around
this

circular
both

remnant

cohort

cohort)

collars.

of

arrangement
which

stumps.

(hereafter

is reflected

the five stands

Additionally,

are

The

or

had

of

to

ramets

crescent

indicative

extent

referred

many

of

the

shaped

development

ramification

as

were

within

sprout-origin

in the high ramet to genet

ratios

for

(2:1-3:1).

M ea n age of individuals in the sprout-origin cohort was
not

significantly

6.4).

different

Additionally,

among

stands

(P>

0.5,

there was not a significant difference

between the age of the bigtooth aspen and red maple
all individuals
t s= 0 .479,
examined,
four

older

0.5;

than

An additional

that was
aspen.

P>

on each plot

Table

6.5).

(mean of

(paired t - t e s t :

Of

the

20

plots

eight had at least one red maple that was three or

years

plot.

in the cohort)

df=19,

Table

one

or

two

the

sampled

bigtooth

aspen

five plots had at least one
years

older

than

the

sampled

on

the

red maple
bigtooth

Only one plot did not have a red maple at least as

old as the sampled bigtooth aspen.

224

Figure 6.1.
The number of new red maple individuals
establishing as
a
function of stand age
in
five
bigtooth aspen-dominated
forests within the HuronManistee National Forest.

225
Figure 6.1

180
160

180

■i STAND 1

- STAND 2

160

140

140

120
100
80

80
60
40

20

20

0

1
0

5

r ••

i — i— i— f -

0'

10 15 20 25 30 35 40 45 50 55 60 65 70

fTh

- H1 1 1 1 1 1 i- P1
1 I 1 1
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

160 -

STAND

4

120
100
80

NUMBER

NEW

140

OF

STEMS

180

STAND 3

10 15 20 25 30 35 40 45 50 55 60 65 70

^ T T
i- m
i
r T
5 10 15 20 25 30 35 40 45 50 55 60 65 70

YEARS AFTER STAND INITIATION
STAND 5

5

t
i
r~~lp =l P
l
p —r
10 15 20 25 30 35 40 45 50 55 60 65 70

YEARS AFTER STAND INITIATION

226

Table 6.4.
Mean age and height of red m a p l e sproutorigin cohorts in five bigtooth aspen-dominated
stands within the Huron-Manistee National Forest and
analysis of variance for comparisons among stands.

Height
Stand

Age

67
67
66
66
66

1
2
3
4
5

(years)

after stand initiation

(0.6)—
(0.6)
(0.1)
(0.4)
(0.4)

— Values are the means

(m) at 70 years

17.4
17.3
18.6
18.6
16.3

(0.6)
(1.7)
(1.3)
(1.0)
(0.6)

(+ SE) of four o b s e r v a t i o n s .

Age
Source

Stands
Error

df—

4
15

SS

15.292
76.233

^df= degrees of freedom,

Height
P > F

0. 57

SS

2.257
13.313

SS= sum of squares.

P > F

0.64

227

Table 6.5.
M e a n age and height of bigtooth aspen and
sprout-origin red maple on 20 plots within the HuronManistee National Forest.

Height
Years
Species

Bigtooth aspen
Red maple (all stems
pooled by plot)
Red ma p l e (tallest)
Red maple (shortest)

— Values are the means

Age

(years)

(m) at 70

After Stand Initiation

66
66

(0.2)a
(0.2)

24.8
17.6

(0.5)
(0.5)

66
66

(0.3)
(0.5)

21.5
14.4

(0.8)
(0.6)

(+ SE) of 20 observations.

228

The second red maple cohort was composed of individuals
(> 1.5
years

m tall)
after

that began

stand

F i g . 6.1).

establishing approximately

initiation

Across

all

(assuming

five

stands,

stand
there

age=70
w ere

surviving individuals that established between
years after stand initiation
All

individuals

in

s ee d l i n g - o r i g i n .
portion

of

the

the

The

red

(Fig.
the

6.1).

total

cohort

to

establishment
all

standard

errors

among stands
The

each
<

of

number

the mean
(from

plot)
1.7)

(F= 2.017,

age

only

of

were

six

10 and 30-35
6.1).

apparently

individuals

in

of

this

(hereafter referred

to

was obviously variable among stands

establish
on

cohort

maple population

However,

years;

(stands 2, 4 and 5; Fig.

second

at the seedling cohort)

30-35

number of years
the

was
and

year

similar
not

of

for

10%

bigtooth

(34.3-38.3

significantly

of

aspen
years,

different

P= 0.17).

individuals

within

the

subsample

of

stems

less than 1.5 m tall ranged from 1 to 3 0 years. M ean age for
these

individuals,

stratified by

shown in Table 6.6.

red

maple

sampling.
exceeding
of

35

cohort

Additionally,
1.5 m

years,

individuals

less

is
than

the establishment period

had

continued

through

the

since so few surviving

in height
it

height classes,

are

It is clear that with inclusion of this

segment of the population,
second

0.5 m

for the
time

of

individuals

established prior to a stand age

highly
1.5 m

unlikely
tall

established more than 35 years ago.

at

that

the

time

any
of

unsampled
sampling,

229

Table 6.6.
Mean age of red maple seedlings less than
1.5 m tall, stratified by 0.5 m height classes, in
five bigtooth aspen-dominated stands within the
Huron-Manistee National Forest.

Height Class
Stand

0-0.5

0.5-1.0

Age

1
2
3
4
5

11
9
11
13
10

(o.i) —
(0.6)
(1.3)
(2.5)
(2.6)

— Values are the means

15
13
16
19
17

(m)
1.0-1.5

(years)

(1.5)
(1.0)
(4.6)
(3.0)
(2.7)

20
17
17
23
20

(2.1)
(2.1)
(7.1)
(4.6)
(1.5)

(+ SE) of four observations.

230

Red Maple Height Structure
Mean height
not

of the

significantly

6.4) .

different

Within

variable,
for one

red maple

stands

among

and

as shown by the

(ts= 0 .894,
SE)

from

the

d f = 1 8 , P>

height

range

for

O v e r a l l , differences
height variation.
h eight were
for three

from each

(rs= -0.42,

Table

were

highly

this

stand

(Fig.

6.2).

6.5),

cohort

on

although

comparison

was

each

plot

the mean

7.3

(0.1)

(+
m.

in stem age contributed v e r y little to

five

between stem

(rs= 0.56,

P< 0.04),

age and

(rs= 0.09-0.25,

s t a n d s , while

correlations were wea k
sign

Table

Correlations

the

0.6,

in age of the tallest and

low and non-significant

of

(P>

heights

sprout-origin

0.3;

cohort was

individual height growth profiles

There was virtually no difference
stem

stands

plots,

randomly selected plot

shortest

sprout-origin

in the

P< 0.02)

total

P> 0.2)

other two,

the

or of the wrong

assuming that older stems should

have been taller.
On

most

of

the

plots

examined,

the

sampled

bigtooth

aspen equaled and usually exceeded most of the red maple
height throughout the 70 year period examined
see Fig.
highly

6.2) .

(for examples,

The m ean height difference at 70 years was

significant

(bigtooth

aspen

versus

the

red maple sprout-origin cohort on each plot;
19, P< 0.001; Table 6.5). However,
four had

at

exceeded

(> 0.3 m)

70 years.

in

least

of

the

t s =11.08,

df=

of the 20 plots examined,

one red maple that
the

mean

equaled

sampled bigtooth aspen

(± 0.3 m)

or

in height by

Eleven plots had at least one red maple that was

231

Figure 6.2.
Height growth reconstructions of individual
bigtooth aspen and sprout-origin red maple w i t h i n one
randomly selected plot
from each of five bigtooth
aspen
dominated-stands within
the
Huron-Manistee
National Forest.
Each line represents an individual
stem.

232
Figure 6.2

30-

B ig to o th A spen
Red M aple
20-

1510-

STAND 1

STAND 2

5-

PLOT 1

PLOT 3

r=j | | | | | | | | ! p

CUMULATIVE

STEM

HEIGHT

(M )

0

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

0

30

30

25-

25-

20

20-

15-

15-

10-

10-

5-

0

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

STAND 3

STAND 4

PLOT 1

PLOT 2

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

YEARS AFTER STAND INITIATION
30
2520 1510-

STAND 5

5-

PLOT 2
0

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

YEARS AFTER STAND INITIATION

233

only 0.3-3 in shorter than the sampled bigtooth aspen on the
plot.

While heights

tallest

red

different

maple

(ts=

of the sampled bigtooth aspen and the

on

each

4.99,

df=

plot
19,

comparison w e r e muc h closer,

were

P<

still

0.001),

means

stronger

heights

within

relationship

sprout-origin

to

the
stem

individuals.

stands

(Fig.

correlation between
(including the
was

0.63

seedling
age

This

6.3).

cohort

than

can

was

be

All

upper two-thirds
youngest

distributions,

by

case

examining

plot-level

height of
<

correlations

for

from each of the
(±sd)

individuals

a

1.5 m

seedlings
in height)

coefficients

were

In g e n e r a l , stems from roughly the

of the

stems.

showed

the

seen

mean

and total

subsample of

significant at P< 0.01.

the

The

stem age

(0.09).

this

(Table 6.5).

one randomly selected stem age-height plot
five

for

relative to the entire sprout-

origin red maple-bigtooth aspen comparison
Total

significantly

age distributions

Within

variability

these

in

were taller than

portions

height

among

of

the

age

similar

aged

individuals was a significant feature of the seedling-cohort
(Fig.

6.3).

Red Maple Height Growth Patterns
Obviously,
specific

stem

seedling

differences
ages

cohorts

Height

growth

detail

by

within
led

rate

examining

to

in
the
the

height
both

the

growth

relationships

was

about

sprout-origin

variable

variability

rate

height

assessed

between

the

and

patterns.
in

greater

competitive

234

Figure 6.3.
The relationship between stem age (years)
and total height (meters) for individual seed-origin
red maple
in five bigtooth aspen-dominated
stands
within
the
Huron-Manistee
National
Forest.
One
randomly selected plot from each of the five stands is
shown.
In each plot, height classes above the dotted
line were completely sampled. Height classes be l o w the
dotted line were randomly subsampled.

235
Figure 6.3

7

7

STAND 1
6
5

STAND 2

PLOT 1

6

rs= 0.58
P < 0.0001

5-

4

4

3

3

2

2

-

VAA a
1

1

0

0
0

5

10

15

20

25

30

35

40

7

rs= 0.68
P < 0.0001

5

A

fc*AaA
aA
A
^a

&

- A -A
--a Aa A
£A ^
AAAAA
T--

STAND

PLOT 1

6

rs= 0.44
P < 0.0001

5-

4

4-

3

3-

2

2-

1

1

0

0
0

5

10

15

20

25

30

35

40

~. 4. . . A
A A

STAND 5
PLOT 2

A ^

A

rs= 0.80
P < 0.0001

^
S

A
A .

3

A

&*A

A

A

4

2

&

A

#

A
S

A

aA

AA A

..........

1
a

0

A

a

10

25

20

30

35

40

35

40

4

PLOT 2
rs= 0.59
P < 0.0001

A

ti"i^ A ^ iA|? iA ii Ii riI
-'IiIII
0
5
10
15
20
25

STEM AGE (YRS)

IIIiiIi-ri
30

35

40

A ^A
A

A aa
A A

A

*

A A

A

S A

i-i i 11111 ' T 11 r" '" r 1
3
5
10
15
STEM

7

5

AA a

7

6

6

A

M

AA£

5

STAND 3
STEM HEIGHT (M)

PLOT 3

20
AGE

25
(Y R S )

30

236

status

of

an

height,

and recent patterns of height increment.

Within
height

individual,

the

reflected

sprout-origin

increment

immediately

as

prior

cohort,

(in

meters)

over

to

sampling

and

beginn i n g of this period were low
significant

(P= 0.07-0.76)

same correlation
0.02)

in

the

o verstory

red

relative

correlations
the

between

five-year

initial

period

height

(rs= -0.08-0.40)

(rs=

stand,

maple

its

at

0.57)

were

This

and significant

indicating

that

the

and non­

in four of the five stands.

was higher

fifth

by

some

(P=

taller

growing

faster

than

shorter

cohort,

correlations

between

conspecifics.
Within
initial

the

height

m o d e ra t e l y

that

shorter

lowest

and

high

indicating
than

seedling
height
(mean
some

stems

(0.8-1.9

with
m) ,

(for

the

(±sd)
17

0.0001)

plots

examples

see

to

low

the

each

plot

(+0.24)),
often

Fig.

again

grew

6.4).

were

faster

The

three

(rs= 0 .19-0.37)
m aximum

remaining

were

initial
plots

heights
(initial

After excluding these three plots,

correlation coefficient was 0.68

remaining

correlations

were

(0.08).

significant

(P<

the

All of
0.001-

.

w i t h i n the

were

rs= 0 .60

individuals

having

relative

R elationships

detail

within

coefficients

heights= 2.2-5.5 m ) .
mean

(±sd)

taller

correlation

associated

increment

by

at

between

stem

seedling-origin
focusing

least

1.5

only
m

in

on

age

and

cohort were
the

height

fastest
at

the

height

growth

examined
growing
time

of

rate

in greater
stems

that

sampling.

237

Figure 6.4.
The relationship between height increment
(meters) over the five-year period immediately prior
to
sampling
and
initial
height
(meters)
at
the
beginning of this period
for individual
seedlingorigin red maple
in
five bigtooth
aspen-dominated
stands within the Huron-Manistee National Forest. One
randomly selected plot from each of the five stands is
shown.

238

Figure 6.4

3.5-

3.5STAND 1
PLOT 1

3.0H
2.5-

2.5-

rs- 0.37
P < 0.0001

2.0-

rs- 0.74
P < 0.0001

2.0-

1.5-<i

1.5-

1.0-<1

A

a

A

1.0-

. .a..A

A A A

0.5- A

0.5- 4.A/VS & V A A
0.0

STAND 2
PLOT 3

3.0-

a

^ A

a ^ l l ^ A

a

A

A

A

A
h■

W

r"'""l - l

T

r

r-T—p-T- r r-

3.5-

0.0-

— i— l— r-A— i— l— |— i— i— l— |— l—l—l— |— i—i— i—

3 .5 -j

STAND 3
PLOT 1

Z

2 .5 -

rs= 0.67
P < 0.0001

ce u j

<2

STAND 4
PLOT 2

3 .0 -

rs— 0.81
P < 0.0001

A
2 .0 -

wlu

>-cd
UjU
>5

A
<1
<5

4

1 .5 -

A
A
A
A A^ A A

1 .0 -

X
o
LU

0 .5 -

X

0 .0 -

A A

AA

A
A

^

A A . ................. ..
INITIAL HEIGHT (M )

3.5-

STAND 5
PLOT 2

3.02.5-

rs= 0.78
P < 0.0001

2.0 1.5

aa
a A ^
4A

1.0 -

0 .5 -

4

A

A.A

A

Aa
AA

4CA
AAa
A . &A A

0.0

INITIAL HEIGHT (M )

A

239

These individuals are those forming the u p p e r bounds of the
age-height diagrams
growing

(Fig.

individual

6.3).

on a plot was

five-year stem-age classes
years

old,

increment
ii)

etc.)
(over

For each stand,

and

the

growth

rate

interval

shared

by

examined

for

period

over

all

from

the

i)

prior

first

stems.

its

Plots

recent

to

age

classes

relationship
was

0.24,
low

between

inconsistent.

stands

were

the

and

1.5

m,

were

growth were

not

distributions.

age

negative

plot.
and

height

pooled

within

from the fivestands,

height

coefficients
(rs=

the

Among

recent

-0.59,

height

for
P=

the

increment
two of

0.01;

the

r s=

-

In the o t h e r s , the coefficients were either
(rs=

P= 0.005).

individuals

0.95)

stem

non-significant

0.54 and 0.58,

each

Correlation

w eak

P= 0.42).

and

w i thin

6-10

s a m p l i n g ) , and

stands because of the low number individuals
year

consecutive

(for example 0-5 years old,

5-year

its

selected

the fastest

with

the

P>

0.50)

or

These correlations
fastest

restricted
In contrast,

and significant

0.17,

rates

of

(rs=

indicate that
recent

to specific portions
there was a strong

(P< 0.006-0.0001)

weak

height

of the

age

(rs= -0.69-

inverse relationship

between stem age and growth rate over the first 1.5 m in all
stands

(Fig.

6.5),

indicating that the

individuals with the

fastest rates of early height growth were often the youngest
stems.
Temporal Patterns of Overstory Height Increment
The

pattern

of

red

maple

seedling

beginning 35-38 years after stand initiation

establishment,
(Fig.

6.1),

and

240

Figure 6.5.
The relationship between stem age (years)
and height growth rate (m/yr) over the first 1.5 m for
the fastest growing seedling-origin red maple in five
bigtooth
aspen-dominated
stands
within
the
HuronManistee National Forest.
For each stand, the fastest
growing individuals from consecutive five-year stemage classes (i.e. 5-10 years old, 10-15 years old,
etc.)
within
each of
four plots were pooled
for
analysis.

241
Figure 6.5

0 .3 5 -

0 .3 5 -

0 .3 0 -

A

STAND 1

A

rs = - 0 . 8 6
P < 0.0001

0 .2 5 A A

0 .2 0 -

A

A

A

0 .1 0 -

.A

A

A
aa

a

<3

(M /Y R )

A
A

<1

RATE

P = 0.001
A3

0 .1 5 -

A

0 .0 5 -

0 .0 5 -

* 11' ) 11 r ’ i 1TT'' i 1r r r i 11 tt 11 t 7 11 i_r ' 1'T 1111
5
10
15
20
25
30
35
40
45

0.35-]

0

i n i | i i r i | i i i i | i i"i q-iT i i-prr-i I p-r-ri-|T n 'i | i i i I
5
10
15
20
25
30
35
40
45

0.35-1

0 .3 0 -

STAND

A A

3

STAND

0 .3 0 -

rs - - 0 . 8 6
0 .2 5 -

A

0 .2 0 -

A

P < 0.0001

A
A

.

<

&

0 .0 5 -

A
A

A

&
£

%

0 .0 5 -

n i i | i—
i1n | i i-i i | i ri i | i ■r i | i i i t | i i i r-| i i i i | i i i i
5
10
15
20
25
30
35
40
45

1.5

0

P = 0 .0 06
A

0 .1 0 -

A

U.UV

0 .2 5 -

0 .1 5 -

0 .1 0 -

-rrr-rp i
0
5

| i i ii| i
10
15

it

|i
20

iii

| n I i j-ri'i i | ct i i | i i m
25
30
35
40
45

i ii

ST E M A G E ( Y R S )
0.35
0.30

STAND 5

A

0.25

rs— -0.95
P < 0.0001

A
AA

0.20

A
A

0.15

AA A
0.10

a

A

aaa

A

0.05

0.00

tt t t t"i t t i

0

5

A

I I IT. ■i i t i i i i rrp i . T| ri i r p r ■i p r r
10
15
20
25
30
35
40
45

STEM AGE (Y R S )

4

rs » - 0 .6 9

0 .2 0 -

%

0 .1 5 -

A
A
A

AA
A
0

GROWTH

A

<

A A

A

aaa

M HEIGHT

0 .2 5 -

t

0 .1 0 -

O

r s = - 0 .7 3

0 .2 0 -

0.15 -

n rw~i

STAND 2

0 .3 0 -

242

the

inverse

rate
be

(Fig.

relationship between
6.5),

negatively

related

to

w ith

individuals,

particularly

was

by

of

periods.

patterns

understory

concurrently

increment

and early

overstory

reinitiation

decreasing

height

bigtooth

examining

bigtooth

of

growth

aspen.

temporal

aspen

over

of

This

occur
overstory

hypothesis

patterns

in

consecutive

height

five-year

The height growth patterns of the fastest growing

also

begun

growth.

might

and the median growing sprout-origin red maple
were

growth

suggests that understory reinitiation might

Specifically,

tested

stem age

10

examined.
years

Analyses

after

stand

of

growth

initiation,

in each plot

increments

the

stand

were

age

by

which all sampled bigtooth aspen and overstory red maple had
established.
stand

ages

years

and

were

Contrasts of overstory height increment between
of:

1)

30-35

conducted

10-35

years

y e a r s ; and
in

an

and

3)

30-35

attempt

to

35-70
years

clarify

y e a r s ; 2)

25-30

and

years

35-40

overstory

growth

patterns relative to patterns of understory development
In
aspen

general,

and

red maple

increasing
stand

growth

initiation,

remainder

of

significant
height

the

trend
over

in

the

increment

period

over

a

10

(Fig.

and

6.6).

age

30

Time

of

bigtooth

was

one

years

increment

(P< 0.025;

stand

of

period

decreasing

in all three cases

increment

growth

sixty-year

between

followed by

the

height

after

for

effects

the
were

Table 6.7).
10-35

of

Mean

years

was

significantly greater than height increment over 35-70 years
in

all

cases

(P<

0.01-0.001;

Table

6.7).

M ean

height

243

Figure 6.6.
Height increments (m) over consecutive fiveyear stand-age classes, beginning 10 years after stand
initiation, for bigtooth aspen and sprout-origin red
maple.
Values
are
back-transformed
means
(+
95%
confidence intervals) of 20, 18 and 19 log-transformed
observations for bigtooth aspen, the fastest growing
red maple on each plot and the median growing red
maple on each plot, respectively. In each graph, the
bottom point of the inverted triangle indicates the
year after stand initiation (mean of five stands) by
which 10% of the seed-origin red maple > 1.5 m tall
had established. The single open triangle at the right
end of each graph is the mean
(+ 95% confidence
intervals) height increment of the fastest growing
seedling-origin red maple over a stand age of 65-70
years.

244
Figure 6.6

B ig to o th A spen

0H

i i i i i r ~~i
i i i i r
1 0 -1 5 2 0 - 2 5 3 0 - 3 5 4 0 - 4 5 5 0 - 5 5 6 0 - 6 5

Red M aple

5

(fa s t g ro w in g)

x

U
J
LU

g

2
Z
I—
O H
L
U
X
_1

!

!

!

|

|

|

|

|

|

|

p

1 0 -1 5 2 0 - 2 5 3 0 - 3 5 4 0 - 4 5 5 0 - 5 5 6 0 - 6 5

Red M aple
(m e d ia n g ro w in g)

1

-

~i—

i

1—

i—

i—

i—

i—

i—

i—

i—

i—

1 0 -1 5 2 0 - 2 5 3 0 - 3 5 4 0 - 4 5 5 0 - 5 5 6 0 - 6 5

YEARS AFTER STAND INITIATION

r

Table 6.7.
Repeated measure analysis of variance and specific contrasts for height
increment across time in bigtooth aspen and sprout-origin red maple.

Species
Red maple

Bigtooth aspen

Source

df—

Plot
Time

19
11 (1)209

(19)

0.777
9.094

P > F

df

17
11 (1)

<.0001
(.0001)^

187

8.081

Bigtooth Aspen

Stand ages
(years)

10-35 vs 35-70
25-30 VS 30-35
30-35 vs 35-40

fcb18.148
4.482
-0.011

SS

(17)

1.475
5.766

<.0001
(.01)

10.136

Red Maple

P > F

< 0.001
< 0.001
> 0.500

P > F

df

(fast)

< 0.01
< 0.01
> 0.10

(median)

SS

18
11 (1)
198

P > F

3.584
3.511
-2.245

Red maple

(18)

0.827
3 .805

4.046
0.173
1.411

<•0001
(.025)

10.721

Red Maple

^b

P > F

(median)

P > F

< 0.005
> 0.500
> 0.250

— df= degrees of f r e e d o m , S S = sum of squares*
— Values in parentheses are adjusted df and probabilities associated with the GreenhouseGiesser conservative F-test.
— tfc-values are for Bonferroni contrasts.
Critical value= +t^ alpha/2, 3 ,dfe r r o r .

245

Error

SS

(fast)

246

increment

over

25-30

years

was

significantly

greater

than

height increment over 30-35 years for the bigtooth aspen and
the faster growing red maple
not

for the

Mean height

slower growing

(P< 0.01-0.001; Table 6.7), but
red maple

(P> 0.50;

Table

6.7).

increment over stand ages 30-35 years and 35-40

years were not significantly different

for any contrast

(P>

0.10-0.50).
Figure 6.6 also shows the mean recent height
(over a

stand

understory

age

red

of

65-70 years)

maple

relative

overstory red maple.
5

to

35

years

for
to

the

increment

fastest

bigto o t h

growing

aspen

and

The understory individuals ranged from

of

age

at

the

time

of

sampling.

Height

increment of the understory stems was significantly greater
than

that

of

both

age of 65-70 years
t s=

2.73,

df=

overstory

red maple

groups

(fast: tg= 2.30, df= 17,

18,

P<

0.02) ,

but

was

different from that of bigtooth aspen

over

a

stand

P< 0.05; median:

not

significantly

(ts= 0.83,

df= 19,

P>

0.40).
DISCUSSION
Origin of the Red Maple Sprout Cohort
Red maple was historically a minor overstory component
of

the

presettlement

Lakes region.
and

Wisconsin,

indicate

that

pine

forests

of

Reconstructions of these
using
red

General

maple

Land

comprised

the

forests

Office
no

northern

in Michigan

survey

g reater

Great

records,

than

seven

percent of the b earing and line trees used to locate section
lines

(Whitney

1986;

Nowacki

et

a l . 1990;

Chapter

3),

but

247

was

more

abundant

in

the

understory

of

these

forests

(Whitney 1986).
In
cohort
H igh

all

of

was

the

stands

largely

composed

proportions

overstory
mature

red

maple

mi x e d

of

the

older

sprout-origin

sprout-origin
populations

red

been

forests

maple

individuals.

individuals

have

aspen-dominated

northern Michigan
1985),

of

bigtooth

examined,

within

reported

for

elsewhere

in

(Roberts and Richardson 1985; Sakai et a l .

hardwood

forests

in the

Southern

Appalachians

(Beck and Hooper 1986), and presumably is true for red maple
in

oak

forests

that

developed

northeastern United States

after

clearcutting

in

the

(Lorimer 1984).

Red maple genets survive and resprout readily from root
collars and stumps following fire or cutting
Swan

(1970)

found

surface

fires

(1988)

reported

experimental
bigtooth

in

no

mortality

oak-dominated
high

root

forest.

and

systems

forests.

survival

clearcutting

aspen

of

(Fowells 1965).

Scheiner

of

genets

burning

of

Additionally,

following

a

they

et

al.

following

70

year

old

no

new

found

genets establishing over the first four post-fire years the
stand was examined,

despite a nearby seed source.

Sprout-origin individuals examined in the present study
may

have

original
sites,

developed

presettlement

rather

seed after
is

from

than

removal

suggested by

root

pine

from

forests

data

remnant

that

individuals which

of the pine

the

systems

overstory.

from Swan

(1970)

from

occupied
established

the
these
from

This hypothesis
and

Scheiner

et

248

al.

(1988) and observations on the spatial pattern of ramets

in the current study, which indicated resprouting from large
diameter stumps or root collars.
of

the

sprout-origin

provide
height

some
of

(mean ±
values
red

additional

0.59

reported

maple

individuals
evidence

sprout-origin

SE=

+

red

0.04

m)

on

dry,

in

for

maple
was

this

at

rocky

and

growth

are

related

negatively

stump or root system

rate
to

present

three

soils

in

years

age

of

one-half

The
age
of

sprout-origin

the

northeastern

It is known that both

of red maple vegetative
the

study

hypothesis.

smaller

(Kays and Canham 1991).

production and

the

approximately

for much younger

growing

United States

Early height growth rates

and

size

of

sprouts

the

parent

(Solomon and Blum 1967; Kays and Canham

1991).

Red Maple Age Structure
The

age

distribution

examined in this study
of

even-aged

In all
by

of the

rapid

within

a

sampling.

stands

ten-year

(Fig.

examined,
of

period,

the

6.1)

development

establishment

the

red

maple

populations

clearly reflect the model

proposed
stand

by

Oliver

(1981).

initiation was marked

initial

beginning

overstory

red

70

prior

years

maple
to

This occurred concurrently with establishment of

bigtooth aspen
5)

stand

of

(and all other overstory species; Chapters 4,

predominantly

via

resprouting.

Stand

initiation

was

followed by a period of stem exclusion lasting approximately
25 years,

after w h i c h time red maple,

and virtually no other

249

species,

began

Red maple

to

establish

seedling

time of sampling
delayed

from

establishment had

establishment

initiation

have

reported

forests

northern

1985;

Roberts

in the
and

been

and Richardson

northeastern

Ree d

1924;

Peet 1981),
Northwest

Michigan

Peet

and
and

in

the

understory.

continued

(stand age= 70 y e a r s ) .

seedling

in

seed

through

Similar patterns of

subsequent

for

mesic

(Sakai
1985),

to

stand

aspen-dominated

1990a;

Sakai

et

pine-dominated

southeastern U n ited
Christensen

the

1987;

al.

forests

States

(Tarbox

C h ristensen

and

and coniferous forests in Alaska and the Pacific

(Alaback 1 9 8 2 a ; Oliver et a t . 1985; Harcombe 1986;

Fried et al.

1988).

Understory Reinitiation and Overstory Development
Roberts

and

Richardson

(1985)

discuss

relationships

between overstory development and understory reinitiation in
aspen-dominated forests of northern Michigan.
that on xeric or highly disturbed
stocking,

sites

with

They suggest
low

canopy closure is delayed or incomplete,

turn allows continuous understory e s t a b l i s h m e n t .
earlier

overstory

study

found

continuous

establishment

which

in

Indeed,

an

of

several

later successional species in a understocked bigtooth aspendominated
grazing,
2).

forest

that

had

initiated

following

and fire in a northern hardwood community

cutting,
(Chapter

On more favorable sites they suggest that seedlings of

tolerant,
maple,

later

successional

can establish and

tree

increase

species,
in height

including
under

a

red

closed

250

canopy.

However,

indicates that
extended

(Zeide

should

be

theoretical

canopy closure

periods

stands

stand

recent

of

time,

1987,

reached

does not

even

1991).

in

followed

decrease

is

attributed

thinning

and

a

by

to

decreasing

stocked

max i m u m
in

the

of

closure

course

decrease.

combination

of

canopy

for

even-aged

canopy

subsequent

rate

evidence

remain complete

early

a

a

empirical

fully

Rather,

asymptotically,

development,

and

of
The

overstory

expansion

in

surviving dominant individuals.
In

the

increment

present

in

s t u d y , the

bigtooth

aspen,

significantly

and

in

lower

overstory

height

red

maple,

beginning immediately prior to understory reinitiation

(Fig.

6.6),

that,

suggests

coupled
canopy

wit h

a

overstory

closure

and

understory.

Red

understories

may

closure.

thinning,

increased

maple
be

crown-growth
may

resource

seedling

dependent

reduction

result

should

be

availability

this

reduction

correlated

with

in

the

in

aspen

in

canopy

and that

in even-aged stands should begin to decrease

canopy dominants

(Zeide 1987,

resource
is

overstory

Oliver and Larson 1990)

immediately after the period of maximum height

examined,

decreased

reinitiation

on

negatively

(Alaback 1982a;

canopy closure

understory

in

This is consistent with the belief that understory

development
production

stand-level

by the

in

1991). A continued increase in

availability,

suggested

increment

over

inverse

the

time

period

relationship

between

age and early height growth rates of the fastest growing red
maple seedlings

(Fig.

6.5). Changes in seed rain,

as sprout-

251

origin

red

maple

become

reproductively

mature,

may

also

influence the pattern of seedling establishment.
Increasing
stand

light

development

availability

may

be

understory reinitiation

and

even-aged
Fried

forests

et

al.

trenching
1931)

peak
This

availability
factor

that
of

new

stems

Experimental
availability

understory

in

that

stand

in

1985;

the

development
if

root

is

understories

manipulation
be

of

required

of

light
to

of
and

can

other

1982a,b;
classic
Kienholz

occur

even

competition

increasing

development

and

Ala b a c k

Tourney

soil

also

important

recruitment

even-aged
soil

adequately

is

resource

an

and

and

of

controlling

results

1929;

course

these

e s t a b l i s h m e n t , survival
the

will

al.

exclusion,

suggests

influencing

et

the

factor

development

(Tourney

stem

during

major

However,

experiments

the

reduced.

of

1988).

indicate

during

(Oliver

a

during

forests.
resource

define

the

mechanism driving understory reinitiation.

Red Maple Height Structure
Once established,

height growth among individual genets

was highly variable in both the sprout and seedling cohorts.
For

the

former,

heights

at

the

time

particularly dependent on stem age.

of

sampling

were

not

Height differentiation

within the even-aged overstory was m ore likely a function of
differences
vigor

of

in

parent

competitive

microsite-related
root

status

of

systems,

resource

genetic

individuals

availability,

potential,

(Peet

and

and

the

Christensen

252

1987; O l i v e r and Larson 1990).
be

a

p a r t i cularly

differ e n t i a t i o n
populations
1978;

Wierman

Kelty

1936;

and

factor

Oliver
1986;

trees

(1990)

between

relative

found

1979;

Guldin

Hix and

Lorimer

significant,

height

and

of

of

an

the

positive

radiation and height

height

increment

position
factor

weak

within

in
the

influencing

in the canopy

probably

b ecause

competitively
Competitive
physical

to

an

inhibition
of

may

the

maple by bigtooth aspen,

late

in

taller

terminal

for

light
of

Hutchison

height

suggests

overly

stand

tallest

have

height

levels

initial

not

by

Hix and

In the present study,

between

the

Oliver

of

1970;

overstory

but

1985;

relationships

between

maple

all

Oliver

1991;

compete

(Assmann

was

and

rates

red

inhibited

abrasion

1990,

sprout-origin

rate

among

Relative height can be

relationship

growth

and

Lorimer

positive

ability

relationship

height

For example,

and Matt 1977; Cole and Newton 1986).
the generally

and

short-term

individuals

to

(Stubblefield

growth for several hardwood s p e c i e s .

because

with i n

Oliver and Larson 1990).

Lorimer

indicative

leading

stratification

similar-aged

Smith

et a l . 1990;

important

and

of

The latter has been shown to

important

maple

bigtooth

leaders

were
aspen.

through
of

that

development,

red

occurred

and

direct

shorter

red

a mech a n i s m that has been found to

limit height growth in both deciduous and coniferous species
(Oliver 1978; Wierman & Oliver 1979; Kelty 1986).
H eights of individuals in the seedling cohort were more
dependent

on

stem

age.

This

is

not

surprising

given

the

253

greater age range of the cohort
sprout-origin

individuals

observed v ariation

(> 30 years)

(10 year

age

compared to the

range).

in height among similar-aged

Still,

the

individuals

within the cohort suggests that competitive ability differed
greatly among s t e m s .
an

competitive

In general,

advantage

in

the tallest stems were at

terms

of

recent

rate, but these were not always the oldest
fact,

among the

individuals

fastest growing members

with

the

highest

rates

sometimes the youngest s t e m s .

This

growth

individuals.

of

of

height

the

height

In

c o h o r t , the
growth

were

indicates that position

beneath the overstory may be an important factor influencing
height

growth

of

understory

formation of large treefall

stems,

gaps.

It

even

prior

to

the

is clear that height

variation developed prior to the formation of large treefall
gaps,

since

all

the

stands

very high basal areas
very

few

large

downed

that

availability

of

examined

be

related

boles

were

resources,

to

the

distribution,

dominant overstory
mortality

in

interactions
Larson

1990).

with

understory

Differential

This

levels

and early
other

and

suggests

required

recruitment,

than

by

mortality

for
may
of

These additional factors could
growth

individuals,

overstory

observed.
at

factors

mature overstory individuals.
include

characterized

(33-47 m 2 at 66 y e a r s ; Table 6.2)

seedling e s t a b l i s h m e n t , survival
initially

were

rates

patterns

dominants,
herbs
browsing

and

and

species

of

of branch and root
and
shrubs

pressure

competitive
(Oliver
may

also

and
be

import a n t . Similar conclusions were reached by Fried et a l .

254

(1988)

to

explain

height

variation

within

populations

of

big-leaf maple in even-aged Douglas-fir stands.
It

is

important

stand development

to

distinguish

during wh i c h

between

seedlings

of

the

period

overstory

of

tree

species begin to establish and are able to survive and grow
slowly

in

height

reinitiation

sensu

in

the

Oliver

period characterized by
even-aged

overstory

forest
1981),

understory

and

a

later

(understory
developmental

increased mortality of the original

(old-growth

phase

sensu

Oliver

1981).

It is during the latter period that a significant amount of
recruitment
from the

of

more

tolerant

understories

of

species

even-aged

to

stands

(Alaback 1 9 8 2 a ; Oliver 1981; Oliver et al.
Richardson 1985;
However,

as

Harcombe 1986;

the

data

show,

canopy

positions

typically

occurs

1985; Roberts and

Peet and Christensen
initial

1987).

establishment

and

recruitment to larger seedling or sapling height classes may
occur much
not

well

earlier.

This mode

addressed

by

overstory mortality

and

stand
the

development
availability

late in an even-aged sequence,
reinitiation

and

early

of even-aged development
models
of

is

stressing

treefall

gaps,

as the trigger for understory

recruitment

(see,

for

example,

Bormann and Likens 1979; Peet and Christensen 1987).

Landscape Patterns
The five stands examined in this research were
within

an

Overstory

18
age

km2
data

bigtooth
clearly

aspen-dominated
show

that

all

located

landscape.
stands

were

255

approximately 66 years old and may have initiated after the
same

major

population
height

post-logging
size

was

structures

wildfire.

variable

of

the

Although

among

stands,

populations

red
the

were

maple

age

ver y

and

similar.

This was true despite some difference in bigtooth aspen site
index among s t a n d s .
The similarity of red maple population structure among
the

stands

examined

similarities
stand

in

may

the

initiation.

be

the

result

historical
Pickett

of

landscape-wide

circumstances

(1989)

has

leading

emphasized

to
the

important role historical circumstances have on successional
pathways.
of

the

While the present study examined

total

historical

forest

community,

circumstances

five stands,
subsequent

including

fire,

loss

leading

recruitment of

of

seed

sources

and

occurred
historical

to

several

(Chapter

to the development
of

pine

apparently

compared

that

intensity

rootstocks,
study

suggest

and

survival

this

data

a portion

frequency

regene r a t i o n , and
were

the

only

of bigtooth aspen

earlier

tree

2),

circumstances

very

similar.
work,

The
where

species beneath

further
play

support

a

major

and

role

of

advance

red maple
results

of

continuous

notion
in

the

cutting,

bigtooth

the

the

how

aspen
that
stands

develop through time.

SUMMARY AND CONSLUSIONS
The results of several studies from the eastern United
States

indicate

that

red maple

is becoming an

increasingly

256

abundant

overstory

dominated by oak
and

Lorimer

1991b;
should

forests

the

much

shorter

potential
dry-mesic

and

occur

Great

inability

been

overstory

cited

region

slow

as

in

to

growth

without

the

factors

that

red

of

While

height

has

seedling
and

of

prevent

the

data

clearly

on

been

a l . 1963)

growth

the

maple

both

protection

will

a

taxon's

a

an

heat-

(Aber et a l . 1982; Pastor et al.

recruitment.

individuals,

the

region

et

Such

However,

by

rate

(Graham

of

oak.

Lakes

Hix

aspen-dominated

because

maple in the present study were variable,
some

historically

1992) .

dominance

Great

individuals

survive

Nowacki

relative

the

reducing nurse canopy
have

forests

faster

overstory

Inherently

to

much

span

in

sprout-origin

and

Lakes

future

sites

questioned.

Abrams

life

for

in

(Lorimer 1984; Heiligmann et a l . 1985;

transition
of

component

1982)

significant

rates

of

red

and quite low for

show that

sprout-origin

red maple can reach dominant-codominant canopy positions on
the sites examined.
tallest

Further,

individuals

Unlike

rates of height growth for the

approached

sprout-origin

stems,

those

of

bigtooth

seedling-origin

red

fail to recruit to dominant canopy positions.
growth rates
cohort
aspen

did
and

for the most vigorous members

equal

or

exceed

overstory

red

height

maple

growth

over

the

maple

may

While height

of the
of

aspen.

both

same

seedling
bigtooth

stand

ages,

these growth rates w ere still much lower than those of the
overstory
the

low

individuals
growth

rates

over
of

comparable

seedlings

in

stem
the

ages.

However,

understory

may

257

simply be the result of resource limitation.
red

maple

positions

may
in

be
gaps

successful
produced

bigtooth aspen overstory.
may also be
increased

favored by

growth

upon

by

of

seed

attaining

the

breakup

Additional

red maple's
release

Stephens 1977? Lorimer 1980).
composition

at

rain,

Seedling-origin
dominant
of

the

canopy
current

overstory recruitment

ability to respond with

(Gerrard

1969;

Oliver

and

Barring future changes in the
increasing

later

successional

importance and self-replacement of shade-tolerant red maple
is

suggested

by

its

overwhelming

understories of the stands examined.

dominance

in

the

BIBLIOGRAPHY
Aber,

J. D . , Pastor, J. , and Melillo, J. M.
1982.
in
forest
canopy
structure
along
a
site
gradient in southern Wisconsin.
The American
Naturalist 108: 256-265.

Changes
quality
Midland

A b r a m s , M. D . , and N o w a c k i , G. J. 1992. Historical variation
in fire, oak recruitment, and post-logging accelerated
succession in central Pennsylvania.
Bulletin of the
Torrey Botanical Club (in p r e s s ) .
Alaback, P. B.
1982a.
Dynamics of understory biomass in
Sitka
spruce-western
hemlock
forests
of
southeast
Alaska.
Ecology 63: 1932-1948.
________.
1982b.
Forest community structural changes during
secondary succession in southeast Alaska.
In Forest
succession
and
stand
development
research
in
the
n o r t h w e s t : Proceeding of the symposium.
Edited by J.
E. M e a n s . Oregon State University, Corvallis,
pp. 7079.
A s s m a n n , E.
1970.
Pergamon Press.

The principles

Barnes, B. V.
1966.
The clonal
a s p e n s . Ecology 47: 439-447.

of

forest

growth

habit

yield

study.

of American

Barnes, B. V. , P r e g i t z e r , K. S., Spies, T. A. , and Spooner,
V. H.
1982.
Ecological forest site classification.
Journal of Forestry 80: 493-498.
Beck,

D. E . , and Hooper,
R. M.
1986. Development of a
southern Appalachian hardwood stand after c l e a r c u t t i n g .
Southern Journal of Applied Forestry 10: 168-172.

Benninghof f ,
W.
S.,
and
Cramer,
K.
J.
1963.
Phytosociological
analysis
of
aspen
communities
on
three site classes for Populus grandidentata in western
Cheboygan County, M I . Vegetatio 11: 253-264.
Bormann, F. H . , and Likens, G. E.
1979. Pattern and process
in a forested ecosystem.
S p r i n g e r - V e r l a g , New York.
B r o w n , R. T. and Curtis, J. T.
1952.
The upland coniferhardwood
forests
of northern Wisconsin.
Ecological
Monographs 22: 217-234.

258

259

Christensen, N. L. and Peet, R. K.
1981.
Secondary forest
succession on the North Carolina piedmont, in Forest
succession: concepts and application.
Edited by D. C.
West, H. H. Shugart, and D. B. Botkin. Springer-Verlag,
N e w York.
pp. 230-245.
Cole,

E. C. and Newton, M.
1986.
Nutrient, moisture, and
light relations in 5-year-old Douglas-fir plantations
under variable competition.
Canadian Journal of Forest
Research 16: 727-732.

Crow,

T. R.
1988.
Reproductive mode and mechanisms for
self-replacement of northern red oak (Ouercus r u b r a l— a
review.
Forest Science 34: 19-40.

Gerrard, D. G.
1969.
Competition quotient: a new measure
of the competition affecting individual forest trees.
Michigan
State
University
Agricultural
Experiment
Station Research Bulletin 20.
D o d g e , S. L. and Harman, R. J.
1985.
W o o d l o t composition
and
successional
trends
in
south-central
Lower
Michigan.
Michigan Botanist 24: 43-54 .
Fowells, H. A.
1965.
Silvics of forest trees of the United
S t a t e s . USD A Agricultural H a n dbook N u m b e r 271.
Fried, J . S., T a p p e i n e r , J . C . , II, and Hibbs, D. E.
Bigleaf maple seedling establishment and early
in Douglas-fir forests.
Canadian Journal of
Research 18: 1226-1233.

1988.
growth
Forest

G i l l , J. L.
1978.
Design and analysis of experiments in
the animal and medical sciences.
V o l u m e 1.
Iowa State
University Press, Ames.
________ .
1988.
Design and analysis of experiments in the
animal and medical sciences.
V o lume
2.
Iowa State
University P r e s s , Ames.
Gleason, H. A . , and C r o n q u i s t , A.
1963.
Manual of vascular
plants
of Northeastern United
States
and
adjacent
C a n a d a . P. Van Nostrand, N e w York.
Graham, S. A . , Harrison, R. P . , Jr., and W e s t e l l , C. E . , Jr.
1963.
Aspens: phoenix trees of the Great Lakes region.
University of Michigan Press, Ann Arbor.
Guldin,
J.
M.,
and
Lorimer,
C.
G.
1985.
Crown
differentiation in even-aged northern hardwood forests
of the Great Lakes region, U.S.A.
Forest Ecology and
Management 10: 65-86.

260
G y s e l , L. W. and A r e n d , J. L.
1953.
Oak sites in southern
Michigan:
their
classification
and
evaluation.
Michi g a n
Agricultural
Experiment
Station
Technical
Bulletin 236.
Harcombe, P. A.
1986.
Stand development in a 130-year-old
spruce-hemlock forest based on age structure and 50
years of mortality data.
Forest Ecology and Management
14: 41-58.
Heiligmann, R. B . , Norland, E. R. , and Hilt, D. E. 1985.
28-year-old reproduction on five cutting practices in
upland oak.
Northern Journal of Applied Forestry 2:
17-22.
Heinen, J. T. and Sharik, T. L. 1990. The influence of
mammalian browsing on tree growth and mortality in the
Pigeon River State
Forest,
Michigan.
The American
M idland Naturalist 123: 202-206.
Heinselman, M. L.
1973.
Fire in the virgin forests of the
Boundary
Waters
Canoe
Area,
Minnesota.
Quaternary
Research 3: 329-382.
H i b b s , D. E.
1983.
Forty years of forest succession
central New England.
Ecology 64: 1394-1401.
Hix,

in

D. M . , and Lorimer, C. G. 1990a. Growth-competition
relationships
in
young
hardwood
stands
on
two
contrasting sites in southern Wisconsin. Forest Science
36: 1032-1049.

________ .
1991.
Early stand development on two former oak
sites of southwestern Wisconsin.
Forest Ecology and
Management 42: 169-193.
Host,

G. E . , Pregitzer, K. S., Ramm, C. W . , Hart, J. B . , and
Cleland, D. T.
1987.
Landform-mediated differences in
successional pathways among upland forest ecosystems in
northwestern Lower Michigan.
Forest Science 33: 445457.

Hutchison. B. S., and Matt, D. R.
1977. The distribution of
solar radiation within a deciduous forest. Ecological
Monographs 47: 185-207.
K a y s , J. S., and C a n h a m , C. D.
1991.
Effects of time and
frequency of cutting on hardwood root reserves and
sprout growth.
Forest Science 37: 524-539.
K e l t y , M. J.
1986.
Development patterns in two
hardwood
stands
in
southern New
England.
Journal of Forest Research 16: 885-891.

hemlockCanadian

261

Kilburn, P. D.
1957.
Historical d e velopment and structure
of the aspen, jack pine, and oak vegetation types on
sandy
soils in
northern
lower
Michigan.
Ph.D.
dissertation,
University of Michigan, Ann A r b o r .

________. 1960.
Effects of logging and fire on xerophytic
forests in northern Michigan.
Bulletin of the Torrey
B o t a n i c a l . Club 87: 402-405.
Larsen, J. A.
1953.
A study of an invasion by red maple of
an
oak woods
in southern Wisconsin.
The American
Midland Naturalist 49: 908-914.
Lorimer, C. G.
1980.
Age structure and disturbance history
of a southern Appalachian virgin forest. Ecology 61:
1169-1184.
________. 1984.
Development of the red ma p l e understory in
northeastern oak forests.
Forest Science 30: 3-22.
N o w a c k i , G. J. , Abrams, M. D. , and Lorimer, C. G. 1990.
Composition, structure, and historical development of
northern red oak stands along an edaphic gradient in
north-central Wisconsin.
Forest Science 36: 276-292.
Oliver, C. D.
1978.
The development of northern red oak in
mixed stands in central New England. Yale University
School of Forestry and Environmental Studies Bulletin,
91.
Yale University, New Haven, Connecticut.
________.
1981.
Forest development
in North America
following
major
disturbance.
Forest
Ecology
and
Management 3: 153-168.
Oliver, C. D. and Stephens, E. P.
1977.
Reconstruction of
a mixed species forest in central N e w England.
Ecology
58: 562-572.
Oliver, C. D . , A d a m s , A. B. , and Z a s o s k i , R. J.
1985.
Disturbance
patterns
and
forest
development
in
a
recently deglaciated valley in the northwestern Cascade
Range of Washington, U.S.A.
Canadian Journal of Forest
Research 15: 221-232.
Oliver, C. D. , Clatterbuck, W. K. , and B u r k h a r d t , E. C.
1990. Spacing and stratification patterns of cherrybark
oak and American sycamore in mixed, even-aged stands in
the southeastern United States.
Forest Ecology and
Management 31: 67-79.
Oliver, C. D. and Larson,
McGraw Hill.

B. C. 1990.

Forest stand dynamics.

262

Padley, E. A.
1989.
Associations among glacial landforms,
soils, and vegetation in northeastern Lower Michigan.
Ph.D.
dissertation,
M i c higan State University,
East
Lansing.
Pastor, J . , Aber, J. D . , and M c C l a u g h e r t y , C. A. 1982.
Geology,
soils and vegetation of Blackhawk Island,
Wisconsin.
The American Midland Naturalist 108: 266277.
Peet,

R. K. and
tree death.

Christensen, N. L.
1987.
BioScience 37: 586-595.

Competition

and

Peet,

R. K. and L o u c k s , O. L.
1977.
A gradient analysis of
southern Wisconsin forests.
Ecology 58: 485-499.

Pickett, S. T. A.
1989.
Space-for-time substitution as an
alternative to long-term studies.
In Long-term studies
in ecology: approaches and a l t e r n a t i v e s .
Edited by G.
L i k e n s . Springer, New York.
pp. 110-156.
P r e g i t z e r . K. S. and B a r n e s , B. V.
1984. Classification and
comparison of upland hardwood and conifer ecosystmes of
the
Cyrus
H.
M cCormick
Experimental
Forest,
upper
Michigan.
Canadian Journal of Forest Research 14: 362375.
Pyne,

S. J . 1982.
Fire in America.
Press, Princeton New Jersey.

Princeton

University

R o b e r t s , M. R. and Richardson, C. J.
1985.
Forty-one years
of p o p u lation change and community succession in aspen
forests on four soil types in northern Lower Michigan,
U.S.A.
Canadian Journal of Botany 63: 1641-1651.
Rudolph,
V. J.
and Lemmien,
W.
A.
1976.
Silvicultural
cuttings in an oak-hickory stand in Michigan: 21-year
results.
In Proceedings of the Central Hardwood Forest
Conference.
Edited by J. S. Fralish, G. T. Weaver and
R.
C. S c h l e s i n g e r .
Southern
Illinois
University,
Carbondale, Illinois,
pp. 431-453.
Sakai, A. K. 1990a.
Sexual reproduction of red
northern
Lower
Michigan.
The
American
Naturalist 123: 309-318.

maple in
Midland

________ .
1990b.
Sex ratios of red maple (Acer r u b r um)
populations in northern Lower Michigan.
Ecology 71:
571-580.
S a k a i , A.
K . , R o b e r t s , M.
R.
and Jolls,
C.
L.
1985.
Successional
changes
in a mature
aspen
forest
in
n o r thern
Lower
Michigan:
1974-1981.
The
American
M i d l a n d Naturalist 113: 271-282.

263
Scheiner,
S.
M . , Sharik,
T.
L . , Roberts,
M.
r .,
and
VandeKopple, R.
1988.
Tree density and inodes of tree
recruitment in a Michigan pine-hardwood forest after
clear-cutting and burning.
Canadian Field Naturalist
102: 634-638.
Sharik,
T.
L . , Ford,
R.
H . , and
Davis,
M.
L.
1989.
Repeatability of invasion of eastern white pine on dry
sites in northern Lower Michigan.
The A m e rican Midland
Naturalist 122: 133-141.
Smith, D. M.
1986.
The practice
Wiley and Sons, New York.

of

silviculture.

John

Solomon. D. S., and Blum, B. M.
1967.
Stump sprouting of
four northern h a r d w o o d s . U S D A Forest Service Research
Paper N E - 5 9 .
Stubblefield, G. W . , and Oliver, C. D.
1978.
Silvicultural
implications
of
the
reconstruction
of
mixed
alder/conifer stands.
Utilization and management of
red alder, pp. 307-320.
USDA Forest Service General
Technical Report, PNW-70.
Swan,

F. R. , Jr.
1970.
Post-fire response of four plant
communities in south-central New York State. Ecology
51: 1074-1082.

Tarbox, E. E. and Reed, P. M .
1924.
Quality and growth of
white
pine
as
influenced
by
density,
site,
and
associated species.
Harvard Forest Bulletin 7.
Tourney, J . W.
1929.
The vegetation of the forest floor;
light
versus
soil
moisture.
Proceeding
of
the
International Congress of Plant Sciences 1: 575-590.
Tourney, J. W. and K i e n h o l z , R.
1931.
Trenched plots under
forest canopies.
Yale University School of Forestry
Bulletin 30.
Wierman, C. A . , and Oliver, C. D. 1979. Crown stratification
by species in even-aged mixed stands of Douglas-fir western hemlock.
Canadian Journal of Forest Research
9: 1-9.
Whitney, G. G.
1986.
Relation of Michigan's presettlement
pine forests to substrate and disturbance history.
Ecology 67: 1548-1559.
________.
1987.
An ecological history of the Great Lakes
forest of Michigan.
Journal of Ecology 75: 667-684.
Zeide, B.
1987.
Analysis of the 3/2 power
thinning.
Forest Science 33: 517-537.

law

of

s elf­

264

•
1991.
Self-thinning
S cience 37: 517-523.

and

stand

density.

Forest