l" " )
._ ' ’ .
O

my
magaé
Utb‘ 021%.

1 .1
~~ 10". 4
,- twin

. 0:1 9.7 {66" ~.

9?."

ABSTRACT

HERBACEOUS COMPOSITION AND PRODUCTIVITY
AS INFLUENCED BY CANOPY REMOVAL
IN AN OAK-MAPLE STAND

BY

Douglas Nelson McEwen

Changes in the productivity and Species composi-
tion of the herbaceous layer caused by timber harvesting
of an oak-maple stand in southwestern Michigan were in-
vestigated. The stand was divided into 1.1 ha areas.
Four cutting treatments, each replicated four times were
applied: undisturbed control areas, clearcut areas,
group selection areas, and shelterwood areas.

Five 3 m x 3 m plots were randomly established in
each treatment area the year before cutting. After
cutting, four of these plots were expanded to 20 x x 20 m
and five m2 subsamples were taken in each plot. Leaf
area of the 37 most abundant Species was estimated with

the aid of a 0.01 m2

template at each plot. These

surveys were conducted during the mid-summer before

cutting and the early and late summer after cutting.
Thirty two additional m2 plots (8 in each of the

4 cutting treatment areas) were randomly established to

Douglas Nelson McEwen

measure throughfall precipitation. Self-recording
instruments were randomly placed in each treatment area
to measure sunlight, relative humidity, and temperature.
Areas in which trees had been harvested had a
significantly higher level of herbaceous net productivity
as compared to control areas. However there were no
significant differences among the clearcut, shelterwood,
and group-selection harvest techniques. Slash distribu-
tion created very heterogeneous microenvironmental patterns
in areas of harvesting. Throughfall precipitation was not
significantly different among the four cutting treatments,
and net herbaceous productivity was not significantly
correlated with the amount of throughfall precipitation.
Species composition remained unchanged in control
areas. In areas of harvesting, some shift in species
composition did occur. Eight principle new invading

Species were observed. Parthenocissus quinguefolia, the

 

dominant herbaceous Species before harvesting, decreased
in its proportion of total leaf area, but other resident
Species increased their prOportions.

Species diversity increased in harvested areas,
but remained unchanged in control areas. The relation-
ship between diversity and net productivity is not clear.
Plots with very low or very high net productivity had

low diversity, while plots of intermediate net productivity

Douglas Nelson McEwen

had high diversity. In comparison with the herbaceous
layer in other forest ecosystems, the net productivity

of the Russ Forest site is of an intermediate value.

HERBACEOUS COMPOSITION AND PRODUCTIVITY
AS INFLUENCED BY CANOPY REMOVAL

IN AN OAK-MAPLE STAND

BY

Douglas Nelson McEwen

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Forestry

1973

ACKNOWLEDGMENTS

I wish to express my appreciation to all members
of my guidance committee for their help in preparation
of this thesis: Dr. William Cooper, Dr. Peter Murphy,
Dr. George Coulman. Special appreciation goes to my
major professor, Dr. Gary Schneider, for his extensive
support, financial, intellectual, and moral, in the
completion of this project.

My wife Kiva did all of the preliminary typing
and made many helpful editorial comments. As with any
project of this type the support of one's wife is a key

factor and I gratefully acknowledge her help.

ii

LIST OF

LIST OF

Chapter
I.

II.

III.

IV.

V.

TABLE OF CONTENTS

TABLES O O O O O O O O O

FIGURES O O O O O O O 0

INTRODUCTION . . . . . . .
LITERATURE REVIEW . . . . .

Response of Tree Species to
CanOpy Removal . . . . .
Herbaceous Productivity on
Forest Sites . . . . . .
Succession in the Herbaceous
Stratum . . . . . . .
Factors Influencing Herbaceous
Response . . . . . . .

THE STUDY AREA . . . . . .

Site Description . . . . .
Treatment Design . . . . .

FIELD METHODS . . . . . . .

Definition of Herbaceous Layer
Plot Establishment . . . .
Productivity Measurement . .
Measurement of Shifts in
Composition and Structure .

ESTIMATES OF PRODUCTIVITY . . .
Results for Cutting Treatment

and Control Areas . . . .
Seasonal Patterns of Productivity

iii

Page

vii

11

11
12

15
15
15
18
23

24

24

Chapter Page

Effect of Cutting on Productivity . . . . 28
Productivity Differences Among
Cutting Treatments . . . . . . . . 31

VI. THE INFLUENCE OF THROUGHFALL PRECIPITATION
ON NET PRODUCTIVITY . . . . . . . . 34

Introduction . . . . . . . . . . . 34
MethOdS O O O O O O O O O O O O 3 5
Results . . . . . . . . . . . . 36

VII. SHIFTS IN SPECIES COMPOSITION
AND STRUCTURE . . . . . . . . . . 43

Species Dominance and Leaf Area in

Cutting Treatment Areas . . . . . . 43
Species Dominance and Leaf Area

in Control Areas . . . . . . . . . 48
Shifts in Species Composition . . . . . 50
Species Diversity . . . . . . . . . 53

VIII. SUMMARY . . . . . . . . . . . . . 61

LITERATURE CITED 0 O O O O O O O O O O O 66

iv

Table

1.

LIST OF TABLES

Regression equaEions converting leaf
area (cm x 10 ) to dry weight (gm) . . .

Mean weight (gm m'z) of selected Species for
two sampling periods . . . . . . .

Analysis of variance of herbaceous
productivity (leaf area) for all
treatment areas before cutting-~197l . .

Analysis of variance for herbaceous
productivity (leaf area) for all
treatment areas after cutting--1972 . .

Significant differences among treatment
means as determined by Tukey (leaf area
expressed as mZ/mz) . . . . . . . .

Comparison among treatment means for
various environmental parameters at
the .01 level . . . . . . . . . .

Correlation coefficients describing the
relationship between percent open
canOpy, the percent of throughfall
precipitation reaching the forest floor,
and net herbaceous productivity . . .

Mean throughfall precipitation (mm) reaching
the forest floor during the period July 19-
September 14, 1972 . . . . . . . .

A record of summer precipitation (inches)
and percent soil moisture by weight
taken Simultaneously in various cutting
treatment areas . . . . . . . . .

Page

22

26

29

30

30

32

37

38

41

Table Page

10. Textural analysis and moisture content
(% by weight) at several tension
‘levels for a composite soil sample
of all plots . . . . . . . . . . . 42

11. Changes in mean leaf area (percent of
l m 2) for each Species in cutting
treatment areas between the years
1971 and 1972 . . . . . . . . . . 44

12. Changes in mean leaf area (percent of l
l m 2) for each Species in control
areas between the years 1971 and 1972 . . 49

13. Changes in evenness and three measures
of diversity; number of Species,
Shannon-Weiner function, and
Simpson's index, in relation to
changes in net productivity . . . . . . 57

14. Comparisons of above ground net
herbaceous productivity (gm m 2
yr'l) between Russ Forest and
other ecosystems. All data concerns
only the herbaceous strata . . . . . . 62

vi

Figure

1.

2.

LIST OF FIGURES

Arrangement of plots in the Fred Russ

Forest hardwood study area . .

Determination of plot number in one
block by graphical control method

Portable m2 sampling frame with .01 m2

and .02 m2 templates used in
estimating leaf area . . . .

Number of quadrants occupied by species

in cutting treatment areas during

1971 (black line) and 1972 (dash line)

Number of quadrants occupied by various

Species in control areas during
1971 (black lines) and 1972 (dash
lines). . . . . . . . .

Variation in diversity, as measured by

the Shannon-Weiner function and
Simpson's index, with increasing
above ground net productivity .

vii

Page

14

17

19

54

56

59

CHAPTER I

INTRODUCTION

The study of the herbaceous layer's reSponse to
overstory removal in eastern deciduous forests has been
largely neglected. While vegetative reSponse to various
cutting regimes has been examined, most studies have
ignored the herbaceous species and concentrated on tree
reproduction (Metzger and Tubb 1967, Rudolph and Lamine,
Minckler 1965, Ray 1932, Boivin 1971, Wendel and Trimble
1968, 1971, Church 1960). The usual procedure is to wait
three to seven years and then tabulate the number of young
trees that have become established since the cutting. The
first few years after cutting are often described as being
dominated by weedy invaders that are detrimental to tree
reproduction.

The response of the herbaceous layer to overstory
removal needs to be examined for several reasons. Marks
and Bormann (1972) pointed out that the rapid growth of
early successional Species is important in preventing loss
of nutrients from the site of disturbance. From a forest

management point of view, the herbaceous response is

important since these Species are competitors with the
newly established tree Species. Also, many of the herba-
ceous Species growing in these forest openings have food
value for wildlife (McCaffey and Creed 1968) and docu-
mentation of their abundance and distribution provides
input for a wildlife management proqram.

The objective of this study is to document changes
occurring in the herbaceous layer after varying degrees
of canOpy removal. There were two main hypotheses con-
cerning changes in the herbaceous layer: (1) that there
would be a Significant increase in herbaceous productivity
and a shift in species composition; (2) that environmental
parameters of sunlight, relative humidity, temperature,
and the amount of throughfall precipitation reaching the

forest floor would be significantly changed.

CHAPTER II

LITERATURE REVIEW

Res onse of Tree Species to
CanoEy RefiEvaI ._'

The effects of canOpy removal on tree reproduction
has been well documented. However, the reSponse of herba-
ceous Species has been neglected. This response is
usually recorded as Shifts in density and composition of
Species. For example, Tubbs (1968) found that the amount

of canOpy removal in Acer saccharum stands in Michigan was

 

positively correlated with the number of seedlings. About
one half of these seedlings germinated after cutting. On
the other hand, Church (1960), studying a hardwood stand

in upper Michigan, found that Acer saccharum seedlings

 

develOped in the same numbers regardless of what quantity
of canOpy was removed.

Most authors report a Shift in tree Species com-
position after cutting. Winget (1968) observed that
clearcut and partially cut stands in Quebec were dominated

by Acer saccharum, Fagus grandifolia, Abies balsamea,

 

while Prunus serotina, Betula alleghaniensis, Tilia

 

americana, and Quercus rubra were eliminated. Leak and
Wilson (1958) found that for old growth hardwoods in New
Hampshire, selection cutting favored Shade tolerant Acer

saccharum; patch cuttings favored intermediate tolerant

 

Acer rubra; and clearcutting favored intolerant POpulus

tremuloides. Trimble and Hart (1961) also found that the

 

species composition measured five years after cutting was
correlated with the amount of canOpy removal. The excep-
tion to the above studies was the one by Tubbs (1968) in
which he found that various amounts of canOpy removal in

an Acer saccharum stand did not Significantly affect the

 

tree Species composition, Acer saccharum being predominant.

 

The original stand was also dominated by Acer saccharum,

 

and these results illustrate that the amount of change in
Species composition depends on the composition of the
original stand.

Stump sprouts and advance reproduction (seedlings
established before cutting) constitute a large percentage
of the reproduction seen after cutting. Wendel and
Trimble (1968) found 53 percent of the reproduction to be
of sprout origin. Trees derived from this type of repro—
duction frequently outgrows seedlings and may dominate the
stand for many years (Johnson 1971, Winget 1968). Trimble
and Hart (1961) found that sprouting was positively corre-

1ated with the amount of canopy removal.

Herbaceous Productivity 22

Productivity estimates of the herbaceous layer
on forest sites have only recently been made. Early
studies of the herbaceous layer were concerned with such
structural parameters as density, composition, pattern,
and frequency distributions under different amounts of
canOpy coverage (Gysel 1951, McIntosh 1962, Smith and
Cottam 1957, Struik and Curtis 1962, Pace and Hurd 1957,
Sanders 1969).

Much of the early work of estimating herbaceous
productivity on forest Sites was done by R. H. Whittaker
(1966, 1968). Recently Siccama, Bormann and Likens (1970)
reported on another important productivity study on the
Hubbard Brook Experimental Forest. These authors esti-
mated herbaceous productivity by destructive sampling
techniques. Herbaceous material from field plots was
clipped, dried, weighed, and converted into productivity
estimates per unit area.

The level of herbaceous productivity varies within
a forest site. Usually a negative correlation exists
between percent canOpy coverage and productivity (Siccama,
Bormann, and Likens 1970, Hitherington 1969, Gainer et al.
1954). A study by Anderson et al. 1969 attempted to
correlate productivity to light and soil moisture condi-

tions as controlled by canOpy coverage. Soil moisture was

found to have a greater influence on herbaceous production
than light, and was highly dependent on canOpy coverage.

Direct harvesting is a common method of estimating
plant dry weight production. Such destructive sampling
techniques for estimating productivity are both time
consuming and unattractive since repeated measurements
cannot be made on the same plots. A nondestructive tech-
nique such as recording plant volume or leaf area, with a
subsequent conversion of these parameters into dry weight
by the use of a regression equation, would be a better
alternative if sufficient accuracy could be attained.

This type of sampling technique has been applied for years
by foresters in estimating merchantable timber, and lately
by Whittaker (1966) and Whittaker and Woodwell (1969) to
estimate total productivity of standing timber.

A number of workers have estimated dry production
in forest understories by use of the harvest techniques
(U.S. Forest Service, 1958). There have also been a
number of indirect sampling techniques applied to estimate
the coverage in forest understories (Tayle 1959, 0.8.
Forest Service 1958, Daubenmire 1959). However, only a
few studies (Cristofoline 1970 and Anderson 22 31. 1969,
Whittaker 1966, Siccama, Bormann, and LikenS 1970,
Whittaker and Woodwell 1968) were found that applied an
indirect sampling technique to estimate the dry weight

production by the herbaceous understory layer in a forest

community. Anderson 35 El. (1969) measured leaf width
and converted these parameters into dry weight via a
regression equation. Cristofoline (1970) visually esti-
mated leaf area on 0.1m2 plots and converted this para-
meter into dry weight in the same manner. The remaining
authors converted percent plant cover into dry weight by
the same regression technique. All the above studies
found strong relationships (r .74 to .96) between the

independent parameter and dry weight.

Succession in the Herbaceous Stratum

 

 

Studies documenting the course of herbaceous
develoPment after overstory removal have been well docu-
mented in the Pacific Northwest (Isaac 1940, Kienholz
1929, Mueller-Dombois 1965, Yerkes 1960, Steen 1966).
Similar studies are reported for eastern deciduous forest
communities (Gysel 1951, McCaffery and Creed 1969,
McIntosh 1957, Rogers 1959, Kittredge 1934).

The above studies have described successional
stages based on reconnaissance data. The investigators
inventoried a number of Sites ranging in age from recently
logged to the oldest available rather than following
vegetative changes on the same Site over a period of years.
In attempting to reconstruct successional stages from data
on a wide range of sites, they were able to describe vege-

tative changes only in broad terms. This is because

vegetation on disturbed Sites is influenced not only by
age of disturbance, but also by differences in soil,
microclimate, and severity of the disturbances.

No studies were found for the eastern deciduous
forest similar to Dryness' (1973) in which vegetative
develOpment after logging in the Pacific Northwest was
followed for seven years on permanent plots. Much of the
data collected on old field succession (Odum 1960, Evans
and Cain 1952, Evans and Dahl 1955) is probably non-
applicable. Old field soils have been greatly altered
by cultivation (Keever 1950, Booth 1941). Also, slash
left on forest Sites after cutting greatly alters the
microclimate, resulting in a mosaic of different
microsites.

Usually there is a great increase in herbaceous
productivity the first few years after overstory removal,
but as the remaining tree species begin to form a new
canopy, herbaceous productivity declines. Dryness (1973)
found average herbaceous coverage to increase from 26
percent to 45 percent by the fourth year after cutting.
Studies like those of Ahlgren and Ahlgren (1960) and Lyon
(1971) noted large increases of herbaceous coverage after
forest fires. In some cases after intensive Site prepara-
tions, productivity is greatly curtailed and re-population
by herb and shrub species is delayed (Schultz, personal
communication).1

1Robert Schultz, U.S.F.S., Olusee, Florida.

Factors Influencinngerbaceous

ResEonse

 

Removal of the canOpy undoubtedly changes a
plethora of biotic and abiotic factors to which the
herbaceous layer reSpondS. Few of these factors seem to
have been studied directly in forest Sites. Whipple
(1968) found higher soil temperatures after clearcutting,
and Larson (1970) noted that growth in Quercus £2253
seedlings responds to total daily degree hours rather
than to the differential between day and night tempera-
ture. After cutting, soil moisture decreases and has a
higher rate of fluctuation than before cutting (Dixon
1969, Anderson at 31. 1970, Whipple 1968).

After a clearcutting, slash will often cover 21-26
percent of the ground surface (Wendel and Trimble 1968).
Microsites in and around this slash will undoubtedly have
different environmental parameters than more exposed
microsites. However, no references were found that dis-
cussed microclimate conditions.

There is some evidence that annual weed Species
have an inherently higher growth rate than tree Species,
and given Optimum conditions, weeds have a higher net
productivity. Coombe and Hadfield (1966) found dry matter
production in fast growing woody species from secondary
trOpical rain forests to be Significantly lower than

herbaceous Species. Growth and leaf area are highly

10

correlated, and the percent distribution of new growth to
leaves is important (Newhouse 1968). Jarvis and Jarvis
(1964) took published values of maximum assimilation rates
for temperate Species and found that woody plants lie in

the range of 20-50 gm m"2 wk-l, while herbaceous plants

lie in the range of 70-150 gm m-2 wk-l. They speculate
the existence of some inherent corollary of wood formation
which is incompatible with a high rate of net assimilation.
Unfortunately none of the above authors considered the
rapid growth of stump and root Sprouts.

The role of allelochemics in regulating the
herbaceous response is probably important, but at this
point little is known of this subject. Allelochemic
relationships among the various competing Species have
been studied, but unfortunately none of these Species

occurred in this study (Whittaker and Feeny 1971, Kohno

1969, del Moral and Muller 1970, Tubbs 1970).

CHAPTER III
THE STUDY AREA
Site Description

The study Site is a 17.6 ha stand of oak and mixed
hardwoods in the Fred Russ Experimental Forest, a property
of Michigan State University. The study area is part of
a 235 ha forest located in southwestern Michigan, Cass
County, near Dowagiac.

Prior to 1930, the stand had been used for grazing
cattle and a source of firewood, but during the last 40
years, it has remained essentially undisturbed. As a re-
sult of grazing and lack of forest management, stand
structure was primarily of very large and very small

2 ha-l. Major species

trees. Basal area averaged 31.2 m
in the stand were Age; saccharum, Quercus £2252, Quercus
31b3, Quercus velutina, and Prunus serotina.

The stand is generally level to gently undulating
in topography. One minor depression on the border of

the stand leads to a swamp located outside the stand. Two

major soil series are present in the area: the Kalamazoo

11

12

and Oshtemo. These are quite Similar in soil properties,
being well—drained typic hapludalf soils which have
develOped on gravelly loam, sandy loam, and loamy sand
material. The Oshtemo soils differ from the Kalamazoo
soils primarily in the thickness and the amount of clay

in the B2t horizon. This horizon has a thickness of 20-31
cm in the Oshtemo and 25-56 cm in the Kalamazoo. The
Kalamazoo series has the higher percentage of clay.

Annual precipitation over a 30 year period averages
1092 mm of which 16 mm is snow. Rainfall during the
growing season of May 1 to September 15 is about 417 mm.
During the two growing seasons of the study, rainfall was
317 mm (1971) and 376 mm (1972).

Spring vegetative growth begins between April 20
and May 1. However, trees are not in full leaf until
about May 15, and much of the herbaceous growth does not
appear before May 30. Growth measurements were arbi-
trarily terminated on September 15. The growing season
does continue to the beginning of October, but few Species
show any gains of net productivity during these last few

weeks.

Treatment Design

 

The present study was conducted within the
framework of a larger experimental improvement cutting

initiated by the Department of Forestry in 1970. The

13

stand was surveyed and divided into 16 rectangular treat-
ment areas, each 1.1 ha in Size (Fig. 1). There were
four types of treatment areas: control, in which the
trees were left undisturbed; clearcut, in which all trees
were removed; Shelterwood, in which 20 to 25 large trees
(6.2 m2 of basal area) were left undisturbed; and group
selection, in which all trees were removed within the
small areas (0.25 ha) indicated in Figure 1. These four
types of treatment areas were arranged in a randomized
block design, comprised of four blocks, four treatments
per block.

The need for blocking was based on observed dif-
ferences in the density, canopy coverage, and species
composition of tree Species within the whole stand.
Herbaceous cover followed a Similar pattern, being high
in areas of low tree canOpy coverage and low in areas of
high tree canOpy coverage. Thus the same randomized
block design was used in studying the herbaceous stratum.

The stand was inventoried during the Spring and
summer of 1971. A11 trees larger than 1.27 cm diameter
breast height (d.b.h.) were catalogued according to
species, d.b.h., merchantable height, and form class.
Cutting operations were carried out by a commercial logger

from October to December, 1971.

14

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.

 

 

 

hardwood study area.

 

u

u

n 0 Mixed Fina:

:: snmenwooo CLEAR 0 0 WWW“ "-0"

n 0 CUT O a” “‘- ,_ SOIL ma: aounoanv

:: .3 O 0...: n-a,m. Rzmcmou aaLocx menu'r.

ll .- ==== wwsmm

3 ' 0 eye A

a 00 o 135 270 405 540 @

I .
/ ‘1: wgyo GROUP _ £_ f-eld scat/w rttr I
.1: 1 CONTROL 0 cut ‘\ W/
|_ O O __ \ L

h' 0 cu o 0 0 comm cur

" .1 4-I o 4-2

:' , __—/ o o o

—r': 6'8“”, OCLEAR O 3." ‘

n CUT cut 0 GROUP 0 OSHELEEGWOOO O /

I. 0 CUE 2 I

:: saw 20" 2-2 ‘ at? 4-3 0 o 4-4 0 o l

O I
u 0 - ‘ .
n Ljfimiw ’1”, onson
‘ I) arm. 0 O f

a 0 New 0 0‘ T CLE¢R 00%

n suzmnwooo M 03_2 cu. Red Pine ,0 /

u o cur cormm 34 Plantation ,/

II 0 o ’1

2-3 _

n O 2 ‘ O O 0 [’0

:1 o O o O I,”

i: o ocomnm. 0 0,2 ’0’,

- Stan I
a 3 4 O ,” lav/and
" 0’ INN-vow:
/
u [I
\\ I
\\ ’l
\\ //

"’r"‘:LL——h—“"’—__753757?:VEEEIUEIWT“-——-—Q____________________
I E] roars! no.

Arrangement of plots in the Fred Russ forest

CHAPTER IV
FI ELD METHODS

Definition of Herbaceous Layer

 

 

The herbaceous layer was defined as all annual-
biannual Species plus all perennial Species less than 30
cm in height. The latter category included tree seedlings,

small Shrubs, and plants with a rhizomatous habit.

Plot Establishment

 

The question of initial sample size posed a prob-
lem as no estimate of sample variance was available. A
species area curve Showed that four 3 m x 3 m plots per
block would be adequate. However, forest understory
Species are known to be distributed in a patchy type of
pattern (Smith and Cottam 1967, McIntosh 1962) and a
visual survey Showed this to be the case on the study site.
Thus 20 3 m x 3 m plots per block, 80 plots in total
were used. To insure some uniformity in the arrangement
of sample plots, one was randomly located in each quadrant

of each treatment area, and one was located randomly in

15

16

respect to the whole treatment area (see Fig. 1). For
statistical purposes, plots needed to be randomly located
with an equal number in each treatment area.

Following the 1971 herbaceous layer survey, the
data was analyzed for purposes of determining the in
adequacies of the sample Size. The graphical control
method proposed by Grieg-Smith (1964) was applied for this
(Fig. 2). The abscissa indicates progressive increases of
plot numbers and the ordinate indicates mean leaf area or
standard error expressed as a percentage of the mean. At
first, the ordinate fluctuates widely, but later stabilizes
as the number of samples is increased. Although this
graphical technique is strictly empirical and only of sug-
gestive value, it indicates that a sample Size of 20 plots
is a reasonable number for one block.

After cutting, microsite conditions became very
heteroqeneous due to slash deposits. To increase sample
accuracy of productivity estimates, 64 of the original
plots were expanded from 3 m x 3 m to 20 m x 20 m and
then subsampled. A one meter square sample was taken at
the original plot Site, and four other meter square
samples were randomly taken in the 20 m x 20 m area sur-
rounding the original plot. Subsamples were then condensed
into a one plot total. The remaining 16 3 m x 3 m plots

were left unchanged. These plots, one in each treatment

17

(ueem go 3)
10119 pzepueqs

.ponume Houucoo Hmowcmmuw an xooan mco cw Hones: uoam mo coaumcflsumumo

om

muon mem mo nonesz

OH

.N musmwm

. ooh

 

0H

cm i

 

uouuo pumdeum

comm mama sum:

11‘

1 com

I omm

v ova

1 com

I own

1 com

a cam

r ova

u owm

1 0mm

T oooa

I ONOH

v oeoa

 

f coca

(am) 9319 JEaI usew

18

area, were used to estimate Shifts in Species composition

and structure.

Productivitereasurement

 

This study employed an indirect, nondestructive
sampling technique involving leaf area measurement and its
subsequent coversion via regression analysis into dry
weight. The most common means of determining leaf area
iS to measure maximum width and length of each leaf, and
convert this to surface area. Such a tedious technique
would be impossible for the number of samples involved in
this study. To provide the needed speed with a reasonable
level of accuracy, a visual method was adOpted similar to
one used by Cristofoline (1970) in a mixed hardwood under-
story community at Oak Ridge, Tennessee. Leaf area was
estimated with the aid of the portable m2 frame and tem-
plates Shown in Figure 3. The borders of a plot were
defined with the m2 frame. Next, either a circular .01 m2
template or a rectangular .02 m2 template was placed on
.the plants to estimate leaf area of each of the major
Species. In cases where leaves were larger than or smaller
than the templates the judgment of the author was impor-
tant. An attempt was made to group small leaves or com-
bine portions of large leaves in order to arrive at an
estimate of total leaf area. The use of templates worked

quite well in the case of multi-layer canOpieS Since the

19

 

2

Figure 3. Portable m2 sampling frame with .01 m and
.02 m2 templates used in estimating leaf
area. Templates were placed on each leaf
of all plants within the plot.

20

templates could be repeatedly moved so as to measure all
leaves in the canopy.

A series of samples was taken on 10 of the major
Species to determine the accuracy of the above technique.
Leaf area, as measured by the templates, was compared to
the value obtained by placing a grid of squared centi-
meters over the foliage and counting the number of squares
occupied. The difference between the leaf area estimated
by these two methods varied from Species to Species.
Large broadleaf Species were easier to measure and had
more accurate estimates of leaf area than small multi—leaf
Species. For example, the percent error (the direct count
minus the template estimate + template estimate) was only

+0.2 percent for PodOphyllum peltatum. For Circaea quad-

 

risucata it was +9.3 percent and for Parthenocissus

 

quinquefolia, it was +1.9 percent. Smaller leaf Species
had higher errors: Erigeron canadensis, +17.7 percent;
Quercus rubra, -16.5 percent; Viola Spp., -18.0 percent.

The small multi-leaf Species of Geranium robertianum had

 

an error of +30.3 percent. Rubus spp. had an error of
-29.1 percent. The greatest errors were found in the

grasses, +122.9 percent, and in Galium circaezans, +161.2

 

percent. The templates badly underestimated the leaf
area of these Species. This is probably due to their

small laminar type foliage which greatly overlaps. Based

21

on these observed errors, corrections were made in the
original field estimates of leaf area.

Table 1 shows the regression equations develOped
for the 29 most abundant Species. This was done by re-
peated harvesting of the above-ground parts of a species
after its leaf area had been recorded. The plant material
was oven-dried at 85° C for 48 hours and dry weight deter-
mined. Leaf area was then regressed upon dry weight to
give the equations. These Species comprise 93 percent of
the total leaf area measured over all plots. Development
of a regression equation for each Species present was not
possible and although 7 percent of the total leaf area is
ignored, the dry weight value calculated on the remaining
leaf area should approximate annual productivity.

For each Species, mean sample weight and standard
error are given. Standard errors were, for the most part,
less than 10 percent of the means. The standard error of
a single measurement taken at mean sample weight is also
shown. This standard error gives the maximum amount of
deviation one would expect to find. In some cases, it is
up to 60 percent of the mean, but this deviation would,
of course, decrease with an increase in sample size.

The coefficient of variation shows the amount of
variation in the dependent variable, weight, that is
included in the regression equation. The coefficient of

variation is analagous to a correlation coefficient and

22

.unoa A.q. nomads-u asuucaasw noosHoaHv

.a sedasauuub unsung IonaHoan

.4 nausouasdo casuwndm nonnaosH .A asses. unusuaucsu cusaocu

N H

 

 

H5 go. «6.5 .om. . «.5 5o. new. I 05.HI I 5 III: ..uu:z. mavens. Innuuunnm ma
«H Ho. mm. .65. c o.~ m5. unv. + vu. I a .q wanna Inca-so am
an no. m5.v .No.Hc ov.ns no. xmo. + 05. I I 5 .q usaasuuonoa laden-o 55
on so. m~.~ .om. . «5.5 mm. xvo. + oc.HI I 5 .uqu ..a. Ic-«eamu«> qupoe om
.>¢m vs- .nusauh
am so. ea." .va. 3 oa.oa am. now. + nu.~ I 5 1.a«uqxc saunas-aueasu II-ouau m5
orgasm ..Hnazc
5H Ho. mm." .mm. . «m.n we. now. + 55. I I a Ina-anodaueao asuaamosaom vn
4H Ho. am.” 1mm. . Hm.n no. x5o. + am. I I a .a Iago-unease” «sauna nu
ca 30. so.” .vn. c vm.~ o5. xnm. + ov. I I a .na- quoa> «a
.858
ma no. mo.« 15m. c ve.aa ma. xmn. + me. I 5 ..qc «Adan-sucqav n=.-wooconuuam an
scam can
a Ho. ~m.~ 1m5. c o5.o ma. xmo. + ow. I 5 .uaoxuqx aumaunuaoaaa uses-mam 05
as so. oa.~ .ov. . 5m." we. xoe. + nv. I I 5 .a unease.» gags me
55 Ho. n5.5 .mH.~. 5m.HA «5. x65.” I m¢.HI I 5 .ma. .Ion a”
Iguaau .n .0
ma do. on.“ .m5. 1 av.“ mm. x5~. + '5. I » 1.xauazc Iaeoasnao Iuaeuuoauo 5H
as do. o~.~ .m~. c o~.m no. xnm. + oo.~I I 5 .an- nuns: as
ma #0. an.” 1mm. . 5a.. 59. xoo. + oo.AI I » .n-uI: nan-gun‘s uaoa as
as so. on.“ .55. . am.a we. xv«.« + mo.nI I 5 Anyone uneven-c Iaooaom ca
am so. no.5 low. a «.5 5m. goo. + o5.HI I a .q neauoau uaeuoo as
as Ho. on.« 1H5. . o.~a om. non.” + on. I I 5 .uuoa ..qc noouoaooasm Isa-co: «a
55 so. «a. Iva. . «.n «o. x54. + mu. I » n.sunu Icauouog usnsua as
ea Ho. on." .o5. . n.oa ma. ~54.“ I no.5I I 5 .a Iguana. uaauxo as
v" Ho. 5.. 1 a. . o.a Hm. unm. + o5. I I u .q saunas nooa a
Isa-do .n .u
a mo. om.n .mu.dc m.oa me. new.” + ac. I I 5 1.xauqzc Icauocuuqu Insane a
#5 do. we.” 15m. . v.5H em. xmw. + on. I a .98. Inns: 5
vs so. HH.H 155. . m.5 om. xom. + oH.~ I » «.uoun nanacuanuo 15.0 o
«H «o. mm. .on. . o.n o5. xav. + «a. I a .4 and. Inna-ac m
05 Ho. n~.w~ .mo.mc o.~v m5. xmm.a + 5o.n I 5 .q neuuauuau couaaousnm v
«a Ho. mm.“ x v.o. va.v o5. xov.a + 5o.HI I » .aau asauuwo n
ma so. 5o.H« 165.5. o.ma vo. xom.n + vm.o I » H.q Iaacquaauo couoowuu a
ma do. av.“ 15M. . o.n an. xo5. + no. I u a nan-uses aaaassaauom a
onau sowudsvo usage: nuouuo canvas»: and sawudau¢> uo Amoa a so. can.
oamacm caduceuucu oanuuu sacs an anode) Ignace can: ucoaoauuooo «sea-x .Asu..um3I>v
onu no acoaouandoa caved. scuuanul s0«-ouua¢
ooaaoauacuam uo uoauo oaavcnum

 

.Asuv anodes mum on fine” x no. can: used ucwuuo>soo acowudsvo scauuouoca .a can-a

23

Shows the strength of association between leaf area and
dry weight. Most coefficient of variation values are
high and only equations 1, 6, and 17 had low values.

During the period of June lZ-July 7, 32 of the 20
m x 20 m plots were sampled, two in each treatment area.
The remaining 32 plots (20 m x 20 m) were sampled during
the period of September S-September 24.

Measurement of Shifts in_
Composition and Structure

 

 

 

Estimates of leaf area for each species were made
on the 16 3 m x 3 m permanent plots using the methods
developed earlier for estimating herbaceous productivity.

Measurements were taken the year before and after
cutting, but no attempt was made to record all Species
present. The year before cutting, the 29 most abundant
Species, as determined by frequency of occupied quadrants
were tabulated. The year after cutting, 37 species were
tabulated. Due to the difficulty of Species identifica-
tion, members of the family Poaceae were recorded as one

multi-Species group. Geum canadensis and Sanicula

 

 

canadensis also were considered as one species group as

 

was Polygonatum canaliculatum and Smilacina racemosa.

 

CHAPTER V

ESTIMATES OF PRODUCTIVITY

Results for Cutting Treatment
and ControI Areas

 

Net productivity is the amount of biomass produced,
and is usually estimated by the standing crop of biomass
that has accumulated over a growing season (Westlake 1963).
The general formula for net productivity given by Newbould
(1967) is Pn= W max.- W min.+S+G+D, where W equals biomass,
5 equals the accumulation of reserves, G equals the amount
consumed by animals, and D equals the amount lost to
mortality during an interval when standing crOp is in-
creasing to its peak level (W min to W max).

The factor 8 was ignored as was increases in root
biomass Since this study estimated only above-ground pro-
duction. Large accumulations of root biomass or reserves
could influence above-ground production but this factor was
not considered here. The amount of net primary productivity
of herbaceous Species consumed by herbivores (G) is thought
to be minor. Golley (1960) reported that Microtus consume
.3 percent of net primary production while Wiegert and
Evans (1967) reported that the combined consumption of all

herbivores is approximately 1 percent. The latter authors

24

25

estimate that herbivorous consumption rarely exceeds 5 per-
cent of net primary production and is usually lower. The
factor D is also felt to be minor. This fact is based on
the author's knowledge of the species involved in this
study and daily observations during the growing season.
Ignoring the factors S, G, and D reduces Newbould's expres—
sion to Pn= W max.- W min. Assuming W min. equals zero for
the above-ground parts of herbaceous Species, all that is
needed to estimate above—ground productivity is the value
of peak biomass during the growing season.

By applying the regression equations developed in
Table 1, leaf area was converted into dry weight (gm m-Z).
Next, the average dry weight per square meter was calcu-
lated for each individual species surveyed during the first
sampling period (June lZ-July 7). The same was done for
the second sampling period (September 5-September 24), and
Table 2 lists the results. Peak biomass for each Species
was determined by selecting the larger value recorded in
the two sampling periods. These peak biomass values were
in turn summed to give the net productivity per square
meter. More frequent sampling periods would have increased
data accuracy, but the bulk of productivity occurred during
the above two sampling periods.

During the 1972 growing season, net above-ground
productivity of the herbaceous layer averaged 142.6 gm m-2
in the cutting treatment areas where the tree canOpy had

been removed and 34.4 gm m"2 in control areas where no

26

Table 2. Mean weight (gm n-2) of selected species for two sampling periods.

 

Mean weight and
standard error
June 12 «July 7

Mean weight and
standard error
Sept . Séept . 24

Species

 

Increasing Biomass

 

 

*+ Phytolacca americana L. 6.00 (.93) 33.25 (3.75)
8+ Poaceae (species group) 18.71 (5.50) 26.10 (4.50)
*+ Erigeron canadensis L. 2.52 (.72) 12.74 (2.33)
9+ Circaea quadrisulcata (Maxin.)

French. and Sav. 3.10 (.65) 8.17 (1.26)

+ Geranium robertianum L. 1.90 (.37) 5.95 (.74)
Rubus app. .97 (.57) 3.14 (.58)

+ Cirsium spp. 1.00 (.46) 3.87 (.83)
Quercus rubra L. 1.75 (.46) 3.14 (.53)
Oxalis stricts L. .00 (.00) 3.13 (.73)
Geum canadensis Jaoq. and

Sanicula canadensis L. .36 (.12) 2.79 (.68)
Ulmus spp. 1.13 (.30) 2.99 (.32)
Sasstrass elbidum mutt.) Bees .96 (.31) 1.92 (.54)
Physalis subglabriata

Heckens. and Bush .20 (.08) 1.72 (.60)
Hedeona pulegioides (L.) Pere. .00 (.00) 1.73 (.73)
Cornus florida L. .26 (.10) .78 (.41)
Ace: rubrum L. .09 (.04) .26 (.06)
Polygonatum canaliculatum (Muhl.)

Pursh and Prunus Virginians L. .12 (.06) .38 (.14)

Decreasing Biomass
9+ Parthenocissus quinquefolia
(L.) Planch. 13.20 (1.38) 7.69 (.69)
' Podophyllum peltatum L. 2.88 (.55) .48 (.25)
Viola spp. 1.84 (.28) 1.26 (.16)
* Galium circaezans Michx. 5.58 (1.33) 3.34 (1.27)
Osmorrhisa Claytonia (Michx.)

C. 8. Clark 1.57 (.17) 1.26 (.19)
Quercus elba L. .44 (.13) .22 (.06)
Anelanchier spp. .43 (.11) .20 (.07)
Phryea leptostachya L. .19 (.06) .43 (.16)

Constant Biomass

Ribee spp. 1.23 (.34) 1.76 (.43)
Rhus radicans L. .39 (.14) .45 (.15)
Ace: saccharm Marsh. .42 (.17) .67 (.17)
Tovera Virginians (L.) Ref. .31 (.21) .51 (.14)
Prunus serotina Ehrh.

Ina Prunus Virginians L. .73 (.11) .52 (.10)

Total 68.30 130.84

 

'These species oospriee 764 of the standing

+These species ooeprise 754 of the standing crop Sept. S-Sept. 24.

crop June lZ-July 7.

27

canopy was removed. Both these values are probably under-
estimates of the true values Since 7 percent of the leaf
area was not converted to a dry weight basis (see page 21).
By assuming that the species not accounted for in the re-
gression equations had a leaf area-dry weight equal to the
average leaf area-dry weight ratio for all other Species,
the original productivity values could be inflated by 7
percent. This would give an average of 152.6 gm m-2 in
treatment areas and 36.8 gm m-2 in control areas. These
values, although dependent on the above assumption, probabhr

give a more accurate estimate of above-ground productivity.

Seasonal Patterns of Productivity

 

 

Using means and standard errors, species in Table 2
were grouped according to the period when maximum biomass
was observed, i.e., early or late summer. It can be seen
that these Species can be categorized into three different
groups. In the first group, biomass is continually in-
creasing throughout the summer, reachingits maximum in
the fall. The second group attains maximum biomass by
early summer, and then Slowly decreases. Species in the
third group maintained a more or less constant biomass
throughout the summer period.

Cristofolini (1970) found groups of species that
had growth patterns Similar to those above. He conducted
an additional survey in late Spring and consequently found

a group of Species, spring ephemerals, that had peak

28

biomass during Spring and then decreased throughout the
summer.

Both Cristofolini's (1970) study and this study
emphasize the important point that in a natural assemblage
of plants, a variety of growth patterns is present. This
fact has great bearing on sampling techniques. In order
to achieve Optimum accuracy, each Species group must be
sampled during the period of its peak biomass. Compare,
in Table 2, the dry weight for all Species in the first
sample period (68.3 gm m-z) with the second sample period
(130.8 gm m'z) to that of the peak dry weight (142.6 gm
m-z) determined by summing the maximum values found for
each species. The error inherent in a Single sampling
period is easily seen.

Cristofolini (1970) suggests four sampling periods:
late winter, midspring, early summer and late summer. In
this study, samples were not taken during late winter and
midspring. However, based on the phenology of species
encountered in this study, the bulk of productivity occurs

during the summer sampling periods.

Effect of Cutting gn_Productivity

 

To determine if cutting treatments significantly
affected herbaceous productivity, leaf area data from
surveys of 1971 and 1972 was subjected to analysis of
variance. Leaf area rather than dry weight was chosen to

represent productivity because dry weight data ignored 7

29

percent of the total leaf area. Table 3 Shows the analysis
for the 1971 survey. Treatment areas had been laid out
(Fig. 1) before the survey was taken, but no cutting had
been done. That no significant differences existed among
treatment areas indicates that the herbaceous productivity
was essentially the same in all areas before cutting. A
Significant difference between blocks implies that blocking

was done correctly.

Table 3. Analysis of variance of herbaceous productivity
(leaf area) for all treatment areas before

 

 

Degrees of Mean Level of
Source freedom square F significance
Treatment 3 110485.245 1.0559 .373
Block 3 3282649.846 31.3714 .0005
Error 73 104638.474

 

Table 4 Shows the analysis of variance for the
1972 survey. As expected, cutting treatments significantly
increased herbaceous productivity and treatment means (leaf
area. m-z) are shown in Table 5. These means were compared
using Tukey's method, and none of the cutting treatments
differed significantly from control areas. For 1971 and
1972; average leaf areas (m’z) were 413, 445; 451, 665;
456, 586; 462, 601; for control, clearcut, shelterwood

and group treatments, respectively.

30

Table 4. Analysis of variance for herbaceous productivity
(leaf area) for all treatment areas after
cutting--l972).

 

 

Degrees of Mean Level of

Source freedom square F significance
Treatment 3 137937.307 3.368 .05
Block 3 205677.099 4.534 .01
Sample

period 1 264507.797 6.459 .01
Treatment x

time 3 176310.923 4.305 .01
Treatment x

block 9 57431.918 1.250 .30
Error 44 38456.970

additivity l 781959.000 18.814 .0005

residual 43 7331.661

 

Table 5. Significant differences among treatment means as
determined by Tukey (leaf area expressed as

 

 

mz/mz) .
Clear Level of
Group Shelterwood cut Control significance
1.330 1.204 1.1731 .890 .01 level

 

 

1All figures not Significantly different are

connected by the underline.

31

Analysis of variance of the 1972 data provided an
opportunity to test the effect of sampling at two time
periods. As Shown in Table 4, both time and the time x
treatment interaction were Significant sources of varia-
tion. This analysis of variance confirms the earlier
results Shown in Table 2 that the amount of standing crop
varies greatly throughout the growing season.

It is notable that the F statistic for block in
Table 4 is much lower than the same statistic in Table 3.
Before cutting there were great differences of herbaceous
productivity between blocks; block 2 being high and block 4
being very low. Cutting treatments eliminated much of this
difference, all blocks being maintained at a high level
of productivity. This effect is reflected in the lower F
statistic seen in Table 4.

Productivity Differences Among
CuttinggTreatmentS

 

Why the various cutting treatment means did not
differ among themselves is an interesting question. A
partial explanation is offered by looking at the environ-
mental data in Table 6. The parameters shown were moni-
tored throughout the summer using self-recording pyrohelio-
meters and hygrothermographs (Belfort instruments). Three
hygrothermographs and four pyroheliometers were randomly
placed in a control area. An equal number were placed in

one treatment area. Each week the instruments in the

32

mosflauooss Sn announces mus ucmuomuwo haucsowmwcmwm uoc asses Ads

N

unease oncogene an vowssmsooos asses

H

 

1mm.mec «.mm

eoozumuamrm

AN.¢NV v.5v

coauooaom macho

Amh.HNV ¢.mm

usousoau

Amm.mmv H.bh

Houucou

 

 

Abm.mv h.mm

Am5.vv m.vm

Avm.mv m.~m

ANO.NHV v.0m

“H0.HHV v.¢m

 

.mm.mv m.am

AmH.bV b.Nm

Amm.ov oo.Hm

 

 

Aav .SHE
massage: .Hum

any .803
5uaeaane .aoa

Auv .xss .man

 

 

amo.mv 0.4 aem.mv H.5H aam.mc H.ma 1mm.mac m.m~ Aaaux.auo .auc arses
Houucoo coauooaom macho ooozuouaonm usousoao usuafiuusm
Ned Hm>oH Ho.
0:» us muouossusm Houcoficoufl>co msoaus> Hem asses nauseous» moose conansmsoo .o canoe

33

treatment area were moved to another treatment area. After
three weeks, all instruments were moved to a new block.

By concentrating instruments in one treatment area and
control area, microsite variability could be better
sampled.

There appears to be no clear cut differences
between the environmental factors measured in the cutting
treatments. However, all the environmental parameters,
except maximum relative humidity, did Show Significant
differences between controls and cutting treatments.
Variability was extremely high in the environmental data,
but it appears that the differential effect of the three
cutting treatments on microsite conditions is negligible.

It is highly probable that slash distribution had
the most profound effect on microsite conditions. Slash
appeared to be distributed randomly across all treatment
areas, and this may account for the lack of differences
between cutting treatments in regard to both environmental
parameters and productivity. Clearcut areas had larger
amounts of slash than group-selection and shelt erwood
areas. However, the excessive slash in clearcut areas
was concentrated in a few large piles with the remaining
slash being distributed like that in the group-selection

and Shelterwood areas.

CHAPTER VI

THE INFLUENCE OF THROUGHFALL PRECIPITATION

ON NET PRODUCTIVITY

Introduction

 

A study by Anderson at El. (1969) revealed that
the understory herbaceous cover of a pine forest in
northern Wisconsin was more responsive to differences in
throughfall precipitation than to differences in light.
Both light and throughfall precipitation were regulated
by the amount of canOpy opening. ‘However, light was not
found to be a limiting factor in herbaceous cover even
in the dimmest sites.

Removal of the tree canopy and its capacity to
intercept rain, will increase the amount of throughfall
precipitation. It was hypothesized that results of the
above study could be extended to this study where various
prOportions of the canOpy had been removed. The increased
amount of throughfall precipitation should be Signi-
ficantly associated with increases of herbaceous

productivity.

34

35

Methods

The methods used in this chapter were adapted from
the Anderson study. Eight meter Square plots were assigned
to the four treatment areas, giving a total of 32 plots.
Plots were randomly located in the clearcut area. In the
Shelterwood and group selection areas, plots were located
so each had at least a partial coverage of canOpy. In the
control area, with canopy intact, plots were chosen to
include a wide range of herbaceous cover, from moderately
dense to very sparse.

CanOpy cover was estimated by use of photographs.
Prints covering a canOpy area of about 0.04 he were en-
larged to 18.4 cm square. A dot grid having 10 rows of
10 dots was placed over the photograph. The number of
dots falling on undarkened portions of the photoqraph was
divided by 100 to obtain percentage "open canOpy" for
each plot.

Rainfall reaching the forest floor as throughfall
was measured with four rain gauges located 1 m from the
center of the plot, one at each cardinal direction. The
catchments of all gauges at each point were averaged to
give an estimate of throughfall for each plot for a single
rain period. Rainfall was measured within a few hours
after storms during the daylight, or before 9:00 A.M. for

precipitation having fallen during the night. Gross

36

precipitation measurements were made fron a control gauge
in a large Opening near the plots.

The herbaceous reSponse to changes in throughfall
was measured by estimating net productivity. All above-
ground plant material was clipped from each m2 plot at
the end of the 1972 growing season. The plant material

was oven-dried at 85° C for 48 hours and weighed.

Results

The first set of correlation coefficients in
Table 7 describes the relationship between percent through-
fall precipitation and net herbaceous productivity. All
coefficients were nonsignificant. Table 8 Shows the mean
of throughfall precipitation (mm) reaching the forest
floor during the period of July l9-September 14. Although
the amount of throughfall precipitation is slightly lower
in the control area as compared to cutting areas, no signi-
ficant differences exist between them.

These results indicate that throughfall precipita-
tion had little influence upon the various levels of net
herbaceous productivity. The amount of throughfall pre-
cipitation in the control area was approximately the same
as in cutting areas. However, as stated earlier, herba-
ceous productivity was significantly higher in the latter

areas 0

37

 

 

mm.| Houusou
S . me . 050905
on. poozuouaocm
amosmo ammo ucmouomIIHHsmcosounu unmoumm
m5. NO.I aouusoo
mv. usomsouu
m~.I oooaumuamnm
amosso ammo unmouoquauw>HuosOoum uoz
om. nonhuman
em. -.I aouusoo
mo . 03905
no. @003uouaunm
Hasmnmsounu useouomuuwufi>auosvoum umz
Amomav .Hs um coaumocc tha seam usoausoua
acmfloammooo cofiumaouuoo ucoaowmmooo coaumHouuoO

 

.muw>fiuosooum uncoosnuon
no: use .uoon umouom on» meanness coausuwmflooum Hammnmooucu mo usouuom ecu
.mmocso ammo unmouom smashes manmcoausaou on» mcfinauommp mucoaoammooo cOwusHouuoo .5 «Home

38

Table 8. Mean throughfall precipitation (mm) reaching the
forest floor during the period July 19-
September 14, 1972.

 

Control Shelterwood Groupcut Clearcut

 

Ppt. 47 54 53 55

 

The remaining two sets of correlation coefficients
in Table 7 describe the relationship between percent Open
canOpy, percent throughfall precipitation, and net pro-
ductivity. These coefficients also proved nonsignificant.

Since Anderson 35 31. (1969) found very strong
correlation coefficients compared to the weak coeffi—
cients in this study, a careful re-examination of the
Wisconsin study was done. It showed that several of the
field conditions were similar to those in the present
study area. Light intensities, 7-19 percent of full sun-
light, were the same for both sites. Percent throughfall
precipitation and the amount of herbaceous cover were
also very similar.

There were also some differences between the two
studies. A much greater range of percent open canopy was
observed in Wisconsin, 15-50 percent, as compared to the
control areas at Russ Forest at 7-18 percent. The
Wisconsin study took place in a pine forest while Russ

Forest is a deciduous hardwood stand. The recording of

39

the herbaceous response also differed. Total leaf area
was measured in cm2 In.2 in Wisconsin while net produc-
tivity was recorded in gm m-2 at the Russ study. However,
these two sets of data should be comparable Since leaf
area can be converted to dry weight.

Soil on the Wisconsin site, is classified as
Vilas sand, a soil of pitted outwash origin, overlying
crystalline bed rock. Weathering and eluviation have
improved the texture of the B horizon to a loamy sand.
The Kalamazoo and Oshtemo soils at Russ Forest are well
drained sandy loams. These soils have a much higher
percentage of Silt and clay than does the Vilas sand.

Differences in predominant overstory vegetation
and soils, as well as greater variability in percent of
Open canOpy coverage partially account for the lack of
good comparisons between these two study sites. These
Site differences probably influence soil moisture greatly
and whereas soil moisture was limiting in the Wisconsin
study, it was not limiting in the Russ Forest study.

A basic tenet underlying the Anderson study is
that the distribution of rhizomes and roots of herbaceous
plants is mainly in the upper mineral or lower organic
horizons where drying from the surface is minimal and
recharge from showers is frequent. Since the roots of
these herbaceous plants cannot draw on moisture from the

water table, one would expect the productivity of these

40

plants to be very responsive to patterns of surface
recharge in a soil where moisture in the upper horizons
is frequently in critical amounts. The Wisconsin study
never explicitly states whether soil moisture frequently
reached limiting amounts, but the Vilas sand of the study
Site is a droughty soil, and soil moisture could indeed
become limiting.

At Russ Forest, numerous gravimetric moisture
samples of the upper 15 cm of soil were taken throughout
the summer of 1972 (Table 9). The lowest percent moisture
reading recorded in the control areas was about 8.3
percent by weight. This is approximately 4.0 atmospheres
of tension (see Table 10). Fifteen atmospheres of tension
is considered the wilting point. All other soil moisture
values ranged from 14.5 to 21.7 percent by weight; 21.7
percent being field capacity. Thus it appears that during
the study period, soil moisture in the upper 15 cm of the
profile was rarely in limiting supply.

In addition, 1972 was a dry year, only 16.2 cm
of rain recorded for June, July and August compared to
the 30 year average of 27.2 cm. Therefore, the relative
abundance of soil moisture cannot be attributed to an
unusually wet season. Of course, gravimetric soil
moisture data alone cannot conclusively prove whether

soil moisture was or was not in limiting supply.

41

Table 9. A record of summer precipitation (inches) and percent soil moisture by weight
taken simultaneously in various cutting treatment areas.

 

Group
Date Pot. Control Selection Shelterwood Clearcut

 

June 29 9.
1

July 21.74 19.64

14.44

11 14.15
13.98

17 19.12

19 ' 20.80

24 20.40 14.94 18.32 22.68

26 19.57 23.74
27 10.9

31 15.06 13.41 14.80 21.97
August

16 14.50 14.87 13.94

18 15.65 11.50 18.73 13.30

22 8.28 8.26 10.22 14.06

 

42

Table 10. Textural analysis and moisture content (% by
weight) at several tension levels for a
composite soil sample Of all plots.

 

SOil Moisture Tension
Texture AtmOSpheres
Sand % Clay

 

Silt 1/3 1/2 2/3 1 5 10

 

63 26 11 21.7 18.1 16.5 13.3 7.9 6.5

 

If soil moisture was not limiting at Russ Forest,
this would account for the weak correlation between
canOpy Openings, throughfall precipitation and herbaceous
productivity. It appears that the levels Of herbaceous
productivity are randomly distributed in relation tO
canOpy Openings. Therefore, it is very likely that the
strong relationship Anderson 3E_3£. (1969) found between
percent Open canOpy, percent throughfall precipitation,
and herbaceous productivity cannot be routinely extra-

polated tO other areas.

CHAPTER VII

SHIFTS IN SPECIES COMPOSITION

AND STRUCTURE

Species Dominance and Leaf Area
in Cutting Treatment Areas

 

 

By taking successive measurements on the permanent
3 m x 3 m plots before and after cutting, two types Of
data are available: increases or decreases in leaf area
for each species and the prOportion Of total leaf area
occupied by that species. Table 11 shows the results Of
the 1971 and 1972 surveys for all treatments. Mean leaf
areas, expressed as percent Of one m2 accompanied by
their standard errors are given for each Of the 37 species
recorded. ”T" tests Of differences between the 1971 and
1972 means are also shown.

Parthenocissus guinguefolia dominated the herba-
ceous layer the year before cutting (1971). This species
accounted for 63 percent Of the total herbaceous leaf
area. Four other predominant species: HEEE§.£EEES'

Viola spp., Osmorrhiza claytonia, and Circaea guadrisul-

 

cata accounted for another 18 percent Of the leaf area.

43

‘4«4

Table 11. Changes in mean leaf area (percent Of 1 m2) for each species in cutting treatment
areas between the years 1971 and 1972.

 

 

Mean leaf area Mean leaf area Changes in
. and standard and standard r ortion
SPBCIes error error Difference "T" 1971 to
Herbs 1971 1972 1971-1972 819.1 1972
1 Circaea quadrisulcata 3.86 (.98) 10.81 (1.92) 6.95 ** + 3.8
' 2 Cirsium spp. .00 (.00) .34 (.32) .34 + .2
8 3 Erigeron canadensis .00 (.00) 2.05 (.68) 2.05 ** + 1.7
4 Galium circaezans .51 (.20) 1.26 (.38) .76 + .4
5 Geranium robertianum 1.40 (.67) 2.51 (.46) 1.10 + .3
6 Geum canadensis and
Sanicula canadensis .65 (.20) 2.61 (.57) 1.96 " + 1.5
* 7 Hedeoma pulegioides .00 (.00) .29 (.14) .29 + .2
8 Hydrophyllum caradenses .00 (.00) .01 (.01) .01
‘ 9 Lactuca saligna .00 (.00) .71 (.21) .71 *' + .6
10 Osmorrhiza Claytonia 2.25 (.33) 3.61 (.91) 1.35
'11 Oxalis stricts .00 (.00) 1.25 (.33) 1.25 *' + 1.1
12 Parthnocissus quinquetolia 46.51 (4.49) 37.94 (3.74) 8.57 o32.2
13 Phryma leptostachya 1.31 (.35) .94 (.28) .37 - 1.0
'14 Physalis subglabriata .00 (.00) .17 (.08) .17 ' + .1
'15 Phytolacca americana .00 (.00) 12.55 (1.93) 12.55 " +10.6
16 Poaceae (species group) .62 (.30) 3.04 (.54) 2.42 " + 1.7
17 Podophyllum peltatum .26 (.15) 3.57 (.99) 3.30 8* + 2.6
'18 Polygonatum canaliculatum
and Smilacina racemosa .22 (.09) 1.09 (.45) .87 + .6
19 Polygonatum sagittatum .00 (.00) 1.24 (.63) 1.24 ** + 1.1
20 Rhus radicans .85 (.25) 1.02 (.27) .17 — .3
21 Tovara virginiana 1.76 (.80) 4.19 (1.23) 2.43 + 1.2
22 Viola app. ' 3.80 (.83) 3.38 (.47) .41 - 2.4
Trees
23 Acer rubrum .33 (.11) .56 (.16) .23 + .1
24 Acer saccharum 1.03 (.24) 3.28 (.63) 2.25 *‘ + 1.4
25 Carya cordiformis .23 (.13) .00 (.00) .23 - .3
26 Cornus florida .66 (.33) 2.72 (.77) .29 + .2
27 Ostrya Virginians .00 (.00) .40 (.26) .40 + .3
28 Prunus serotina .18 (.08) 1.10 (.19) .92 '* + .3
29 Prunus Virginians .68 (.22) 1.05 (.38) .37
30 Quercus alba .13 (.05) .10 (.04) .03 - .1
31 Quercus rubra .13 (.06) 4.59 (.66) 4.46 '* + 3.7
32 Sassfrass albidum .17 (.07) 1.73 (.68) 1.56 * + 1.2
33 Ulmus rubra 3.06 (.80) 3.82 (.63) .76 - 1.0
Shrubs
34 Crataegus spp. .04 (.03) .08 (.08) .04
35 Ribes spp. .54 (.15) 1.19 (.50) 1.36 9 + .9
36 Rubus spp. .02 (.20) 1.94 (.56) 1.92 8* + 1.6
37 Virbunum lentago .71 (.22) .00 (.00) .71 " - 1.0

 

*Invading species.

1Test of significance between 1971 and 1972 means.

45

The remaining 28 percent was distributed among the other

33 Species present. Although Parthenocissus quingue-

 

folia dominated the herbaceous layer in terms of total
leaf area, it was positively correlated at the 0.01 level,

with all the major herbaceous Species. Parthenocissus

 

quinguefolia and all other herbaceous species were nega-

 

tively correlated with woody Species but not at a signi—
ficant level. Judging from the correlation matrix for all
Species and knowledge of the area, it appears that the
special distributions of all species are arranged in a
generally random pattern. After cutting (1972), Partheno-
cissus quinquefolia still remained the largest component
of the herbaceous leaf area, 32 percent. However, two

other Species: Phytolacca americana and Circae guad-

 

risulcata were major codominants, collectively comprising

 

20 percent Of the total leaf area. Four other major

species: Osmorrhiza claytonia, Tovara virginiana, Ulmus

 

rubra and Quercus rubra contributed another 14 percent.

 

The above 7 species accounted for 56 percent of the total
leaf area.

A comparison of the above data with the 1971
survey indicates that canopy removal caused a shift in
the proportion Of leaf area occupied by each species.
This Shift is Shown in the right hand column of Table 11.
Most resident species (those present before cutting) and

invading Species increased their prOportion of total leaf

46

area at the expense of Parthenocissus guinquefolia which

 

decreased 32 percent. These proportional shifts of leaf
area resulted in a more equitable distribution of herba—
ceous leaf area among all Species.

The three dominant species: Parthenocissus

 

guinguefolia, Circaea guadrisulcata, and Phytolacca

 

 

americana were negatively correlated with one another,

 

but these correlations were not significant. The above
three Species were negatively correlated with all woody
species and positively correlated with all herbaceous
species, both correlations significant at the 0.01 level.
Aside from these generalizations, there are no apparent
groups of herbaceous or woody species associated with any
of the dominant species. Plot size greatly influences cor-
relations and a smaller plot than the 3 m x 3 m size used
here would probably produce different results. The corre-
lation matrix for the 1972 survey is similar to that of
the 1971 survey.

Invading Species (those not present before cutting)
Showed a significant increase in leaf area (Table 11).
Hedeona pulegioides is the only exception. Many of the res-
ident species also had a significant increase in leaf area.

Total leaf area increased from 0.72 m2 . m-2 in 1971 to

1.18 m2 . m-2 in 1972. It is interesting to note that of

the total increase of 0.46 m2 . m'z, 57 percent (0.26 m2 .

m-Z) can be attributed to invading species. The remaining

2

43 percent (0.20 m . m-z) came from resident species.

47

Those specific resident herbaceous Species with
significant increases of leaf area (numbers 1, 6, l6, 17)
were examined further. They are all Species described as
occurring in Open woods, thickets, borders, and small
Openings (Fernald 1950). The study area is a second growth
stand and had been grazed up until 40 years ago. There-
fore, it is not surprising that the above herbaceous
species were present and that canopy removal did not
present an unacceptable Shock. Perhaps the herbaceous
Species were being slowly eliminated as the stand pro-
ceeded toward a climax state. Cutting treatments changed
the environment to a more Optimal state for these herba-
ceous Species, and thus we observed a large increase in
their leaf area.

Eight resident tree and shrub species also had
Significant increases in leaf area. All Of these Species
except Prunus serotina are classified as high or inter-
mediate in tolerance (Harlow and Harrar 1958). It is well
known that these Species reSpond to canopy removal and an
increase in their leaf area was expected. The large in-
crease in Quercus rubra leaf area (number 31) was due to
the large acorn crOp in the fall of 1971 rather than to

an increase in growth of established seedlings.

48

S ecies Dominance and Leaf
Area £2 ControIAreas

 

 

Table 12 Shows the results of the 1971 and 1972
surveys for control areas. The format of data presenta-
tion is the same is in Table 11. Characteristics of the
herbaceous layer in control areas were identical to those
found in treatment areas for the 1971 pre-cutting survey.
Parthenocissus guinguefolia was the Single dominant
species. Spatial distributions of all Species are
arranged in a random pattern.

Little change was Observed in control areas

between the 1971 and 1972 surveys. Parthenocissus

 

guinquefolia was again the dominant Species. Only three

 

Species (numbers 16, 21, 33) had significant shifts in
leaf area, and all three were resident Species. In fact,
none Of the invading Species found in treatment areas
were present in control areas.

The data on Podophyllum peltatum is Somewhat mis-
leading. The large increase Of leaf area is due to a
sample plot being located in the midst of a clone of these
species. In the 1971 survey, this plot was not surveyed
until late August, and most of the plant tOps had already
died. The 1972 survey was conducted in July, and the
PodOthllum peltatum plant tOps were still alive.

There were some shifts in the proportion of total

coverage occupied by each species, possibly in the

49

Table 12. Changes in mean leaf area (percent of 1 m2) for each species in control areas
between the years 1971 and 1972.

 

 

Mean leaf area Mean leaf area Changes in
and standard and standard Difference ”T" proportion
SP°°1°’ error error 1971-1972 sig.1 1971 to
Herbs 1971 1972 1972
1 Circaea quadrisulcata 10.89 (2.35) 7.39 (1.71) 3.50 - 1.2
' 2 Cirsium spp.
' 3 Erigeron canadensis
4 Galium circaezans .14 (.14) .06 (.06) .08 - .2
5 Geranium robertianum 2.33 (1.16) .67 (.26) 1.66 - 1.5
6 Geum canadensis and
Sanicula canadensis 1.67 (.48) 2.56 (.68) .89 + 1.9
‘ 7 Hedeona pulegioides
8 HydrOphyllum canadense 4.17 (1.96) 2.00 (.83) 2.17 - 1.6
' 9 Lactuca saligna
10 Osmorrhiza claytonia 2.53 (.70) 1.58 (.32) .95 - .5
*11 Oxalis stricts
12 Parthenocissus quinquefolia 57.39(10.57) 43.56 (7.34) 13.83 + .1
13 Phryma leptostachya .42 (.31) .64 (.26) .22 + .5
'14 Physalis subglabriata
'15 Phytolacca americana
16 Poaceae (species group) .11 (.11) .03 (.03) .08 - .1
17 Podophyllum peltatum .00 (.00) 2.06 (.72) 2.06 *' + 3.0
18 Polygonatun canaliculatum
and Smilacina racemosa .91 (.14) .22 (.15) .04 + .1
‘19 Polygonatum sagittatum
20 Rhus radicans .14 (.14) .00 (.00) .14 - .2
21 Tovara virginiana .69 (.69) 1.06 (.64) .37 + .7
22 Viola spp. 2.28 (.76) 1.55 (.41) 1.87 ' - .3
Trees
23 Acer rubrum .06 (.06) .06 (.04) .00
24 Acer saccharum .69 (.27) 1.00 (.31) .31 + .7
25 Carya cordiformis .28 (.19) .00 (.00) .28 - .3
26 Cornus florida 1.69 (.94) 2.86 (1.30) 1.17 + 2.3
27 Ostrya virginiana
28 Prunus serotina .97 (.37) .11 (.09) .86 - .9
29 Prunus virginiana .75 (.45) .14 (.09) .61 - .6
30 Quercus alba .28 (.17) .17 (.12) .11 - .1
31 Quercus rubra
32 Sassfrass albidum
33 Ulmus rubra 1.56 (.48) .92 (.36) .64 * - .4
Shrubs
34 Crataegus spp.
35 Ribes spp. 1.67 (.49) 1.06 (.44) .61 - .3
36 Rubus spp.
37 Viburnum lentago .58 (.30) .00 (.00) .58 - .6

 

*Invading species.

1Test of significance between 1971 and 1972 means.

50

direction of a climax community. However, shifts in
herbaceous species Similar to these have been found in a

mature undisturbed Acer saccharum-Fagus grandifolis

 

 

stand (Schneider 1966). Permanent plots had been Ob-
served in the previous stand for 30 years and it was
concluded that the changes were non-directional.

Total leaf area in control areas drOpped from
about 0.92 m2 . m"2 in 1971 to 0.70 m2 . m'2 in 1972.
There is no apparent explanation for this drop except
to attribute it to annual variation.

These data from the control plots Show that in
the absence of canOpy removal, a Shift in species composi-
tion is slight. The lack of long term data, and the fact
that grazing was excluded only 40 years ago, makes it
difficult to characterize this herbaceous vegetation as
approaching a climax state. The usefulness of data from
these areas is in comparison with cutting treatment areas.
From this comparison it can readily be seen that canOpy
removal definitely causes a Shift in Species composition

and dominance.
Shift in Species Composition

CanOpy removal had a very significant effect in
increasing herbaceous productivity. One might expect
this higher productivity to be associated with a collec-

tion of new invader Species not present before cutting.

51

However, such a large Shift in species composition did not
take place. In a comparable Old, second growth hardwodd
stand near East Lansing, Michigan, it was found that of
the total of 144 Species recordedover the whole stand,
only 35 species were common to both clearcut and undis-
turbed areas (S. N. Stephenson, personal communication).2

A number of possible explanations exist as to why
the Species composition Shift was only of moderate propor-
tions in this particular study:

1. Much of the ground in the treatment cutting areas
was not available for colonization. The 1971
survey Showed that each plot contained a leaf

2 2

area of about 0.92 m . m- . This is not to

2 . m-2 is full site occupancy.

Values as high as 3.2 m2 . m"2 were recorded, but

imply that 1.0 m

it does Show that the forest floor is not com-
parable to the bare ground found in abandoned
fields. Also, lOgging slash physically occupied
much of the site. In many cases, this Slash
created microsites that appeared not too unlike
the control areas in regard to shade, temperature,

and relative humidity.

2. Time is an important factor. Dryness (1973),

working in a clearcut area in the Cascade

28. N. Stephenson, Dept. of Botany, Michigan

State University.

52

Mountains of Oregon, found that invading herba-
ceous Species dominated the site from the second
through the fourth growing seasons. By the fifth
year, residual herbaceous Species had regained
dominance. Perhaps a Similar sequence will take

place in this eastern deciduous forest.

McIntosh (1970) suggests that Species do not grow
exponentially and saturate a site. The occupancy

of a Species in any area is restricted by a combina-
tion of suppression and competition among all
Species having access to the Site. As long as the
resident Species of a Site can respond to changes

in the microenvironment, they will surely be suffi-
ciently competitive to prevent a large influx of
invading Species. Such seems to be the case in

this study.

Advance reproduction, saplings and stump Sprouts,
quickly occupied large areas providing shade and

competition for available resources.

Certainly, allelochemical reactions, internal
regulation of germination, and seed availability
all have an effect on what Species will invade a
Site, but little is known of how these factors

work.

 

53
Species Diversity

Diversity is property that can be used to measure
changes in Species composition and structure. Such changes
are reflected in the number of species (the variety com-
ponent) and how the individuals are distributed among the
Species (the equitability component). The two measures of
diversity used here, the Shannon-Wiener function (H') and
Simpson's index (D), combine both of these components into
one value.

CanOpy removal increased the mean diversity of
herbs in treatment areas from H'=2.48 before cutting to
H'=3.88 after cutting. These areas had a rather large

increase in the number of Species per plot; 4.46 m"2 in

1971 to 8.69 m‘2

in 1972. Since only 8 new invading Species
were included in the 1972 survey, increasing the total
number of Species from 29 to 37, it seems likely that the
large increase in number of Species after a cutting is not
entirely due to the new invaders. The Spatial patterns
of the resident Species have also shifted, spreading these
species more equally over all plots.

Figure 4 illustrates this Spatial Shift in species
composition. Resident Species, numbers 1, 5, 7, l6, 17,
22, 23, 24, 26, and 28 greatly increased in the number of

quadrants occupied after cutting. These Species are quite

adaptable to successfully invading newly available sites

created by canopy removal.

54

 

 

 

 

 

 

 

 

nun-unno-nu-on-n-ussg
.nsss-I-I-E
.nn-aué

.u................EE

f

 

 

 

 

 

 

 

 

 

90?

80-

70..

60
0

5 w m

powmsooo mussupssv Mo “03592

20
10
0

 

Number of quadrants occupied by species in cutting treatment

areas during 1971 (black line) and 1972 (dash line).

Figum 4.

55

Mean diversity in the undisturbed control areas
remained very stable; H'=2.27 in 1971 and 2.26 in 1972.
These areas did have a Slight increase in number of Species

2 in 1971 to 4.89 m'2 in 1972. This

per plot: 4.11 m-
rather small shift is probably due to random new establish-
ments of resident Species. Figure 5 shows minor shifts in
the number of quadrants occupied by each species. These
shifts probably account for the increase in Species number.

The increase in mean diversity and productivity in
the same treatment areas indicates a positive relationship
between diversity and productivity. However, Margalef
(1969) suggests that these two parameters vary in an inverse
lOgarithmic relationship. Odum (1971) feels that while
productivity affects Species diversity, the two are not
related in any simple manner. He gives examples of very
productive communities with high diversity: a coral reef,
or low diversity: a temperate estuary.

To further examine the relationship between diver-
sity and productivity, 14 meter square plots, ranging in
productivity from 11 gm m-2 to 507 gm m'z, were selected
from the 1972 herbaceous survey. Diversity was calculated
for each plot using both the Shannon-Wiener function (H')
and Simpson's index (D). The results are shown in
Table 13.

The number of Species per plot first increases

with increasing productivity, then levels Off at a value

Number of quadrants occupied

 

 

 

 

 

56

 

   

 

 

 

30 '1
25 - E
201 5
15
10- ;
5. t i l : :
0 _ 7 E E , 1 E I E E l E
12 1 10 22 6 24 26 33 5 2888 133129 3) 37 18 2523 21 1620 4
Species (numbers correspond to Table 12)
Figure 5. Number of quadrants occupied by various species in control

areas during 1971 (black lines) and 1972 (dash lines).

57

Table 13. Changes in evenness and three measures of diversity;
number Of species, Shannon-Weiner function, and
Simpson's index, in relation to changes in net

 

 

productivity.

Dry Number Shannon-Weiner Simpson's Evenness
weight of function index H'/H'max.
gm.m’2 Species (H') (0) (J)

11 5 1.827 .732 .786
19 5 1.901 .732 .822
40 6 1.339 .844 .518
50 6 2.973 .634 1.150
75 9 2.485 .749 .784
106 11 1.551 .548 .409
132 12 3.033 .435 .846
165 16 3.651 .915 .913
196 14 2.727 .786 .716
229 17 4.362 .918 1.067
292 11 1.458 .569 .348
303 12 2.641 .784 .736
425 10 1.192 .321 .359

507 12 1.538 .448 .429

 

58

Of 17, and finally drOps off Slightly as productivity
reaches its highest levels. The two diversity indices
follow a similar pattern. The highest diversity is found
in the intermediate plots, where productivity is between
132-229 gm m-2. As productivity increases to its highest
levels, diversity drops Off sharply. Some of the lowest
diversity values are found in the two most productive
plots. These plots still contain a rather large number of
Species, but productivity is concentrated in one Species,
Phytolacca americana, as revealed by the low evenness
values (J). This is why the diversity indices are so low
in these same plots.

The two indices of diversity coincide rather
closely (Figure 6). It is only in the three least produc-
tive plots that the two indices widely diverge. Both
indices are rather insensitive to rarer species. Perhaps
some inherent characteristic Of each index causes them to
diverge in such cases.

The decrease in diversity on the most productive
Of the 14 selected plots seems to contradict the general
Observation that mean diversity increases with produc-
tivity. Diversity, in this study, was calculated only
for the 37 most abundant Species. The data here only
serves to emphasize Odum's generalization that the rela—

tionship between diversity and productivity is not clear.

59

(0) reps; s.uosdm13

om.

awn-.suv hvwswuoathn as:

.mufi>wuosmoum use
Ocsoum m>onm mcfimsmuocw muw3 .xmpcfl m.c0mmafim was coauocsm
umcwmzucoccmnm emu ha commemoe mm .huwmuw>wp cw coaumwus>

 

 

 

 

 

 

 

 

 

I ...~.

      

 

'8: '

.m mummwm

. o.”
9” m
m
m
r 9~ n
O
J
m
m
fine )
m
T can
.nw

U 004

 

60

It should be noted that the data from these 14
selected plots again illustrates the large amount of
spatial variability in productivity and diversity that
can be encountered in cut-over areas. This variability

was previously noted in regard to environmental parameters.

CHAPTER VIII

SUMMARY

Comparisons of above-ground net productivity of the
herbaceous stratum at Russ Forest with other ecosystems of
various ages, densities, and locations (Table 14). The
uncut control areas Of Russ Forest had a high-intermediate
productivity value of 37 gm m-2 yr -1. This value is
Slightly lower than for other forest sites in Tennessee and
Nova Scotia (numbers 2, 3, 4, 5, 6) and much higher than in
forest Sites at Brookhaven Hubbard Brook, Santa Catalina
Mts., Louisiana, and Germany (numbers 7, 8, 9, 10, 11).

The cutting treatment areas at Russ Forest had an
average productivity of 153 gm m-2 yr-l. This figure is
high compared to partially cut stands in Louisiana
(number 11), but low compared to an old field the first
year after abandonment (number 14).

These comparisons, although general, do illustrate
that the productivity of the herbaceous layer at Russ
Forest was above average before cutting. It must still be

determined if this high productivity had a direct influence

upon the number of new invading Species appearing the year

61

62

Table 14. Comparisons of above ground net herbaceous productivity
(gm m"2 yr'l) between Russ Forest and other ecosystems.
All data concerns only the herbaceous strata.

 

Aboveground productivity

 

Ecosystem g m_2 Yr.1

1. Russ Forest (control areas), 80 yrs.,

31 m2 h-1 37
2. Mixed hardwood forest (Oak Ridge,

Tenn.), 4 yr. old opening 45(1)
3. Red-white oak forest (Smokey Mts.,

Tenn.), climax stand, 22 m2 h-1 35(2)
4. Red oak forest (Smokey Mts.,

Tenn.), climax stand, 24 m2 h-1 52(2)
5. Dense hardwood (N. Brunswick-

Nova Scotia) 51(3)
6. Open mixed hardwood-conifer

(Nova Scotia) 44(3)
7. Fagetum forest

(west Germany), 120 yrs. dense 1(4)

8. Oak-pine woodland (Brookhaven,
N.Y.), 56 yrs, 16 m2 h-1 2(5)

9. Pine-oak woodland (St. Catalina
Mts., Calif.), climax stand,
26 m2 h-1 3(6)

10. Sugar maple, beech, birch forest
(Hubbard Brook, N.H.), 67 yrs.,

23 m2 h-1 lO-l6(max. 29) (7)
ll. Pine-hardwood forest (Louisiana)

evenage plantation, 17 m2 h"1 10(8)
12. Russ Forest (cutover areas) 153

13. Pine-hardwood forest after
elimination of hardwoods

 

(Louisiana) 62(8)
14. Oldfield first year after

abandonment (Georgi a) 494 (9)
l. Cristofolini, 1970 6. Ibid.
2. Whittaker, 1966 7. Siccama st 31., 1970
3. Telfer, 1971 8. Blair, 1971
4. fiber, 1972 9. Odum, 1960
S. Whittaker and Woodwell, 1969

63

after cutting. Do areas of high productivity have little
species change after cutting, while areas of low produc-
tivity.have a high Species change after cutting?

Data reported by Dryness (1970) implies that areas
of high productivity more fully occupy a site and there is
less available Space for invading Species. In an unpro-
ductive Site, 11 Species were not present in both Of the
periods: one year before cutting and 7 years after cutting.
In a highly productive Site, only 5 Species were not pre-
sent in both the same two periods. At Russ Forest, 8
principle invading Species were Observed. These Species
constitute 22 percent of the total species observed.
Although this data does not suggest that 8 new Species is
a high or low number, it can be seen that there was no
large change of Species. A11 29 of the resident species
were found after cutting.

Analysis of the correlation matrix for all Species
showed no distinct groupings of Species in the herbaceous
strata. This was true for both the precutting and post—
cutting communities. Species composition remained very
stable in the control area, while in cutting treatment
areas, 8 invading Species and a number Of resident Species
increased their prOportion of the total coverage at the

expense of the dominant Species, Parthenocissus

 

quinquefolia.

 

64

It is difficult to project future shifts in species
composition based only on two years' data. A visual survey
in June, 1973, the second growing season after cutting,
Showed that grasses had greatly increased their coverage,
especially in the more disturbed areas such as roads, skid
trails and loading sites. Stump Sprouts of Prunus serotina
and glmgg_rub£3_have now attained heights of 2.5 m, and will
soon begin to capture large portions of the solar nutrient
and water resources. These Sprouts can completely dominate
a site as shown by Johnson (1971). In southern Wisconsin
stump Sprouts grew to a height of 5.8 m in four growing
seasons.

At Russ Forest, stump sprouts may dominate portions
of the cutting areas by the fourth or fifth growing season.
What effect this will have on the herbaceous strata is
difficult to say. On other areas, such as roads and loading
Sites which have no stump Sprouts, herbaceous Species
should dominate longer.

It was emphasized that the cutting areas contained
a mosaic of different microsites created by the distribu-
tion of Slash. The conditions in these microsites range
from hot and dry in exposed Sites to cool and humid in
very sheltered sites. This study did not specifically
investigate the relationship between microsite conditions
and herbaceous productivity. It was observed that under

very large slash piles and in very exposed sites, the

65

level of herbaceous productivity was lower. It would be
interesting to see the effect microsite conditions has on
tree reproduction. Perhaps the manipulation of slash
distribution could be a very useful tool in forest manage-
ment creating microclimates to encourage desirable species
and discourage undesirable ones.

The amount of basal area removed in each of the
three treatments resulted in a similar herbaceous response.
Herbaceous productivity was not Significantly different
between these treatments. Such a differential response of
herbaceous productivity to canopy removal would probably
have been observed had lesser amounts of basal area been
removed as in Single tree selection cuttings.

The dimension of time is a most important factor in
a study of herbaceous response to canOpy removal. It
remains to be seen at just what time various herbaceous
species will reach their peak abundance and then are
replaced by other species. Results of this type will not
be available until four to six years of data have been
collected. Continuing annual surveys at Russ Forest are

planned.

LITERATURE CITED

LITERATURE CITED

Ahlgren, I. F. and C. E. Ahlgren. 1960. Ecological
effects of forest fires. Bot. Rev. 26:483-533.

Anderson, R. C., O. L. Loucks, and A. M. Swain. 1969.
Herbaceous response to canOpy cover, light
intensity, and throughfall precipitation in coni-
ferous forests. Ecology 50:255-262.

Blair, Robert M. 1971. Forage production after hardwood
control in a southern pine-hardwood stand. Forest

Boivin, Jean-Louis. 1971. A study of regeneration after
clearcutting in mixed and hardwood cover types
in eastern Quebec. For. Chron. 47(2):82-85.

Bormann, F. H. 1953. The stastical efficiency of sample

plot size and shape in forest ecology. Ecology
34:477-487.

Church, Thomas W., Jr. 1960. Residual stand density and
the early develOpment of northern hardwood repro-
duction in upper Michigan. U.S. Dept. of Ag.
Forest Service, Lakes States For. Expt. Sta.,
Tech. Note 593, 2p.

Clapham, A. R. 1932. The form of the observational unit
in quantitative ecology. J. Ecology. 20:192-197.

Coombe, D. V. and W. Hadfield. 1962. An analysis of
growth of Mersanya cerOpiOideS. J. Ecology 50:
221-234.

Cristofolini, Giovanni. 1970. Biomassa e produttivita

dello strato erbaceo di un ecosistema forestale.
Giornale Botanico Italiano. 104(1):1-34.

66

67

Daubenmire, Rexford. 1959. A canOpy-coverage method of
vegetational analysis. Nthwest Sci.
33(1):43-64.

. 1968. Plant Communities: A Textbook Of Plant
Synecology. Harper and Row, New York, 300 P.

del Moral, Rodger, and Cornelius H. Muller, 1970. The

allelopathic effects of Eucalyptus camaldulensis.

 

Dixon, H. D. 1969. Growth of lodgepole pine in the
Colorado forest range as related to environment.
Disst. Absts. 30:6, 256.

Dryness, C. T. 1973. Plant succession in the Oregon
Cascades. Ecology 54:57-69.

Eber, W. 1972. The primary productivity of the ground
vegetation of the Luzulo-Fagetum. In Integrated
Experimental Ecology, Heinz Ellenberg (ed.),
Springer-Verlag, New York, 214 P.

Evans, Francis C. and Stanley A. Cain. 1952. Preliminary
studies on the vegetation of an Old field community
in southeastern Michigan. Contributions from the
Laboratory of Vertebrate Biology University of
Michigan NO. 51. 19 P.

Evans, Francis C. and Eilif Dahl. 1955. The vegetational
structure of an abandon field in southeastern
Michigan and its relation to environmental
factors. Ecology 36:685-706.

Fayle, D. C. F. 1959. The point contact method as a three
dimensional measure of ground vegetation. For. '
Chron. 35(2):135-141.

Fernald, Merritt Lyndon. 1950. Gray's Manual of Botany.
American Book Co., New York, 1632 P.

Gainer, E. M. et al. 1954. Forage production on a
longleEf'EIne stand of southern Alabama.
Ecology 35:59-62.

Getz, L. L. 1960. Standing crOps of herbaceous vegetation
in southern Michigan. EOOIOgy 41:393-395.

Golley, Frank B. 1960. Energy dynamics of a food chain
of an old-field community. Ecol. Mono.
30:187-206.

68

Greig-Smith, P. 1964. Quantitative Plant Ecology. 2nd
ed. Butterworth and Co., London, 256 P.

GySel, Leslie W. 1951. Boarders and openings of beech-
maple woodlands in southern Michigan. Jour. For.
49:13-19.

Harlow, William H. and Ellwood S. Harrar. 1958. Textbook
Of Dendrology. 4th ed. McGraw-Hill Book Co.,
Inc., New York, 561' P.

Hetherington, J. C. 1969. An economic evaluation of
alternative stand treatments in relation to the
develOpment of understory vegetation and sub-
sequent regeneration costs. Jour. For. 42:47-68.

Isaac, L. A. 1940. Vegetative succession following
logging in the Douglas-fir region with special
reference to fire. Jour. For. 38:716-721.

Jarvis, P. G. and M. S. Jarvis. 1964. Growth rates of
woody plants. Physiologia Pl. 17:654-666.

Johnson, Paul S. 1971. Growth and survival of inter-
planted hardwoods in southern Wisconsin oak clear-
cuttings. U.S. Dept. Ag. Forest Service, North
Central For. Expt. Sta., Res. Note NC-118, 4 P.

Keever, C. 1950. Causes of succession on Old field of
the Piedmont, North Carolina. Ecol. Monogr.
20:229-250.

Kershaw, Kennith A. 1964. Quantitative and Dynamic
Ecology. Amer. Elsevier Pub. Co., New York, 183 P.

Kienholz, Raymond. 1929. Revegetation after logging and
burning in the Douglas-fir region of western
Washington. Ill. State Acad. Sci. Trans.
21:94-108.

Kittredge, Joseph, Jr. 1934. Evidence of the rate of
forest succession on Star Island, Minnesota.
Ecology 15:24-35.

Kohno, M. A. 1969. Allelopathic properties of fallen
leaves of Acer spp. and their role in natural
regeneration. Ukr. Bot. Z. 26(1):?8-83.

Larson, M. M. 1970. Root regeneration and early growth
of red oak seedlings: Influence of soil
temperature. For. Sci. 16:442-446.

69

Larson, M. M. and Gilbert H. Schubert. 1969. Root compe-
tition between ponderosa pine seedling and grass.
U.S. Dept. Ag. Forest Service, Rockey Mt. For.
and Range Expt. Sta. Res. Paper, RM-54, 11 P.

Leak, William B. and Robert w. Wilson, Jr. 1958.
Regeneration after cutting of old growth northern
hardwoods in New Hampshire. U.S. Dept. Ag.

Forest Service, Northeast For. Expt. Sta.,
Sta. Paper 103, 8 P.

Lyon, L. Jack. 1971. Vegetal develOpment following
prescribed burning of Douglas-fir in south central
Idaho. U.S. Dept. Ag. Forest Service, Inter-
mountain For. and Range Expt. Sta. Res. Paper,
INT 105, 30 P.

Margalef, R. 1969. Diversity and stability--a practical
prOposal and a model of interdependence. Brook-
haven Symp. Biol. 22:25-37.

Marks, P. L. and F. H. Bormann. 1972. Revegetation follow-
ing forest cutting: echanism for return to steady
state mutrient cycling. Science 176:914-975.

McCaffery, Keith R. and William A. Creed. 1969. Signifi-
cance of forest Openings to deer in northern
Wisconsin. Tech. Bull. Wisc. Dept. of Natural
Resources, 44, 104 P.

McIntosh, Robert P. 1957. The York woods, a case history
of forest succession in southern Wisconsin.
Ecology 38:29-37.

. 1962. Pattern in a forest community. Ecology
43:25-33.

. 1970. Community, competition, and adaptation.
Quarterly Rev. of Biol. 45:259-280.

Metzger, Fred T. and Carl H. Tubbs. 1971. The influence
of cutting methods on regeneration of second
growth in northern hardwoods. Jour. For.
69:555-564.

Milner, C. and R. Elfyn-Hughes. 1968. Methods for the
Measurement of the Primary Production of Grass-

land. IBP. Handb. n. 6, Blackwell, Oxford,
70 P.

70

Minckler, L. S. and J. D. Woerheide. 1965. Reproduction
of hardwood 10 years after cutting as affected by
site and Opening Size. Jour. For. 63:103-107.

Muller-Dombois, Dieter. 1965. Initial stages of secondary
succession in the coastal Douglas-fir and western
hemlock zones. Ecology of Western North America,
Vol. 1, ed. by V. J. Krajima, Univ. Brit. Columbia,
Vancouver, pp. 38-41.

Newbould, P. J. 1967. Methods for Estimating the Primary
Productivity of Forests. IBP Handb. n. 2, Black-
well, Oxford, 62 P.

Newhouse, Michael, Jr. 1968. Some physiological factors
affecting seedling growth of hardwoods. Dissts.
Absts. 29:8, 2698.

Odum, Eugene P. 1960. Organic production and turnover in
Old field succession. Ecology 41:33-49.

. 1971. Fundamentals of Ecology. 3£d_ed.
W. B. Saunders Co., Philadelphia, 547 P.

Oosting, H. J. 1956. The Study of Plant Community.
W. H. Freeman and Co., San Francisco, 440 P.

Pace, C. P. and R. M. Hurd. 1957. Understory vegetation
as related to basal area, crown cover, and letter
produced by immature ponderosa pine stands in
the Black Hills. Proc. Soc. Amer. For. 156-158.

Ray, R. G. 1932. Natural reproduction on abandoned land.
For. Chron. 3(4):l97-205.

Rogers, Dilwyn. 1959. Ecological effect of cutting in
southern Wisconsin woodlots. Dissts. Absts.
20:4, 1994.

Rudolph, V. J. and Walter Lamine. Reproduction in the
Fred Russ forest oak-mixed hardwood areas.
Working paper on file in Dept. of Forestry,
Michigan State University.

Sanders, Dana Ray. 1969. Structure and pattern of
herbaceous understory Of deciduous forest in
central Iowa. Dissts. Absts. 30:6, 2567.

Schneider, G. 1966. A twenty year ecological investiga-
tion in a relatively undisturbed sugar maple-beech
stand in southern Michigan. Mich. State Univ.
Expt. Sta. Res. Bull. 15, 58 P.

71

Siccama, T. G., F. H. Bormann, and G. E. Likens. 1970.
The Hubbard Brook ecosystem study: Productivity,
nutrients, and phytosociology of the herbaceous
layer. Ecol. Mono. 40:389-402.

Smith, Bryce E. and Grant Cottam. 1967. Spatial relation-
ships Of mesic forests in southern Wisconsin.
EcolOgy 48:546-558.

Steen, Harold K. 1966. Vegetation following Slash fires
in one western Oregon locality. Nthwest Sci.
40:113-120.

Struik, G. J. and J. T. Curtis. 1962. Herb distributions
in an Acer saccharum forest. Amer. Mid. Nat.

Telfer, E. S. 1971. Forage yield in two forest zones
Of New Brunswick, Nova Scotia. Jour. Range
Mgt. 25:446-449.

Trimble, George Jr. and George Hart. 1961. An appraisal
of early reproduction after cutting in northern
hardwood stands. U.S. Dept. Ag. Forest Service,
Northeastern For. Expt. Sta., Station Paper
162, 22 P.

Tubbs, Carl H. 1968. The influence of residual stand
densities on regeneration in sugar maple stands.
U.S. Dept. Ag. Forest Service, North Central
For. Expt. Sta. Res. Note, NC47, 4 P.

. 1970. The competitive ability of yellow birch
(Betula alleghaniensis Bretton) seedlings in the
presence of sugar maple (Acer saccharum Marsh.)
with specific references to the role Of allelo
pathic substances. Dissts. Absts. 31:6, 7025.

 

U.S. Forest Service. 1958. Techniques and methods of
measuring understory vegetation. Proceedings of a
symposium at Tifton, Ga. Oct. 1958. U.S. Dept.
Ag. Forest Service, Southeastern For. Expt. Sta.
104 P.

. 1967. Ceanothus retards Douglas-fir.
Extr. from U.S. Dept. Ag. Forest Service, Pacific
Nthwest For. and Range Expt. Sta., Ann. Report: 19.

Van Dyne, G. M. 1963. Influence of small plot Size on

range herbage production estimates. Ecology
44:746-759.

72

Wendel, George W. and George R. Trimble, Jr. 1968.
Early reproduction after seed--tree harvest
cuttings in Appalachian hardwoods. U.S. Dept.
Ag. Forest Service, Northeastern For. Expt. Sta.
Research Paper, NE-99, 16 P.

Westlake, D. F. 1963. Comparison of plant productivity.
Biol. Rev. 38:385-429.

Whipple, D. S. 1968. Yellow pOpular regeneration after
seed-tree cutting and Site preparation. Bull. Ala.
Agri. Expt. Sta. NO. 348, 15 P.

Whittaker, R. H. 1966. Forest dimensions and production
in the Great Smokey Mts. Ecology 44:103-121.

. 1968. Vegetation of the Santa Catalina Mts.,
Arizona IV limestone and acid soils. J. Ecology
56:523-544.

, and P. P. Feeny. 1971. Allelochemics;
chemical interactions between Species. Science
171:757-770.

 

, and G. M. Woodwell. 1969. Structure, production
and diversity of oak-pine forest at Brookhaven,
New York. J. Ecology, 44:155-174.

Wiegert, R. G. and F. C. Evans. 1967. ,Investigations of
secondary productivity in grasslands. In Secondary
Productivity of Terrestrial Ecosystems.

K. Petrasewiez ed. Warszawa, Panstwowe Wydawn,
Naukowe, 879 P.

Winget, Carl H. 1968. Species composition and develOpment
of second growth hardwood stands in Quebec.
For. Chron. 44(6):31-35.

Yerkes, Vern P. 1960. Occurrence of shrubs and herbaceous
vegetation after clearcutting Old growth Douglas-
fir in the Oregon Cascades. U.S. Dept. Ag.

Forest Service, Pacif. Nthwest For. and Range
Expt. Sta. Res. Paper 34, 12 P.

VITA

Douglas Nelson McEwen

Final Examination: July 16, 1973

Guidance Committee: Dr. Gary Schneider, Dept. of Forestry
Dr. Peter Murphy, Dept. of Botany
Dr. William COOper, Dept. of Zoology
Dr. George Coulman, Dept. of Chemical
Engineering

Biographical Items: Born September 4, 1943 in Youngstown,
Ohio; married June 13, 1966 to
Kiva Ann Scholnik; one child,
Ilana (born February 24, 1970)

Education: Miami University, 1961-1963

Michigan State University
B.S. Forestry, 1966
M.S. Resource DevelOpment, 1970

Work Experience: U.S. Forest Service (summers 1964-
1966--research assistant)

Peace Corps (l967-1969--volunteer,
Chile)

Organizations: Alpha Zeta

xi Sigma Pi

Society of American Foresters

Ecological Society of America

American Institute of Biological
Science

American Association for the
Advancement of Science

American Botanical Society

HICHIGRN STATE UNIV. LIBRRRIES
\H‘lNIIIIHHIIIIIIIIIIHIHIWIIHNlll!WWIWIIIWI
31293010576894