 

LIBRARY
Mink», . State

VI nau-

University

 

 

 

This is to certify that the
thesis entitled

Inﬂuence of host vigor on larval distribution, development, and
mortality of Agn’lus planipennis Fainnaire (Coleoptera:
Buprestidae) in North America

presented by

Andrew Roy Tluczek

has been accepted towards fulﬁllment
of the requirements for the

degree In EntomologL

 

 

 

1Major ﬁfésor’s‘v Si'ghatu e ’

34/1944. 41001

Date

MSU is an Afﬁrmative Action/Equal Opportunity Employer

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KlProi/AccanPresIClRCIDateDueindd

 

INFLUENCE OF HOST VIGOR ON LARVAL DISTRIBUTION, DEVELOPMENT,
AND MORTALITY 0F AGRILUS PLANIPENNIS FAIRMAIRE (COLEOPTERA:
BUPRESTIDAE) IN NORTH AMERICA
By

Andrew Roy Tluczek

A THESIS

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

MASTERS OF SCIENCE
Entomology

2009

ABSTRACT

INFLUENCE OF HOST VIGOR ON LARVAL DISTRIBUTION, DEVELOPMENT,
AND MORTALITY OF AGRILUS PLAN/PENNIS FAIRMAIRE (COLEOPTERA:
BUPRESTIDAE) IN NORTH AMERICA AS INFLUENCED BY HOST VIGOR

By

Andrew Roy Tluczek

Effects of host stress on the density, development, and mortality of Agn'lus
planipennis Fairmaire (Coleoptera: Buprestidae) larvae were evaluated in ﬁeld
and laboratory settings from 2006 to 2008. In ﬁeld studies, A. planipennis larval
density was consistently higher on girdled trees than non-girdled trees. A.
planipennis larval development was determined by tree vigor, and not larval
density.

Field data were used to construct a life-table for A. planipennis larvae for
2006 and 2007. Woodpecker predation was the largest source of A. planipennis
mortality. Woodpeckers fed on larvae most commonly near the tree top and
most predation occurred after the month of November. Mortality from unknown
causes, such as desiccation, pathogen, or infection, and from cannibalism was
rare. A higher percentage Of larvae died from unknown causes and cannibalism
below the girdle compared with above the girdle on girdled trees.

Laboratory studies Of A. planipennis larvae showed that cannibalism was
rare, even at high densities. Individual larvae each consumed about 10 cm2 Of

phloem before reaching the prepupal stage.

ACKNOWLEDGMENTS

I thank, ﬁrst and foremost, my advisor; Deborah G. McCullough. She has
skillfully mentored me through studies, presentations, and writing. Deb gave me
direction and advice, but I always felt like she treated me with respect. I felt like
she gave my ideas and goals the space tO develop on their own. I would also
like to thank my committee, Therese Poland and Jim Miller. Their guidance was

helpful for developing my thesis and to my growth as a scientist.

My studies would never have been accomplished with the same efﬁciency
or detail without the help of many people. Nate Siegert, Jake Baker, James
Wieferich, David Cappaert, Sarah Smith, Tara Dell, Bob McDonald, Stephanie
Blumer, Chis Pell, Keali Chambers, Ben Schmitt, Erin Burkett, Steven Burr, and
Allison Wroblewski were indispensible in the ﬁeld and lab. I don’t know how I
could have ﬁnished my projects without the many hours they helped me with my
projects and the foil for ideas that they provided me with. They were all a valued
source of advice and inspiration. I would also like to thank my fellow graduate
students, Andrea Anulewicz, Rodrigo Mercader, Sara Tanis, and Chenin
Limback, who gave me valuable help in the ﬁeld and lab as I worked on my

studies and as I wrote my thesis.

I appreciate the help and cooperation from John Durling, John Lief, Dave
Burgdorf, and the staff of the USDA-NRCS at the Rose Lake Plant Materials

Center. They gave us access to the ash plantation that was so critical to my

iii

thesis. I also appreciate the maintenance they performed at the plantation. Paul
Bloese and Randy Klevickas at the Tree Research Center at Michigan State
University gave me crucial advice they gave me regarding raising seedlings and
cut sections of ash and they were always willing to give me a little bit more space
to work in. Debbie Miller and Tina Kuhn were very helpful in answering
questions related tO beetle rearing and oviposition, and even shared beetles in

an emergency.

I could not have done this without the love, support, and patience of my
wife Claire. NO matter what, she was always there, giving me strength and
resilience. I also thank all of my family and friends for their support and

encouragement for the past three years.

Funding for this work was provided by the USDA Forest Service,
Northeastern Area Forest Health Protection, and Michigan States University’s

Project GREEEN.

iv

PREFACE

Emerald ash borer, Agrilus planipennis Fairmaire (Coleoptera:
Buprestidae), a phloem-feeding buprestid native to Asia, was ﬁrst discovered in
Southeastern Michigan and Essex County, Ontario in June 2002. Initial
Observations in Michigan indicated that A. planipennis larvae had a univoltine life
cycle; eggs laid in June and July, larvae fed in the summer, overwintered as
prepupae, and then emerged as adults the following spring. However,
subsequent studies have found that at least some A. planipennis larvae take two
years to develop. There is limited information available on the factors that
inﬂuence A. planipennis mortality and no life-tables have been developed.
Understanding factors that affect larval development and inﬂuence mortality
would be helpful for modeling A. planipennis spread and population dynamics,

and for improving survey activities Of program managers.

This thesis focused on evaluating effects Of host vigor on the density,
development, and mortality Of larval A. planipennis in Michigan. In Chapter One,
ﬁeld studies were conducted to evaluate the effects of host stress on density and
development of larvae on green ash, Fraxinus pennsylvanica Marshall
(Oleaceae), trees. In Chapter Two, I assessed the factors that affect A.
planipennis larval mortality in the same ﬁeld site and constructed life-tables in
2006 and 2007. Chapter Three addressed the mechanisms that may affect larval
development using larvae produced by adult beetles caged on trees. The goal

was to determine whether larval density or host stress determined the rate of

larval development. I also calculated the amount Of phloem required for A.

planipennis larvae to complete development.

The goal Of Chapter 4 was to determine if larval cannibalism occurred,
whether cannibalism was related to population density, and whether it increased
rate of development of the surviving larvae. For this laboratory study, white ash,
Fraxinus amen'cana Linnaeus (Oleacea), phloem arenas sandwiched between
two pieces Of Plexiglas were used to Observe A. planipennis larvae. In a second
study, larval feeding and mortality on small sections of white ash stems were
Observed. Results were used to calculate the amount of phloem necessary for A.

planipennis larvae to complete development.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. ix
LIST OF FIGURES ................................................................................ ix
CHAPTER 1
Effects of Host Vigor on the Density, Development, and Distribution of Agrilus
planipennis Adults and Larvae .................................................................. 1
Introduction ................................................................................. 1
Materials and Methods ................................................................... 3
Study Site ........................................................................... 3
2006 Study .......................................................................... 3
2007 Study ......................................................................... 6
Statistical Analysis ................................................................ 7
Results ....................................................................................... 9
2006 Study ......................................................................... 9
2007 Study ....................................................................... 12
Discussion ................................................................................. 16
Adult and Larval Distribution ................................................. 16
Larval Development ............................................................ 20
Tables ....................................................................................... 26
Figures ..................................................................................... 27
CHAPTER 2
Mortality Of Agrilus planipennis larvae in green ash trees .............................. 41
Introduction ............................................................................... 41
Materials and Methods ................................................................. 43
Statistical Analysis ............................................................. 45
Construction Of Life-Tables ................................................... 46
Results ...................................................................................... 47
2006 Study ........................................................................ 47
2007 Study ....................................................................... 50
Discussion ................................................................................. 53
Tables ....................................................................................... 59
Figures ..................................................................................... 61
CHAPTER 3
Development of Agrilus planipennis larvae (Coleoptera: Buprestidae)
on stressed and healthy Fraxinus pennsylvanica trees ................................. 65
Introduction ................................................................................ 65
Materials and Methods ................................................................. 67
2006 Study ....................................................................... 67
2007 Study ....................................................................... 70
Phloem Disruption — Pilot Study ............................................ 71

vii

Statistical Analysis .............................................................. 71

Results ..................................................................................... 72
2006 Study ....................................................................... 73
2007 Study ....................................................................... 74
Phloem Disruption - Pilot Study ............................................ 76
Discussion ................................................................................. 77
Tables ....................................................................................... 84
Figures ..................................................................................... 85

CHAPTER 4 ‘
Development, survival and phloem consumption of Agrilus
planipennis Fairmaire larvae phloem arenas and potted ash

sections .............................................................................................. 87
Introduction ................................................................................ 87
Materials and Methods ................................................................. 89

Phloem Arenas .................................................................. 89
Potted Fraxinus americana Sections ...................................... 91
Results ..................................................................................... 95
Phloem Arenas .................................................................. 95
Potted Fraxinus americana Sections ....................................... 96
Discussion ................................................................................. 98
Phloem Arenas .................................................................. 98
Potted Fraxinus americana Sections ....................................... 99
Tables ..................................................................................... 102
Figures .................................................................................... 103

APPENDICCES

APPENDIX 1

Record of Deposition of Voucher Specimens .......................................... 107

LITERATURE CITED .......................................................................... 1 10

viii

LIST OF TABLES

Table 1.1. Mean (:I: SE) tree DBH, height, surface area sampled per tree,
phloem thickness, and the percent Of larvae developing in one

year for three treatments of plantation trees in 2006 and 2007 .............. 26
Table 2.1. Age speciﬁc life table Of A. planipennis larvae for 2006 .................. 59
Table 2.2. Age speciﬁc life table of A. planipennis larvae for 2007 .................. 60

Table 3.1. Mean (:I: SE) tree diameter at breast height (DBH), height,

wild larval density and cages and trees and A. planipennis larvae

under cages from adults caged on Fraxinus pennsylvanica in 2006

and 2007 ............................................................................................. 84

Table 4.1. Summary Of variables recorded on phloem arenas with
A. planipennis larvae and mean (t SE) number Of larvae per arena,

and percentage Of larvae that were 4th instar ............................................ 102

ix

LIST OF FIGURES

Figure 1.1. Mean (:I: SE) density of A. planipennis larvae per m2 on

control and methyl jasmonate trees, and above and below the girdle

on girdled trees in 2006 and 2007. Within year means with the

same letter are not signiﬁcantly different (T ukey protected LSD test;

P <0.05) ............................................................................................. 27

Figure 1.2. Log1o(x+1) relationship between A. planipennis larvae

per m2 and total adult capture on sticky bands on trees in 2006 (A)
and 2007 (B) ....................................................................................... 28

Figure 1.3. (A) Mean (:I: SE) larval density by height in trees in 2006.

(B) Mean (:I: SE) percent of larvae by height in trees in 2006. Means

within different heights of the same treatment with the same letter

(a, b, and c) are not signiﬁcantly different. Means within the same

height of different treatments with the same letter(x and y) are not

signiﬁcantly different (T ukey protected LSD test; P <0.05) ............................. 29

Figure 1.4. Mean (:I: SE) percent of sections by height that were

infested with A. planipennis larvae for 2006 (A) and 2007 (B).

Means within different heights of the same treatment with the same

letter (a, b, and C) are not signiﬁcantly different. Means within the

same height of different treatments with the same letter(x and y) are

not signiﬁcantly different (Tukey protected LSD test; P <0.05) ........................ 31

Figure 1.5. Mean (:I: SE) percentage of A. planipennis larvae that

had reached at least 4th instar on control and methyl jasmonate trees,

and above and below the girdle on girdled trees in 2006 and 2007.

Within year, means with the same letter are not signiﬁcantly different

(T ukey protected LSD test; P <0.05) ......................................................... 33

Figure 1.6. The log1o+1 relationship between the percent of 1-yr
larvae and larvae per m square on control (A), methyl jasmonate
(B) and girdled (C) trees in 2006 (GLM; Only girdled trees were
signiﬁcant P <0.05) ............................................................................... 34

Figure 1.7. Mean (:I: SE) width Of A. planipennis prepupal galleries

on control and methyl jasmonate trees, and above and below the

girdle on girdled trees in 2006 and 2007. Within year, means with

the same letter are not signiﬁcantly different (T ukey protected LSD

test; P <0.05) ....................................................................................... 36

Figure 1.8. (A) Mean (1: SE) larvae density m2 by height in tree in 2007.

(B) Mean (:I: SE) percent Of total larvae by height in tree in 2007.

Means within different heights of the same treatment with the same

letter (a, b, c, and d) are not signiﬁcantly different. Means within the

same height Of different treatments with the same letter(x and y) are

not signiﬁcantly different (Tukey protected LSD test; P <0.05) ........................ 37

Figure 1.9. The log1o+1 relationship between the percent Of 1-yr

larvae and larvae per m square on control (A), methyl jasmonate (B)

and girdled (C) trees in 2007 (GLM; No relationships were

signiﬁcant, P <0.05) ............................................................................. 39

Figure 2.1. Regression of A. planipennis larva mortality by woodpecker

predation versus larval density in 2006. Log1o(x+1) model is signiﬁcant
(GLM; P<0.05) ..................................................................................... 61

Figure 2.2. Mean (:I: SE) proportion Of A. planipennis larvae killed

by woodpecker predation per m2 by log height up to 8 m above ground
in 2006 and 2007. Within year means with the same letter are not
signiﬁcantly different (Tukey protected LSD test; P <0.05) ............................. 62

Figure 2.3. Mean (:I: SE) proportion of A. planipennis larvae killed by

woodpecker predation per m by month from October through

March in 2006 and 2007. Within year means with the same letter are

not signiﬁcantly different (Tukey protected LSD test; P <0.05). Trees

were not felled until December in 2006. Tree felling was completed by late
February in 2007 ................................................................................................. 63

Figure 2.4. Regression of A. planipennis larvae cannibalized by larval density in
2006 (A) and 2007 (B). Log1o(x+1)

models are signiﬁcant (GLM; P <005) ............. 64
Figure 3.1. Mean (t SE) density of A. planipennis larvae per m2
produced by caged beetles on untreated control trees, trees exposed

to methyl jasmonate, and girdled trees in 2006 and 2007. Within

each year, means with the same letter are not signiﬁcantly different

(Tukey protected LSD test; P <0.05) ......................................................... 85

Figure 3.2. Mean (:I: SE) percentage Of A. planipennis larvae produced

by cages per m2 that reached 4th instar or prepupal stage on untreated

control trees, trees exposed to methyl jasmonate, and girdled trees in

2006 and 2007. Within each year, means with the same letter are

not signiﬁcantly different (Tukey protected LSD test; P <0.05) ........................ 86

xi

Figure 4.1. Accumulated mortality of A. planipennis larvae on phloem
arenas by week for A) high density phloem arenas (n=120 larvae) and
B) low density phloem arenas (n=15 larvae) ............................................. 103

Figure 4.2. Percentage of surviving larvae by instar by week for A) high
density phloem arenas (n=120 larvae) and B) low density phloem
arenas (n=15 larvae) ........................................................................... 104

Figure 4.3. Percentage Of A. planipennis larvae by instar on potted

F. amen'c’ana sections debarked three to 11 weeks after adult beetles

were placed in cages. n=13 larvae (Week 3), n=14 larvae (Week 4),

n=13 larvae (Week 6), n=12 larvae (Week 7), n=2 larvae (Week 8), and

n=9 larvae (Week 10) .......................................................................... 105

Figure 4.4. Mean (i SE) area phloem consumed and mean

horizontal gallery dimension of A. planipennis larvae by instar on potted F.
americana sections ............................................................................. 106

xii

CHAPTER 1

Density, Development, and Distribution of Agrilus planipennis

Adults and Larvae as Inﬂuenced by Host Vigor

INTRODUCTION

Emerald ash borer, Agn’lus planipennis Fairmaire (Coleoptera:
Buprestidae), is a phloem-feeding beetle native to Asia that was ﬁrst discovered
in southeastern Michigan, USA, and Essex County, Ontario, Canada in June
2002. To date, the host range Of A. planipennis in North America is known to
include Fraxinus americana Linnaeus (white ash), F. pennsylvanica Marshall
(green ash), F. nigra Marshall (black ash), F. quadrangulata Michaux (blue ash)
and F. pmfunda (Bush) (pumpkin ash). A. planipennis is thought to be able to
colonize all ash species native to North America (Anulewicz et al. 2006a, 2008a).

Initial studies in southeast Michigan indicated that the life cycle of A.
planipennis was univoltine (Bauer et al. 2004). Adults oviposited during the
summer, larvae fed from late summer tO fall, then oven/vintered as pre-pupal
larvae. Pupation occurred in spring and adults emerged in summer (Cappaert et
al. 2005).

Recent Observations, however, have indicated that at least some larvae
feed for two summers. For example, in newly established infestations with low
densities of A. planipennis, only 18% Of 282 larvae had developed in one year

(Cappaert et al. 2005). Additionally, data from a detection tree site indicated that

larvae in vigorous trees developed more slowly than larvae in girdled trees
(Siegert and McCullough, unpublished data). Such prolonged development, if
common, would strongly inﬂuence A. planipennis rate Of spread, population
dynamics, survey and control activities Of program managers. More information
Is necessary to determine what effects host vigor and larval density have on
larval development.

This study had three Objectives including (1) to determine the role of host
vigor on A. planipennis larval development; (2) to determine if density affects the
rate Of development Of A. planipennis larvae and (3) to assess the distribution of

captured adults and larvae among trees experiencing different levels Of stress.

MATERIALS AND METHODS

Study Site

This study was conducted using green ash (Fraxinus pennsylvanica) trees
in a well-stocked plantation established in 1980 and located in the Rose Lake
State Wildlife Research area in Clinton County, Michigan. The plantation was
organized into 54 rows of 10 trees each, spaced at 1.5 meters. Occasional trees
were missing, but 363 trees with an average DBH of 10.8 :I: 0.20 cm were
available for this study. Original planting records were lost, but the plantation
was presumably established as a replicated provenance test (J. Durling, Rose
Lake Plant Material Center, pers. comm.) Initial Observations in fall 2005 and
spring 2006 indicated that emerald ash borer was not yet established in this
plantation. In May 2006, upon closer inspection, however, 11 trees infested with
a very low-density population of emerald ash borers (mean larval density 12.8 :t
7.2 larvae per m2, range 1 to 197 larvae per tree) were identiﬁed, felled and

removed.

2006 Study

In May 2006, ten blocks Of nine trees each were established for a total Of
90 trees. Since all trees were in similar growing conditions and were of similar
size, blocks were assigned based on the proximity Of trees to each other. Trees

were chosen in a way to ensure that their canopies did not contact or overlap

with each other. Diameter at breast height (DBH) was measured in May of 2006
(Table 1.1).

Three trees within each block were randomly assigned to one Of three
stress treatments. Trees were either girdled (30 trees), exposed to the stress-
elicitor methyl jasmonate (30 trees), or left as untreated controls (30 trees).
Girdled trees had a 20 cm band Of outer bark and phloem removed between 0.85
m and 1.0 m on 10 May. Methyl jasmonate a volatile derivative of jasmonic acid,
can elicit stress responses from plants (Gols et al. 2003, Rodriguez-Sacha and
Thaler 2005), including ash seedlings (Poland 2005, Rogriguez-Saona et al.
2006). TO expose trees to methyl jasmonate, a line with 20 bubble caps spaced
20 cm apart was suspended in the canopy. Each bubble cap had 150 pl of
methyl jasmonate with a release rate (per bubblecap) of 0.38 mg per day
determined in the laboratory at 20 °C (Con Tech Inc., Delta, BC). Bubble caps
were suspended through the canopy on 17 May, using a Big Shot Line
LauncherTM (WesSpur Tree Equipment, Bellingham WA). DBH and tree height
did not vary signiﬁcantly among treatments and so were not included in further
analyses.

During the summer of 2006, experimental trees were qualitatively ranked
from 1 to 5 based on the amount of canopy exposure. A score Of 1 indicated an
Open grown tree, 2 indicated that the tree was super-dominate and most of the
leaves extended above the rest of the canopy, 3 indicated that the tree was

bordered by one or two trees, 4 indicated that the tree was bordered by at least

three trees, and 5 indicated that the tree was suppressed, and little or no foliage
extended into the canopy.

I captured adult A. planipennis by wrapping a 20 cm band of plastic wrap
between 1.3 m and 1.5 m above the ground on each tree. TanglefootTM (The
Tanglefoot Company, Grand Rapids, MI) was applied tO the band. Over the
summer, the bands become less sticky, so TanglefootTM was reapplied in July.
We collected adult A. planipennis from the sticky bands once a week from 21
June until 16 August. Adults were Cleaned with 70% ethanol and sexed under a
stereoscope. The ﬁrst and second abdominal sections of the male are narrower
than those of the female. A male also has a visible layer Of setae on the ventral
side of the thorax which females lack (Rodriguez-Saona et al. 2007).

Trees were felled and debarked from December 2006 through March 2007
to quantify larval density and development. Trees were felled in blocks so that
trees in each block were felled within a few days Of each other. After trees were
felled, tree height was measured, then each tree was bucked into one m
sections, up tO at least 8 m aboveground. Bark area was estimated by
measuring the diameter of each section at the midpoint then calculating the
surface area. The one m logs were returned to the lab and debarked with Chisels
and drawknives.

The numbers and life stage Of A. planipennis larvae were recorded and
density was standardized per m2 of exposed surface area for each tree. lnstars
were identiﬁed by head capsule width (Cappaert et al. 2005b). It was assumed

that larvae found as 15‘, 2nd, or 3rd instar would probably not reach the prepupal

stage by spring, and would require another summer Of feeding resulting in a two
year life cycle. Observations by Cappaert et al. (2005a) supported this; 2"d and

3"I instars present in April did not complete development before the end of the
summer. Larvae that are 4th instars or prepupae in late fall or winter are

expected to emerge in the spring as adults, resulting in a one year life cycle
(Cappaert et al. 2005a).

Up to three galleries Of prepupal larvae were randomly selected on each
log by choosing a prepupa that was Closest tO the top, the bottom and the middle

Of the log. The maximum width and length of each gallery was measured.

2007 Study

The study was replicated in 2007 with a few modiﬁcations. Ten blocks Of
nine trees were selected using the same parameters as in 2006 for a total Of 90
trees. Trees were girdled on 10 May. A line with ten methyl jasmonate bubble
caps spaced 20 cm apart was suspended through the canopies on 25 May, and
a second line with 10 methyl jasmonate bubble caps was suspended on 25 June.
While this did not Change the total amount of methyl jasmonate released, it was
expected that this would create a more sustained level over the summer. Sticky
bands were checked for A. planipennis adults once a week from 29 May until 29
August. Trees were felled and debarked from October 2007 through February
2008. Phloem thickness was measured on the cut ends Of each log in 2007

using a caliper. Two measurements were taken on each end and the mean of

the four measurements was used for analysis. Bark thickness did not vary among

treatments and so was not used in further analyses (Table 1.1).

Statistical Analysis

All data were tested for normality using the Shapiro-Wilk test (Shapiro and
WIlk 1965) and residual plots. Four response variables were evaluated including
number Of captured adults, larval density, the proportion of larvae that developed

in one year, and the horizontal gallery dimensions. Number of captured adults,
larval density, and horizontal gallery dimensions were log1o(x+1) transformed tO

normalize data. Proportion of larvae that developed in one year required no
transformation.

All response variables were analyzed using the GLIMMIX procedure for
mixed models in SAS statistical software (PROC GLIMMIX; SAS Institute 2003).
Effects Of three factors: tree vigor with three levels (control, methyl jasmonate, or
girdled), canopy exposure with ﬁve levels (1, 2, 3, 4, 5), and height in tree (in one
m increments) were analyzed. Block and block crossed with treatment were
used as random factors. Differences among treatment means were tested as
unplanned comparisons and multiple comparison tests were applied only when
overall analysis of variance (ANOVA) was signiﬁcant (P<0.05). When signiﬁcant
differences among levels occurred, the Tukey protected least signiﬁcant

difference (LSD) test was used to evaluate those differences (Ott and
Longnecker 2001). The Iog1o(x+1) relationship between larval density and adult

catch was analyzed using a general linear model (PROC GLM, SAS institute

2003). The linear relationship between larval density and the proportion of one
year larvae on each of the three vigor treatments was analyzed separately using

a general linear model with transformed data (PROC GLM, SAS institute 2003).

RESULTS
2006 Study

Overall, a total of 81 adults were captured on sticky bands in 2006. At
least one A. planipennis adult was captured on 5, 9, and 24 of the control, methyl
jasmonate, and girdled trees, respectively. Most adult beetles were captured in
July; 16, 64, and 4 adults were captured in June, July, and August, respectively.
Peak captures occurred during the week preceding 7 July, with 27.4% of all
beetles captured, and during the week preceding 19 July when 26.2% of all
beetles were captured. Girdled trees captured signiﬁcantly more adults than

trees in other treatments, with an average Of 0.2 :I: 0.1, 0.4 :I: 0.1, and 2.1 :t 0.5
adults on control, methyl jasmonate, and girdled trees, respectively (F2,13=23.70,
P<0.001). Adult capture did not vary among canopy exposure classes
(F3,57=2.69, P=0.055). Of the total beetles captured, 79.1% were female.

A total Of 4,621 A. planipennis larvae were recorded when the 90 trees
were debarked; counts ranged from 0 to 514 larvae per tree. Overall, the mean
density Of A. planipennis larvae was 34.5 :t 5.8 larvae per m2. At least one larva
was found on 24, 21, and 30 Of the control, methyl jasmonate, and girdled trees,
respectively. Larval density was signiﬁcantly higher on girdled trees than on
trees treated with methyl jasmonate or untreated control trees (F2'13=85.02,
P<0.001) (Fig. 1.1). Larval density averaged 3.7 t 0.9, 6.7 :I: 4.4, and 71.1 :I: 10.9

larvae per m2 on control, methyl jasmonate, and girdled trees respectively.

Average larval density was not signiﬁcantly lower below the girdle than above it

on girdled trees (T 27=-2.32, P=0.118) (Fig. 1.1). Larval density did not vary

among canopy exposure ranks (F3,54=2.55, p=0.066). A log1o relationship

existed between larval density and adults captured was signiﬁcant (F1,37=59.64,

P<0.001) but relatively weak (Fig 1.2A).

Height signiﬁcantly affected larval density on girdled trees only
(F7544=10.46, P<0.001). Signiﬁcantly lower densities of A. planipennis larvae
were found between 2-4 m above ground compared with densities found at other
heights (Fig 1.3A). The interaction between height and treatment was not

signiﬁcant with respect to larval density (F14, 647=1.48, P=0.115). The

percentage Of total larvae in the tree also varied by height (F7,474=11.53,

P<0.001) for all treatments. The highest proportions of larvae were found

between two to four meters above ground (Fig. 1.3B). The interaction between

height and treatment signiﬁcantly affected the percent of total larvae (F14,

424:1.83, P=0.032). Larvae were more uniformly distributed with height on

girdled compared to control and methyl jasmonate trees, which had a stronger
peak between three and four m above ground (Fig. 1.38). The proportion of logs
with at least one larva did not vary signiﬁcantly by height among vigor treatments
(F14,859=2.67; P<0.001) (Fig. 1.4A). After larval mortality was taken into account,
the mean density of A. planipennis larvae that survived until the trees were
debarked averaged 2.9 :I: 0.7, 3.6 :I: 1.4, and 47.9 1 6.8 larvae per m2 on control,
methyl jasmonate, girdled trees, respectively.

Overall, an average Of 28.1% :t 3.2% larvae per tree had reached the 4th

instar or were prepupal larvae when trees were debarked in fall or winter.

10

Densities of larvae were higher on girdled trees, and had a larger proportion Of

larvae reached the 4th instar or prepupal stage on these trees than on control and
methyl jasmonate trees (F2,13=19.95, P<0.001). On average, 16.0% 1: 3.8%,

12.1% :t 4.2%, and 49.0% :t 4.3% of larvae were 4th instars or prepupae on

control, methyl jasmonate, and girdled trees, respectively (Fig. 1.5). Within

girdled trees, larvae below the girdle developed at more than two times the rate
of larvae feeding above the girdle (F3,27=40.16, P<0.001) (Fig. 1.5). Interestingly,

larvae above the girdle did not develop signiﬁcantly faster than larvae on control

and methyl jasmonate trees (Fig. 1.5). Larval development did not vary among

canopy exposure levels (F3,42=1.80, P=0.162). When all trees were analyzed,

larval development was not signiﬁcantly related to with larval density (F2,34=2.35,
P=0.103). However, a signiﬁcant relationship did occur between larval density
and development when only girdled trees were analyzed (F1,29=13.03, P=0.001)
(Fig 1.60). There was no signiﬁcant relationship between larval density and

development on control (F1,13=1.22, P<0.285) (Fig 1.6A) or methyl jasmonate

trees (F1'19=0.00, P<0.949) (Fig. 1.68).

Dimensions Of galleries made by prepupal larvae averaged 3.8 :I: 0.2 cm

wide and 7.8 i 0.4 cm long. Gallery width was correlated with gallery length

(F1.46=31.98, P<0.001 r2=0.42). Gallery width did not vary by treatment
(F2,11=1.96, P=0.187) or canopy exposure (F3,21=1.04, P=0.395). However,

galleries were signiﬁcantly wider when larvae fed below the girdle than above the

girdle (F3'57=52.74, P=0.001) (Fig. 1.7). Similarly, average length of galleries
11

was greater below the girdle than above, with mean lengths of 13.7 :I: 1.1 cm and

6.9 i 0.2 cm, respectively.

2007 Study

The number of A. planipennis adults captured on sticky bands increased
substantially in 2007. A total Of 426 adults were captured and at least one A.
planipennis adult was captured on 20, 21, and 30 control, methyl jasmonate, and
girdled trees, respectively. The largest number of beetles were captured in June,
with 0, 269, 145, and 12 captured in May, June, July, and August, respectively.
Peak capture occurred during the week preceding 21 June when 261 adults, or
35.7% Of all beetles were captured. Girdled trees captured signiﬁcantly more
adults than control and methyl jasmonate trees. On average, 1.3 :I: 0.2, 2.0 1: 0.4,

and 10.9 :t 2.0 adults were captured on control, methyl jasmonate, and girdled

trees, respectively (F2'13=52.63, P<0.001). The number of adults captured did

not vary by canopy exposure (F4.56=1.48, P=O.222). Overall, 55.8% of adult

beetles captured were female.

A total of 18,891 larvae were recorded on the 90 trees that were
debarked, ranging from 0 to 1,862 larvae per tree. Overall, the mean density of
A. planipennis larvae was 79.6 19.8 larvae per m2, a two-fold increase over
2006. At least one larva was found on 29, 29, and 30 of the control, methyl
jasmonate, and girdled trees, respectively. A signiﬁcantly higher density Of

larvae occurred on girdled trees than on control or trees treated with methyl

jasmonate (F2,13=30.87; P<0.001) (Fig. 1.1). Larval density averaged 38.4 :I:

12

10.4, 25.8 :I: 9.7, and 174.5 :I: 14.3 larvae per m2 on control, methyl jasmonate,

and girdled trees, respectively (Fig. 1.1). The average larval density was not

signiﬁcantly lower below than above the girdle on girdled trees (T 27=-0.64,
P=0.920) (Fig 1.1). Larval density did not vary by canopy exposure (F5,5o=1.08;
P=0.386). A signiﬁcant relationship existed between larval density and adults
captured (F1,89=66.40; P<0.001) (Fig 1.28).

Signiﬁcantly lower densities of A. planipennis larvae were feeding near the
top of the tree than in other sections (F7,859=26.09, P<0.001) (Fig 1.8A). With
respect to density of A. planipennis larvae, there was no interaction between

treatment and tree height (F14, 859:1.34, P=0.178). The percentage Of total

larvae in the tree varied by height (F7'583=26.53, P<0.001) for all treatments; the
highest proportion Of larvae was consistently found 2-4 m above ground (Fig.
1.88). The interaction between height and treatment signiﬁcantly affected the
distribution Of larvae by height (F14, 53352.67, P=0.001). As in 2006, larvae on
girdled trees were more evenly distributed along the trunk and leader than on
control and methyl jasmonate trees (Fig 1.8B). The proportion of logs that had at
least one larva varied on the control and methyl jasmonate trees with height
(F14,359=2.67; P<0.001), but not on girdled trees (Fig 1.48). The highest
proportion of infested logs was from trunk sections 3-4 m above ground and the

lowest proportion Of infested logs were from the basal trunk section, within 1 m

above ground (Fig. 1.48). After larval mortality was accounted for, the density Of

13

A. planipennis larvae averaged 25.1 :I: 4.6, 19.0 i 7.3, and 111.2 :I: 9.8 larvae per
m2 on control, methyl jasmonate, and girdled trees, respectively.

Overall, 88.7% of larvae that began feeding reached the 4th instar or were

prepupal larvae when trees were felled. Larvae developed signiﬁcantly faster on

girdled trees than on control and methyl jasmonate trees (F2,18=18.07, P<0.001).

On average 69.0% :I: 4.2%, 64.1% :I: 4.8%, and 81.3% 1 3.4% Of larvae were 4th

instars or prepupae on control, methyl jasmonate and girdled trees, respectively.
Unlike in 2006, larvae below the girdle did not develop signiﬁcantly faster than

larvae above the girdle on girdled trees (T -value27=1 .92, P=0.244) (Fig 1.5).
Larval development did not vary among canopy exposure Classes (F4,54=2.24,
P=0.077) and development rate was not signiﬁcantly related to larval density

(F2,36=O.39, P=O.676) (Fig. 1.9). One data point appeared to be an outlier and

was removed from the analysis, but this did not alter the result (F2,36=1.05,

P=0.356).
Galleries formed by larvae that had reached the prepupal stage averaged
3.5 1: 0.1 cm wide and 8.6 :I: 0.3 cm long. Similar to 2006, there was a signiﬁcant

correlation between gallery width and length, but this relationship only accounted

for a small amount of the variation (F1,33=7.03; P=0.009; r2=0.08). Gallery width

did not vary by treatment (F2.11=1.96; P=0.187). As in 2006, larvae that
developed below the girdle on girdled trees had galleries that were signiﬁcantly

wider than larvae above the girdle (F3.27=78.92; P<0.001) (Fig. 1.7). Similarly,

14

mean gallery length was longer below the girdle than above, with a length of 15.5

:I: 1.5 cm and 7.3 :I: 0.1 cm, respectively.

15

DISCUSSION

Adult and larval distribution

A. planipennis adults, including ovipositing females, were more attracted
to girdled trees than to control or methyl jasmonate trees. These ﬁndings are
similar to those from other ﬁeld studies of A. planipennis and Other buprestids.
Agrilus bilineatus (Weber) (two-lined Chestnut borer) (Haack and Benjamin 1982,
Dunn et al. 1986a, 1986b), Agrilus anxius (bronze birch borer) (Barter 1957), and
Agn'lus burkei (ﬂatheaded borer) (Svihra and Koehler 1993), are attracted to
declining oaks (Quercus sp.), birch trees (Betula), and alders (Alnus),
respectively. A meta-analysis found that wood boring insects were attracted to
trees stressed by drought at higher rates than vigorous trees (Koricheva and
Larsson 1998). A. planipennis follows this pattern in its native range China,
attacking primarily stressed trees (Akiyama and Ohmomo 2000, Schaefer 2005,
VIﬁlliams et al. 2005, 2006). Field studies in North America using girdled trees tO
attract A. planipennis adults have produced similar results; girdled trees attracted
signiﬁcantly more adults than non-girdled trees (Poland et al., 2004, 2005,
McCullough et al. 2006, Fraser and Mastro 2007, Anulewicz et al. 2007,
Anulewicz et al. 2008b). However, these differences are obscured in sites with
high densities of A. planipennis (McCullough et al. 2009 submitted). In this
‘ study, the relationship between the number of adults captured and the density of

larvae was weak and highly variable.

16

Adult capture rates did not vary between control and methyl jasmonate
trees. An earlier trap tree study also found that girdled trees captured
signiﬁcantly more larvae than either control or methyl jasmonate trees, which did
not differ from each other (McCullough et al. 2006). While methyl jasmonate had
no affect on A. planipennis distribution or development in this study, it can induce
stress responses in other plants. Methyl jasmonate induced volatile emissions in
cotton plants (Rodriguez-Saona et al. 2001) and ash seedlings in a laboratory
setting (Poland et al. 2006). Subsequent tests showed A. planipennis adults
were attracted to volatiles produced by ash seedlings treated with methyl
jasmonate (Poland et al. 2006). Methyl jasmonate generates stress responses in
conifer saplings by activating the jasmonic pathway (Hudgins and Franceschi
2004). It is possible that the effect Of methyl jasmonate is minimal on large trees,
or the dose used was too low to elicite a stress response. Many Previous studies
using methyl jasmonate generally involved treating seedlings or small,
herbaceous plants while the trees in this study were about 10 m tall. Aerts et al.
(1994) found that methyl jasmonate vapor increased the synthesis Of protective
alkaloids in the herbaceous plant Catharanthus roseus L. G. Don, but only while
the plant is a seedling. Henery et al. (2007) found that methyl jasmonate did not
increase the production of defensive chemicals in 15-month-Old Eucalyptus
grandis Hill ex Maiden trees. Further studies will be needed to determine why
methyl jasmonate had no effect.

Overall, adult and larval density increased by a factor of ﬁve from 2006 to

2007. This increase has been observed at other sites where adult A. planipennis

l7

captures increase between three- to ten-fold between 2006 and 2007 (Anulewicz
let al. 2008). There was a drop in the proportion of captured adults that were
female from about 75% in 2006 to about 55% in 2007. Other trapping studies
have not found similar divergences in sex ratios (McCullough et al. 2009). The
sex ratio in 2006 may have been due to random chance, or more females were
captured on the sticky bands because they were ovipositing.

McCullough and Siegert (2007) reported that the average density of A.
planipennis in killed trees varied between 105 adults per m2 on trees with a DBH
Of 313 cm and 69 adults per m2 on trees with a DBH of <13 cm. Similar results
were found in 2007 on girdled trees after taking mortality into account. Larval
densities on non-girdled trees were lower than the adult yield reported by
McCullough and Siegert (2007). The control and methyl jasmonate trees did not
appear to be declining when they were felled, and would certainly have produced
additional A. planipennis adults in later years. Larval densities were also lower
on girdled trees in 2006 than the adult yield reported by McCullough and Siegert
(2007). While the girdled trees in 2006 would not likely survive for a second year
to be infested again because Of the artiﬁcial wounding, the phloem was probably
under-utilized by A. planipennis larvae due to the low density that existed that
year.

While adult catches and larval density did not differ due to canopy
exposure in this study, previous studies have shown that buprestids are attracted
to trees exposed to sun or in open areas (Anulewicz et al. 2007, Wermelinger et

al. 2007). Anulewicz et al. (2007) found about 90% Of all captured A. planipennis

18

adults were on trees fully exposed to the sun or exposed on three of four sides.
Wermelinger et al. (2007) found that buprestid species such as Anthaxia nitidula
(Linné), Agrilus Olivicolor Kiesenwetter, Agrilus vin'dus (Linnaeus), Agn'lus
convexicollis Redtenbacher, and Agrilus Iaticollis Kiesenwetter prefered edge
habitats to the interior Of the forest. The plantation in my study was small and
intermixed with canopy gaps. Sunlight was able to penetrate most Of the stand
and likely obscured effects from canopy exposure rank that were found to be
signiﬁcant in other studies.

Larval densities were signiﬁcantly lower near the top Of the tree than in
lower portions of the tree. Previous studies found no differences in buprestid
adult capture by height (Wermelinger et al. 2007). McCullough et al. (2006)
placed sticky bands at the height of 1.5 m, and at 3-5 m above ground in trap
trees, as well as sticky panel traps in the canopies. They found no differences in
adult capture of A. planipennis adults at different heights on trap trees. The
proportion Of logs infested varied by height for control and methyl jasmonate
treated trees; signiﬁcantly higher proportion of logs from the trunk between 2-5 m
were infested compared with logs from other heights within trees. Differences in
larval densities by height were not as pronounced on girdled trees. This may be
because the high population density of A. planipennis larvae utilized more Of the
phloem on girdled trees. Additionally, more larvae were found on the lowest one-
meter log from girdled trees than on one-meter logs from control or methyl
jasmonate treated trees. This may have been more attractive to A. planipennis

adults, therefore leading to higher oviposition. Larval survival may also be higher

19

on below the girdle. Haack and Benjamin (1982) found that A. bilineatus adults
were captured at seven times the rate on girdled oaks compared with controls
and that A. bilineatus larva were also found at higher densities on oak trees that
had lower reserves Of stored starch. Starch reserves in oak trees have been
found to be lower after defoliation events, drought, and on trees infested with
high levels of A. bilineatus (Haack and Benjamin 1982). Haack and Benjamin
(1982) suggested that lower starch levels either led to higher levels of infestation,
were the result of larval girdling, or a mix Of the two. A. planipennis found below
the girdle might have developed faster because the girdle cut off the ﬂow Of
nutrients from leaves down the tree, creating a lower level of starch.

The trees used in this experiment were similar in size tO trees used as
detection trees in surveys for A. planipennis adults in Michigan and other states
(Flint 2005, Storer et al. 2005, Harrison 2005). While trees used in this study
were in a plantation, and detection trees are commonly located along roads,

patterns of larval distribution are probably similar.

Larval Development

Overall, treatment was the primary factor affecting the larval development
of A. planipennis in both 2006 and 2007, with larvae developing faster on girdled
trees than on control and methyl jasmonate trees. Additionally, in 2006, larvae
below the girdle developed signiﬁcantly faster than larvae found above the girdle.
Larvae found above the girdle in 2006 did not develop at signiﬁcantly different

rates than larvae on control and methyl jasmonate trees. Furthermore, larval

20

development rates do not appear to be heavily inﬂuenced by larval density. Even
though a higher proportion Of larvae reached the fourth instar or prepupae below
the girdle than above the girdle, the larval density was lower below the girdle
than above the girdle.

While A. planipennis larval density did not signiﬁcantly affect larval
development at the scale Of the individual tree, there may be a relationship
between larval density and development at a larger scale. The number of adults
and larvae in the entire plantation increased by a factor of ﬁve from 2006 to 2007
which coincided with a ﬁve-fold increase in the proportion of larvae that reached
4th instar or the prepupal stage. This pattern has been found in other ﬁeld
studies as well. Low density sites had large numbers of ovenrvintering early
instar, while larvae only occasionally overwintered as early instars in moderately
to heavily infested sites (Cappaert et al. 2005b).

The paradox Of different rates of larval development at the scale of the
individual tree and the entire site might be explained by variables associated with
the hosts or by changes in adult A. planipennis behavior at different population
densities. Individual trees might vary genetically or inhabit a different
microclimate, which may cause some trees to be under more stress, and
therefore more susceptible to attack. A. planipennis adults might also behave
differently in high density populations compared with low density populations.
For example, Wallin and Raffa (2004) found that host preferences Of adult spruce
beetles, Dendroctonus ruﬁpennis (Kirby), attacking Picea glauca (Moench) and

Picea engelmannii (Parry and Engelm), differed between endemic and epidemic

21

populations. D. ruﬁpennis in low population densities were found to only attack
felled trees and D. ruﬁpennis in high population densities attacked vigorous hosts
as well (Wallin and Raffa 2004). A. planipennis might exhibit a similar adult
density—dependent behavior, with larvae produced by adults in high adult density
populations developing faster than larvae found in low adult density populations.
Further studies will be needed to determine why larval development appears to
be affected by larval density at the scale of the plantation but not at the scale Of
the individual tree.

Average gallery width was dramatically greater below the girdle, matching
the faster larval development rate. Galleries found above the girdle and on non-
girdled trees formed the standard serpentine pattern, while galleries found below
the girdle did not follow this pattern. Larvae below the girdle usually fed in
random directions. This pattern was also found by Barter (1957) in A. anxius
larvae on birch trees. Barter (1957) speculated that larvae form serpentine
galleries to help slow cambial activity, which may interfere with larval
development. By blocking the downward ﬂow of nutrients or defensive
compounds, the gallery functions like a small girdle. It would be interesting to
determine if A. planipennis larvae feed downwards more commonly than they
feed upwards, and if feeding direction is affected by girdling. Future studies are
needed to elucidate the effects that damage from artiﬁcial wounding and from A.
planipennis feeding has on the nutritional and chemical content Of phloem.

In North America, native buprestids are known to attack and kill stressed

and wounded trees at far higher rates than vigorous trees. Barter (1957) found

22

that A. anxius had survival rates Of only 7-25% on healthy birch trees and that all

surviving larvae took at least two years to develop. Barter (1957) also found that
as more larvae attacked a tree, with densities Of 5 and 12 larvae per m2, the

mortality decreased and about 1-5% of larvae developed in one year.

Drought and water stress may have played a role in our study.
Precipitation patterns varied during the ﬁrst four to six weeks Of larval feeding
between 2006 and 2007. Larvae typically begin to feed in mid to late July
(Cappaert 2005b). First instars were found in Jackson County on 17 July in 2006
(Chapter 4). Weather records for nearby Haslett, Michigan
(www.wunderground.com) indicated a much drier July in 2007 compared with
2006, with 13.6 cm, and 4.1 cm of rain falling in all of July and the ﬁrst two weeks
Of August in 2006 and 2007, respectively. Precipitation in the last two weeks of
August was 5.4 cm and 13.9 cm for 2006 and 2007, respectively. Therefore,
total rainfall was greater in 2007 compared with 2006, but it occurred later in the
season. For comparison, the 30-yr precipitation average for the months Of July
and August are 8.5 :l: 0.7 cm and 8.7 :I: 0.9 cm, respectively. Further studies are
needed to clarify whether short-tenn drought can affect the development of A.
planipennis. Studies Of other buprestids have found that water stress plays an
Important role in larval development Of buprestids. Trees that are stressed by
drought had higher densities Of buprestid larvae (Koricheva and Larsson 1998).
Mattson and Haack (1987) suggested that hosts may become more suitable
during droughts because plant nutrients become either more concentrated or

better balanced. Trees under drought stress have been known to form localized

23

zones Of cambial and phloem tissue degeneration (Santamour 1990). Water and
carbon manipulations using irrigation and girdling affect stern diameter growth in
Jug/ans nigra (black walnut) trees (Daudet et al. 2005). Tree diameter was
shown to grow at lower rates below the girdle and to slow dramatically during a
drought. Stem diameter growth was also shown to recover quickly after a
drought.

While no studies have been performed directly linking tree traits such as
bark moisture or callus growth to A. planipennis development, several studies
have been performed on other wood boring insects. Hanks et al. (1999) found
that moisture content of bark affects feeding behavior of Phoracantha
semipunctata (Fabricius) (Cerambycidae), with fewer larvae reaching the
cambium on logs with moist bark compared with logs that had dry bark in a
laboratory setting. Additionally, Hanks et al. (1999) found that P. semipunctata
larvae feeding on an artiﬁcial substrate survived at higher levels in dry compared
with wet conditions. Fierke and Stephen (2008) found that the number Of attacks
by Enaphalodes rufulus (red oak borer) larvae in the ﬁeld were not affected by
the level of bark moisture, but were affected by the level of callus overgrowth on
wounds. This was also found in A. bilineatus on oaks, with adults attacking trees
at higher rates that had lower levels Of callus overgrowth (Dunn 1990). It is
possible that A. planipennis development is responding in a similar fashion to
tree stress.

Climate was found to be the strongest determining factor Of whether A.

planipennis larvae in China would develop in one or two years. Speciﬁcally,

24

larvae required at least 150 frost-free days in order to develop (Wie et al. 2007).
Therefore, larvae took two years tO develop in the northern province Of
Heilongjiang 45.45 °N (Yu 1992), one year to develop in the more southerly
Liaoning Province 41.8 °N (Zhao et al. 2005), and either one or two years to
develop in the transitional area of Jilin Province (Wie et al. 2007). My site in
Clinton County is at 42.7 °N, corresponding to the transitional area in China
where larvae take 1-2 years to develop.

Similarly, development Of A. anxius has been found to vary by latitude. A.
anxius larvae reportedly take at least two years to develop In New Brunswick and
Quebec (Barter 1957) but one year to develop in Connecticut and New York.

The development Of A. anxius varies in Minnesota and Maine, taking one or two
years to develop depending on host vigor (Anderson 1944, Nash et al. 1951).
Barter (1957) found that even in northern locations, larvae could develop in one
year on plant tissue that was fairly moribund, including on girdled trees that had
been attacked repeatedly. To date, development of A. planipennis has not been
shown to vary with climate or latitude in North America. Both one and two-year
development has been found in sites in Michigan. It may be possible that as A.
planipennis spreads further north and south, climate may have a larger impact on

its larval development rate.

25

33de Bow 5 88 556 5 names, >8 Le Enigma a: was

26

as? A e. a $3 A 28 $3. A QB 85. an.» Io as 522
md A _..m v.0 A Nd Nd A 0N AAEV coEEmm mOLm Oomtzm cam—2
no A RN 3 A nu ed A 3.. Es A852,: .823 cams.

no A 2: to A S: no A 2: E £99. 8: :85.
no A RS 3 A Re no A ANA E3 1mg :85.
on on on 8e: .02
8e: Bow
$3. A Rev $3 A ANA $3 A 3: mean. can; to as ewes.
Nd A md Nd A RN to A v.m ANEV 336mm new Oomtzm com—2
ed A 0.2 no A RS no A A. 3 E 292 we. 58.2
no A RNA to A 2: ed A «d Ass :mo 582
on om om wow: .Oz
89. 8cm

 

8.30 9838238.). 6:80

 

Bow new moon 5 mesh 20:85...
ho 8:253: 3.5 Lo.— ..ug one 5 EEO—goo 02:2 be 23.8.. 2: new £35.35

802.5 60.: Lou OOEENO no.5 costs» .290: .750 ooh Em A. cues. ._. 032.

200
180

7:5 160
'3 140
120
100
80
60
4o
20

Larval Density

El Control

I Methyl Jasmon
I Above Girdle

l Below Girdle

 

2006 2007

2

Fig. 1.1 Mean (:I: SE) density Of A. planipennis larvae per m on control

and methyl jasmonate trees, and above and below the girdle on girdled

trees in 2006 and 2007. Within year means with the same letter are not
signiﬁcantly different (T ukey protected LSD test; P<0.05).

27

 

400

 

 

y= 54.056ln(x) + 4.9516
R2: 0.419 P<0.001

 

0)
0|
0

 

 

 

 

O
O

 

NNO)
co:
00
Cl

 

150 O/

Larval density per m2
8
o

 

 

C,"
O O
J I

 

A) 0 10 20 30 40 50
No. of Adult Beetles

 

 

y = 61 .884|n(x) + 2.4895 _.

o 0 R2: 0.4077 P<0.001
O

O

 

 

 

 

 

 

 

 

Larval density per m2
_. N
01 o
o o
O
968900
O
X

 

01
oo
GD

 

20 30 40 50
NO. of Adult Beetles

U)
v
O
.3
C

Fig. 1.2 Log1o(x+1) relationship between A. planipennis larvae per m2 and

total adult capture on sticky bands on trees in 2006 (A) and 2007 (B).

28

Fig. 1.3 A Mean (:1: SE) larval density by height in trees in 2006. 3 Mean (1 SE)
percent of larvae by height in trees in 2006. Means for different heights within Of
the same treatment with the same letter (a, b, and c) are not signiﬁcantly
different. Means within the same height of different treatments with the same
letter(x and y) are not signiﬁcantly different (T ukey protected LSD test; P <0.05).

29

3292..
6-4 Tm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.2.0 I
OE :oEmmq. _> 50: I r
.6580 U

 

 

 

 

 

 

 

 

 

65

$0

$3

$2..

e\eom

$94

aeiusl to wasted ueaw

$om

00
N

 

C
‘-

 

 

 

IInm LW
00:85 I

Be :oEmms $592 I

 

«I».
>11
1.—

.3806

 

 
 

l
0
no

I cow

0
CO
zuI .Iad Ausuap lame-I

 

 

_ a.

 

30

Fig. 1.4 Mean (:I: SE) percent Of sections by height that were infested with A.
planipennis larvae for 2006 (A) and 2007 (B). Means for different heights within
the same treatment with the same letter (a, b, and c) are not signiﬁcantly
different. Means within the same height Of different treatments with the same
letter(x and y) are not signiﬁcantly different (T ukey protected LSD test; P <0.05).

31

 

 

 

 

Height (m)

 

 

 

aeruel ;0 tuaaiad ueaw

Height (m)

60‘

.6o.ov n_ ”.8“ cm; 86205 >935 EOLOEO $56536 .oc mum Loco.
OEmm OE 5:5 mcmoE 36> :_£_>> Room ccm ooom 5 wow: 8.26 :o 0.25 05 323 Can 93% cam now: camcoEwﬂ

359: new _o=coo co BEE Ev “mac. 8 venues. um: :65 6636. 9:69:63 .< *0 33588 Gm Av cams. m4 .9".

 

        

 

 

 

 

 

Room meow
- .B a
- :8 m
- <8 m
. w
- *8 m.
- e\eov w.ﬂ
- e O M
\oom w a
6.25328- . $8 M p
6.266665- #2 m
96856625623 $8 A6”
a _ 65:000. *8 W.

 

$09.

33

Fig. 1.6 The log1o+1 relationship between the percent Of 1-yr larvae and larvae
per m square on control (A), methyl jasmonate (B) and girdled (C) trees in 2006
(GLM; Only girdled trees were signiﬁcant P <0.05).

34

100%
90% y = 0.0855ln(x) + 0.0453
80% I"?2 = 0.1644
70%
60%
50%
40%
30%
20%
10%

0%
A) 0 50 100 150 200 250

 

100%
90% y = 0.0287|n(x) + 0.0678
30% R2 = 0.0245
70%
60%
50%
40%
30%
20%
10%

0%
B) 0 50 100 150 200 250

 

100%
90%

Percentof larvae reaching at least the 4th instar

 

80%
70%
60%
50% CI
40%
30%
20% y = 0.121ln(x) 4' 0.0351
1 0% R2 = 0.3178
0%
C 0 50 100 1 50 200 250
) Larvae per m2

35

 

.3
.5

El Control b
I Methyl Jasmon ate
IAbove girdle
l Below Girdle

 

.3
N

 
   

   

 

.3
O

 

 

 

 

 

 

 

Mean horizontal gallery
dimenslon (cm)

 

 

 

 

 

 

GIG-503m

Fig. 1.7 Mean (:I: SE) width of A. planipennis prepupal galleries on control and
methyl jasmonate trees, and above and below the girdle on girdled trees in 2006
and 2007. Within year, means with the same letter are not signiﬁcantly different
(T ukey protected LSD test; P<0.05).

36

Fig. 1.8 A Mean (:I: SE) larvae density m2 by height in tree in 2007. B Mean (:I:
SE) percent Of total larvae by height in tree in 2007. Means for different heights
within the same treatment with the same letter (a, b, C, and d) are not signiﬁcantly
different. Means within the same height Of different treatments with the same
letter(x and y) are not signiﬁcantly different (T ukey protected LSD test; P <0.05).

37

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

      

       

 
   

 

 

 

     

 

 

 

  

 

 

 

  

 
 

 

 

 

 

 

 

 

 

 

 

€3.96... Am
3. Wm 9m 3. I” m-~
«co
m
$9. a
U
a
$om m
m
«.8 m.
86.5. L a mi.
2m:oEm2..>£O_2I Sow e
69:86 a
$8
.5296: A
m.» he 3 m4 Tm «A . w.-. To <
o
I I
II- II- x x x .I
coax m on e
x x 3 A an n 2: m
O n n > M.
Ba on? m
> s
8 8 com M
> OBEOI I 0mm .m
I...“ > A 922.8236:- 4 8.4. w...
e a .eEcoOO

 

 

 

 

0mm

38

Fig. 1.9 The log1o+1 relationship between the percent of 1-yr larvae and larvae
per m square on control (A), methyl jasmonate (B) and girdled (C) trees in 2007
(GLM; No relationships were signiﬁcant, P <0.05).

39

 

1 00%

Y = 0.0406|n(x) + 0.5653

T

 

o 1
90/0 DD

'-' D U R== 0.0486

 

 

 

80%

 

I /.~—~——-T—
70%

 

60%

 

50%

 

40%

ED

 

30%

 

20% Q

 

10%

 

 

0% ["1

A)0

1 00 200 300 400

 

100%

 

Y = 0.1049|n(x) + 0.3797

 

90%

 

 

 

._. R2= .1759
80% U/ 0

 

70%

 

 

60%

 

50% -

 

40%

 

30% '3

 

20%

 

10%

 

 

0% "”2

B)0

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%

0%

Percentof larvae reaching at least the 4th instar

 

100 200 300 400

Y = 0.0141In(x) + 0.8577
R2= 0.0049

100 200 300 400

Larvae per m2

40

CHAPTER 2

Mortality of Agrilus planipennis larvae in green ash trees

INTRODUCTION

Like other insects, Emerald ash borer, Agrilus planipennis Fairmaire
(Coleoptera: Buprestidae), larvae can die from a variety Of causes including
woodpecker predation, pathogens, cannibalism, and parasitism. After the
discovery Of A. planipennis in southeastern Michigan in June 2002, about ten
years after its introduction from northwest China, studies were conducted to
determine the types and rate of mortality. Bauer et al. (2004) found that less
than 2% of larvae were infected with entomopathogenic fungi. Cappaert et al.
(2005b) and Lindell et al. (2008) reported that woodpeckers consume highly
variable proportions of larvae; e.g., ranging from 0-95% of larvae (Cappaert et al.
2005b). Lindell et al. (2008) found that native woodpeckers have responded to
new infestations of A. planipennis by spending more time foraging on ash trees
than they did before the trees became infested.

Parasitoids that attack A. planipennis larvae and eggs have been
identiﬁed in China and in North America (Yang et al. 2005, 2006, Bauer et al.
2005, Bauer and Liu 2006, Liu and Bauer 2007a, Gould et al. 2007, Liu et al.
2008, Cappaert and McCullough 2008). The parasitoids Oobius agn'li Zhang and
Huang, and Tetrastichus planipennisi Yang were estimated to reduce the total

population Of A. planipennis by an estimated 73.6% in four sites near Changchun

41

city, in the Jilin Province Of China (Liu et al. 2007b). Trial releases of these
parasitoids have now occurred in Michigan. Parasitoids native to North America
infest less than 1%, Of larvae and eggs (Bauer et al. 2004).

A comprehensive survey of A. planipennis larval mortality in a population
in North America has not yet been performed. Data on the factors and rates Of
larval mortality would be useful for researchers developing models Of A.
planipennis rate Of spread and population dynamics and for control or
management efforts.

This study had four Objectives including (1) identify the factors causing A.
planipennis larvae to die; (2) determine the mortality rates for each factor; (3)
determine whether host vigor affects mortality factors or rates; and (4) determine

whether mortality factors and rates vary with A. planipennis larval density.

42

MATERIALS AND METHODS

For this study, I used the green ash, Fraxinus pennsylvanica Marshall
(Oleaceae), trees at the Rose Lake plantation described in Chapter 1. As
explained in Chapter 1, ten blocks of 9 trees each (total Of 90 trees) were
selected and treated in spring 2006, then felled and debarked the following
winter. In 2007, an additional ten blocks of 9 trees each were similarly selected,
treated, then felled and debarked the following winter. Trees within each block
were randomly assigned to one of three stress treatments each year. Trees
were either girdled (30 trees), exposed to the stress-elicitor methyl jasmonate (30
trees), or left as untreated controls (30 trees) as described in Chapter 1.

The number and instar of each A. planipennis larvae on each log were
recorded. Larval density was standardized per m2 Of exposed surface area per

tree as in Chapter 1. Developmental instar was determined by head capsule
width (Cappaert et al. 2005b). When a larva was missing from its gallery, instar
was estimated based on the gallery size compared to galleries constructed from
larvae of known instar. The month in which blocks Of trees were felled was
recorded.

Larval mortality caused by woodpecker predation, unknown causes, and
cannibalism was recorded for each log. A. planipennis larvae preyed on by
woodpeckers were identiﬁed by matching holes made by woodpeckers with
galleries Of removed larvae. Mortality from unknown causes was recorded when

the larva apparently died from disease or desiccation but was still present in the

43

gallery. Cannibalism was previously deﬁned by Elgar and Crespi (1992) as the
killing and consumption of intraspeciﬁc individuals. A larvae was considered to
have been cannibalized when either; a remnant of a larva was discovered with a
gallery from another larva intersecting its body or when a gallery abruptly ended
at another gallery and the associated larva was absent.

I assumed that larvae ovenrvintering as 1“, 2"d or 3rd instars would
probably not reach the prepupal stage by the spring, and would require another
summer Of feeding, resulting in a two year life cycle (Cappaert et al. 2005a).
Larvae overwintering as 4th instars or as prepupae would likely emerge in the
spring as an adult, resulting in a one year life cycle (Cappaert et al. 2005a).
Larvae hatching in 2006 but requiring two years tO develop would not contribute
to the adult population in 2007. Therefore, larvae with delayed development
would not reproduce in the subsequent year and were effectively “lost” from the
current years’ cohort. A few larvae were found in 2006 that had hatched and
begun developing during the summer of 2005. Larvae that required two years to
develop were identiﬁed by gallery appearance and callus tissue that had begun
to grow over the Older section of the gallery (Cappaert et al. 2005b, Siegert et al
2007). These larvae were recorded separately from larvae that hatched and
began developing in 2006.

The 2006 study was replicated with trees felled and debarked trees from

October 2007 through February 2008.

44

Statistical Analysis

All data were tested for normality using the Shapiro-Wilk test (Shapiro and
WIlk 1965) and residual plots. Three response variables were evaluated
Including the proportion Of larvae fed upon by woodpeckers, the proportion Of
larvae that died due to unknown causes, and the proportion of larvae that died

due to cannibalism. Larval density was used as a covariate and required a
log1o(x+1) transformation to achieve normality.

All response variables were analyzed using the GLIMMIX procedure for
mixed models in SAS statistical software (PROC GLIMMIX; SAS Institute 2003).
Effects Of three factors: tree vigor with three levels (control, methyl jasmonate, or
girdled), canopy exposure rank with ﬁve levels (1, 2, 3, 4, 5), and height in tree
with eight levels (one m increments abOVe ground) were analyzed. Block and
block crossed with treatment were used as random factors. Differences among
treatment means were tested as unplanned comparisons and multiple
comparison tests were applied only when overall analysis Of variance (ANOVA)
was signiﬁcant (P<0.05). When signiﬁcant differences occurred among levels,

the Tukey protected least signiﬁcant difference (LSD) test was used to evaluate
those differences (Ott and Longnecker 2001). The log1o relationship between
larval density and the three Classiﬁcations Of mortality was analyzed using a
general linear model with a log1o(x+1) transformation (PROC GLM, SAS institute

2003).

45

Life-tables

Life-tables were constructed using A. planipennis larvae recorded when
trees were debarked each year. The total number Of larvae found was used as
the initial cohort for the population of A. planipennis for that year. My study
focused on A. planipennis larvae and did not attempt to quantify eggs. A.
planipennis adults oviposit in bark crevices or under bark ﬂaps, making detection
difﬁcult on trees with ﬂaky periderm (Anulewicz et al. 2008a). The apparent loss,
or age-speciﬁc mortality, was a measure of the mortality within a life stage.
Apparent loss was calculated by dividing the total number of dead individuals for
each stage by the total number Of larvae in that life stage. The real mortality, or
real loss, measured the total contribution Of mortality at a life stage to the total
population. Real mortality was calculated by dividing the total number Of dead
individuals for each stage by the total number of larvae found. Live larvae
ovenrvintering as 1"", 2”“, or 3rd instars were counted as a loss for the adult
population of the following year but were added to the potential adult population

for the following year.

46

RESULTS

2006 Study
Mean density Of A. planipennis larvae was 34.5 :t 5.8 larvae per m2 per
tree and at least one larva was found on 24, 21, and 30 Of the control, methyl

jasmonate, and girdled trees, respectively. Larval density averaged 3.7 :I: 0.9,
6.7 :I: 4.4, and 71.1 :t 10.9 larvae per m2 on control, methyl jasmonate, and

girdled trees respectively. Larval density was signiﬁcantly higher on girdled trees
than control and methyl jasmonate trees (F2,13=85.02, P<0.001).

Out of 4,621 larvae recorded in 2006, 1438, or 30.1%, of all larvae died.
On average, 17.1% :I: 2.1% of the larvae per tree died (Table 2.1). Mortality

varied by vigor treatment, with a slightly higher proportion Of larvae dying on
girdled trees than control and methyl jasmonate trees (F2,13=5.62, P=0.013).

Average larval mortality was 10.3% :t 2.9%, 15.0% :I: 3.9% and 24.0% :I: 3.0%
dying on control, methyl jasmonate, and girdled trees, respectively. The
proportion Of apparent mortality, or the total loss within an instar, was highest 4th
instars; overall, 9.7%, 36.9%, 71.6%, and 30.4% of 1St and 2"“, 3'“, 4th instars,
and prepupae died, respectively (Table 2.1). The real loss, or the total loss in the
population, by instar was highest at the 4th instar and the prepupae stage; 3.0%,
6.5%, 10.3%, and 11.4 Of 1“ and 2"“, 3'”, 4th instars, and prepupae died,
respectively (Table 2.1).

Woodpecker predation caused the highest larval mortality, accounting for
9.9% :I: 2.0% of all larvae. A signiﬁcantly higher proportion Of woodpecker

predation occurred on girdled trees than on control or methyl jasmonate trees

47

(F2,13=16.30, P<0.001). On average, woodpeckers killed 4.8% :I: 2.2%, 2.6% :I:
1.3%, and 19.1% :I: 2.7% of larvae on control, methyl jasmonate, and girdled

trees, respectively. A signiﬁcant log1o(x+1) relationship existed between larval
density and woodpecker predation (F1,4o=21.82, P<0.001) (Fig. 2.1).
Woodpeckers preyed on 4th instars at signiﬁcantly higher rates than prepupae,
3'd instars, and 1"“t and 2"“I instars (F3,175=18.80, P<0.001) (Table 2.1). Height in

tree signiﬁcantly affected woodpecker predation rates (F7,230=7.42, P<0.001).

More predation occurred near the top Of the tree than near the bottom of the tree

(Fig. 2.2). The proportion of larvae killed by woodpeckers was not affected by

month (F3i42=0.82, P=0.492) (Fig. 2.3). Canopy exposure rank also did not

signiﬁcantly affect woodpecker predation (F3,42=2.92, P=0.045).
Larval mortality from unknown causes accounted for 6.9% :I: 1.5% of all
larvae and did not vary by vigor treatment (F2'18=2.62, P=0.100). The proportion

of mortality from unknown causes averaged 5.5% :I: 1.9%, 9.2% :t 2.6%, and
4.3% :I: 0.9% Of larvae on control, methyl jasmonate, and girdled trees,

respectively. On girdled trees, however, larvae died at signiﬁcantly higher rates
from unknown causes below the girdle than above the girdle (F3,27=3.21,

P=0.003). The proportion Of mortality from unknown causes averaged 3.9% i

1.2% above the girdle and 12.3% :I: 4.3% below the girdle. Larval density and
mortality from unknown causes were not signiﬁcantly related (F1,42=2.84,
P=0.100). Larvae in the prepupal stage died due to unknown causes at

signiﬁcantly lower rates compared to 4‘“, 3'“, and 1"t and 2"d instars (F3'171=5.18,

48

P=0.002) (Table 2.1). Height in tree (F7.280=1.27, P=0.267), month of felling

(F3_42=1.04, P=0.383) and canopy exposure rank (F3,42=1.92, P=0.141) did not
signiﬁcantly affect mortality from unknown causes.

Cannibalism killed only 0.3% :I: 0.1% of all larvae and did not signiﬁcantly
vary among vigor treatments (F2.13=1.28; P=0.302). The proportion of

cannibalized larvae averaged 0.0%, 0.2% :I: 0.1%, and 0.6% :I: 0.2% on control,

methyl jasmonate, and girdled trees, respectively. Larval density and

cannibalism had a weak but signiﬁcant relationship (F1,4o=12.59, P=0.001) (Fig.
2.4A). Rates of cannibalism did not vary signiﬁcantly by instar (F2I114=1.69,
P=0.190), height in tree (F6,274=1.96, P=0.072), month Of tree felling (F3.4o=1.65,

P=0.192) or canopy exposure rank (F3,4o=1.76, P=0.169).

Mortality from parasitism or pathogens was not Observed. While mortality
from unknown causes could have been caused in part by pathogens, external
symptoms were not apparent. We did not conﬁrm the presence Of pathogens in
the larvae.

A total Of 1791, or 38.8%, Of live larvae from the 2006 cohort had not
reached the 4th instar or the prepupal stage by winter and were expected to
require an additional year to develop (Table 2.1). We also recorded 81 larvae
that hatched and began feeding in 2005. These individuals would contribute to
the adult population in 2006 and were included in the life table (Table 2.1).
Woodpeckers predated 67.9% of the 81 larvae from 2005 larvae in 2006 (Table

2.1).

49

2007 Study
Mean density Of A. planipennis larvae was 79.6 :I: 9.8 larvae per m2 per
tree and at least one larva was found on 29, 29, and 30 Of the control, methyl

jasmonate, and girdled trees, respectively. Larval density averaged 38.4 t 10.4,
25.8 :t 9.7, and 174.5 1: 14.3 larvae per m2 on control, methyl jasmonate, and

girdled trees respectively. Larval density was signiﬁcantly higher on girdled trees
than control and methyl jasmonate trees (F2,13=30.87, P<0.001).
Out Of 18,891 larvae recorded, 6738, or 35.7%, of all larvae died; average

mortality was 34.2% :I: 2.6% per tree (Table 2.2). Mortality did not vary
signiﬁcantly by vigor treatment (F2,13=0.30, #0745). Larval mortality averaged

33.3% :I: 4.3%, 31.5% :I: 4.9%, and 34.0% :I: 3.5% on control, methyl jasmonate,
and girdled trees, respectively. The proportion of apparent mortality, or the total
loss by instar, was highest for 4th instars; 44.8%, 69.7%, 88.0%, and 21.1% of 1“t
and 2"“, 3'“, 4th instars, and prepupae died, respectively (Table 2.2). The real
loss, or the total loss in the population, was highest 4th instars and prepupae.
Overall, 1.1%, 5.8%, 13.3%, and 15.7% of 1“t and 2"“, 3'“, 4th instars, and
prepupae died, respectively (Table 2.2). This level Of stage-speciﬁc mortality
was similar to that observed in 2006.

Woodpecker predation accounted for most Of the larval mortality. Overall,

30.1% :I: 2.5% Of all larvae were preyed upon by woodpeckers. Unlike 2006,
woodpecker predation did not vary by vigor treatment (F2,13=0.17; P=0.843).

Woodpecker predation averaged 29.9% :I: 4.2%, 29.5% :I: 5.1% and 28.2% :I:

3.2% Of larvae on control, methyl jasmonate, and girdled trees, respectively.

50

Larval density and the rate Of woodpecker predation were not signiﬁcantly related
(F1,51=0.57, P=0.454). Woodpeckers killed 4th instars at higher rates than
prepupae and 3rd instars, which were predated at higher rates than 1"t and 2nd
instars (F3'21g=70.18; P<0.001) (Table 2.2). As in 2006, a signiﬁcantly higher
percentage of larvae were consumed by woodpeckers near the top of the tree
compared to larvae near the bottom of the tree (F7'5g4=28.04, P<0.001) (Fig.
2.2). Woodpecker predation rates were signiﬁcantly higher on trees felled in late

winter, compared with trees felled in fall (F4_51=9.83, P<0.001) (Fig 2.3). Canopy

exposure rank (F4'51=0.43, P=0.789) did not signiﬁcantly affect woodpecker

predation.

Mortality from unknown causes accounted for 3.3% i 0.4% of all larvae
and did not vary by vigor treatment (F2,13=0.45, P=0.643). As in 2006, mortality

from parasitism or pathogens was not Observed. On average, of mortality from
unknown causes killed 3.3% :I: 0.7%, 2.0% :I: 0.6%, and 3.6% :l: 0.7% of all larvae
on control, methyl jasmonate, and girdled trees, respectively. However, as in

2006, larvae died at signiﬁcantly higher rates from unknown causes below the
girdle than above the girdle on girdled trees (F3.27=7.34, P=0.001). Larval

mortality above the girdle averaged 2.6% :I: 0.4% and 11.1% :I: 2.3% below the

girdle. Larval density and mortality from unknown causes were not signiﬁcantly

related (F1,51=1.74, P=0.193). Prepupal larvae died from unknown causes at
signiﬁcantly lower rates compared to 4th and 3rd instars, which died at
signiﬁcantly lower rates than 1St and 2nd instars (F3,229=17.34, P<0.002) (Table

51

2.2). Height in tree (F7.562=1.97, P=0.057), month of tree felling (F3,51=0.86,

P=0.496), and canopy exposure rank (F3,51=0.07, P=0.990) did not signiﬁcantly

affect mortality from unknown causes.
As in 2006, mortality from cannibalism was rare, accounting for 0.8% :I:
0.3% of all larvae. Cannibalism did not signiﬁcantly vary with vigor treatment

(F2'13=1.50; P=0.249), averaging 0.1% :I: 0.0%, 0.0% :I: 0.0%, and 2.1% :I: 0.7% Of

all larvae on control, methyl jasmonate, and girdled trees, respectively.
Cannibalism did occurr at signiﬁcantly higher rates below than above the girdle

on girdled trees, with 9.2% :t 2.6% and 0.2% :t 0.1% of larvae cannibalized below

and above girdle, respectively (F3,27=7.44, P=0.001). Larval density and

cannibalism had a weak, but signiﬁcant relationship (F1.53=10.32, P=0.002) (Fig.
2.4A). Cannibalism occurred at signiﬁcantly higher rates during 1’3t and 2"d instars

than during 3rd and 4th instars (F2,177=8.35, P<0.001) (Table 2.3). Rates of
cannibalism did not signiﬁcantly vary by height in tree (F7,567=0.74, P=0.639),

month Of tree felling (F3,53=0.45; P=0.771) or canopy exposure rank (F4,53=1.17;
P=0.149).

A total Of 697, or 3.7% of the total 2007 cohort, ovenrvintered as 1"“, 2"“, or
3'“ instars (Table 2.2). We also recorded 710 larvae that began feeding in 2006
and would have contributed to the adult population in 2007 (Table 2.2).
Woodpeckers predated 43.4% Of the 710 larvae that had begun feeding in 2006

(Table 2.1).

52

DISCUSSION

Overall, 30.1% of the total larval cohort died in 2006 and 35.7% of the total
larval cohort died in 2007. Mortality at the prepupal stage accounted for most the
total mortality, followed by fourth instars and then third instars. Woodpeckers
accounted for most of the mortality, followed by unknown causes and
cannibalism.

Woodpecker predation was the largest source of A. planipennis larval
mortality in both 2006 and 2007. Woodpeckers preferred 4th instars and
prepupas, probably because these stages are always more than three times the
size of smaller instars. A much larger proportion Of the total 4th instars were
predated compared with prepupae, so it appeared that woodpeckers focused on
4th instars compared with prepupae. Predation increased after the month Of
November, and continued through the winter. Because all trees were felled by
March and February in 2006 and 2007, respectively, additional predation by
woodpeckers could have occurred before adults emerged in the spring.
Cappaert et al. (2005b) also found that woodpecker predation caused higher A.
planipennis larval mortality than any other factor. Data collected from 24 sites in
MI showed that woodpecker predation ranged from 9% tO 95% Of all larvae,
however, no reason was given for this variation (Cappaert et al. 2005b). Lindell
et al. (2008) found that woodpeckers predated about 30% of all larvae when
comparing woodpecker attack sites and A. planipennis exit holes on ash trees.

The rate Of woodpecker predation was signiﬁcantly related larval density.

A higher proportion of larvae were predated by woodpeckers in 2007 (30.1% of

53

all larvae) than in 2006 (9.9% of all larvae), coinciding with a ﬁve-fold increase in
larval density. Similar correlations between the woodpecker predation rates and
larval density were reported in previous studies on woodboring insects (Koplin
and Baldwin 1970, Lindell et al. 2008). In a meta-analysis of predation of spruce
bark beetle larvae (Dendroctonus sp.) by downy woodpeckers (Picoides
pubescens), a common woodpecker in our study site, predation did not vary
between high density epidemic populations and low density non-epidemic
populations (Fayt et al. 2005).

A larger proportion of larvae were predated near the top (58 m) of the tree
compared with the rest Of the tree. This pattern is different the distribution of
larvae found in the tree, with larval density highest between 3-4 In (Chapter 1).
This pattern suggests that woodpeckers base feeding preferences on height, and
do not focus their attention solely on larval density within an individual tree.

Canopy exposure rank did not affect rates Of woodpecker predation at this
site and Lindell et al. (2008) reported a similar result. They found a signiﬁcant
negative correlation between the proportion of forest cover and predation rates,
but forest cover explained little Of the variation in woodpecker predation
(R2=0.024).

Multi-year development did not enable to larvae to avoid woodpecker
predation. About 50% Of larvae that had begun feeding the previous year were
attacked by woodpeckers, as prepupae, compared with about 25% of the

prepupal larvae that began feeding in the current year.

54

Mortality from unknown causes was recorded infrequently in both 2006
(6.9% Of all larvae) and 2007 (3.3% Of all larvae) and most Of this mortality
occurred below the girdle on girdled trees and a higher proportion Of early instars
died than late instars. This was probably because the phloem below the girdle
was dying while larvae were feeding in mid-late summer. Barter (1957) found
higher mortality Of Agrilus anxius (bronze birch borer) larvae from unknown
causes on browning phloem tissue and early instars died at higher rates due to
unknown causes than later instars.

The reason that early instars die at higher rates than late instars is still
poorly understood. A meta-study by Zalucki et al. (2002) found that the most
commonly reported cause Of mortality of early Lepidoptera instars is unknown
factors. Because the exact cause of mortality is unknown, further research is
required to determine why earlier instars die at higher rates than later instars.
However, the fact that more larvae die below the girdle on girdled trees than
above the girdle creates a starting point for new studies. Moisture, nutritional, or
defensive compound content Of phloem may affect early instars more than late
instars. Early instars are probably more susceptible to the inner host
environment because they are small, and hence more susceptible to smaller
amounts Of defensive compounds, or less robust when encountering poor levels
of moisture or nutrition within the phloem.

While there was a difference in mortality from unknown causes below and
above the girdle, no signiﬁcant difference in mortality from unknown causes

occurred between larvae on control and methyl jasmonate trees. Methyl

55

jasmonate has been known to induce defense responses in other tree species,
including Picea abies (nonivay spruce) (Franceshi et al. 2002) and Pseudotsuga
menziesii (douglas ﬁr) and Sequoiadendron gn'ganteum (giant redwood)
(Hudgins and Franceshi 2004). Methyl jasmonate is also known to induce stress
volatile emissions in cotton plants (Rodriguez—Saona et al. 2001) and ash
seedlings in a laboratory setting (Poland et al. 2006, Rodriguez-Sacha et al.
2006). While ash species native to North America have no known constitutive or
inducible defenses against A. planipennis, Fraxinus mandshun‘ca (Manchurian
ash) in China experiences little A. planipennis - induced mortality (Rebek et al.
2007). A total of 11 potential defensive allelochemicals were found in higher
rates in F. mandshun'ca than in F. americana and F. pennsylvanica (Eyles et al.
2007). The lack Of allelochemicals found in American ash trees may explain the
lack of response to methyl jasmonate and the low rates of mortality from
unknown causes.

Cannibalism occurred at very low levels in both 2006 (0.3%) and 2007
(0.8%). This was different then the rates Of cannibalism that occur in
cerambycids. Ware and Stephen (2006) studied Enaphalodes mfulus (red oak
borer) in the laboratory with small phloem arenas; 84% of larvae exhibited
cannibalistic behavior, and cannibals gained ﬁve times the weight Of non-
cannibals. However, my results are consistent with my laboratory studies Of
larvae placed in small phloem arenas; 1.5% of A. planipennis larvae died due tO
cannibalism (Chapter 4). Previous studies of buprestids and cerambycids have

found that cannibalism generally occurs in high density populations. Cannibalism

56

was previously Observed in high densities of A. planipennis larvae (Bauer et al.
2004), but the exact proportions and densities were not reported. Additionally,
cannibalism was Observed in high density populations Of cerambycid
Phoracantha semipunctata (Hanks et al. 2005), but exact proportions and were
not reported.

A. planipennis cannibalism appears to be accidental, and is determined by
the probability that larvae will come in contact with each other. The rate Of
cannibalism in my study had a weak positive relationship to larval density, with
larvae having more contacts with other larvae in high population densities.
Additionally, galleries formed by larvae were larger below the girdle than above
(Chapter 1), leading to more contacts and signiﬁcantly higher proportions of
cannibalism occurring below the girdle than above the girdle on girdled trees in
both 2006 and 2007. Rates of cannibalism are higher below the girdle than
above even though larval population densities were smaller below the girdle than
above the girdle (Chapter 1). However, the rates Of cannibalism that I am
reporting may be underestimated at very high larval densities because individual
galleries became indistinguishable from each other, and therefore the presence
of cannibalism was harder to detect.

NO parasitism was Observed in this study. However, parasitoids are
known to attack the native bruprestid A. anxius (Barter 1957). Some of these
native parasitoids of buprestids, such as Heterospilus sp. and Atanycolus sp.
(Braconidae), Phasgonophora sulcata (Chalcidae), Balcha sp. and Eupelmis sp.

(Eupelmidae), and three ichneumonids, attacked less than 1% of A. planipennis

57

larvae (Bauer et al. 2004). Recently Atanyco/us sp., was reported to parasitize
A. planipennis larvae at much higher levels; in one site the average parasitism
rate ranged from 2% to 73% per tree (Cappaert and McCullough 2008). This and
other native parasitoids may become more important mortality factors for A.
planipennis in the future. If parasitoids become an important source of mortality,
additional research will be required to assess how parasitoids will interact and

compete with woodpeckers.

58

 

 

 

 

.263 .65
3 6936 .6... c. 66.... 6626. .33 .o 33:66.66 65 m. 662 .66. 6.7.3 3.3.2: 6566666036 62.66656. 52 €6.39...
exendo noun gndn. 2.3. .38.
$5.2. com .3335
.xN... exemNo mm 6626. SN 53666.6 65.666303
saw. 7 5- 6626. SN :5... 6662.6...
$1:
e\ev.o exec; 9 666366 565...:
$9 3 £6de mom 5.666... 6.66383
33 66:66....
«and? one .3325
$2. $0.? 3 66:69:50
$1? $3.. no 666366 5.65.5
$55 $0.06 3v c2666.“. 63.666682,
8mm .36... 5v
$6.2. new .3635
$0. 3 $0.8 mom 2.6.5.2656 66.6.60
exevd aid 5 Em..ma_cc60
aid £6. 3 mm 666366 :26...ch
oboe e\eo.m~ mow con—666.6 60.666683
mm...” .36... En
$93 «a: .3225
$5.5 $18 now. .c6an_6>6u 66.6.60
exevd {one 2. Em_.6n.ccmo
$6M Rem.» A: 666366 :26...ch
$2. $6.0 m 5.666... 63.66383
53 .36... 2.8.3
662 .66”. $2 €6.69? Inc}. 62 acag .6 9.3.. .oz .263 5.83.06. 8.63%..

..

_.

 

moon .0. 6626. 6.5.6353 .< .O 633 6.... 6566A.» 6u< ﬂu 636...

59

.266. .6...

.5 666.6 .62. c. 66:... 6626. .36. .6 663666.66 6... 6. 666. .66. 6.....3 3:63... 6....6666-666.6 63666.66. 666. 6.6.6665.

 

 

 

..

$6....” 2.2. 3.6.6... 666: .33
$6.6. .666 .3635
$6.. 6...? 86 6626. 6.6 86636... 6.666383
$6.6- 6...- 6626. SN 6.6.. 66366.6
$6.6 $6.6 6v 666366 6365...:
$6.6. $6.66 6666 36666.6 .6..666666>>
66...... 66366.6
$6.6. 6....“ .3635
$6.6 $6. . 66 86:66.6660
$2 $6... EN 666366 6365...:
$6. 3 $6.66 6666 ..6..666.6 .6..666666>>
3‘66. .366. ...v
$6.6 3.6.. .3635
$.66 $6.36 66.. 2658653 8366
$6.6 $6.6 «6.. E6__66.c..6o
$6.. $6.3. .66 666366 636..ch
$6.6 $6.66 66. ..6..666.6 .6x666666>>
6666. .366. 6.6
$6.6 66.. .3635
$6.. $6.66 666 56862966 36566
$6.6 $6. _.N 66 E6__66.c..6o
$6.6 $6.6m 66 666366 6365...:
$..6 $6.6 6.. 56666.6 8.6666662.
.666. .36... 66666.
666. .66... 666. ...6.666< 6:36 .62 9.6.6.666 .6 66.2. .62 .6.66. 62.6366. 66.6.3666

_.

 

.66" .6. 6626. 63666.36... .< .6 633 6... 656666 66.4. «.6 636...

60

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%

0%

Y= 0.121ln(x) + 0.0351
R2= 0.3178

0 50 100 150 200 250

 

Percent of larval mortality

Larvae density per m2

Fig. 2.1 Regression of A. planipennis larval mortality by woodpecker predation
versus larval density in 2006. Log1o(x+1) model is signiﬁcant (GLM; P <0.05).

61

 

70%

 

‘<

60% *- 02006
50% J

 

 

 

 

 

40% bc

 

30%
20%
10%

0%

 

 

Percent of Larvae

 

 

 

 

 

 

 

 

 

0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8
Height(m)

Fig. 2.2 Mean (:l: SE) proportion of A. planipennis larvae killed by woodpecker

predation per m2 by log height up to 8 m above ground in 2006 and 2007.
Within year means with the same letter are not signiﬁcantly different (Tukey
protected LSD test; P<0.05).

62

 

80%

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

70% ‘i‘ 02006
g 60% ‘ I2007
g 50%
Q-
3 40%
g 30% y
g
a, 20% a
xy
10%
0% "J I ﬁ l 1

 

 

 

 

 

 

 

 

October November December January February March

Fig. 2.3 Mean (1: SE) proportion of A. planipennis larvae killed by woodpecker
predation per m2 in the month from October through March in 2006 and 2007.
Within year means with the same letter are not signiﬁcantly different (Tukey
protected LSD test; P<0.05). Trees were not felled until December in 2006. Tree
felling was completed by late February in 2007.

63

 

16%
14%
12%
10%
8%
6%
4%
2%

A) 0%

Percent of total larvae

16%
14%
12%
10%
8%
6%
4%
2%
0%

Percent of total larvae

EB

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

y = 0.0026In(x) - 0.0023 _

R2=0.2233

C: 8
DD B I;
W U

0 50 100 150 200 250 300 350
‘3 y= 0.0054ln(x) - 0.0103 _

R==0.1244

a
£1 Cl

 

 

 

 

 

 

Larval density per m2

 

Fig. 2.4 Regression of A. planipennis larvae cannibalized by larval density in
2006 (A) and 2007 (B). Log1o(x+1) models are signiﬁcant (GLM; P <0.05).

64

 

CHAPTER 3

Development of Agrilus planipennis larvae (Coleoptera:
Buprestidae) on stressed and healthy Fraxinus pennsylvanica

trees

INTRODUCTION

Emerald ash borer, Agn'lus planipennis Fairmaire (Coleoptera:
Buprestidae), was initially thought to have a univoltine life cycle in North America
(Cappaert et al. 2005a). However, recent studies have indicated that some
larvae feed for two summers (Cappaert et al. 2005b), a pattern most common in
trees with low densities of A. planipennis. For example, in an outlier site with a
one-year old infestation, only 18% of 282 larvae developed in one year (Cappaert
et al. 2005b). In a second, newly established outlier site, tree vigor appeared to
play an important role in determining larval development time. When trees were
debarked in February, the ﬁrst winter after the initial infestation, the 36 larvae on
a stressed tree were all 4th instars or prepupae. They would have emerged after
a single summer of feeding. Only 23% of the larvae feeding on vigorous trees at
the site were 4th instars or prepupae. The rest would likely have fed for an
additional summer to complete development, emerging after two years (Cappaert

et al. 2005b).

65

In this study, I determined the effects of host vigor and larval population
density on the rate of larval development. I hypothesized that larvae feeding on
stressed trees develop faster than healthy trees. I also hypothesized that larvae
in a high density population develop faster than larvae feeding in a low density
population. I monitored the relationship of development of larvae produced by
adults caged on trees with varying densities of A. planipennis larvae produced by

wild adults.

66

MATERIALS AND METHODS

2006 Study

For this study we used the same green ash, Fraxinus pennsylvanica
Marshall (Oleaceae), trees described in Chapter 1 and Chapter 2. From the 90
trees in the 2006 study and the 90 trees in the 2007 study, we randomly selected
two trees per treatment in each of the ten blocks. Thus, a total of 60 trees were
used for this study including 20 trees girdled, 20 trees exposed to the stress- r

elicitor methyl jasmonate, or 20 trees left as untreated controls. The trees used

 

in the 2006 study and the 2007 study were felled and debarked the following L'.
winter as described in Chapter 1.

Adult A. planipennis were reared from heavily infested bolts of ash trees
collected from sites in Washtenaw County, MI from October through November
2005. The trees were bucked into logs, each 0.5 m long (approximately16 cm
diameter) and placed in cold storage at 38° F. Logs were pulled from cold
storage in early May 2006 and placed in 25 cm diameter cardboard tubes with
screened caps in a room maintained at 267° F, 50% RH. Adult beetles began to
emerge from logs three weeks later. After emergence, adult beetles were sexed
and placed in clear plastic 120 ml cups in groups of six adult beetles (three males
and three females) per cup for a two week period. Sex was determined by
examining the beetles under a stereoscope. Males were identiﬁed by the ﬁrst
and second abdominal sections, which are narrower than those of females.
Males also have a visible layer of setae on the ventral side of the thorax which

females lack. Leaves of F. pennsylvanica, collected from the ﬁeld in lngham

67

County, were provided to the beetles and replaced twice a week. Cups with
beetles were kept in growth chambers kept at 24° C, 75% RH and 16:8 lightzdark
photoperiod. After about a week, we observed adult beetles beginning to mate.
Adult female beetles were assumed to have mated after fertile eggs appeared in
the cups, typically two to three weeks after emergance. Eggs are fertile when
they change from creamy white to reddish-brown.

On 13 June, we attempted to establish high and low densities of larvae on
our study trees by caging adult A. planipennis on the trunks. Cages consisted of
Medegen Gent-L-Kare® (Medegen Medical Products, Gallaway, TN) 4 oz.
specimen containers painted white. The bottom was cutoff and covered with
plastic mesh attached with hot-glue. A 5 cm diameter hole was cut in the lid,
leaving only the threads. Cages were attached to the bark of the trees at a
height of 1.5 m. Lids were attached to the bark using GE Silicone ll* Window &
Door (Clear) (GE Sealants & Adhesives, Huntersville, NC) with the threads facing
outwards. After two days, the silicone had set and the cage was screwed onto
the lid.

To generate locally high larval densities, we placed four cages on each of
ten trees per stress treatment. Cages were placed at the same height and
spaced 2 cm apart, around the trunk. To generate low larval densities, we
placed one cage at the same height on the north aspect on each of ten trees per
stress treatment. In other words, there were 30 trees with four cages each, on
ten control, ten methyl jasmonate, and ten girdled trees and 30 trees with one

cage each, on ten control, ten methyl jasmonate, and ten girdled trees (total of 60

68

trees). When a trunk branched below 1.5 m, an additional set of cages were
placed on the second trunk as above. The results from the two trunks were
averaged for that tree. Petioles of F. pennsylvanica leaves were pushed into
water-ﬁlled microcentrifuge tubes and placed into each cage. Foliage was
replaced twice a week. One male and one female beetle that were at least two
weeks old were placed in each cage. Using the same criteria as Anulewicz
(2006), if a female beetle died during the ﬁrst three days of the study, she was
replaced with another female beetle to ensure adequate time for egg laying.

I waited an average of 74.9 :I: 1.0 days (range 66 to 87 days) after adults
were placed on trees before peeling so larvae could feed and develop. Between

18 August and 8 September, we peeled the bark under the cages with a
drawknife. A mean area of 0.13 :l: 0.01 m2 was debarked under each cage. We

recorded the number, life stage, and mortality of larvae in each window. lnstars

were identiﬁed by head capsule width (Cappaert et al. 2005b). As described in
Chapter 1, we assumed that larvae found as 18‘, 2nd, or 3rd instars would

probably not reach the prepupal stage by spring, and would require another
summer of feeding before pupating (Cappaert et al. 2005a).

Although we removed 11 trees that had external symptoms of A.
planipennis infestation in May 2006, a population of A. planipennis persisted in
the plantation and surrounding area (Chapter 1 and Chapter 2). These wild
beetles laid eggs on the trees. The 90 trees in this study were felled and
debarked from December 2006 through March 2007 as described in Chapter 1.

Larvae associated with the wild A. planipennis population, found when study

69

 

trees were debarked. Larval diversity was standardized per m2 of exposed area

on each tree.

2007 Study

We replicated the study in 2007 with a few modifications. Ten blocks of
nine trees were selected using the same parameters as in 2006 for a total of 90
trees. Trees were girdled on 10 May. A line with ten methyl jasmonate bubble
caps spaced 20 cm apart was suspended through the canopies on 25 May, and
a second line with 10 methyl jasmonate bubble caps was suspended through the
canopies on 25 June as described in Chapter 1.

Adult A. planipennis in 2007 were reared in a similar way as in 2006 with
the exception that Fraxinus udhei (Wenzig) leaves, a tropical species, were
provided in cups instead of F. pennsylvanica foliage.

F. udhei foliage was also placed in the cages on trees. To assure a higher
rate of successful gallery formation, we placed two cages on each of two trees
per treatment in randomly selected blocks (total of 60 trees) (20 trees per
treatment). One cage was attached on the north aspect and the other on the
south aspect. Adult beetles were caged on 26 June.

In 2007, we waited an average of 80 j: 1 days (range 76 to 101 days) after
adult beetles were caged before beginning to debark areas under the cages.

Between 10 September and 5 October, we debarked an average of 0.12 :l: 0.003
m2 under each cage. Larval galleries were traced onto transparencies, which

were then scanned. The area of phloem consumed per larva was calculated

7O

using the computer program WinfoliaTM (Winfolia 2004a). Trees were then felled

during between October 2007 and February 2008, as in the 2006 study.

Phloem Disruption - Pilot Study

In 2007, we initiated a pilot study with 10 additional trees in the same
plantation. On 10 June, a 2 cm wide horizontal band of phloem was removed
from half of the circumference of the trees at 1.38 m above ground on the south
aspect. On 18 June, we selected 20 female and 20 male adult beetles that were
ﬁeld collected from a heavily infested site in lngham country, Ml. These beetles
laid fertile eggs in a related study (Anulewicz et al. 2008) and were re-used for
this pilot study. On each of the 10 trees, we placed one cage 4 cm above the
phloem disruption and one cage 4 cm below the phloem disruption, directly in
line with each other. Areas under the cages were debarked an average of 84 i 1
days (range 80 to 86 days) after the adult beetles were caged. On average, 0.14
i 0.005 m2 were debarked per cage. We recorded the number, instar and

mortality of larvae produced by caged females and standardized larval density
per m2 of exposed area. Galleries were traced onto transparencies which were

scanned and the area of phloem consumed per larva calculated as described

above.

Statistical Analysis
All data were tested for normality using the Shapiro-Wilk test (Shapiro and

Wilk 1965) and residual plots. We evaluated three response variables, including

71

larval density from caged adults, the proportion of those larvae that reached at
least the 4th instar, and the amount of phloem consumed per larva by instar in
2007. Larval density and phloem consumption were log10(x+1) transformed to
normalize the data. Proportion of larvae that reached at least the 4th instar

required no transformations. The same response variables were tested in the
phloem disruption pilot study, following transformations to normalize the data.
All response variables were analyzed using the GLIMMIX procedure for
mixed models in SAS statistical software (PROC GLIMMIX; SAS Institute 2003).
Effects of three factors: tree vigor (control, methyl jasmonate, or girdled), density
treatment (high density with four cages per tree and low density with one cage
per tree) in 2006 only, and canopy exposure with ﬁve levels (1, 2, 3,4, 5) were
analyzed. Block and block crossed with treatment were used as random factors.
Differences among treatment means were tested as unplanned comparisons and
multiple comparison tests were applied only when overall analysis of variance
(ANOVA) was signiﬁcant (P<0.05). When signiﬁcant differences among levels

occurred, the Tukey protected least signiﬁcant difference (LSD) test was used to
evaluate those differences (Ott and Longnecker 2001). The log10(x+1)

relationship between density of the wild A. planipennis population and the
proportion of one year larvae and the log10(x+1) relationship between density of

the caged A. planipennis population and the proportion of larvae under cages

that had reached at least 4"1 instar was analyzed using a general linear model

with a log1o(+1) transformation (PROC GLM, SAS institute 2003).

72

RESULTS

2006 Study

Overall, a total of 174 larvae were found under the 171 cages in 2006.
Adult female beetles survived in cages for an average of 9.5 :l: 0.3 days. At least
one larva was found on 31 of the 60 trees that had one or more cages. Larvae
were associated with 56 of the total 171 cages (Table 3.1). Twice as many larval
galleries were associated with cages on girdled trees compared with cages on
control and methyl jasmonate trees (Table 3.1). There was a signiﬁcantly higher

number of galleries on trees with four cages compared to trees with a single

cage, with an average of 12.3 t 2.5 larvae per m2 compared with 3.2 :I: 1.3 larvae

per m2, respectively (F1,23=15.98, P<0.001). Signiﬁcantly more larvae were
produced by beetles caged on girdled trees than control or methyl jasmonate
trees (F2,13=3.65, P=0.047). There were three times as many larvae on girdled

trees compared with control and methyl jasmonate trees (Fig. 3.1). Larval
densities were not signiﬁcantly different between control and methyl jasmonate

trees (Fig. 3.1). The interaction between larval density and treatment was not
signiﬁcant (F2,23=2.01, #0156). Canopy exposure (F3,23=2.09, P=0.130) did

not signiﬁcantly affect larval density below cages.

On average, 16.5% :l: 3.1% of larvae found below the cages had reached
the 4th instar or were prepupal larvae at the time of debarking. Larvae developed

signiﬁcantly faster on girdled trees than on control or methyl jasmonate trees

(F2,11=7.99, P=0.007) (Fig. 3.2). Development of larvae on trees with four cages

73

and larvae on trees with one cage did not differ signiﬁcantly (F1,1=0.83, P=0.529).
An average of 19.5% :I: 5.4% and 8.8% :t 6.3% of larvae reached the fourth instar
on trees with four cages and larvae on trees with one cage, respectively. The
interaction of stress treatment and density was not signiﬁcant (F2,1=0.80,
p=0.621). Density of larvae from the wild A. planipennis population did not
signiﬁcantly affect development rates of larvae (F1,1=8.91, P=0.206). Larval
development was not signiﬁcantly affected by canopy exposure rating (F3.1=1.61,
P=0.512).

Mortality of larvae under cages was very rare. Only three, one, and two

18’t instars died on control, methyl jasmonate and girdled trees, respectively, all

from unknown causes. Cannibalism, woodpecker predation or pathogens were
not observed. All late instar larvae were alive when the area under the cages

was debarked.

2007 Study

Overall, a total of 278 larvae were found under the 106 cages in 2007.
Adult females survived for an average of 10.7 1: 0.5 days. At least one larva was
found on 32 of the 53 trees that had cages and larvae were associated with 50 of
the 106 cages (Table 3.1). Larval galleries were associated with three times as
many cages on girdled trees compared with control and methyl jasmonate trees
(Table 3.1). There was no signiﬁcant difference in the number of cages
associated with larvae on control and methyl jasmonate trees (Table 3.1).

Average larval density was signiﬁcantly higher when beetles were caged on

74

girdled trees than on control or methyl jasmonate treatments (F2,13=3.77,

k0.043). Six times as many larvae were found under cages on girdled trees
compared with cages on control and methyl jasmonate trees (Fig. 3.1). Larval
density did not vary signiﬁcantly between control and methyl jasmonate trees
(Fig. 3.1). Additionally, larval densities from beetles caged on control and methyl
jasmonate trees in 2007 were similar to the larval densities on control and methyl
jasmonate trees in the 2006 study (Fig. 3.1). Larval densities from beetles caged

on girdled trees were twice as high in 2007 as they were on girdled trees in 2006.
Larval density was not signiﬁcantly affected by canopy exposure (F4. 5o=1.39,
P=0.249).

Overall, an average of 56.5% i: 7.7% of larvae had reached 4th instar or

the prepupal stage when debarking occurred. The proportion of larvae that had

reached 4th instar or the prepupal stage did not vary signiﬁcantly among
treatments (F2,12=1.23, P=0.327) (Fig 3.2) or by canopy exposure rating (F4,
4=3.22, P=0.142). No signiﬁcant relationship existed between the density of
larvae from wild populations and the development of larvae produced by caged
adults (F1,4=1.25, P=0.326).

Mortality of larvae under cages in 2007 was again rare. Only one 1$t
instar, one 19’t and two 2"d instars, and one 2'“, two 3”, and one 4th instar died

on control, methyl jasmonate, and girdled trees, respectively, all from unknown

causes. Cannibalism, woodpecker predation or pathogens were not observed.

75

The area of phloem consumed per larva per instar averaged 1.18 :t: 0.23

cmz, 2.11 :l: 0.44 cm2, 3.20 :l: 0.33 cmz, 7.56 1 0.52 cm2, and 12.45 i 1.10 cm2
for 18‘, 2nd, 3rd, and 4th instar or prepupae, respectively. Treatment (F2,1o=1.28,

P=0.319) and canopy exposure (F4,39=2.02, P=0.099) did not signiﬁcantly affect
phloem consumption. Most phloem consumption occurred in later stages of
development, with feeding by 4th instars and prepupal larvae accounting for

about 75% of all phloem consumed.

Phloem Disruption - Pilot Study

Adult females caged above and below the thin strip of removed phloem
survived for a mean of 14.6 1: 1.9 days. Galleries were found beneath only two of
the ten cages above the phloem disruption for a total of six larvae, with a mean of
0.7 :l: 0.4 larvae per tree. Seven of the ten cages below the phloem disruption
produced 45 larvae, with a mean of 5.0 :l: 1.4 larvae per tree. No larvae above
the phloem disruption had reached the 4th instar by the time bark was removed in
September 2007. Below the phloem disruption, a mean of 43.1% :t 12.7% of
larvae were 4th instars or prepupae. The mean phloem consumed per larva per
instar, derived mostly from larvae below the phloem disruption, was 1.34 :l: 0.04
cmz, 2.09 1 0.14 cmz, 4.82 i 0.72 cmz, 5.03 :l: 2.14 cm2, and 12.39 cm2 for 1*"t
instar, 2'“1| instar, 3rd instar, 4th instar, and prepupae, respectively, similar to the

consumption rates recorded for larvae under the cages in the main study.

76

DISCUSSION

In both 2006 and 2007, larvae developed faster on girdled trees than on
non-girdled trees, and while tree exposure to methyl jasmonate had no effect on
larval development. Density of larvae produced by wild A. planipennis beetles
was also higher on girdled trees but we found no signiﬁcant relationship between
density of wild larvae and development rate of larvae from caged beetles.

Interestingly, larvae were found more frequently and at higher densities
under cages on girdled trees compared with control and methyl jasmonate trees.
This occurred even though the same number of adult female A. planipennis was
caged for the same amount of time on trees of all treatments. Cages on girdled
trees had two to three times as many larvae associated with them compared with
control and methyl jasmonate trees in 2006. In 2007, cages on girdled trees had
three to ﬁve times as many larvae associated with them compared with control
and methyl jasmonate trees. The population density of wild A. planipennis larvae
was also higher on girdled trees. This suggests that both larval density and
development rates are affected by host vigor and that larval development is not
strongly affected by larval density. Additionally in the phloem disruption pilot
study, more larvae were associated with cages below than above the phloem
disruption.

In similar studies with native Agrilus sp., larvae associated with caged
adults were found in higher densities on girdled trees compared with non-girdled
trees. When adult Agrilus anxius (bronze birch borer) were caged on girdled

trees and on non-girdled trees, 30% and 82% of cages yielded larvae on non-

77

girdled and girdled trees, respectively (Nash et al. 1941). Barter (1957) did not
report the exact number of larvae produced by adults in a four-year study using
caged adults. His qualitative results, however, indicate that the larval population
under the cages increased in density each year as tree vigor declined due to
girdling from the larval feeding.

At least two factors may explain the difference in larval density among
cages on girdled, control, and methyl jasmonate trees. First, adult A. planipennis
and A. anxius may be making a choice about whether to lay or not to lay eggs
depending on host suitability. Alternatively, more neonates may have died on
control and methyl jasmonate trees without leaving detectable galleries. l was
very careful in dissecting the bark and phloem, and it is very unlikely that neonate
galleries were missed, so the second hypothesis can probably be rejected.

Other reports suggest A. planipennis adults may be selective in where they
choose to lay their eggs. Anulewicz et al. (2008a) found that when wild
populations of adult A. planipennis were presented with logs from host species,
non-host species and black plastic drain pipes, signiﬁcantly more eggs were laid
on ash logs than non-ash logs, and no eggs were found on plastic drain pipes.

Host selection by A. planipennis beetles likely involves adult attraction to
host volatiles. Host selection is important because wood-boring larvae rely on
the oviposition choices of adult females and larvae are unable to disperse to
more suitable host (Hanks et al. 1993). Studies performed in Chapter 1 reported
that wild adult beetles were captured more commonly on girdled trees than

control and methyl jasmonate trees. Stressed trees are known to produce foliar

78

volatiles that are attractive to A. planipennis adults in laboratory tests (Rodrig uez-
Saona et al. 2001, Poland et al. 2006). Adult A. planipennis in cages were not
directly exposed to foliar volatiles, but volatiles produced by bark from trees
stressed by girdling can also produce antennal responses from A. planipennis
adults (Crook et al. 2008). Some of the volatiles produced by bark on trees
stressed by girdling are also found in Manuka oil distilled from the New Zealand
tea tree (Leptospennum scopan‘um Myrtaceae). Field studies have found that
traps baited with Manuka oil attract A. planipennis adults at signiﬁcantly higher
rates than unbaited traps (Crook et al. 2006, Poland and McCullough 2007a,
Poland et al. 2007b, Crook et al. 2007, Anulewicz et al. 2007, McCullough et al
2008, Crook et al. 2008a, Crook et al. 2008b).

The hypothesis that female beetles lay few eggs on unsuitable hosts is
supported by no-choice laboratory bioassays with A. planipennis. Anulewicz et
al. (2006) reported that A. planipennis eggs were associated with Fraxinus spp.
at four to ten times the rate compared with non-host species. Still, some eggs
were laid on unsuitable hosts, including the sides of the cages. These eggs were
laid at far lower rates compared with the number of eggs laid on the host,
suggesting that another factor, albeit with less inﬂuence, also induces oviposition
behavior.

Girdling affected larval development as well as ovipostion. Larvae from
caged adults developed faster on girdled trees compared to larvae found on
control and methyl jasmonate trees in 2006 but not in 2007. Buprestids generally

develop faster and have a higher rate of developmental success on host trees

79

that are stressed by environmental factors such as drought or artiﬁcially by
girdling (Koricheva and Larsson 1998). Data from A. planipennis outlier sites
similarly showed larvae on stressed trees developed faster than larvae on non-
stressed trees (Cappaert et al. 2005b, Tluczek et al. 2008).

While the exact mechanism for how girdling or phloem disruption affects
larval development is still unknown, we can speculate on possible mechanisms.
Girdling may affect the nutritional content, the defensive chemical response,
moisture levels, or the ability of the tree to produce callus tissue. Available
carbohydrates are known to increase in phloem tissue above branch-girdles on
fruit trees by growers attempting to increase their yield (Jordan and Habib 1995).
We did not compare the nutritional or chemical defenses on girdled with non-
girdled trees, but future studies should evaluate them.

Girdling may also affect water pressure in the phloem. Water pressure is
an important component of cell division and growth in trees, with a speciﬁc
pressure required for continued cell growth (Boyer and Silk 2004). The level of
moisture in an environment has been found to inﬂuence the development of
Phoracantha semipunctata (Fabricius), a cerambycid in eucalyptus (Eucalyptus
rudis Endlicher) (Hanks et al. 1999). A similar process might be affecting the
development of A. planipennis. Damage from girdling or feeding can disrupt
phloem tissue not directly affected by the injury. This is because phloem tissues
are elastic and adjust to pressure changes caused by disruptions in the phloem
(Lee 1981a). While phloem and xylem water-potential has been shown to be

highly correlated and water can ﬂow between the phloem and xylem laterally

80

(Lee 1981b), girdling does not affect the xylem and therefore probably has
minimal, direct impact on moisture levels. Additionally, anecdotal observations in
the ﬁeld found that moisture levels under the bark were still quite high on girdled
trees, so moisture variation may not result from phloem girdling.

Fierke and Stephen (2008) reported that the ability of a tree to produce
callus tissue was correlated with the level of attack the tree sustained from
cerambycids. Trees that could not produce callus tissue were successfully
attacked at higher rates than trees that could produce callus tissue. However, no
callus tissue was observed around galleries formed by A. planipennis larvae
under the cages on any tree, so tissue callus is probably not a contributing factor
to larval development in this study.

Methyl jasmonate had no effect on population density, development, or
mortality of larvae associated with larvae in this study. A. planipennis adults
have been shown to be attracted to volatiles produced by ash seedlings exposed
to methyl jasmonate (Rodriguez-Saona et al. 2006, Poland et al. 2006).

Similarly, McCullough et al. (2006) found trees exposed to methyl jasmonate
were no more attractive than untreated trees to wild adult beetles. Development
of larvae produced by wild beetles on methyl jasmonate trees did not signiﬁcantly
vary from controls as well (Chapter 1). We suspect that that methyl jasmonate
had no effect on our trees either because they were too large and mature, or we
did not use enough methyl jasmonate to elicit a stress response. Leaves from
trees in this study were clipped then aerated, but no volatiles were collected.

Future studies should include measuring volatiles directly from trees.

81

 

Development of larvae from caged adults was not signiﬁcantly related to
the density of larvae produced by the wild A. planipennis population. The
average larval density associated with cages did not change on control and
methyl jasmonate trees from 2006 to 2007. However, the proportion of larvae
that had reached the 4th instar or prepupal stage in 2007 was two to eight times
the proportion found in 2006. Larval development was also not signiﬁcantly
related to the population density of larvae produced by wild beetles when
comparing individual trees. However, if you compare the average larval
development at the study site with population density between 2006 and 2007,
there appears to be a relationship. This was similar to the results reported in
Chapter 1.

In both the caged study and the phloem pilot disruption study, larvae
consumed a total of about 12 cm2 of phloem before developing into a prepupa,

regardless of treatment. Using this we can calculate that the total potential yield
of A. planipennis is about 833 adults per m2 of phloem. This supports the
ﬁndings by McCullough et al. (2007) who observed densities of young larvae in
heavily infested logs ranging from 300 to 1000 larvae, but produced an average
of 89 adult beetles per m2 of surface area after high rates of mortality occurred.
The population density of larvae on girdled trees produced by wild adults
averaged 71 larvae per m2 in 2006 and 175 larvae per m2 in 2007 (Chapter 1).
This matches observations in the ﬁeld that unutilized phloem was still available.

About 75% of the phloem was consumed by later instars. This is relevant for

82

 

insecticide control because if feeding can be halted before larvae develop into

late instars, feeding damage could be minimized.

 

83

86.on .50... cm... 880on
>935 95.5mm: 95:5 220:6 3:85ch «o: 05 .033 9:3 05 55. megs. 50> c_£_>>

 

 

mm m w omZm. 5.3 898 .02
or w m £30m 398 So... 095
0 cm5 0.0:. 5.3 wow: .02
ow ow { as 898 .02
85.2 « ~32 $3 a 9.. S «8.2 a B. a £5 3.2% .85. 2.; :85.
«no a a? «so a m2 and a 2: 95 £99. 8.. :85.
mm... a of who a RF and 4 <8 .53 1mm :85.
2 t 2 89. .02

soon
mm 8 as 82m. 5.; «.88 .oz
2 3 m mgacm ocmmo E0: mats
o cm... 0.5:. 5.? wow: .oz
.8 8 mm 880 .oz
83 H 91: «88 4 Ba 83 a 3!. £5 3.2.8 .093. 2.; :85.
an... a 0.2 «no a m9 and a F. 2 as 299. 8.. :85.
mm... a 3? «to H 4.8 mod 0 Nd .23 :mo cams.
om on on 89. .oz

ooou

8.2.0 05858.. 350.2 .228

 

3cm new ooou c. aﬁcwéxmctoq
95.98". co tonne 3.35 Eat mouse .895 main. $583563 .< can dock .393 2.3

E88 .25. 2.; .292. .....mo. £92. .235 :3 Ewen... 8: .mm a. :8... w... 2%»

84

40 El Control
I Methyl Jasmon ate
I Girdled

Larval Density per m2
- _. 16 to co co
(II O 01 0 U1 0 U1

 

O

2006 2007

Fig. 3.1 Mean (:l: SE) density of A. planipennis larvae per m2 produced by caged
beetles on untreated control trees, trees exposed to methyl jasmonate, and
girdled trees in 2006 and 2007. Within each year, means with the same letter
are not signiﬁcantly different (T ukey protected LSD test; P <0.05).

85

100%
90% El CODtl‘OI

30% I Methyl Jasmon ate
70% I Girdled
60%
50%
40%
30%
20%
10%
0%

Percent of One Year Larvae

 

2006 2007

Fig. 3.2 Mean (:l: SE) ercentage of A. planipennis larvae produced by cages
4’.

per m2 that reached 4 instar or prepupal stage on untreated control trees, trees
exposed to methyl jasmonate, and girdled trees in 2006 and 2007. Within each
year, means with the same letter are not signiﬁcantly different (Tukey protected
LSD test; P <0.05).

86

CHAPTER 4

Development, survival and phloem consumption of Agrilus
planipennis Fairmaire larvae phloem arenas and potted ash
sections

INTRODUCTION

After the discovery of emerald ash borer, Agn'lus planipennis Fairmaire ,
(Coleoptera: Buprestidae) in southeast Michigan in 2002, the life cycle of A.
planipennis larvae was found to vary from one to two years in North America

(Cappaert et al. 2005b). However, factors determining the rate of larval

 

development remain unknown (Cappaert et al. 2005b). One possible mechanism
may be that at higher densities intraspeciﬁc cannibalism provides more nutrients
leading to faster larval development and a univoltine life cycle. When
Enaphalodes rufulus (Haldeman) (Coleoptera: Cerambycidae), the red oak borer
larvae cannibalize conspeciﬁc larvae, they grew at nearly ﬁve times the rate of
larvae developing only on phloem and wood (Ware and Stephen 2006).
Cannibalism was observed in high density populations of A. planipennis (Bauer
et al. 2004) and larvae in high density sites generally develop in a single year.
(Cappaert et al. 2005b). If cannibalism is common when A. planipennis larvae
are at high densities, it may partially explain why larvae develop in a single year
in high density environments, but require two years in trees with low densities.
The main goal of this study was to determine if A. planipennis larval
cannibalism was related to larval density. If cannibalism occurred frequently, l
hypothesized that, A. planipennis larval development rates would increase,

similar to E. rufulus. I conducted two experiments. The ﬁrst experiment involved

87

rearing larvae at different densities on phloem arenas, consisting of small pieces
of ash phloem sandwiched between Plexiglas. Speciﬁcally, I expected
development of A. planipennis larvae that cannibalized other larvae to be
enhanced and cannibalism to be common in high density settings. In the second
experiment, I attempted to establish varying densities of A. planipennis larvae on

small sections of potted ash in a greenhouse.

88

 

METHODS

Phloem Arenas

On 17 July 2006, I collected 200 A. planipennis 1”t instars from a white
ash tree, Fraxinus americana L. (Oleaceae), in Jackson County, MI by chiseling
the underlying cambium and sapwood, creating a wood chip that held a larva.
Wood chips were placed in a 25.4 cm Petri dish lined with a damp paper towel
and returned to the laboratory. Larvae from the wood chips were randomly
assigned to either a low density or high density treatment.

On 17 July 2006, a 7.6 cm diameter key-hole saw on a 19 volt cordless
drill was used to remove circular sections of bark with intact cambium and
phloem from a single, F. Americana tree. Each piece of phloem was placed in a
separate plastic. bag in a cooler and transported back to the lab. The pieces of
bark and phloem were sandwiched between two pieces of 5 cm by 5 cm
Plexiglas (0.25 cm thick) with the outer bark facing down and the cambium facing
up.

We randomly assigned 15 phloem arenas to high density treatments and

15 to low density treatments. Each high density phloem arena recieved eight 1St

instars and each low density arena recieved one 1st instar. Larvae were placed

in 0.3 cm deep grooves carved with a tapering 5-mm-wide metal spatula. High
density arenas had eight grooves in two rows of four spaced 3 mm apart in the
center of the arena, while the low density arenas had one groove in the center of
the phloem arena. The two pieces of Plexiglas were then bolted together using

four 5-cm-long, 0.8 cm bolt inserted in holes in the corners of the plexiglas and

89

 

secured with a wing-nut over a locking washer. Once secured, a layer of
paraﬁlm was wrapped around the edge of the arena to seal it. The arenas were
placed in a growth chamber and kept at 24 °C and 75% RH in the dark.

Once a week, fresh phloem was harvested from a new F. americana tree.
Larvae were extracted from the phloem arenas and moved onto the fresh
phloem. The Plexiglas and all equipment were sterilized with 95% ethanol.
Developmental instar, phloem consumption, and mortality were recorded each
time larvae were transfered, but individual larvae could not be marked or
distinguished. Larval instar was determined by measuring the width of the head
capsule (Cappaert et al. 2005b). To measure phloem consumption, we traced
the galleries onto transparencies, which were scanned and the area of phloem
consumed was calculated using the computer program WinfoliaTM (Winfolia
2004a). The area of phloem consumed per larva per week was calculated by
dividing the total area of phloem consumed in an arena by the number of live
larvae feeding that week on each arena. Some larvae burrowed below the
surface of the top of the phloem arena. When I moved these larvae to a new
phloem arena, they were dug out of the old arena using a knife. Galleries below
the surface were traced onto the transparency to ensure all feeding was
included.

Causes of larval mortality included handling, cannibalism, and mold.
Handling mortality occurred when larvae were killed as they were transferred to a
fresh phloem arena or if the larva was not placed properly into the new arena and

failed to feed. Cannibalism was deﬁned by Elgar and Crespi (1992) as the killing

90

 

and consumption of intraspeciﬁc individuals. We recorded a cannibalism event
when we found the remnants of a larva where two galleries intersected, or when
one gallery ended at another gallery and there was no trace of one of the
associated larvae. Mold was recorded when fruiting bodies were observed on a
larval cadaver. Some larvae were missing. Missing larvae may have been a
result of cannibalism where the surviving larva completely consumed the missing
larva and took over its gallery.

Larvae that died during the experiment were not replaced, and as time

went on, the number of phloem arenas with live larvae decreased (Table 4.1).

 

By 14 August, intact pieces of phloem could no longer be removed from trees,
because trees were preparing for dormancy. Larvae remained in arenas for two

ﬁnal weeks of feeding before the last measurements were recorded.

Potted Fraxinus americana Sections

In December 2006, two uninfested F. americana trees (DBH 10 cm) were
felled and trimmed in Riley Township, Clinton County, MI. Branches were
removed and the trunks were waxed on both ends, then placed into cold storage.
In January 2007, we cut the trees into 40 sections, each with a mean length of
29.4 :t 0.2 cm and diameter of 5.0 :l: 0.2 cm. One end of each ash section was
treated with the plant hormone Rootone® (Garden TechTM, Lexington KY)
following methods of Herard et al. (2004). The other end and any bark wounds
were waxed. The ash sections were then planted at a depth of 15 cm in Fafard 2

Mix (Conrad Fafard lnc., Agawam MA) potting soil in seven l pots made of thick

91

black plastic with an upper diameter of 22.9 cm and a height of 21.6 cm, ﬁlled to
3 cm from the top with soil (P. Bloese, MSU Tree Research Center, pers. comm.
2005). The pots were placed in pairs in plastic trays with 5 cm of water. Trays
were reﬁlled once or twice and were allowed to dry completely twice a month.
The trays were kept in a greenhouse at the Tree Research Center at Michigan
State University with an average temperature of 19°C, yielding 63 degree days
per week (base 10°C). Two weeks after planting, callus tissue began to form
along the top of the section and a variable number of sprouts formed on most
stems.

Adult A. planipennis were reared from heavily infested bolts of ash trees
collected from sites in Washtenaw County, MI from October through November
of 2005. The trees were bucked into pieces that were each 0.5 m long (16 cm
diameter) and placed in cold storage at 38° F. Sections were pulled from cold
storage in early May 2006 and placed in 25 cm diameter cardboard tubes with
screened caps in a room maintained at 267° F, 50% RH. Adult beetles began to
emerge from logs three weeks later. After emergence, adult beetles were sexed
by examining the beetles under a binocular dissecting microscope as described
in Chapter 1. Groups of six adult beetles (three males and three females) were
placed in clear plastic 120 ml cups with a snap-on lid. Leaves of Fraxinus udhei
(evergreen ash), a tropical ash, were provided to beetles for feeding and were
replaced twice a week. Cups were kept in growth chambers at 24° C, 75% RH
and 16:8 lightzdark photoperiod. After about a week, adult beetles began to

mate. Adult females were considered to be mated after fertile eggs were laid in

92

 

the cups. Eggs are known to be fertile when they change from a cream to a red-
brown.

On 7 March, we attempted to establish high and low densities of larvae by
caging adult beetles that were at least two weeks old on our cut sections. Before
cages were attached, a one cm wide strip of plastic tree wrap was wrapped
around the stem to facilitate egg-laying. Cages consisted of Medegen Gent-L-
Kare® (Medegen Medical Products, Gallaway, TN) 4 oz. specimen containers
painted white. The bottom was cutoff and covered with plastic mesh attached
with hot-glue. A 5 cm diameter hole was cut in the lid, leaving only the threads.
The lid was attached to the sections using clear window silicone caulk. Petioles
of F. udhei were inserted into water-ﬁlled micro-centrifuge tubes. Foliage was
replaced twice a week. One reared male and female that were at least two
weeks old were placed in each cage. If a female beetle died during the ﬁrst three
days of the study, she was replaced with another female beetle to ensure
adequate time for egg laying, following methods of Anulewicz et al. (2006).

Cages were attached to the bark of the cut sections. Lids were attached
to the bark using GE Silicone II* Window & Door (Clear) (GE Sealants &
Adhesives, Huntersville, NC) with the threads facing outwards. After two days,
the silicone had set and the cage was screwed onto the lid. To generate high
larval densities, we placed three cages on each of 20 ash sections. The cages
were spaced evenly around the stem and staggered in height from the top to the
bottom. To generate low larval densities, I placed one cage on each of 20 ash

sections at the same height on a south aspect. Initially, a large number of adults

93

 

 

died before they could be placed in cages. Only 32 sections, 16 high density and
16 low density treatments, had female adult A. planipennis that survived for 4
days. On 10 April, additional egg-laying adults became available and were
placed on the remaining eight ash sections (four high density and four low

density sections).

 

Larvae were allowed to develop for two weeks after all adult A. '1‘
planipennis died. At that time, two high density and two low density sections
were randomly selected and debarked. This was repeated once a week for ten
weeks. A mean of 464.3 :I: 13.8 cm2 was debarked per ash section. We 8*

recorded the number, life stage, mortality, and phloem consumption of larvae on
sections as described for the phloem arenas. Phloem consumption, after tracing
larval galleries onto transparencies, was calculated per larva. Because, the ash
sections began to die during the study, the area of dead phloem was also

measured on each section.

94

RESULTS

Phloem Arenas
A total of 90.0%, or 108 out of 120 larvae, died on high density phloem
arenas after six weeks (Fig. 4.1A). Handling was the greatest cause of larval

mortality, accounting for over half of all larvae (Fig. 4.1A). Ten to 25% of larvae

 

..
were missing after 6 weeks (Fig. 4.1A). Cannibalism and mold were rare on high

density arenas, with two and ﬁve larvae dying from cannibalism and mold,

respectively (Fig. 4.1A). Most mortality occurred in the ﬁrst two weeks of the
experiment, with 24.2% of all larvae dying after one week, and 53.3% of all i

larvae by the second week (Fig. 1A).

A total of 80.0%, or 12 out of 15 larvae, died on low density phloem
arenas (Fig. 4.1 B) and all mortality was attributed to handling. As with the high
density arenas, most mortality occurred in the ﬁrst two weeks of the experiment.
Overall, 13.3% of larvae died after one week and 46.7% of larvae were dead by
the second week (Fig. 13).

After six weeks, the 15 surviving larva on both high density and low

density arenas had reached the 4th instar. There was little variation between the

total percentages of larvae that had reached the 4th instar on high and low

density arenas each week (Fig. 4.2). A total of 588 degree days (base 10 °C)
were accumulated during the six weeks of the experiment (Table 4.1). Eight out
of the 120 larvae on high density arenas had reached the 4th instar by the second

week (Table. 4.1).

95

Most larvae remained near the top of the phloem arena and were visible
through the Plexiglas. When larvae were moved to a fresh arena, an average
72.6% :I: 3.8% were feeding on the surface of the arena, while the rest of the
larvae were burrowed below the surface. On average, surviving larvae

consumed a total of 24.5 :t 5.5 cm2 and 22.1 i 3.0 cm2 of phloem per larva over

the course of six weeks for low density and high density treatments, respectively F
E

(Table 4.1). if

Potted Fraxinus americana Sections £3

 

A. planipennis adults survived in cages for an average of 4.5 :I: 1.3 days.

A total of 40 and 23 larvae were recorded on seven high density and seven low
density sections, respectively. Larval density on all stems, regardless of date
peeled, averaged 2.0 :1: 0.9 and 1.2 :I: 0.6 larvae per section on high and low
density stems, respectively. Therefore, no statistical comparisons of larval
density could be made. Little larval mortality occurred on the ash sections; a
total of six larvae died on three out sections. Five early instars died when they
tried to feed on dead and dessicated tissue. One second instar was cannibalized
on a stem with a total of eight larvae.

Larvae on ash sections debarked three and four weeks after adults were
placed on cages were all 2"“ instars (Fig. 4.3). Most larvae recovered six to eight
weeks after adult beetles were placed in cages were 3rd instars (Fig. 4.3). Ten
weeks after adult beetles were placed in cages, all recovered larvae were either

4th instars or had reached the prepupal stage (Fig. 4.3). No larvae were found on

96

sections debarked on weeks 5, 9, 12, or 13 (Fig. 4.3). A total of 126 degree days

(base 10 °C) accumulated between the time when A. planipennis adults were ﬁrst

caged and when 1St instars were ﬁrst recovered. This increased to 443
accumulated degree days before 2nd and 3rd instars were recovered and to 630
degree days before 4th instars and prepupae were recovered.

Larvae consumed roughly 2 cm2 of phloem per instar through the third
instar, then consumed about 4 cm2 of phloem to reach fourth instar and the

prepupal stage, for a total mean of 9.9 :l: 1.4 cm2 of phloem consumed per larva

(Fig. 4.4). Interestingly, the maximum horizontal dimension of galleries did not

vary signiﬁcantly between early and late instars (F4,2=10.16, P<0.092) (Fig. 4.4).

Larval gallery areas averaged 2.49 :I: 0.24 cmz, 5.38 i: 1.23 cmz, 5.88 :l: 0.45 cm2,

6.55 i 0.84 cm2, and 9.23 :I: 1.22 cm2 for galleries formed by 13‘, 2nd, 3rd, 4th

instars and prepupae, respectively.

97

 

DISCUSSION

Phloem Arenas
Little variation existed between larval density and the rate at which larvae
developed. These results are similar to the data reported in Chapters 1 and 3,

which also showed that larval density did not affect the rate of larval

 

development. Interestingly, A. planipennis larvae in phloem arenas consumed h
twice the area of phloem as larvae feeding on sections of F. americana or larvae

feeding on live trees (Chapter 3). There may be two possible reasons for this.

The nutritional quality of the phloem may have been poorer in the phloem arenas Ir

compared with living trees or in the cut sections of ash, so that larvae had to
consume more tissue to fully develop. Another reason could be that the phloem
in the arenas may have been missing a key developmental trigger that would
cause the larva to stop feeding and begin to pupate. Additional studies on the
nutritional and chemical composition of ash phloem and their effects on A.
planipennis larvae are needed.

Cannibalism occurred at very low levels on phloem arenas, similar to
results of studies of low levels in other A. planipennis populations in the ﬁeld
(Chapter 2, Bauer et al. 2004). Cannibalism has been observed at higher rates
in other wood boring insects. For example, Ware and Stephen (2006) found that
cannibalism in E. rufulus occurred about 50% of the time when larvae were
placed in phloem arenas. Additionally, E. rufulus larvae grew at ﬁve times the
normal rate after cannibalizing another larva. A. planipennis larvae did not,

however, experience a similar increase in growth or in development rate.

98

The phloem arena design was a modiﬁcation of the design used by Dodds
(2001) and Ware and Stephen (2006) to study E. rufulus (Haldeman). The
arenas used to study the E. rufulus had a large hole drilled into them for larval
placement (Dodds 2001, Ware and Stephen 2006). This did not work for A.
planipennis larvae because they need to push ventrally and dorsally against a
substrate to move and feed. Arenas for Dendroctonus frontalis Zimmerman, the
southern pine beetle, had holes to allow adult beetles to enter the arena and form
nuptial galleries (Grosman 1992, Taylor 1992). The larvae were placed into the
phloem in the arenas used in this study by hand, and while adult A. planipennis
might have laid eggs through the holes on the bark, fresh phloem was needed
weekly and eggs would not have had time to develop.

Most mortality occurred when larvae were handled. If phloem arenas are
used in the future to study A. planipennis, then the size of the phloem arena
should be increased so larvae will not have to be transferred weekly to fresh
phloem. Care must be taken to ensure that the phloem is solid so that A.
planipennis larvae can move and feed. We have observed that larvae must
press against the substrate within a tight space to move efﬁciently. Even a small

depression in the phloem may cause a larva to stop feeding and die.

Potted Fraxinus americana Sections
Overall, 4"1 instars and prepupal larvae were not found on ash sections
until ten weeks after the adult beetles were caged. This coincided with

development times observed in ﬁeld populations (Chapter 3). Compared with

99

 

phloem arenas, larvae on cut sections developed much more slowly. Larvae that
reached the 4th instar or prepupal stage consumed about the same area of
phloem as larvae in Chapter 3, or about 10 cm2 per larva.

Larval mortality was low on the cut sections and mostly occurred where

portions of the stems had desiccated. As in other studies (phloem arenas,

 

Chapter 2 and Chapter 3), cannibalism was very rare, with only one of 63 larvae t

being cannibalized. I
I could not determine in this study if the difference in larval development

rates between phloem arenas and ash sections was caused by degree day i;

accumulation. Larvae used in the phloem arena experiment were collected in the
ﬁeld, we could not determine the initial age of the larvae.

Moisture levels may affect larval development of A. planipennis and other
buprestids. Buprestid larvae on stressed trees perform better than on vigorous
trees (Koricheva 1998), with examples including Agn'lus bilineatus (Weber) (two-
lined chestnut borer) on oaks (Haack 1982, Dunn 1986a), Agrilus burkei Fisher
(ﬁatheaded borer) on alders (Svihra 1993), Agn'lus anxius Gory (bronze birch
borer) on birch (Barter 1957) and on A. planipennis in its native range in China
(Akiyama and Ohmomo 2000, Schaefer 2005, Williams 2005, 2006). A.
planipennis larvae developed faster when feeding in areas under girdled phloem
(Chapter 1 and Chapter 3) and on stressed compared to non-stressed trees
(Siegert et al. 2007). All surviving larvae in this study would have likely
developed into adults in one year. Both the phloem arenas and the cut sections

created environments that were relatively dry compared to phloem found in living

100

trees. Further studies are needed to determine the optimum moisture level for

larval development.

101

 

 

 

 

 

 

8m 8m 3m 8. 8 8° 2
63$ 98: 00..ch cc.m_:E:oo<
E... H 33. R.» H 8.2 mm... H 8.. 8.0 H 8.. 2... H Ed @5389“ .8.
mac. .0: 38:28 EcoEa :82
$8. $5.8 can...» so so .98. £4 8.88.
omen. @5223 .o 33:85...
_. F v F .. mccﬁ .0: 695. .02 :35.
m m o m 2. 695. 5.2. 8:05 62
once; 3.6.3.8 26..
8m 8m 8m 8. 8 .oo 2
ommm. go: 6906: 35.36394
2 H E. E H m... m... H am No H :N to H o... m8. 808 8..
mac. .0: ccEamcoo £023 :85.
nx52. o\o r .8 $08 $5.? $0 56:. :3. 95:80..
695. m:_>_>.:m .0 33:00.9”.
de H «we 56 H .89 ovd H ovN 93 H oo.m 3.0 H 9.0 265 .0: cmZm. .oz :32
o w or m? m— 82m. 9.. 5.2. 8:05 .02
3:93 33:28 :9...
o w m N ..
v.c....c>>
.83. iv 22> an...“ eats. ho oWﬂceohoo E: 6:98 :2. cuts. .0 .BEE

.mw H. :38 En cute. £553.53 .< 5.3 once..." 88...: :o .3232 «6333., no ENE—cam ....v 03m...

102

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100%
90? DAlive BHandling
° IMissing ICannibalism

8 80% 1 IMoldL'
E. 70% -—
1-3 60% '7 "': 5:5
2 50% -- _ ——4
*2 40% ..
§ 30% ~—
g 20% 1'-

10°/o "'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A) 0%

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 2 3 4 6
Week
100% DAI' EJH dl'
rve an Ing
90% _ IMissing ICannibalism
g 30% -l— *- IMOId ,2:
E 70% -~
E 60% "—
.2. 40% ~— ”5
g 30% -— __
g 20% -—
10% -— r
o/ ,5; as:
B 0 0 r ' ’ '
) 1 2 3 4 6
Week

Fig. 4.1. Accumulated mortality of A. planipennis larvae on phloem arenas by
week for A) high density phloem arenas (n=120 larvae) and B) low density
phloem arenas (n=15 larvae).

103

100%
90% c11stlnstar Banlnstar

8 3rd In star I 4th In star

,1, 80%
g 70%
g 60%
3 50%
8 40%
§ 30%
°' 20%
10% ..

0% 3: 2:3 ::
A) 1 2 3 4 6

Week

 

100%
90% D1stlnstar B2nd lnstar

a, 30% I3rd lnstar I4th lnstar
g 70%
¢.v 60%
h
g 50%
8 40%
g 30%
°' 20%
10%
0%

B) 1 2 3 4 6

Week

Fig. 4.2. Percentage of surviving larvae by instar by week for A) high density
phloem arenas (n=120 larvae) and B) low density phloem arenas (n=15 larvae).

 

104

 

100%

2:81:78:
0 n ns l'

i :30? I3rd lnstar

g 600; I4th lnstar

«5 ° I

g 50%

E 40%

g 30%

g 20%

g 10%

 

0%

3 4 5 6 7 8 9 1O 1O 11

Week
Fig. 4.3. Percentage of A. planipennis larvae by instar on potted F. americana
sections debarked three to 11 weeks after adult beetles were placed in cages.
n=13 larvae (Week 3), n=14 larvae (Week 4), n=13 larvae (Week 6), n=12 larvae
(Week 7), n=2 larvae (Week 8), and n=9 larvae (Week 10).

 

105

 

—L
.h

 

 

IArea of phloem consumed
El Horizontal dimension of gallery

A
N
l

r:

 

 

 

.3
O

 

 

 

FBI
FIB-Pr

 

 

”I
1. ,
{t t

. r r r I
0 1 2 3 4 Prepu pae

Life Stage .
Fig. 4.4: Mean (1: SE) area phloem consumed and mean horizontal gallery

 

 

Phloem area (cm2) and
Horizontal Dimension (cm)

ores-moo

 

 

 

dimension of A. planipennis larvae by instar on potted F. americana sections.

106

APPENDICES

107

Appendix 1
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.: 2008-12

Title of thesis or dissertation (or other research projects):

Inﬂuence of host vigor on larval distribution, development, and mortality of Agn'lus planipennis
Fairmaire (Coleoptera: Buprestidae) in North America as inﬂuenced by host vigor

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

Entomology Museum, Michigan State University (MSU)

lnvestigator’s Name(s) (typed)
Andrew Tluczek
Date March-5-2009

*Reference: Yoshimoto, C. M. 1978. Voucher Specimens for Entomology in North America.
Bull. Entomol. Soc. Amer. 24: 141-42.

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

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

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

108

R x 9mm. .CO\ 0

m SuwNMVmV III -............ 88-9585. 28

      

 

 

 

 

 

 

 

Appendix 1.1
Page 1 of 1 Page

Voucher Specimen Data

 

 

 

 

 

 

 

 

 

 

 

-. s. .626. o m
3.825 as 8.8.2 9.. c. 88.8
.0. 29:6on 3%.. 93m c... 328$. 88:... 39:5.
988 .oz .22; .89... .882 $98.82.
$88.00: .. Becca .m:o....uum cm...
:22 or do 62.9.. .o8~.§u:< 80. so: 8.8.30 c.2595... w:..:m<
:ms. or or 8862. :8 89.8.. 880...... 8o. 82. 8995 8.5535... 3.5:.
m m 6 0+ e m e 352.3 05 cam:
m s r S .
m m m m Mus m m m. m w .o 88:8 285.8% 6. s8 .83 858503 8.8%
M w w 0 A A P N m E
”.o .8522

 

 

109

LITERATURE CITED

110

LITERATURE CITED

Akiyama, K. and S. Ohmomo. 2000. The Buprestid Beetles of the world.
Iconographic series of insects 4. Gekkan-Mushi, Tokyo, Japan.

Anderson, R. F. 1944. The relation between host condition and attacks by the
bronzed birch borer. J. Econ. Entomol. 37: 588-596.

Aerts, R. J., D. Gisi, E. De Carolis, V. De Luca, and T. W. Baumann. 1994.
Methyl jasmonate vapor increases the developmentally controlled synthesis
of alkaloids in Catharanthus and Cinchona seedlings. The Plant Journal 5:
635-643.

Anulewicz, A. C., D. G. McCullough, and D. L. Miller. 2006. Ovipostion and
development of emerald ash borer (Agrilus planipennis) (Coleoptera:
Buprestidae) on hosts and potential hosts in no-choice bioassays. Great
Lakes Entomol. 59: 99-112.

Anulewicz, A. C., D. G. McCullough, T. M. Poland, and D. L. Cappaert. 2007.
Attraction of emerald ash borer to trap trees: can MeJa or manuka oil
compete with girdling? Pp. 83-84. In Mastro et al. (eds), Emerald Ash Borer
and Asian Longhorned Beetle Research and Technology Development
Meeting. Cincinnati, OH. 29 October to 2 November 2006. US. Department
of Agriculture, Forest Service publication FHTET-2007-04, Morgantown, WV.

Anulewicz, A. C., D.G. McCullough, D. L. Cappaert, and T. M. Poland.
2008a. Host Range of the Emerald Ash Borer (Agn'lus planipennis Fairmaire)
(Coleoptera: Buprestidae) in North America: Results of Multiple-Choice Field
Experiments. Environ. Entomol. 37(1): 230-241.

Anulewicz, A. C., D. G. McCullough, T. M. Poland, and D. L. Cappaert.
2008b. The ’06 trap trees in ’07. Pp. 71-72. In Mastro et al. (eds), Emerald
Ash Borer and Asian Longhorned Beetle Research and Technology
Development Meeting. Pittsburg, PA. 23-24 October, 2007. US. Department
of Agriculture, Forest Service Publication FHTET-2008-07, Morgantown, WV.

Barter, G. W. 1957. Studies of the bronze birch borer, Agn'lus anxius Gory, in
New Brunswick. Can. Entomol. 89: 12-36.

111

Bauer, L. 8., L. Houping, R. A. Haack, T. R. Petrice. and D. L. Miller. 2004.
Natural enemies of emerald ash borer in southeastern Michigan. Pg. 33-34.
In Mastro and Reardon (eds), Emerald Ash Borer Research and Technology
Development Meeting. Port Huron, MI. 30 September — 1 October, 2003.
US. Department of Agriculture, Forest Service Publication FHTET-2004-02,
Morgantown WV.

Bauer, L. 8., L. Houping, R. A. Haack, R. Gao, T. Zhao, D. L. Miller, and T. R.
Petrice. 2005. Update on emerald ash borer natural enemy surveys in
Michigan and China. Pp. 71-72. In Mastro and Reardon (eds), Emerald Ash
Borer Research and Technology Development Meeting. Romulus, MI. 5-6
October, 2004. US. Department of Agriculture, Forest Service Publication
FHTET-2004-15, Morgantown WV.

Bauer, L. S. and L. Houping. 2006. Egg and larval parasitoids of EAB from
China: potential for biocontrol in North America. Pp. 48-49. In Mastro et al.
(eds), Emerald Ash Borer Research and Technology Development Meeting.
Pittsburg, PA. 26-27 September, 2005. US. Department of Agriculture,
Forest Service Publication FHTET—2005-16, Morgantown WV.

Boyer, J. S. and W. K. Silk. 2004. Hydraulics of plant growth. Functional Plant
Biology 31: 761-773.

Cappaert, D., D. G. McCullough, T. M. Poland, and N. W. Siegert. 2005a.
Emerald ash borer life cycle: a reassessment. Pp. 19-20. In Mastro and
Reardon (eds), Emerald Ash Borer Research and Technology Development
Meeting, Romulus, MI, 5-6 October 2004. US. Department of Agriculture,
Forest Service Publication F HTET-2004-15, Morgantown WV.

Cappaert, D., D. G. McCullough, T. M. Poland, and N. W. Siegert. 2005b.
Emerald ash borer in North America: A research and regulatory challenge.
Am. Entomol. 51: 152-165.

Cappaert, D. L. and D. G. McCullough. 2008. The anticipated host switch: a
new braconid parasitoid in Michigan. Pp. 51-55. In Mastro et al. (eds),
Emerald Ash Borer and Asian Longhorned Beetle Research and Technology
Development Meeting. Pittsburg, PA. 23-24 October, 2007. US. Department
of Agriculture, Forest Service Publication FHTET-2008-O7, Morgantown WV.

Crook, D. J., l. Fraser, J. A. Francese, and V. C. Mastro. 2006. Chemical
ecology of the emerald ash borer, Agrilus planipennis Fairmaire (Coleoptera:
Buprestidae), in relation to tree volatiles. Pg. 63. In Mastro et al. (eds),
Emerald Ash Borer Research and Development Meeting. Pittsburgh, PA, 26-
27 September. US. Department of Agriculture, Forest Service publication
FHTET-2005-16, Morgantown, WV.

112

Crook, D. J., A. Khrimian, J. A. Francese, I. Fraser, T. M. Poland, and V. C.
Mastro. 2007. Chemical ecology of emerald ash borer. Pg. 79. In Mastro et
al. (eds), Emerald Ash Borer Research and Technology Development
Meeting. Cincinnati, OH. 29 October - 2 November 2006. US. Department
of Agriculture, Forest Service publication FHTET—2007—04, Morgantown, WV.

Crook, D. J., A. Khrimian, J. A. Francese, I. Fraser, T. M. Poland, A. J.
Sawyer, and V. C. Mastro. 2008a. Development of a host-based
semiochemical lure for trapping emerald ash borer Agrilus planipennis
(Coleoptera: Buprestidae). Environ. Entomol. 37:356-365.

Crook, D. J., J. Francese, A. Khrimian, I. Fraser, V. Mastro. 2008b. Field
responses of emerald ash borer to multicomponent lures. Pg. 78. In Mastro et
al. (eds) Emerald Ash Borer Research and Technology Development
Meeting, Pittsburg PA. 23-24 October 2008. US. Department of Agriculture,
Forest Service publication FHTET-2008-07, Morgantown, WV.

Dodds, K. J., C. Graber, F. M. Stephen. 2001. Facultative intraguild predation
by larval Cerambycidae (Coloeptera) on bark beetle larvae (Coleoptera:
Scolytidae). Environ. Entomol. 30: 17-22.

Dunn, J. P., T. W. Kimmerer, and G. L. Nordin. 1986a. The role of host tree
condition in attack of white oaks by the twolined chestnut borer, Agilus
bilineatus (Weber) (Coleoptera: Buprestidae). Oecologia (Bed) 70: 596-600.

Dunn, J. P., T. W. Kimmerer and G. L. Nordin. 1986b. Attraction of the
twolined chestnut borer. Agrilus bilineatus (Weber) (Coleoptera: Buprestidae),
and associated borers to volatiles of stressed white oak. Can. Entomol. 118:
503-509.

Dunn, J. P., D. A. Potter, and T. W. Kimmerer. 1990. Carbohydrate reserves,
radial growth, and mechanisms of resistance of oak trees to phloem-boring
insects. Oecologia 83: 458-468.

Elgar, M. A., and B. J. Crespi. (eds). 1992. Cannibalism: ecology and evolution
among diverse taxa. Oxford University Press, Oxford UK.

Eyles, A., W. Jones, K. Riedl, D. Cipollini, S. Schwartz, K. Chan, D. A.
Harms, and P. Bonello. 2007. Comparative phloem chemistry of
Manchurian (Fraxinus mandshurica) and two North American ash species
(Fraxinus americana and Fraxinus pennsylvanica). J. Chem. Ecol. 33: 1430-
1448.

Fayt, P., M. M. Machmer, and C. Steeger. 2005. Regulation of spruce bark

beetles by woodpeckers — a literature review. Forest Ecology and
Management 206: 1-14.

113

Fierke, M. K., and F. M. Stephen. 2008. Callus formation and bark moisture as
potential physical defenses of northern red oak, Quercus rubra, against red
oak borer, Enaphalodes rufulus (Coleoptera: Cerambycidae). Can. Entomol.
140: 149-157.

Flint, T. 2005. Michigan’s emerald ash borer response project. Pg. 6. In Mastro
and Reardon (eds), Emerald Ash Borer Research and Technology
Development Meeting. Romulus, MI. 5-6 October 2004. US. Department of
Agriculture, Forest Service Publication FHTET-2004-15, Morgantown WV.

Fraser, l. and V. C. Mastro. 2007. Emerald ash borer attraction to girdled trees:
effect of placement and timing on attraction. Pg. 82. In Mastro et al. (eds),
Emerald Ash Borer and Asian Longhorned Beetle Research and Technology
Development Meeting. Cincinnati, OH. 29 October — 2 November 2006.
US. Department of Agriculture, Forest Service Publication FHTET-2007-04,
Morgantown WV.

GoIs, R., M. Roosjen, H. Dijkman, and M. Dicke. 2003. Induction of direct and
indirect plant responses by jasmonic acid, low spider mite densities, or a
combination of jasmonic acid treatment and spider mite infestation. J. of
Chemical Ecology 29: 2651-2666.

Gould, J., J. Ayer, Y. Zhong-qi, and W. Xiao-yi. 2007. Host speciﬁcity of
Spathius agrili Yang, a parasitoid of the emerald ash borer. Pp. 65-66. In
Mastro et al. (eds), Emerald Ash Borer and Asian Longhorned Beetle
Research and Technology Development Meeting. Cincinnati, OH. 29 October
to 2 November 2006. US. Department of Agriculture, Forest Service
Publication FHTET-2007-04, Morgantown WV.

Grosman, D. M., S. M. Salom, T. L. Payne. 1992. Laboratory study of
conspecific avoidance by host-colonizing Dentroctonus frontalis Zimm
(Coleoptera: Scolytidae). J. of Insect Behavior 5: 263-271.

Haack, R. A. and D. M. Benjamin. 1982. The biology and ecology of the
twolined chestnut borer, Agrilus bilineatus (Coleoptera: Buprestidae), on
oaks, Quercus spp, in Wisconsin. Can. Entomol. 114: 385-396.

Hanks, L. M., T. D. Paine, and J. G. Millar. 1993. Host species preference and
larval performance in the wood-boring beetle Phoracantha semipunctata F.
Oecologia (Berl.) 95 22-29.

Hanks, L. M., T. D. Paine, J. G. Millar, C. D. Campbell, and U. K. Schuch.
1999. Water relations of host trees and resistance to the phloem boring
beetle Phoracantha semipunctata F. (Coleoptera: Cerambycidae). Oecologia
(Berl.) 119: 400-407.

114

Hanks, L. M., T. D. Paine, and J. G. Millar. 2005. Inﬂuence of the larval
environment on performance and adult body size of the wood-boring beetle
Phoracantha semipunctata Entomologia Experimentalis et Applicata 114: 25-
34.

Harrison, T. 2005. Ohio emerald ash borer update. Pg. 9. In Mastro and
Reardon (eds), Emerald Ash Borer Research and Technology Development
Meeting. Romulus, MI. 5-6 October 2004. US. Department of Agriculture,
Forest Service Publication FHTET-2004-15, Morgantown WV.

Henery, M. L., l. R. Wallis, C. Stone, and W. J. Foley. 2008. Methyl jasmonate
does not induce changes in Eucalyptus grandis leaves that alter the effect of
constitutive defenses on larvae of a specialist herbivore. Oecologia 156: 847-
859.

Hudgins, J. W. and V. R. Franceschi. 2004. Methyl jasmonate-induced
ethylene production is responsible for conifer phloem defense responses and
reprogramming of stem cambial zone for traumatic resin duct formation.
Plant Physiology 135: 2134-2149.

Jordan, M-O., and R. Habib. 1996. Mobilizable carbon reserves in young peach
trees as evidenced by trunk girdling experiments. J. of Experimental Botany:
47: 79-87.

Koplin, J. R. and P. H. Baldwin. 1970. Woodpecker predation on an endemic
population of Engelmann spruce beetles. Am. Midl. Nat. 83: 510 - 515.

Koricheva, J. and S. Larsson. 1998. Insect performance on experimentally
stressed woody plants: a meta-analysis. Annu. Rev. Entomol. 42: 195-216.

Lee, D. R. 1981a. Elasticity of phloem tissues. Journal of Experimental Botany
32: 251-260.

Lee, D. R. 1981 b. Synchronous pressure-potential changes in the phloem of
Fraxinus americana L. Planta 151: 304-308.

Lindell, C. A., D. G. McCullough, D. L. Cappaert, and N. M. Apostolou. 2008.
Factors inﬂuencing woodpecker predation on emerald ash borer. Am. Midl.
Nat. 159: 434-444.

Liu, H., L. 8. Bauer, R. Gao, T. Zhao, T. R. Petrice. and R. A. Haack. 2003.
Exploratory survey for the emerald ash borer, Agn'Ius planipennis
(Coleoptera: Buprestidae), and its natural enemies in China. Great Lakes
Entomol. 36: 191-204.

115

Liu, H. and L. S. Bauer. 2007a. Tetrastichus planipennisi (Hymenoptera:
Euplophidae), a gregarious larval endoparasitoid of emerald ash borer from
China. Pp. 61-62. In Mastro et al. (eds), Emerald Ash Borer and Asian
Longhorned Beetle Research and Technology Development Meeting.
Cincinnati, OH. 29 October to 2 November 2006. US. Department of
Agriculture, Forest Service Publication FHTET-2007-04, Morgantown WV.

Liu, H., L. S. Bauer, D. L. Miller, T. Zhao, R. Gao, L. Song, Q. Luan, R. Jin,
and C. Gao. 2007b. Seasonal abundance of Agrilus planipennis (Coleoptera:
Buprestidae) and its natural enemies Oobius agrili (Hymenoptera: Encyrtidae)
and Tetrastichus planipennisi (Hymenoptera: Eulophidae) in China.

Biological Control 42: 61-71.

Liu, H., L. S. Bauer, T. Zhao, and R. Gao. 2008. Population biology of emerald
ash borer and its natural enemies in China. Pp. 59-60. In Mastro et al.
(eds), Emerald Ash Borer and Asian Longhorned Beetle Research and
Technology Development Meeting. Pittsburg, PA. 23-24 October, 2007. US.
Department of Agriculture, Forest Service Publication FHTET-2008-07,
Morgantown, WV.

Loerch, C. R., and E. A. Cameron. 1984. Within-tree distribution and
seasonality of immature stages of the bronze birch borer, Agrilus anxius
(Coleoptera: Buprestidae). Can. Entomol. 116: 147-152.

Mattson, W. J., and R. A. Haack. 1986. The role of drought in outbreaks of
plant-eating insects. Bioscience 37: 110-118.

McCullough, D. G., T. M. Poland, and D. Cappaert. 2006. Attraction of
emerald ash borer to trap trees: effects of stress agents and trap height. Pp.
61-62 In Mastro et al. (eds), Emerald Ash Borer Research and Technology
Development Meeting. Pittsburg, PA. 26-27 September 2005. US.
Department of Agriculture, Forest Service publication FHTET-2005-16,
Morgantown, WV.

McCullough, D. G., and N. W. Siegart. 2007. Estimating potential emerald ash
borer (Agrilus planipennis Fairmaire) populations using ash inventory. J.
Econ. Entomol. 100: 1577-1586.

McCullough, D. G., T. M. Poland, A. C. Anulewicz, and D. L. Cappaert. 2008.
Double-deckers and towers: emerald ash borer traps in 2007. Pp. 73-75. In
Mastro et al. (eds) Emerald Ash Borer Research and Technology
Development Meeting, Pittsburg PA. 23-24 October 2008. US. Department
of Agriculture, Forest Service publication FHTET-2008-07, Morgantown, VW.

116

 

 

 

McCullough, D. G., T.M. Poland, A. C. Anulewicz, and D. L. Cappaert. 2009.
Emerald ash borer (Agrilus planipennis Fairmaire) (Coleoptera: Buprestidae)
attraction to stressed or baited ash (Fraxinus sp.) trees. Environ. Entomol.
(Submitted)

Nash, R. W., E. J. Duda, and N. H. Gray. 1951. Studies on the extensive
drying, regeneration, and management of birch. Maine Forest Service Bulletin
15.

Ott. R. L., and M. Longnecker. 2001. An introduction to statistical methods and
data analysis. Duxbury Thomson Learning. Paciﬁc Grove, CA.

Poland, T. M., P. de Groot, G. Grant, L. MacDonald, and D. G. McCullough.
2004. Developing attractants and trapping techniques for the emerald ash
borer. Pg. 15-16. In Mastro and Reardon (eds), Emerald Ash Borer
Research and Technology Development Meeting, Port Huron, MI, 30
September-1 October 2003. US. Department of Agriculture, Forest Service
Publication FHTET-2004-02, Morgantown WV.

Poland, T. M., D. G. McCullough, P. D. Groot, G. Grant, and D. L. Cappaert.
2005. Progress toward developing trapping techniques for the emerald ash
borer. Pp. 50-51. In Mastro and Reardon (eds), Emerald Ash Borer
Research and Technology Development Meeting, Romulus, MI, 5-6 October
2004. US. Department of Agriculture, Forest Service Publication FHTET-
2004-15, Morgantown WV.

Poland, T. M., Rodriguez-Saona, G. Grant, L. Buchan, P. de Groot, J. Miller,
and D. G. McCullough. 2006. Trapping and detection of emerald ash borer:
identiﬁcation of stress-induced volatiles and tests of attraction in the lab and
ﬁeld. Pp. 64-65. In Mastro et al. (eds), Emerald Ash Borer Research and
Technology Development Meeting. Pittsburg, PA. 26-27 September 2005.
US. Department of Agriculture, Forest Service publication FHTET—2005-16,
Morgantown, WV.

Poland, TM. and D. G. McCullough. 2007a. Evaluation of a multicomponent
trap for emerald ash borer incorporating color, silhouette, height, texture, and
ash leaf and bark volatiles. Pp. 74-76. In Mastro et al. (eds), Emerald Ash
Borer Research and Technology Development Meeting. Cincinnati, OH. 29
October - 2 November 2006. US. Department of Agriculture, Forest Service
publication FHTET-2007-04, Morgantown, WV.

117

Poland, T. M., D. S. Pureswaran, G. Grant, and P. de Groot. 2007b. Field
attraction of emerald ash borer to antennally and behaviorally active ash
volatiles. Pp. 80-81. In Mastro et al. (eds), Emerald Ash Borer Research and
Technology Development Meeting. Cincinnati, OH. 29 October — 2
November 2006. US. Department of Agriculture, Forest Service publication
FHTET-2007-04, Morgantown, WV.

Nash, R. W., E. J. Duda, and N. H. Gray. 1951. Studies on the extensive
drying, regeneration, and management of birch. Maine Forest Service Bulletin
15.

Rebek, E. J., D. A. Herms, and D. R. Smitley. 2007. lnterspecific variation in
resistance to emerald ash borer (Coleoptera: Buprestidae) among North
American and Asian ash (Fraxinus spp.). Environ. Entomol. 37: 242-246.

Rodriguez-Saona, C. S., J. Crafte-Brandner, P. W. Pare, and T. J.
Henneberry. 2001. Exogenous methyl jasmonate induces volatile emissions
in cotton plants. J. of Chemical Ecology 27: 679-694.

Rodriguez-Saona, C. S., J. S. Thaler. 2005. The jasmonate pathway alters
herbivore feeding behaviour: consequences for plant defences. The
Netherlands Entomological Society Entomolodia Experimentalis et Applicata
1 15: 125-134.

Rodriguez-Saona, C. S., T. M. Poland, J. R. Miller, L. L. Stelinski, G. G.
Grant, P. de Groot, L. Buchan, and L. MacDonald. 2006. Behavioral and
electrophysiological responses of the emerald ash borer, Agrilus planipennis,
to induced volatiles of Manchurian ash, Fraxinus mandshunca.
Chemoecology 16: 75-86.

Rodriguez-Saona, C. 8., JR. Miller, T. M. Poland, T. M. Kuhn, G. W. Otis, T.
Turk, D. L. Ward. 2007. Behaviors of adult Agrilus planipennis (Coleoptera:
Buprestidae). The Great Lakes Entomologist 40: 1-16.

Santamour, F. S. 1990. Rhododendrol and susceptibility to the bronze birch
borer. Journal of Arboriculture 16: 260-263.

SAS Institute. 2003. PROC user’s manual, version 9.1 SAS Institute, Cary, NC.

Schaefer, P. W. 2005. Foreign exploration for emerald ash borer and its natural
enemies. Pp. 67-68. In Mastro and Reardon (eds), Emerald Ash Borer
Research and Technology Development Meeting. Romulus, MI, 5-6 October
2004. US. Department of Agriculture, Forest Service publication FHTET-
2004-15, Morgantown WV.

118

Shapiro, 8.3., and M. B. Wilk. 1965. An analysis of variance test for normality,
Biometrika 52: 591-599.

Siegert, N. W., and D.G. McCullough. 2005. Reconstructing the temporal and
spatial dynamics of emerald ash borer in black ash: a case study of an outlier
site in Roscommon County, Michigan. Pp. 21-22. In Mastro and Reardon
(eds), Emerald Ash Borer Research and Technology Development Meeting,
Romulus, MI, 5-6 October 2004. US. Department of Agriculture, Forest
Service Publication F HTET-2004-15, Morgantown WV.

Siegert, N. W., and D. G. McCullough and A. R. Tluczek. 2007. Two years
under the bark: towards understanding multiple-year development of emerald
ash borer larvae. Pg. 20. In Mastro et al. (eds), Emerald Ash Borer and
Asian Longhorned Beetle Research and Technology Development Meeting.
Cincinnati, OH, 29 October to 1 November 2006. US. Department of
Agriculture, Forest Service Publication FHTET-2007-04, Morgantown WV.

Storer, A. J., E. E. Graham, M. D. Hyslop, R. L. Heyd. 2005. Michigan emerald
ash borer detection survey. Pp. 7-8. In Mastro and Reardon (eds), Emerald
Ash Borer Research and Technology Development Meeting. Romulus, MI.
5-6 October 2004. US. Department of Agriculture, Forest Service Publication
F HTET-2004-1 5, Morgantown WV.

Sugiura, N. 1999. The family Buprestidae in Fukushima Prefecture: the genus
Agrilus
(http://www1 .linkclgb.or.igl~sugirinIfukusima/nagatama/nagatama2.html.

Svihra, P. and C. S. Koehler. 1993. Flatheaded borer in white alder landscape
trees. J. of Arboriculture 19: 260-265.

Taylor, A. D., J. L. Hayes, L. Roton, and J. C. Moser. 1992. A phloem
sandwich allowing attack and colonization by bark beetles (Coleoptera:
Scolytidae) and associates. J. Entomol. Sci 27: 311-316.

Tluczek, A. R., D. G. McCullough, T. M. Poland, and A. C. Anulewicz. 2008.,
Effects of host stress on emerald ash borer development: what makes a good
home? Pp. 32-33. In Mastro et al. (eds) Emerald Ash Borer Research and
Technology Development Meeting, Pittsburg PA. 23-24 October 2008. US.
Department of Agriculture, Forest Service publication FHTET-2008-07,
Morgantown, WV.

Wallin, K. F. and K. F. Raffa. 2004. Feedback between individual host selection
behavior and population dynamics in an eruptive herbivore. Ecological
Monographs 74: 101-116.

119

 

Ware, V. L. and F. M. Stephen. 2006. Facultative intraguild predation of red oak
borer larvae (coleoptera: cerambycidae). Environ. Entomol. 35: 443-447.

Wei, X., Y. Wu, R. Reardon, T. Sun, M. Lu, and J. Sun. 2007. Biology and
damage traits of emerald ash borer (Agn'Ius planipennis Fairmaire) in China.
Insect Science. 14: 367-373.

Wermelinger, B., P. F. Fluckiger, M. K. Obrist, and P. Duelli. 2007. Horizontal
and vertical distribution of saproxylic beetles (Col., Buprestidae,
Cerambycidae, Scolytinae) across sections of forest edges. J. Appl. Entomol.
131: 104-114.

Williams, D., H. P. Lee, and Y. S. Jo. 2005. Exploration for natural enemies of
emerald ash borer in south Korea during 2004. Pg. 66. In Mastro and
Reardon (eds), Emerald Ash Borer Research and Technology Development
Meeting. Romulus, MI. 5-6 October 2004. US. Department of Agriculture,
Forest Service Publication FHTET-2004-15.

Williams, D., H. P. Lee, and Y. S. Jo. 2006. Exploration for emerald ash borer
and its natural enemies in South Korea during May-June 2005. Pg. 52. In
Mastro et al. (eds), Emerald Ash Borer Research and Technology
Development Meeting. Pittsburgh, PA, 26-27 September 2005. US.
Department of Agriculture, Forest Service Publication FHTET—2005-16.

Winfolia. 2004a. Winfolia for leaf analysis, user’s manual, Regent Instruments
Inc., www.renetinstruments.com, Quebec, Canada.

Yang, Z. 0., J. S. Strazanac, P. M. Marsh, C. Van Achtergerg, and W. Y.
Choi. 2005. First recorded parasitoid from China of Agrilus planipennis. A
new species of Spathius (Hymenoptera: Braconidae: Doryctinae). Annals of
the Entomological Society of America. 98: 636-642.

Yang, 2. Q., J. S. Strazanac, Y. X. Yao, and X. Y Wang. 2006. A new species
of emerald ash borer parasitoid from China belonging to the genus
Tetrastichus planipennisi (Hymenoptera: Euplophidae). Proceedings of the
Entomological Society of Washington. 108: 550-558.

Yu, C. 1992. Agn'lus maroopoli Obenberger, pp. 400-401. In G. Xia (ed.), Forest
insects of China, 2nd ed. China Forestry Publishing. Beijing, China.

Zalucki, M. P., A. R. Clarke, and S. B. Malcolm. 2002. Ecology and behavior of
ﬁrst instar larval Lepidoptera. Annu. Rev. Entomol. 47: 361-393.

Zhao, T., R. Gao, H. H. Liu, L. S. Bauer, and L. Sun. 2005. Host range of

emerald ash borer, Agrilus planipennis Fairmaire, its damage and
countermeasures. Acta Entomol. Sin. 48: 594-59.

120

  

   
  

 
 

AAAAAAAAAAA

lllllljljlllljljjljlljjjljlll“

31 30