. X2. .33
.,n..d,..‘ 4.!

unnuwx. - ﬁrm.
1"

 

‘3 . .ﬁv.
5.2.. a.
$1.13.. .2

.139! l n 1%
~19]. It} al
1 1v! 3‘.

{uh-nun“ «nun r .

 

.V ‘IQID.

roll: 1. ,iﬁuﬂl

. .u 5. ‘I. \ III-
07.!

ovl.“
L 31A.

 

uh. 3'12“] ‘
iii b

L\
.l‘
‘I‘l
|||| l

 

.. . 1:.H395w.

..n.V.ﬁ

lift A

mesa

3 lllllllllWillllllllllllllllll[Illllllllllllllllllllllllll

31293 01413 8873

LIBRARY

Michigan State
Unlverslty

 

 

 

 

This is to certify that the
thesis entitled
THE IMPACTS OF GYPSY MOTH (LEPIDOPTERA:
LYMANTRIIDAE) ON NATIVE ARTHROPOD ABUNDANCE,
SPECIES RICHNESS, AND DIVERSITY IN TWO
HARDWOOD ECOSYSTEMS IN NORTHERN LOWER
M I C H I G A N presented by

TIMOTHY THERON WORK

has been accepted towards fulﬁllment
of the requirements for

MASTERS OF SCIENCE degreein ENTOMOLOGY

l I 1114/1; ,0

Major pr essor /

Date 2—AUGUST-1996

 

0-7639 MS U is an Aﬂ'mnau've Action/Equal Opportunity Institution

 

 

 

PLACE IN RETURN 30X to remove this checkout from your record.
TO AVOID FINES rotum on or before date duo.

DATE DUE DATE DUE DATE DUE

     
     

  

 

 

J“: 09 -
L..L, HL‘

 

 

.1 1"»
i
.___._~.' 1.... ".‘U-’
.“ x a . 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU I. An Afﬁrmative Action/Equal Oppommlty Imam
Wan-9.1

 

THE IMPACTS OF GYPSY MOTH (LEPIDOP’I'ERA: LYMANTRHDAE) ON NATIVE
ARTHROPOD ABUNDANCE, SPECIES RICHNESS, AND DIVERSITY IN TWO
HARDWOOD ECOSYSTEMS IN NORTHERN LOWER MICHIGAN

by

Timothy Theron Work

A THESIS

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

MASTER OF SCIENCE

Department of Entomology

1 996

ABSTRACT

THE IMPACTS OF GYPSY MOTH (LEPIDOPI'ERA: LYMANTRIIDAE) ON NATIVE
ARTHROPOD ABUNDANCE, SPECIES RICHNESS, AND DIVERSITY IN TWO
HARDWOOD ECOSYSTEMS IN NORTHERN LOWER MICHIGAN

by

Timothy Theron Work

Insects play a critical role in many forest processes. However, little is known about
the extent or importance of insect diversity within forests. Ecosystems dominated by red
oak and northern hardwoods were surveyed between 1993 and 1995 to obtain baseline data
on insect diversity and determine whether gypsy moth (Lymantria dispar), an exotic forest
defoliator, impacted native arthropods. A method for quantifying effects of defoliation
using canopy transmittance was also tested in both ecosystems and stands that experienced
differing levels of defoliation. Estimating leaf area index through canopy transmittance was
useful in quantifying effects of severe defoliation. Northern hardwood and red oak
ecosystems differed in species composition, diversity and abundance of carabid beetles and
lepidoptera but not in overall community structure of arthropods. Gypsy moth defoliation
negatively impacted abundance of the carabid, C. limbatus, and species richness and

diversity of lepidoptera, particularly overstory and oak-feeding Noctuidae.

To Julie, Mom, Dad and Katie

ACKNOWLEDGMENTS

I thank my advisor, Deborah G. McCullough, for her seemingly never-ending
enthusiasm and dedication as a teacher, a scientist, and most of all a friend. Deb's
guidance and motivation inspired me to endeavor for excellence and creativity in my
research and in my non-academic life. I would also like to thank Bill Mattson, Carl Ramm
and Mark Scriber for their advice and consultation pertaining to this project.

This project could not have been completed without the help of others. I am
grateful to Brian Kopper for his assistance sorting, identifying, and preparing the daunting
number of insect specimens collected during this project. I also thank Brian for his effort
and company during the back-breaking, mosquito—bitten days packing entirely too much
equipment all over the Manistee National Forest. I am also grateful to Liz Weber, Katie
Albers, and Kevin Elwood for their assistance throughout the summer. I also thank Tom
Ellis for his help in thermal fogging after Kevin's 'accident.‘ I thank John Wilterding,
Robert Kriegel, Mogens Nielson, and Fred Stehr for their hard work identifying the
multitude of Lepidoptera specimens collected. I thank Dennis Gilliland for his helpful
comments and insight into the statistical analysis of this project.

I greatly appreciate the support and comments of my fellow graduate colleagues;
Brian Bishop, Lyle Buss, Sean Clark, Beth Dankert, and Cathy Papp. I also appreciate the
opportunities and distractions provided by my fellow students in Entomology and
Ecology/Evolutionary Biology.

This research was funded by the McIntire-Stennis Cooperative Foresz Research

Grant program. Additional funding was provided through the Professor Hutson Student

Research Award from the Department of Entomology and student support funds from the

Ecology and Evolutionary Biology program at Michigan State University.

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. ix
LIST OF FIGURES ................................................................................ xv
LIST OF SYMBOLS, ABBREVIATIONS, AND NOMENCLATURE ................. xviii
INTRODUCTION .................................................................................. 1
CHAPTER 1
Quantifying Gypsy Moth (Lepidoptera: Lymantriidae) Defoliation in Two Hardwood
Ecosystems Using a Portable PAR sensor ..................................................... 5
Introduction ............................................................................... 5
Methods ...................................................................................... 6
Study Area .......................................................................... 6
Estimation of Gypsy Moth Populations ......................................... 8
Estimation of Defoliation ......................................................... 8
Statistical Analysis ............................................................... 10
Results ...................................................................................... 10
Discussion ................................................................................. 12
Ecosystem Differences .......................................................... 12
Effects of Gypsy Moth .......................................................... 14
Tables ..................................................................................... 16
Figures ..................................................................................... 18
CHAPTER 2

The Impact of Gypsy Moth (Lepidoptera: Lymantriidae) Outbreaks on Carabid Beetle
(Coleoptera: Carabidae) Diversity in Two Hardwood Ecosystems in Northern Lower

Michigan ............................................................................................. 22
Introduction ................................................................................ 22
Methods .................................................................................... 23

Study Area ........................................................................ 23
Beetle Collection ................................................................. 25
Statistical Analysis ............................................................... 27
Results ...................................................................................... 28
Ecosystem Differences in Diversity ............................................ 28
Impacts of Defoliation on Carabid Species Richness, Abundance, and
Diversity ........................................................................... 30
Response of Carabid Species to Gypsy Moth ................................ 31
Discussion ................................................................................. 32

vi

Ecosystem Differences .......................................................... 3 2

Effects of Gypsy Moth and Natural Variation ................................ 3 3
Implications of Gypsy Moth and Defoliation ................................. 3 3
Tables ..................................................................................... 3 7
Figures ..................................................................................... 45
CHAPTER 3
The Impact of Gypsy Moth (Lepidoptera; Lymantriidae) Outbreaks on Native Lepidoptera
in Two Northern Hardwood Ecosystems in Northern Lower Michigan ................... 5 7
Introduction ................................................................................ 57
Methods .................................................................................... 59
Study Area ........................................................................ 5 9
Collection of Lepidoptera ....................................................... 61
Canopy Strata ............................................................ 61
Shrub Strata .............................................................. 63
Statistical Analysis ............................................................... 63
Results ...................................................................................... 65
Effects of Gypsy Moth on Lepidoptera Collected During the Early
Season ............................................................................. 66
Effects of Gypsy Moth on Lepidoptera Collected from the Late
Season ............................................................................ 68
Discussion ................................................................................. 69
Differences in Lepidopteran Fauna Between Ecosystems .................. 69
Effects of Gypsy Moth on Native Lepidoptera Diversity .................... 71
Effects of Gypsy Moth on Oak-Feeders ....................................... 74
Management Implications ....................................................... 76
Tables ..................................................................................... 77
Figures .................................................................................... 93
CHAPTER 4

The Impact of Gypsy Moth (Lepidoptera: Lymantriidae) Outbreaks on the Community
Structure of Arthropods in Two Northern Hardwood Ecosystems in Northern Lower

Michigan ........................................................................................... 108
Introduction ............................................................................... 108
Methods ................................................................................... 1 10

Study Area ....................................................................... l 10
Collection of Arthropods ...................................................... 1 12
Canopy Strata .......................................................... 1 l3

Shrub Strata ............................................................ 1 14

Ground Strata .......................................................... 1 l4

Arthropod Identiﬁcation and Guild Assignments ........................... 1 14
Statistical Analysis .............................................................. 116
Results .................................................................................... 1 16
Abundance of Arthropods Within Guilds .................................. 1 17
Diversity of Families Within Guilds ......................................... 1 18
Discussion ................................................................................ 1 l9
Justiﬁcation for Guild Assignments ......................................... 1 19
Effects of Gypsy Moth on Community Structure ........................... 121
Management Implications ..................................................... 123
Ecosystem Differences in Community Structure ............................ 123
Temporal Variation in Guilds ................................................. 124

Guilds and Constancy ......................................................... 125

vii

High Variability of Saprophages .............................................. 126

Tables ..................................................................................... 127
Figures .................................................................................... 157
APPENDICES .................................................................................... 1 60

APPENDIX A- Table A1. Total number of noctuid species collected from four red oak
sites and four northern hardwood sites over three years .................................... 160

APPENDIX B- Table B1. Total number of tortricid species collected from four red oak
sites and four northern hardwood sites over three years ..................................... 165

APPENDD( C- Table C 1. Total number of geometrid species collected from four red oak
sites and four northern hardwood sites over three years ..................................... 167

APPENDIX D- Table D]. Total number of pyralid species collected from four red oak
sites and four northern hardwood sites over three years ..................................... 169

APPENDIX E— Table E1. List of species collected from families other than Noctuidae,

Tortricidae, Geometridae, and Pyralidae ....................................................... 171
APPENDIX F- Record of Deposition of Voucher Specimens ............................. 174
LIST OF REFERENCES ........................................................................ 186

viii

LIST OF TABLES

CHAPTER]

Table 1.1. Visual estimates of peak defoliation from two locations in red oak ecosystems
and two locations in northern hardwood ecosystems .......................................... 16

Table 1.2. Split-plot ANOVA table for leaf area index (LAD with two levels of defoliation
(severe or none) over two or three months in red oak sites during 1994 and 1995. ...... 17

Table 1.3. Split-plot AN OVA table for leaf area index (LAI) analyzing between site
differences (Harrietta and Mesick) over three months in northern hardwood sites during
1994 and 1995. ................................................................................... l7

CHAPTER2

Table 2.1. Visual estimation of peak defoliation and estimation of mean LAI (iSE) using
photosynthetically active radiation transmittance of sites in red oak and northern hardwood
ecosystems .......................................................................................... 37

Table 2.2. Abundant carabid species collected using suspended UV traps and pitfall traps
in sites dominated by red oak (R0) and northern hardwood (NH) species over three

years1 ............................................................................................... 38

Table 2.3. Uncommon carabid species collected in UV and pitfall traps from red oak (R0)
and northern hardwood (NH) ecosystems over three years1 ................................. 39

Table 2.4. Statistical signiﬁcance of treatment effects from split-plot ANOVA of ﬁve
measures of carabid diversity in four red oak sites. ........................................... 41

Table 2.5. Means (:tSE) of ﬁve measurements of diversity pooled over three years from
red oak and northern hardwood sites (n=2). .................................................. 42

Table 2.6. Statistical signiﬁcance of treatment effects from split-plot ANOVA of ﬁve
measures of carabid diversity in four northern hardwood sites. ............................ 43

ix

Table 2.7. Means (:tSE) of species richness, beetle abundance, and the Brilloun index
from four northern hardwood sites that varied signiﬁcantly over three years .............. 44

CHAPTER3

Table 3.1. Visual estimation of peak defoliation and estimation of mean LAI (iSE) using
photosynthetically active radiation transmittance of sites in red oak and northern hardwood
ecosystems .......................................................................................... 77

Table 3.2. Total number of insects/family that comprised greater than 1% of the total
number of lepidoptera collected from two sites near Branch and two sites near Freesoil
within red oak ecosystems ........................................................................ 78

Table 3.3. Total number of insects/family that comprised greater than 1% of the total
number of lepidoptera collected from two sites near Harrietta and two sites near Mesick
within northern hardwood ecosystems .......................................................... 79

Table 3.4. Total number of insects/family that comprised less than 1% of the total number
of lepidoptera collected each year from two sites near Branch and two sites near Freesoil
within red oak ecosystems ....................................................................... 80

Table 3.5. Total number of insects/family that comprised less than 1% of the total number
of lepidoptera collected each year from two sites near Harrietta and two sites near Mesick
within northern hardwood ecosystem. ......................................................... 81

Table 3.6. Total number of species within each lepidopteran family collected over three
years from two sites near Branch and two sites near Freesoil within a red oak
ecosystem ........................................................................................... 82

Table 3.7. Total number of species within each lepidopteran family collected over three
years from two site near Harrietta and two sites near Mesick within northern hardwood
ecosystems .......................................................................................... 84

Table 3.8. Means (:tSE) of total abundance, species richness, Brilloun’s index and
Simpson’s index for Lepidoptera collected from red oak and northern hardwood
ecosystems over three years (means calculated from data pooled over four months) ...... 86

Table 3.9. Statistical signiﬁcance of treatment effects from split-plot ANOVA of ﬁve
measures of Lepidoptera diversity in four red oak sites in 1993 ............................ 87

 

Table 3.10. Statistical signiﬁcance of treatment effects from split-plot AN OVA of ﬁve
measures of Izpidoptera diversity in four red oak sites in 1994 ............................ 8 8

Table 3.11. Statistical signiﬁcance of treatment effects from split-plot AN OVA of ﬁve
measures of Lepidoptera diversity in four red oak sites in 1995 ............................ 8 9

Table 3.12. Statistical signiﬁcance of treatment effects from split-plot AN OVA of ﬁve
measures of Lepidoptera diversity in four northern hardwood sites in 1993 ................ 90

Table 3.13. Statistical signiﬁcance of treatment effects from split-plot AN OVA of ﬁve
measures of Lepidoptera diversity in four northern hardwood sites in 1994 ................ 91

Table 3.14. Statistical signiﬁcance of treatment effects from split-plot AN OVA of ﬁve
measures of Lepidoptera diversity in four northern hardwood sites in 1995 ................ 92

CHAPTER4

Table 4.1. Visual estimation of peak defoliation and estimation of mean LAI (iSE) using
photosynthetically active radiation transmittance of sites in red oak and northern hardwood
ecosystems ........................................................................................ 127

Table 4.2. Arthropod families assigned to six guilds based on feeding habits and
corroboration with previous classifications ................................................... 128

Table 4.3. Total number of arthropods collected within red oak (R.O.) and northern
hardwood (N .H.) sites over 1993, 1994, 1995 and overall ................................. 131

Table 4.4. The total number of arthropod families collected within red oak (R.O.) and
northern hardwood (N.H.) during 1993, 1994, 1995 and overall .......................... 132

Table 4.5. Mean values (iSE) of the relative proportional abundance of six feeding guilds
collected from two sites near Branch and two sites near Freesoil between 1993 and
1995 ................................................................................................ 133

Table 4.6. Statistical signiﬁcance of treatment effects from split-plot ANOVA of the
relative abundance of insects within six feeding guilds in four red oak sites in 1993.134

Table 4.7. Statistical signiﬁcance of treatment effects from split-plot ANOVA of the
relative abundance of insects within six feeding guilds in four red oak sites in 1994. 1 35

xi

 

Table 4.8. Statistical signiﬁcance of treatment effects from split-plot AN OVA of the
relative abundance of insects within six feeding guilds in four red oak sites in 1995. 1 36

Table 4.9. Mean values (18E) of the relative abundance of arthropods in guilds that varied
significantly by month in red oak sites (n=4) .................................................. 137

Table 4.10. Statistical signiﬁcance of treatment effects from split-plot AN OVA of the
relative number of insects within six feeding guilds in four red oak sites over three
years .............................................................................................. 1 3 8

Table 4.11. Mean values (iSE) of the relative proportional abundance of six feeding
guilds collected from two sites near Harrietta and two sites near Mesick between 1993 and
1995 ................................................................................................ 139

Table 4.12. Statistical signiﬁcance of treatment effects from split-plot AN OVA of the
relative abundance of insects within six feeding guilds in four northern hardwood sites in
1993. .............................................................................................. 140

Table 4.13. Statistical signiﬁcance of treatment effects from split-plot AN OVA of the
relative abundance of insects within six feeding guilds in four northern hardwood sites in
1994. .............................................................................................. 141

Table 4.14. Statistical signiﬁcance of treatment effects from split-plot ANOVA of the
relative abundance of insects within six feeding guilds in four northern hardwood sites in
1995. .............................................................................................. 142

Table 4.15. Mean values (iSE) of the relative abundance of arthropods in guilds that
varied signiﬁcantly by month in northern hardwood sites (nﬁ) ............................ 143

Table 4.16. Statistical signiﬁcance of treatment effects from split-plot AN OVA of the
relative number of insects within six feeding guilds in four northern hardwood sites over
three years ........................................................................................ 144

Table 4.17. Statistical signiﬁcance of treatment effects of ecosystem and year from split-
plot ANOVA of the relative number of insects within six feeding guilds ................. 145

Table 4.18. Statistical signiﬁcance of treatment effects from split-plot ANOVA of the
relative number of families within six feeding guilds in four red oak sites in 1993 ....... 146

Table 4.19. Statistical signiﬁcance of treatment effects from split-plot ANOVA of the
relative number of families within six feeding guilds in four red oak sites in 1994 ....... 147

xii

APPENDIX C

Table C1. Total number of geometrid species collected from four red oak sites and four
northern hardwood sites over three years ...................................................... 167

APPENDIX D

Table D1. Total number of pyralid species collected from four red oak sites and four
northern hardwood sites over three years ...................................................... 169

APPENDIX E

Table El. List of species collected from families other than N octuidae, Tortricidae,
Geometridae, and Pyralidae. .................................................................... 171

xiv

LIST OF FIGURES
CHAPTER 1

Figure 1.1. Mean density (t SE) of gypsy moth egg masses from four stands 1n A) red
oak ecosystems (ELTP 20) and four stands in B) northern hardwood ecosystems (ELTP
45). Means were calculated from ﬁve 0.1 ha ﬁxed-radius plots ............................. 1 8

Figure 1 ..2 Mean LAI (:tSE) from June, July, and August for undefoliated red oak sites
(Branch) and northern hardwood sites (Harrietta and Mesick) during 1994 and 1995.
Means were calculated from n=2 for each site. Error bars were small and may be
obscurred by data points .......................................................................... 19

Figure 13 Comparison of mean LAI (:tSE) from defoliated (Freesoil) and undefoliated
(Branch) sites in red oak ecosystems during 1994 and 1995. Means were calculated each
month from two sites located near Branch and two sites located near Freesoil ............. 20

Figure 1.4. Comparison of mean LAI (iSE) from two northern hardwood sites during
1994 and 1995. Means were calculated each month from two sites located near Harrietta
and two sites located near Mesick ............................................................... 21

CHAPTER 2

Figure 2.1. Mean density of (iSE) of gypsy moth egg masses from four sites in A) red
oak ecosystems (ELTP 20) and four sites in B) northern hardwood ecosystems (ELTP 45).
Means were calculated from ﬁve 0.1 ha ﬁxed-radius plots within each site ................ 45

Figure 2.2. Hierarchical clusters created using beta—ﬂexible linkage (B=c0.25) and
Sorenson’s quantitative index of similarity for total carabids collected from eight sites
during A) 1993, B) 1994 and C). 1995 ......................................................... 46

Figure 2.3. Hierarchical clusters created using beta-ﬂexible linkage (B=—O.50) and
Sorenson’s quantitative index of similarity for total carabids collected from eight sites
during A) 1993, B) 1994 and C). 1995 ......................................................... 48

Figure 2.4. Hierarchical clusters created using average linkage and Sorenson’s quantitative
index of similarity for total carabids collected from eight sites during A) 1993, B) 1994 and
C). 1995 ............................................................................................. 50

XV

Figure 2.5. Means (iSE) of the alpha log series index from red oak sites that experienced
differential amounts of defoliation over three years. Means were calculated from n=2.
Standard errors were small during some years and may be obscurred by size of data
points ................................................................................................ 52

Figure 2.6. Means (:tSE) of species richness from red oak sites that experienced
differential amounts of defoliation over three years. Means were calculated from n=2.
Standard errors were small during some years and may be obscuned by size of data
points ................................................................................................ 53

Figure 2.7. Means (:tSE) of the alpha log series index from Harrietta and Mesick sites in
the northern hardwood ecosystem. Means were calculated from n=2 ...................... 54

Figure 2.8. Means (iSE) of species richness from Harrietta and Mesick sites in the
northern hardwood ecosystem. Means were calculated from n=2 .......................... 55

Figure 2.9. Total number of C. limbatus collected from A) Branch and Freesoil sites and
B), Harrietta and Mesic sites ..................................................................... 56

CHAPTER 3

Figure 3.1. Mean density of (iSE) of gypsy moth egg masses from four sites in A) red -
oak ecosystems (ELTP 20) and four sites in B) northern hardwood ecosystems (ELTP 45).
Means were calculated from ﬁve 0.1 ha ﬁxed-radius plots within each site ................ 93

Figure 3.2. Means for species richness in A) 1993, B) 1994, and C) 1995. Data from
1993 and 1994 were backtransformed and error bars depict 95% conﬁdence intervals.
Error bars depict standard errors for 1995 data. Means were calculated from n=2 in each
month ............................................................................................... 94

Figure 3.3. Means for species richness in A) 1993, B) 1994, and C) 1995. Error bars
depict standard errors for 1993 and 1994 data. Data from 1995 were backtransfonned and
error bars depict 95% conﬁdence intervals. Means were calculated from n=2 in each
month ............................................................................................... 95

Figure 3.4. Hierarchical clusters created using beta-ﬂexible linkage (B=-0.25) and
Sorcnson’s quantitative index of similarity for total lepidoptera collected from eight sites
during A) 1993, B) 1994 and C). 1995 ........................................................ 96

Figure 3.5. Hierarchical clusters created using beta-ﬂexible linkage (B=-O.SO) and
Sorenson’s quantitative index of similarity for total lepidoptera collected from eight sites
during A) 1993, B) 1994 and C). 1995 ......................................................... 98

xvi

Figure 3.6. Hierarchical clusters created using average linkage and Sorenson’s quantitative
index of similarity for total carabids collected from eight sites during A) 1993, B) 1994 and
C). 1995 ............................................................................................ 100

Figure 3.7. Species richness of noctuids collected during A) early season months of May
and June and B) late season months of July and August from red oak sites (Branch and
Freesoil) and northern hardwood sites (Harrietta and Mesick) over three years ........... 102

Figure 3.8. Total abundance of noctuids collected during A) early season months of May
and June and B) late season months of July and August from red oak sites (Branch and
Freesoil) and northern hardwood sites (Harrietta and Mesick) over three years ........... 103

Figure 3.9. Abundance of overstory noctuids collected during A) early season months of
May and June and B) late season months of July and August from red oak sites (Branch
and Freesoil) and northern hardwood sites (Harrietta and Mesick) over three years ...... 104

Figure 3.10. Abundance of understory noctuids collected during A) early season months
of May and June and B) late season months of July and August from red oak sites (Branch
and Freesoil) and northern hardwood sites (Harrietta and Mesick) over three years ...... 105

Figure 3.11. Abundance of oak feeding noctuidae collected over three years during A)
early season months of May and June and B) late season months of July and August from
defoliated and undefoliated red oak sites. Branch sites were heavily defoliated in 1993,
while Freesoil sites were moderatelt defoliated in 1994 and heavily defoliated in 1995... 106

Figure 3.12. Abundance of oak feeding Noctuidae excluding Hypostrotia pervetrens
collected over three years during late season months of July and August from defoliated
and undefoliated red oak sites. Branch sites were heavily defoliated in 1993, while
Freesoil sites were moderatelt defoliated in 1994 and heavily defoliated in 1995 .......... 107

CHAPTER4

Figure 4.1. Mean density of (iSE) of gypsy moth egg masses from four sites in A) red
oak ecosystems (ELTP 20) and four sites in B) northern hardwood ecosystems (ELTP 45).
Means were calculated from ﬁve 0.1 ha fixed-radius plots within each site ................ 157

Figure 4.2. Relative number of arthropods within six feeding guilds collected from four
northern hardwood and four red oak sites from 1993 to 1995 ............................... 158

Figure 4.3. Relative proportions of arthropod families within six feeding guilds collected
from four northern hardwood and four red oak sites from 1993 to 1995 ................... 159

xvii

SPP

LIST OF SYMBOLS, ABBREVIATIONS, AND NOMENCLATURE

Bacillus thuringiensis var.
kurstaki

centimeter

diameter breast height
ecological classiﬁcation
system

ecological land type phase
hours

hectare

leaf area index

landtype association
meter

nitrogen
Nucleopolyhedrous virus
photosynthetically active
radiation

standard error

species

total light interception
ultraviolet

xviii

 

INTRODUCTION

In terms of species richness, insects are by far the most abundant group on Earth.
The number of described species of insect have been estimated at over 750,000, distantly
followed by multicellular plants (248,428) and non-insect arthropods ( 123,161) (Wilson
1985). The total number of described and undescribed insect species present on Earth has
been estimated at 30 million (Erwin 1982). Much of the diversity of insects is concentrated
in forests (Erwin 1982).

In light of the tremendous abundance of species, it is not surprising that insects play
a critical role in forest ecosystem processes. Insects can be sensitive to changes in
ecosystems and can be used to evaluate effects of disturbances such as forest management
and land conversion (Kremen et al. 1993). Insects are intertwined in food webs where
endemic populations of herbivorous insects are regulated by higher trophic consumers such
as predatory and parasitoid insects (Hairston et al. 1960; Lawton 1986; Price 1987).
Decomposition of woody material and detritus is facilitated by microarthropods and bark
and wood boring insects (Schowalter 1991). Outbreaks of herbivorous insects have been
shown to reduce tree growth and increase tree mortality (Kulman (1971), alter the relative
composition of stands (Attiwill 1994; Schowalter 1981; Schowalter 1985), increase
productivity (Mattson and Addy 1975), and increase nutrient cycling (Risley and Crossely
1993) through defoliation. Outbreaks of defoliating insects have been shown to decrease
diversity and abundance of other insects in West Virginia (Sample et al. 1993) presumably

by reducing the amount of foliage available for food, shelter, and oviposition sites.

Insect outbreaks represent a disturbance that can directly alter ecosystem processes,
but may also indirectly alter ecosystem processes by altering diversity. Diversity has been
shown to have a direct effect on ecosystem processes such as resistance to disturbance and
primary productivity (T ilman 1994; Tilman 1996; Tilman and Downing 1994). Insect
outbreaks represent a disturbance of both intermediate spatial and temporal scales (Risser
1995). Examination of diversity during disturbances of intermediate scales is appropriate
because interactions between diversity and ecosystem processes are greatest (Risser 1995).
Invasion of exotic organisms, particularly animals, also provides an opportunity to examine
ecosystem processes because invaders alter the trophic structure of the native community
(V itousek 1990).

The gypsy moth (Lymantria dispar L.), an exotic forest insect, was responsible for
defoliating over 10 million ha of forest between 1985 and 1994 in the northeastern United
States (Butalla 1996). Gypsy moth will feed on over 400 North American woody plant
species including economically valuable species such as Quercus and Populus spp.
(Mauffette et al. 1983). Preferred tree species such as Quercus experienced 30% mortality
after an average of 37% defoliation over 10 years in New England (Baker 1941). Trees
that have been stressed by defoliation can be killed by invasion of secondary mortality
agents like two-lined chestnut borer (Agrilus bilineatus Weber) and fungal pathogens like
Armillaria root diseases (W argo 1977).

In addition to altering the relative composition of forests, gypsy moth defoliation
can change nutrient and hydrological patterns. While nitrogen in frass deposited by gypsy
moth is immobilized by soil microbes, insect bodies and leaf fragments may be a source of
nitrogen loss through leaching (Lovett and Ruesink 1995). Defoliation can also reduce
transpiration and increase soil moisture, resulting in increased denitriﬁcation (Lovett and
Ruesink 1995). Gypsy moth increased water yield and fecal coliforrn and streptococci
concentrations in heavily defoliated watersheds in Pennsylvania during 1984 and 1985

(Corbett and Lynch 1987).

The effects of gypsy moth on native forest insects are relatively unknown,
especially in the north central region of the United States. Low levels of defoliation
decreased abundance of native lepidoptera in West Virginia, but speciﬁc data was not
reported and the lack of high levels of defoliation makes interpretation of the effects of
gypsy moth difﬁcult (Sample et al. 1993). Heavy defoliation was associated with reduced
numbers of predatory insects, but neither defoliation or number of insects collected were
quantiﬁed in West Virginia (Muzika 1994).

With the current emphasis on ecosystem management, practical and quantitative
information about diversity of an area is needed before proper management strategies can
be applied (Anonymous 1991). Studies evaluating diversity at regional scales are needed
before political, economic, and biological concerns can be integrated into effective
guidelines and policy for management of public forests (Probst and Crow 1991; Wilcove
1989).

The purpose of this research was two-fold. The ﬁrst objective was to obtain
baseline information on the abundance, diversity and community structure of native
arthropods from two forest types that varied in productivity, tree species composition, and
susceptibility to the gypsy moth. The second objective was to quantify the effects of
defoliation by gypsy moth on native insect abundance, diversity and community structure
within a susceptible forest ecosystem.

Diversity of organisms can be deﬁned at many scales, varying from the level of
genetic diversity of a population to the diversity of many species from many trophic levels
across a landscape (Probst and Crow 1991; Risser 1995). In this study, I have chosen to
evaluate native insect diversity at the species and community level in northern hardwood
and red oak stands in northern lower Michigan. I have presented this study as separate
chapters to reﬂect relevant ecological processes that may occur during an outbreak. In
chapter 1, I have tested the utility of a portable light sensor to quantify defoliation in terms

of leaf area index (LAI) during an outbreak of gypsy moth. Chapter 2 examined the effect

of gypsy moth on carabid beetles, a group of predators representative of higher trophic
level consumers as well as a group useful as indicators of ecosystem change. The impact
of defoliation and resource depletion on native lepidoptera, a group likely to be in
competition with gypsy moth, was analyzed in chapter 3. The overall effect of gypsy moth
on the community structure of insect guilds at multiple trophic levels was evaluated in

chapter 4.

 

Uh

Chapter 1
QUANTIFYING GYPSY MOTH (LEPIDOPTERA:LYMANTRIIDAE)
DEFOLIATION IN TWO HARDWOOD ECOSYSTEMS USING A
PORTABLE PAR SENSOR

11111321111511”

Severe defoliation by forest pest insects can have great economic and ecological
impacts on commercial and public forests. Defoliation caused by insects can result in
growth loss, tree mortality (Gross 1992; Kulman 1971), increased risk of mortality from
secondary pests and pathogens (Leonard 1974; Mallett and Volney 1990; Wargo 1977;
Wargo and Montgomery 1983) on a variety of tree species. By decreasing the amount of
canopy cover, defoliation can alter the abiotic and biotic characteristics of a stand.
Increased light interception on the forest ﬂoor can increase temperature and reduce available
water through evaporation (Klein and Perkins 1988; Perkins et a1. 1987; Spurr and Barnes
1980). Plant species normally absent or repressed in‘an undefoliated stand may be able to
invade and compete with the shade tolerant ﬂora (Klein and Perkins 1988).

Gypsy moth (Lymantria dispar L.) is a major forest defoliator in the northeastern
United States. This exotic pest feeds on over 400 North American woody plants including '
economically valuable species such as Quercus spp. and Popular spp. (Mauffette et al.
1983) and was responsible for over 10.4 million ha of defoliation in the United States

between 1985 and 1994 (Butalla 1996).

Obtaining efﬁcient, accurate, and consistent assessments of gypsy moth defoliation
can be problematic. Methods of estimating defoliation such as visual observation can be
inaccurate, subjective and biased due to inter-observer variation in ability and experience
(Cooper et al. 1987; Talerico 1981). Although extensive training and repetitive testing of
observers can increase precision and accuracy of defoliation estimates (Millers and
Lachance 1988), such resources may be unavailable or observer accuracy may vary among
cover types. Use of canopy photography with a ﬁsh eye lens (Liebhold et al. 1988) and
point intercept sampling using a sighting tube (Cooper et al. 1987) have been proposed as
alternatives to visual estimates, but have not been widely used.

A consistent and accurate method for quantifying defoliation is needed to provide a
standardized measure in stands that experience partial defoliation. Such a method should
be accurate, repeatable, provide consistent results, and be applicable to a variety of spatial
scales and canopy species.

A portable light meter called a ceptometer (Decagon Devices, Pullman WA.) has
been used to estimate leaf area index (LAI) (Pierce and Running 1988). The ceptometer
measures photosynthetically active radiation (PAR). It is sensitive to phenological and
seasonal changes in LA] (Vose and Swank 1990) and has been recommended for
evaluating primary productivity (Pierce and Running 1988). In this study, we examined
the ability of the ceptometer to quantify defoliation in oak-dominated sites and the relative
ability of the ceptometer to detect limited or patchy defoliation in northern hardwood

stands.

Mﬂhnsls

Study Area
This research was conducted in eight stands in the Manistee National Forest in

northern lower Michigan. Manistee National Forest has been described using an ecological

_ PJ'- ‘1

 

classiﬁcation system (ECS) (Cleland et al. 1993). The ECS is a hierarchical classiﬁcation
that groups ecosystem components at spatial scales ranging from the landscape to the stand
level. Ecological landtype phases (ELTPs) are classiﬁed on the basis of soils, landscape
position, and natural vegetation. ELTPs are subsets of larger spatial scale units such as
ecological landtypes (ELTs) and landtype associations (LTA). Because stands within the
same ELTP will have similar overstory and understory vegetation, soil hydrology, soil
types, nitrogen cycling, productivity and geological history, use of the ECS allows
ecosystems to be replicated in experimental designs with a high level of conﬁdence that
experimental plots are similar.

We randomly chose four sites classiﬁed as ELTP 20 and four sites classiﬁed as
ELTP 45 from maps provided by the Manistee National Forest. All sites were between
approximately 12 to 16 ha in area. ELTP 20 was characterized by an overstory dominated
by northern red oak and white oak (Quercus alba L.) and a non-diverse ground ﬂora
(Cleland et al. 1993). Soils on ELTP 20 were typically sandy, xcric soils with low
productivity (Host et al. 1987; Host et al. 1988). ELTP 45 had an overstory dominated by
northern hardwood species such as sugar maple (Acer saccharum Marsh) and American
basswood (Tilia americana L.) (Cleland et al. 1993). This ELTP was notable for its rich
herbaceous understory, well drained mesic soils, and high productivity (Host et al. 1987;
Host et al. 1988; Zak et al. 1989; Zak and Pregitzer 1990; Zak et al. 1986).

Two oak-dominated sites (ELTP 20) were located near the town of Branch (44° 00’
N, 86° 00' W) and two other sites were located near the village of Freesoil (44° 08' N,
86° 09' W). Although northern red oak and white oak dominated the overstory, big
toothed aspen (Populus grandidentata Michx.) and red maple (Acer rubrum L.) together
comprised up to 30% of the canopy in each of the four stands. Overstory trees were
approximately 20 m tall. A sparse understory of red maple saplings and witch hazel,

Hamamelis virginiana L., of diameter breast height (DBH) less than 3 cm was also present.

 

 

Two northern hardwood sites (ELTP 45) were located near the town of Mesick
(44° 22' N, 86° 44' W) and two similar sites were located near the village of Hanietta
(44° 19' N, 86° 44' W). Northern hardwood species including sugar maple and American
basswood dominated the overstory. Overstory trees were approximately 30 m tall, and

sites had a rich understory of 1 m tall sugar maples with a DBH of less than 3 cm.
Estimation of gypsy moth populations

The size of gypsy moth populations was estimated using mean egg mass counts
from four to ﬁve 0.01 ha plots within each site (Kolodny-Hirsch 1986). Percentage of
current-year egg masses was determined by inspecting all egg masses between ground level
and 2 m high. Ratio of current-year egg masses to total egg masses was determined and

multiplied by the total number of egg masses counted in the plot.
Estimation of Defoliation

Defoliation was visually assessed within each site by the same observer to eliminate
potential inter-observer variance. Visual estimates were expressed as ﬁve classes of 0-
20%, 21-40%, 41-60%, 61-80%, and 81-100% defoliation.

Photosynthetically active radiation (PAR) (400-700 nm) was measured under the
canopy in each site using a "Sunﬂeck Ceptometer" (Model SF-80), a light sensor
manufactured by Decagon Devices, Inc. The ceptometer had eighty individual sensors
along a 1 meter wand. Mean light measurements were stored in an attached memory
storage unit. PAR below the canopy was compared to PAR measurements taken in
adjacent open areas and used to determine total light interception values (TI). TI is deﬁned

as the light interception by trunks, boles, branches and leaves. T1 was calculated using a

 

 

derivation of the Beer-Lambert law which incorporates correction factors for time of day.

date, and diffusion of light (Decagon Devices 1987). T1 was expressed as the relation:

PAR under canopy * Constant

 

TI: -ln
( PAR above canopy

The interception value of non-leaf portions of trees (i.e.. the trunks, boles and branches)
was subtracted from the TI to get the LAI for foliage alone. To obtain interception values
of non-leaf portions, PAR readings were taken in the northern hardwood stands on 6 April
1995 and in the red oak stands on 26 April 1995, after all snow had melted but before
budbreak.

PAR samples were taken at 45° angles until a complete 360° was sampled for a
total of 8 measurements. Readings were taken by holding the ceptometer level, parallel to
the ground, at approximately 1 M above the forest ﬂoor. When understory saplings were
present, readings were taken by holding the ceptometer level, parallel to the ground, above
the shrub foliage. Readings were then averaged and stored in the ceptometer memory.
This was repeated 10 times at 10 m intervals along a transect radiating from the center of
the site. Four transects, running north, south, east and west, were sampled in this manner.
This resulted in a total of 2560 individual measurements, stored as 320 average values,
collected in each stand on a single sampling period.

During 1994, visual estimates of defoliation and PAR were measured in northern
hardwood sites on 21 June, 28 July, and 16 August and in the red oak sites on 13-14 June,
12 July, and 9 August. During 1995, visual estimates of defoliation and PAR were
measured in the northern hardwood sites on 24 May, 15 June, 19 July, and 21-22 August
and in the red oak sites on 17 May, 13 June, 11 July, and 16 August. Sample dates
roughly corresponded to signiﬁcant points in gypsy moth and stand phonology. PAR was

measured during the typical time of Bacillus thuringensis Berliner var. kurstaki application

 

 

10

for gypsy moth suppression in northern lower Michigan (May), peak fourth instar
develOpment (June), peak defoliation (July), and following stand refoliation (August). All
sampling was conducted between 1000 and 1400 h. This was done to minimize artifactual
sampling of trees outside the site because of the small angle of the sun relative to the plot at
early and late hours of the day. PAR was measured on either uniformly sunny days or
overcast days, but not on partly sunny days due to the difﬁculty in obtaining consistent

PAR measurements within and outside the stand (Decagon Devices 1987).

Statistical Analysis

Differences in mean density of egg masses between the red oak and northern
hardwood ecosystems for 1994 and 1995 were analyzed separately by year using the
Mann-Whitney U-test (Conover 1971) because values were non-normally distributed and
heteroscedastic. LAI values collected from red oak sites in 1995, and northern hardwood
sites in both years, were normalized using the Box-Cox transformation (Sokal and Rohlf
1995). Differences in LAI between sites were analyzed separately within ecosystems and
year using a one-way AN OVA with repeated measures to account for monthly variation.
Differences in LAI between non-defoliated red oak stands and northem hardwood stands
were analyzed within each year using the Mann-Whitney U-test (Conover 1971). All
monthly observations were used in the nonparametric ecosystem comparison, resulting in 6
observations from red oak sites and 12 observations from northern hardwood sites. All

statistical tests were conducted using or =0.05.

Results

Mean gypsy moth egg mass densities were signiﬁcantly higher in red oak
ecosystems than in northern hardwood ecosystems in both 1994 (x2: 4.39, df=l, p<

 

11

0.04) and 1995 (752: 10.33, df=1, p<0.001) (Figure 1.1). In the red oak ecosystem in
1994, Branch sites had more egg masses/ha than Freesoil, but this difference was not
statistically signiﬁcant (x2: 1.13, df=1, p<0.29) (Figure 1.1). In 1995 however, egg
mass densities were signiﬁcantly higher at Freesoil sites than at Branch sites (x2: 12.66,
df=1, p<0.004) (Figure 1.1). Differences between the northern hardwood sites were not
signiﬁcant in 1994 (x2: 0.28, df=1, p<0.60) but were signiﬁcant in 1995 (12: 5.62,
df=1, p<0.02) (Figure 1.1). During 1995, differences in mean egg mass density between
northern hardwood sites was small (250.3 egg masses/ha) compared to differences
observed in red oak sites (11,540.4 egg masses/ha).

Visual estimates indicated generally higher amounts of defoliation in the red oak
sites than in the northern hardwood sites (Table 1.1). Spatial distribution of defoliation
within sites varied between ecosystems. Red oak sites were dominated by many preferred
host plant species of the gypsy moth and most trees were defoliated during outbreaks.
Within the redoak ecosystem, Freesoil sites experienced more defoliation than Branch sites
in both years (Table 1.1). In the northern hardwood sites where preferred species were
scattered, defoliation was only observed on American basswood. This created a patchy
distribution of defoliation within these sites.

Mean LAI for June, July, and August were signiﬁcantly greater in northern
hardwood ecosystems than in undefoliated red oak sites in both 1994 (x2: 11.37, df=1,
p<0.0007) and 1995 (x2: 4.25, df=1, p<0.04) (Figure 1.2). Mean LAI determined
monthly for undefoliated red oak sites during June, July and August ranged from 2.18 to
6.71 in 1994 and 4.51 to 4.85 in 1995 (Figure 1.3). Mean LAI determined monthly for
northern hardwood sites during June, July and August ranged from 10.17 to 12.89 in 1994
and from 4.13 to 11.89 in 1995 (Figure 1.4).

Comparisons of LAI between red oak sites showed LAI initially increased in both
sites as leaves expanded. In July, LAI was reduced in defoliated sites when defoliation

was severe (Figure 1.3). In 1994, a signiﬁcant defoliation by month interaction was

12

observed (F=10.31, p<0.03) (Table 1.2). During June 1994 when early instar larvae were
feeding, the difference in LAI between sites was small. However, Freesoil sites had
signiﬁcantly less LAI than Branch sites in July and August when defoliation peaked
(Figure 1.3). In 1995, the effects of defoliation were not statistically signiﬁcant (F=3.77,
p<0.19). However, when the month of June was excluded from the analysis, Freesoil
sites had lower LAI than did Branch sites (F=l6.40, p<0.055) (Table 1.2) although the
difference was marginally insigniﬁcant. No difference in LAI occurred in June 1995 when
early stage gypsy moth larvae were beginning to feed. In July 1995 during peak
defoliation, LAI in the undefoliated Branch sites was almost 3 times higher than in Freesoil
sites, where gypsy moth population were high (Figure 1.3). Defoliated red oak sites had
not refoliated by 15 August in 1994 or 1995.

Within the northern hardwood ecosystem, a signiﬁcant site by month interaction
was observed in 1994 (F=35.67, p<0.003) (Table 1.3 and Figure 1.4), but differences
_ between northern hardwood sites were relatively smaller compared to differences observed
in the red oak sites (Figure 1.3 and 1.4). Only monthly variation was signiﬁcant in 1995
(F=164.16, p<0.0001).

D' .

EcoSystem Differences

Our results indicating that red oak ecosystems have canopies with lower LAI than
northern hardwoods ecosystems are consistent with the relative differences in productivity
and LAI reported previously in the literature (Bolstad and Gower 1990; Burton et al. 1991;
Waring 1983). However, the magnitude of differences we observed was comparatively

large. While estimates for red oak sites were similar to previous studies (Bolstad and

 

I__——a._oI-h.' I I.

l3

Gower 1990), our estimates of maximum LAI (12.88 in 1994 and approximately 12.00 in
1995) in the northern hardwood sites greatly exceeded those of Burton et al. (1991) who
estimated a maximum LAI of 8.3 using PAR transmittance.

Plausible explanations for the larger estimates we obtained could include differences
in sensitivity of the ceptometer under sunny and cloudy conditions. Under cloudy
conditions, PAR is diffuse and "above canopy" readings are dependent on the size of the
clearing where the reading was taken (Huston and Isebrands 1995). Furthermore, the
ceptometer was shown to overestimate LAI under low light conditions in stands with high
LAI such as Picea abies (Bolstad and Gower 1990). This could account for some of the
variability in PAR readings we observed in northem hardwood sites.

The high variability of monthly readings in northern hardwood stands during 1995
makes seasonal variation in LAI difﬁcult to interpret. Low LAI in the month of May was
attributed to new leaves not yet fully expanded. The decrease in LAI in July, the usual time
of peak defoliation by gypsy moth in northern lower Michigan, may be attributed to
defoliation or leaf drop not detected visually. This suggests the ceptometer could be useful
in detecting low level defoliation or defoliation at a small spatial scale within the stand.

However, this hypothesis should be interpreted cautiously. If defoliation caused
the decrease in LAI, we would expect to see LAI maintained at approximately 4.5 unless
the defoliated trees fully refoliated by August Trees rarely refoliate unless they experience
heavy defoliation. We would also expect to see differences in LAI between the northern
hardwood sites related to the relative density of gypsy moth populations. Instead, we saw
a decline in LAI in July followed by an increase in LAI in August in both sites, regardless
of the density of gypsy moth. It is also possible that the variable pattern of monthly LAI
was attributable to defoliation caused by other leaf feeders and pathogens that we did not
observe.

In a previous study, when methods of LAI estimation based on PAR transmittance

were compared to allometric and litterfall methods, PAR methods yielded results more

14

consistent with litterfall than allometric methods (Burton et al. 1991). Our results were
more consistent with the allometric results obtained by Burton et al. (1991). However,
while Burton et al (1991) used the Beer-Lambert law directly with extinction coefﬁcients
calculated from litterfall data, we used the derivation of the Beer-Lambert law described
previously. This difference in calculation may have accounted for our relatively high LAI

estimates for northern hardwood sites.

Effects of Gypsy Moth

Within the red oak sites, gypsy moth defoliation reduced cumulative LAI (sum of
all monthly values of LAD by 39% in 1994 and 42% in 1995 at Freesoil when compared to
Branch. This corresponds with the increasing amount of defoliation obtained from visual
estimates between 1994 and 1995. However, the ceptometer detected the effects of
defoliation only during peak defoliation in July and August, not while early larvae were
feeding in June. This suggests that the ceptometer was most useful in quantifying
defoliation at a broad spatial scale, when defoliation was moderate to severe.

Our results are consistent with other studies that have used PAR transmittance
methods to examine the effects of gypsy moth defoliation. Liebhold et al. (1988) observed
a signiﬁcant decrease in LAI between June and peak defoliation in late July using inverted
canopy photography to estimate LAI.

While the ceptometer appears to be limited in its sensitivity to spatial scale and
defoliation intensity, it still may be a useful tool for quantifying the effects of defoliation.
In addition to eliminating biases inherent in visual estimation, the ceptometer expresses
defoliation as a reduction in LAI, which is readily comparable to physiological processes
within defoliated trees. Using relationships between defoliation and growth loss in oak
species previously reported in the literature (Baker 1941; Kulman 1971; Minott and Guild
1925) it may be possible to estimate growth loss using LAI reduction. Plant defensive

15

responses to defoliation such as delayed induced resistance could be evaluated in terms of

damage thresholds expressed as reduced LAI.
The ability to predict growth loss due to forest defoliators could be useful to land

managers not only for economic assessment of a pest outbreak but also in decision models
for pest management. Growth loss incurred from defoliation could be quantiﬁed and
losses readily compared with costs of pesticide application or biological control to

determine an appropriate management strategy.

16

Table 1.1. Visual estimates of peak defoliation from two locations in red
oak ecosystems and two locations in northern hardwood ecosystems.

 

1995

 

 

l 9 9 4
Wk
Branch < 20%
Freesoil 40-60%
W
Harrietta < 20%

Mesick < 20%

< 20%
80-100%

< 20%
< 20%

 

17

Table 1.2. Split-plot ANOVA table for leaf area index (LAI) with two
levels of defoliation (severe or none) over two or three months in red oak
sites during 1994 and 1995.

 

1534 19951

 

df Mean F—value p value W Mean F -value p value
Square Square

Defoliation 1 9.19 1.293 0.3733 1 3116.84 16.404 0.0559
Error a 2 7.10 2 190.00

Month 2 13.34 59.273 0.0011 1 0.23 0.001 0.9762
Defoliation 2 2.32 10.305 0.0264 1 36.95 0.179 0.7133
x Month

Error b . 4 0.23 2 206.30

v—-z._

 

 

lValues transformed using the Box-Cox transformation.

Table 1.3. Split-plot ANOVA table for leaf area index (LAI) analyzing
between site differences (Harrietta and Mesick) over three months in
northern hardwood sites during 1994 and 1995.

 

 

 

19941 19951
df Mean F -value p value (If Mean F—value p value
Square Square
Site 1 4.0455 52.797 0.0184 1 695.17 0.630 0.5106

Error a 7659.09 1 103.84

2 2
Month 2 3.4555 58.993 0.001 1 2 1.7355 164.158 0.0001
2 2 4965.23 4.709 0.0889

Sitex 2.0855 35.667 0.0028

Month

4:.

Error b 4 5841.94 1054.50

 

lValues transformed using the Box-Cox transformation.

18

 

 

 

 

 

 

 

 

 

 

II]

m A A

f: 16000 .. A. -

a

E 1 '

E 12000 . - 0 Branch 1
an 0 Branch 2
on if '

3

>. 8000 - -

’5 n Freesoil l
g ' o Freesoil 2
g 4000 .. i-

E

ii I 9 °

= 0 . 1...:

3 1994 1995

2 Year

III

a: 1 . .

‘H 16000 d B. ’-

’3

E

«2’ 12000 - '. Harrietta 1
e0 ,0 Harrietta 2
an

3

>. 80001 -

’2' _a Mesickl
g o Mesick 2
a 4000 - -

a

E

00

25° 0 ‘ n _ P 0 M

g 1994 1995

0

2 Year

Figure 1.1. Mean density (iSE) of gypsy moth egg masses from four sites
in A) red oak ecosystems (ELTP 20) and four sites in B) northern
hardwood ecosystems (ELTP 45). Means were calculated from five 0.1 ha
ﬁxed-radius plots.

‘19

 

 

 

 

 

l3 ‘
121 0 r
11 - o - .
a ‘ ' Undefolrated
a 10 -: fRed Oak
: 9‘. a F. Branch
3 s- 6 -
g 7 -: {Northern
g 5J [Hardwood
5 .. . - El Harrietta
4.‘ ' Lo Mesick
3 5 -
2 . i
1994 1995
Year

Figure 1.2. Mean LAI (:l:SE) from June, July, and August for undefoliated
red oak sites (Branch) and northern hardwood sites (Harrietta and Mesick)
during 1994 and 1995. Means were calculated from n=2 for each site.
Error bars were small and may be obscurred by data points.

20

 

 

8 . . 1 v f . .
’ +Branch

7' -o—Freesoil
E16: '
85. -
2
Q4:- -
S
03- .
2

2- .

1- -

0

 

 

Jun-94 Jul-94 Aug-94 May-95 Jun-95 Jul-95 Aug-95
(163-164) (193) (221) (137) (164) (193) (221)

Date (Julian Day)

Figure 1.3. Comparison of mean LAI (:l:SE) from defoliated (Freesoil) and
undefoliated (Branch) sites in red oak ecosystems during 1994 and 1995.
Means were calculated each month from two sites located near Branch and
two sites located near F reesoil.

21

 

14 - a .
* + Harrietta
12 - —o— Mesick '
a . -
U) 10
:11 .
I-t .. -1
< 8
1.1 . J
g 6 - -
u D
2
4 - -
p
2 - .

 

 

 

0 A A n 4 n A A
Jun-94 Jul-94 Aug-94 May-95 J un-95 Jul-95 Aug-95
(172) (209) (228) (144) (164) (200) (232-233)

Date (Julian Day)

Figure 1.4. Comparison of mean LAI (:l:SE) from two northern hardwood
sites during 1994 and 1995. Means were calculated each month from two
sites located near Harrietta and two sites located near Mesick.

 

Chapter 2
THE IMPACT OF GYPSY MOTH (LEPIDOPTERA: LYMANTRIIDAE)

 

OUTBREAKS ON CARABID BEETLE (COLEOPTERA: CARABIDAE) 3'
DIVERSITY IN TWO HARDWOOD ECOSYSTEMS IN NORTHERN :1
LOWER MICHIGAN ;

Introduction .

Sensitive suites of species are useful as indicators of ecosystem health in
conservation planning and land management (Kremen et al. 1993). Useful indicator
species are deﬁned as well-known taxa that are easily measured, sensitive to environmental
change, widely distributed, and able to provide an assessment of ecosystem properties over
a wide range of stress (Kremen et a1. 1993; Noss 1990). Ground beetles (Coleoptera:
Carabidae) satisfy these criteria in that they are sensitive to habitat fragmentation, well-
known taxonomically, and vary locally across regional landscapes (Niemela et a1. 1988;
Niemela et al. 1990; Niemela et a1. 1993). Carabids have been used as indicators of
ecosystem health across a broad spectrum of disturbance regimes, including impacts of
grassland management techniques in agricultural systems (Erye et al. 1989) and clear-
cutting in forested ecosystems (Lenski 1982; N iemela et al. 1993).

Generally, carabids are considered Opportunistic predators which further
emphasizes their importance in ecosystem monitoring. They have been employed as
biological control agents in efforts to regulate pest insect outbreaks such as spruce

budworrn (Choristoneura funuferana Clem.) and gypsy moth (Lymanm'a dispar L.) (Kelly

22

23

and Regniere 1985). Gut analysis of native carabids in Pennsylvania indicated that 31 of
35 species collected had ingested gypsy moth during an outbreak (Cameron and Reeves
1990). The exotic carabid Calosoma sychophanta L., was widely released in the
northeastern United States in programs to establish biological control of gypsy moth

(Weseloh 1990).
While inventory and observational studies of ground beetles in the Great Lakes

region are available (Jeffords and Case 1987; Leibherr and Mahar 1979; Purrington et al.
1989), little quantitative work has addressed changes of native carabid assemblages during
a gypsy moth outbreak. Severe defoliation can alter light penetration through the canopy
(see chapter 1) and other nricroclirnatic factors. Decreased canopy cover can result in
elevated temperatures (Klein and Perkins 1988; Perkins et al. 1987) and decreased
humidity in the understory (Perkins et al. 1987). Gypsy moth invasion may also alter
biological interactions by acting as an increased pool of prey resource for predatory
carabids and displacing other lepidopteran prey species (see chapter 3).

The purpose of this study was to determine whether large gypsy moth populations
and the resulting severe defoliation altered the abundance and species composition of native
carabids in two ecosystems, one dominated by preferred host plants and the other
dominated by non-preferred host plants of the gypsy moth. Speciﬁcally, we tested the null
hypothesis that carabid abundance and species composition were not effected by gypsy
moth defoliation against the alternate hypothesis that carabid abundance and species

composition were effected by gypsy moth defoliation.

Methods

Study Area

This research was conducted in eight sites in the Manistee National Forest in

northern lower Michigan. Manistee National Forest has been described using an ecological

24

classiﬁcation system (ECS) (Cleland et al. 1993). The ECS is a hierarchical classiﬁcation
that groups ecosystem components at spatial scales varying from the landscape to the stand
level. Ecological landtype phases (ELTPs) are classiﬁed on the basis of soils, landscape
position and natural vegetation. ELTPs are subsets of larger spatial scale units such as
ecological landtypes (ELTs) and landtype associations (LTA). Because sites within the
same ELTP have similar overstory and understory vegetation, soil hydrology, soil type,
nitrogen cycling, productivity and geological history, use of the ECS allows ecosystems to
be replicated in experimental designs with a high level of conﬁdence that experimental plots
are similar.

We randomly chose four sites classiﬁed as ELTP 20 and four sites classiﬁed as
ELTP 45 from maps provided by the Manistee National Forest. All sites ranged from
approximately 12 to 16 ha in area. ELTP 20 was characterized by an overstory dominated
by northern red oak (Quercus rubra L.) and white oak (Quercus alba L.) and a non-diverse
ground ﬂora (Cleland et al. 1993). Soils on ELTP 20 were typically sandy, xeric soils
with low productivity (Host et al. 1987; Host et al. 1988). ELTP 45 had an overstory
dominated by northern hardwood species including sugar maple (Acer saccharum Marsh)
and American basswood (Tilia americana L.) (Cleland et al. 1993). This ELTP was
notable for its rich herbaceous understory, well drained mesic soils, and high productivity
(Host et al. 1987; Host et al. 1988).

Tree species typical of ELTP 20 and 45 also differed in susceptibility to gypsy moth
defoliation (Houston and Valentine 1985; Mauffette et al. 1983; Twery 1990). Oak
species, which dominated ELTP 20, are highly preferred by gypsy moth and oak stands
often experience severe defoliation during outbreaks (Houston and Valentine 1985;
Mauffette et al. 1983; Twery 1990). Northern hardwood stands, such as the ELTP 45, are
less susceptible to gypsy moth defoliation because the dominant canopy species, sugar

maple, is only moderately preferred by gypsy moth (Houston and Valentine 1985;

 

25

Mauffette et al. 1983; Twery 1990). However, highly preferred species such as American
basswood (Mauffette et al. 1983; Twery 1990) were present within this ELTP.

In the red oak ecosystem (ELTP 20), two sites were located at 44° 00' N, 86° 00'
W near the town of Branch and two sites were located at 44° 08' N, 86° 09' W near the
village of Freesoil. In the northern hardwood ecosystem (ELTP 45), two sites were
located at 44° 22' N, 86° 44' W near the town of Mesick and two sites were located at 44°
19' N, 86° 44' W near the village of Harrietta.

Before 1993, none of our study sites had ever experienced any observable gypsy
moth defoliation (F. Sapio, MI DNR, Forest Health Management, 1993 pers com, S.
Katovich, USDA For. Serv. NA. S&PF, Forest Health Management, 1992, pers comm).
We surveyed these sites in 1992 and found no visible defoliation. Density of gypsy moth
egg masses, visual estimates of defoliation, and leaf area index were measured in all stands
as part of a related study (see chapter 1). Red oak sites experienced more defoliation (Table
2.1) and had higher densities of gypsy moth egg (Figure 2.1) masses than the northern
hardwood sites from 1993 to 1995. Red oak sites had overall lower leaf area index (LAI)
than northern hardwood sites, and showed a particular reduction in LAI in areas with large
gypsy moth populations in 1994 and 1995 (see chapter 1) (Table 2.1). Within red oak
sites, gypsy moth populations ﬂuctuated anually: outbreak populations at Branch in 1993
were followed by a population 'crash’ due to NPV infection in 1994. Populations at

Freesoil increased in 1994 and reached outbreak levels in 1995 (Figure 2.1).

Beetle Collection

Carabids were collected in 1993, 1994 and 1995. Beetles were collected four times
each summer. Collection times roughly corresponded to signiﬁcant points in gypsy moth
management or biology. For example, typical application time of the microbial insecticide

Bacillus thuringensis Berliner var. kustaki (Btk) for gypsy moth suppression in northern

 

26

lower Michigan occur in mid-May, fourth instar gypsy moth in mid-June, peak defoliation
in mid-July, and following stand refoliation in mid-August. During 1993, carabids were
collected on 12-13 May, 16 June, 13 July, and 10 August from red oak sites and on 21
May, 22 June, 20 July, and 18 August from northern hardwood sites. In 1994, carabids
were collected on 10 May, 13-14 June, 11-13 July, and 9 August from red oak sites and on
18 May , 20 June, 19-20 July, 15-16 August from northern hardwood sites. In 1995,
collection dates were 15-16 May, 13-14 June, 10 July, 14—15 August at red oak sites and
22-23 May, 19 June, 18-19 July, and 21-22 August at northern hardwood sites.

Carabids were collected using unbaited pitfall traps and ultraviolet (UV) light traps
suspended in the forest canopy. Pitfall traps consisted of plastic containers 11.5 cm in
diameter ﬁlled with soapy water. Each trap was covered by a four-legged 'roof‘ to
minimize capture of non-ground arthropods and to protect the contents from predation by
birds. Ten pifall traps were placed at 10 m intervals along a random transect through each
stand. Traps were open for 24 hours each sampling date.

Ultraviolet traps consisted of a 22 watt UV bulb connected to a photosensor and
powered by a 6 volt battery. A collection funnel and bucket containing insecticidal pest
strips (Vapona TM) were attached below the bulb. The trap was raised into the canopy
approximately 1 to 2 M above the lower edge of the canopy in the center of each stand. UV
traps were activated for approximately 8 hours during each sampling date. Use of a low
wattage bulb and a rain cover over the bulb restricted potential long-distance attraction of
nocturnal insects, resulting in a localized sample of insect populations (Bowden 1982).

These trapping methods were used in combination to sample species present all
above-ground strata and to represent the total ecosystem fauna as completely as possible
given available resources. We recognize trapping biases inherent to these methods, such as
the differential trapping ability of UV traps for carabids capable of ﬂight and the
confounding of abundance and activity measures involved in pitfall trapping (Leibherr and
Mahar 1979). These biases could not be eliminated, but were minimized through use of a

 

27

consistent, standardized sampling period for each trap type in each site (Coddington et al.
1991).

Beetles collected from the UV traps were packed in plastic containers lined with
tissue paper. Beetles collected from pitfall traps were transferred to plastic jars containing
90% ethanol. All insects were stored in a freezer until they could be identiﬁed. Voucher
specimens were prepared and identiﬁed by T. Work at the Center for Biological Diversity at
Michigan State University.

Statistical Analysis

Monthly trap catch was pooled for statistical analysis because the relatively short
duration of trapping was not adequate to calculate meaningful indices of diversity or to
evaluate seasonal changes in carabid species. Numbers of beetles, numbers of species,
alpha log series index, Brillouns‘s and Simpson's indices were evaluated at each site
(Magurran 1988). The alpha log series index was used because it provided an estimate of
diversity based on species richness. Simpson's index was used because it estimates
diversity based on species dominance. Because UV trapping does not sample
independently, Brilloun's index, an index that is used when randomness of a sample is in
question, was also calculated (Magurran 1988; Southwood 1994). A combination of
diversity indicies, species richness and beetle abundance were evaluated because each
particular measure was subject to biases related to sample size and distribution of insects
(Magurran 1988). We believe that a consistent pattern of diversity that is reﬂected by
multiple measurements is a more reliable estimate than an estimate of diversity based on a
single index (Magurran 1988).

Effects of gypsy moth and severe defoliation were examined only in the red oak
ecosystem because of the consistently large gypsy moth populations. Site differences

within the northern hardwood ecosystem were analyzed to provide baseline data on species

28

composition as well as provide an estimation of natural variation through time in sites
without large populations of gypsy moth. Within each ecosystem, differences between
sites and among year variation were evaluated using a one-way AN OVA with repeated
measures (also known as a split-plot AN OVA). Measurements that were not normally
distributed were transformed using the Box-Cox transformation (Sokal and Rohlf 1995).
Analyses were conducted using the software package JMP (SAS Institute Inc.).

Similarity of carabid fauna between ecosystems (beta diversity) were examined
using Soresnon's quantitative index of similarity (also known as Bray-Curtis index)
(Ludwig and Reynolds 1988; Magurran 1988). Multiple dendograms were generated using
Bray-Curtis distance measures and four different linkage methods (Ludwig and Reynolds
1988). Clusters were generated using these alternative linkage methods: beta-ﬂexible
method (0 = -0.25 and -0.50) (Milligan 1989); average linkage (group average/unweighted
pair group); and centroid linkage (Ludwig and Reynolds 1988). Cluster analysis was
performed using BASIC software developed by Ludwig and Reynolds (1988). Because
varying the value of [3 between -0.25 and -0.5 resulted in percent similarity values greater
than 1, values expressed on dendograms have been scaled by dividing all similarity values
by 1.2 and 1.5, respectively.

Results

Ecosystem Differences in Diversity

A total of 47 carabid species were collected in the red oak sites and 38 in the
northern hardwood sites. Of the 17 most abundant species, 11 were consistently caught in
UV traps, while the remaining 6 were consistently caught in pitfall traps (Table 2.2).

Uncommon species (total catch less than 20 individuals) are listed in Table 2.3.

 

29

Dendograms generated using the centroid linkage method were not meaningful
because of the presence of reversals and therefore were not presented. Consistent patterns
were observed in dendograms generated by beta-ﬂexible and group average linkage
methods within each year except in 1994. In 1993, three clusters were apparent within all
dendograms (Figures 2.2, 2.3, and 2.4). Red oak sites that experienced no defoliation
during 1993 (Freesoil 1 and 2) formed a single cluster, as did three northern hardwood
sites (Harrietta l, Harrietta 2, and Mesick l). Defoliated red oak sites (Branch 1 and 2)
clustered with one northern hardwood site (Mesick 2). In 1994, beta-ﬂexible methods

formed two distinct clusters one containing three of the red oak sites and one northern

 

hardwood site and the other containing three northern hardwood sites and one red oak site _
(Figure 2.2 and 2.3). The group average method showed no appreciable pattern in —-
clustering in 1994 (Figure 2.4). In 1995, both beta-ﬂexible methods generated two
clusters; one containing all red oak sites, the other containing all northern hardwood sites
(Figure 2.3). The group average method generated four separate clusters in 1995 (Figure
2.4). The four distinct clusters were formed by defoliated oak sites (Freesoil 1 and 2),
undefoliated oak sites (Branch 1 and 2), three northern hardwood sites (Harrietta 2, Mesick
1, and Mesick 2), and by one northern hardwood site (Harrietta 1). Results of the cluster
analyses for 1993 and 1994 were ambiguous in terms ecosystem differences. However, in
1995, results of cluster analysis indicate that the carabid fauna of the red oak sites were
more similar to one another than they were to northern hardwood sites.
Red oak sites had greater species richness and total abundance than northern
hardwood sites in every year. Twenty-one species were unique to the red oak ecosystem.
Nine of these species are known to have particular habitat requirements with respect to
moisture and light availability (Lindroth 1969). Amara convexa Lee. and Carabus serratus
Say are restricted to xeric habitats. Pterostichus mutus (Say) and P. melanarius (111.) are
commonly found in well-lighted, open forests (Lindroth 1969). Chlaenius tricolor tricolor
Dej. and Pterostichus chaleitis Say are commonly collected in well-lighted ﬂood plain

30

forests (Lindroth 1969). Dyschirius politis (Dej.), Agonum harrisi Lec., and Bembidion
americanum Dej. are known hygrophiles and are commonly found near pond edges
(Lindroth 1969). The remaining 14 species unique to the red oak stands were eurytropic
(Lindroth 1969).

Twelve species were unique to the northern hardwood sites, nine of which were
hygrophagous (Lindroth 1969). No strictly xerophilous species were collected from
northern hardwood stands. Snail predators, including Dicaleus teter Bon., Scaphinotes
bilobus (Say), and Spheroderus stenostanus lecontei Dej. (Lindroth 1969) were found

only in this ecosystem.

Impacts of Defoliation on Carabid Species Richness, Abundance, and

Diversity

Because gypsy moth populations and the amount of defoliation varied between
Branch and Freesoil sites over time, defoliation by year interactions better described the
effects of gypsy moth than main effects alone. Defoliation by year interactions were
signiﬁcant in the alpha log series index (F=6.98 p<0.05) (Table 2.4 and Figure 2.5) and
only marginally insigniﬁcant in species richness (F=6.29 p<0.06) (Table 2.4 and Figure
2.6). In 1993, no differences in carabid species richness or alpha log series index were
observed (Figure 2.5 and 2.6). However in 1994, richness and diversity declined in
Branch sites one year following defoliation at Branch sites. Likewise, in 1995, richness
and diversity increased in Freesoil sites one year following defoliation at Freesoil sites
(Figures 2.5 and 2.6). No signiﬁcant differences between Branch and Freesoil sites were
observed in terms of beetle abundance, Simpson's or Brilloun's indices (Table 2.4 and
2.5). Among year variation was only signiﬁcant in the Brilloun index (F=21.17 p<0.01)
(Table 2.4). Brilloun's index increased each year in the red oak sites and ranged from a

mean value (:l:SE) of 0.58 ($0.04) in 1993 to 1.6 (10.10) in 1995.

31

In 1993 and 1995 when differences in defoliation and size of gypsy moth
populations were greatest, similarity of carabid fauna differed between Branch and Freesoil
sites (Figures 2.2, 2.3, and 2.4). Differences in carbid fauna between Branch and Freesoil
were reﬂected in the number of species and genera collected. Like the number of species,
the number of genera increased one year following defoliation at Freesoil in 1995. Mean
number of genera were greater at Freesoil sites (12.5 21:15) than at Branch sites (7 .5i0.5)
in 1995.

Within the northern hardwood ecosystem, no signiﬁcant site by year interactions
were observed in any measurement (Table 2.6). Differences between northern hardwood
sites were less variable in terms of alpha log series index (Figure 2.7) and species richness
(Figure 2.8) than red oak sites (Figure 2.5 and 2.6). Among year variation was observed
in species richness (F=15.08 p< 0.01), beetle abundance (F=9.19 p<0.03), and Brilloun's
index (F=68.47 p<0.001) (Table 2.5 and Table 2.7)

Response of Carabid Species to Gypsy Moth

While species such as Bembidion partreule Dej., B. vericolor (Lec.), Stenopholus
comma (F .), and Synuchus impunctata (Say) were abundant in one particular year, their
populations appeared to be highly variable and showed no trend with gypsy moth
populations. However, Carabus limbatus Say consistently accounted for at least 10% of
the catch each year in red oak sites.

The relative difference in total abundance of C. limbatus between Branch and
Freesoil suggests that C. limbatus was negatively effected by gypsy moth one year
following defoliation (Figure 2.9). In 1993, the difference in total catch of C. limbatus
between Branch and Freesoil was only 6 beetles. However, in 1994 after severe

defoliation had occurred at Branch the previous year, Branch sites had 17 fewer C.

32

limbatus than did Freesoil sites. In 1995, following defoliation at Freesoil sites during the
previous year, Freesoil had 25 less C. limbatus than did Branch sites.

In northern hardwood sites where gypsy moth populations were low, the relative
differences in total abundance of C. limbatus did not change over three years (Figure 2.9).

Mesick sites always had roughly twice as many C. limbatus than did Harrietta sites.
Discussion ,

Ecosystem Differences I

 

Differences in carabid species composition between northern hardwood and red oak
sites were likely a result of differences in plant diversity and microclirnate between
ecosystems. Leibherr and Mahar (1979) concluded that increased plant diversity creates
heterogeneity of microhabitat and more opportunities for habitat specialists to ﬁnd suitable
conditions for oviposition. Microclimatic components such as temperature and humidity
are directly related to light penetration through the canopy. The amount of light penetration
through the canopy of each ecosystem differed considerably (see chapter 1). Humidity and
light preference are important factors in determining habitat preference of a forest dwelling
carabids (Allen 1976). Likewise, the amount of canopy closure has also been attributed to
the amount of free moisture available in the litter (Leibherr and Mahar 1979) Northern
hardwood sites had a denser canopy and more leaf area index (LAI) than red oak sites (see
chapter 1).

While not directly tested in our study, these results may suggest an association
between the diverse understory ﬂora of the northern hardwood sites and the amount of
hygrophilous carabid species present. Furthermore, the dense canopy of the northern
hardwood sites may have provided a cooler, moister microclimate capable of supporting a

relatively large proportion (75%) of hygrophilic species and specialist species that rely on

33

snails, a moisture-dependent prey animal. Red oak sites had lower LAI and smaller relative
proportion (15%) of hygrophilous species. Species that preferred well-lighted xeric
conditions were present only in red oak stands and comprised 15% of the species unique to
the ecosystem. It is likely that a combination of the effects of understory ﬂoral diversity

and canopy density explain the patterns in carabid species composition that we observed.

Effects of Gypsy Moth and Natural Variation

Because all red oak sites experienced severe defoliation by gypsy moth over the
three years in which this experiment was conducted, comparisons of undisturbed red oak
and undisturbed northern hardwood sites cannot be made, nor does an optimal control
treatment exist for the effects of defoliation in the red oak sites. However, lack of
differences between Mesick and Harrietta in species richness and alpha log series provided
an estimate of the amount of natural variation in richness and diversity of beetles over time.
When red oak and northern hardwood sites are compared in terms of species richness and
alpha log series index, red oak sites varied more than northern hardwood sites. While not
optimal, we believe that the consistent patterns within the northern hardwood sites provided
a useful comparison to assess effects of defoliation on native carabid richness and

diversity.

Implications of Gypsy Moth and Defoliation

Disturbance events have been thought to disrupt processes of competitive exclusion
and result in an increase in species diversity (Connell 1978; Lenski 1982). Gypsy moth
populations and high defoliation may have represented a major disturbance in the red oak
sites and could presumably have both positive and negative effects on native carabid

species. The presence of high populations of gypsy moth may act as an increased prey

34

source for some carabids, while defoliation may alter rnicrohabitat and decrease availability
of other prey items through interspeciﬁc competition with other Lepidoptera and foliage
feeders (see chapter 3).

The increase in species richness and diversity that we observed one year following
severe defoliation could be attributed to exploitation of gypsy moth through predation or

scavenging by carabids. Univoltine carabid species that used gypsy moth as a food

 

resource would presumably be more fecund and show increased abundances in the next a:

generation during the following year. However, responses of individual carabid species to

large gypsy moth populations, and presumably increased prey abundance, were j

inconclusive in our study because of the high variability and small numbers of individuals *
l

collected between years. We did not collect caterpillar-specialist species such as those in
the genus Calosoma in great enough numbers to make conclusions about the effects of
increased prey abundance. However, because carabid food preference in the ﬁeld is largely
unexplored and because carabids are opportunistic feeders, we do not believe the
hypothesis that carabids may be exploiting gypsy moth through predation and scavenging
can be disrrrissed.

Signiﬁcant defoliation by year interactions were not observed in either Brilloun or
Simpson's index. Simspon's index has been characterized as less discriminating than the
alpha log series index because of its dependence on abundance of common species
(Magurran 1988). Similarly, the Brilloun index has been characterized as moderately
indiscrinrinant and dependent on sample size (Magurran 1988). The variability in the
abundance of carabids collected may have accounted for the lack of differences observed
using these indices.

While not directly tested in this study, two hypotheses explaining the increase in
carabid diversity in terms of relaxed competition for food resources during a gypsy moth
outbreak exist (Connell 197 8; Lenski 1982). One hypothesis suggests that carabid species

diversity increased because more gypsy moth were available for consumption. This would

35

result in relaxed competition among species with similar feeding habits, such as members
of the same genus (Lenski 1982) and increased diversity of these sirrrilar feeding-carabids.
Our second hypothesis is that carabids may be exploiting gypsy moth through scavenging.
Scavenging for gypsy moth is a strategy available to all carabids regardless of body size
and predation ability because of the decreased cost of subduing prey. If carabids are
scavenging gypsy moth cadavers, a greater number of potential competitors from different
genera will be able to utilize the inﬂux of gypsy moth as a food resource, and between-
genera diversity should increase.

We observed both an increase in the number of species and genera one year
following a severe defoliation event. This suggests that carabids may be exploiting gypsy
moth as food resource either through predation and or scavenging. This conclusion should
be interpreted conservatively because of the lack of available knowledge of carabid feeding
habits in the ﬁeld and the small sample size. Moreover, information on the acceptability of
gyspy moth cadavers infected with NPV by carabid species is unavailable.

Effects of defoliation were complicated by the confounding effects of decreasing
canopy closure and the increase of potential prey items, e.g. gypsy moth larvae and pupae.
Reduction of LAI by gypsy moth defoliation could augment the trapping ability of the
canopy UV traps by increasing the visual range at which UV light is detected. This could
result in higher estimates of richness, abundance, and diversity in defoliated stands relative
to undefoliated stands during the same year of a defoliation event. However, this would
not be consistent with the increased species diversity we observed one year after a
defoliation event.

The responsive nature of the species and the ease with which large numbers of
individual beetles were collected suggests C. limbatus may be an appropriate candidate for
an indicator species (Kremen et a1. 1993) that could be used to monitor environmental
change in red oak sites. Defoliated stands had a lower LAI and presumably less available

moisture than undefoliated stands (see chapter 1). C. limbatus, a moist forest dweller, was

 

36

less abundant during severe defoliation but recovered to pre-defoliation levels the next
year. Disturbances that altered forest canopy caused similar responses in C. limbatus in a
previous study. For example, C. limbatus was intolerant of clearcut conditions in
Appalachian forests (Lenski 1982). The high sensitivity and rapid response to disturbance
exhibited by C. limbatus is similar to the environmental sensitivity of ground beetles
described in previous studies (Desender and Turin 1989; Dijk 1986; Holliday 1991;
Niemela et a1. 1988; Niemela et al. 1990; Turin and denBoer 1988).

37

Table 2.1. Visual estimation of peak defoliation and estimation of mean
LAI (:l:SE) using photosynthetically active radiation transmittance of sites
in red oak and northern hardwood ecosystems.

 

 

 

 

 

 

 

 

 

 

Peak Visual Estimation of Defoliation

1993 1994 1995
RsrLQak
Branch 80-100% < 20% < 20%
Freesoil < 20% 40-60% 80-100%
mm
Hanietta < 20% < 20% < 20%
Mesick 20-40% < 20% < 20%
Estimated Leaf Area IndexI

1994 1995

Rﬂﬂalt
Branch 4.48 i 0.221 4.71 i 0.115
Freesoil 2.73 i 0.076 2.72 i 0.151
Winged
Harrietta 10.9 i 0.226 8.61 :1: 0.229
Mesick 11.9 :t 0.189 8.14 i 0.222

 

1Means and standard errors were calculated from two sites (n=2)

38

Table 2.2. Abundant carabid species collected using suspended UV traps
and pitfall traps in sites dominated by red oak (R0) and northern hardwood

(NH) species over three years1

 

 

 

 

 

 

1 9 9 3 l 9 9 4 l 9 9 5
Species RD N H RO N H RO N H Total
W
Stenopholus comma (F.) 88 3 227 9 3 151 481
Stenopholus ochropezus (Say) 7 0 159 0 27 4 197
Bembidion versicolor (Lec.) 0 17 5 161 6 190
Stenopholus lineola (F.) 81 2 0 0 1 24 108
Bembidion partreule Dej. 1 0 l9 7 61 2 90
Cymindis platicollis (Say) 0 0 7 20 4 40 71
Agonum placidum (Say) 6 0 6 2 3 28 45
Trichotichnus dichrous (Dej.) 0 0 24 5 3 12 44
Clivinafossor (Lime) 1 0 36 0 6 0 43
Harpalusfaunus Say 0 0 l3 9 9 4 35
Cymindes limbatus (Dej.) 3 1 7 11 3 4 29
WW
Carabus limbatus Say 36 4 73 34 123 75 345
Synuchus impunctatus (Say) 65 30 22 72 0 45 234
Pterostichus pensylvanicus Lec. 0 5 26 66 l l 17 125
Pterostichus sp A 0 0 25 64 0 0 89
Myas cyanescens Dej. l 15 0 23 0 l 1 50
Pterostichus corvinus (Dej.) 10 6 1 3 l 2 23

 

Total 300 66 662 330 416 425

 

lSpecies were termed abundant if more than 20 individuals were
collected over three years.

39

Table 2.3. Uncommon carabid species collected in UV and pitfall traps
from red oak (R0) and northern hardwood (NH) ecosystems over three

years1

 

Species

1993

1994

1995

 

RO NH

RO NH

RO NH

Total

 

WM
Agonum decorum (Say)
Agonum galvestonicum
(Csy-)

Agonum harrisi Lec.
Agonum sordens Kby.
Amara convexa Lec.
Anadaptus discoides Dej.
Bembidion americanum Dej.
Brachinus cordicollis Dej.

Chlaenius tricolor tricolor Dej.

Clivina dentipes Dej.
Dyschirius erythrocerus Lec.
Dyschirius politis (Dej.)
Harpalus lewisi Lec.
Harpalus pensylvanicus
(1360-)

Harpalus ruﬁpes (DeG.)
Lebia atriventalis Say

Lebia ornata Say

Lebia tricolor Say

Lebia viridis Dej.
Pterostichus chalcites Say
Stenopholus conjunctus (Say)
Tachys sp A

Tachys sp B

Tachys sp C

Trechus discus F.

OOOOOHOOONOH

OOOOOOOOOt—O

CO

OOOOOOOOOOOO

OOOOOOOOOOO

O

NOOOHNOD—UIONO

H
NNANv—oo

COCO

O

HMCOCOOOOOOO

OOOOOOOOOO-d

O

MOW—‘OQOOOMv—O

N—‘UJOO—‘H—‘OOh—O

OOOwOOt—OOOOO

OOOOUIO—‘HO-RO

-—--u.\rm—-

18

Nv-‘UJW

Table 2.3. (cont'd)

Eli" Ell IIS .
Calosoma frigidum Kby.
Calosoma scrutator (F .)
Calathus gregarius (Say)
Carabus serratus Say
Carabus sylvosus Say
Cymindis americana Dej.
Cymindis cribicollis Dej.
Cymindis pilosa Say
Dicaelus teter Bon.

Myas coracanus (Say)

Pterostichus adstrictus Eschz.

Pterostichus adoxus (Say)
Pterostichus melanarius (111.)
Pterostichus mutus (Say)
Pterostichus sp

Scaphinotus bilobus (Say)
Sphaeroderus stenostanus
lecontei Dej.

NNOOOOOOOO~

OOUJt-‘OO

40

WNOOOUIQON-hUIOOONOO

y—a
N

—c
ooo-ooeo~ooo~o~

wNOOOOOOwON—‘OOOOO

OOOONOOOOO—‘OOHOON

«Ii—‘OOOOOOOOOOv-‘OOO-d

u—a
O\

OMMOOH—‘WN—

ﬂ
MO

12

 

lSpecies were termed uncommon if less than 20 individuals were collected over three

years.

41

Table 2.4. Statistical signiﬁcance of treatment effects from split-plot
ANOVA of five measures of carabid diversity in four red oak sites.

 

 

Source of Variation

Gypsy Moth
Year

Gypsy Moth x Year

Gypsy Moth
Year

Gypsy Moth x Year

Error a

Error b

Species Abundance Alpha Brilloun's Simpson's

Richness1 Log Index1 Index

Series

F value of significant treatment effects

0.629 3.414 0.008 0.002 0.002
3.806 0.607 2.410 21.169 0.558
6.287 1.623 6.981 1.504 0.457

Probability of significant treatment effects

0.51 10 0.2059 0.9539 0.9702 0.9676
0.1186 0.5887 0.2057 0.0075 0.61 13
0.0582 0.3048 0.0496 0.3257 0.6626

Mean square

0.0006 0.2030 0.9700 0.1450 0.0040
0.0004 0.6440 1.1 160 0.0630 0.0130

 

lValues transformed using the Box-Cox transformation

 

42

Table 2.5. Means (:l:SE) of ﬁve measurements of diversity pooled over
three years from red oak and northern hardwood sites (n=2).

     
   

Freesoil

 
 

Branch

Variable

 

 

Species Richness 12.3 (2.50) 14.2 (2.32) 12.8 (2.21) 10.8 (1.25)

Abundance 129.0 (75.75) 123.3 (14.88) 70.8 (28.92) 29.3 (11.96)
Alpha Log Series 4.14 (0.57) 4.19 (0.76) 4.4 (0.40) 4.2 (0.46)
Brilloun's Index 1.0 (0.17) 1.0 (0.24) 1.0 (0.22) 1.0 (0.19)

Simpson's Index 4.1 (0.72) 4.1 (0.72) 5.5 (0.70) 6.1 (0.93)

 

43

Table 2.6. Statistical signiﬁcance of treatment effects from split-plot
ANOVA of ﬁve measures of carabid diversity in four northern hardwood
sites.

 

Source of Variation df Species Abundance Alpha Brilloun's Simpson's
Richness Log Index Index
Series

F value of signiﬁcant treatment effects

Site 1 1.756 1.122 0.288 0.243 0.074
Year 2 15.077 9.188 1.294 68.467 0.09]
Site x Year 2 3.692 1.235 2.405 0.928 0.039

Probability of signiﬁcant treatment effects

Site 1 0.3162 0.4006 0.6454 0.6711 0.8115
Year 2 0.0137 0.0320 0.3687 0.0008 0.9147
Site x Year 2 0.1234 0.3823 0.2061 0.4665 0.9625

Mean square

6.833 2912.833 0.095 0.037 13.301

N

Errora

Error b 4.333 948.333 0.924 0.017 3.250

4:.

 

 

Table 2.7. Means (iSE) of species richness, beetle abundance, and the
Brilloun index from four northern hardwood sites that varied signiﬁcantly

over three years.

 

 

Variable 1993 1994 1995

 

Species Richness 7.5 (0.65) 12.5 (1.85) 15.5 (1.55)
Beetle Abundance 23.5 (2.53) 87.0 (20.89) 114.5 (28.74)

Brilloun Index 0.6 (0.04) 0.8 (0.03) 1.6 (0.9)

 

45

 

12000 - .
0 Branch 1

f H ' 0 Branch 2

 

 

 

 

Mean egg mass density (egg masses/ha) :t SE

 

 

 

 

 

8000 T ' n Freesoil 1
. - 0 Freesoil 2
4000 " )-
1 ° °
0 — . I ‘ .
1993 1994 1995
Year
Ill . . .
to
+1 16000- B. -
”a?
g
E 12000 ‘ '. Harrietta]
3 . O Harrietta 2
3
g 8000 - -
5 . _ u Mesick 1
E o Mesick 2
E 4000 - ..
90
DD
0
g 0 _ . . . D 0 . . D Q - . f A
E 1993 1994 1995
Year

Figure 2.1. Mean density (iSE) of gypsy moth egg masses from four sites
in A) red oak ecosystems (ELTP 20) and four sites in B) northern
hardwood ecosystems (ELTP 45). Means were calculated from ﬁve 0.] ha
ﬁxed-radius plots

46

Figure 2.2. Hierarchical clusters created using beta-ﬂexible linkage
(13:41.25) and Sorenson's quantitative index of similarity for total carbids
collected from eight sites during A) 1993, B) 1994 and C) 1995.

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A. Percent Similarity
0 10 20 30 40 50 60 70 80 90
I I I I I I I I I J
1993

|

I
B. Percent Similarity
0 10 20 3040506070 8090
I I I I J I I I I I
1994
C. Percent Similarity
0 10 20 30 40 50 60 70 80 90
L I L I I I L I I I
1995

 

 

 

 

 

 

 

 

 

 

 

 

 

Branch 1

Branch 2
Mesick 2

Freesoil 1
Freesoil 2
Harrietta 1
Mesick l
Hanietta 2

Branch 1
Branch 2
Freesoil 1
Harrietta 1
Freesoil 2
Mesick l
Harrietta 2
Mesick 2

Branch 1
Branch 2
Freesoil 1
Freesoil 2
Harrietta 2
Mesick 1
Mesick 2
Hanietta 1

48

Figure 2.3. Hierarchical clusters created using beta-ﬂexible linkage
(B=-0.50) and Sorenson's quantitative index of similarity for total carbids
collected from eight sites during A) 1993, B) 1994 and C) 1995.

49

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A. Percent Similarity
0 10 20 3O 40 506070 80 90100
I l I I I I I I I I
1993 Branchl
Branch 2
Mesick 2
Freesoill
Freesoil 2
* Harrietta]
l I Mesick]
I Hanietta 2
B, Percent Similarity
0102030405060708090100
l l I I I I J I J I
1994 [ Branchl
Branch 2
I ‘ Freesoill
Harrietta]
Freesoi12
Mesick]
‘ I Harrietta 2
Mesick 2
C, Percent Similarity

0 10 20 30 40 50 60 70 8090100

I I I I I I I I I

1995

 

- Branch 1

- Branch 2

 

 

Freesoil 2

J
|__.
|__
I r——- Freesoil 1
l..—
——l: “W“
‘ Mesick 1
J Mesick 2

Harrietta 1

 

 

 

 

 

50

Figure 2.4. Hierarchical clusters created using average linkage (B=-0.25)

and Sorenson's quantitative index of similarity for total carbids collected
from eight sites during A) 1993, B) 1994 and C) 1995.

A.

0 10 20

I I

4

Percent Similarity

30 40 50 60 70

I I I J I

51

80 90

 

1993

0 10 20

I I

1994

C.
0
r J

1995

 

__L_.

 

 

 

 

 

 

 

__l

 

 

 

 

Percent Similarity

3040506070

I I I I I

80 90

 

 

10 20

I

 

 

 

 

 

 

 

 

 

 

 

 

Percent Similarity
30 40 50 60 70

I I i I I

80 90

I

 

 

 

 

 

 

 

——4

 

 

 

__|
_{_‘

 

100

Branch 1

Branch 2
Mesick 2

Freesoil 1

Freesoil 2

Harrietta 1
Mesick 1

Harrietta 2

Branch 1

Harrietta 1
Branch 2
Freesoil 1
Freesoil 2
Mesick 1
Mesick 2

Harrietta 2

Branch 1
Branch 2
Freesoil 1
Freesoil 2
Harrietta 2
Mesick 1
Mesick 2
Harrietta 1

52

 

5-

4 .1 —-0— Branch

—0— Freesoil

2-

Alpha Log Series Index
0

 

 

 

I I 1
1993 1994 1995

Year

Figure 2.5. Means (:l:SE) of the alpha log series index from red oak sites
that experienced differential amounts of defoliation over three years.
Means were calculated from n=2. Standard errors were small during some
years and may be obscured by size of data points.

53

 

 

 

 

 

25
g 20- -
ﬂ
5
E 15- —.— Branch
g —-o-— Freesoil
£- 104
5 I I I
1993 1994 1995
Year

Figure 2.6. Means (:l:SE) of species richness from red oak sites that
experienced differential amounts of defoliation over three years. Means
were calculated from n=2. Standard errors were small during some years
and may be obscured by size of data points.

54

 

 

+ Harrietta

—-o—- Mesick

 

Alpha Log Series Index
A
I

 

 

 

1
1903 1934 1995

Year

Figure 2.7. Means (iSE) of the alpha log series index from Harrietta and
Mesick sites in the northern hardwood ecosystem. Means were calculated

from n=2.

55

 

 

 

 

25
m
3 20-
e:
.1:
.2
a: 15- —o— Harrietta
In
.3 —o— Mesick
0
g- 10-

5 I ’1 I

1993 1994 1995
Year

Figure 2.8. Means (iSE) of species richness from Harrietta and Mesick

sites in the northern hardwood ecosystem. Means were calculated from
n=2.

 

56

 

 

 

 

 

 

 

Total numbers of C. limbatus collected

 

 

 

 

 

   

I Branch
E] Freesoil
1993 1994 1995
Year

3 80
‘5 B.
2
§
,6 60-
a
‘6
a
.5. 40_ I Harrietta
B B Mesick
e
E
0
.e
E
:1
r:
'5
e
p.

1993 1994 1995

Year

Figure 2.9. Total number of C. limbatus collected from A) Branch and
Freesoil sites and B) Harrietta and Mesick sites.

Chapter 3

THE IMPACT OF GYPSY MOTH (LEPIDOPTERA: LYMANTRIIDAE)
OUTBREAKS ON NATIVE LEPIDOPTERA IN TWO NORTHERN
HARDWOOD ECOSYSTEMS IN NORTHERN LOWER MICHIGAN

W

Herbivorous insects play an important role in forest ecosystems. During outbreaks,
herbivores insects can remove up to 100% of available foliage causing reduced tree growth
and tree mortalilty (Kulman 1971). Forest defoliators can alter plant community structure,
nutrient cycling, and regulate primary productivity (Mattson and Addy 1975; Swank ct al.
1981). Foliage allelochemicals can be affected by insect feeding, subsequently affecting
other plant feeders (Schultz and Baldwin 1982).

In the northeastern United States, the gypsy moth, Lymantria dispar L., is a major
forest insect that defoliated over 10 million ha of forest between 1985 and 1994 (Butalla
1996). This exotic pest feeds on over 400 North American woody plant species including
economically valuable species such as Quercus and Populus spp. (Mauffette et al. 1983).
Trees that have been stressed by defoliation can be killed by invasion of secondary
mortality agents like two-lined chestnut borer (Agrilus bilineatus Weber) and fungal
pathogens like Armillaria root diseases (W argo 1977).

57

58

Gypsy moth can alter species composition of a stand by reducing the relative
abundance of oaks or other preferred species. Repeated defoliation from 1911 to 1931 by
gypsy moth not only increased the mortality rates of oaks ﬁve years following defoliation,
but also decreased the relative proportion of oaks within defoliated stands in New England
(Campbell and Sloan 1977). Likewise, relative proportion of oaks declined by 10% in
forests defoliated by gypsy moth in Pennsylvania between 1971 and 1979 (Gansner et a1.
1983).

In addition to altering the relative composition of forests, gypsy moth defoliation
can change nutrient and hydrological patterns. While nitrogen accrued via deposition of
gypsy moth frass is immobilized by soil microbes, insect bodies and leaf fragments may be
a source of nitrogen loss (Lovett and Ruesink 1995). Defoliation can also reduce
transpiration and increase soil moisture resulting in increased denitriﬁcation (Lovett and
Ruesink 1995). Severe defoliation increased water yield and concentration of fecal coliform
and streptococci occurred in watersheds in the northeastern United States (Corbett and
Lynch 1987).

However, effects of gypsy moth on native forest insects are relatively unknown,
especially in the north central region of the United States. One study in West Virginia
showed low levels of defoliation decreased abundance of native lepidoptera species, but
species-speciﬁc data was not reported and severe defoliation did not occur in the area of
study (Sample et al. 1993). In another West Virginia study, gypsy moth defoliation was
associated with reduced numbers of predatory insects, but few quantitative data were
reported (Muzika 1994).

The ﬁrst objective of this research was to obtain baseline information on the
abundance, species richness, and diversity of the native lepidopteran fauna from two forest
ecosystems that varied in productivity, tree species composition, and susceptibility to the
gypsy moth. The second objective was to determine whether gypsy moth invasion affected

native lepidopteran abundance, species richness and diversity within the susceptible forest

59

ecosystem. We hypothesized that gypsy moth would reduce the abundance and diversity

of native lepidoptera, particularly those that feed on oak.

Methods
Study Area

This research was conducted in eight sites in the Manistee National Forest in
northern lower Michigan. Manistee National Forest has been described using an ecological
classﬁcation system (ECS) (Cleland et a1. 1993). The ECS is a hierarchical classiﬁcation
that groups ecosystem components at spatial scales varying from the landscape to the stand
level. Ecological landtype phases (ELTPs) are classiﬁed on the basis of soils, landscape
position and natural vegetation. ELTPs are subsets of larger spatial scale units such as
ecological landtypes (ELTs) and landtype associations (LTA). Because stands within the
same ELTP have similar overstory and understory vegetation, soil hydrology, soil type,
nitrogen cycling, productivity and geological history, use of the ECS allows ecosystems to
be replicated in experimental designs with a high level of conﬁdence that experimental plots
are similar.

We randomly chose four sites classiﬁed as ELTP 20 and four sites classiﬁed as
ELTP 45 from maps provided by the Manistee National Forest. All sites ranged from 12 to
16 ha in area. ELTP 20 was characterized by an overstory dominated by northern red oak
(Quercus rubra L.) and white oak (Quercus alba L.) and a non-diverse ground ﬂora
(Cleland et al. 1993). Soils on ELTP 20 were typically sandy, xeric soils with low
productivity (Host et al. 1987; Host et a1. 1988). ELTP 45 had an overstory dominated by
northern hardwood species particularly sugar maple (Acer saccharum Marsh) and, to a

lesser extent, other overstory species including American basswood (Tilia americana L.),

60

white ash (F raxinus americana L.), red maple (Acer rubra L.) and red oak (Cleland et al.
1993). This ELTP was notable for its rich herbaceous understory, well-drained mesic
soils, and high productivity (Host et al. 1987; Host et al. 1988; Zak et al. 1989; Zak and
Pregitzer 1990; Zak et al. 1986).

Dominant tree species within ELTP 20 and 45 also differed in susceptibility to
gypsy moth defoliation. Oak species, which dominated ELTP 20, are highly preferred by
gypsy moth and oak stands often experience severe defoliation during outbreaks (Mauffette
et al. 1983; Twery 1990). Northern hardwood stands, such as ELTP 45, are less
susceptible to gypsy moth defoliation because the dominant canopy species, sugar maple,
is only moderately preferred by gypsy moth (Mauffette et al. 1983; Twery 1990).

In the red oak ecosystem (ELTP 20), two sites were located at 44° 00' N, 86° 00'
W near the town of Branch and two sites were located at 44° 08' N, 86° 09' W near the
village of Freesoil. In the northern hardwood ecosystem (ELTP 45), two sites were
located at 44° 22' N, 86° 44' W near the town of Mesick and two sites were located at 44°
19' N, 86° 44' W near the village of Harrietta.

Gypsy moth were ﬁrst discovered in the eastern central region of Michigan's lower
peninsula in 1954, but defoliation due was not apparent until 1979 (Gage et al. 1990).
P0pulations slowly radiated westward reaching central Michigan by 1977 (Dreistadt 1983);
signiﬁcant defoliation did not occur in this area until 1987 (Gage et al. 1990). Before
1993, none of our study sites had ever experienced any observable gypsy moth defoliation
(F. Sapio, MI DNR, Forest Health Management, 1993 pers com, S. Katovich, USDA
For. Serv. NA. S&PF, Forest Health Management, 1992, pers comm). We surveyed
these sites in 1992 and found no visible defoliation or evidence of previous infestation.

Number of gypsy moth egg masses/ha, visual estimates of defoliation, and leaf area
index were measured in all stands as part of a related study (see chapter 1). Red oak sites
experienced more defoliation (Table 3.1) and had higher gypsy moth egg mass densities
(Figure 3.1) than the northern hardwood sites from 1993 to 1995. Red oak sites had lower

61

leaf area index (LAI) than northern hardwood sites, and leaf area was signiﬁcantly reduced
by large gypsy moth populations in 1994 and 1995 (see chapter 1) (Table 3.1). Gypsy
moth populations in red oak sites ﬂuctuated annually. Outbreak populations at Branch sites
in 1993 were followed by a population 'crash' due to NPV infection in 1994 (Table 3.1).

Populations at Freesoil sites increased in 1994 and reached outbreak levels in 1995 (see

chapter 1) (Figure 3.1).
Collection of Lepidoptera

Lepidoptera were collected during 1993, 1994 and 1995. Lepidoptera were
collected four times a summer. Collection times roughly corresponded to signiﬁcant points
in gypsy moth management or biology. For example, typical application time of the
microbial insecticide Bacillus thuringensis Berliner var. kurstala' for gypsy moth
suppression in northern lower Michigan occur in mid-May, fourth instar gypsy moth in
mid-June), peak defoliation in mid-July, and stand refoliation in mid-August. During
1993, lepidoptera were collected on 12-13 May, 15-17 June, 13-15 July, and 10-12
August from red oak sites and on 21-22 May, 22-23 June, 20-22 July, and 18-20 August
from northern hardwood sites. In 1994, lepidoptera were collected on 10-12 May, 15-16
June, 11-13 July, and 9-10 August from red oak sites and on 18-20 May , 20-21 June, 19-
20 July, 15-17 August from northern hardwood sites. In 1995, collection dates were 15-
17 May, 12-14 June, 10-11 July, 15-16 August at red oak sites and 22-24 May, 19-20
June, 17-19 July, and 21-22 August at northern hardwood sites. Lepidoptera were
collected from the canopy and shrub strata using a variety of sampling methods in an effort

to adequately represent the total lepidopteran diversity within each site.

Canopy strata- In 1993, canopies were sampled using a pole pruner with a 9 m

extendible handle, modiﬁed sweep nets with a 10 m extendible handle, and suspendable

62

UV traps. In 1994 and 1995, pole pruners and extended sweep nets were replaced with a
suspendable thermal fogger to facilitate collection of diurnal canopy and immature
lepidoptera.

Ultraviolet traps consisted of a 22 watt UV bulb connected to a photosensor and
powered by a 6 volt motorcycle battery. UV traps were operated approximately 8 hours
each sampling period. A collection funnel and bucket loaded with insecticidal pest strips
(V apona TM) were attached below the bulb. Use of a low wattage bulb and a rain cover
over the bulb restricted potential long-distance attraction of noctumal insects, presumably
resulting in a localized sample reﬂecting endemic insect populations (Bowden 1982). The
trap was raised into the canopy approximately 1 to 2 m above the lower edge of the canopy
foliage in the center of each site.

Pole pruners were used to clip a 1 m length branch tip from the lower canopy six
times along a random transect through each stand at 20 m intervals. Samples were
collected from the lower canopy of the nearest tree to each 20 m interval. Samples were
clipped over a large plastic tarp spread on the ground. Branch cuttings were collected off
the tarp and quickly placed in plastic bags until they could be examined for lepidoptera.
Extended sweep net samples were collected from lower canopy foliage six times along a
random transect through each stand at 20 m intervals. Four sweeps with a 50 cm diameter
net were made from lower canopy foliage at each interval.

A thermal fogger (Dyna-fogTM) (Erwin 1983) modiﬁed with a radio-controlled
switch was raised into the canopy ﬁve times along a randomly selected transect through the
stand at 20 m intervals. The fogger was suspended by ropes and pulleys placed over large
tree limbs with an arrow-line. A 25 m2 diameter column of the canopy was then fogged
with 1% pyrethroid (Pyrocide Fogging Concentrate 73387“, McLaughlin Gormley King
Company) applied for 15 seconds at a rate of 0.377 L/minute. Pyrethroid was selected for
its quick "knock-down ability" and minimal persistence. The fogged foliage was
thoroughly shaken using additional suspended r0pes. Insects were collected in 25 1 m2

63

funnels arranged in a grid below the fogger. All fogging was done between 0400 and 0800

h when winds were calm.

Shrub strata- Lepidoptera were also collected using malaise traps and sweep nets. One
malaise trap with an insecticidal pest strip was set up in the center of each stand. Malaise
traps were left open for 24 hours. Sweep samples were collected using a 38 cm diameter
net. Sixteen sweeps (4 at each cardinal direction) were made six times along a random

transect through the stand at 20 m intervals.

We assumed that the location where insects were captured was typical of the
location they occupy within the site. For example, lepidoptera collected from UV traps
were considered overstory dwellers while lepidoptera collected in malaise traps were
considered understory dwellers. All lepidoptera were placed into plastic containers lined
with tissue paper to minimize damage of specimens between collection and identiﬁcation.
Voucher specimens were prepared from each site and sample period. Lepidoptera were
identiﬁed to species or morpho-species by J. Wilterding, R. Kriegel, and M. Neilson at the

Center for Insect Diversity at Michigan State University.

Statistical Analysis

Number of insects, number of species, and Brilloun and Simpson's indices of
diversity were detemrined for each site (Magurran 1988). Brilloun's diversity index was
selected for use because the majority of insects were collected with UV traps, a non-
random sampling technique (Magurran 1988). Furthermore, Brilloun's index provided an
estimate of diversity based on species richness (Magurran 1988). Simpson's index was
chosen because it provided an estimate of diversity based on species dominance (Magurran

1988). The combination of diversity indices, species richness and abundance was used to

64

evaluate differences among sites because each particular measure was subject to biases
related to sample size and distribution of insects (Magurran 1988). We believe that a
consistent pattern of diversity that is reﬂected by multiple indices is a more reliable indicator
than any single index (Magurran 1988).

Diversity indices, species richness and insect abundance were calculated from
yearly totals of all lepidoptera collected. Within each ecosystem, all measurements were
evaluated for normalin and homoscedascity. Site differences in variables that approached
norrnailty were analyzed using a one-way ANOVA with repeated measures (also known as
a split -plot ANOVA) (Sokal and Rohlf 1995). Site differences in variables with non-
norrnal distributions were analyzed using the nonparametric Kruskal-Wallis test (Conover
1971). An alpha value of 0.05 was used for both parametric and nonparametric tests.

Similarity of lepidoptera fauna among all sites was examined using cluster analysis
of the Sorenson's quantitative index of similarity (also known as the Bray-Curtis index)
(Ludwig and Reynolds 1988; Magurran 1988). Dendograrns were generated using the
Bray-Curtis distance measures and four different linkage methods: the beta-ﬂexible method
with B: -0.25 and -0.5 (Milligan 1989); the average linkage grouping method; and centroid
linkage method (Ludwig and Reynolds 1988).

While species and morpho-species identiﬁcations were available for all lepidoptera,
life history data, including host plant preference and patterns of voltinism, were most
available for lepidoptera in the family Noctuidae. For this reason, additional analysis of the
noctuids was conducted by examining the early (May and June) and late season (July and
August) catches. Extensive sampling of N octuidae was prohibited by limited resources,
and thus many species were represented by small numbers of insects. Because of this
relative paucity of insects, diversity indices and formal inferential statistical analysis were
inappropriate. We have presented our data as totals numbers of insects collected, with the
caveat that while differences may illustrate effects of gypsy moth, conclusions should be

interpreted cautiously.

65

Results

Over the course of this study 12,116 lepidoptera were collected from 38 families.

A total of 35 families were collected from the red oak sites and 30 families were collected
from the northern hardwood sites. While three trapping methods were used, over 90% of
lepidoptera collected each year were collected in the UV traps. Of the families collected, 21
families accounted for 99% of the total lepidoptera collected within each year. Abundance
of tortricids, noctuids, gelechiids, geometrids, and pyralids was consistently high in both
ecosystems over all three years (Table 3.2 and 3.3). Of the 38 families collected, 17 rarely
caught families accounted for less than 1% of the total lepidoptera collected within each
year (Table 3.4 and 3.5). Species richness in both ecosystems was dominated by four
families; the Noctuidae, Torticidae, Geometridae, and Pyralidae (Table 3.6 and 3.7).
Species caught within these families and their total abundance within each ecosystem are
listed as Appendix A through Appendix D. Species collected from families other than
Noctuidae, Tortricidae, Geometridae, or Pyralidae are listed in Appendix E. Abundance of
individual species was not great enough to determine whether gypsy moth directly affected
particular species.

Within the red oak ecosystem, defoliation by month interactions signiﬁcantly
affected total abundance and total species richness in 1993 (Table 3.8 and 3.9) and total
species richness again in 1994 (Table 3.8 and 3.10 and Figure 3.2). Marginally signiﬁcant
defoliation by month interactions were observed in Brilloun's index in 1993 (Table 3.8 and
3.9) and in total species richness in 1995 (Table 3.8 and Table 3.11).

Within the northern hardwood ecosystem, total abundance, species richness and
Simpson's index in 1993 signiﬁcantly differed among months (Table 3.12). Marginally
signiﬁcant monthly differences were apparent in total species richness in 1994 (Table
3.13). Differences in species richness between Mesick and Harrietta were not signiﬁcant in

any year (Table 3.8 and 312-314 and Figure 3.3). During 1993, total species richness

66

increased throughout the summer (Figure 3.3). In 1994 and 1995, species richness was
generally greatest in the months of June and July (Figure 3.3).

Hierarchical cluster analysis was used to compare species composition of the two
ecosystems. Dendograrns created using the beta ﬂexible method with B: -0.25, revealed
two distinct clusters, one containing all the red oak sites, and the other containing all the
northern hardwood sites in 1993 and 1995 (Figure 3.4). In 1994, both Branch sites
formed a single cluster while Freesoil sites and the northern hardwood sites comprised the
remaining cluster (Figure 3.4).

Dendograrns created using the beta ﬂexible method with B: -0.5 (Figure 3.5), gave
similar results to the clustering using [3: -0.25. However, the dendogram created for 1994
data using [3: -0.5 could have been interpreted as two clusters or three clusters with equal
conﬁdence. Northern hardwood sites formed one cluster, while Branch sites and Freesoil
sites could have been interpreted as one or two clusters.

When the group average/unweighted pair-grouping method was used, Branch sites
consistently clustered separately from Freesoil sites and the northern hardwood sites in
1993 and 1994 (Figure 3.6). In 1995, red oak sites clustered separately from northern
hardwood sites (Figure 3.6). The centroid/median method produced dendograms that were
fraught with reversals and had little biological meaning; these were not considered further.

Effects of Gypsy Moth on Lepidoptera Collected During the Early Season

Adult lepidoptera collected in the early season months of May and June were
presumably larvae the previous year. Consistent patterns of decreased species richness and
abundance in early season adult lepidoptera were observed in red oak sites one year
following severe defoliation. In the red oak ecosystem, species richness of all adult
lepidoptera was marginally lower in June 1994 at sites that experienced almost 100%

defoliation the previous year (Branch) (Figure 3.2). Likewise, in 1995, total species

 

67

richness was lower in the early season months of May and June in Freesoil sites that

experienced 40-60% defoliation the previous year (Figure 3.2).

Total species richness of early season Noctuidae was lower in Branch sites than in
Freesoil sites one year following heavy defoliation at Branch in 1993. Similarly, the year
after heavy defoliation at Freesoil sites in 1994, species richness of early season noctuids
was lower in Freesoil sites than in Branch sites (Figure 3.7). Total abundance of noctuids
during May and June was lower at Branch sites than at Freesoil sites during 1994 (Figure
3.8). Likewise, abundance of noctuids collected during May and June was lower at
Freesoil sites than at Branch sites during 1995 (Figure 3.8). Abundance of overstory
(Figure 3.9) and understory noctuids (Figure 3.10) showed similar patterns to total
abundance of noctuids. Abundance of oak-feeding noctuids collected in the early season
months of May and June (Figure 3.11) also suggested negative interactions with gypsy
moth one year after severe defoliation in 1993.

However, the difference in abundance of oak-feeding noctuids between Branch and
Freesoil sites is small in 1995, one year following severe defoliation at Freesoil. Few oak-
feeding noctuids were collected in 1995 relative to the numbers collected in 1993 and 1994.
This may indicate that if early season oak feeders in Branch sites were affected in 1993,
they did not recover to pre-defoliation levels even two years after a defoliation event.

Within the northern hardwood ecosystem, the magnitude of differences between
sites in species richness, total noctuid abundance, and over and understory abundance were
generally smaller than differences among red oak sites and were not related to gypsy moth
defoliation (Figure 3.7-3.10). However species richness was signiﬁcantly greater in

Harrietta than in Mesick sites in 1995 (Figure 3.8).

68

Effects of Gypsy Moth on Lepidoptera Collected from the Late Season

Most adult lepidoptera collected from late season months of July and August were
presumably larvae earlier during the same year. Species richness and abundance of adult
lepidoptera collected from the late season months of July and August declined in oak-
dominated sites that experienced gypsy moth defoliation in 1993 and 1994 during the same
year as the defoliation occurred. In the red oak ecosystem, species richness of all
lepidoptera was lower in defoliated sites (Branch) during the late season months of July
and August of 1993 than in non-defoliated (Freesoil) sites (Figure 3.2). Similarly, in
1994, defoliated sites (Freesoil) had lower species richness during July than did non-
defoliated sites (Branch) (Figure 3.2). Species richness of late season noctuids was lower
in Branch than in Freesoil in 1993 (Figure 3.7). In 1994, total species richness of noctuids
was lower in Freesoil than in Branch (Figure 3.7). Similar patterns were observed for the
total abundance of noctuids (Figure 3.8).

In 1995, however, species richness, abundance of total lepidoptera and total
Noctuidae collected from late season months was greater at Freesoil than at Branch, even
though Freesoil experienced nearly 100% defoliation (Figure 3.2, 3.7 and 3.8).
Understory Noctuidae were also more abundant in Freesoil than in Branch in 1995 (Figure
3.10).

Overstory and oak-feeding Noctuidae also appeared to be more abundant in Freesoil
than in Branch during July and August in 1995. However, one species, Hyperstrotia
pervetrens, comprised 23% (35 individuals) of the total noctuid catch. This species had not
been caught during our study in any site previous to July and August 1995, and is at the
northwestern edge of its distribution in Michigan (I. Wilterding, MSU, pers comm). If
H. pervetrens is excluded from the analysis, the difference in overstory noctuids between
Branch and Freesoil becomes much smaller. Likewise, when H. pervetrens was excluded

from analysis, Branch and Freesoil had the same amount of oak-feeding lepidoptera in the

69

months of July and August in 1995 (Figure 3.12). The decline in Noctuidae in Freesoil in
1995 suggests oak-feeding Noctuidae may be negatively affected during the same year as a
defoliation event.

In the northern hardwood ecosystem, differences in species richness, total noctuid
abundance, and over and understory abundance between sites were again smaller than
differences in the red oak ecosystem (Figure 3.7-3. 10). However, Mesick did have greater
numbers of overstory noctuids in 1993 than did Harrietta (Figure 3.9).

D' .

Differences in Lepidopteran Fauna Between Ecosystems

It was surprising that red oak sites had more species diversity and generally greater
abundance of lepidoptera than did northern hardwood sites. The high productivity typical
of northern hardwood sites (Zak et al. 1989) suggests that insects would not be as
nutritionally limited as insects within the less productive red oak sites (Mattson 1980;
Mattson and Scriber 1987; Scriber and Slansky 1981). Likewise, greater diversity of
overstory and understory ﬂora within the northern hardwood sites also suggests greater
diversity of herbivores than would be expected in the non-diverse ﬂora of the red oak sites.

One explanation for the higher species richness in red oak sites may be related to the
differences in light penetration and microclimate between each ecosystem. Red oak sites
typically had more light reaching the forest ﬂoor and were presumably warmer than
northern hardwood sites. This warmer microclimate may have promoted insect
development and activity and may account for the greater diversity and abundance collected

from red oak sites.

 

 

70

Another potential explanation for the greater diversity and abundance at red oak
sites may be related to the current successional stage of each ecosystem. In plant
communities, species richness rapidly increases in early successional stages through
immigration and colonization (Huston 1994). As succession continues, species richness
begins to decrease due to competition for resources (Huston 1994). In late successional
types, diversity is relatively low and the community is dominated by fewer species than in
intermediate successional stages (Huston 1994).

Likewise, species richness and diversity of insects also increase during early stages
of succession because of the increased spatial heterogeneity and architectural diversity of
the plant community (Brown 1984). Later in succession when plant diversity decreases,
insect communities no longer increase in diversity although species turnover does still
occur (Brown 1984).

The highly productive northern hardwood sites represent late successional seres.
Predicted compositional changes for the northern hardwood sites included replacement of
the already minor component of basswood by more sugar maple or beech (Host et al.
1987). This decrease in plant species diversity may help explain the relatively low species
richness of lepidoptera we observed. The insect community, like the plant community may
have been reﬂecting competitive exclusion typical of late stage succession as decreased
species richness and diversity.

Red oak sites may be representative of a more intermediate successional sere than
northern hardwood sites. Future compositional changes to red maple or northern
hardwood species are predicted for red oak sites (Host et al. 1987). The relatively greater
species richness of insects we observed in red oak sites may have been related to this earlier
successional stage of the plant community.

Our ﬁndings have interesting implications for ecosystem management and forest
health. Information about diversity at regional scales is needed before political, economic,

and biological concerns can be integrated into effective guidelines and policy for

 

1w r......a.-...- . _

71

management of public forests (Probst and Crow 1991; Wilcove 1989). It should also be
realized that diversity within an ecosystem is likely to change over time through succession.
The northern hardwood sites we studied are under consideration for designation as sites to
be set aside to become old-growth forest (R. Ingram, Huron-Manistee National Forest
Silviculturalist, pers. comm). While these sites represent a unique habitat type in terms of
plant diversity and productivity, they had lower lepidopteran diversity than less productive
red oak sites. This illustrates that ecosystems that are initially managed to promote

diversity may eventually become less diverse over time.

Effects of Gypsy Moth on Native Lepidoptera Diversity

The lepidopteran fauna collected from red oak sites was different than the fauna
collected at northern hardwood sites. This baseline information has implications for the
pool of species potentially affected by gypsy moth. Species common to the northern
hardwood ecosystem are less likely to interact with gypsy moth because sugar maple, the
dominant species, is not a preferred host of gypsy moth. Lepidopteran species common in
the red oak ecosystem were more likely to interact with gypsy moth because the dominant
species, red oak, is a preferred host of gypsy moth.

Because all red oak sites experienced severe defoliation by gypsy moth over the
three years in which this experiment was conducted, comparisons of undisturbed red oak
and undisturbed northern hardwood sites cannot be made, nor does an optimal control
treatment exist for the effects of defoliation in the red oak sites. However, the lack of
differences between Mesick and Harrietta in species richness of all lepidoptera, Noctuid
abundance, and the overall similarity of lepidopteran fauna provide an estimate of the
amount of natural variation in abundance and diversity of lepidoptera over time. When red
oak and northern hardwood sites were compared in terms of total species richness, red oak

sites had signiﬁcantly more variation between sites than did northern hardwood sites.

72

When the abundance of noctuids was compared, red oak sites varied more than did
northern hardwood sites. While not optimal, we believe that the consistent patterns within
the northern hardwood sites provided a realistic and adequate control treatment to assess
whether defoliation affected lepidoptera richness and abundance.

While speciﬁc mechanisms responsible for the patterns in species richness and
abundance we observed were not directly tested in our experiment, we have posed
potential explanations for the negative effects of gypsy moth on native lepidoptera,
particularly the Noctuidae. Consistent patterns of decreased species richness and
abundance in early season lepidoptera one year following a defoliation event may be
attributed to a lag effect apparent in univoltine and second generation multivoltine
lepidoptera. This suggests that univoltine and second generation multivoltine lepidoptera
were negatively affected by gypsy moth in immature lifestages and consequently reﬂected
this interaction as decreased numbers of adult insects the following year. It is possible that
other non-noctuid lepidoptera that consume oak foliage, such as tortricid leafrollers, could
also face resource depletion during a defoliation event. However, without speciﬁc life
history data on each species collected and without greater sample size of each species,
interpretation of species compositional changes are difﬁcult.

Species richness and abundance of lepidoptera collected from the late season
months of July and August, during peak gypsy moth defoliation, declined in sites that
experienced gypsy moth defoliation in 1993 and 1994. This decline occurred during the
same year as the defoliation. This decline was also reﬂected in the abundance of noctuids.
The decrease in adult lepidoptera during these periods of intense defoliation may reﬂect
unsuitable nricrohabitat conditions created by gypsy moth. Microclimatic factors and
structural features of the site are potentially more important than food resources to adult
lepidoptera collected during the late season. During a defoliation event, an open canopy
increases temperature and decreases the relative humidity within the site (Klein and Perkins

1988; Perkins et al. 1987; Spurr and Barnes 1980). Furthermore, defoliation depletes

73

potential ovipositional sites and can result in egg or larval mortality caused by poor
oviposition site selection (Dethier 1959).

However, the species richness and abundance of lepidoptera and Noctuidae was
higher in the severely defoliated Freesoil sites than in undefoliated Branch sites in 1995
during July and August. This increase in diversity is not easily explained. Our method of
trapping could be biased during a defoliation event. The UV trap would be more visible
when foliage was removed than in undefoliated sites. This seems an unlikely explanation,
however, because similar conditions existed in Branch in 1993 and Freesoil in 1994 and
there was a consistent decrease in abundance and richness in these defoliated sites.

Another explanation may be found upon further examination of the understory
noctuids. By defoliating overstory trees, herbivorous insects can increase the growth of
understory ﬂora through increased light penetration and/or increased nitrogen, phosphorus,
and potassium availability in the form of insect frass, corpses, and unused food (Mattson
and Addy 1975). Freesoil experienced heavy defoliation and an opened canopy two
consecutive years, while Branch only experienced one year of open canopy. The two years
of open canopy at Freesoil could have promoted availability and/or quality of understory
herbaceous plants and increased the abundance and diversity of understory feeders. More
understory noctuids were collected from Freesoil sites than Branch sites. The relative
difference of understory noctuids between Branch and Freesoil in 1995 is greater than the
difference between the abundance of overstory feeders in these same sites, particularly after
H. pervetrens is excluded from analysis. The increase in understory noctuids may be a
response to increased understory plant resource availability or quality and may explain the

increase in noctuid abundance in Freesoil during late season 1995.

 

 

74

Effects of Gypsy Moth on Oak-Feeders

Decreased abundance of oak-feeding noctuids collected from the early season
months of May and June may be evidence of negative interactions with gypsy moth one
year after severe defoliation in 1993. However, the difference between Branch and
Freesoil sites was small in 1995, one year following defoliation at Freesoil. Abundance of
oak-feeding noctuids collected during early season months at Branch increased from 1994
to 1995. This may be evidence of a gradual recovery of the oak-feeders. This conclusion
should be interpreted cautiously because any lag effect present at Freesoil during 1995 will
be confounded with the defoliation that also occurred during that same year. The potential
effect of gypsy moth on oak-feeding noctuids present during early season months warrants
further study to determine the rate of recovery of native noctuid populations.

Late season oak-feeders may have been negatively affected by gypsy moth during
the same year as a defoliation event, but again this conclusion should be interpreted
cautiously. Abundance of oak-feeders was lower in sites that experienced defoliation in
1993 and 1994, but no effects of gypsy moth were apparent in Freesoil sites during 1995.
even after the removal of H. pervetrens from the analysis.

Interspeciﬁc competition for oak foliage with gypsy moth may have been
responsible for the observed decline in adult oak-feeding Noctuidae during May and June at
Branch in 1994 and Freesoil in 1995. Defoliation removed between 40 and 100% of
foliage otherwise available to oak-feeding noctuids. Any immature oak-feeder that was
present between the end of June and July was presumably in direct competition with gypsy
moth for oak foliage. Gypsy moth has been shown to negatively impact native lepidoptera
even when defoliation was light (Sample et al. 1993).

While competition is disfavored by some investigators as the ecological principle
governing community organization of herbivorous insects (Lawton and Strong 1981),

insect outbreaks may be an exception to this vieWpoint (Denno et al. 1995). Leaﬂroppers

 

75

were occasionally out-competed for sycamore foliage by lacebugs and other leafhopper
species when large competitor populations exhausted food reserves of leafhoppers
(McClure 1974; Ross 1957). In rangeland ecosystems where grasshoppers can consume
large proportions of available plant resources, eight species overlapped signiﬁcantly in food
items consumed (Hansen and Ueckert 1970). In forested ecosystems, pine needle scale
(Chionaspis pinifoliae (Fitch)) outcompeted black pineleaf scale (Nuculaspis califomica
(Coleman)) on ponderosa pine growing along roadways when parasites were eliminated by
increased dust concentrations (Edmunds 1973). Likewise, four species of pine-defoliating
moths experienced interspeciﬁc competition for food during outbreaks on rare occasions

(V arley 1949). Monarch butterﬂies (Danaus plexippus (L.)) eliminated the milkweed bug
(Oncopeltus spp.) from the island of Barbados by depleting the shared food plant Asclepias
curassavica. (Blakley and Dingle 1978).

The apparently minor and short-term effects of gypsy moth on native lepidopteran
abundance and diversity may be related to the spatial scale of gypsy moth outbreaks.
Spatial scale can affect the recolonization of a disturbed area (Huston 1994). Aerial
surveys in northwestern lower Michigan by the Michigan DNR indicated that between 1993
to 1995, severe defoliation occurred in areas less than 115 contiguous ha in size.
Approximately 80% of the severe gypsy moth defoliation that occurred during this period
was concentrated in areas of no larger than 120 ha (Sapio 1996). Similarly, approximately
80% of moderate defoliation was concentrated in areas no larger than 130 ha (Sapio 1996).
Larger outbreaks that exceeded 480 contiguous ha accounted for only 4% of heavy
defoliation and 3% of moderate defoliation (Sapio 1996). In our study sites, severe
defoliation did not exceed 65 ha and lasted no more than 2 years before populations
collapsed due to NPV infection. Lepidoptera may have been temporarily displaced to
surrounding unaffected forest, but were presumably able to recolonize disturbed areas

because of the relatively small area and short duration of severe defoliation.

 

,l.. a.

76

Management Implications

Our conclusions have implications for forest pest management, including the use of
the microbial insecticide Bacillus thuringiensis kurstaki. (Btk). While the oak-feeding guild
may be negatively impacted by gypsy moth, other lepidoptera collected from the early
season appeared to recover within two years after a defoliation event. Similarly, non-oak
feeders such as understory Noctuidae may even be positively affected by gypsy moth. In
contrast, ﬁeld studies indicated Btk persisted for up to 30 d following application and
negatively affected survival of two Papilio spp. and Callosamia promethea (Drury)
(Johnson et al. 1995). Application of Btk to control gypsy moth populations reduced
native lepidoptera species richness and abundance for up to two years after application
(Miller 1990; Sample et al. 1993). While one feeding guild in particular (the oak-feeders)
was negatively affected by gypsy moth, comparatively more species may be at risk from

application of Btk than are likely to be at risk from .gypsy moth.

 

77

Table 3.1. Visual estimation of peak defoliation and estimation of mean
LAI (iSE) using photosynthetically active radiation transmittance of sites
in red oak and northern hardwood ecosystems.

 

Peak Visual Estimation of Defoliation

 

 

 

 

l 9 9 3 l 9 94 l 9 9 5
RM
Branch 80-100% < 20% < 20%
Freesoil < 20% 40-60% 80-100%
magnum
Harrietta < 20% < 20% < 20%
Mesick 20-40% < 20% < 20%

Estimated Leaf Area IndexI

 

 

1994 1995
m1:
Branch 4.48 i 0.221 4.71 :I: 0.115
Freesoil 2.73 :1: 0.076 2.72 i 0.151
Nurthsrmﬂarzdmud
Harrietta 10.9 :1: 0.226 8.61 i 0.229
Mesick 11.9 i 0.189 8.14 i 0.222

 

1Means and standard errors were calculated from two sites (n=2)

78

Table 3.2. Total number of insects/family that comprised greater than 1 %
of the total number of lepidoptera collected from two sites near Branch and
two sites near Freesoil within red oak ecosystems.

   

1993 1994 1995

 

 

 

 

Family Branch Freesoil Branch Freesoil Branch Freesoil
Tortricidae 104 419 328 224 487 599
Noctuidae 133 379 127 273 176 204
Lyonetiidae 1 - 2 - 56 33
Gelechiidae 29 5 2 17 1 67 127 l 16
Geometridae 145 76 40 48 24 46
Pyralidae 8 56 55 55 24 83
Blastobasidae 13 11 47 18 34 28
Oecophoridae 4 5 2 83 3 l4
Arctiidae - 19 9 - 23 24
Lasiocampidae 5 4 4 - 1 9
Lymantriidae 46 17 - - 1 -
Argyresthiidae - 1 1 - 1 14
Notodontidae 6 10 12 15 1 -
Nepticulidae - - - 2 24 44
Tineidae 12 24 1 3 - 5
Coleophoridae - - 3 - 7 21
Gracillariidae - 7 l2 1 7 4
Hydroptilidae - - 33 3 - -
Lirnacodidae - 1 l 1 2 l
Eriocraniidae 1 - 16 7 -

Cochylidae 1 - - - - -

 

79

Table 3.3. Total number of insects/family that comprised greater than 1 %
of the total number of lepidoptera collected from two sites near Harrietta
and two sites near Mesick within northern hardwood ecosystems.

 

 

 

 

 

 

1 9 9 3 1 9 9 4 1 9 9 S
Family Harrietta Mesick Harrietta Mesick Harrietta Mesick
Tortricidae 451 996 305 260 960 725
Noctuidae 59 92 78 76 137 143
Lyonetiidae - 1 - - 584 226
Gelechiidae 20 47 35 59 49 44
Geonretridae 41 31 50 71 51 83
Pyralidae 25 41 43 30 27 60
Blastobasidae 5 l 1 l2 2 77 80
Oecophoridae - 3 2 12 10 6
Arctiidae 13 14 6 3 5 1 1
Lasiocampidae 7 2 14 19 26 17
Lymantriidae 18 11 - - - -
Argyresthiidae 2 3 20 6 22 21
Notodontidae 28 8 5 5 - -
Nepticulidae - 2 l - 4 4
Tineidae 3 2 2 l4 2 -
Coleophoridae - 1 1 - l4 1 l
Gracillariidae - 4 1 1 9 10
Hydroptilidae - - - - - -
Limacodidae 2 1 8 2 2 10
Eriocraniidae - - 4 1 - -
Cochylidae 2 6 12 2 - -

 

 

80

Table 3.4. Total number of insects/family that comprised less than 1 % of
the total number of lepidoptera collected each year from two sites near
Branch and two sites near F reesoil within red oak ecosystems.

 

1993 1994 1995

 

 

 

Family Branch Freesoil Branch Freesoil Branch Freesoil

 

Sphingidae 1 - - 6 1 -
Cosmopterigidae - 3 - - -
Satyridae - - 4 2 1
Hesperiidae - - - - 2
Olenthreutinae - - - 1 -
Plutellidae - - - - 2
Yponomeutidae - - - - -
Glyphipterigidae - - - - l
Pieridae - 1 - - - -
Psychidae 3 - - - - -
Sesiidae - - - 3 - -
Agonexenidae - - - - - 1

IUJNOO

-~

Cossidae 1 - - - - -
Thyatiridae 1 - - - . - -
Zygaenidae - 1 - - - -

 

81

Table 3.5. Total number of insects/family that comprised less than 1 % of
the total number of lepidoptera collected each year from two sites near
Harrietta and two sites near Mesick within northern hardwood ecosystem.

 

1993 1994 1995

 

 

Harrietta Mesick Harrietta Mesick Harrietta Mesick

 

Sphingidae 2 1 1 - l 3
Cosmopterigidae - - - - - l
Satyridae 1 - - - - 1
Olenthreutinae - - 1 3 - -
Plutellidae - - - - 2 -
Yponomeutidae - - 2 - 1 -
Glyphipterigidae - - - - - l
Pieridae 2 - - - - -
Agonoxenidae - 1 - - - -
Decophoridae - - - - - 1

Drepanidae - - - - 11 -

 

 

82

Table 3.6. Total number of species within each lepidopteran family
collected over three years from two sites near Branch and two sites near
Freesoil within a red oak ecosystem.

 

 

 

 

Family Branch Freesoil Branch Freesoil Branch Freesoil

 

Noctuidae 40 62 55 49 37 68
Tortricidae 15 44 39 17 57 61
Geometridae 12 23 18 23 15 22
Pyralidae 6 25 10 17 29
Gelechiidae 8 8 22 19
5
1

N
u—a

Blastobasidae 3 12 12
Lyonetiidae - 21 16
Arctiidae -
Notodontidae 5
Oecophoridae
Coleophoridae - -
Gracillariidae -
Tineidae 2
Nepticulidae -
Argyresthiidae —
Lasiocampidae 2
Limacodidae -
Lymantriidae 5
Sphingidae 1 - - 3
l 3
1

H
ﬂ

OWQ
I—dHNNO‘UIF-‘Mm
Nd

axe—nu—au—sro‘m
I
ION-—
I
t—NWGUJAMO

Eriocraniidae
Cochylidae

Satyridae - - 2 2 1 1
Cosmopterigidae - - - - . - 5
Hesperiidae - - - - 2 3
Pieridae - l - - 1 1
Olenthreutinae - - - 1 - -
Hydroptilidae - - 2 l - -
Yponomeutidae - - - - - 1

Table 3.6 (cont'd.)

Psychidae 2

Agonexenidae
Cossidae
Glyphipterigidae
Sesiidae
Thyatiridae
Zygaenidae

83

84

Table 3.7. Total number of species within each lepidopteran family
collected over three years from two site near Harrietta and two sites near
Mesick within northern hardwood ecosystems.

 

 

Family Harrietta Mesick Harrietta Mesick Harrietta Mesick
Noctuidae 32 46 42 33 68 83
Tortricidae 30 43 29 3 l 57 82
Geometridae 15 18 20 15 25 31
Pyralidae 12 15 16 18 16 18
Gelechiidae 5 7 6 7 11 17
Blastobasidae 4 5 2 19 22
Lyonetiidae - 1 - - 18 20
Arctiidae 3 7 3 2 5 5
N otodontidae 5 5 3 4 - 2
Oecophoridae - 3 2 1 6 3
Coleophoridae - l l - 5 9
Gracillariidae - 2 1 4 5
Tineidae 2 1 2 3 2 -
Nepticulidae - 2 l - 4 4
Argyresthiidae l 1 3 1 4 5
Lasiocampidae 2 1 2 2 2 2
Limacodidae 2 1 4 2 2 2
Lymantriidae 2 3 - - - -
Sphingidae 2 1 1 - l 3
Eriocraniidae - - 3 1 - -
Cochylidae 2 2 2 l - -
Satyridae 1 - - - - l
Cosmopterigidae - - - - - 1
Pieridae 1 - - - 1 -
Olenthreutinae - - l 2 - -
Glyphipterigidae - - - - - 1
Yponomeutidae - - 1 - 1 -
Agonoxenidae - l - - - -

Decophoridae - - - - - 1

Table 3.7 (cont'd.)

Drepanidae

85

 

86

Table 3.8 Means (iSE) of total abundance, species richness, Brilloun’s
index, and Simpson’s index for Lepidoptera collected from red oak and
northern hardwood ecosystems over three years (means calculated from
data pooled over four months).

 

Freesoil Harrietta Mesick
Mean (:tSE) Mean (:tSE) Mean (iSE)

     

Branch
Mean (iSE)

 

Abundance
1993 64.25 (25.181) 137.00 (34.927) 85.25 (29.364) 161.50 (55.885)
1994 108.50 (50.188) 101.50 (45.797) 74.25 (22.162) 69.25 (21.641)
1995 321.38 (113.304) 206.38 (50.314) 206.88 (55.170) 129.88 (44.372)

Species

Richness
1993 13.63 (2.235) 25.25 (5.793) 15.13 (4.514) 20.88 (5.709)
1994 21.88 (4.980) 17.13 (5.386) 18.50 (2.860) 15.75 (3.447)
1995 46.88 (4.845) 40.38 (6.533) 35.75 (7.158) 31.13 (3.925)

Brilloun’s

Index
1993 0.70 (0.067) 0.87 (0.125) 0.58 (0.147) 0.61 (0.138)
1994 0.86 (0.111) 11.31 (7.214) 0.84 (0.061) 0.74 (0.087)
1995 1.03 (0.052) 1.02 (0.062) 0.95 (0.068) 1.02 (0.026)

Simpson’s

Index
1993 5.93 (1.369) 9.35 (2.578) 5.60 (2.537) 3.77 (1.120)
1994 12.33 (3.413) 10.08 (1.926) 9.47 (1.979) 8.04 (2.378)
1995 8.21 (1.687) 8.85 (1.254) 7.23 (1.236) 13.69 (2.654)

 

 

Table 3.9. Statistical significance of treatment effects from split-plot
ANOVA of ﬁve measures of Lepidoptera diversity in four red oak sites in

1993.

 

 

Source of Variation

 

   

Species Abundance1 Brilloun's Simpson's

  

 

 

df

Richnessl Index Index
F value of signiﬁcant treatment effects

Gypsy Moth 1 42.596 1 1.090 5.080 6.606

Month 3 1.790 3.225 1.214 0.873

Gypsy Moth x Month 3 8.862 6.823 4.512 0.288

Probability of signiﬁcant treatment
effects
Gypsy Moth 1 0.0227 0.0796 0.1529 0.1239
Month 3 0.2491 0.1035 0.3827 0.5054
Gypsy Moth x Month 3 0.0127 0.0232 0.0555 0.8326
Mean square

Error a 2 0.122 4.514 0.025 7.080
Error b 6 0.778 7.581 0.047 48.822

 

lValues transformed using the Box-Cox transformation

 

88

Table 3.10. Statistical significance of treatment effects from split-plot
ANOVA of ﬁve measures of Lepidoptera diversity in four red oak sites in

 

 

1 9 94 .
Source of Variation df Species Abundancel Brilloun's Simpson's
Richnessl Index2 Index1
F value of signiﬁcant treatment effects
Gypsy Moth 1 8.721 3.106 x2=0.7l 0.637
Month 3 16.926 7.979 x2=9.26 5.635
Gypsy Moth x Month 3 5.209 3.020 - 1.910
Probability of signiﬁcant treatment
effects
Gypsy Moth 1 0.0981 0.2201 0.4008 0.5085
Month 3 0.0025 0.0162 0.0260 0.0352
Gypsy Moth x Month 3 0.0415 0.1157 - 0.2291
Mean square
Error a 2 0.015 0.012 - 0.015
Error b 6 0.020 0.128 - 0.033

 

1Values transformed using the Box-Cox transformation
2Values tested using the Kruskal-Wallis test

89

Table 3.11. Statistical signiﬁcance of treatment effects from split-plot
ANOVA of ﬁve measures of Lepidoptera diversity in four red oak sites in
1995.

 

Source of Variation df Species Abundancel Brilloun's Simpson's
Richness Index2 Index1

     

F value of signiﬁcant treatment effects

Gypsy Moth 1 1 1.965 4.460 x2=0.01 0216
Month 3 1.486 0.523 x2=4.37 0.876
Gypsy Moth x Month 3 4.184 1.503 - 0.460

Probability of signiﬁcant treatment

effects
Gypsy Moth 1 0.0744 0.1691 0.9162 0.6880
Month 3 0.3105 0.6820 0.2242 0.5042
Gypsy Moth x Month 3 0.0644 0.3064 - 0.7206

Mean square

Error a 2 14.125 0.027 - 7.683
Error b 6 159.792 0.146 - 23.189

 

lValues transformed using the Box-Cox transformation
2Values tested using the Kruskal-Wallis test

90

Table 3.12. Statistical signiﬁcance of treatment effects from split-plot
ANOVA of ﬁve measures of Lepidoptera diversity in four northern
hardwood sites in 1993.

 

Source of Variation df Species Abundance1 Brilloun's Simpson's
Richness Index2 Index

 

F value of signiﬁcant treatment effects

Site 1 1.164 1.675 x2=0.04 0.650
Month 3 9.969 6.650 x2=14.20 76.661
Site x Month 3 0.658 0.450 - 0.806

Probability of signiﬁcant treatment

effects
Site 1 0.3935 0.3249 0.8832 0.5407
Month 3 0.0095 0.0230 0.0026 0.0001
Site x Month 3 0.6069 0.7268 - 0.5348

Mean square

Error a 2 1 13.625 22.732 - 0.007

0\

Error b 72.292 20.748 - 0.003

 

lValues transformed using the Box-Cox transformation
2Values tested using the Kruskal-Wallis test

91

Table 3.13. Statistical signiﬁcance of treatment effects from split-plot
ANOVA of ﬁve measures of Lepidoptera diversity in four northern
hardwood sites in 1994.

   

 
 

Source of Variation df Species Abundance‘ Brilloun's Simpson‘s
Richness Index Index1

 

 

F value of signiﬁcant treatment effects

Site 1 6.050 0.327 10.589 0.082
Month 3 4.370 3.1 16 3.095 0.156
Site x Month 3 0.700 0.373 0.431 1.233

Probability of signiﬁcant treatment

effects
Site 1 0.1331 0.6249 0.0829 0.801 1
Month 3 0.0592 0.1097 0.1 1 10 0.9224

Site x Month 3 0.5855 0.7762 0.7388 0.3770
Mean square

Error a 2 5.000 4.856 0.004 0.026
Error b 6 52.500 4.019 0.038 0.012

 

1Values transformed using the Box-Cox transformation

92

Table 3.14. Statistical signiﬁcance of treatment effects from split-plot
ANOVA of ﬁve measures of Lepidoptera diversity in four northern
hardwood sites in 1995.

    
   

Source of Variation df Species Abundance2 Brilloun's Simpson's
Richnessl Index Index1

 

F value of signiﬁcant treatment effects

Site 1 0.1 14 x2=2.50 0,451 3.638
Month 3 0.928 x2=3.26 2.388 3.034
Site x Month 3 0.318 - 2.016 1.584

Probability of signiﬁcant treatment

effects
Site 1 0.7678 0.1141 0.5710 0.1967
Month 3 0.4827 0.3530 0.1676 0.1 147
Site x Month 3 0.8123 - 0.2133 0.2887

Mean square

Error 3 2 2.585 - 0.047 0.009
Error b 6 1.986 - 0.01 1 0.003

 

lValues transformed using the Box-Cox transformation
2Values tested using the Kruskal-Wallis test

93

 

 

 

 

 

 

 

 

 

 

III . .
m
+l 16000- A "
’3‘
g
E 120004 .. 0 Branch]
a3 ' O Branch2
>3
'2 8000 " I f ' a Freesoil 1
8 r O Freesoi12
2 4000- -
an
8‘
= i Q a
3 O , _ , :k. ,
2 1993 1994 1995
Year
Eﬂ . . .
a:
+1 16000- B. -
Q
g
5 12000" "a Harrietta]
§ r. Harrietta2
g; 8000- _
5 _ n Mesick]
g o Mesick2
E 4000' :-
M
M
U
E
3 0__._A . a 0 ‘ A a O u . A
2 1993 1994 1995
Year

Figure 3.1. Mean density (18E) of gypsy moth egg masses from four sites
in A) red oak ecosystems (ELTP 20) and four sites in B) northern
hardwood ecosystems (ELTP 45). Means were calculated from ﬁve 0.1 ha
ﬁxed-radius plots

94

.55.:— 53 3 Nu: :3...— eoaa.=u_ao 0.5.5 2502 .33.
mag 3.. 89:0 Eat—:3.“ .3.—cc 9:5 .3.—an.— .m_a>._3:_ 3:02.38 95mm «930—. 9::— ueto 33 625828323
ouo? :3 ES m2: Eat 3a: .maam AU 23 .33 a £23 3. 3 $233.. 360% .5.— 332 .N.n 0.53m

 

 

 

 

 

 

 

 

 

 

 

 

 

=8on 10' :05me III!
mam—2. .23. 0:3. .32 3:35.. .33. 0:3. .32 3:93 :3. 2:3. .82
— - — n p — — - b p _
S o o
:2 1o—
S
.8 .8 m.
.8 ﬁom m
a
.I llllllllo I In.
ow W ow m.
low low .s.
s
I now
2:: .m 8 2.: .<
on E.

 

95

.530:— 3000 3 NH: 50.... 10.33.00 0:03 05:02
6.0283 00:02.38 §ma 0030.. 0...:— ..P...0 1:: 50530055030: 0:03 man 50.: 80: .33. 23— ES 33 :8
0.8.30 33.5.: «030—. 0:0.— ..etm .33 6 EB .33 a .33 2 3 000.38.. 0.300% :8 0502 .n.n 0.5mm..—

x0_002 IOII 020.50: +

 

 

 

   

.392 .23 0:3. >02 6:33. .33. 0:3. >02 30:32 :3. 0:3. .32
— p n n — — — b p n b I.
O_ O O
I ON T O_ r O— .«03
a
1 ON T ON m.
I on s
.. Om \ I Om m
1| CV \\\s m
-9. :9. w
w
. on I On .. on
maan U vaau .m— maa— .<

 

 

 

 

 

 

 

 

 

96

Figure 3.4. Hierarchical clusters created using beta-flexible linkage (B:
-0.25) method and Sorenson's quantitative index of similarity for total

lepidoptera collected from eight sites during A) 1993, B) 1994 and C)
1995.

97

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A, Percent Similarity
O 10 20 30 _ 4O 50 60 70 8O 90 100
l l l l l l l 4 l l
1993
264 species Branch 1
I Branch 2
l | Freesoil 1
I Freesoi12
Harrietta 1
I I Harrietta 2
I Mesick 2
, Mesick 1
B. Percent Similarity
O 1020 3040506070 8090100
I l 1 l l l l 4 l I
1994 I Branch 1
197 |
species * Branch 2
Freesoil l
Freesoil 2
I Harrietta l
Harrietta 2
l I Mesick 1
Mesick 2
C. Percent Similarity
O 10 20 3O 4O 50 6O 7O 80 90 100
l l l 1 I J l l l l
1995
411 species I Branch 1
Branch 2
HA Freesoil 1
Freesoi12
I Harrietta l
* Mesick 2
L—|___I Harrietta 2
Mesick 1

98

Figure 3.5. Hierarchical clusters created using beta-ﬂexible linkage (B:
-0.50) method and Sorenson's quantitative index of similarity for total

lepidoptera collected from eight sites during A) 1993, B) 1994 and C)
1995.

99

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A. Percent Similarity
0102030405060708090100
l l l l J l l l l l
1993 , Branchl
264 species I
I ' Branch 2
I I Freesoill
i I Freesoi12
Harrietta]
Harrietta 2
Mesick 2
I Mcsickl
B. Percent Similarity
0102030405060708090100
l l l l l l l l i l
1994 , Branchl
196 species I
Branch 2
I———- Freesorll
I |——- Freesoi12
I I Harriettal
— Harrietta 2
I ~—- Mesick]
*—-—- Mesick 2
C. Percent Similarity
0 1020 304050 6070 8090100
L l l J l l l l l l
1995
411 species I Branch 1
Branch 2
|—_I——I Freesoill
Freesoi12
Harrietta]
* Mesick 2
‘ I I Ham'ctta 2
Mesick]

100

Figure 3.5. Hierarchical clusters created using average linkage and
Sorenson's quantitative index of similarity for total lepidoptera collected
from eight sites during A) 1993, B) 1994 and C) 1995.

101

A. Percent Similarity
0102030405060708090100

1 _ 1 I I I l 1 1 I I
1993 B h 1
264 species J ranc

I Branch 2

Freesoil l

 

 

 

 

 

 

_L_.

Freesoil 2

 

 

Harrietta 1

 

Harrietta 2
Mesick 2
I Mesick l

 

 

 

 

 

B. Percent Similarity
0102030405060708090100

l I I I I J I I J I
1994 . Branch 1

196 species I

, Branch 2

r——- Freesoil l
L—- Freesoi12

I Harrietta l

 

 

 

 

 

 

 

 

- Harrietta 2
- Mesick 1
- Mesick 2

 

C. Percent Similarity
0 1020 304050 6070 8090100

I J I l l I I l I

l

1995 A

411 species Branch 1
Branch 2

Freesoil l

 

 

 

 

 

 

Freesoil 2

 

 

Harrietta 1
J Mesick 2
- Harrietta 2
- Mesick]

 

 

 

102

$661
17661
1' £661
S661
t7661

 
  

 

 

 

_

_ 18
0.0.82 I _
_

5.3.5: B _ 19.
_

- 8
Sun—gm E _

_ Au
8
can: 05 2.5 :83 35

 

9661

m

m 0

J
a
a
a
w

.930.» 0005 :02. 30.002 6:: 0:02:35
00.? 60:39.0.— :..05..:: 6:: 230.02% 6:: 30:35 00:0 0.0: 10.. EP... «0:33. 6:: .23. a: 05:0:— :8000
0.0. a .30 0:3. .05 .32 a: 05:0:— :0m:00. .2.—:0 A< ”3:3. c0000=00 0.3300: g: 80:33.. 0300.5 gun 0.53%

 

. S661

 

I
6
6
.7

E661

 

 

.<

 

 

 

3:3. 6:: .32v :080@ .2.—am—

o
.8 s
m.
Io... m
a
m
'8
8

103

.0000» 00...: 00.5 38002 6:0 0:0_....0=v
00:0 6005000: 5050:: 6:0 £00000“.— ::0 5:005 00:0 :00 .00.. 50.: gnaw—5. 0:0 .33. 00 05:0:— 500%
0.0. a ::0 0:3. e:0 >02 00 05:2: :00000 ~$.00 A< u:_.=:. 10:00:00 03500: 00 00:01:30 .0005. .0...” 00:3..—

    
  

 
  

 

 

 

 

I a a I a a a I I I m a
m 0 0 m 0 0 0 m m m 0 0
m u m m __
m .8 00 m .00
n 0 u N
0.0.82 I m .8. m 1.2 m
u n a
3.35: E m -02 m -02 W.
- -
.3000; . . m
I m -80 m -80 m.
5:0:— ! m m a
" ucmm " Iona
“ .0 n 4
. Sm . 80

 

 

 

 

 

 

Cam—E 6:0 2:3 :800m 0.0.— A0::_. 0:0 >03: :0000m .2.—0m

104

.0000.» 00...: 00.5 20.002 0:0 0:00.03:
00:0 00030.0.— :..0._...0: 0:0 2.000000— 0:0 5:005 00:... 0.00 00.. 50... 0000.5. 0:0 .23. ..0 05:0... :0000... 0.0—
:— 0:0 05... 0:0 >02 ..0 05:0... :0000... .2..00 A< 0:500 0800—.00 00.0.00: 30.9.0.5 ._0 00:00:25. dd 000w...—

 

  
   

 

 

 

 

 

 

 

ﬂ a "m a mu. am am a am am I mm

0 0 0 0 0 0 0 0 0 0 m ,0 w

:0 o
a a m
w v _ _ 0 m
_ unm _ mm m
0.0.82 I _ _ f m
A
gov—ha: Z _ 1cm _ Om m
_ _ m
=83... I _ -2 _ -0» m.
5:0...— ! _ _ m.
_ loo. _ loo. 3
N
_ .m _ .< w
mm. mm. m
Gaga. 0:0 .203 :800m 302 5:2. 0:0 .503: :0000m 2.0”.— m.

105

.0000» 000... 00.5 30.00.). 0:0 030.003..
0030 0003000.. :00500: 0:0 2.00000... 0:0 5:005 000.0 0.00 000 E00.. .00u:< 0:0 .2.... .0 05:05 :00000 0.0.
A: 0:0 05... 0:0 .00.). .0 05:2: :00000 .2000 3. u:.0:0 00.02.00 00.0.00: 000000005. .0 00:00:00.0. .36 00:»...—

I I
6 6
6 6
C... C.

 

O

 ‘ V661
£661

0.0.003. I

 

ION

l
O
N

VIII/11111.. 966‘
l
g
smmoon 3mg Mmmapun 10 ”WWW

 

 
  

 

030.000.. 0
000000... I

5:00.. i

 
 

é .

_
_
_
_
_
_
_ -8
_
_

 

 

_
_
_
_
low _
_
_
_
_

 

 

 

 

 

 

w .<
8
90:000. 0:0 .203 500$ 000‘. .05... 0:0 .00.): :0000m .00”.—

C
00

106

  

nma. 00¢. 80.

 

:88... I
55.0 E

 

Io.

ION

I?“

1...V

IOW

 

 

A.0:u:< 0:0 2:... :0000m 0.0..

00

.maa. :. 00.0..0.00
20.00.. 0:0 2.0. :. 00.0..0.00 20.00000... 000.5 00..0 ..00000... 0......» £0... :. 00.0..0.00 20.00.. 000.3 00.00 5:00..
.00..0 0.00 000 00.0..0.00:: 0:0 00.05.00 E00. .0:w:< 0:0 2:.. .0 0...:0E :00000 0.0. a 0:0 0::.. 0:0 .02

.0 05:05 :00000 2000 3. 050:0 0000.. 000... 00>: 00.02.00 000.300: u:.000....00 .0 00:00:25.

30. 00m. mom.

 

 

....n 2.0....

I
O

I
2

1cm

I
O
M

 

 

A0::.. 0:0 :02. :0000@ 200m.

aupgmaoN Bugpaag 3‘80 10 aauupunqv

107

 

 

 

 

 

 

3 Late Season (July and August)
2 60
3 '8
U U
2 § 50 -
GD I.
\l
g ﬁ- 40.
0 .3
[a ‘0
a: g 30- Branch ,
8 E . _
. a 20_ . Freesoxl T
° 5‘
8
52° 10-
gE _
.D g 0_ .. . :
< 0 1993 1994 1995 E

Figure 3.12. Abundance of oak-feeding Noctuidae excluding Hypostrotia
pervetrens collected over three years during late season months of July and
August from defoliated and undefoliated red oak sites. Branch sites were
heavily defoliated in 1993, while Freesoil sites were moderately defoliated
in 1994 and heavily defoliated in 1995.

Chapter 4
THE IMPACT OF GYPSY MOTH (LEPIDOPTERA:LYMANTIIDAE)
OUTBREAKS ON THE COMMUNITY STRUCTURE OF ARTHROPODS
IN TWO NORTHERN HARDWOOD ECOSYSTEMS IN NORTHERN
LOWER MICHIGAN

Introduction

Since the term 'guild' was deﬁned by (Root 1967; Root 1973) to mean 'a group of
species that exploit the same class of environmental resources in a similar way,‘ the concept
has been used in many ecological studies to describe arthropod community structure (see
Hawkins and MacMahon (1989) for a review). The guild concept is typically applied in
systems where inter-trophic interactions or species diversity are too complex to be
explained using a species by species comparison.

Guilds have been used to compare the arthropod community structure in forest
canopies over large spatial and temporal scales. Community structure of arthropods was
found to be surprisingly consistent in tropical and temperate forests regardless of
taxonomic relatedness of canOpy trees (Moran and Southwood 1982; Stork 1987).
Regenerating forests in western Oregon were shown to have relatively more sap-sucking
arthropods than did old growth forests (Schowalter 1989).

Guilds have also been used to examine the response of forest arthropod

communities to natural and anthropogenic disturbances. Hurricanes decreased the amount

108

109

of folivores, predators, and detritivores but increased the amount of sap-feeders in
disturbed areas in a tropical rain forest in Puerto Rico (Schowalter 1994). Similarly, forest
management practices such as clearcutting (Schowalter et al. 1981) and partial harvesting
(Schowalter 1995) also decreased the amounts of defoliators, predators, and detritivores
but increased the relative amounts of sap-feeders in disturbed areas.

In the northeastern United States, gypsy moth, Lymantria dispar L., is a major
forest pest responsible for defoliating over 10 million ha of forest between 1985 and 1994
(Butalla 1996). By 1994, gypsy moth infestation had extended as far west as Michigan
and as far south as Virginia (Anonymous 1995). This exotic pest feeds on over 400 North
American woody plant species including economically valuable species such as Quercus
and Populus spp. (Mauffette et al. 1983). Defoliation can make trees susceptible to
mortality agents like two-lined chestnut borer (Agrilus bilineatus Weber) and fungal
pathogens like Armillaria root diseases (W argo 1977).

Gypsy moth can alter species composition of a stand by reducing the relative
abundance of oak species. Preferred tree species such as Quercus experienced 30%
mortality after an average of 37% defoliation over 110 years in New England (Baker 1941)
Repeated defoliation from 1911 to 1931 by gypsy moth not only increased the mortality
rates of preferred host plant species such as oaks in stands with a greater diversity of non-
preferred host plant species ﬁve years following defoliation, but also decreased the relative
proportion of oaks within defoliated stands in New England (Campbell and Sloan 1977).
Likewise, relative proportion of oaks decreased in forests defoliated by gypsy moth in
Pennsylvania between 1971 and 1979 (Gansner et al. 1983).

In addition to altering the relative composition of forests, gypsy moth defoliation
can change nutrient cycling and hydrological patterns. While nitrogen (N) in gypsy moth
frass is immobilized by soil microbes, insect bodies and leaf fragments may be a source of
nitrogen that is lost through leaching (Lovett and Ruesink 1995). Defoliation can also

reduce transpiration and increase soil moisture resulting in increased denitrification (Lovett

110

and Ruesink 1995). Severe gypsy moth defoliation has been demonstrated to increase
water yield and fecal colliforrn concentrations in watersheds in the northeastern United
States (Corbett and Lynch 1987).

Gypsy moth populations and subsequent defoliation may further disturb forest
stands by decreasing canopy cover and altering the microclimate of the stand, increasing
competition within the folivorous feeding guilds, and disrupting inter-trophic interactions
by altering the abundance of prey items available to higher trophic consumers (see chapters
1, 2 and 3). However, little quantitative information is available on the overall effects of
gypsy moth on the native arthropod community structure.

In this study, our objectives were to examine the effects of gypsy moth populations
and defoliation on community structure of native arthropods in Michigan forest stands
experiencing the ﬁrst invasion of gypsy moth. Our goals were to evaluate differences in
community structure of native arthropods between ecosystems that have high productivity
and available N, but low susceptibility to gypsy moth defoliation (northern hardwood) and
ecosystems that have moderate productivity but are very susceptible to defoliation (red

oak).
Malinda
Study Area

This research was conducted in eight sites in the Manistee National Forest in
northern lower Michigan. Manistee National Forest has been described using an ecological
classiﬁcation system (ECS) (Cleland et al. 1993). The ECS is a hierarchical classiﬁcation
that groups ecosystem components at spatial scales varying from the landscape to the stand
level. Ecological landtype phases (ELTPs) are classiﬁed on the basis of soils, landscape

position and natural vegetation. ELTPs are subsets of larger spatial scale units such as

lll

ecological landtypes (ELTs) and landtype associations (LTA). Because stands within the
same ELTP have similar overstory and understory vegetation, soil hydrology, soil type,
nitrogen cycling, productivity and geological history, use of the ECS allows ecosystems to
be replicated in experimental designs with a high level of conﬁdence that experimental plots
are similar.

We randomly chose four sites classiﬁed as ELTP 20 and four sites classiﬁed as
ELTP 45 from maps provided by the Manistee National Forest. All sites ranged from
approximately 12 to 16 ha in area. ELTP 20 was characterized by an overstory dominated
by northern red oak (Quercus rubra L.) with a minor component of white oak (Quercus
alba L.) and a non-diverse ground ﬂora (Cleland et al. 1993). Soils on ELTP 20 were
typically sandy, xeric soils with low productivity (Host et al. 1987; Host et al. 1988).
ELTP 45 had an overstory dominated by northern hardwood species including sugar maple
(Acer saccharum Marsh) with a minor component of American basswood (Tilia americana
L.) (Cleland et al. 1993). This ELTP was notable for its rich herbaceous understory, well
drained mesic soils, and high productivity (Host et al. 1987; Host et al. 1988; Zak et al.
1989; Zak and Pregitzer 1990; Zak et al. 1986).

Dominant tree species within ELTP 20 and 45 also differed in susceptibility to
gypsy moth defoliation. Oak species, which dominated ELTP 20, are highly preferred by
gypsy moth and oak stands often experience severe defoliation during outbreaks (Mauffette
et al. 1983; Twery 1990). Northern hardwood stands, such as ELTP 45, are less
susceptible to defoliation because the dominant canopy species, sugar maple, is only
moderately preferred by gypsy moth (Mauffette et al. 1983; Twery 1990).

In the red oak ecosystem (ELTP 20), two sites were located at 44° 00' N, 86° 00'
W near the town of Branch and two sites were located at 44° 08' N, 86° 09' W near the
village of Freesoil. In the northern hardwood ecosystem (ELTP 45), two sites were
located at 44° 22' N, 86° 44' W near the town of Mesick and two sites were located at 44°

19' N, 86° 44' W near the village of Harrietta.

112

Before 1993, none of our study sites had ever experienced any observable gypsy
moth defoliation (F. Sapio, MI DNR, Forest Health Management, 1993 pers comm, S.
Katovich, USDA For. Serv. NA. S&PF, Forest Health Management, 1992, pers comm).
We surveyed these sites in 1992 and found no visible defoliation. Number of gypsy moth
egg masses/ha, visual estimates of defoliation, and leaf area index were measured in all
stands as part of a related study (see chapter 1). Red oak sites experienced more defoliation
(Table 4.1) and had higher gypsy moth egg mass densities (Figure 4.1) than the northern
hardwood sites from 1993 to 1995. Red oak sites had overall lower leaf area index than
northern hardwood sites, and mid-summer leaf area index was reduced in areas with large
gypsy moth populations in 1994 and 1995 (see chapter 1) (Table 4.1). Within red oak
sites, gypsy moth populations ﬂuctuated annually: outbreak populations at Branch in 1993
were followed by a population 'crash' due to larval infection by Nucleopolyhedrous virus
(NPV) in 1994. Populations at Freesoil increased in 1994 and reached outbreak levels in

1995 (see chapter 1) (Figure 4.1).

Collection of Arthropods

ArthrOpods were collected during 1993, 1994 and 1995 growing seasons.
ArthrOpods were collected four times a summer. Collection times roughly corresponded to
signiﬁcant points in gypsy moth management or biology. For example, typical application
time of the microbial insecticide Bacillus thuringensis Berliner var. kurstaki (Btk) occurs in
mid-May, fourth instar gypsy moth in mid-June, peak defoliation in mid-July, and stand
refoliation in mid-August. During 1993, arthropods were collected on 12-13 May, 15-17
June, 13-15 July, and 10-12 August from red oak sites and on 21-22 May, 22-23 June,
20-22 July, and 18-20 August from northern hardwood sites. In 1994, arthropods were
collected on 10-12 May, 15-16 June, 11-13 July, and 9-10 August from red oak sites and
on 18-20 May , 20-21 June, 19-20 July, 15-17 August from northern hardwood sites. In

H3

1995, collection dates were 15-17 May, 12-14 June, 10-11 July, 15-16 August at red oak
sites and 22-24 May, 19-20 June, 17-19 July, and 21-22 August at northern hardwood
sites. Arthropods were collected from the canopy, shrub and ground strata using a variety

of sampling methods in an effort to adequately represent the total insect diversity within

each site.

Canopy strata- In 1993, canopies were sampled using a pole pruner with a 9 m
extendible handle, modiﬁed sweep nets with a 10 m extendible handle, and suspendable
UV traps. In 1994 and 1995, pole pruners and extended sweep nets were replaced with a
suspendable thermal fogger to facilitate collection of diumally active canopy arthropods.

Pole pruners were used to clip one branch tip, 1 m in length, from the lower canopy
of six trees along a random transect at 20 m intervals through each stand. Branch cuttings
were then placed in plastic bags until they could be examined for arthropods. Extended
sweep net samples were collected along a random transect from six trees at 20 m intervals.
Four sweeps were made with a 50 cm diameter net at each interval.

Ultraviolet traps consisted of a 22 watt UV bulb connected to a photosensor and
powered by a 6 volt motorcycle battery. UV traps were operated approximately 8 hours
each sampling period. A collection funnel and bucket containing insecticidal pest strips
(V apona TM) were attached below the bulb. Traps were raised into the canopy in the center
of each stand. Use of a low wattage bulb and a rain cover over the bulb restricted potential
long—distance attraction of nocturnal arthropods, resulting in a localized sample reﬂective of
endemic arthropod populations (Bowden 1982). The trap was raised into the canopy
approximately 1 to 2 m above the lower edge of the canopy foliage in the center of each
site.

A thermal fogger (Dyna-fogTM) (Erwin 1983) modiﬁed with a radio-controlled
switch was raised into the canopy at ﬁve points (20 m intervals) along a 100 m transect

through the stand. The fogger was suspended by ropes and pulleys placed over large tree

ll4

limbs with an arrow-line. A 25 m2 diameter column of the canopy was then fogged with
1% pyrethroid (Pyrocide Fogging Concentrate 73381“, McLaughlin Gorrnley King
Company) applied for 15 seconds at a rate of 0.377 Uminute. Pyrethroid was selected for
its quick "knock-down ability" and low persistence. The fogged foliage was thoroughly
shaken using additional suspended ropes. Arthropods were collected in 25 l m2 funnels
arranged in a grid below the fogger. All fogging was done between 0400 and 0800 h when

winds were calm.

Shrub strata- Arthropods were collected from shrub and understory vegetation less than
3 m tall using sweep nets. Sweep samples were collected using a 38 cm diameter net.
Sixteen sweeps (4 from each cardinal direction) were made six times along a random
transect through the stand at 20 m intervals. In addition to sweep netting, one malaise trap
with an insecticidal pest strip was set up in the center of each stand. Malaise traps were left

open for 24 hours primarily to collect diurnal ﬂying insects and Hymenopteran parasitoids.

Ground strata- Ground-dwelling arthropods were sampled using 10 unbaited pitfall
traps located along a randomly selected transect through the stand at 10 m intervals. Pitfall
traps contained soapy water and were covered with four legged 'roofs' to minimize capture
of non-ground dwelling arthropods and prevent predation by birds. Traps were left open
for 24 hours. Arthropods collected in pitfall traps were placed in plastic containers with

70% ethanol as a preservative.

Arthropod Identiﬁcation and Guild Assignments

All samples were placed in portable coolers in the ﬁeld and transferred to freezers
Upon arrival at the laboratory. Arthropods were identiﬁed to family by T. Work and

parataxonmists at Michigan State University.

115

Guilds were assigned using systematic literature and corroboration from previous
studies, (LaVigne et al. 1990; Moran and Southwood 1982; Stork 1987). A hierarchical
approach was employed when assigning insect families to guilds. If the guild classiﬁcation
of a family was consistent in the three guild studies cited above, that classiﬁcation was also
used in our study. If a guild classiﬁcation was suggested by only one or two of the studies
or conﬂicts in guild assignment developed, references to natural history from systematic
literature were used to determine guild assignment (Amett et al. 1980; Borror et al. 1989;
Erwin 1983; Goulet and Huber 1993).

Collembola were collected only sporadically and in disproportionately large
numbers relative to other families. Our results were expressed as number of arthropods
and guilds represented the relative proportions of arthropods collected. Collembola were
excluded from analyses to prevent them from overestimating the relative importance of
other groups.

Because our sampling units were large spatial areas (forest stands), we assumed
most of the arthropods spent both their immature and adult life within the stand. For this
reason, members of Moran and Southwood's (1982) and Stork‘s (1987) tourist guild were
reassigned to other guilds. Guild assignments used by LaVigne et al. (1990) for immature
insects were used. Guilds were assigned according to immature life-stages because
members of higher trophic level guilds such as parasitoids and some predators most likely
impacted other insect groups only in the juvenile stage. Likewise, plant tissue-feeders such
as folivorous lepidoptera most likely affect ecosystem processes during immature stages
and experience more effects from direct competition with gypsy moth as immatures.

Guilds included in this study were ants, parasitoids, predators, plant sap-feeders,
plant tissue-feeders and saprophages (Table 4.2). These guilds were included because they
comprised a major proportion of the total arthropods collected each year. Arthropods from
guilds that fed on algae, fungus, animal tissue, pollen and nectar were rarely collected and

were not included in our analyses.

116

Statistical Analysis

Numbers of insects and numbers of families within each guild were recorded at
each sampling period for three years (12 sample periods total). Numbers of insects and
families were expressed as relative proportions of collected totals. Within the red oak
ecosystem, guild data were analyzed using a split-plot ANOVA (e. g. a one-way ANOVA
with repeated measures) to test for differences between sites with high and low defoliation,
and monthly variation (Sokal and Rohlf 1995). Differences in arthropod guilds between
Mesick and Harrietta sites and monthly variation within the northern hardwood ecosystem
were similarly analyzed using a split-plot AN OVA. Data from all months were analyzed
collectively to determine if effects of defoliation and variation among years were signiﬁcant
using a split-plot AN OVA within each ecosystem. Differences between ecosystems were
analyzed using yearly totals with a one-way ANOVA with repeated measures. Variables
that did not meet the assumptions of normality or heteroscedasticity were transformed using
the arcsine transformation (Sokal and Rohlf 1995). Variables that could not be normalized
or homoscedastic through transformation were tested separately for differences between
sites and differences in time using the nonparametric Wilcoxin t-test (Conover 1971). All
analyses were conducted using the software package JMP (SAS Institute Inc. 1995) and
with a: 0.05.

Results

A total of 49,973 arthropods from 207 families were collected and assigned to six
guilds (Table 4.3 and 4.4). Number of total arthropods collected declined from 1993 to
1994 presumably because of adverse weather conditions during collection dates.

Collection dates in 1994 were generally cooler and wetter when compared to either 1993

and 1995.

117

Abundance of Arthropods within Guilds

Within the red oak ecosystem, the relative number of individuals within each guild
did not differ signiﬁcantly between high and low levels of defoliation except in the ant guild
during 1993 (Table 4.5-4.8). Ants comprised a greater proportion of total insects captured
in Freesoil (6.2%) than in Branch (2.7%) in 1993 (Table 4.6). Monthly variation within
guilds was signiﬁcant for ants in 1995, sap-feeders in 1994 and 1995, and saprophages in
1993 and 1995 (Table 4.6-4.9). When data from all months were analyzed, only the
saprophage guild was signiﬁcantly affected by defoliation (Table 4.10). Saprophages
comprised a greater proportion of total insects collected at Branch (28%) than Freesoil
(17%) (Table 4.10). Variation among years was signiﬁcant only for the sap-feeder guild
(Table 4.10).

Within the northern hardwood ecosystem, the relative proportion of arthropods did
not differ between Harrietta and Mesick sites, except for the saprophage guild during 1994
(Table 4.11-4.14). Saprophages comprised a larger proportion of total insects captured in
Mesick (36%) than in Harrietta (22%) during 1994 (Table 4.13). Monthly variation within
guilds was signiﬁcant for ants in 1995 and sap feeders in 1994 and 1995 (Table 4.13-
4.15). When total annual collection data were examined, saprophages were signiﬁcantly
affected by the interaction between site and year (Table 4.16). Relatively more
saprophages were collected from Harrietta (32%) than Mesick (22%) in 1993. However,
Mesick had more saprophages than Harrietta in 1994 (40% and 24% respectively) and
again in 1995 (36% and 20% respectively). Among year variation was signiﬁcant in the
parasitoid guild (Table 4.16).

When differences between ecosystems were examined, only predators and tissue-
feeders signiﬁcantly differed in relative proportions between red oak sites and northern
hardwood sites (Table 4.17 and Figure 4.2). Predators comprised a signiﬁcantly greater
proportion of arthropods at red oak sites (18%) than at northern hardwood sites ( 12%)

118

(Table 4.17). A signiﬁcant interaction between ecosystem and year was apparent for the
tissue-feeder guild (Table 4.17). Tissue-feeders increased in red oak sites from 35% to
41% between 1993 and 1994 respectively, but decreased to 35% by 1995. In the northern
hardwood sites, tissue-feeders decreased from 50% in 1993 to 28% in 1994, then
increased to 41% in 1995. The drop in relative proportion of tissue-feeders during 1994 in

the northern hardwood sites reﬂected a large catch of saprophagous insects, particularly

Sciarid ﬂies.

Diversity of Families within Guilds

Within the red oak sites, the relative number of families within each guild did not
differ between high and low levels of defoliation for any guild (Table 4.18—4.20). Monthly
variation was signiﬁcant for ants in 1995, the parasitoids in 1993, and sap-feeders in 1993
and 1994 (Table 4.18-4.19 and 4.21). When all data were examined, there were
signiﬁcant differences between the relative number of families of predators and tissue-
feeders between Branch and Freesoil sites (Table 4.22). Greater relative proportions of
predator families were observed in Freesoil (26%) than in Branch (23%). However, there
were greater proportions of tissue-feeders in Branch (37%) than in Freesoil (34%) (Table
4.22). No signiﬁcant interactions between defoliation and year were observed for any
guild (Table 4.22). Furthermore within the red oak ecosystems, there was no signiﬁcant
variation among year within any guild (Table 4.22).

Within the northern hardwood sites, no differences in the relative number of
families within each guild were observed between Mesick and Harrietta, except for the
saprophage guild in 1994 (Table 4.23-4.25). In 1994, Mesick sites had more families in
the saprophage guild than did Harrietta sites (30% and 24% respectively) (Table 4.24).
Monthly variation in the number of families within each guild was much higher in the

northern hardwood sites than in the red oak sites. Ants, sap-feeders, and saprophages all

119

varied signiﬁcantly among month in 1994 and 1995 (Table 4.26). Predators varied
signiﬁcantly among months in 1993 and tissue-feeders varied among months in 1995
(Table 4.26). When total annual catch was examined all samples were pooled, no
differences between sites were observed, and variation among year was only signiﬁcant for
the parasitoid and saprophage guild (Table 4.27).

The relative number of families within each guild did not vary signiﬁcantly between
ecosystems except for the saprophage guild (Table 4.28 and Figure 4.3). More
saprophagous families were observed within northern hardwood sites (21%) than in red
oak sites (19%).

D . .
Justification for Guild Assignments

While guilds are commonly used in entomological studies, concerns have been
raised about how the term 'guild' is deﬁned and used in community ecology. The original
deﬁnition of guild was "...a group of species that exploit the same class of environmental
resources in a similar way." (Root 1967; Root 1973). Many investigators have emphasized
the 'same class of environmental resources' e. g. resource utilization rather than 'in a similar
way' e.g. resource acquisition (Simberloff and Dayan 1991). This was seen as
problematic when the role of competition, which is implied by Root's deﬁnition of guild,
was further examined (Simberloff and Dayan 1991). Individuals that exploit the same
resource may not be directly in competition if they are separated from accessing resources
in time or by morphological differences that promote different methods of exploitation

(Simberloff and Dayan 1991).

120

However, when feeding methods and speciﬁc food preference are unknown, or if
species are omnivorous, both of which are true in the case of many arthropods, deﬁning
the phrase 'in a similar way' (resource acquisition) becomes much more difﬁcult. Hawkins
and MacMahon (1989) dismiss the matter of resource acquisition with the reasoning that
the consequences to the ecosystem and the community are the same, regardless of how a
resource is used.

Likewise, guild assignments are only as accurate as are the taxonomic
identiﬁcations (Stork 1987). Arthropod families such as Pentatomidae and Miridae include
both both predatory and sap-feeding species. In this study we classiﬁed both these families
as sap-feeders, recognizing that some predatory species may be potentially incorrectly
assigned to the sap-feeding guild. However, the effects of erroneous guild assignments of
these families may be negligible because pentatomids and mirids accounted for less than
1% and 4% respectively, of the total arthropods collected in every year.

Obviously, the term 'guild' is somewhat elusive and open to interpretation
depending on the objectives of the investigator. We have deﬁned our guilds with more
emphasis on resource utilization than resource acquisition. We justiﬁed our interpretation
of guilds because of the paucity of information on the speciﬁc feeding habits of most forest
arthropods and because our objectives were to examine the effects of a disturbance rather
than to speciﬁcally determine mechanisms by which communities are formed or
maintained.

Another concern regarding the use of guilds is whether the guilds chosen by the
investigator truly reﬂect the community structure (Simberloff and Dayan 1991; Stork
1987). While we could not directly demonstrate vertical interactions between higher and
lower trophic level consumers, we assumed that higher level consumers such as ants,
predators, and parasitoids, interacted with lower level guilds such as tissue-feeders, sap-

feeders, and saprophages. Similar classiﬁcations from previous studies in other temperate

121

forests have shown a high degree of association between predacious and herbivorous

arthropods (Schowalter 1995).

Effects of Gypsy Moth on Community Structure

Because gypsy moth populations and the amount of defoliation observed varied
among years in the red oak ecosystem, effects of gypsy moth on arthropod communities
are relevant only when defoliation by year interactions are signiﬁcant. Defoliation by year
interactions did not signiﬁcantly affect either relative abundance of arthropods or relative
number of families within in any guild. The consistent proportionality of guilds over time
within the red oak ecosystem indicate that gypsy moth did not signiﬁcantly alter the
community structure of arthropods in these sites although individual species could have
been affected.

Previous studies have found gypsy moth populations can affect speciﬁc groups
such as oak-feeding lepidoptera (see chapter 3 and Sample et al. 1993), lepidoptera families
including notodontids and lasiocampids (Sample et al. 1993), and some ground-dwelling
arthropods (Muzika 1994). All of these studies demonstrated an effect of gypsy moth, but
none evaluated the effects of gypsy moth at the community level.

The apparent lack of effect of gypsy moth defoliation on the community structure of
native arthropods may result from the nature of gypsy moth as a disturbance. Intensity and
area of disturbance both play an important role in determining the effects of disturbance
(Huston 1994). The gypsy moth defoliation we observed, while intense when compared to
background levels of defoliation, is relatively benign when compared to disturbances such
as hurricanes or catastrophic ﬁres. Defoliation removes foliage from particular host plants
temporarily during the months of June and July, but trees usually refoliate by August.
While defoliation can lead to mortality of host trees (Kulman 1971) or altered host plant
quality through delayed induced resistance (Schultz and Baldwin 1982), the removal of

122

foliage physically alters forest stands for a relatively short amount of time. Spatial scale of
disturbance can also affect the recolonization of a disturbed area (Huston 1994). Between
1993 and 1995 in the four counties where our surveys were conducted, severe defoliation
did not typically exceed 120 contiguous ha (Sapio 1996). During these years,
approximately 80% of the severe gypsy moth defoliation that occurred was concentrated in
areas no larger than 120 ha (Sapio 1996). Similarly, approximately 80% of moderate
defoliation occurred in areas no larger than 130 ha in size (Sapio 1996). Larger outbreaks
of greater than 480 ha accounted for only 4% of the total area of heavy defoliation and 3%
of moderate defoliation (Sapio 1996).

Intensity and area of disturbance caused by gypsy moth defoliation may provide
explanations for the effects of gypsy moth seen in previous studies and the lack of effect
we observed. In our study sites, outbreaks did not exceed 65 ha and lasted no more than 2
years before they collapsed due to larval infection by NPV. While outbreak areas
encompassed entire stands, defoliation did not extend over the entire range of surrounding
forest stands. We believe that the removal of oak foliage during gypsy moth outbreaks
may be detrimental to those organisms that rely speciﬁcally on oak for food or as habitat,
but these organisms may be capable of recolonizing disturbed areas from surrounding
undisturbed forest. Within red oak sites (ELTP 20), many organisms will rely on oak
foliage to some extent as habitat. These organisms are probably less affected by defoliation
than those that require oak foliage for nutrition. The relatively short time in which habitat is
altered through defoliation may be insufﬁcient to affect community structure as a whole.
Gypsy moth presumably has greater impacts on individual species which rely only on oak.
Guilds such as tissue-feeders presumably utilize multiple species of host plants, and thus
may be less affected by the removal of oak foliage.

Higher trophic level guilds were not affected by gypsy moth in this study. Relative
abundance of higher trophic guilds would be expected to decrease only if there was a

decrease in the abundance of available prey items such as tissue and sap-feeders.

123

However, we observed no effect of defoliation on plant-feeding guilds. We would not
expect the relative abundance of higher trophic guilds to increase during defoliation, unless
generalist predators and parasitoids began to exploit gypsy moth as a resource. While
some native species do exploit gypsy moth as a resource (Cameron and Reeves 1990),
overall increased abundance of the guild may be unlikely because native predators and

parasitoids have not coevolved with the gypsy moth.

Management Implications

The apparent resistance of the arthropod community to gypsy moth invasion has
implications for current management strategies employed for controlling gypsy moth.
Management of gypsy moth using Btk has been shown to reduce the abundance and
diversity of non-target lepidoptera (Johnson et al. 1995; Miller 1990; Sample et al. 1993).
Application of Btk may not only decrease the abundance of native lepidoptera, but may alter
the balance of the trophic community. A decrease in the amount of tissue-feeders would
presumably increase the relative proportion of other guilds and could potentially concentrate
mortality from higher trophic consumers. The abundance of parasitoids decreased
following application of microbial insecticides to gypsy moth in a laboratory setting
presumably due to premature death of hosts and reduced abundance of available hosts
(Vinson 1990). Arthropod orders associated with the soil such as Acarina, Nemotoda,
Collembolla, and Hymenoptera may be adversely affected by residual build-up of Btk in
the soil (Addison 1993).

Ecosystem Differences in Community Structure

Community structure of the red oak and northern hardwood forests were similar

and signiﬁcantly differed only in the predator and tissue-feeder guild. Difference in the

124

relative number of predatory arthrOpods was small (6%) and the biological relevance of this
difference should be interpreted cautiously. However, northern hardwood sites had overall
greater abundance of tissue-feeders than red oak sites after the effects of large catches of
saprophages like Sciarids were removed.

Greater tissue-feeder abundance within northern hardwood sites may reﬂect the
high productivity and plant species diversity typical of this ecosystem. Northern hardwood
ecosystems represented in this study by ELTP 45 have higher rates of net N mineralization
and nitriﬁcation than oak ecosystems such as ELTP 20 (Zak and Pregitzer 1990). These
rates translate into more available nitrogen available to be utilized by trees (Binkley 1986).
Herbivorous insects in nitrogen poor environments generally have slower growth rates
(Mattson 1980) and lower survivorship (Strong et al. 1984) than herbivores in nitrogen
rich environments. Furthermore, northern hardwood ecosystem had richer diversity and
abundance of understory herbaceous plants than did the red oak ecosystem (Cleland et al.
1993). Herbivores that feed on forbs generally have greater growth rates than those that
feed on older foliage from woody plants (Mattson and Scriber 1987; Scriber and Slansky
1981). Presumably, greater abundance and diversity of understory plants supplied tissue-
feeders with an exploitable source of nitrogen which was not available in the red oak

system.
Temporal Variation in Guilds

Numbers of insects and numbers of families in the predator and tissue-feeder guilds
remained relatively constant through time. Within both ecosystems, however, we observed
variation within and among years in the numbers of insects and the number of families in
the ant, parasitoid, sap feeder, and saprophage guilds. This variation was statistically

signiﬁcant, although differences were small and their biological relevance is questionable.

 

125

Seasonal and phenological changes over time may have accounted for much of the
within year variation, but temporal variation may have been partly attributable to our use of
multiple types of insect collection methods. Particular collection methods may have been
differentially affected by environmental factors such as weather. For instance, malaise
traps captured a large proportion of parasitoids such as parasitic Hymenoptera, while
sweep nets tended to catch larger proportions of sap feeders such as aphids. However,
malaise traps may be less effective during high winds (Southwood 1994) because of
decreased Hymenopteran ﬂight activity, whereas sweep netting is less affected by wind.
This particular example would result in over-representation of sap feeders on a particular

sampling date, and would contribute to the variation we have observed.

Guilds and Constancy

Moran and Southwood (1982) deﬁned the term 'proportional constancy' when they
observed that the number of arthropod species within predator and phytophage guilds from
arboreal communities was relatively constant in communities dominated by broad- and
narrow-leaved trees. Proportional constancy was stressed by Moran and Southwood as an
interesting phenomenon that had been previously observed by Simberloff and Wilson
(1969) during arthropod recolonization of six islands where all arthropods had been
previously killed with insecticide. Since then, proportionality of arthropod guilds has also
been demonstrated in tropical forests (Stork 1987). While we cannot deﬁne mechanisms
by which communities are formed or maintained with our observations, we observed
similar 'proportional constancy' within all guilds except saprophages between ecosystems
with different levels of productivity and over a period of three years. Our observations
provide additional evidence that the phenomenon of proportional constancy exists not only

in communities with different evolutionary origins (Moran and Southwood 1982; Stork

126

1987), but also over time in communities with different nitrogen availability, a resource

critically important to insect development and survival.

High Variability of Saprophages

We cannot explain the high variability observed in the saprophage guild.
Saprophages varied between levels of defoliation in the red oak ecosystem, between sites
within the northern hardwood ecosystem, and between ecosystems. These differences
appeared to be independent of the effects of defoliation because the amount of saprophages
was always higher in Branch sites than in Freesoil sites. Saprophages were mostly
represented by Diptera. Our sampling regime may have been spatially too small to
adequately and consistently represent the Diptera, but other factors such as differential
attraction to traps due to environmental heterogeneity may have played a role. Also,
heterogeneity may have been caused by stochastic mortality patterns of animals, which

would be an attractive resource to many saprophagous Diptera, within each site.

127

Table 4.1. Visual estimation of peak defoliation and estimation of mean
LAI (:l:SE) using photosynthetically active radiation transmittance of sites
in red oak and northern hardwood ecosystems.

 

Peak Visual Estimation of Defoliation

 

 

 

 

 

1 9 93 l 994 l 9 95
RerLQak
Branch 80-100% < 20% < 20%
Freesoil < 20% 40-60% 80—100%
mm
Harrietta < 20% < 20% < 20%
Mesick 20-40% < 20% < 20%

Estimated Leaf Area Index!

 

 

1994 1995
RM
Branch 4.48 :1: 0.221 4.71 i 0.115
Freesoil 2.73 i 0.076 2.72 :t 0.151
W
Harrietta 10.9 i 0.226 8.61 :1: 0.229
Mesick 11.9 i 0.189 8.14 :1: 0.222

 

1Means and standard errors were calculated from two sites (n=2)

128

Table 4.2. Arthropod families assigned to six guilds based on feeding

habits and corroboration with previous classifications.

 

 

 

 

 

 

 

 

A n ts
W
Forrnicidae
Parasitoids
Hymenoptera
Aphelinidae Chrysididae Eupelrnidae Platygasteridae
Aulacidae Diapriidae Figitidae Pteromalidae
Bethylidae Dryinidae Ichneumonidae Scleonidae
Braconidae Elasrnidae Mutillidae Scoliidae
Ceraphronidae Encyrtidae Mymaridae Torymidae
Chalcididae Eucoilidae Ormyiidae Trichogrammatidae
Chalcidoidea Eulophidae Perilampidae
Diptera
Acroceridae Ptychopteridae
Pipunculidae Sarcophagidae

Tachinidae
Predators
Wm Diptera
Carabidae Histeridae Asilidae Empididae
Cincindelidae Hydrophilidae Athericidae Rhagionidae
Cleridae Larnpyridae Bombyliidae Syrphidae
Coccinellidae Lycidae Charnaemyiidae Tabanidae
Colydiidae Meloidae Dolichopodidae Xylophagidae
Cucujidae Psephalidae
Derodontidae Pyrochroidae
Dytiscidae Staphylinidae

Table 4.2 (cont'd)

129

 

 

 

 

W [imam (ileum Qtltaﬂtcdcters
Mymeleontidae Anthocoridae Chrysopidae Aranae
Pompilidae Corixidae Corydalidae Chilopoda
Sphecidae Gerridae Hemerobiidae
Vespidae Nabidae Mantispidae

Reduviidae

Saldidae
Plant Sap Feeders
Wm WM W
Coreidae Aphididae Derbidae Phlaeothripidae
Lygaeidae Cercopidae Membracidae
Miridae Cicadellidae Psyllidae
Pentatomidae Cicadidae
Plant Tissue Feeders
Mm 212ch W
Anobiidae Curculionidae Agromyzidae Cephidae
Anthicidae Elateridae Anthomyiidae Cimbicidae
Anthribidae Eucnemidae Argomyzidae Cynipidae
Apionidae Haliplidae Bibionidae Diprionidae
Bostrichidae Mordellidae Cecidomyiidae Symphyta
Brentidae Ostomatidae Chironomidae Tenthredinidae
Buprestidae Phalacridae Chloropidae Xiphyrididae
Byrrhidae Scolytidae Tephritidae Xyelidae
Cerarnbycidae Trogossitidae
Chrysomelidae

Curculionidae

Table 4.2 (cont'd)

130

 

 

 

mm Qcthgeteza— ﬂramtm— Acari
All Lepidoptera Acrididae Phasmatidae Tetranyctidae
Tetrigidae
Tettigoniidae
S apropha ges
CQIQQQLQEQ Diptera
Alleculidae Anisopodidae Heleomyzidae Phoridae
Elmidae Calliphoridae Lauxaniidae Scatopsidae
Endomychidae Camidae Lonchopteridae Sciaridae
Erotylidae Clusiidae Milichiidae Sepsidae
Helodidae Culicidae Muscidae Simulidae
Heteroceridae Curtonotidae Mycetophilidae Sphaeroceridae
Nitidulidae Dixidae Neottiophilidae Stratiomyidae
Oedemeridae Drosophilidae Odiniidae
Ptinidae Dryomyzidae Otitidae
Tenebrionidae
Elam—— demurre— Magenta.— WW
Blattelidae Largidae Panorpidae Diplopoda
Blattidae Rhinoterrnitidae

 

131

Table 4.3. Total number of arthropods collected within red oak (R.O.) and
northern hardwood (N.H.) sites over 1993, 1994, 1995 and overall.

W

 

 

 

 

 

1993 1994 1995 Total

11.0. N.H. 11.0. N.H. 11.0. N.H.
Ants 258 72 ' 121- 44 506 182 1183
Parasitoids 315 449 219 218 672 418 2291
Predators 939 493 1032 907 3071 2076 8518
Sap Feeders 377 381 194 205 1398 1041 3596
Tissue Feeders 2011 3363 2350 1786 4948 6407 20865
Saprophages 1667 1838 979 2044 2565 4427 13520

5567 6596 4895 5204 13160 14551 49973

 

132

 

 

 

.0. we. 00. 00. an. 00. 00.
a. R mm .m . e. R 2 0. 00 00 80208080
5 9. cm 8 0m 3. mm 0. s. a. 0880 2.3:.
e 2 o. _ _ e w 0 e w w 88660 9.0
em 2 mm mm mm 00 em 2 o. N. .2880
a E 2 Na 2 2 w. 2 M: 2 2608.00
_ _ 0 _ _ _ _ _ _ _ 05..

85:30 8.05... 80.03...
cahaﬂ-m 0:02 0O.“ ego-aF-m 0:02 000‘ eukaF—m 0:02 0O.“

 

 

 

 

.=000>0 0:0 ma..— .va... .2: ”0.000 01.7.. 0003000..
=00...00.. 000 00.1. 0.00 000 0......» 00.03.00 00......0. 02.00500 .0 00.....0: .0.0. 0:... .0... 0.00...

133

 

 

5.8 0...: 03.8 8.... 20.... .0... 20.8 8.8 .00.... 00.8 80.8 00... 8028.80
0.0000...

30.... a... 80.8 00.00 .03.. 00.0.. 30.8 8.00 30.8 0:. .00... 02.0 2.8.0
E. c .00 .008 2.0 60.8 0.0.. 30.8 00.. 30.8 8.. 60.8 :0 888... 000
60.8 .000 5.8 $.00 80.8 2.: 5.8 00.8 80.8 0...: 80.8 00.0. 82820
.8: 00.0 80.8 .0... .38 e5. 0...: 3.... 80.8 mm... 2.0.: 3.0 8.2.8...
04.8 e... 5.8 00.0 .8... 00.0 .008 0... 30.8 3.0 5.8 0... 0:...
..00000.. 00000.. ..00000r.. 00000.. ..00000... 00000.. 0......

 

.maa. 000 m3. 0003.00 000000... 0000 00..0 0.3. 000 00000.. 0000 00..0
00.. 0.00.. 00.00..00 00....» w0.000. 0.0 .0 000000000 .000..00..00.. 02.0.00 0... .0 Gumﬁ 000.0.» 000.). .m... 0.00...

134

Table 4.6. Statistical signiﬁcance of treatment effects from split-plot
ANOVA of the relative abundance of insects within six feeding guilds in
four red oak sites in 1993.

@
Source of df Antsl Parasitoidsl Predatorsl Tissue Sap Sapro-

Variation Feeders Feeders2 phages

 

F value of significant treatment effects

Gyspy Moth 1 28.162 2.1 15 1.068 8.550 x2=l.10 8.965
Month 3 0.722 0.626 1.938 2.105 12:] 1.05 7.753
Gypsy Moth 3 0.329 2.836 2.406 6.007 - 1.907
x Month

Probability of significant treatment effects

Gyspy Moth 1 0.0337 0.2831 0.4100 0.0998 0.2936 0.0958

Month 3 0.5744 0.6242 0.2248 0.2010 0.01 15 0.0173
Gypsy Moth 3 0.8053 0.1283 0.1658 0.0307 - 0.2296
x Month

Mean square

Error a 0.0012 0.0056 0.016 52.729 - 138.963
Error b 6 0.0036 0.0032 0.013 177.293 - 138.346

N

 

1Values transformed using arcsine 04)"-5
2Values tested using the Kruskal-Wallis test

135

Table 4.7. Statistical significance of treatment effects from split-plot
ANOVA of the relative abundance of insects within six feeding guilds in

four red oak sites in 1994.

 

 

Source of df Ants Parasitoidsl Predators Tissue Sap Sapro-
Variation Feeders Feeders phages
F value of signiﬁcant treatment effects
Gyspy Moth 1 0.334 0.053 5.135 0.140 0.801 0.022
Month 3 1.495 0.399 0.834 1.088 1 1.881 0.824
Gypsy Moth 3 0.309 2.022 1.111 0.422 1.868 0.169
xMonth
Probability of significant treatment effects
Gyspy Moth 1 0.6215 0.8390 0.1517 0.7444 0.4653 0.8956
Month 3 0.3083 0.7591 0.5225 0.4231 0.0062 0.5268
Gypsy Moth 3 0.8189 0.2124 0.4153 0.7439 0.2359 0.9134
x Month
Mean square

Error a 2 6.643 0.0054 7.738 221.634 5.109 142.378
Error b 6 9.058 0.0048 131.048 335.578 1.224 130.1 14

 

lValues transformed using arcsine (y)°-5

136

Table 4.8. Statistical significance of treatment effects from split-plot
ANOVA of the relative abundance of insects within six feeding guilds in
four red oak sites in 1995.

 

Parasitoids Predators Tissue Sap Sapro-

Souree of df Ants
Variation Feeders Feeders phagesl

       

 

F value of significant treatment effects

Gyspy Moth 1 0.042 0.448 0.119 7.355 0.034 2.375
Month 3 7.274 3.682 1.011 3.885 4.639 5.775
Gypsy Moth 3 0.590 0.287 0.201 2.217 0.515 0.958
x Month

Probability of significant treatment effects

Gyspy Moth 1 0.8567 0.5720 0.7629 0.1133 0.8708 0.2632
Month 3 0.0201 0.0818 0.4506 0.0740 0.0526 0.0334
Gypsy Moth 3 0.6437 0.8336 0.8922 0.1868 0.6868 0.4709
x Month

Mean square

Errora 2 0.0007 0.0009 0.012 0.002 0.002 0.0001
Errorb 6 0.0001 0.0004 0.009 0.006 0.002 0.0001

 

lValues transformed using arcsine 000-5

137

Table 4.9. Mean values (iSE) of the relative abundance of arthropods in
guilds that varied significantly among month in red oak sites (n=4).

      
     

F ratio P value

 

August

 

Guild May June July

 

Ants 1995 3.5 (0.5) 2.3 (0.9) 2.6 (0.8) 5.1 (0.4) 7.27 0.020
Sap Feeders 1994 0.0 (0.0) 3.3 (1.1) 4.4 (0.7) 3.2 (0.6) 11.88 0.006
Sap Feeders 1995 3.8 (2.7) 7.2 (2.2) 13.9 (1.0) 13.0 (1.7) 4.64 0.053
Saprophages 4.9 (0.7) 30.4 (6.3) 41.4 (8.8) 37.3 (12.0) 7.75 0.017

1993
Saprophages 23.7 (7.8) 30.5 (6.7) 11.9 (1.1) 13.9 (2.3) 11.20 0.020

1 995

 

138

Table 4.10. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of insects within six feeding guilds in four
red oak sites over three years.

Source of df Ants Parasitoids Predators Tissue- Sap- Sapro-

   

 

Variation Feeders Feeders phages
F value of significant treatment effects
Gyspy Moth 1 2.272 0.621 0.510 3.968 3.613 60.832
Year 2 6.817 1.341 0.836 3.134 8.977 3.901
Gypsy Moth 2 5.376 2.030 0.090 3.650 1.507 0.761
x Year

Probability of significant treatment effects
Gyspy Moth 1 0.2707 0.5132 0.5494 0.1846 0.1977 0.0160
Year 2 0.0515 0.3584 0.4972 0.1518 0.0332 0.1 149
Gypsy Moth 2 0.0735 0.2463 0.9161 0.1253 0.3252 0.5245

xYear

Mean square
Error a 1655-4 2888-5 1.27E-3 2905-3 1685-4 6.3 113-4
Error b 4 8. 1413—4 3895-5 4398-3 1.73E-3 5.22E-4 5.0415-3

N

 

 

139

 

 

 

sod 3.2 and 8.2 can mean 3.: EQ and 3.2 SE 3.3 mowaﬁoaam
who—wool

80.8 8.8m are 8.? 21.8 3.3 a; 8.? 83: an? :88 3.2 2.3:.
a. . : mg 84.: S... 88.8 43 Go. : M: .m 84.8 3: :2; 5.2 388m 98
as: men. :90 8.2 sod 8.2 $19 8.8 cad N; am: one 3.285
$08 an 5.8 mom 62: 2% 2.8.8 8w an: S.“ an: 5.0 8235.
$4.3 3o 308 was 64.8 :5 82: 85 ﬁne :2 :2: mg 3.2
6.332 «:35: 6.382 532.3: :38: 3.2.5: 226

 

e3. 59: 033:3 “we—Eu «5.38 in a: 3:25.55 .aceuueneua 3:20.. 05 he $53 3...?» =32

.maa— ES n.3— 5952. 3332 .39: «8% 95 6:: 332.5: .30: mega

.u—é 03:9

140

Table 4.12. Statistical significance of treatment effects from split-plot
ANOVA of the relative abundance of insects within six feeding guilds in
four northern hardwood sites in 1993.

 

Source of df Antsl Parasitoids2 Predators2 Tissue Sap Sapro-
Van'ation Feeders Feedersl phages
F value of significant treatment effects
Site 1 x2=3 .43 0.514 5.182 1.439 12:00] 2.553 f
Month 3 38:20] 1.743 1.647 1.132 12:12.20 3.009 4
Site x Month 3 - 1.189 0.025 0.058 - 3.942
Probability of significant treatment effects '
E
Site 1 0.0641 0.5447 0.1506 0.3532 0.9164 0.2512 E
Month 3 0.5705 0.2575 0.2757 0.4048 0.0067 0.1 163
Site x Month 3 - 0.3902 0.9939 0.9801 - 0.0721
Mean square
Error a 2 - 0.0035 0.0007 287.921 — 18.645
Error b 6 - 0.0066 0.0140 570.325 - 90.687

 

 

1Values tested using the Kruskal-Wallis test
2Values transformed using arcsine (3))"-5

141

Table 4.13. Statistical significance of treatment effects from split-plot
ANOVA of the relative abundance of insects within six feeding guilds in
four northern hardwood sites in 1994.

 
 
 

 
 

 
 

 

 
 

 

Source of df Ants1 Parasitoids2 Predators Tissue Sap Sapro-
Variation Feeders Feeders2 phages1

 

 

F value of significant treatment effects

 

I‘-

Site 1 x2=0.29 1.939 1.083 0.966 0.1 10 x2=7 .46 -

Month 3 x2=7.49 1.497 1.566 1.321 14.566 x2=2.51

Site x Month 3 - 1.643 0.035 0.351 3.412 -

Probability of significant treatment effects ..

it

Site 1 0.5895 0.2984 0.4072 0.4293 0.7721 0.0063

Month 3 0.0577 0.3078 0.2926 0.3518 0.0037 0.4726

Site x Month 3 - 0.2765 0.9905 0.7907 0.0937 -

Mean square
Error a 2 - 0.0030 109.688 100.445 0.0038 —
Error b 6 - 0.0019 77.935 246.214 0.0021 -

 

lValues tested using the Kruskal-Wallis test
2Values transformed using arcsine (y)0.5

142

Table 4.14. Statistical significance of treatment effects from split-plot
ANOVA of the relative abundance of insects within six feeding guilds in
four northern hardwood sites in 1995.

      

   

Source of df Antsl Parasitoids Predators Tissue Sap Saprophages
Variation Feeders Feeders

 

F value of significant treatment effects

p—s

Site 0.059 3.434 0.318 0.361 1.652 4.997
Month 3 29.565 1.026 1.444 1.052 86.119 3.703

SitexMonth 3 0.919 0.182 1.044 0.761 3.341 3.774

Probability of significant treatment effects

 

Site 1 0.8310 0.2050 0.6299 0.6090 0.3275 0.1549
Month 3 0.0005 0.4450 0.3203 0.4358 <0.0001 0.0809
Site x Month 3 0.4862 0.9048 0.4386 0.5558 0.0973 0.0781

Mean square

Errora 2 19713-5 0.00009 0.0130 0.0210 4.9613—5 0.0050
Errorb 6 4528-6 0.00016 0.0051 0.0098 7.318—5 0.0045

 

1Values transformed using arcsine 000-5

     

143

Table 4.15. Mean values (:tSE) of the relative abundance of arthropods in
guilds that varied significantly by month in northern hardwood sites (n=4).

Guild May June July August F ratio P value

 

 

 

Ants 1995 0.0 (0.0) 2.0 (0.5) 0.7 (0.1) 1.1 (0.3) 29.35 0.001

Sap Feeders 1994 0.2 (0.2) 5.3 (1.3) 2.0 (0.1) 3.8 (1.8) 14.57 0.004

Sap Feeders 1995 0.5 (0.2) 10.0 (0.7) 3.4 (0.3) 4.5 (0.6) 86.12 0.000

 

 

144

Table 4.16. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of insects within six feeding guilds in four
northern hardwood sites over three years.

 
 
 

 

Parasitoids Predators Tissue- Sap Sapro-

   

Source of (if Ants

 

Variation Feeders Feeders phages
F value of significant treatment effects
Site 1 1.034 0.022 0.948 1.902 0.01 1 1 1.513
Year 2 1.209 10.200 3.029 5.413 5.925 2.658
Site x Year 2 0.822 0.846 0.593 0.903 0.555 22.629
Probability of significant treatment effects
Site 1 0.4163 0.8949 0.4330 0.3018 0.9268 0.0770
Year 2 0.3885 0.0269 0.1582 0.0728 0.0637 0.1844

 

Site x Year 2 0.5022 0.4938 0.5947 0.4747 0.6125 0.0066

Mean square

Errora 2 32313-5 3.05E-5 1.6413-3 1.57E-3 2.49E-4 1.41E-3
Errorb 4 3.02E-6 8745-6 2528-3 8.6lE-3 2.0015-4 1.03E-3

 

145

Table 4.17. Statistical significance of treatment effects of ecosystem and
year from split-plot ANOVA of the relative number of insects within six
feeding guilds.

         

df Ants Parasitoids Predators Tissue- Sap- Sapro-

Source of

 

Variation Feeders Feeders phages
F value of significant treatment effects
Ecosystem 1 23.639 3.246 16.108 1.187 5.350 2.848
Year 2 3.545 4.683 3.629 1.981 13.254 0.802
Ecosystem x 2 1.631 1.224 0.689 6.992 1.790 2.615
Year

Probability of significant treatment effects
Ecosystem 1 0.0028 0.1 165 0.0070 0.3178 0.0600 0.1425
Year 2 0.0617 0.0338 0.0585 0.1805 0.0009 0.4710
Ecosystem x 2 0.2363 0.3313 0.5207 0.0097 0.2088 0.1142

Year

 

Mean square
Error a 6 1308-4 7 .00E-5 1.34E-3 3.91E-3 2.4OE-4 9.79E-3
Error b 12 12015-4 1.725-4 2638-3 5805-3 3.91E-4 6.56E-3

 

146

Table 4.18. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within six feeding guilds in four
red oak sites in 1993.

Source of df Ants Parasitoids Predators Tissue Sap Sapro-
Variation Feeders Feeders phages

   

 

F value of significant treatment effects

Gyspy Moth 1 1.249 5.973 0.176 0.110 0.142 0.952
Month 3 0.330 6.038 0.353 0.989 4.917 0.138
Gypsy Moth 3 1.135 1.598 0.525 1.302 1.852 2.724
x Month

 

Probability of significant treatment effects

Gyspy Moth 1 0.3800 0.1345 0.7157 0.7715 0.7424 0.4321
Month 3 0.8045 0.0304 0.7889 0.4588 0.0468 0.9336
Gypsy Moth 3 0.4075 0.2858 0.6809 0.3572 0.2385 0.1369
x Month

Mean square

Errora 2 1805-4 9908-4 0.0036 0.0088 3.42E-4 0.0065
Errorb 6 70115-5 0.0019 0.0087 0.0016 9.01E-4 0.0038

 

147

Table 4.19. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within six feeding guilds in four
red oak sites in 1994.

Tissue Sap

     

Source of df Antsl

 

.“"""'."‘-“"

Variation Feeders Feeders phages
F value of significant treatment effects
Gyspy Moth 1 12:1 .73 0.689 x2=0.33 12.492 1.025 2.77E-4
Month 3 x2=6.56 0.084 x2=4.09 0.715 18.054 1.912
Gypsy Moth 3 - 1.165 - 0.783 0.688 0.389
x Month
Probability of significant treatment effects
Gyspy Moth 1 0.1889 0.4938 0.5632 0.0716 0.4179 0.9882
Month 3 0.0873 0.9661 0.2522 0.5778 0.0021 0.2288
Gypsy Moth 3 - 0.3979 - 0.5453 0.5917 0.7654
x Month
Mean square

Error a 2 - 0.0033 - 6.54E-5 0.0016 0.0022
Error b 6 - 0.0025 - 0.0033 0.0004 0.0025

 

lValues tested using the Kruskal-Wallis test

 

148

Table 4.20. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within six feeding guilds in four
red oak sites in 1995.

 

Source of df

Sapro-

Sap

 

 

Variation Feeders Feeders phages
F value of significant treatment effects
:-
Gyspy Moth l x2=0.54 1.829 0.347 0.666 0.474 0.551
Month 3 78:10.86 1.712 1.596 1.702 1.117 0.402
Gypsy Moth 3 - 0.156 2.223 0.617 0.301 0.765
xMonth
Probability of significant treatment effects
Gyspy Moth 1 0.4616 0.3088 0.6157 0.5003 0.5624 0.5352
Month 3 0.0125 0.2632 0.2861 0.2650 0.4135 0.7570
Gypsy Moth 3 - 0.9217 0.1862 0.6290 0.8242 0.5536
xMonth
Mean square

Error a 2 - 3005-5 7398-4 0.0014 1.76134 0.0016
Error b 6 - 0.0016 7.2lE-4 0.0015 5.69E-4 0.0013

 

lValues tested using the Kruskal-Wallis test

149

Table 4.21. Mean values (iSE) of the relative number of families of
arthropods in guilds that varied significantly among month in red oak sites
(n=4).

 

Guild May June July August F ratio P value

 

 

Ants 19951 2.6 (0.1) 2.0 (0.2) 1.5 (0.0) 1.5 (0.0) x2=10.87 0,013
Parasitoids 1993 8.2 (3.3) 12.4 (2.2) 20.2 (1.7) 17.8 (2.4) 6.03 0.030 I..-

Sap Feeders 1993 5.0 (1.7) 13.0 (2.2) 8.7 (1.0) 7.5 (0.6) 4.92 0.047

Sap Feeders 1994 0.0 (0.0) 8.2 (1.2) 7.0 (1.1) 8.6 (1.9) 18.05 0.002

 

 

lMonthly variation was tested with the Kruskal-Wallace test.

150

Table 4.22. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within five feeding guilds in
four red oak sites over three years.

 
 

 
 

      

Source of df Parasitoids Predators Tissue- Sap- Sapro-

 

Variation Feeders Feeders phages
F value of significant treatment effects
Gyspy Moth 1 1.039 68.177 229.407 0.357 0.1 12
Year 2 2.860 2.735 1.104 2.283 0.723
Gypsy Moth 2 1.761 2.213 1.203 1.931 2.070
xYear
Probability of significant treatment effects
Gyspy Moth 1 0.4152 0.0144 0.0043 0.6110 0.7700
Year 2 0.1694 0.1784 0.4153 0.2180 0.5393
Gypsy Moth 2 0.2827 0.2253 0.3899 0.2589 0.2415
xYear
Mean square

Error a 2 2.641 1 0.5253 0.1323 2.6584 0.7544

Error b 4 2.0585 6.4668 4.4117 1.0097 7.5342

 

"-— n Janna-.1

 

151

Table 4.23. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within six feeding guilds in four
northern hardwood sites in 1993.

  
  

 
 

    
 

 

Source of df Antsl ParasitoidsI Predatorsl
Variation Feeders Feeders phagesl

 

Tissue Sap Sapro-

 

F value of significant treatment effects

 

I?

Site 1 x2=1.26 x2=0.71 12:1.33 0.159 0.455 x2=0.72 -

Month 3 x2=2.20 x2=6.64 x2=9.09 4.096 0. 125 12:1 .04

Site x Month 3 - - - 1.244 0.135 -

Probability of significant treatment effects .

{i

Site 1 0.2623 0.4005 0.5135 0.7284 0.5696 0.3962

Month 3 0.5329 0.0842 0.0281 0.0670 0.9418 0.7907

Site x Month 3 - - - 0.3736 0.9356 -

Mean square
Error a 2 - - - 0.0018 7.74E-4 -
Error b 6 - - - 0.0025 0.0020 -

 

lValues tested using the Kruskal-Wallis test

152

Table 4.24. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within six feeding guilds in four
northern hardwood sites in 1994.

 

 

Source of df Antsl Parasitoidsl Predators Tissue Sap Sapro-
Variation Feeders Feeders phages

 

 

F value of significant treatment effects

 

B
Site 1 x2=0.06 x2=1.28 1.085 1.524 0.030 63.244
Month 3 38:20.75 x2=5 .02 0.444 2.964 5.635 6.159
Site x Month 3 - 1.165 2.226 1.215 1.622 2.340
Probability of significant treatment effects
5.
Site 1 0.8035 0.2578 0.4070 0.3424 0.8787 0.0154
Month 3 0.0001 0.1702 0.7305 0.1 193 0.0352 0.0291
Site x Month 3 - 0.3979 0.1858 0.3823 0.2808 0.1728
Mean square
Error a 2 - 0.0033 0.0028 5.9013-4 9.3 115-4 1.9OE-4
Error b 6 - 0.0025 0.0026 8. 1313-4 6875-4 3.04E-4

 

lValues tested using the Kruskal-Wallis test

 

153

Table 4.25. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within six feeding guilds in four
northern hardwood sites in 1995.

m
Source of df Antsl Parasitoids Predators Tissue Sap Sapro-

Variation Feeders Feeders phages

 

F value of significant treatment effects

Site x2=0.10 1.315 0.193 0.728 5.083 3.138
Month 3 12:14.37 0.872 2.329 12.202 7.248 8.526

~

Site x Month 3 - 0.799 3.420 2.354 0.890 0.229
Probability of significant treatment effects
Site 1 0.7506 0.3701 0.7033 0.4834 0.1529 0.2185

Month 3 0.0024 0.5058 0.1740 0.0058 0.0202 0.0139

Site x Month 3 - 0.5381 0.0934 0.1713 0.4984 0.8253
Mean square

Error a 2 - 0.0016 0.0029 8.53E-4 7.30E-4 4.0015-4

Error b 6 - 9015-4 7025-4 3258-4 8.82E-4 8.0215-4

 

lValues tested using the Kruskal-Wallis test

 

154

Table 4.26. Means values (iSE) of the relative number of families of
arthropods from guilds that varied significantly among month from four
northern hardwood sites.

 

u.

 

August

 
 

F ratio

 
 

 

 

Guild May P value
Ants 19941 0.0 (0.0) 2.0 (0.2) 1.6 (0.5) 2.3 (0.8) x2 =20.75 0,000
Ants 19951 0.0 (0.0) 1.4 (0.1) 2.5 (0.1) 3.0 (0.8) x2=l4.37 0,002
Predators 44.4 (2.2) 24.6 (0.8) 22.1 (1.8) 23.8 ( 1.9) X2 =6.64 0.028
19931
Sap Feeders 1.7 (1.7) 5.0 (0.7) 7.4 (1.3) 8.7 (1.6) 5.63 0.035
1994
Sap Feeders 3.8 (1.5) 8.1 (1.2) 10.1 (1.0) 13.4 (1.6) 7.25 0.020
1995
Tissue Feeders 21.2 (1.6) 18.7 (1.0) 15.7 (0.5) 14.1 (1.2) 12.20 0.058
1995
Saprophages 37.1 (5.5) 21.1 (2.4) 25.0 (0.4) 27.1 (2.8) 6.16 0.029
1994
Saprophages 35.7 (1.6) 27.2 (0.4) 29.6 (1.2) 26.7 (1.5) 8.53 0.014
1995

 

1Monthly variation was tested with the Kruskal-Wallace test.

155

Table 4.27. Statistical significance of treatment effects from split-plot
ANOVA of the relative number of families within five feeding guilds in
four northern hardwood sites over three years.

Tissue- Sap- Sapro-

Source of df Parasitoids

         

Predators

 

Variation Feeders Feeders phages
F value of significant treatment effects

Site 1 12: 1.59 1.412 x2: 0.33 1.923 0.000

Year 2 12: 7.54 0.108 12: 3.08 1.965 15.689

Site x Year 2 - 0.015 - 0.183 0.6718

Probability of significant treatment effects

Site 1 0.2069 0.3567 0.5632 0.2993 0.9952
Year 2 0.0231 0.9001 0.2142 0.2544 0.0128
Site x Year 2 - 0.9854 - 0.8396 0.5604

Mean square

Error 3 2 - 0.1980 - 1.2265 6.6931
Error b 4 - 9.7993 - 2.8246 1.6386

 

156

Table 4.28. Statistical significance of treatment effects of ecosystem and
year from split-plot ANOVA of the relative number of families within five
feeding guilds.

   

Source of df Parasitoids Predators Tissue- Sap- Sapro-

 

Variation Feeders Feeders phages
F value of significant treatment effects
Ecosystem 1 3.514 0.010 0.156 0.740 13.009
Year 2 6.663 1.316 1.522 4.282 4.066
Ecosystem x 2 2.657 1.077 0.300 0.370 1.269
Year

Probability of significant treatment effects

Ecosystem 1 0.1 100 0.9257 0.7065 0.4228 0.01 13
Year 2 0.01 15 0.3042 0.2575 0.0395 0.0448
Ecosystem x 2 0.1 109 0.3713 0.7465 0.6987 0.3162
Year

Mean square

Error 3 6 4.2087 6.2561 8.6454 1.8736 2.4897
Error b 12 2.7176 7.8309 6.6115 0.1689 5.8398

 

 

157

 

 

 

 

 

 

 

 

 

 

{I} . . .
m
+1 16000“ A. ’
i
E 12000. - o Branchl
:3 i I? ‘ 0 Branch 2
>6
"2" 8000 ' I _ a Freesoil 1
.8 . o Freesoi12
E 4000- .
ED
3
r: I D O
3 . — s I ' .
2 1993 1994 1995
Year
If} . . .
m
+1 16000- B. r
’6?
i
a 12000 - -
E o Harriettal
E3 .0 Harrietta2
a; 8000 4 -
5 _ n Mesick 1
E o Mesick2
E 4000 4 .
90
M
o r
r:
3 0 _ - . . D O C . D O - ‘ ' A
5 1993 1994 1995
Year

Figure 4.1. Mean density (:tSE) of gypsy moth egg masses from four sites
in A) red oak ecosystems (ELTP 20) and four sites in B) northern
hardwood ecosystems (ELTP 45). Means were calculated from five 0.1 ha
fixed-radius plots

158

«a. 3 m3. :8... 8.... 185:2

52:3: .58 E... :3 e9. .58 Ea... 688:8 3....» 95.8.. in .3.—=3 mix—35.... a: .35.... 95.23— .N... 9.53%

am
.

Emswgeasta. he oozaccaa< 9523— 98

ca
.

c.
.

o

 

23.

T

mowanaoaam

 

0| I p.83". new

Encoun— can;

 

E239...

 

ol . 833...

or. 22

mowagnoaam

 

01 1 8300". cam
Encoun— 2.3....

 

‘T

8235....

.1 5=<

 

mowasaoaaw

 

0'. I eognm QNW

 

 

n.3—

ceeB—Za: 505.82

whoVool gunman

Oli I 8932..
I «Ragga.

0|

4 as.
a.

 

on

on
.

ow
.

 

mg—

0 83.3053
oi I838". new

£300... 25%..

 

o 8232..
0| .. 3633......

01.. 8=<

 

 

cam

mommienam

9| lav—wool mum

mac—econ 25$...

 

 

 

o 8232..
ol . ”22.3....

oi $=<

 

 

n.3—

‘1

momacaoaam

 

oi nae—.8... new

 

‘1

£300.... 25$...

832..
.. £2.25.

 

0

 

I 3.3.

 

Jan. to“

«£11

159

 

Emanuxmcﬁﬁah Ec~=t< a: 33:52 0.523— Q»
9.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.33 3 n2: 5?: m8...” 3936...:— 595.8:
.58 1:: “.3 to.— .58 SP... vogue—.8 3....» ”5.50.. in 5:33 85:5. 62.9.5.3 a: agar—23.5 033.3— .m... «hang

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 om cw am am o._ c 8 awn . mm c
.1! momagnoaam mow—£353
9| I ﬂown..."— aam all . Eouoou mam
Eouoon— 035. ? “accoun— can?
I .. mumogmﬂam ell - 86:32am
T 39395 0 3852;
m2: of 2.2 HP. L 3.2.
0 «unannoaam mowasaohqam
0| I who—goo..— gam 0| .. “Cocoon— mam
£38m 33$. 7 ﬂovoom osmmﬁ.
all I mEozmaam olIIIL- mEoszam
9 E2305 0 3980.5
3a 0+ 9.2 3.: 0+ .3
or mowasaoaam momazaoaam
cl- 388m .5 elbows; ea
who—Bum can»? 7 whouoom can?
.. 3.26.5; all I muﬁzmﬂum
of ﬂea—BE o ﬂea—got
33 0.. 2... ma .3
$836.33 505.52 .m sac com .<

APPENDICES

APPENDIX A

Appendix A

Table A1. Total number of Noctuid species collected from four red oak
sites and four northern hardwood sites over three years.

Re ak orthern
Hardwood

 

1 19 19 l l 4199

__

Abagrotis altemata (GrL) - l 1 l 1
Achatia distincta an. l9 1 3 - l 4
Achatia spp. 1
Acronicta dactylina GrL -
Acronicta haesitata (Gn.) -
Acronicta hamamelis Gn. -
Acronicta hasta Gn. -
Acronicta impleta Wlk. -
Acronicta inclara Sm. 1
Acronicta increta Morr. 3
Acronicta modica Wlk. -
Acronicta momla G. & R. -
Acronicta ovata Grt. 12
Acronicta tritona (an.) - -
Agriopodesfallax (H. -S.) l
Agriopodes teratophora (H. -S.) -
Agrotis ipsilon (Hufn.) l - -
Amphipoea americana (Spcyer) - - 11 - - -
Amphipoea velata (Wlk.) - - 11
Amphipyra tragopoginis (CL) - - - 1 - -
Anagrapha falczfera (Kby.) l - - - - -
Anaplectoides prasina (D. & S.) - 2 -
Anathix ralla (G. & R.) - - -
Autographa precationis (Gn.) - - -
Baileya australis (Grt.) - -
Baileya ophthalmica (Gn.) - 1 -
Bleptina caradirinalis Gn. - - 2
Bomolocha baltimoralis (Gn.) - 1 2
Bomolocha deceptalis (Wlk.) - - - -
Bomolocha madefactalis (Gn.) - - -
Callopzls'tria cordata (Ljungh) 1 3 4 1
Callopistria mollissima (Gn.) - - - -
Catocala amica (an.) - - 2 -

U.) h— I —n N A I I

N I U) -—- I I I
I I I I
I I I
I

u=nt—bt-—ANI—II
U)
I
I
I

I—5 I
I
I

II—tN—I
I—INII—tl u—A
w—d—l—bo—Oh-‘I

160

Table A1. (cont’d)

Catocala antinympha (I-Ibn.)
Catocala ilia (Cram)
Catocala unijuga Wlk.
Ceramic tenebrifera (Wlk.)
Charadra deridens (Gn.)
Chrysanympha formosa (Grt.)
Chytonix palliam'cula (Gn.)
Cissusa spadix (Cram)
Capivaleria grotei (Morn)
Cosmic calami (l-Iarv.)
Crocigrapha normani (Gn.)
Crymodes devastator (Brace)
Cryphia pallida (B. & L.)
Egira alternans (Wlk.)

Egira dolosa Grt.

Elaphria festivaides (Gn.)
Elaphria georgii (Moore & Rawson)
Elaphria veriscolor (Grt. )
Enargia decolor (Wlk.)
Eueretagrotis attenta (Grt.)
Euplexia benesimilis McD.
Eupsilia monisoni (Grt.)
Eupsilia m’stigmata (Grt.)
Eutolype rolandi Grt.

Euxoa declarata (Wlk.)
Euxoa messoria (Ham)
Euxoa perpolita (Morn)
Ewcoa redimicula (Morn)
Euxoa tessellata (Ham)

F eltia geniculata G. & R.

F eltia jaculifera (Gn.)
Galgula partita Gn.
Graphiphora hampica (Grt.)
Hadeninae spp.
Homohadena inﬁxa (W1k.)
Homorthodesﬁufurata (Grt.)
Hypenodes caducus (Dyar)
Hypenodesfractilinea (Sm.)
Hypenodes sombrus F gn.
Hypenodes spp.
Hyperstrotia pervertens (B. & McD.)

161

NWII

DJ

I D—i w I I—I I

—I h—I I

h—II—‘H'

— I—a I

lt—ID—OuprU—II NIWNI

III-I I

NIIHHr—th—NI

how—AMI

I—iht—II

II—wal

.b.

Table A1. (cont’d.)

Hyperstrotia villiﬁcans (B. & McD.)
Hyppa xylinoides (Gn.)

Idia aemula an.

Idia americalis (Gn.)

Idia diminuendis (B. & McD.)
Idia julia (B. & McD.)

Idia lubn'calis (Gey.)

Idia new species

Idia rotundalis (Wlk.)
Ipimorpha pleonectusa Grt.
Lacanobia lilacina (Ham)
Lacinipolia implicata McD.
Lacinipolia lorea (Gn.)
lacinipolia lustralis (Grt.)
Lacinipolia meditata (Grt.)
Lacinipolia olivacea (Mom)
Lacinipolia renigera (Steph.)
Lacinipalia vicina (Grt.)
Leucania commoides Gn.
Leucania pseudargyria Gn.
Leuconycta dipteroides (Gn.)
Lithacodia albidula (Gn.)
Lithacodia cameola (Gn.)
Lithacodia concinnimacula (Gn.)
Lithacodia muscosula (Gn.)
Lithacodia synochitis (G. & R.)
Lithophane hemina Grt.
Lithophane innominata (Sm)
Lithophane patefacta (W1k.)
Lithaphane petulca Grt.
Lithophane semiusta Grt.
Lithophane spp.

Marathyssa inﬁcita (Wlk.)
Meganola minuscula (Zell)
Melanchra adjuncta (Gn.)
Melipotus spp.

Morrisonia conﬁasa (an.)
Morrisonia evicta (Grt.)
Nedra ramosula (Gn.)
Nephelodes minians Gn.
Nigetia formosalis Wlk.

162

I—awI

H I

t—I I

u—a H |

Nt—tl

II—It—OHu—ol

II-5 I

thNu—ou—I—II

I I—ap—IIAI—AII

NNI

Table A1. (cont’d.)

Noctuidae I

Noctuidae 2

Noctuidae 3

Noctuinae spp.

Nolinae spp.

Ochropleura plecta (I...)
Oligia exhausta (Sm.)
Orthodes crenulata (Butler)
Orthodes cynica Gn.
Orthodes spp.

Orthosia hibisci (Gn.)
Orthosia revicta (Morn)
Orthosia rubescens (Wlk.)
Palthis angulalis (an.)
Peridroma saucia (I-Ibn.)
Phlogophora iris Gn.
Phlogophora perculosa Gn.
Phoberia atomaris an.
Plathypena scabra (F .)
Polia delecta B. & McD.
Polia detracta (Wlk.)

Polia latex (Gn.)

Polia nimbosa (Gn.)
Protorthodes oviduca (Gn.)
Proxenus miranda (Grt.)
Psaphida resumens Wlk.
Pseudorthodes vecors (Gn.)
Pyreferra ceromatica (GrL)
Raphia frater Grt.
Rhynchagrotis adulta (Gn.)
Rhynchagrotis anchoceloides (Gn.)
Rhynchagrotis cupida (Grt.)
Rivulinae spp.

Schim'a arcigera (Gn.)
Spaelotis clandestina (Ham)
Spodopterafrugiperda (J . E. Smith)
Sunira bicolorago (Gn.)
Synedoida grandirena (Haw)
Tarachidia erastrioides (Gn.)
Ulolonche culea (Gn.)
Ulolonche modesta (Morn)

I—III—II

wI—SI

163

INNII

h-‘ I

I— h-i I H I

—| I

NI-‘D—OII—ll

NIIr—lNIu—AI

ﬂ

I—twlbl

Iu—pINI

N

Table A1. (cont'd.)

Xestia adela Franc.

Xestia dolosa Franc.

Xestia normaniana (Grt.)
Xestia smithii (Snell)

Xestia tenuicula (Morn)

Xestia spp.

Zale duplicata (Bethune)

Zale helata (Sm)

Zale lunifera (an.)

Zale minerea (Gn.)

Zale submediana Strand

Zale undularis (Drury)

Zale unilineata (Grt.)
Zanclognatha jacchusalis (Wlk)
Zanclognatha laevigata (Grt.)
Zanclognatha ochreipennis (Grt.)
Zanclognatha pedipilalis (Gn.)
Unidentiﬁed species

164

I—aNII

tow—I

At—Ib—II

 

APPENDIX B

Appendix B

Table Bl. Total number of Tortricid species collected from four red oak
sites and four northern hardwood sites over three years.

   

_ rt em
Hardwood

W 1993 1%? 19%

Acleris I - - 1 - - -
Acleris 2 - - 1 - - -
Aclerisforbesana (McD.) - - - 2 - -
Acleris oxycoccana (Pack) - - - 3 - -
Acleris spp. l7 - - 163 - -
Ancylis burgessiana (Zell) 1 1 - - l l
Ancylisfuscociliana (Clem.) - 1
Ancylis metamelana (Wlk.) 1 -
Ancylis nubeculana (Clem.) - 1 - -
Archips argyrospila (Wlk.) 44 3 4 2
Archips cerasivorana (Fitch) - - 1 l
Archipsfervidana (Clem.) 37 - - 1 -

5

4

 

‘ Archips purpurana (Clem.) 6 - -
Archips rosana (L.) 69 - -
Archips semiferana (Wlk.) - l 2

Argyrotaenia alisellana (Rob.) - - l - - -
Argyrotaenia mariana (Fem) - - l -
Argyrotaenia quercifoliana (Fitch) 4O 2 4 1
Argyrotaenia spp. - — - 5 -
Choristoneura conﬂictana (Wlk.) -
Choﬁstoneuraﬁactivittana (Clem.) 2 2 - - -
Choristoneura fumiferana (Clem.) - - - 5 - -
Choristoneura obsoletana MIR.) 7
Choristoneura pinus Free. -
Choristoneura rosaceana (Ham) 22
Choristoneura spp. 6 - - - -
Clepsis melaleucana Mk.) -
Clepsis persicana (Fitch) -
Croesia albicomana (Clem.) 12
Croesia semipurpurana (Kft.) -
Ecdytolopha punctidiscana (Dyar) -
Epinota timidella (Clem.) -
Eucosma dorsisignatana (Clem.) - - 1 - - -
Melissopus lanferreanus Wlsm. 1 - 1 2 - 1

AM!
I
I

—n
I
U.)
\O
—o
D.)
NJ)!

H # H |
I
I
g
I

165

Table Bl. (cont’d.)

Olethreutes 1

Olethreutes 2

Olethreutes 3

Olethreutes 4
Olethreutesfasciatana (Clem.)
Olethreutes nigrana (Heinr.)
Olethreutes pennundana (Clem.)
Olethreutes spp.

Olethreutinae spp.
Olethreutresfootiana (Fem)
Olethruete appendicea (Zell)
Pandemis canadana Kft.
Pandemis limitata (Rob.)
Petrova gemistrigulana (KfL)
Phaneta raracana (KfL)
Platynotus idaeusalis (Wlk.)
Proteateras aesculana (Riley)
Proteateras moﬁatiana Fern.
Pseudexentera costomaculana (Clem.)
Pseudosciaphila duplex (Wlsm.)
Pytcholoma peritana (Clem.)
Sparganothis diluticostana (Wlsm.)
Sparganothis directana (Wlk.)
Sparganothis'niveana (Wlsm.)
Sparganothis pettitana (Rob.)
Sparganothis reticulatana (Clem.)
Sparganothis sulﬁireana (Clem.)
Sparganothis tristriata Kft.
Sparganothis xanthoides (Wlk.)
Sparganothis spp.

Spargonothis unifasciana (Clem.)
Syndemis aﬁlictan‘a (W 1k.)
Tom'icinae

Unidentiﬁed species

166

(3....

”wt

h—iNl

Wh—‘Nh

 

APPENDIX C

Appendix C

Table Cl. Total number of Geometrid species collected from four red oak
sites and four northern hardwood sites over three years.

 

N ort . ern
Hardwood

   

 

 

1993 1994 1995 1993 1994 1995

 

Anavitrinelia pampinaria (Gn.) - - -1 7 -
Besma endropiaria (G. & R.) - - 4 -
Besma quercivoraria (Gn.) l
Biston betularia (L.)
Cabera variolaria Gn.
Cabera spp.

Campaea perlata (Gn.)
Caripeta divisata Wlk. - - 1 - -
Cladara anguilineata (G. & R.) - - - - -
Cladara atroliturata (Wllc) - - - - -
Cladara limitaria (Wllc) - - - - -
Cleora projecta (Wlk.) -
Ectropis crepusculan'a (D. & S.) 1 - - - -
Ennominae spp. 1 -
Ennomos subsignaria (an.) - 1 - 5 -
Eubaphe mendica (Wlk.) - - - 1
Euchlaena johnsonaria (Fitch) - - - -
Euchlaena tigrinaria (Gn.) - -
Eugonobapta nivosaria (Gn.) -
Eupithecia spp. 3
Geometridae I -
Geometridae 2 -
Geometridae 3 -
Geometridae 4 -
Geometridae 5 -
Geometridae 6 -
Geometridae 7 -
Geometridae 8 -
Heliomata cycladata G. & R. -
Hydria prunivorata (an.) - - - -
Hydria spp. - - - -
Hydromena spp. 1 18
[tame argillacearia (Pack) - - 1 - -

I-d I
I
”

MHI—tI
I
I
I
I

HNNNIMI

i—‘t—OHIINNII

II IIUJUJI
I—HI—II—II—ININNAIu—AI

hi ~ g I
I
I

N

U.)

I

Nv—st I
IIIwu—IIIIIu—nI—nu—IAII—IIIII—INII—IIu—AIINI

167

Table C1. (cont'd.)

Itame coortaria (Hulst)

[tame evagairia (Hulst)

Itame pustularia (Gn.)
Lobophora montanata Pack.
Lomographa glomeraria (Grt.)
Lomographa semiclarata (Wlk.)
Lomographa vestaliata (Gn.)
Melanolophia canadaria (Gn.)
Metanema inatomaria Gn.
Metarranthis refractaria (Gn.)
Nematocampa limbata (Haw)
Nepytia canosaria (Wlk.)
Orthonama centrostrigaria (W 011.)
Orthonama obstipata (F .)
Pero hubneraria (Gn.)
Plagodis alcoolaria (Gn.)
Plagodisfervidaria (H. -S.)
Plagodis phlogosaria (Gn.)
Plagodis serinaria H. -S.
Plagodis spp.

Probole alienaria H. -S.
Proﬁtame virginalis (Hulst)
Protoboarmia porcelaria (Gn.)
Scopula inductata (Gn.)
Scopula limboundata (Haw)
Semiothisa bisignata (Wlk.)
Semiothisa ocellinata (Gn.)
Synchlora aerata (F .)

Tetracis cachexiata Gn.
Tetracis crocallata Gn.
Xanthorhoeferrugata (CL)
Unidentiﬁed species

168

NI—‘I

I—ﬁ I

Nt—oINILpI—AI

ﬂ

NII NI—tlu—at

N

H H |

N I

H H I

INNII—I hINIu—bImIINr—IAI

H

 

APPENDIX D

Appendix D

Table D1. Total number of Pyralid species collected from four red oak
sites and four northern hardwood sites over three years.

   

_, Norther
Hardwood

 

 

1993 1994 1995 1993 1994 1995

 

u—s
I

Acentria nivea (Olivier)
Acrobasis spp.

Aethiophysa lentzﬂualis (Zell.)
Aglossa cuprina Zell.
Blephanomastix ranalis (Gn.)
Chrysoteuchia topiaria (Zell.)
Crambinae I

Crambinae 2 -
Crambinae 3 -
Crambinae spp. -
Crambus agitatellus Clem. -
Crambus albellus Clem. 1
Crambus spp. -
Eulogia ochrifrontella (Zell.) 14
F umibotys fumalis (Gn.) -
Herculia olinalis (Gn.) 1
Herpetogramma pertextalis (Led.) -
Microcrambus elegans (Clem.) 1 -
Munroessa gyralis (Hulst) - -
Munroessa icciusalis (Wlk.) 1 -
Nephopterix vetustella (Dyar) - 1 - - 2
Nephopterix virgatella (Clem.) - - - - -
Nephopterix spp. - -
Nymphalinae spp. - 1
Oneida lunulalis (Hulst) - -
Ostrinia nubilalis (l-lbn.) - -
Palpita arsaltealis (Wlk.) - - - - - -
Parapoynx maculalis (Clem.) - -
Petrophiliaﬁdicalis (Clem.) - - - -
Phycitinae spp. 11 - - 2 1
Platytes vobisne Dyar - - - 4
Pyla fusca (Haw.) - - 2 - -
Pyralidae I - - - 1 -

- 2 - - 2

\Or—AI
AI
DJ
~
I
I

NNI
I
I
I
I
I

NIIIu—at—tu—II
I—Iu—AmII
t—‘I I I
”I I I
I—au—IINUJII

t—ﬁI—‘I
N
I
I—iNI
I

h -—o I I L» N I I
I I I I
I v—a I I
u—a U.) I N b) ._. I

N
I
I

I
H H ~ ~ H I

169

 

Table D1. (cont'd.)

Pyralidae 2

Pyralidae 3

Pyralidae 4

Pyralidae 5

Pyralidae 6

Pyralidae 7

Pyrausta acrionalis (Wlk.)
Pyrausta signatalis (Wlk.)
Pyrausta spp.
Pyraustinae I

Pyraustinae 2

Salebriaria twpidella (Rag)
Scoparia basalis Wlk.
Scoparia biplagialis Wlk.
Scoparia spp.
Scoparia/Eudonia spp.
Scoparinae spp.
Sitochroa chortalis (Grt.)
Synclita tinealis Mun.

Thaumatopsis gibsonella Kft.

Udea rubigalis (Gn.)
Urola nivalis (Drury)

Unidentiﬁed species

170

 

APPENDIX E

Appendix E

Table E1. List of species collected from families other than Noctuidae,
Tortricidae, Geometridae, and Pyralidae.

 

Agonoxenidae Glyphipteryx linneella (Bsk.)
Arctiidae Apantesis parthenice (W. Kirby)
Arctiidae Clemensia albata Pack.

Arctiidae Crambidia pallida Pack.

Arctiidae Cycnia oregonensis (Stretch)
Arctiidae Dryocampa rubicunda (F.)
Arctiidae Halysidota tesselaria (J. E. Smith)
Arctiidae Haploa reversa (Stretch)

Arctiidae Holomelina aurantiaca (I-Ibn.)
Arctiidae Hypoprepia ﬁicosa an.
Arctiidae Phragmatobia assimilans Wlk.
Arctiidae Spilosoma congrua Wlk.
Arctiidae spp.

Argyresthiidae Argyresthia oreasella Clem.
Argyresthiidae spp.

Blastobasidae Gerdana can'tella Bsk.
Blastobasidae spp.

Cochylidae spp.

Coleophoridae Batrachedrinae spp.
Coleophoridae spp.

Cosmopterigidae Euclemensia bassettella (Clem.)
Cosmopterigidae spp.

Cossidae Acossus centerensis (Lint)
Decophoridae spp.

Drepanidae Drepana arcuata Wllc
Eriocraniidae Mnemonica auricyanea (Wlsm.)
Eriocraniidae spp.

Gelechiidae Aristotelia spp.

Gelechiidae Chionodes spp.

Gelechiidae Coleotechnites spp.
Gelechiidae Dichomeris ligulella an.

171

 

172

Table El. (cont’d.)

Gelechiidae Dichomeris spp.

Gelechiidae spp.

Gelechiidae Trichotaphe setosilla Clem.
Gelechiidae Walshia spp.

Gelechioid spp.

Glyphipterigidae Abrenthia cuprea Bsk.
Gracillariidae Caloptilia spp.

Gracillariidae spp.

Hesperiidae Erynnis brizo (de. & LeConte)
Hesperiidae Poanes habomok (Ham)
Hydroptilidae spp.

Lasiocampidae Malacosoma americanum F.
Lasiocampidae Malacosoma disstria an.
Lasiocampidae Phyllodesma americana (Ham)
Lasiocampidae spp.

Limacodidae Apoda biguttata (Pack)
Limacodidae Lithacodesfasciola (H. -S.)
Limacodidae spp.

Limacodidae Tortricidia ﬂexuosa (Grt.)
Limacodidae Tortricidia pallida (H. -S.)
Limacododae Lithacodes testacea Pack.
Lymantriidae Lymantria dispar (L)
Lymantriidae Orygia leucostigma (J. E. Smith)
Lymantriidae Orygia

Lymantriidae spp.

Lyonetiidae Bucculatrix

Lyonetiidae spp.

N epticulidae spp.

Notodontidae Dasylopha thytiroides (Wlk.)
Notodontidae Ellidia caniplaga (Wlk.)
Notodontidae F urcula modesta (Hudson)
Notodontidae Gluphisia septentrionalis Wlk.
N otodontidae Heterocampa guttivita (Wlk.)
Oecophoridae Polex coloradella (Wlsm.)

 

173

Table El (cont'd.)

Oecophoridae Psilocorsis spp.
Oecophoridae Semiosopis inornata Wlsm.
Oecophoridae spp.

Olenthreutinae spp.

Pieridae Phoebis sennae (L.)

Plutellidae Plutella xylostella (L.)
Psychidae spp.

Satyridae Megisto cymela (Cram)

Sesiidae Syanthedon acemi (Clem.)
Sphingidae Ceratomia undulosa (Wlk.)
Sphingidae Pachysphinx modesta (Ham)
Sphingidae Paonias excaecatus (J. E. Smith)
Sphingidae Paonis myops (J. E. Smith)
Sphingidae Sphecodina abbom'i (Swainson)
Thyatiridae Euthyatira pudens (Gn.)
Tineidae Acrolophus spp.

Tineidae spp.

Tinicoid spp.

Yponomeutidae Atteva punctella (Cram)
Yponomeutidae Yponomeuta multipunctella Clem.
Yponomeutidae Zelleria haimbachi Bsk.
Zygaenidae spp.

 

APPENDIX F

APPENDIX F

Record of Deposition of Voucher Specimens*

The specimens listed on the following sheet(s) have been deposited in
the named museum(s) as samples of those species or other taxa'which were
used in this research. Voucher.recognition labels bearing the Voucher
No. have been attached or included in fluid-preserved specimens.

Voucher No. : 1996-6
Title of thesis or dissertation (or other research projects):
THE IMPACTS OF GYPSY MOTH (LEPIDOPTERA:LYMANTRIIDAE)

ON NATIVE ARTHROPOD ABUNDANCE, SPECIES RICMNESS, AND
DIVERSITY IN TWO HARDWOOD ECOSYSTEMS IN NORTHERN LOWER

MICHIGAN
Museum(s) where deposited and abbreviations for table on following sheets:

Entomology Museum, Michigan State University (“50)

Other mseums:

Investigator's ﬂame (s) (typed)
Timothy Thergn ﬂgrk

 

 

 

Date 8/1196

*Reference: Yoshimoto, C. N; 1978. Voucher Specimens for Entomology in
North America. Bull. Entomol. Soc. Amer. 24:141-62.

Deposit as follows:

Original: Include as Appendix 1 in ribbon copy of thesis or
dissertation.

Copies: Included as Appendix 1 in copies of thesis or dissertation.
Mbseum(s) files.
Research project files.

This form is available from and the Voucher No. is assigned by the Curator,
Michigan State University Entomology Museum.

174

 

a at Ave.) 33

 

 

 

 

 

 

175

AJWHMWDEKF'

Voucher Specimen Data

Page __J_ of __1_1 Pages

baggage.» 33m sonar—ca: 0.5 ca uauonov sung...
now ucosaooao vuuoaa o>onn «no wobaouux
IIIIIIIImnAﬁwma..oz nososo> Anonhuv Aavoamz u.uousaauoo>ca
Ahumomuuos ma ouoosm aacoauavvm many
mama a2< ma . mcmunom Eacoa<
. vmma 2:3 aw
mama 223 cm camauzzu am amazoa maamatm:
«mma 2:3 am vmma.z:ﬁ am umump anammoao
mam.“ 2:1 ma mnﬂwuuom manmhmo
mama a2< Nu mamo>a>m manmnou
vama mam ma upcmzaoo>m meomoaoo
«mad 2:3 ma «mma 2:1 ma encamaum oeomoamo
mama 2:1 ma

mama ><z ma mama ><z ma
vmma ><z am vmma ><z om
mama m=< mm vmma ><= om
vmma ><z om mama ><z ma
mama m:< Nu mama 2:1 ma
wmma #1: aa mmma 2:3 mm
mmma m=< Nu mama ><z 5N
mmma m:< aa mama ><z ow
vmma ><z aa vmma ><z om
vama ><z aa mama m2< Nu

 

 

 

 

 

 

 

 

 

cmma 2:3 ma emma.a=< ma manageaa manmnmo
Jun? wouaaonov was can: no vouuoaaou soxou assoc no uuauunm
m e r r m m e .m u ocosauoao you muav aonma
.amamaawm:
.._t d Au u .a u
axially“...

 

 

.0
“H

nvuﬂﬁﬁz

 

 

176

AJHHHWDEKF‘

Voucher Specimen Data

Page __2__ of __1_1 Pages

nun

noumuso

 

. ,\ ﬁg

Dana—03.5 33m smaasoaz «5 ca uauoaov
now usuaauoao voueaa o>opm ozu vo>aouux

j 38

 

 

 

snagged...

 

 

 

 

 

 

 

 

 

 

 

 

nae—52

 

 

llaulmlmwﬂséz .3.—25> Goa—5v Revue—oz Pneuamaumgca
. Amumuouoos «a muoozu ascoauavvm mama
mama a=< mm
vama a=< ma mama mam am mamoaam mmpocazamum
mama m:< mm aopcooma
vaaa ><z ow vmma,><z am macmumocoum waamuoammcam
aama 4:3 am amma mam Na maaamamaa macamaaa
eaaa man ma maoaaa mascaeaa
vama 4:: am ocooanmso maccaeao
vaaa 22m am aaaaooamaao maucaEaa
mama a=< aa
emaa 2:3 am mama 4:3 am
mama a=< ma mama 2:: am amummEaa maucaeaa
mama a=< mm
mama ><z mu vaaa 2:1 aw
mmma 2:: aa mama mam aa aaaaooaamaa maucaeaa
J a... vouauomov was can: no vouuoaaou noxuu uoﬁo uo «vacuum
me... r u m e .m a 256.30% you 3mm amend
r 0.0 e «m .M a W 8
ma pm a a m. m . a
u wk... 0 A M P .L 8
NO

 

at: N 1. 38

 

 

 

 

unanno>aea uuuua caaasomz «an :a uaoonov

 

 

177

AJWWHWDDKF‘

of 11 Pa

3

voucher Specimen Data

Page

now asoaauoao vouuaa £2? of .5333 xnoz .._. xnposah
mmmda .oz 3:33 9.25: 3.232 Bnouaaanauéa
Ahnoaoooos ma mucosa amcoauavvm om:v
mama m:< aa _
vaaa 4:: ma mama a:< ma noaooanp maacmmacu
mama m:< Nu
vaaa m:< ma mmaa m:< aa maca>noo macoaumonmua
vmma 2:: ma mea 4:: ma mavae manuaumoumua

mama a:< mm namesannooa macoaumonopa
mama 4:: aa mama 4:: aa azanmcwame manuauaonmua

Qaaa ><2 ow
vaaa ><2 aa mama 4:: am
vmaa ><2 ow vama ><2 ow
vama ><z aa vama 2:: am
vaaa 4:: ma vmma ><2 ow
vama ><2 om vaaa ><2 aa

 

where
depos-
ited

Museum

 

 

 

vama 2:: am maoaccm>a>mccma m: oapmonmn
«maa 4:: «a c . a
vmma 2:: ma mama m:< ma mmuamawco macoapmonmpa
Jan? vuuaooaov man man: no monuoaaou canon nonuo no uuauoaa
m m e .m u ocosauoau now and: aonsa
mmmwaw
“up u...

Other

 

 

 

 

 

 

 

ﬂuno£E;_

 

 

178

AJWHHWDEKF‘

voucher Specimen Data

Page J— of _1_1_ Pages

 

huauno>aea sauna amaasoaz any ca nanoao:
now usoanoao vouuaa u>onu any mo>aouom

 

 

¢$M~ dﬁ)‘ mama

 

 

 

 

m-maaa .oz :53» can:

and: aauunH anuaaah

onusaz u.n0nuaanuo>:a

Ahnuonuuoa ma ouoogu amaoauamvm om:v

 

mama m:< ma mama 4:: Na
mama a:< ma mama 2:: am
mama 4:: Na

mama 4:: ca
mama 4:: ma mmaa m:< ma

mama 2:: mm mama 2:: mm
mama 2:: ma mama 2:: ma

mama 2:: mm mmaa 2:: mm

mama mam ma mama mam aa
mama zam ma mama zam an

mama 2:: ma mama 2:: ma

mama 4:: mm mmma.m:< ma
mama 4:: ma mama 4:: ma

eacooanoea coauangaa
maamnpnoa coauanema

noaooawno> coauanemm
mapaaoa maanacomxa

anoncoau maczoapocoana

meson maaozaocmpa

maaoocaa mnaozaocmua

mmuaoomau mauamumc<

monopocaaea manoacaa

 

depos-
ited

Museum
where

 

Other

 

Amdxs¢f

 

Adults 9

 

 

 

Pupae
Nymphs

Larvae
Eggs

 

vouaoonov and man: no vouooaaou
unusauoau now «no: ampua

 

soxun nonuo no ooaoonm

 

 

O

nomesz

 

 

 

 

um? 38
P 3mm . Jr: .0

wow. .N. 13‘

 

 

 

 

Daonopasp 33a cmaacoaz 0.5.... ca namoao: $340353.
now usoanoau mouuaa o>opu «no vubaouom
mldmaaoz non—26> Goon—buy Amvusmz u.noumaaumo>ca

Ahnuamoooc ma unoono aucoauavmn um:v

 

179
11 Pages

.APPENEHX]?

of

5

Voucher Specimen Data

Page

 

 

 

 

 

 

 

 

 

mama 2:: am
mama 2:: am mama 4:: am
mama 4:: cm mama 4:: ma mcoommcm>o an»:
mama 4:: ma
mama mam aa amaa 4am ma aacauanoo aaaz
mama 2:: mm maaaooaunoo macazomna
mama ><2 mm mama 2:: ma noaooanu manm4
mama 2:: ma mumcno manm4
mama a:< ma mauana> manm4
mama a:< ma manuco>anum manm4
mama a:< ma mama a:< ma mxm>coo mnms<
mama 2:: ma mama 2:: mm
mama 2:: Nu mama 2:: cm Eauaooaa encoa<
mama 2:: Nu sanooou socoa<
1.0m couaoonom new mom: no vonooaaoo eoxou nozuo no aoauoaa
mm...“ d a m m u .m u s 2553: now 3m: ago.“
mama a m a me a
mu 4. o_.A .A .r . a.

 

 

.0
“4

o non—292

 

 

 

mums nonunoo mam; A. 6 38

 

 

:3“ alum. .Smsam ,
£502.: 053

 

 

Cannabis 33a coma—moat «.5 ca nauonuv Jada;
now nsoaauoau moumaa o>opm on» mo>aooom
mummmH .oz nozoso> Goa—Ana A3882 o.noumaaumo>ca

Amnsmoouoa ua muuocm amcoauavvu omav

 

180
ges

APPENDIX F

of 11 Pa

6

Voucher Specimen Data

Page

mama 4:: ma mama 4:: ma

 

 

 

 

 

 

 

 

 

mama a:< ma mama a:< ma aoeaxonq mason»
mama a:< ma noamon aca>oao
mama zam ma aaaancau aca>maa
J 0.. vouaooaov mom can: no vouooaaoo =88» noﬁo no uuaooaa
e r r w m e .m 0. 23.503» now «no: ammo.“
“aaaamMWa
.w .m i O A A P N M E

 

 

 

 

 

 

- u m
nub. neumnao QVN: N W n a

 

 

 

 

 

 

181

.APPENEHX]?
Page 1 of 1; Pages

Voucher Specimen Data

 

 

 

 

 

 

 

 

 

 

 

\a
3a a V /
.Esom: oa ucu
Daungaca 33a cmaazuaz 05 ca uaooaov Jug... acne—ca...
now unusauono vouoaa o>onm ecu vo>aoooa .
llllllmuuludad .oz non—25> Goa—hay Anya—32 u.noumaanoo>ea
. “unannouoc ua onooca amcoauammo «may
mama ><z Ma manocae mamN
mama 4:: aa maamnconmancoo oaaoaom
mama 4:: aa onaaoca onoacono<
mama a:< ma oaaa oaoooumo
mama ><z ma anaoaaaoo mama
mama ><z ma wanano>aomac osmma
mama 4:: aa oaampocoapcomno meaaamo
mama ><z mm manocmaaw maonona
maaa ><z mm onoa>mn oaaonnna
mama a:< mm onoaaoacoa aapaoa
mama a:< mm oaonoaooan anacom
mama 4:: ma manuanao manooonua4
mama 4:: aa aaaoanac macanuma
mama 2:: am mamcmcoamno oacu>o
mama 23m an aaaoannaaaaaa xacoaaca
mama 2:: am noumnm omcpcoa
mama 42m aa aaaacoanacmoaaa aaaamaaaa
mama 4:: ma mmcaaapoonm mouocma>z
mama 2:: am moaaouonaaa massano
mama ><z mm moaeaocuzao o>oaaaa
.mmaa a:< mm ouooaaasa oaaoaocaoo4
mama ><z mm moanuoa mcocaozua4
J 6.. monauonu: mam moms no vouooaaoo aoxmu nosuo no uoaooaa
mm wd u a a u m u unusaooao now «no: ammo.—
m . pa a m a. a. m a m...
u m a 1 o A m r u a
"mo nun—5:2

 

 

 

 

00¢“ ROUGH:

 

2.: WEN :3

 

£50m: hmoa o u

 

 

Daunonaas: 33a cmaasoaz «in ca namoaom
now ososaooao monuaa u>opo osu mobaouom

xnoz .H >5002ah

mummma .oz noses; €099; 3882 a.nonmaauoo>:a
. Annunuouoc ma mucosa ausoauamvo many

 

ges

182

APPENDIX F

of 11 Pa

8

Voucher Specimen Data

Page

 

 

 

 

 

 

 

 

 

 

 

mama ad: «a !Nddddlddmdmﬂd
mama 2:: am muoaxocuoo maoannmn
mama ><z um onoocaaac: oaoN
mama 2:: cm mao5>o macoananm
mama 4:: aa oconmooapnaa: maznocwanmoa
mama 4:: aa oconoaaoauon macnocoanoaa
mama 4:: aa omoaauca masononoo
mama 4:: ma manoooooxo aoacoma
mama mam ma acasacancm aaaoonaa
mama a:< ma aaaonocaam unmounxn
mama a:< ma maaaeamocon maxoaoom
mama a:< ma annooa ocomapoaa
mama a:< ma ooacmznnoa mammncoa<
maaa mam ma maaacoamaa aaaaonaeaazn
mama a:< ma omoa32on annoz
mama a:< ma wuama>amonomanwo
mama a:< ma onoaaono oacacoa
mama a:< ma muaaaoa mauansmno
mama 4:: aa ancanam aauonaaoanaam
mama 4:: aa mama 4:: aa maaomoenom manoaaz
mama 423 aa aaaaaaoaaam macanaaaaacm
mama 4:: aa manxanwuaomn macoono4
J or vouauoaov man man: no monuoaaoo aoxmu noﬁo no uuauoma
m e r r m m e .m u ocusaooao now «no: aomma
eaWam.uu w as
”bk... t d d u w a a
M w i O A A P .L E
«mo nae—52

 

 

 

ﬂouma noumnﬁb

a
at: 6...? 38

 

 

 

 

3.2%de KAN I
.anm mac 060 u

 

 

 

183

AJWHHNDDKF‘

Voucher Specimen Data

Page 3__ of _11__ Pages

 

 

where
depos-
ited

Museum

 

 

 

 

 

 

Other

 

 

 

Dannaaaaca 33a nmaacoaz «.5 ca uaooao: and;
now oaoaaooao mouoaa o>omo can mo>auoo¢ .
mWwIaImJa .oz nonoao> Goa—hay Anvuamz u.n0umaauoo>=a
Ahnmomoooa ma mnuozn amsoauavvu away
mama ><z ma acoeaoon.oua;aoma
mama 4:: aa ocmano> monanoaa
mama 4:: aa onocnonao manonman<
mama 4:: ma onoaoaamn ooxom
mama 4:: ma mama: onoacono<
mama ><z mm mcmaaeammo manonoemmncn
mama 4:: aa mcoaoaomac: masnocoanoaa
mama 4:: aa aaaocaao oaaoonm:
mama 4:: aa mama 4:: aa oeooaaao acaoanpnon

mama 4:: aa upmcaaaamnou osmooam
mama 4:: aa maanm mama
mama 2:: ma noeono: aocooa
mama a:< ma cannon mnoacocam
maaa a=< ma aaamaaaama anocanmaa
mama ><z ma onann maccanm
mama ><z ma aannoomm acaoooocam
mama 2:: ma aooao>anaon aanzamam
mama mam «a maaanmmna> manapaozamz
mama a=< an apacnmnaa aanonmmmm
mama 42m ma aaaaaamaa nanosecoaa
mama 2:: am anono>ac=na want»:

mace monaoono: and mom: no monuoaaoo aoxuu nosuo no uoauoaa

u m e .m u oaosauoao now and: among

mamaaa

A A r .u z

”ﬂungﬁaz

 

 

 

.xrﬁiouum nonmnso

want an mam) . QR

.anmoz aaoa cm
unannobaaa onmua caaasoaz «so ca namoaov

 

a,
abb~ .n mo‘ 33

 

 

 

 

184
Pages

APPENDIX F

of 11

10

voucher Specimen Data

Page

.23: can;
now osoaaooam vouuaa u>opm «nu mo>aooum
, mumaaa .oz 3:052» Goa—h; A3832 u.noumaauoo>ca
Ahnmomoooa ma mucosa aucoauammm «may
mama ><z um onanonaaonna onoooao
mama ><z am aouaocoaax man»:
mama 4:: am onoaa>ooa ocpocaoaocoN
mama mam aa mama mam aa manaaaoaanm aaamunm
mama 2:: ca aanazoocaa aaoooagna4
mama 2:: am oaoaou aanmmx
mama 4:: aa mama 4:: aa aanpmmau amoeooaamz
mama ><z mm ocoanoe aacoanonxan<

mmaa 4:: aa mama 2:: am
mama 2:: am

mama 2:: am

mama z:: ow

mama z:: am

mmaa 2:: mm

mama a:< um

mama.><z ma mama ><z mm
mama ><z mm
mmaa 2:: am
mama 2:: am mama 4:: aa
mama 2:: on

anamnoo mannaaaoaamo
mummwao>u apneoaao:
oaoocnoo mauooacua4
acmnamcana anaoumc>a
acooano: onuanozo
mcooamnma mamamau
maammamno macnamn>an<
onocanaa: canaco<
oanmmoaoaca mauoaoaa
manaaoooao aauoaoaa
acaooomon annoconmanozu
maoaxoam mauaoannnoa

 

 

 

 

 

 

 

 

Museum
where
depos-
ited
Other
Aanxscf
Adams 9
Pupae
Larvae
Eggs

mouaooaom man man: no monuoaaoo

oaosauoao now «no: ampua

 

canon nonuo no uuaooam

 

“mo nomssz

 

 

 

imam a...

 

 

 

 

 

 

185

APPENDIX F

of 11. Pa

Voucher Specimen Data

Page 11

 

 

depos-
ited
Other

where

Museum

 

 

 

 

 

 

 

 

 

wnmm.. , » nanmnao
mama H {ax $3
.Eooa oao nan .
Saunas—n: 33m coma—anar— uzu ca namono: snail—maﬁa...
now ocosaooao monoaa o>oam «so mu>auoum
mammma .oz nososo> €09.95 A3252 o.nouoaanoo>ca
Ahnmonouo: ma mucosa aucoanamma mama
maaa azm aa uaaomEac aaaoa
mama 4:: aa uconomaeoo anacon<
mama 4:: aa ocanasaa maeomcaa
mama 4:: Na asoaau oasaoo
mama a:< a oamoaoaco oeonuanoa
mmaa 4:: Na aaoooa aeonmanad
mama a:< ma anuoaa annoaaonzuo
mama 2:: Na oaaomocae oaocomoz
maaa mam ma oaamnocaa a>ann<
mama a:< ma mamavoocooaa mnqnoeaoa
mama zam ma anaamoe xcacaaacoaa
mama 2:: Na aocaoanan anaoo>no
mama >mz ma acaoanoEm asaamoaaama
mama ><z Na aaaanoganaon ozooaosoa
mama ><z aa anaeaanaann aaaaaaam
mama ><z mm acoenoc acaonaaoonu
mama ><= mm oaonosoaa ocaonaoso4
mama m:< mm aaoaooanoa anozaoaoaca
mama 4:: aa mama 4:: ma acmnanaa aacpocamnaaa
mmaam4=m ma mama 4mm aa mucosa aaamnaoaa:
. mmaa 2:: am mqoxe awacoaa
mama 2:: am acaoaoaoaoe aaaomau
mumm monaoomom and man: no mouuoaaoo coxou nunuo no moauoaa
m u e .m u unusaooao now and: among
mmmama
.A A v. .n 8
«mo nupeaz

 

 

LIST OF REFERENCES

Addison, J .A. 1993. Persistence and nontarget effects of Bacillus thuringensis in soil: a
review. Canadian Journal of Forest Research 23:2329-2342.

Allen, RT. 1976. The occurence and importance of ground beetles in agricultural and
surrounding habitats. In: Carabid Beetles: Their Evolution, Natural History, and
Classiﬁcation. Proceedings. Smithsonian Institution, Washington D. C.

Anonymous. 1991. Task Force Report on Biological Diversity in Forest Ecosystems:
Society of American Foresters.

Anonymous. 1995. Gypsy Moth Management in the United States: a cooperative approach
Final Environmental Impact Statement: USDA, USPS, and APHIS.

Amett, R.H., N .M. Downie, and HE. Jaques. 1980. How to Know the Beetles. Second
ed. Dubuque, IA: Wm. C. Brown Company.

Attiwill, P.M. 1994. The disturbance of forest ecosystems: The ecological basis for
conservative management. Forest Ecology and Management 63:247-300.

Baker, W.L. 1941. Effect of gypsy moth defoliation on certain forest trees. Joumal of
Forestry 39: 1017-1022.

Binkley, D. 1986. Forest Nutrition Management. New York: John Wiley and Sons.

Blaldey, N.R., and H. Dingle. 1978. Competition: Butterﬂies eliminate milkweed bugs
from a Caribbean island. Oecologia 37:133-136.

Bolstad, P.V., and ST. Gower. 1990. Estimation of leaf area index in fourteen southern
Wisconsin forests stands using a portable radiometer. Tree Physiology 7:115-124.

Borror, D.J., C.A. Triplehom, and N.F. Johnson. 1989. An Introduction to the Study of
Insects. Sixth Edition ed. Fort Worth: Saunders College Publishing.

Bowden, J. 1982. An analysis of factors affecting catches of insects in light-traps. Bulletin
of Entomological Research 72:535-556.

186

 

187

Brown, V.C. 1984. Secondary succession: insect-plant relationships. BioScience 34:710-
716

Burton, A.J., K.S. Pregitzer, and DD. Reed. 1991. Leaf area and foliar biomass
relationships in northern hardwood forests located along an 800 km acid deposition
gradient. Forest Science 37:1041-1059.

Butalla, H.A. The GMDigest. <http://www.fsl.wvnet.edu/fhp/gm_digest/gmdigest.html>
May 22, 1996. (Accessed June 12, 1996).

Cameron, E.A., and R.M. Reeves. 1990. Carabidae (Coleoptera) associated with gypsy
moth, Lymantria dispar (L.)(LepidoptermLymantriidae), populations subjected to
Bacillus thuringiensis Berliner treatments in Pennsylvania. Canadian Entomologist
122:123-129.

Campbell, R.A., and R.J. Sloan. 1977. Forest stand responses to defoliation by the gypsy
moth. Forest Science Monograph 19:1-35.

Cleland, D.T., J.B. Hart, G.E. Host, K.S. Pregitzer, and CW. Ramm. 1993. Ecological
classiﬁcation and inventory system of the Huron-Manistee National Forests. In
USDA Forest Service, Huron-Manistee National Forest

Coddington, J .A., C.E. Griswold, D.S. Davila, E. Penaranda, and S.F. Larcher. 1991.
Designing and testing sampling protocols to estimate biodiversity in tropical
ecosystems. pp. 1048. In The unity of evolutionary biology: Proceedings from
the fourth international congress of systematic and evolutionary biology, edited by
E. C. Dudley. Portland, Oregon: Dioscorides Press.

Connell, J .H. 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302-
1310.

Conover, W.J. 1971. Practical Nonparametric Statistics. New York: John Wiley and Sons,
Inc. .

Cooper, R.J., K.M. Dodge, and RC. Whitmore. 1987. Estimating defoliation using
stratiﬁed point intercept sampling. Forest Science 33:157-163.

Corbett, E.S., and J .A. Lynch. 1987. The gypsy moth-does it affect soil and water
resources? pp. 39-46. In Coping with the gypsy math in the new frontier,
Proceedings. Workshop for Forest Managers. 1987, West Virginia University,
Morgantown, WV.

 

188

Decagon Devices, 1. 1987. Sunﬂeck PAR Ceptometer Plant Canopy Measurement
Operator's Manual. Pullman, Washington USA.

Denno, R.F., M.S. McClure, and J .R. Ott. 1995. Interspecific interactions in
phytophagous insects: competition reexamined and resurrected. Annual Review of
Entomology 40: 297-331.

Desender, K., and H. Turin. 1989. Loss of habitats and changes in the composition of the
ground and tiger beetle fauna in West European countries since 1950 (Coleoptera:
Carabidae, Cincindelidae). Biological Conservation 48:277-294.

Dethier, V.G. 1959. Egg-laying habitats of lepidoptera in relation to available food.
Canadian Entomologist 91:554-561.

Dijk, T.S.V. 1986. Changes in the carabid fauna of a previously agricultural ﬁeld during
the ﬁrst twelve years of impoverishing treatments. Netherlands Journal of Zoology
36:4 13-437.

Dreistadt, SH. 1983. An assessment of gypsy moth eradication attempts in Michigan
(Lepidoptera: Lymantriidae). Great Lakes Entomologist 16:143-148.

Edmunds, G.F. 1973. Ecology of black pineleaf scale (HomopterazDiaspididae).
Environmental Entomology 2:765-777.

Erwin, TL. 1982. Tropical forests: Their richness in Coleoptera and other arthropod
species. Coleopterists Bulletin 36:74-75.

Erwin, T .L. 1983. Beetles and other insects of tropical forest canopies at Manuas, Brazil,
sampled by insecticidal fogging. 59-76. In Tropical Rain Forest Ecology and
Management, edited by S. L. Sutton, T. C. Whitrnore and A. C. Chadwick:
Blackwell, Oxford.

Erye, M.D., M.L. Luff, S.P. Rushton, and CJ. Topping. 1989. Ground beetles and
weevils (Carabidae and Curculionoidea) as indicators of grassland management
practices. Journal of Applied Entomology 107:508-517.

Gage, S.H., T.M. Wirth, and GA. Simmons. 1990. Predicting regional gypsy moth
(Lymantriidae) population trends in an expanding population using pheromone trap
catch and spatial ananlysis. Environmental Entomology 19:370-377.

Gansner, D.A., O.W. Herrick, P.S. DeBald, and RE Acciavatti. 1983. Changes in forest
condition associated with gypsy moth. Journal of Forestry 81: 155-157.

 

 

 

189

Goulet, H., and J.T. Huber. 1993. Hymenoptera of the world: An identiﬁcation guide to
families. In Research Branch Argriculture Canada Bulletin 1894/15.

Gross, H.L. 1992. Impact analysis for a jack pine budworm infestation in Ontario.
Canadian Journal of Forestry 22:818-831.

Hairston, N.G., F.E. Smith, and LB. Slobodkin. 1960. Community structure, population
control, and competition. American Naturalist 94:421-425.

Hansen, R.M., and D.N. Ueckert. 1970. Dietary similarity of some primary consumers.
Ecology 51:640-648.

Hawkins, CR, and J .A. MacMahon. 1989. Guilds: the multiple meanings of a concept.
Annual Review of Entomology 34:423-451.

Holliday, NJ. 1991. Species responses of carabid beetles (Coleoptera: Carabidae) during
post-ﬁre regeneration of boreal forest. Canadian Entomologist 123:1369-1389.

Host, G.E., K.S. Pregitzer, C.W. Ramm, J .B. Hart, and D.T. Cleland. 1987. Landforrn-
mediated differences in successional pathways among upland forest ecosystems in
northwestern lower Michigan. Forest Science 33:445-457.

Host, G.E., K.S. Pregitzer, C.W. Ramm, D.P. Lusch, and D.T. Cleland. 1988. Variation
in overstory biomass among glacial landforms and ecological land units in
northwestern Lower Michigan. Canadian Journal of Forest Research 18:659-668.

Houston, DR, and HT. Valentine. 1985. Classifying forest suscetibility to gypsy moth
defoliation. Agriculture Handbook No. 542. USDA Combined Forest Pest
Research and Development Program 19 pp.

Huston, M.A. 1994. Biological Diversity: The Coexistence of Species on Changing
Landscapes. New York: Cambridge University Press.

Huston, M., and J .G. Isebrands. 1995. Monitoring interception of photosynthetically
active radiation in forests. Bulletin of the Ecological Society of America 76:95-97.

Jeffords, M.R., and LJ. Case. 1987. Effect of prey density on diurnal activity and
development in Calosoma calidum (Coleoptera: Carabidae): Implications for
biological control of the gypsy moth, Lymantria dispar (Lepidoptera: Lymantriidae)
in the Midwest. Great Lakes Entomologist 20:93-97.

 

190

Johnson, K.S., J .M. Scriber, J.K. Nitao, and DR. Smitley. 1995. Toxicity of Bacillus
thuringiensis var. kurstaki to three nontarget Lepidoptera in ﬁeld studies.
Environmental Entomology 24:288-297.

Kelly, B., and J. Regniere. 1985. Predation on the pupae of the spruce budworm
(Lepidoptera: Tortricidae) on the forest floor. Canadian Entomologist 117:33-38.

Klein, R.M., and TD. Perkins. 1988. Primary and secondary causes and consequences of
contempory forest decline. The Botanical Review 54:1-43.

Kolodny-Hirsch, D.M. 1986. Evaluation of methods for sampling gypsy moth
(lepidopteran Lymantriidae) egg mass populations and development of sequential
sampling plans. Environmental Entomology 15:122-127.

Kulman, HM. 1971. Effects of insect defoliation on growth and mortality of trees. Annual
Review of Entomology 16:289-324.

Kremen, C., R.K. Colwell, T.L. Erwin, D.D. Murphy, R.F. Noss, and M.A. Sanjayan.
1993. Terrestrial arthropod assemblages: Their use in conservation planning.
Conservation Biology 7:796-808.

LaVigne, R.J., J .A. Lockwood, and T. Christiansen. 1990. Insect response to the 1988
ﬁres in Yellowstone National Park: Final Report 1989-1990: University of
Wyoming and the National Park Service Research Center. Unpublished Report.

Lawton, J .H. 1986. The effects of parasitoids on phytophagous insect communities. 265-
287. In Insect Parasitoids, edited by J. Wage and D. Greathead. London:
Academic Press.

Lawton, J .H., and DR. Strong. 1981. Community patterns and competition in folivorous
insects. American Naturalist 1 18:317-338.

Leibherr, J ., and J. Mahar. 1979. The carabid fauna of the upland oak forest in Michigan:
Survey and analysis. Coleopterists Bulletin 33:183-197.

Lenski, RE 1982. The impact of forest cutting on the diversity of ground beetles
(Coleoptera: Carabidae) in the southern Appalachians. Ecological Entomology
72385-390.

Leonard, DE. 1974. Recent developments in ecology and control of the gypsy moth.
Annual Review of Entomology 19:197-229.

191

Liebhold, A.M., J .S. Elkington, D.R. Miller, and Y.S. Wang. 1988. Estimating oak leaf
area index and gypsy moth, Lymantria dispar (L.) (Lepidoptera: Lymantriidae),
defoliation using canopy photographs. Environmental Entomology 17:560-566.

Lindroth, CH. 1969. The ground-beetles of Canada and Alaska. Opuscula Entomologica
Supplenta 20, 24, 26, 29, 33, 34, 3521-1192.

Lovett, G.M., and A.E. Ruesink. 1995. Carbon and nitrogen mineralization from
decomposing gypsy moth frass. Oecologia 104:133-138.

Ludwig, J .A., and J .F. Reynolds. 1988. Statistical Ecology: A Primer on Methods and
Computing. New York: John Wiley and Sons.

Magurran, A.E. 1988. Ecological Diversity and Its Measurement. Princeton: Princeton
University Press.

Mallett, KL, and W.J.A. Volney. 1990. Relationships among jack pine budworm
damage, selected tree characteristics, and Armillaria root rot in jack pine. Canadian
Journal of Forest Research 20: 1791-1795.

Mattson, W.J. 1980. Herbivory in relation to plant nitrogen content. Annual Review of
Ecological Systematics 1 1:1 19-61.

Mattson, W.J., and ND. Addy. 1975. Phytophagous insects as regulators of forest
primary production. Science 190:515-522.

Mattson, W.J., and J .M. Scriber. 1987. Nutrional ecology of insect folivores of woody
plants: Nitrogen, water, ﬁber, and mineral considerations. 105-146. In Nutrional
Ecology of Insects, Mites, and Spiders, edited by F. Slansky and J. G. Rodriguez.
New York: John Wiley and Sons Inc.

Mauffette, Y., M.J. Lechowicz, and L. Jobin. 1983. Host preferences of the gypsy moth,
Lymantria dispar (L.), in southern Quebec. Canadian Journal of Forest Research
13:53-60.

McClure, MS. 1974. Biology of Erythroneura lawsoni (Homoptera: Cicadellidae) and
coexistence in the Sycamore leaf-feeding guild. Environmental Entomology 3:59-
68.

Milligan, G.W. 1989. A study of the beta-ﬂexible clustering method. Multivariate
Behavioral Research 24: 163- 176.

 

192

Miller, J .C. 1990. Field assessment of the effects of a microbial pest control agent on
nontarget Lepidoptera. American Entomologist 36: 135-139.

Millers, I., and D. Lachance. 1988. North American sugar maple decline project: Appendix
7.3, Quality assurance and quality control plan. In Eastern Hardwoods Research
Cooperative- National Acid Precipitation Assessment Program of U. S. A. Canada
Agriculture Forestry Service and USDA Forest Service. 13 pp.

Minott, C.W., and LT. Guild. 1925. Some results of the defoliation of trees. Journal of
Economic Entomology 18:345-348.

Moran, V.C., and T.R.E. Southwood. 1982. The guild composition of arthropod
communities in trees. Journal of Animal Ecology 51.

Muzika, R.-M. 1994. Defoliation and thinning effects on gypsy moth predators. Gypsy
Moth News 36:2-3.

Niemela, J., Y. Haila, E. Halme, T. Lahti, T. Pajunen, and P. Punttila. 1988. The
distribution of carabid beetles in fragments of old coniferous taiga and adjacent
managed forest. Annales Zoologici F ennici 25: 107-1 19.

Niemela, J ., Y. Haila, E. Halme, T. Pajunen, and P. Punttila. 1990. Diversity variation in
carabid beetle assemblages in the southern Finnish taiga. Pedobiologia 34: 1-10.

N iemela, J ., D. Langor, and J .R. Spence. 1993. Effects of clear-cut harvesting on boreal
ground-beetle assemblages (Coleoptera: Carabidae) in Western Canada.
Conservation Biology 7:551-561.

Noss, R.F. 1990. Indicators for monitoring biodiversity: A hierarchical approach.
Conservation Biology 4:355-364.

Perkins, T.D., H.W. Vogelmann, and R.M. Klein. 1987. Changes in light intensity and
soil temperature as a result of forest decline on Camels Hump, Vermont. Canadian
Journal of Forestry Research 17:565-568.

Pierce, L.L., and SW. Running. 1988. Rapid estimation of coniferous forest leaf area
index using a portable integrating radiometer. Ecology 69: 1762-1767.

Price, PW. 1987. The role of natural enemies in insect populations. 578. In Insect
Outbreaks, edited by P. Barbosa and J. C. Schultz. San Diego: Academic Press.

193

Probst, J .R., and TR. Crow. 1991. Integrating biological diversity and resource
management: An essential approach to productive, sustainable ecosystems. Journal

of Forestry 89:12-17.

Purrington, F.F., J.B. Bater, M.G. Paoletti, and ER. Stinner. 1989. Ground beetles from
remnant oak-maple-beech forest and its surroundings in northeastern Ohio
(Coleoptera: Carabidae). Great Lakes Entomologist 22: 105-1 10.

Risley, LS, and DA. Crossley. 1993. Contribution of herbivore-caused greenfall to
litterfall nitrogen ﬂux in several southern Appalachian forested watersheds.
American Midland Naturalist 129:67-74. I.

Risser, PG. 1995. Biodiversity and ecosystem function. Conservation Biology 9:742-
746.

 

Root, RB. 1967. The niche exploitation pattern of the blue-gray gnatcatcher. Ecological
Monographs 4.

!”

Root, RB. 1973. Organization of plant-arthropod association in simple and diverse
habitats: The fauna of collards (Brassica oleracea). Ecological Monographs 43:95-
124.

Ross, H.H. 1957. Principles of natural coexistence indicated by leafhopper populations.
Evolution 11:113-129.

Sample, B.E., L. Butler, C. Zivkovich, and RC Whitmore. 1993. Evaluation of Bacillus
thuringiensis and defoliation effects on native lepidoptera. In AIPM Preliminary
Report.

Sapio, F. Michigan DNR Forest Management Division, Forest Health Web Page.
<http:l/www.dnr.state.mi.us/www/fmd/pest/forheal.html> May 1, 1996.
(Accessed June 13, 1996).

SAS Institute, 1995. JMP Statistics and Graphics Guide Version 3.1. Cary, North
Carolina, USA.

Schowalter, TD. 1981. Insect herbivore relationship to the state of the host plant: Biotic
regulation of ecosystem nutrient cycling through ecological succession. Oikos
37:126-130.

Schowalter. TD. 1985. Adaptations of insects to disturbance. 235-252. In The Ecology of
Natural Disturbance and Patch Dynamics, edited by S.T.A. Pickett and P.S. White.
New York: Academic Press.

194

Schowalter, TD. 1989. Canopy arthropod community structure and herbivory in old-
growth and regenerating forests in western Oregon. Canadian Journal of Forestry

Research 19:318-322.

Schowalter, T. D. 1991. Roles of insects and diseases in sustaining forests. Paper read at
Siviculture, Forest Genetics and Tree Improvement, Forest Pest Management, and
Soils section, SAF National Convention, at San Francisco, CA.

Schowalter, TD. 1994. Invertebrate community structure and herbivory in a tropical rain
forest canopy in Puerto Rico following Hurricane Hugo. Biotropica 26:312-319.

Schowalter, TD. 1995. Canopy arthropod communities in relation to forest age and
alternative harvest practices in western Oregon. Forest Ecology and Management

78:115-125.

Schowalter, T.D., J .W. Webb, and DA. Crossley. 1981. Community structure and
nutrient content of canopy arthropods in clearcut and uncut forest ecosystems.
Ecology 62: 1010-1019.

Schultz, J.C., and LT. Baldwin. 1982. Oak leaf quality declines in response to defoliation
by gypsy moth larvae. Science 217: 149-151.

Scriber, J .M., and F. Slansky. 1981. The nutritional ecology of immature insects. Annual
Review of Entomology 26:183-211.

Simberloff, D., and T. Dayan. 1991. The guild concept and the structure of ecological
communities. Annual Review of Ecology 22:115-43.

Simberloff, D.S., and ED. Wilson. 1969. Experimental zoogeography of islands: The
colonization of empty islands. Ecology 50:278-296.

Sokal, RR, and F.J. Rohlf. 1995. Biometry: the principles and practice of statistics in
biological research. Third Edition ed. New York: W. H. Freeman and Company.

Southwood, T.R.E. 1994. Ecological Methods: With Particular Reference to the Study of
Insect Populations. Second Edition ed. London: Chapman and Hall.

Spurr, S.H., and B.V. Barnes. 1980. Forest Ecology. Third ed. New York: John Wiley
and Sons.

 

195

Stork, N .E. 1987. Guild structure of arthropods from the Bomean rain forest trees.
Ecological Entomology 12:69-80.

Strong, D.R., J.H. Lawton, and T.R.E. Southwood. 1984. Insects on Plants: Community
Patterns and Mechanisms. Cambridge: Harvard University Press.

Swank, W.T., J.B. Waide, D.A. Crossely, and R.L. Todd. 1981. Insect defoliation
enhances nitrate export from forest ecosystems. Oecologia 51:297-299.

Talerico, R.L. 1981. The gypsy moth: Research toward integrated pest management. In
Forest Service Science and Education Agency Technical Bulletin.

Tilman, D. 1994. Competition and biodiversity in spatially structured habitats. Ecology
75:2-16.

Tilman, D. 1996. Biodiversity: Population versus ecosystem stability. Ecology 72:350-
363.

Tilman, D., and J .A. Downing. 1994. Biodiversity and stability in grasslands. Nature
367:363-364.

Turin, H., and R]. denBoer. 1988. Changes in the distribution of carabid beetles in the
Netherlands since 1880. II. Isolation of habitats amd long-term trends in the
occurence of carabid species with different powers of dispersal (Coleoptera,
Carabidae). Biological Conservation 44: 179-200.

Twery, M.J. 1990. Effects of defoliation by gypsy moth. pp. 27-39. In: USDA
Interagency Gypsy Moth Research Review 1990, Proceedings. General Technical
Report NE-l46. USDA Forest Service. 1990, East Windsor, CT.

Varley, SC. 1949. Population changes in German forest pests. Journal of Animal
Ecology 18:117-122.

Vinson, SB. 1990. Potential impact of microbial insecticides on beneﬁcial arthropods in
the terrestrial environment. 259. In Safety of Microbial Insecticides, edited by M.
Laird, L. A. Lacy and E. W. Davidson. Boca Raton: CRC Press Inc.

Vitousek, P.M. 1990. Biological invasions and ecosystem processes: towards an
integration of population biology and ecosystem studies. Oikos 57:7-13.

 

196

Vose, J .M., and W.T. Swank. 1990. Assessing seasonal leaf area dynamics and vertical
leaf area distribution in eastern white pine (Pinus strobus L.) with a portable light
meter. Tree Physiology 7: 125-134.

Wargo, P.M. 1977. Armillaria mellea and Agrilus bilineatus and mortality of defoliated oak
trees. Forest Science 23:485-492.

Wargo, P.M., and ME. Montgomery. 1983. Colonization by Armillaria mellea and
Agrilus bilineatus of oaks injected with ethanol. Forest Science 29:848-857.

Waring, RH. 1983. Estimating forest growth and efﬁciency in relation to canopy leaf area.
Advances in Ecological Research 13:327-354.

Weseloh, RA. 1990. Experimental forest releases of Calosoma sycophanta (Coleoptera:
Carabidae) against the gypsy moth. Journal of Economic Entomology 83:2229-
2234.

Wilcove, D.S. 1989. Protecting biodiversity in multiple-use lands: lessons from the US
Forest Service. Trends in Ecology and Evolution 4:385-388.

Wilson, ED. 1985. The biological diversity crisis. BioScience 35:700-706.

Zak, D.R., K.S. Pregitzer, and G.E. Host. 1986. Landscape variation in nitrogen
mineralization and nitriﬁcation. Canadian Journal of Forest Research 16: 1258-
1263.

Zak, D.R., G.E. Host, and K.S. Pregitzer. 1989. Regional variability in nitrogen
mineralization, nitriﬁaction, and overstory biomass in northern Lower Michigan.
Canadian Journal of Forestry Research 19: 1521-1526.

Zak, D.R., and K.S. Pregitzer. 1990. Spatial and temporal variability of nitrogen cycling
in northern lower Michigan. Forest Science 39:367-380.