.3.
.._ 252m.
«MEMHE 231

“a”? r urn“. a fist... u...
.5}... 11.2,...qu ..| . I
a. :«uiﬁutnﬁ. .ﬁﬁwwﬂ: .ﬁ...“ 2.3.”.

.FINPJ. but}!!! 31.4%.]...

. .r.. .4}
a

arr!!!-

r. s
.5 Warm...”

a 20......r I
yrs-tuawﬂndﬂuaﬁﬂ

‘11.... I
:1. 1.11....
”31 :21...

 

3:. . .
{1.3.3.1112 9.3..
1.1.1:] .1). x :4

 

, . , . 9.2.4???” .1921)”.

 

./J

This is to certify that the

thesis entitled

ALTERNATIVE CONTROLS FOR TWO IMPORTANT
INSECT PESTS OF CHRISTMAS TREES

presented by
Kirsten Marie Fondren

has been accepted towards fulfillment
of the requirements for

M° 3- degree in JEE‘MLQEX

QJ [HM/74f

 

Major} ofessor

Date 39/315: 2002.

0-7639 MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State

University

 

 

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

 

W

PR 1 1 2334

 

‘30? 04

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDue.p65-p.15

ALTERNATIVE CONTROLS FOR TWO IMPORTANT INSECT PESTS OF
CHRISTMAS TREES

By

Kirsten Marie Fondren

A THESIS

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

MASTER OF SCIENCE
Department of Entomology

2002

ABSTRACT

ALTERNATIVE CONTROLS FOR TWO IMPORTANT INSECT PESTS OF
CHRISTMAS TREES

By

Kirsten Marie F ondren

Balsam twig aphid (Mindarus abietinus Koch) and pine needle scale (Chionaspis
heterophyllae (Cooley)) are two important pests of Christmas trees that are typically
controlled with broad-spectrum insecticides. Our objectives were to evaluate potential
alternative control methods for each insect and describe their biology on Michigan

Christmas tree plantations. The phenology of both insects was related to accumulated

degree-days base 50°F. Balsam twig aphid began to hatch at approximately 70 DD50, and

produced the second generation at approximately 150 DDSO. Flushing time of the host

plant had a signiﬁcant effect on aphid damage levels. Larvae of Chrysoperla ruﬁlabris
Burmeister reduced the population of M. abietinus when applied in the ﬁeld.
Economically, M. abietinus damage did not affect retail or wholesale value until it
reached approximately 30%. The second generation of pine needle scale hatched at
approximately 1280 DD50 and continued for three weeks. The coccinellids Chilocorus
stigma (Say) and Microweisia misella (LeConte) were important scale predators. We
tested horticultural spray oil and found it worked as well as an organophosphate

insecticide when applied at the peak of the second instar, or approximately 1500 DD50.

To Deb

iii

ACKNOWLEDGEMENTS

I can only begin to thank the many people who gave freely of their time, talent
and wisdom to help me along the way. So many people helped in big and small ways to
make this project become a reality, and I am grateful to them all. Most of all to my
advisor, Dr. Deborah G. McCullough, who often kept me going by strength of will alone;
and to all the Entomology graduate students who gave me so much support just by being
there. They may never know how much a kind word of encouragement can do. Dr. Ed
Graﬁus went above and beyond his role as a member of my committee to encourage me
and stand with me as a friend. Dr. Chris DiFonzo helped me learn the ﬁne points of
working with aphids, and inspired me with her constant energy and enthusiasm. Dr. Mel
Koelling patiently answered my questions and resigned himself to letting me destroy his
Christmas trees. His dry sense of humor and realistic approach always helped me to gain
perspective. The support staff of the Department of Entomology helped in so many small
but important ways to make things go smoothly.

Of course this project would not have been possible without the support and
cooperation of Michigan’s Christmas tree growers: Jody Wilkes of Happy Holiday Tree
Farms, Dan Wahmhoff of Wahmhoff Farms, Gary and Mike of Specialty Tree Farms,
Richard Burdo of Polar Equator Tree Farms, Richard Farnsworth of F amsworth Tree
Farms, John Bevier of B&G Tree Farms, Mike Morin of Hramour Nursery, and last but
certainly not least Mel and Laurie Koelling of Tannenbaum Farms. Thanks to Jill

O’Donnell for helping me locate growers and obtain permission to use their ﬁelds. Each

iv

was kind enough to let me work in their ﬁelds, in some cases modifying their busy
schedules to accommodate my inconvenient requests, and even pitching in to help. Also,
Greg Kowalewski and the staff of WK. Kellogg Forest, and Randy Klevickas and the
staff of the Tree Research Center, let me infest their trees with additional pests.

Anne Henderson and Abigail Sommers helped with many time-consuming tasks
in the ﬁeld and laboratory. For help with identiﬁcations, I am grateful to Gary Parsons,
David Voegtlin, Douglass Miller, Jim Stimmell, John Luhman and Michael Gates.
Thanks to Sinthya Penn of Beneﬁcial Insectary for advice. Thanks also to Jeff Andresen
and the Michigan State Agricultural Weather Station website. Thanks to David Wright
for advice on registering the customer surveys with the University Committee on
Research Involving Human Subjects, IRB# 99-660. This project was supported by a
grant from Project GREEEN, for which I am especially appreciative.

Eileen Eliason (now Buss) deserves a special thanks for her words of
encouragement, both professionally and personally. Linda Williams always made me
smile. Joe Zeleznik was both a counselor and a mentor, and an unwitting role model.
The MSU Counseling Center was an invaluable help.

Of course my family and friends deserve a medal for their support, love and
patience, especially my three-member fan club: my little niece Morgyn helped me ﬁnd
joy in life and see the important things, my mom was always eager to do what she could
to help, and Mr. Mike Robinson was an unshakable pillar of support. I thank God for this

opportunity.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................................. ix
LIST OF FIGURES ............................................................................................................... xi
CHAPTER 1

PHENOLOGY AND IMPACT OF THE BALSAM TWIG APHID (MINDAR US
ABIE T IN US KOCH) (HOMOPTERA: APHIDIDAE) ON FIR CHRISTMAS TREE

PLANTATIONS .................................................................................................................... 1
Abstract .................................................................................................................................. 1
Introduction ............................................................................................................................ 2
Methods .................................................................................................................................. 5
Study sites .................................................................................................................. 5
Balsam twig aphid phenology .................................................................................... 6
Tree phenology .......................................................................................................... 9
Balsam twig aphid damage ........................................................................................ 12
Economic impact ....................................................................................................... 12
Statistical Analysis ..................................................................................................... 14
Results .................................................................................................................................... 1 5
Balsam twig aphid phenology .................................................................................... 15
Tree phenology .......................................................................................................... 18
Balsam twig aphid damage ........................................................................................ 22
Economic impact ....................................................................................................... 24
Discussion .............................................................................................................................. 26
Literature Cited ...................................................................................................................... 31
Tables .................................................................................................................................... 35
Figures .................................................................................................................................... 39
CHAPTER 2

POTENTIAL FOR AUGMENTATIVE BIOLOGICAL CONTROL OF THE BALSAM
TWIG APHID (MINDAR US ABIE T IN US KOCH) (HOMOPTERA: APHIDIDAE) IN

MICHIGAN CHRISTMAS TREE PLANTATIONS ............................................................ 50
Abstract .................................................................................................................................. 50
Introduction ............................................................................................................................ 51
Methods .................................................................................................................................. 54
Study sites .................................................................................................................. 54
Objective 1. Natural enemy complex ........................................................................ 54
Objective 2. Evaluation of C. ruﬁlabris in lab and ﬁeld cages ................................. 57
Objective 3. Field test of C. ruﬁlabris ...................................................................... 59
Statistical Analysis ..................................................................................................... 61
Results .................................................................................................................................... 61
Objective 1. Natural enemy complex ........................................................................ 61
Objective 2. Evaluation of C. ruﬁlabris in lab and ﬁeld cages ................................. 64

vi

Objective 3. Field test of C. ruﬁlabris ...................................................................... 65

Discussion .............................................................................................................................. 67
Objective 1. Natural enemy complex ........................................................................ 67
Objective 2. Effectiveness of C. ruﬁlabris in lab and ﬁeld cages ............................ 68
Objective 3. Open ﬁeld release of C. ruﬁlabris ........................................................ 70
Practical applications ................................................................................................. 71

References Cited .................................................................................................................... 73

Tables .................................................................................................................................... 78

Figures .................................................................................................................................... 82

CHAPTER 3

PHENOLOGY AND NATURAL ENEMIES OF THE PINE NEEDLE SCALE
(CHIONASPIS PINIFOLIAE (FITCH))(HOMOPTERA: DIASPIDIDAE) IN

CHRISTMAS TREE FIELDS ............................................................................................... 90
Abstract .................................................................................................................................. 90
Introduction ............................................................................................................................ 91
Materials and Methods ........................................................................................................... 95
Field sites ................................................................................................................... 95
Objective 1. Phenology of the second generation ...................................................... 96
Objective 2. Natural enemies ..................................................................................... 97
Objective 3. Rates of predation and parasitism ........................................................ 98
Statistical Analysis ..................................................................................................... 98
Results .................................................................................................................................... 99
Objective 1. Phenology of the second generation ...................................................... 99
Objective 2. Natural enemies ..................................................................................... 101
Obj ective3. Rates of predation and parasitism .......................................................... 102
Discussion .............................................................................................................................. 103
References Cited .................................................................................................................... 107
Tables .................................................................................................................................... 111
Figures .................................................................................................................................... 121
CHAPTER 4

POTENTIAL EFFICACY OF HORTICULTURAL OIL FOR CONTROL OF PINE
NEEDLE SCALE (CHIONASPIS PINIFOLIAE (FITCH)) (HOMOPTERA:

DIASPIDIDAE) IN CHRISTMAS TREE FIELDS .............................................................. 123
Abstract .................................................................................................................................. 123
Introduction ............................................................................................................................ 124
Materials and Methods ........................................................................................................... 128
Study sites .................................................................................................................. 128
Objective 1. Phenology of the second generation ...................................................... 129
Objective 2. Efﬁcacy of horticultural oil ................................................................... 130
Objective 3. Effectiveness of commercial application .............................................. 132
Statistical Analyses .................................................................................................... 133
Results .................................................................................................................................... 133
Objective 1. Phenology of second generation scales ................................................. 133
Objective 2. Efﬁcacy of horticultural oil ................................................................... 134

vii

Objective 3. Effectiveness of commercial application of horticultural oil ................ 135

Discussion .............................................................................................................................. 136
References Cited .................................................................................................................... 141
Tables .................................................................................................................................... 145
Figures .................................................................................................................................... 149
APPENDIX 1. RECORD OF DEPOSITION OF VOUCHER SPECIMENS ..................... 155

viii

LIST OF TABLES

Chapter 1.

Table 1. Correlations of density of Mindarus abietinus ﬁmdatn'ces with density of
sexuparae in subsequent weeks ................................................................... 35

Table 2. Comparisons between early and late budbreaking trees in Ingham
County. Aphid numbers found on each group were compared using
pooled t tests or Welch tests. Amount of aphid damage (percentage of
shoots damaged) sustained by each group of trees was compared using a
Welch test .................................................................................................... 37

Table 3. Results of chi-square tests of customer preference, 1999 surveys ............. 38

Chapter 2.
Table 1. Potential predators of M. abietinus collected in the ﬁeld in 1999 by hand
or beat samples, from 14 April to 30 June .................................................. 78
Table 2. Adult predators found on the sticky traps in each ﬁeld, 2000 ..................... 79
Table 3. Species of Coccinellidae found on sticky traps in each ﬁeld, 2000 ............ 81
Chapter 3.
Table 1. Study sites used in 1999, 2000 and 2001 for pine needle scale .................. 111
Table 2. Number of eggs present under- female scale armor during oviposition
and hatching period ..................................................................................... 112
Table 3. Pine needle scale phenology, Van Buren County, 1999 (means i 1 SE
where applicable) ........................................................................................ 115
Table 4. Pine needle scale phenology, Montcalm County, 1999 (means :h 1 SE
where applicable) ........................................................................................ 116
Table 5. Pine needle scale phenology 2000 and 2001 .............................................. 117
Table 6. Parasitoids recovered from scale mummies in 2000 .................................. 118

ix

Table 7.

Table 8.

Chapter 4.
Table 1.

Table 2.

Percent predation of pine needle scale in each life stage, Van Buren
County 1999. See Table 3 for the percentage of the population that was
in each life stage on each sample date ........................................................ 119

Percent predation of pine needle scale at each life stage, Montcalm

County, 1999. See Table 4 for the percentage of the population that
was in each stage on each sample date ....................................................... 120

Varieties, ages and numbers of trees used in each trial .............................. 145

Number of eggs present under female scale armor during oviposition
and hatching period ..................................................................................... 146

Chapter 1.

Figure 1.

Figure 2.

Figure 3.

Figure 4.

LIST OF FIGURES

Percentage of M. abietinus eggs hatched in Ingham County in March

2000 (number hatched/100 eggs counted). Cumulative degree days

base 50 degrees F (10 degrees C) are indicated for 21 March, 23 March,
and 30 March ............................................................................................. 39

Phenology of the ﬁrst and second morphs of M. abietinus in 2a) Ingham
and 2b) Grand Traverse Counties, 1999 .................................................... 40

Phenology of the ﬁrst and second morphs of M. abietinus in 3a)
Ingham, 3b) Grand Traverse, and 3c) Antrim Counties, 2000 .................. 41

Phenology of the ﬁrst and second morphs of M. abietinus in 4a)
Ingham, 4b) Grand Traverse, and 4c) Antrim Counties, 2001 .................. 42

Figure 5. Percentage of damaged shoots vs. approximate date of budbreak,

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Ingham Co., 2000 (means +/- SE) ............................................................. 43

Percentage of shoot expansion in early spring for paired early and late
budbreaking trees in Ingham County (n = 12 trees). Pearson’s product-
moment correlation analysis: p < 0.0001, r = 0.92 .................................... 44

Percentage of shoot expansion in 1999 and 2000 for ten randomly
selected trees in Ingham County. Pearson’s product-moment
correlation analysis: p < 0.0001, r = 0.97 .................................................. 45

Percentage of shoot expansion in early spring for the aphid sample trees
in Ingham County (11 = 30 trees measured in both years). Seven trees
were excluded because of missing values. Spearman’s rank correlation

analysis: p = 0.12, rS = 0.33 ....................................................................... 46

Results of 1999 retail customer surveys. The y-axis indicates the

number of customers who stated that they would consider buying each
tree. Within each height group, damage was a signiﬁcant factor (see
Table 3) unless marked with n.s. ............................................................... 47

Figure 10. Results of 2000 customer surveys (n = 50 customers). Trees were

grouped into three categories based on the percentage of damaged
shoots/ 300 shoots examined per tree. The percentage of damaged
shoots is presented as the actual value. Damage categories are

xi

indicated by ‘Light’ (n = 3 trees), ‘Medium’ (n = 2 trees), and ‘Heavy’
(n = 3 trees). The percentage of change in perceived value is presented
as mean +/- 1 SE ........................................................................................ 48

Figure 11. Wholesale grades assigned to trees varying in aphid damage used

Chapter 2.
Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Chapter 3.

Figure 1.

Figure 2.

in 1999 survey (11 = 18). Percentage of damaged shoots is presented as
mean +/- SE. No signiﬁcant differences were found (F = 1.38; df =
2,15; p = 0.28). .......................................................................................... 49

Mindarus abietinus and syrphid larvae counts in clipped branch tip
samples, 1999 ............................................................................................ 83

Numbers of coccinellids, lacewings, and syrphids found in beat and

visual examinations of sample trees, 1999. In Antrim Co., no survey

was done on 2 June because of rain. On 15 June, a survey was done but
no predators were observed ....................................................................... 84

Predators found in beat samples, 2000. Data are presented as the
number of each taxon collected per sample date ....................................... 85

Number of M. abietinus eggs per cm of current-year shoot grth
(mean i 1 SE) on branch tips from ﬁeld cages in 2000. A) Heavily
infested branches (n = 26). B) Moderately infested branches (n = 23) ....86

Mean (:1: 1 SE) number of aphid eggs/cm2 of current-year foliage in

2001 ﬁeld cages. Letters above bars indicate signiﬁcant differences.

A) ‘closed’ cages were kept closed before lacewings were added; B)
‘open’ cages had one end left open enabling access by predators until
lacewings were added ................................................................................ 87

Mean (i 1 SE) aphid counts before and after the lacewing applications
in the open ﬁeld releases in 2001. Data within a column marked with
different letters are signiﬁcantly different (p = < 0.05) ............................. 88

Mean (:1: 1 SE) density of M. abietinus eggs on new foliage in open
releases of C. ruﬁlabris larvae in 2001 ..................................................... 89

Mean (:1: SE) number of eggs present beneath the female scale armor in
2000 in the Van Buren and Ingham County ﬁelds on each sample date...121

Number and stage of Chilocorus stigma and Microweisia misella

xii

Chapter 4.

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

observed in our 1999 ﬁeld study. The total number of each observed
on ten trees in Van Buren County and six trees in Montcahn County ...... 122

Percentage of scale population in the 2nd instar, or hyaline stage, in each
ﬁeld. A) Montcalm and Van Buren Counties, 1999. B) Ingham and

Van Buren Counties, 2000 and 2001. Dotted lines indicate

approximately 1500-1600 degree days. Data were not available for

every sample date ...................................................................................... 149

Mean (:t SE) pine needle scale mortality in the Van Buren County ﬁeld

in 2000. Treatments were applied on 25 July after pretreatment

samples were taken. Signiﬁcant differences among treatment groups

on each date are indicated by different letters. N = 10 trees per

treatment .................................................................................................... 150

Mean (:1: SE) pine needle scale mortality in the Ingham County ﬁeld in
2000. Treatments were applied on 31 July 2000 after pretreatment

samples were taken. There were no signiﬁcant differences among
treatment groups on any date. N = 10 trees per treatment ........................ 151

Pine needle scale mortality in Van Buren County, 2001. Treatments

were applied on 6 August 2001. Means were separated with Fisher’s

LSD where the global ANOVA was signiﬁcant (p < 0.05). Means with
different letters are signiﬁcantly different. Data are presented as mean

:1; 1 SE ........................................................................................................ 152

Percentage of mortality (mean :t 1 SE) in the Montcalm County ﬁeld in
2001. Treatments were applied on 17 May and repeated on 29 May

2001. There were no signiﬁcant differences among treatment groups
(Kruskal-Wallis; n = 4) except on 4 June. Signiﬁcant differences on 4

June are marked with different letters. .................................................... 153

xiii

Chapter 1

Phenology and Impact of the Balsam Twig Aphid (Mindarus abietinus Koch)

(Homoptera: Aphididae) on Fir Christmas Tree Plantations

Abstract. The balsam twig aphid (Mindarus abietinus Koch) is a major insect
pest of balsam and Fraser ﬁrs grown for Christmas trees. In this study, our objectives
were to: 1) monitor the phenology of M. abietinus in ﬁr plantations; 2) assess
relationships among aphid density, tree phenology, and damage to tree foliage; and 3)
develop an aesthetic injury level for M. abietinus on Christmas trees. We monitored the
phenology of M. abietinus and ﬁr trees on three commercial Christmas tree plantations in
central and northern Lower Michigan for three years (1999-2001). Phenology of M.

abietinus was strongly correlated with accumulated degree-days base 50°F (10°C). The

ﬁrst generation matured at approximately 150 DD50 and the second generation began to

occur at approximately 150-200 DD50. In each year, trees that broke bud approximately

one week later than most other trees in the same ﬁeld escaped aphid damage. The rate of
shoot expansion in early spring was often positively correlated with the amount of aphid
damage. We surveyed retail customers at a choose-and-cut plantation in two years to
determine the level of aphid damage that affected the retail value of trees. Customers did
not consistently differentiate between trees with light or moderate damage. Very heavy
damage (mean of 50% damaged shoots) did affect customer perception. Wholesale

grades were assigned to speciﬁc trees with varying levels of M. abietinus damage. Light

to moderate aphid damage (less than 50% of shoots with affected needles) was not a

critical factor in customer choice or wholesale grade.

Introduction

Christmas tree production is a major agricultural industry in Michigan. Roughly
21,853 ha (54,000 acres) in Michigan are in commercial Christmas tree production, and
approximately 3.2 million trees are sold annually, with sales totaling $42.5 million in
1999 (M1 Ag. Stat. Serv. 2000). Managing and preventing insect damage is critical for
maintaining the aesthetic and economic value of Christmas trees. Currently, insect pest
management in Christmas tree plantations is based primarily on use of broad-spectrum
insecticides (McCullough and F ondren 1998) and relatively little research has addressed
alternative controls. Changes in pesticide availability resulting ﬁ'om the federal Food
Quality Protection Act of 1996 may have major impacts on the availability of broad-
spectrum insecticides for minor-use commodities such as Christmas trees (DiFonzo and
McCullough 1998).

Production of ﬁr Christmas trees such as balsam ﬁr (Abies balsamea (L.) Mill.)
and Fraser ﬁr (A. fraseri (Pursh) Poir.) has increased dramatically in recent years. The
area of Fraser ﬁr planted in Michigan has increased by 92% since 1994 (MI Ag. Stat.
Serv. 1998). The high economic value of ﬁrs (US $34-45 retail, $13-19 wholesale)
compared to more traditional species such as Scotch pine (US $21 retail, $8 wholesale) is

increasing the interest in growing ﬁrs and, correspondingly, in pest management in ﬁr

plantations (MI Ag. Stat. Serv. 2000). Fraser ﬁr is one of the most proﬁtable Christmas
tree species produced in Michigan (Jones et a1. 1999).

The balsam twig aphid is a major pest affecting balsam and Fraser ﬁrs, especially
when grown as Christmas trees (Kleintjes 1997a, 1997b, Nettleton and Hain 1982,
Bradbury and Osgood 1986). Fraser and balsam ﬁr Christmas trees have relatively few
insect pests other than M. abietinus (Rather and Mills 1989, McCullough et a1. 1998). In
a 1998 survey of Michigan Christmas tree growers, more insecticide sprays were used on
ﬁrs for control of balsam twig aphid (Mindarus abietinus Koch) than for any other pest
(McCullough and F ondren 1998).

The life cycle of M. abietinus typically begins before budbreak in early spring,
when stern mothers hatch from the overwintering eggs. They do relatively little feeding
and cause little if any damage to the tree (Varty 1966). At maturity they viviparously
produce the second generation (sexuparae), which form colonies that feed on sap in
newly expanding needles (V arty 1966, Bradbury and Osgood 1986, Kleintjes 1997a).
Feeding by the sexuparae causes most of the needle damage. The second generation
typically matures into alate (winged) females, which disperse and produce the ﬁnal
generation of males and oviparous females. A small percentage (6% in Varty 1966) of
the second generation does not develop wings and produces more sexuparae, adding
another generation of parthenogenic females that will become alates (Varty 1966,
Nettleton and Hain 1982). In either case, the cycle is completed by mid-June to mid-July
(Rather and Mills 1989).

High populations of M. abietinus sexuparae cause current-year needles to become

curled and distorted, reducing needle biomass and consequently tree growth (Saunders

1969, Bradbury and Osgood 1986, DeHayes 1981, Berthiaurne et a1. 2000). However,
trees can outgrow up to 55% of the aphid damage by the end of the season, and typical
shearing practices remove some of the damage (N ettleton and Hain 1982). Failure to
account for these factors can lead growers to overestimate the amount of aphid damage
that will be present at harvest time.

Insecticide sprays are often applied only after damaged shoots are observed
(Kleintjes 1997a, 1997b). Once the second generation (sexuparae) appears, however, it is
too late to apply control measures, because the aphids are protected within the new
grth and damage has already occurred (Kleintjes 1997a, Berthiaurne et a1. 2001).
Application of insecticides at this time may kill some aphids and their natural enemies,
but does not reduce needle damage (Kleintjes et a1. 1999).

Host plant resistance to M. abietinus has been suggested on the basis of budbreak
date (Desrosiers 1998, Carter and Nichols 1985). Most observers suggest that if the trees
have not yet broken bud when the sexuparae begin to feed, they will not have new
nutrient-rich growth to feed on and may not survive. The role of genetic control in the
date of budbreak, or initiation of shoot growth in early spring, may be used to selectively
propagate resistant cultivars. Other factors that may indicate potential host plant
resistance include monoterpene levels (DeHayes 1981) and the provenance of the seed
source (DeHayes 1981, Mattson et a1. 1989).

Several authors have reported that M. abietinus reduces the economic value of
Christmas trees, but data supporting this observation is scarce (Saunders 1969, Bradbury
and Osgood 1986, Kleintjes 1997a, Kleintjes et a1. 1999). The exact amount of M.

abietinus damage that results in economic loss has not been determined, although other

workers have estimated it (Kleintj es et a1. 1999). To identify an economic or aesthetic
injury level, the relationships among M. abietinus numbers, the amount of tree damage,
and the resulting loss of economic value need to be deﬁned (Raupp et a1. 1988, Pedi go et
a1. 1986). Developing an action threshold would be an important addition to improving
M. abietinus management.

Our objectives in this study were to: 1) monitor the phenology of M. abietinus in
ﬁr plantations; 2) assess relationships among aphid density, tree phenology, and damage
to tree foliage; and 3) develop an aesthetic injury level for M. abietinus on Christmas

trees.

Methods

Study sites. We monitored the phenology of M. abietinus and associated host
trees in three commercial ﬁelds in northern and central Lower Michigan in 1999-2001.
Fields were located in Ingham County (42°44’N, 84° 33’W), Grand Traverse County
(44°32’N, 85°31’W), and Antrim County (44°59’ N, 85°06’W), and consisted of either
balsam ﬁr or a mixture of balsam and Fraser ﬁr. In 1999, the trees in each ﬁeld were
approximately 8-9 years old. No insecticides, fertilizers or irrigation were used in any
ﬁeld during our study. All ﬁelds were planted to a standard 1.83 m by 1.83 m (6 ft by 6
ft) spacing. The Ingham County ﬁeld vegetation was dominated by herbaceous plants
including crabgrass (Digitaria spp.) and poison ivy (Rhus radicans). The Grand Traverse

County ﬁeld was located on sandy soil with ground cover such as mosses and herbaceous

plants including F ragaria virginiana and Rubus spp. In Antrim County, the plantation
had a thick ground cover composed primarily of crabgrass (Digitaria spp.) and other
grasses (Setaria spp., Bromus spp., Echinochloa spp., etc.).

Balsam twig aphid phenology. Our initial goal was to monitor the phenology of
M. abietinus development weekly in each ﬁeld and year, beginning with egg hatch in
March and continuing until the end of oviposition in late June or early July. However,
logistics made complete observations impossible for every ﬁeld and year, so we focused
on the aspects of M. abietinus phenology that have the most bearing on timing control
measures, i.e. the timing of egg hatch, the duration of the ﬁrst generation (stem mothers)
and the beginning of the second generation (sexuparae).

Egg hatch. In 1999, we began to monitor the aphid population after hatching had
already occurred, so the exact duration of egg hatch could not be recorded. However, the
aphids were still in the ﬁrst generation (stern mother) stage when monitoring began, so
we could estimate the date and degree-day accumulation where egg hatch had been

completed. In the Ingham County ﬁeld in 2000, we were able to monitor egg hatch daily
ﬁ'om 21 March (74 DD50) to 30 March (96 DD50) by counting the percentage of egg

hatch (hatched eggs/ 100 eggs) on branch tips clipped at random from the midcrown level
of ten trees throughout the ﬁeld. Egg hatch could be determined in the laboratory by
examining the branch tips under a microscope and looking for ﬂattened or torn chorions.
Unhatched eggs were turgid and would leak when poked with a minuten pin. In Grand

Traverse County and Antrim County in 2000, we determined the percentage of egg hatch

on 28 March only (50 DD50). Egg hatch was not monitored in 2001.

First and second generations. In 1999, we selected 40 trees in each of the
Ingham, Grand Traverse, and Antrim County ﬁelds in early spring (13 April in Grand
Traverse County, 15 April in Ingham County, 20 April in Antrim County) on the basis of
aphid damage ﬁ'om the previous season, to use for monitoring the phenology of the aphid
population. These 40 trees were tagged and used throughout the three years of the study
for aphid sampling. In 1999, ten of these trees were randomly selected each week and
two branch tips were clipped on either side at midcrown level. The midcrown level was
chosen because M. abietinus tends to colonize the midcrown level most consistently
(N ettleton and Hain 1982, Varty 1966). Clipped branch tips were transported to the lab
in coolers and placed in 70% ETOH. All branch tips were dissected under a microscope
in the laboratory to remove all aphids (per Varty 1966, 1968). Although this method
allowed us to get a close approximation of the actual numbers of aphids on a given
branch tip, it was time-consuming (aphid numbers could exceed 600 per branch tip),
destructive, and could be problematic in active Christmas tree plantations where shape
and fullness are critical. Therefore, we decided to use an additional sample method that
could be used each week on the same trees without being destructive.

In May 1999, we developed a non-destructive method (the ‘beat method’, also
used by Kleintjes et a1. 1999) to sample aphids from a random sample of ten trees in each
ﬁeld, separate from the 40 trees we had already selected. We used a separate sample
population so branch tips would not be clipped off during the season. For this sample
method, we randomly selected ten trees in each ﬁeld by walking in diagonal transects
across the ﬁeld and marking a tree at 20 m intervals. These trees were tagged and given a

unique identiﬁcation number to be used in all three years of the study. On these ten trees,

aphids were sampled weekly. On the north and south sides of each tree at mid-canopy
level, we rapped the foliage three times with a dowel and counted aphids falling onto
black cloth held in an embroidery hoop 22.9 cm in diameter. In 1999, these ten trees
were also used to monitor the rate of shoot expansion (see tree phenology section). In
2000 and 2001, we expanded this sample method from ten trees to include all 40 sample
trees in each ﬁeld. For all data analyses, the sum of the aphid numbers found on the
north and south sides was used as the variable representing aphid numbers per tree.

In 2000, aphid density was sampled weekly on all trees (11 = 50 per ﬁeld) using
the ‘beat method’. All aphids collected in the ﬁeld were placed using a ﬁne camelhair
brush into microfuge tubes ﬁlled with 70% ETOH. During the winter, aphids were
identiﬁed in the laboratory to instar (after Varty 1968).

In 2001, we focused on monitoring aphid phenology intensively only through the
beginning of the second generation, which was most closely tied to our objective of
determining an action threshold. Using Taylor’s Power Law (Hayek and Buzas 1997),
we determined that sampling approximately 30 trees would provide adequate
information. To end up with a sample size of 30, we selected 20 trees at random from
our original sample population, and also used the ten trees that had been used for ‘beat
samples’ in 1999. Trees were sampled weekly in the same manner as in 2000. Because
collecting all of the aphids on each sample was time consuming, we collected aphids
from only the ﬁrst 20 trees in the ﬁeld, or until 2 100 aphids had been collected. In the
laboratory, aphids were examined under a microscope and identiﬁed to instar to
determine phenology. Representative samples of Mindarus were identiﬁed by D.

Voegtlin at the Illinois Natural History Survey. Voucher specimens were deposited in the

A.J. Cook Arthropod Research Collection at Michigan State University, voucher no.
2002-01.

Cumulative degree-days from the nearest weather station to each study site were
obtained weekly from the Michigan State University Agricultural Extension website.
Lansing was used for the Ingham County ﬁeld, Kalkaska and Lake City were used for the
Grand Traverse County ﬁeld, and Kalkaska was used for the Antrim County ﬁeld.
Cumulative degree-days expressed as base 50°F (10°C) were used because degrees
Fahrenheit are more accessible and familiar than degrees Celsius to growers and

extension personnel in the United States (Mussey and Potter 1997, Herms 1990, Pruess
1983). Degree-days are abbreviated here as DD50.

Tree phenology. Our goal in monitoring tree phenology was to determine
whether the timing of bud break affected susceptibility to aphid damage. Two methods
to measure tree phenology were used: date of budbreak and rate of shoot expansion.
However, the exact date of budbreak could not be monitored for all ﬁelds each year
because it would have required us to examine each tree daily. Our ﬁelds were separated
by a three-hour drive, which made that impractical. Also, since the buds on a given tree
break over a period of several days, an exact ‘date of budbreak’ is a relatively subjective
measure. To represent the date of budbreak, we used the sample date when, on average,
50% of the buds had broken on each tree.

In the Ingham County ﬁeld, ten pairs of early and late budbreaking trees were
selected in early May 1999, when it was apparent that some trees had not yet broken bud
while others had completed budbreak and new shoots were beginning to expand. The

trees in each pair were adjacent to each other, to ensure that they were subject to similar

soil temperatures, microclimate, and other factors. These trees were monitored in each
year to determine if the relative order of budbreak and shoot expansion remained similar
between years.

In Grand Traverse County, we identiﬁed ten trees in the ﬁeld that had not yet
broken bud on 26 May 1999. In 2000 and 2001, these trees were observed for
approximate date of budbreak, to determine if they were consistently later than the
surrounding trees.

In 2000 in the Ingham County ﬁeld, we monitored budbreak every few days on
each marked tree (n = 70). On 23 April, 27 April, 30 April, 3 May, 7 May, and 9 May,
we counted the percentage of buds that had broken in each of 12 sectors per tree. Sectors
were designated as the top, middle, and bottom thirds of the tree facing each cardinal
direction. Budbreak was deﬁned as the stage when the cap no longer covered the tip of
the bud, and the new needles were 210% visible (‘Stage 2’ per Osawa et a1. 1983). We
deﬁned the approximate date of budbreak as the date when on average 50% of the buds
had broken. This method was thorough, but was extremely time consuming.

In Grand Traverse County in 2000, we were able to monitor the approximate
status of budbreak for all sample trees in this ﬁeld only on 2 May, 4 May, and 10 May.
All of these trees had completely broken bud by 10 May. Therefore, we divided the trees
into three categories: trees that had begun to break bud on or before 2 May were
categorized as early, trees that had begun on 4 May were categorized as middle, and trees
that broke bud after 4 May were categorized as late budbreakers.

Rates of shoot expansion. The rate of shoot expansion was used in all three years

as a relative indicator of tree phenology and a surrogate for date of budbreak. Since

10

shoot grth and expansion cannot begin until the buds have broken, it seemed
reasonable to assume that trees which broke bud earlier would also begin shoot grth
earlier. This method of calculating the percentage of the ﬁnal shoot length attained on
each sample date was easily quantiﬁable, did not require daily visits to the ﬁeld, and was
objective so different observers could work simultaneously (per Mingo and Dimond
1979)

In 1999, a separate, randomly selected sample population of ten trees per ﬁeld
was designated for the purpose of measuring the rate of shoot expansion and its
relationship with aphid density, and also for measuring aphid density with the ‘beat
method’. These ten trees were tagged permanently and were used each year throughout
the three-year study. To measure shoot expansion in 1999, 12 shoots on each of the ten
trees were marked before or shortly after budbreak in early spring and the length from the
base of the bud collar to the tip of the shoot was measured weekly. Each shoot was
located in one of twelve sectors of the tree, which was divided vertically in thirds and
horizontally by the four cardinal directions. In 2000, we used this method to monitor the
rate of shoot expansion on all of the sample trees in each ﬁeld (n = 50). In 2001, we
selected 20 of the sample trees at random, tagged four shoots on the upper half of the
south side of each tree, and measured them once in early spring and once at the end of the
growing season to determine the percentage of shoot expansion at an early point in the
aphid’s life cycle.

We calculated the percentage of shoot expansion of each tree on each sample date
by dividing the length of each shoot by its ﬁnal length and averaging the results for each

tree. The ﬁnal shoot length was deﬁned as the length when the weekly measurements did

11

not increase beyond the margin of experimental error ( :1: 5 mm), usually in late June
(after Mingo and Dimond, 1979).

Balsam twig aphid damage. In each year, all trees that were used for aphid
sampling and shoot expansion were monitored for aphid damage. In 1999, the
percentage of shoots per tree that exhibited aphid damage was quantiﬁed after aphid
oviposition in June, when damage is most apparent, and again at the end of the growing
season in August or September. Aphid damage was deﬁned as obvious needle curling on
current-year foliage. Each tree was divided into 12 sectors for sampling: vertically by
thirds and horizontally by the four cardinal directions. In 1999, two shoots in each sector
were tagged in June, recorded as damaged or undamaged, and observed again in
September. The percentage of damaged shoots for each tree was calculated as the total
number of damaged shoots/24 sampled shoots per tree. This method did not always
account for the patchy nature of aphid damage, so we made a more thorough
measurement the next year. In 2000, 25 randomly chosen shoots in each sector were
inspected and recorded as damaged or undamaged. The percentage of damaged shoots
for each tree was calculated as the total number of damaged shoots/300 shoots per tree.
This method was time consuming, so we improved our efﬁciency in 2001 by having one
observer visually estimate the percentage of shoot damage in each sector of each tree. To
obtain a measure of damage for each tree, we averaged the results from all 12 sectors for
each tree.

Economic impact. In 1999 and 2000, we conducted surveys of retail customers
on a choose-and-cut farm in Ingham County to determine the impact of aphid damage on

retail tree value. In early December of each year, randomly chosen customers who were

12

planning to buy a ﬁr tree were surveyed to determine their opinion of a select group of
trees with differing levels of aphid damage. Insect damage, insects, or entomology were
not mentioned to the customer until after the survey was completed.

In 1999, two groups of nine live, uncut trees (18 total) were selected on the basis
of height and level of insect damage and designated Group I and Group II. All trees were
located near one end of a section of trees, close to a lane frequented by customers, and
close enough to each other that each customer could complete the survey of nine trees in
less than 15 minutes. Within each group, there were three trees in each of three height
categories. Categories were deﬁned as short (approx. 1.8 m, 6 ft), medium (2.1 m, 7 ft)
or tall (2.4 m, 8 ft). Within each height category, one tree had little or no aphid damage
(0-10% of shoots damaged), one had medium damage (21-56% of shoots damaged) and
one had heavy damage (>56% of shoots damaged). Damage was measured by randomly
selecting a branch tip in each of eight sectors per tree (top and middle thirds by four
cardinal directions) and counting the number of damaged and undamaged shoots (needles
curled). The percentage of damaged shoots in each sector was calculated and averaged
per tree. Trees were free of other obvious defects and relatively uniform in shape. F ifty
customers each were randomly assigned to either Group I or Group 11 (total of 100
customers), to help eliminate bias due to unavoidable subjective differences in the
individual trees. Customers were asked to ﬁll out a form for each tree, with questions
addressing the height, shape and color of the tree and whether or not they would purchase
such a tree.

Methods were revised slightly in 2000. Nine trees of approximately medium (2.1-

2.4 m, 7-8 ft) height and similar taper and a range of aphid damage (3% to 64% damaged

l3

shoots) were selected, cut and placed in a row in front of the ﬁeld in random order. Fifty
customers were shown an undamaged ‘model’ tree ﬁrst and asked to compare the others
to it. To quantify the customers’ response, they were asked what they would pay for the
‘model’ tree (between US $15 and US $50) and whether they would pay more or less for
each of the other trees, in US $5.00 increments. The percentage of change in their price,
relative to what they would pay for the model tree, was used to indicate the customer’s
preference for each tree (the ‘apparent’ or ‘perceived’ value of each tree).

To estimate effects of M. abietinus damage on wholesale value, a grower
experienced with the USDA wholesale grading standards for Christmas trees graded the
18 trees used in the December 1999 retail customer survey. Grades assigned were 1, 2,
or cull (USDA 1997). The ‘premium’ grade was not used in the study because it is
primarily a show grade, used for competitive tree shows, and not normally used for
wholesale sales; wholesale trees are graded as ‘1 or better’, ‘2’ and ‘cull’. (Melvin R.
Koelling, Michigan State University, pers. comm.)

Statistical Analysis. Data were tested for normality with the Shapiro-Wilk test
(PROC UNIVARIATE, SAS v8). Because aphid numbers, percentage of shoot
expansion and the percentage of damaged shoots per tree were all random variables,
Pearson’s correlation analysis was used to assess the linear relationships between aphid
numbers, tree phenology, and damage to the tree. When the data did not ﬁt a normal
distribution, Speannan’s nonparametric correlation analysis was used (Sokal and Rohlf
1995). In all years, whenever the variable ‘aphid numbers’ was used, it represented the

sum of the beat samples on the north and south sides of the tree.

14

Data from the ten pairs of early-late budbreaking trees in Ingham County were
tested for homogeneity of variance (PROC TTEST, SAS v8). We compared aphid
numbers and the percentage of damaged shoots between the ‘early’ and ‘late’ trees with a
pooled t test if the variances were homogeneous, or the Welch test if variances were not
homogeneous. If the Welch test was used, degrees of ﬁ'eedom were adjusted using
Satterthwaite’s procedure (Kuehl 1994, Satterthwaite 1946). To test the categorical
budbreak data from Grand Traverse County, we used the Kappa measure of agreement
(Siegel and Castellan 1988).

The 1999 customer survey data (acceptance or rejection of each tree) was
assessed using chi-square contingency analysis (Siegel and Castellan 1988). For the
2000 customer survey data, the strength of the association between the percentage of
change in the customer’s price and the percentage of shoot damage was assessed using
Spearman’s correlation coefﬁcient (Sokal and Rohlf 1995). All tests were conducted at a

signiﬁcance level of a = 0.05. Data were analyzed using SAS v. 8 (SAS Institute, 1999).

Results

Balsam twig aphid phenology. Egg hatch. In Ingham County in 1999, hatching

was complete by our ﬁrst ﬁeld visit on 16 April, or 95 cumulative degree-days base 50°F

(10°C) (hereafter DD50). In Grand Traverse County in 1999, hatching had already begun

by our ﬁrst ﬁeld visit on 13 April (46 DD50). In Antrim County in 1999, hatching began

15

before 20 April (54 DD50). We took this as evidence of a very early start to the aphid’s
life cycle in early spring.

In 2000, we recorded the phenology of egg hatch more precisely in Ingham
County, and the percentage of egg hatch on one early ﬁeld visit to the ﬁelds in Grand

Traverse and Antrim Counties. In the Ingham County ﬁeld, egg hatch began before our

ﬁrst ﬁeld visit on 21 March (74 DD50), and continued to 30 March (96 DD50) (Figure 1).

In Grand Traverse County on 28 March (50 DD50), eggs were 41% hatched (41
hatched/ 100 eggs examined). In Antrim County, 76% of eggs had hatched on 28 March
(50 DD50)(35 hatched/46 eggs examined).

In 2001, we did not record the duration of egg hatch. Field visits began after egg

hatch had been completed, on 29 April in Ingham County (130 DD50) and 5 May in

Grand Traverse and Antrim Counties (140-150 DD50).

F undatrices. Newly hatched aphids, or stem mothers, were extremely small and
difﬁcult to see in the ﬁeld. As they began to mature, they became more conspicuous,
especially by the third and fourth instars. I observed the new stern mothers feeding on
the previous year’s foliage or through the bud scales. They completed four instars in
approximately four weeks.

In each ﬁeld and year, stem mothers reached reproductive maturity between 100-
150 DD50, although calendar dates varied (Figures 2-4). Under the microscope,
reproductively mature stem mothers could be identiﬁed by observing fully formed
embryos of the sexuparae inside her abdomen. As the stem mothers began to reproduce,

at approximately 150-200 DD50, their number decreased and the second generation,

16

sexuparae, quickly replaced them. The sexuparae could be distinguished ﬁ'om the stem
mothers under the microscope by the greater number of eye facets (per Varty 1968).

The second generation (sexuparae). In each ﬁeld, the second generation was ﬁrst

observed between 150-200 DD50. In 1999, the second generation was ﬁrst observed in
Ingham County on 2 May or 160 DD50 (Figure 2a). In Grand Traverse County, the

second generation was ﬁrst observed on 6 May or approximately 149 DD50 (Figure 2b).
The ﬁrst sample date where the sexuparae had almost completely replaced the stem

mothers was 10 May (229 DD50) in Ingham County and 18 May (247 DD50) in Grand

Traverse County.

In 2000 in Ingham County, the second generation was ﬁrst observed on 27 April
(154 DD50) (Figure 3a). By 3 May (189 DD50), the second generation comprised
approximately 75% of the population. On the next sample date in Ingham County (11
May, 327 DD50), the second generation had completely replaced the stem mothers. In
2000 in Grand Traverse County, the second generation was not observed on 2 May (110
DD50), but by the next sample date on 10 May (228 DD50), sexuparae made up 90% of
the aphid population (Figure 3b). In Antrim County, sexuparae were ﬁrst observed on 2
May (110 DD50) (Figure 3c). By 10 May (228 DD50), sexuparae comprised 50% of the
aphid population in Antrim County.

In 2001 in Ingham County, the second generation was ﬁrst observed on 30 April
(139 DD50), and had completely replaced the stem mothers by 13 May (320 DD50)
(Figure 4a). In Grand Traverse and Antrim Counties, most aphids were mature stem

mothers on the ﬁrst sample date, 5 May (150 DD50) (Figure 4b,c). By the second sample

17

date on 14 May, Antrim and Grand Traverse Counties had reached roughly 200 DD50,
and the sexuparae had replaced the stem mothers. Grand Traverse and Antrim Counties

did not reach 300 DD50 until 30 May, more than two weeks after Ingham County had

reached 320 DD50 (on 13 May). This illustrated the lag in cumulative degree-day

accumulation between the more northern counties and Ingham County, which were
separated by approximately 2 degrees latitude, or roughly 288 km (180 miles).

Predicting aphid numbers. The numbers of stem mothers observed in our
samples in early spring were often positively correlated with aphid numbers in the
following weeks (Table 1). This correlation is important if the stem mothers are to be
used as an indicator of future aphid numbers and potential damage. In 1999, we did not
sample early enough to determine if stem mother numbers could predict later aphid
numbers. In 2000, with an earlier start to aphid sampling, we found signiﬁcant
correlations between numbers of stem mothers (before 27 April in Ingham County, and
before 10 May in Grand Traverse and Antrim Counties) and subsequent density of the
second generation in each ﬁeld. In 2001, we found signiﬁcant correlations in Ingham
County, but not in Grand Traverse County or Antrim County.

Tree phenology. The timing of budbreak for most of the trees in Ingham County
corresponded with 140-189 DD50 (22 April 2000-3 May 2000). Trees with budbreak on

or after 7 May 2000 sustained no aphid damage, while trees which broke bud before 7
May had 10% to 25% damaged shoots (Figure 5).

In general, trees which broke bud approximately 7-10 days later than most of the
surrounding trees tended to exhibit little or no aphid damage, regardless of the number of

aphids found on the trees early in the season. Comparisons of the number of aphids per

18

tree and the percentage of damaged shoots were signiﬁcant between the ten pairs of
early- and late- budbreaking trees in Ingham County (Table 2). Of the late-budbreak
trees in the ten selected pairs, all had less than 1% damage. The amount of damage
ranged from 0% to 70% throughout the ﬁeld. Among the entire sample population in
Ingham County, trees that broke bud late (9 May or later) had signiﬁcantly lower
numbers of aphids than trees that broke bud before 9 May on all sample dates except 30
March (when aphids were newly hatched stem mothers) and 1 June (when most were
sexuales).

Shoot expansion rate and damage. We also wanted to determine if the rate of
shoot expansion was related to the amount of aphid damage. In 1999, the rate of shoot
expansion of the ten ‘beat sample’ trees was not signiﬁcantly correlated with aphid
damage in either Ingham County or Grand Traverse County.

However, in 2000, with a higher sample size, the correlation between percentage

of shoot expansion in early May and aphid damage was positive and highly signiﬁcant in
both Ingham County (10 May: rs = 0.52, n = 67, p < 0.0001) and Grand Traverse County
(13 May: r8 = 0.60, n = 49, p < 0.0001). In Antrim County, this relationship was not
signiﬁcant (13 May: rs = -0.02, n = 49, p = 0.87).

In 2001, the rate of shoot expansion was signiﬁcantly correlated with aphid
damage only in Ingham County (rs =0.40, n = 40, p = 0.01). There was essentially no
relationship between aphid damage and the percentage of shoot expansion on 14 May in
Grand Traverse County (rS = 0.06, n = 23, p = 0.78) or in Antrim County (rs = 0.30, n =

30, p = 0.11). The variability in rates of shoot expansion was highest in Ingham County

(range of 0-98%) than in either Grand Traverse or Antrim Counties (range approx. 7-

l9

39%). Interestingly, the amount of shoot damage was not above 21% on any tree in
Ingham County in 2001, but in Antrim and Grand Traverse Counties shoot damage
ranged widely, from 0 to over 60%.

Shoot expansion rate and aphid numbers. In 1999, the percentage of shoot
expansion was not signiﬁcantly correlated with aphid numbers on any sample date,
although in Ingham County on 17 May (when aphid numbers ranged widely; range of O
to 228/tree; mean 68.6 :1: 23), this relationship was marginally insigniﬁcant (n = 10, p =
0.08, rS = 0.58). In Grand Traverse County, the percentage of shoot expansion was also
marginally insigniﬁcantly correlated with aphid numbers on 18 May (range 3-39; mean

23.7 i 3.5 aphids per tree) (n = 10, p = 0.07, rs = 0.60), but not correlated on the other

sample dates.

In 2000 and 2001, the relationship between rates of shoot expansion and aphid
numbers was variable. In Ingham County in 2000, the percentage of shoot expansion on
all trees (total n = 70 minus 3 missing values) was signiﬁcantly and positively correlated

with aphid numbers on 11 May (when the aphid population was 27% stem mothers and

73% sexuparae) (n = 67, p <0.0001, rS = 0.56) and 15 May (100% sexuparae) (n = 67, p =
0.0002, rs = 0.44). In Grand Traverse County and Antrim County in 2000, this

relationship was not signiﬁcant on any sample date.

In 2001 in Ingham County, the percentage of shoot expansion was positively
correlated with aphid numbers on nearly every sample date. For example, on 30 April
the relationship was highly signiﬁcant (11 = 41, p = 0.003, rS = 0.45). In Antrim County,
the percentage of shoot expansion was positively correlated with aphid numbers only on

14 May (n = 30, p = 0.022, rS = 0.42) when 92.4% of the aphids were sexuparae. No
20

signiﬁcant correlations were observed on other sample dates in Antrim County nor on
any date in Grand Traverse County.

Consistency in tree phenology. We tested the consistency of bud break timing
and shoot expansion between years to determine whether the phenology of our sample
trees was inﬂuenced more by genetic predisposition or environmental factors. If no
correlation existed, then selection of trees with later budbreak tendencies would not be
useful for breeding programs.

First, we looked at the relative order of budbreak for trees that we had observed in
two or more years. In general, trees that broke bud 7-10 days later than the surrounding
trees in 1999 continued this tendency. In Grand Traverse County, ten trees that were
particularly late in breaking bud in 1999 were also up to two weeks later than other trees
in the same ﬁeld in both 2000 and 2001.

To test the relative order of budbreak in Grand Traverse County, we categorized
32 trees that had been observed in both 2000 and 2001 as early, mid, or late budbreakers.
Contingency analysis showed a signiﬁcant measure of agreement between the categories
in 2000 and 2001 (K = 0.29, p = 0.02).

We also correlated the percentage of shoot expansion for the same trees between
years wherever possible. In Ingham County, the ten pairs of trees that had been selected
for differences in budbreak date in 1999 showed a signiﬁcant positive correlation in early
rates of shoot expansion between years 2000 and 2001 (n = 15 because of missing values;
p <0.0001, r = 0.92) (Figure 6). The ten trees that had been selected at random for ‘beat
samples’ in 1999 in Ingham County also exhibited a signiﬁcant and positive correlation

in rate of shoot expansion between 1999 and 2000 (n = 10, p <0.0001, r = 0.97) (Figure

21

7). The 23 trees selected solely on the basis of aphid damage in Ingham County had
neither particularly late or early budbreak (Figure 8). The percent shoot expansion in

early spring for these trees was not signiﬁcantly correlated between years (n = 23, p =
0.12, r5 = 0.33), perhaps due to the lack of variability. In the Grand Traverse County

ﬁeld, we observed a similar pattern. In Antrim County, 19 trees selected on the basis of
aphid damage in 1999 did show a signiﬁcant positive correlation between rates of shoot
expansion in 2000 and 2001 (n = 19, p = 0.01, rs = 0.56).

Balsam twig aphid damage. Establishing a relationship between aphid numbers
and the resulting amount of damage to the tree is essential to our objective of establishing
an aesthetic injury level or action threshold. Such a threshold would only be possible if
aphid numbers early in the season were positively correlated with a quantiﬁable amount
of aesthetic damage (Pedigo et a1. 1986). In 1999 in Ingham County, aphid numbers

sampled using the beat method on 17 May, at the peak of the second generation (mean of

68 aphids/tree), were not signiﬁcantly correlated with aphid damage (rs = 0.62, n = 10, p

= 0.06), but were strongly correlated on 24 May (rs = 0.89, n = 10, p = 0.0006) when

sexuparae had reached the fourth and last adult instar (mean = 99.6 aphids/tree).

In Grand Traverse County in 1999, aphid numbers on 18 May, at the beginning of
the second generation, were not related to aphid damage (rS = 0.02, n = 10, p = 0.95).
This relationship was stronger but still not signiﬁcant on 26 May, when the second
generation was at its height (rs = 0.42, n = 10, p = 0.22).

In 2000, when a much larger sample size was used for sampling aphids with the

beat method, aphid numbers were signiﬁcantly correlated with tree damage in almost

22

every case. In Ingham County, aphid numbers were signiﬁcantly correlated with tree
damage as early as 6 April (r5 = 0.49, n = 65, p <0.0001), when the population was still

entirely stem mothers, and this relationship remained highly signiﬁcant on every sample
date thereafter.

In 2000 in Grand Traverse County, aphid numbers were signiﬁcantly correlated
with tree damage on every sample date. On 2 May, at 110 DDso and just before the

appearance of the second generation, aphid numbers were signiﬁcantly correlated with

tree damage (rS = 0.35, n = 49, p = 0.01). This relationship remained signiﬁcant on 10

May, 19 May, and 26 May.

In 2000 in Antrim County, where aphid numbers were relatively low, the
relationship was still highly signiﬁcant on most sample dates. Even on 28 March, at the
start of aphid sampling and only 50 DD”, aphid numbers were positively correlated with
tree damage (rS = 0.34, n = 50, p = 0.02). This correlation was not signiﬁcant on 18 April
or 25 April, but on 2 May, just before the second generation appeared, it was highly
signiﬁcant (rs = 0.45, n = 50, p = 0.001).

In 2001 in Ingham County, aphid numbers were signiﬁcantly correlated with the
percentage of damaged shoots on almost every sample date: 30 April, 7 May, 12 May,
and 21 May. In Grand Traverse County in 2001, the relationship between aphid numbers
and damaged shoots was not signiﬁcant on any sample date. Aphid numbers in Grand
Traverse County in 2001 tended to be low (highest value of 28.5 :1: 4 per tree on 6 June),
but damage ranged widely, from 1% to 62% damaged shoots. In Antrim County, aphid
numbers were positively correlated with damage on 14 May only. On 14 May in Antrim

County, aphid numbers averaged 31.7 :1: 4.5 per tree, and the sexuparae had almost

23

completely replaced the stern mothers. Damage in Antrim County also ranged widely,
from 0 to 76% damaged shoots per tree, although all except three of the trees had from 0
to 42% damage.

Economic impact. The two replicates in the 1999 customer survey presented
varying results. We analyzed the two groups separately, since the patterns of customer
choice were different between the two. For example, the Group I customers preferred the
tall trees the most (56% of the positive responses from customers were for tall trees), but
the Group II customers preferred the medium height trees (54% of positive responses
were for medium height trees) (Figure 9).

The 1999 Group 1 surveys showed that both height and M. abietinus feeding
damage were signiﬁcant factors in a customer’s decision to accept or reject a particular
tree (height: X2 = 58.9, df= 2, p <0.0001; damage: x2: 10.76, df= 2, p = 0.0046).
However, when the trees were analyzed separately according to height class, damage was
not a signiﬁcant factor in their decision to accept the medium or tall trees (Table 3).
Damage was a signiﬁcant factor for the short trees, but the data were heavily skewed in
that only one customer chose the short tree with no damage (Figure 9). In both groups,
the short trees were the least preferred: 16% and 20% of customers in Group I and Group
II, respectively, accepted short trees, regardless of damage.

The Group 11 surveys also showed that both height and damage were signiﬁcant
factors at p <0.0001, but when heights were considered separately, damage was still a
signiﬁcant factor at each level (Table 3). Trees in this group with medium amounts of
aphid damage (21-5 6%) were apparently preferred over trees with little or even heavy

damage (Figure 9).

24

In the 2000 customer surveys (Figure 10), the percentage of damaged shoots was
a signiﬁcant factor in customer preference if the trees were divided into damage
categories (light = 3-6% of shoots affected; medium = 9-14%, heavy = 23-64%).
Analysis of variance indicated that damage was a signiﬁcant factor affecting customers’
perceived value (F = 23.5; df = 2, 397; p < 0.0001). The light and medium damage trees
did not signiﬁcantly differ from each other in apparent value (Fisher’s LSD, p <0.05), but
both were signiﬁcantly different from the heavily damaged trees.

These data indicated that the effect of aphid damage on apparent retail value is not
a linear relationship. When the percentage of change in value was regressed on the actual
percentage of damaged shoots, the regression explained only 6% of the variability. Using

the Spearrnan rank correlation analysis, the relationship between apparent value and
percentage of shoots with damage was signiﬁcant but not strong (rs = -0.26, p < 0.0001).

In terms of wholesale grades, trees with up to 40% damaged shoots were still
considered a Grade 1, while trees in Grade 2 ranged from 32 to 62% damage (Figure 11).
Most likely due to this high variability, no signiﬁcant differences were found in aphid
damage among the wholesale grades (F = 1.38; df = 2,15; p = 0.28). This would indicate
that the wholesale value of the tree did not necessarily drop because of aphid damage.
The cull trees (unsaleable) consisted of one tree with 42% damaged shoots, and one with
nearly 100% damage. The tree with 42% damage was noted as being ‘not full enough’; it
was likely more sparse in appearance than other trees with the same amount of aphid

damage.

25

Discussion

Our study showed that M. abietinus phenology is more closely tied to degree-day
accumulation than to calendar date, implying that for this aphid, accumulated degree—
days, which reﬂect temperature, are an important cue for development. The temporal
separation of aphid development by about one week between Ingham County and the
more northern counties conﬁrmed this observation. Varty (1966, 1968) also showed that
temperature is an important controlling factor in M. abietinus phenology, and suggested
that degree-days might be useful for predicting development rates in the ﬁeld.

Scouting during the ﬁrst generation of M. abietinus (the stem mother generation),
can indicate the potential size of future M. abietinus populations in upcoming weeks.
Even using our relative sampling method of rapping foliage over an embroidery hoop,
stem mother numbers were oﬁen signiﬁcantly correlated with later aphid sample numbers
(Table 1). Although the fecundity of the stem mothers ranges widely, ﬁom 3 to 60
offspring (Varty 1966), the average fecundity is approximately 22 to 41 offspring per
female (Varty 1968). Hence, applying control measures before the stem mothers mature
and reproduce would likely have a strong effect on subsequent generations.

The relationship between aphid numbers and tree phenology was signiﬁcant for
the ten pairs of early— and late-budbreaking trees in Ingham County. In the other ﬁelds
when tree phenology was measured by percentage of shoot expansion, results were
variable. It is relatively unlikely that M. abietinus is capable of deliberately selecting
trees with earlier ﬂushing tendencies, since the winged adult M. abietinus is a weak ﬂier

and is likely to be dispersed by air currents within a ﬁeld (Amman 1963). This indicates

26

that aphid eggs may be equally likely to be present on trees with earlier or later ﬂushing
times, and the differences in aphid numbers we observed may be the result of differential
survival rates between earlier and later ﬂushing trees. As an example, the ten pairs of
early and late budbreaking trees in Ingham County consistently differed in aphid numbers
and damage in all three years, although the initial number of aphids was not always
signiﬁcantly different. In addition, the strong positive correlation between the rates of
shoot expansion and aphid damage on all sample trees in Ingham and Grand Traverse
Counties in 2000 indicated that trees which broke bud and began to expand new shoots
later in the season escaped at least some aphid damage.

The extent of genetic control of budbreak has been studied for several tree species
(e. g. Li and Adams 1993, Murray et a1. 1994). Although the timing of budbreak is
inﬂuenced by local factors, especially temperature (e. g. Partanen et a1. 1998, Worrall and
Little 1986, Lowe et al.1977), the tendency of balsam ﬁr to break bud at a given time is
determined in part by genetic factors (Lester et al. 1976, Lowe et a1. 1977).

Our data indicated that budbreak and early rates of shoot expansion in balsam and
Fraser ﬁr are at least partly controlled by genetic factors, since trees measured in two or
three consecutive years were almost always consistent in their relative order of budbreak
and early shoot expansion. Although we did not do provenance testing, other workers
have found strong correlation between the origin of the seed source and the date of bud
break for several conifers (Li and Adams 1993, Murray et a1. 1994, Beuker 1994, Lester
et al. 1976). Future research could further explore this, perhaps by selecting and breeding

trees with later ﬂushing tendencies. If host plant resistance can be used in selection or

27

breeding programs, the need for insecticide applications in ﬁr ﬁelds could be
substantially reduced.

The economic impact of M. abietinus appears to be relatively low. Most of the
retail choose-and-cut customers we questioned after they had completed the surveys
stated they would not have noticed light to moderate aphid damage if we had not pointed
it out. This agrees with the ﬁndings of Kleintj es et a1. (1999), who found that choose-
and-cut retail customers still selected trees with up to 50% of the shoots damaged. In our
1999 customer surveys, M. abietinus damage was not a consistent factor in customer’s
choice of tree, but height was important. Kleintjes et a1. (1999) also found that tree
height and shape were signiﬁcant factors affecting customer choice, but that aphid
damage levels did not signiﬁcantly differ between selected and unselected trees. Our
Group 11 surveys in 1999 appeared to show that trees with medium damage were
preferred, which further suggests that a medium level of aphid damage (e. g. 21-56%
damaged shoots) does not have a detrimental effect on perceived tree value. In our 2000
surveys, damage was not a signiﬁcant factor in consumer perception of tree value, which
conﬁrms Kleintj es’ ﬁndings that retail choose-and-cut customers do not consider M.
abietinus damage to be an important factor. In our study, the change in perceived value
varied widely in the light and medium damage levels, indicating that other factors such as
height, shape, fullness, or taper may be equally or more important than the amount of
aphid damage, especially at light or medium levels of damage.

The effect of M. abietinus damage on wholesale grade is also less than might be
expected. The level of damage that caused a tree to drop from a Grade 1 to a Grade 2

rating (representing economic loss) is relatively high, and strongly dependent on its

28

within-tree distribution. The USDA standards for grading wholesale Christmas trees
include 12 characteristics, only one of which is insect damage (USDA 1997). For a tree
to receive a Grade 1 rating, one face may have one noticeable defect (such as moderate to
heavy aphid damage), and each face is allowed one minor defect (such as light aphid
damage). Wholesale trees are unlikely to be reduced from a ‘1 or better’ grade to a ‘2’
grade because of aphid damage unless the damage is moderately heavy on two or more
sides of the tree (USDA 1997). The term ‘moderately heavy’ is subject to interpretation
and is not deﬁned by the USDA, but ‘abnormal curling of needles’ is mentioned, with
‘slightly abnormal’, ‘moderately abnormal’, and ‘severely abnormal’ as categories. Only
trees with ‘severely abnormal curling of needles’ are classiﬁed as culls. Although a few
trees in a given ﬁeld may exhibit such severe damage, the majority of trees in the ﬁelds
we studied did not reach that level. In 2001, 67-100% of the trees used in our study had
damage levels of 25% or less. In 2000, 74-94% of trees used for our study had less than
25% damage. Some other studies on balsam twig aphid have found similar results. In
the study by Nettleton and Hain (1982), trees averaged only 23% damage, with a range of
11-59% damaged shoots. In the study by Kleintjes et a1. (1999), 30 trees selected by
retail customers on a choose-and-cut farm sustained 20-50% damaged shoots or less.

The number of stem mothers that result in trees with an economic level of damage
varies depending on the acceptable economic injury level. If 25% of damaged shoots per
tree is chosen as a conservative economic injury level, the likelihood of a tree sustaining
that much damage can be calculated based on the number of stem mothers sampled. In
the Ingham County ﬁeld in 2000, ﬁve or more stem mothers per tree (11 = 32 trees),

resulted in 12 trees with greater than 25% damage, or a 38 percent chance that tree value

29

would be economically affected. Trees with two or more stem mothers (n = 45) had a 33
percent chance of sustaining 25% or more damage. These probabilities indicate that the
risk of economic damage is relatively low. Although Kleintjes et al. (1999) found that
approximately two or more stem mothers on a given tree would result in about 50% of
damaged shoots, we found that this did not hold true 100% of the time. The action
threshold for balsam twig aphid will vary depending on the economic injury level and the
level of risk the grower is willing to accept.

An interesting possibility for management of M. abietinus could include
developing a method to forecast outbreak years, and planning ahead. In Canada, M.
abietinus has been reported to occur in outbreak cycles of 4-6 years (Varty 1968, Mattson
et al. 1989, Rather and Mills 1989). These outbreaks are reported to occur
simultaneously over large areas (Mattson et a1. 1989, Rather and Mills 1989). Further
research may determine whether M. abietinus follows this pattern in Christmas tree
plantations.

The cost of applying control measures such as insecticide sprays for M. abietinus
may not always be economically justiﬁed. Taken together, our results indicate that
careﬁrl scouting, selecting trees with later budbreak tendencies, and avoiding control
measures that are not economically justiﬁed will help reduce pest control costs,
increasing marginal proﬁts for growers and the overall competitiveness of the Michigan

Christmas tree industry.

30

References Cited

Amman, GD. 1963. A new distribution record for the balsam twig aphid. J. Econ.
Entomol. 56: 113.

Berthiaume, R., Ch. Hébert and C. Cloutier. 2000. Predation on Mindarus abietinus
infesting balsam ﬁr grown as Christmas trees: the impact of coccinellid larval predation
with emphasis on Anatis mali. BioControl 45: 425-438.

Berthiaume, R., Ch. Hébert and C. Cloutier. 2001. Podabrus rugosulus (Coleoptera:
Cantharidae), an opportunist predator of Mindarus abietinus (Hemiptera [sic]: Aphididae)
in Christmas tree plantations. Can. Ent. 133: 151-154.

Beuker, E. 1994. Adaptation to climatic changes of the timing of bud burst in
populations ofPinus sylvestris L. and Picea abies (L.) Karst. Tree Physiol. 14: 961-970.

Bradbury, R.L., and EA. Osgood. 1986. Chemical control of balsam twig aphid,
Mindarus abietinus Koch (Homoptera: Aphididae). Maine Agric. Exp. Stn. Tech. Bull.
124.

Carter, CL, and J.F.A. Nichols. 1985. Some resistance features of trees that inﬂuence
the establishment and development of aphid colonies. Zeitschrift fuer Angewandte
Entomologie 99: 64-67.

DeHayes, D.H. 1981. Genetic variation in susceptibility OfAbies balsamea to Mindarus
abietinus. Can. J. For. Res. 11:30-35.

Desrosiers, N. 1998. Inﬂuence de la fertilisation azotée et de la date de débourrement
sur les populations du puceron des pousses Mindarus abietinus Koch (Homoptere:
Aphididae). Mémoire de maitrise, Université Laval, Quebec, Canada.

DiFonzo, CD. and D.G. McCullough. 1998. Potential impacts of the Food Quality
Protection Act on the Michigan Christmas tree industry. Michigan Christmas Tree
Journal 45(1): 26-29.

Hayek, L-A. C. and M.A. Buzas. 1997. Surveying natural populations. Columbia
University Press, New York, New York. 563 pp.

Herms, D.A. 1990. Biological clocks. Am. Nurseryman 172(8):56-63.
Jones, D.M., L.A. Leefers and M.R. Koelling. 1999. Costs and Returns in Michigan

Christmas Tree Production, 1997. Michigan Agricultural Experiment Station Research
Report 565. Michigan State University, East Lansing, Michigan.

31

Kleintjes, P.K. 1997a. Midseason insecticide treatment of balsam twig aphids
(Homoptera: Aphididae) and their aphidophagous predators in a Wisconsin Christmas
tree plantation. Environ. Entomol. 26: 1393-1397.

Kleintjes, P.K. 1997b. Effects of integrating cultural tactics into the management of the
balsam twig aphid Mindarus abietinus Koch (Aphididae: Homoptera) in balsam ﬁr
Christmas tree plantations. pp. 112-121 In: J .C. Gregoire, A.M. Liebhold, F .M. Stephen,
K.R. Day, and SM. Salom, eds. Proceedings: Integrating cultural tactics into the
management of bark beetle and reforestation pests. USDA Forest Service General
Technical Report NE-23 6.

Kleintjes, P.K., E.E. Lemoine, J .Schroeder, and M.J. Solensky. 1999. Comparison of
methods for monitoring Mindarus abietinus (Homoptera: Aphididae) and their potential
damage in Christmas tree plantations. J .Econ. Entomol. 92(3): 63 8-643.

Kuehl, R0. 1994. Statistical principles of research design and analysis. Duxbury Press,
Belmont, California. 686 pp.

Lester, D.T., C.A. Mohn, and J.W. Wright. 1976. Geographic variation in balsam ﬁr:
ll-year results in the Lake States. Can J. For. Res. 6: 389-394.

Li, P. and W.T. Adams. 1993. Genetic control of bud phenology in pole-size trees and
seedlings of coastal Douglas-ﬁr. Can. J. For. Res. 23: 1043-1051.

Lowe, W.J., H.W. Hacker, Jr., and M.L. McCormack, Jr. 1977. Variation in balsam
ﬁr provenances planted in New England. Can. J. For. Res. 7: 63-67.

Mattson, W.J., R.A. Haack, R.K. Lawrence and D.A. Herms. 1989. Do balsam twig
aphids (Homoptera: Aphididae) lower tree susceptibility to spruce budworrn? Can.
Entomol. 121: 93-103.

McCullough, D.G. and K. Fondren 1998. FQPA and insecticide use in Michigan
Christmas tree ﬁelds: preliminary data from 1998 survey. Michigan Christmas Tree
Journal 45(3): 22-29.

McCullough, D.G., S.A. Katovich, M.E. Ostry, and J. Cummings-Carlson [eds.].
1998. Christmas tree pest manual, 2“d ed. Michigan State University Extension Bulletin
E-2676. East Lansing, MI. 143 pp.

Michigan Agricultural Statistics Service 1998.
URL: http://wwwmda.state.mi.us/mass/index.html

Michigan Agricultural Statistics Service 2000.
URL: http://www.nass.usda.gov/mi/

32

Michigan State University Agricultural Extension website: URL:
<http://www.agweather.geo.msu.edu/>

Mingo, T.M., and J.B. Dimond. 1979. Balsam ﬁr (Abies balsamea (L.) Mill.)
phenology in Maine. Bull. Maine Life Sci. Agric. Exp. Stn. 1979 (759) 13 pp.

Murray, M.B., R.I. Smith, I.D. Leith, D.Fowler, H.S.J. Lee, A.D. Friend and P.G.
Jarvis. 1994. Effects of elevated C02, nutrition and climatic warming on bud
phenology in Sitka spruce (Picea sitchensis) and their impact on the risk of frost damage.
Tree Physiol. 14: 691-706.

Mussey, OJ. and D.A. Potter. 1997. Phenological correlations between ﬂowering
plants and activity of urban landscape pests in Kentucky. J. Econ. Entomol. 90(6):1615-
1627.

Nettleton, W.A., and RP. Hain. 1982. The life history, foliage damage and control of
the balsam twig aphid, Mindarus abietinus Koch (Homoptera: Aphididae) in Fraser ﬁr
Christmas tree plantations of western North Carolina. Can. Entomol. 114: 155-162.

Osawa, A., CA. Shoemaker, and J .R. Stedinger. 1983. A stochastic model of balsam
ﬁr bud phenology using maximum likelihood parameter estimation. Forest Sci. 29(3):
478-490.

Partanen, J., V. Koski, and H. Hanninen. 1998. Effects of photoperiod and
temperature on the timing of bud burst in Norway spruce (Picea abies). Tree Physiol. 18:
8 1 1-8 16.

Pedigo, L.P., S.H. Hutchins, and LG. Higley. 1986. Economic injury levels in theory
and practice. Ann. Rev. Entomol. 31: 341-368.

Pruess, KR 1983. Day-degree methods for pest management. Environ. Entomol.
12:613-619.

Rather, M. and NJ. Mills. 1989. Possibilities for the biological control of the
Christmas tree pests, the balsam gall midge, Paradiplosis tumifex Gagne (Diptera:
Cecidomyiidae) and the balsam twig aphid, Mindarus abietinus Koch (Homoptera:
Mindaridae), using exotic enemies from Europe. Biocontrol News and Information
10(2): 119-129.

Raupp, M.J., J.A. Davidson, C.S. Koehler, C.S. Sadof, and K. Reichelderfer. 1988.
Decision-making considerations for aesthetic damage caused by pests. Bull. Entomol.
Soc. Am. Spring 1988. 27-32.

SAS Institute. 1999. SAS/STAT user’s manual, version 8. Cary, NC.

33

Satterthwaite, F .E. 1946. An approximate distribution of estimates of variance
components. Biometrics 2: 110-114.

Saunders, J.L. 1969. Occurrence and control of the balsam twig aphid on Abies grandis
and A. concolor. J. Econ. Entomol. 62: 1106-1109.

Siegel, S. and NJ. Castellan, Jr. 1988. Nonparametric statistics for the behavioral
sciences. 2nd ed. McGraw Hill, Boston, MA. 399 pp.

Sokal, RR. and F.J. Rohlf. 1995. Biometry: The principles and practice of statistics in
biological research. New York, Freeman. 3rd ed.

United States Department of Agriculture (USDA). 1997. United States Standards for
Grades of Christmas Trees. Agricultural Marketing Service, Fruit and Vegetable
Division, Fresh Products Branch. 10 pp.

Varty, LW. 1966. The seasonal history and population trends of the balsam twig aphid,
Mindarus abietinus Koch, in New Brunswick. Internal report M-12, Forest Research
Laboratory, Canada Department of Forestry, Fredericton, NB.

Varty, LW. 1968. The seasonal history and population trends of the balsam twig aphid,
Mindarus abietinus Koch, in New Brunswick: polymorphism, rates of development and
seasonal distribution of populations. Internal report M-24, Forest Research Laboratory,

Canada Department of Forestry, Fredericton, NB.

Worrall, J. and C.H.A. Little. 1986. An effect of gravity on bud-burst in balsam ﬁr.
Tree Physiol. 1: 47-52.

34

 

 

 

93 88.0 NE. .a N is a 2 om a: N sea: ”N
2.0 Rd NNd N 3 3.0 a ad on Ea... mN €32 N
NE. 58¢ 33 a. can 35 a. ad on E3. 2 532 N SON SEE
aNd 3o N.m a SN 2 a E a. a: N a: N
93 58¢ S a w: 2 a S a. a: a a: N
one :3 N a we 2 .u. E a. a: 2 a: N SON 8.55:. 266
mg Soodv on a Na 3 .a 4.: 8 a: 2 E3 NN
Ned Soodv N .1. n: E .1. a. : 8 a: : Ea... NN
33 Soodv 8.0 a 2. Ed H 3 we E9... NN E3. e
Ed Soodv 2 N NE Ed a 3 we E3 2 E? e
1 Sid N83 is H 3 “N a nNn 3 E3. e pea: eN SON seams
gammy“ & wwwammwom ﬂogged.“ mwob 838mm 338mm 559: 80> 205
n we 2 c: :82 0 Z 08333 3nd 88m 889

 

.8325 “5:333 E 33938 .«0 bacon 515 805325.“ ESEEQ magenta: mo bungee mo mucus—2.80 .— 039—.

35

 

 

 

 

NN.o mNd Na a a? eNd N am on 32 N a: m

m; as 3. H N: eNd N 3 cm :2 E a: m SN 852
3o Noodv 86 a 3 Ned a 3 Na a: _N a: N
N2 585v E a 3 Ned a 3. a. .32 Na a: N
33 So So a S as a 3. a. a: N E? ON
m2 N83 3 a 3 as a an a. a: N_ E9... on

93 :55 Nod a 3. Ned a man 9 a: N E9... om SON can?
$3 886 Ned N N Rd N 3 cm a: N E? N
Ne .26 N._ .1. 3 as a an ON a: S E? M:
$3 886 Ned H eN as a «N 8 32 N E? M:
$3 886 Nd a 3 3o .1. 4N ow E? N E9... 2
£5 255 N._ N 3 Ed N we cm .32 2 Es: NN

eeemwathwham Q wméasxom $0585: mwob @2953. 838mm 852: 80> 22m
0: 502 Ca 532 02 omudQn—xOm 3mm 83m moqu

"wk

 

6.2.8 m 039—.

36

Table 2. Comparisons between early and late budbreaking trees in Ingham County.

Aphid numbers found on each group were compared using pooled t-tests or Welch

tests. The amount of aphid damage (percentage of shoots damaged) sustained by

each group of trees was compared using a Welch test.

 

Aphid numbers (mean :t 1 SE)

 

 

 

Sample date Late trees Early trees t (if p =

30 March 6.0 i 1.2 10.5 i 2.3 -l.77 14 0.099
6 April 1.2 i 0.4 9.1 i 2.5 -3.06 8.4 0.015
22 April 1.9 i 1.1 16.1 i 5.5 -2.53 8.64 0.0331
27 April 0.1 i 0.1 5.5 i 2.0 -2.69 9.04 0.0248
3 May 0.75 i 0.49 5.7 i 1.8 -2.63 9.17 0.0269
11 May 0.1 i 0.1 22.8 i 8.6 -2.62 9.0 0.0276
17 May 5.6 i 0.99 63.1 i 22 -2.61 9.04 0.0284
24 May 3.2 :I: 0.81 31.9 i 8.5 -3.37 9.17 0.008
1 June 8.2 i 2.5 14.2 i 2.9 -1.54 17 0.1432
Damage 0.1 i 0.1 15.0 i- 4.0 -3.71 9.01 0.0048

 

37

Table 3. Results of chi-square tests of customer preference, 1999 surveys

 

 

Group Tree height Chi-square (if p

I Short 26.4 2 <0.0001
Medium 3.77 2 0.15
Tall 4.14 2 0.13

11 Short 19.3 2 <0.0001
Medium 1 1.2 2 0.0038
Tall 7.75 2 0.021

 

38

 

 

100

 

 

 

 

 

 

 

96 E

8 75 a U'—‘ ”8.0 : ’_ ”—7- __ -
f, 74 0 percent hatched
2 ° 130050
,_. 50 _. —-———~- --—--~-—----~—~
: O
o
8
33 25 +0 .ﬁ, N _, m—

o I I I I I I f I I

 

 

 

21 22 23 24 25 26 27 28 29 30
Sample date, March 2000

 

Figure 1. Percentage of M. abietinus eggs hatched in Ingham
County in March 2000 (number hatched/100 eggs counted).
Cumulative degree-days base 50 degrees F (10 degrees C) are
indicated for 21 March, 23 March, and 30 March.

39

 

 

 

 

 

 

 

 

  
   

 

 

F'g' 2“ Ingham County 1999 +stem mothers
100 - A ....... A - - t - -sexuparae
E . .
(D
8 80 a
O.
a:
g 0
a 6 ‘
18.3
“6 40 -
C
.9
1:
8 20 -
9
a I
0 a’
DD 22-Apr 02-May 10-May 16-May 23-May
50 97 141 229 300 380
Fig. 2b Grand Traverse County 1999 +Stem mothers
- - t - - sexuparae
u 100 -' A ........ ‘
c I
d)
8
a 80 -
Q)
E
15 60 d
1.3
‘8 40 -
C
.9
S 20 -
Q
9
0’ 0 rum-1‘- -----"
27-Apr 06-May 12-May 18—May
0050 74 152 179 247

 

 

 

 

Figure 2. Phenology of the ﬁrst and second morphs of M.
abietinus in a) Ingham and b) Grand Traverse Counties,

1999.

40

 

 

 

3a) In ham
00 - 9.

 

—-0— Stem mothers

 

 

 

 

   
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

t e c - - I - -Sexuparae
8) ,I. , -.
53 75 ‘ I' " i
(D .
m
=2 50 —
.E
5 25 - .
2 i
o 1 '
000,9, ' ' *
80 96 107 107 140 154 189 327 441
26-Mar 30—Mar 06-Apr 13-Apr 22-Apr 27-Apr 03-May 11-May 17-May
3b) Grand Traverse —O—Stem Mothers
- - I - - Sexuparae
100 — o——o—¢ e ...... .
a 1' “
,9 75 — '
0)
=33 50 4 .
C X
:3 I
5 25 ~
8 .
8 O I * I I Q
50 57 64 80 1 10 228 255 302
0050
28-Mar 04-Apr 18-Apr 25-Apr 02-May 10-May 19-May 26-May
3c) Antrim —o——Stem Mothers
- - I - - Sexuparae
a, 100 a o e c
a:
E
a) 75 r
92
‘E
o 25 a
n. 0 I I I I ------ '
DD50 50 57 64 80 110 228
28—Mar 04-Apr 18-Apr 25—Apr 02-May 10-May

 

 

 

 

 

 

Figure 3. Phenology of the ﬁrst and second morphs of M. abietinus in
(3a) Ingham, (3b )Grand Traverse, and (30) Antrim Counties, 2000.

41

 

 

 

 

 

 

 

 

 

 

4a) Ingham County —O—Stem mothers

100 — --I - -Sexuparae
81
{3; 75 —
:2
E 50 ~
*6. 2

25 ~ - I

2..g . . .l'
8 O I: "'1 . . r . e .

00,,

130 139 320 420 500
29-Apr 30-Apr 1 3—May 21 -May 04-Jun

 

4b) Grand Traverse County

 

 

 

 

—-o— stem mothers
- - I - -sexuparae

 

 

 

    

 

 

 

    

 

 

 

 

 

 

 

100 — -I.
8
8 75 .
0
g 50 -
C
§ 25 ~
0
a. .'
o I
140 207 234 318 350 475
5—May 14-May 18-May 30-May 6—Jun 14-Jun
4c) Antrim County —o—stem mothers
- - I - - sexuparae
a 100 1 I. ...... I ------ I . i‘
9, 75 J ' ‘
5
E 50 "
5 25 < . ‘i.
0 r k t t t .
150 210 230 290 320 350 475
5-May 14-May 18-May 23-May 30-May 6-Jun 14-Jun

 

Figure 4. Phenology of the ﬁrst and second morphs of M. abietinus in
(4a) Ingham, (4b) Grand Traverse, and (4c) Antrim Counties, 2001.

42

 

 

 

Percent damaged shoots vs. budbreak date, Ingham

 

 

Co., 2000.

30 -
8
g 25 "
v 8 15 - I I
E 5 10 -
3
ca 5 -
a

0 u I i I ._|

27-Apr 30-Apr 03-May 07-May 09-May
Budbreak date

 

 

 

Figure 5. Percentage of damaged shoots vs. approximate date of
budbreak, Ingham Co., 2000 (means +/- SE).

43

 

 

 

 

 

 

 

 

‘_ 80

O

a

g 60 ° 9

.3 ° .0 o

5 40 ’

a

§ 0 0

§ 20 .

.C

m 0 ‘C I I l
0 20 4O 60 8O

Shoot expansion 2000

 

 

 

Figure 6. Percentage of shoot expansion in early spring for paired
early and late budbreaking trees in Ingham County (n = 12 trees).
Pearson's product-moment correlation analysis: p <0.0001, r = 0.92.

44

 

 

g 100
8
c 80
.2
g 60
a
x 40
o
‘6 20
o
.c
‘0 0

 

 

 

 

 

 

 

 

 

T I I

40 60 80 100
Shoot expansion 1999

 

Figure 7. Percentage of shoot expansion in 1999 and 2000 for ten
randomly selected trees in Ingham County. Pearson's product-
moment correlation analysis: p < 0.0001, r = 0.97.

45

 

 

 

 

G)
O

 

 

 

 

 

 

 

5 .

O

‘2 60 4—0’.

O 9 0

g . o ‘0. . o

5 40 p °

:1 o

x

Q 0

g 20

.2

a)

0 I I I
0 20 40 60 80

Sheet expansion 2000

 

Figure 8. Percentage of shoot expansion in early spring for the aphid
sample trees in Ingham County (n = 30 trees measured in both years).
Seven trees were excluded because of missing values. Spearman's rank

correlation analysis: p = 0.12, rs = 0.33.

46

 

 

El none

I medium

.b
O

 

 

El heavy

w
0"
I

Group I

 

 

N 00
01 O
I l
‘7.“{2
JCT ‘\._ x

1b

‘1‘?

\_

15+
101
5-4

'\

Number of 'tree acceptable'
responses
N
c

we

.3
tn

 

 

 

 

 

 

 

 

 

35
30- Groupll
25 -
20 -
15 -
10 -
5 -1

 

Number of 'tree acceptable'
responses

 

 

 

 

 

 

 

 

 

   

Medium

Tree height

Figure 9. Results of 1999 retail customer surveys. The y-axis
indicates the number of customers who stated that they would
consider buying each tree. Within each height group, damage was a
signiﬁcant factor (see Table 3) unless marked with n.s.

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100
IPercent damaged shoots
75 I I T [SPercent change in perceived ' __ I"
l

0
3
g 25 - - e ~-
2
O
a 0 _ w Elg—L

-25 -N. ____. .__-

-50

LIGHT MEDIUM HEAVY

 

Figure 10. Results of 2000 customer surveys (n = 50 customers).
Trees were grouped into three categories based on the percentage
of damaged shoots/ 300 shoots examined per tree. The percentage
of damaged shoots is presented as the actual value. Damage
categories are indicated by 'Light' (n=3 trees), 'Medium' (n=2 trees),
and 'Heavy' (n=3 trees). The percentage of change in perceived
value is presented as mean +/ 1 SE.

48

 

 

Wholesale grades

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 -
19
o
g 75 - a
U) 1
8 1.
8) a
g 50 H ‘.11uum\ 1.1nulmu
3 1a 1
5 in" I'm" .
e 25 ’l 1' ll ‘
COL) .z: """""""""" A 1
11|n=11 |n=5 n=2
0 l “‘ I l
1 2 Cull
Grade

 

 

 

Figure 11. Wholesale grades assigned to trees varying in aphid
damage used in 1999 survey (n=18). Percentage of damaged shoots
is presented as mean +/- SE. Percentage of shoots with aphid
damage did not differ signiﬁcantly among grades (F = 1.38; df = 2,15; p
= 0.28).

49

Chapter 2

Potential for augmentative biological control of the balsam twig aphid (Mindarus

abietinus Koch) (Homoptera: Aphididae) in Michigan Christmas tree plantations.

Abstract. We investigated the effectiveness of natural enemies and augmentative
biological control with Chrysoperla ruﬁlabris Burmeister larvae in ﬁr Christmas tree
plantations to control Mindarus abietinus Koch. Our objectives in this study were to: 1)
describe the natural enemy complex of M. abietinus present in unsprayed ﬁr Christmas
tree ﬁelds; 2) assess the effectiveness of C. ruﬁlabris larvae as predators of M. abietinus
in the laboratory and in ﬁeld cages; and 3) evaluate the potential effectiveness of C.
ruﬁlabris in an open ﬁeld release, as might be done on a commercial Christmas tree
plantation. Results showed that M. abietinus can support a diverse complex of predators
in the ﬁeld. In the laboratory, C. ruﬁlabris proved to be a voracious predator of M.
abietinus. In ﬁeld cages, the presence of a C. ruﬁlabris larva consistently reduced the
mean number of M. abietinus in each cage, as evidenced by a reduced mean number of
M. abietinus eggs present, although the differences were not signiﬁcant. In open ﬁeld
releases, C. ruﬁlabris reduced the M. abietinus population signiﬁcantly in two out of

three ﬁelds.

50

Introduction

The balsam twig aphid (Mindarus abietinus Koch) is a serious pest affecting
balsam (Abies balsamea (L.) Mill.) and Fraser ﬁr (A. fraseri (Pursh) Poir.) Christmas
trees (Kleintjes 1997a, Bradbury and Osgood 1986, Nettleton and Hain 1982). Mindarus
abietinus is native to boreal forests, but has not been a target of large-scale biological
control efforts like the more destructive balsam woolly adelgid (Adelges piceae (Ratz.))
(McGugan and Coppel 1962). Balsam twig aphid does not usually cause tree mortality,
although it may cause chronic changes in tree properties and affect tree growth (Mattson
et a1. 1989, Berthiaume et al. 2000). High populations of M. abietinus cause needles to
become curled and distorted, reducing needle biomass and consequently tree growth
(Berthiaume et a1. 2000, Carter and Nichols 1985). When damage is heavy, the current
year’s needles on the new shoots are tightly curled together and sticky with honeydew,
forming a ‘pseudogall’ around the aphids that can protect them from contact insecticides
as well as some predators (N ettleton and Hain 1982, Berthiaume et a1. 2001).

The life cycle of balsam twig aphid typically begins before budbreak in early
spring, when stem mothers (fundatrices) hatch from the overwintering eggs. At maturity
they produce the second generation (sexuparae), which form colonies that feed on sap in
newly expanding needles. The second generation typically matures into alate (winged)
females, which disperse and produce the last generation of males and oviparous females
(oviparae). Occasionally a portion of the second generation does not develop wings and

produces more sexuparae, adding another generation of parthenogenic females that will

51

become alates. In either case, the last generation (males and oviparae) mates, the females
lay eggs (usually one egg per female) and the cycle is completed by mid to late June.
Currently, broad-spectrum insecticides are the most commonly used control
method for M. abietinus (McCullough and Fondren 1998, Bradbury and Osgood 1986).
However, insecticide sprays to control M. abietinus in Christmas tree ﬁelds are often
applied only after damaged shoots are detected (Kleintj es 1997a). This may kill some
aphids and their natural enemies, but does not prevent or reduce damage to current year
foliage (Kleintjes et a1. 1999). Development of alternative control methods for insect
pests in Christmas tree plantations has been relatively poorly studied, although interest in
alternative methods using integrated pest management, selective insecticides, and
biological control is increasing (F ondren and McCullough 2001, McCullough 1999,
Kleintjes 1997b). The availability of alternative control options may become essential
for the Christmas tree industry as pesticides are reviewed under the Food Quality
Protection Act. Minor-use commodities such as Christmas trees are at particular risk of
losing registrations for commonly used pesticides (DiFonzo and McCullough 1998).
Christmas tree plantations are capable of supporting a complex of natural enemies
due to their year-to-year stability and relative lack of disturbance compared to annual
agricultural crops (McCullough 1999, Raupp et a1. 1992). This makes conservation
biological control a feasible option in tree plantations, and several conservation methods
have been suggested (McCullough 1999, Rather and Mills 1989). However,
augmentative biological control has not been evaluated and is rarely used in Christmas

tree plantations.

52

The natural enemies of the balsam twig aphid tend to appear well after the
ﬁmdatrices hatch in the early spring, allowing them to mature with relatively little
predation pressure until the second generation is already feeding on the new foliage. This
makes predation alone unlikely to prevent shoot damage (Kleintjes 1997b, Rather and
Mills 1989, Fondren and McCullough, unpubl. data). The most common natural enemies
observed preying on M. abietinus include syrphid ﬂy larvae (Diptera: Syrphidae), several
coccinellids (Coleoptera: Coccinellidae), chrysopids (Neuroptera: Chrysopidae), and
hemerobiids (Neuroptera: Hemerobiidae) (Berthiaume 1998, Kleintjes 1997b, Nettleton
and Hain 1982, Varty 1966, Saunders 1969). No parasitoids are known to attack M.
abietinus in North America (Rather and Mills 1989).

We were interested in evaluating an augmentative release of a commercially
available predator for control of M. abietinus. Of the natural enemies associated with M.
abietinus in the ﬁeld, only chrysopids and coccinellids were commercially available. We
decided to experiment with Chrysoperla rufilabris Burmeister larvae. This generalist
species is commercially available, and C. ruﬁlabris larvae have successfully controlled a
closely related aphid (Mindarus kinseyi Voegtlin) on white ﬁr seedlings (Abies concolor
[Gord. & Glend.] Lindl.) in Califomia (Nordlund and Morrison 1990, Ehler and Kinsey
1995). Eggs of C. ruﬁlabris are also commercially available, but have often been shown
to provide inadequate control (Dreistadt et al. 1986, Ehler and Kinsey 1995).

Our objectives in this study were to: 1) describe the natural enemy complex of M.
abietinus present in unsprayed ﬁr Christmas tree ﬁelds; 2) assess the effectiveness of

Chrysoperla rufilabris larvae as predators of M. abietinus in the laboratory and in ﬁeld

53

cages; and 3) evaluate the potential effectiveness of C. ruﬁlabris in an open ﬁeld release,

as might be done on a commercial Christmas tree plantation.

Methods

Study sites. This research was conducted in 1999, 2000 and 2001 on three
commercial Christmas tree farms in Lower Michigan. Fields were located in Antrim
County (44°59’N, 85°06’W), Grand Traverse County (44°32’N, 85°29’W), and Ingham
County (42°44’N, 84°33’W), and consisted of either balsam ﬁr or a mixture of balsam
and Fraser ﬁr. Trees in each ﬁeld were approximately 7-8 years old in 1999. NO
insecticides, fertilizers or irrigation were used in any ﬁeld during our study. All ﬁelds
were planted at the standard 1.83m by 1.83m (6 by 6 ft) spacing. Preliminary studies in
1998 and 1999 (F ondren and McCullough, unpubl. data) indicated that a M. abietinus
population was present in each ﬁeld.

Objective 1. Natural enemy complex. We began to monitor natural enemy
activity in each ﬁeld early in the spring, before aphid eggs hatched. We visited each ﬁeld
weekly throughout the aphid’s life cycle, from late March or early April through late June
or early July, to survey predators and monitor the development of the aphid population.

In April 1999, we selected 40 trees in each ﬁeld based on the presence of needle
curling damage on the year-old shoots. From these 40 trees, we randomly selected ten
trees each week during the aphid life cycle and clipped branch tips at midcrown level in

two randomly selected cardinal directions on the tree. These clipped branch tips were

54

immediately placed in sealed plastic bags, taken to the laboratory in a cooler and ﬁlled
with 70% ETOH. They were stored in the laboratory until they could be examined (per
Varty 1966, Ehler and Kinsey 1985). When time permitted, each branch tip was
dissected under the microscope, and M. abietinus and natural enemies were removed with
a ﬁne brush. The number and stage of aphids and natural enemies were recorded.
Although this method allowed us to get a close approximation of the actual numbers of
aphids and predators on a given branch tip, it favored the discovery of small predators
such as the larvae of syrphid ﬂies, chrysopids, hemerobiids and coccinellids. Adult
predators that could take ﬂight as shoots were clipped were not well represented. Also,
the clipped branch method was destructive, and could have eventually affected the
appearance and value of the trees. Therefore, we decided to use an additional method for
sampling aphids and natural enemies that was not destructive.

In 1999, we used a non-destructive method (‘the beat method’)(Kleintjes et al.
1999) to sample aphids and natural enemies from a random sample of ten trees in each
ﬁeld, in addition to the 40 labeled trees. The ten trees were selected at 20 m intervals
along diagonal transects through each ﬁeld, regardless of aphid damage. Natural enemies
and aphids were sampled weekly on the ten trees by rapping the rrridcrown foliage on two
aspects three times with a dowel, and counting the number of aphids and predators that
fell onto black cotton cloth in a 22.9 cm diameter embroidery hoop. The midcrown level
was chosen because M. abietinus tends to colonize the midcrown level most consistently
(Nettleton and Hain 1982). To add another aspect to our survey of natural enemies in

1999, we also inspected each face (cardinal direction) of 20 trees per ﬁeld each week

55

from 17 May to 21 June, and recorded the number of adult coccinellids, larval
coccinellids, lacewing larvae, and syrphid larvae that were visible on each face.

The ‘beat method’ was more efﬁcient for aphid and natural enemy sampling than
the clipped branch method, and was adopted for all sample trees in 2000 and 2001. In
2000, we sampled all labeled trees (total of n = 50) in each ﬁeld weekly using this
method. However, this sample size was time consuming. We plotted the mean and
variance of the number of aphids sampled per tree and determined that a sample size of
30 trees in 2001 would give us adequate information. Therefore, in 2001, we randomly
selected 20 trees from among the 40 trees with aphid damage that we had originally
marked in 1999, and continued weekly sampling of the other ten trees using the beat
method.

In 2000 and 2001, we also set unbaited yellow sticky traps (22.9 cm x 28.0 cm)
(Pherocon AM unbaited, Trécé Inc., Salinas, CA) in each ﬁeld each week, to trap adult
ﬂying predators. Yellow sticky traps have been previously used to collect adult
chrysopids, coccinellids and syrphid ﬂies (Neuenschwander 1984, Ricci 1986, Bowie et
al. 1999). Four to ﬁve traps, widely spaced throughout the ﬁeld, were tied to an open tree
branch at the top or midcrown level so that the trap could hang freely. Traps were set out
and collected weekly in 2000 and biweekly in 2001, from early April to mid July. Traps
were soaked in Histo—Clear II (National Diagnostics, Atlanta, GA) to free the insects. All
of the insects present on the traps were removed in the lab, and adult chrysopids,
hemerobiids, coccinellids and syrphids were counted and identiﬁed to species when

possible.

56

We identiﬁed adult Chrysopidae to genus using Brooks and Barnard (1990), and
Chrysoperla species were identiﬁed using Brooks (1994). Adult Hemerobiidae were
identiﬁed to genus using Oswald (1993), and Hemerobius species were determined with
Klimaszewski and Kevan (1985). Coccinellids were identiﬁed to species using Gordon
(1985). Adult syrphid ﬂies were identiﬁed with Vockeroth (1969). Voucher specimens
were deposited in the A.J. Cook Arthropod Research Collection, Michigan State
University, voucher number 2002-01.

Objective 2: Evaluation of C. ruﬂlabris in lab and field cages. Chrysoperla
ruﬁlabris larvae were ordered from Beneﬁcial Insectary, Oak Run, CA. First instar
larvae were shipped overnight in plastic bottles of 1,000 larvae mixed with rice hulls.
The bottles were packed in a box with small ice packs to keep them cool.

Laboratory trials. In 2000, we tested C. ruﬁlabris in the laboratory to determine
if it would readily consume M. abietinus. This was important because even generalist
chrysopids may dislike some prey (New 1975, Tauber et al. 2000). Mindarus abietinus
are relatively small aphids and are covered with a white waxy substance, perhaps
dissuading some potential predators.

The number of M. abietinus that C. ruﬁlabris larvae would consume was
determined by placing an uninfested branch tip of Abies balsamea in a water pic set
upright in a 30 x 30 x 30 cm mesh cage. One second-instar C. rufilabris was placed on
the branch tip and provided with 5, 10, 20, or 35 M. abietinus per day. Aphids were
collected from infested ﬁr trees in the Ingham County ﬁeld and were usually fourth-instar
sexuparae, the most abundant stage and largest morph (V arty 1968). Chrysoperla

ruﬁlabris larvae were observed daily and the number of days to the third stadium and

57

pupation were recorded. Live adult C. ruﬁlabris were killed immediately with carbon
dioxide and weighed. Only nine adult C. ruﬁlabris were found alive; the others could not
be weighed accurately because dessication begins very rapidly after death.

Field cage trials. On 22/23 May 2000, we selected 30 trees with a heavily
infested branch and 30 trees with a moderately infested branch in the Ingham County
ﬁeld (60 trees total). At that time, the aphids were still 25% sexuparae, with the rest in
the sexuales stage (males and females). Heavily infested branches were deﬁned as
having more than 65% of the current-year shoots heavily damaged by M. abietinus (30%
or more of the needles curled). Moderately infested branches had less than 50% of the
shoots damaged (30% or more of the needles curled). These branches were caged with
grey nylon mesh sleeve cages. Predators observed on the branches were removed before
the cages were closed by tying a string around the end of the cage. Each moderately and
heavily infested branch was randomly assigned to one of three treatments: zero, one, or
ﬁve C. ruﬁlabris larvae, for a total of 10 trees of each treatment in each of the two
groups. Lacewing larvae were added to the cages on 23 May, by gently placing them on
the foliage of the branches with a camelhair brush. When ﬁve larvae were added, they
were placed in different spots rather than all in one place. The cages were checked
occasionally to make sure other predators did not get in. After aphid oviposition, we
clipped the branch tips with the cages and returned them to the laboratory.

In 2001, methods were modiﬁed slightly. First, we chose 30 infested (2 1 stem
mother present in a beat sample) trees in the Ingham County ﬁeld just after budbreak on
1-2 May. Three infested branches were selected on each tree. To make sure a branch tip

was infested, we observed at least one stem mother on it. Of these three branches, one

58

was tagged but not caged, one was caged but the end was left open (per Grasswitz and
Burts 1995), and one was caged, predators were removed, and the cage was closed (any
remaining predators were removed on 8 May). Our results in 2000 had indicated no
signiﬁcant differences between one and ﬁve lacewings per cage, so on 15 May, when
aphids were all sexuparae, only one lacewing was added to 15 of the 30 closed cages and
to 15 of the 30 open cages. The open cages that received a lacewing were cleared of
other predators and then the cage was closed.

In both 2000 and 2001, branch tips from each of the cages each year were clipped
and returned to the laboratory following aphid oviposition in mid-summer. The density
of aphid eggs per cm of current-year foliage was used to estimate aphid density. The
total number of eggs found in each sample was divided by the total length of the current-
year shoots found in each sample.

Objective 3: Field test of C. ruﬁlabris. In 2001, we conducted open releases of
C. ruﬁlabris in all three ﬁelds. In each ﬁeld, we selected 20 infested trees at opposite
ends of the section of trees, approximately 100 m apart. The initial aphid density at each
end of the ﬁeld was similar. Lacewings were released at one end of the ﬁeld while the
trees at the other end served as controls. In each ﬁeld, the rows ran from north to south.
Lacewings were released on the trees in the south end on a calm day to reduce the
likelihood of drift. We used ﬁrst instar C. ruﬁlabris larvae shipped from Beneﬁcial
Insectary (Oak Run, CA) in bottles of 1,000 each, mixed with rice hulls. We released
approximately 25 larvae (1/20 of a bottle) onto each tree in the lacewing treatment group
by shaking a premeasured amount of the lacewing and rice hull rrrixture onto the upper

canopy of the tree. The possibility of larvae dispersing among trees was limited because

59

trees were spaced widely and there was no contact between foliage on adjacent trees. In
addition, chrysopid larvae tend to have a negative geotactic response (New 1975), which
likely further limits dispersal to neighboring trees.

Samples of M. abietinus were taken in each ﬁeld weekly using the beat sample
method, a few days before treatment and for two weeks afterwards. In Ingham County,
we took pretreatment aphid samples on 10 May 2001. In the other two ﬁelds,
pretreatment samples were taken on 14 May. All aphids were in the second generation
on the release dates. In Ingham County, the release took place on 16 May 2001. In
Grand Traverse and Antrim Counties, the ﬁrst lacewing release was on 18 May.

In Antrim and Grand Traverse Counties, we released additional C. ruﬁlabris
larvae on 30 May, because we had observed very high populations ofM. abietinus on 23
May, especially in Antrim County. We replicated this second release in Grand Traverse
County because aphid phenology and degree day accumulation were similar to Antrim
County. The second release allowed us to see if a larger number of C. ruﬁlabris would
be necessary to effect a change in the aphid population. We added approximately 400
lacewing larvae (1/2 bottle) to ten trees randomly selected from the 20 trees that received
the ﬁrst lacewing release.

In addition to sampling the aphid population using the beat method, we counted
aphid eggs after oviposition had occurred. We clipped branch tips from the midcrown
level of each of the trees used in the open release experiment and examined them under a
microscope. Number of aphid eggs and the length of all current-year shoots was

recorded. Aphid egg density was expressed as the total number of eggs per cm of

.60

current-year foliage. Length and width of each sample branch tip was also recorded, so
egg density could also be expressed as aphid eggs per cm‘.
Statistical Analysis. The number of eggs/cm in the ﬁeld cages in 2000 was

square root transformed to meet the assumptions of ANOVA. In 2001, the number of
eggs per cm2 of new foliage in the ﬁeld cages was transformed by taking the natural

logarithm of the square roots. When the global ANOVA was signiﬁcant, means were
separated using Fisher’s Least Signiﬁcant Difference. For t tests, variances were tested
for homogeneity; if equal, pooled t tests were used. If variances were not homogeneous,
the Welch test was used (Welch 1938). Differences between one and ﬁve lacewing
larvae per cage in 2000 were tested with pooled t tests ((PROC TTEST, SAS v8)(SAS
Institute, 1999)).

The number of eggs/cm of new foliage in the open ﬁeld lacewing releases was
tested for normality using the Shapiro-Wilk test (PROC UNIVARIATE, SAS v8). Given
normality of the treatment groups, analysis of variance or t-tests were used to detect
treatment differences. In Ingham County, the Kruskal-Wallis test was used to determine
differences in the egg density on trees treated with lacewing larvae (Sokal and Rohlf

1995). All tests were conducted at a signiﬁcance level of 01 = 0.05.

Results

Objective 1. Natural enemy complex. Syrphid larvae were rarely observed

until shoots were dissected. Many syrphid larvae were found in the 1999 branch tip

61

Wﬂim M

samples and abundance of syrphid larvae ﬂuctuated over time in synchrony with the
number of aphids in the samples (Figure 1). Although syrphid larvae could not be
identiﬁed to species, we were able to rear some to the adult stage in the laboratory. We
also collected adult syrphid ﬂies hovering near or on ﬁr trees whenever possible. Genera
of adult syrphids found or reared out included Eupeodes spp., Syrphus spp.,
Sphaerophoria spp., Allograpta obliqua, and T oxomerus spp. Each of the genera found
had also been previously reported in association with M. abietinus or the closely related
M. kinseyi Voegtlin (e.g. Ehler and Kinsey 1995).

In 1999, we combined the number of predators found in the beat samples (n = 10
trees per ﬁeld) with visual observations of the number of predators visible on each tree.
The results were variable among ﬁelds (Figure 2). In Ingham County, the abundance of
larval coccinellids observed was 200% and 300% higher than other predators on 30 May
and 7 June, respectively. In Grand Traverse County, the number of larval coccinellids
was also high--130 on 8 June and 87 on 15 June, compared to a total of 6-18 other
predators. The number of adult coccinellids observed in Grand Traverse County was
higher than in Ingham County (a high of 91 in Grand Traverse Co. vs. a high of 5 in
Ingham Co.). In Antrim County, the number of predators observed in visual counts was
low, but lacewing larvae were most frequently observed, especially in beat samples.

We Observed seven species of coccinellids in association with M. abietinus
infestations, including Harmonia axyridis (Pallas), Anatis mali (Say), A. labiculata (Say),
Mulsantina picta (Randall), M. hudsonica (Casey), Coccinella trifasciata perplexa
Mulsant, and C. tranversoguttata richardsoni Brown. Table 1 shows the ﬁrst and last

date of collection for each of these species in each ﬁeld. The most abundant coccinellid,

62

in general, was Harmonia axyridis, an introduced coccinellid. The native coccinellid
Anatis mali was relatively common, but far outnumbered by H. axyridis. Harmonia
axyridis is well known to overwinter as adults, and in the ﬁeld we observed adult H.
axyridis as soon as weather warmed in the spring.

In the beat samples in 2000, lacewing larvae (both Chrysopidae and
Hemerobiidae) were some of the most common predators present (Figure 3). Syrphid
larvae were also relatively common. Notably, most predators did not appear in the beat
samples in 2000 until approximately three weeks after the stem mothers had hatched.
Hemerobiid adults were usually among the ﬁrst aphid predators to be observed in the beat
samples, although their numbers remained low, probably because the sampling method
favored less mobile insects. We identiﬁed the hemerobiid adults as Hemerobius stigma
Stephens, a common species known to prefer arboreal habitats and conifers in particular
(Throne 1971, Stelzl and Devetak 1999). Hemerobius stigma overwinters as adults
(Klimaszewski and Kevan 1985).

In 2000, the sticky traps caught several of the same predators. Adult hemerobiids
were among the ﬁrst aphid predators to appear, particularly in Ingham and Antrim
Counties (Table 2). As in the beat samples, the most abundant hemerobiid was H.
stigma. Adult chrysopids were also present in each ﬁeld, although they were usually
caught on the sticky traps later in the season. The most abundant species was Chrysopa
oculata, which is predaceous as an adult as well as a larva (most green lacewings are not
predaceous as adults). Syrphid adults found ovipositing or hovering near aphid colonies,
or collected on sticky traps, were identiﬁed to species where possible. All were members

of the subfamily Syrphinae, which are homopteran predators in the larval stage (Sadeghi

63

and Gilbert 2000). The most common genus found was Syrphus spp., followed by
Eupeodes spp., Allograpta obliqua (Say), and Sphaerophoria spp. The adult coccinellids
that could be identiﬁed from the 2000 sticky traps included several species present in
more than one ﬁeld (Table 3)

Bumble bees (Bombus spp.) were observed foraging on honeydew in 1999, 2000,
and 2001, especially in Grand Traverse County. We did not take any samples or detailed
observations, but we mention them here because it is highly unusual to observe large
numbers of bumblebees foraging on aphid honeydew (Batra 1993). Future researchers
may wish to address this phenomenon in more detail.

Objective 2: Evaluation of C. ruﬁlabris in lab and field cages. Chrysoperla
ruﬁlabris larvae readily consumed M. abietinus in laboratory trials. To complete
development on a diet of only M. abietinus, C. ruﬁlabris larvae required at least 10
aphids per day. Larvae in the cages consumed the maximum of 35 aphids per day that we
provided. Lacewings fed ﬁve aphids per day survived to the third instar but did not
pupate. Only ten adult lacewings were weighed successfully, but the average adult
weight increased with the number of aphids consumed. On ten aphids per day, the
average weight of the four adults was 4.42 mg i 0.59. The average weight of the two
adults reared on 20 aphids per day was 4.90 mg d: 2.27. On 35 aphids per day, the
average adult weight of four adults was 6.39 mg i 1.08.

2000 ﬁeld cages. In ﬁeld cages in 2000, results differed between moderate and
heavily infested branches. On the heavily infested branches, the mean egg density in
cages with no lacewing was nearly 50% higher than mean egg density in the cages with a

lacewing, but the difference was not statistically signiﬁcant (F = 1.31, df = 2,23; p =

64

0.29) (Figure 4). This may be due to the very high variability of the egg densities. On
the moderately infested branches, mean egg density was signiﬁcantly lower in the
presence of a C. ruﬁlabris larva (F = 3.75, df = 2,20; p = 0.04). There was no signiﬁcant
difference in the number of eggs/cm between one and ﬁve lacewings per cage at either
the heavy density (t = -O.57, df = 16, p = 0.58) or the moderate density (t = 1.15, df = 13,
p = 0.27).

2001 ﬁeld cages. In 2001, the presence of one lacewing larva in a cage did not
signiﬁcantly decrease egg density in either type of cage (‘closed’ or ‘open’). In the cages
that had been open and accessible to predators before lacewings were added, no
signiﬁcant difference was observed in egg density between cages with and without a
lacewing (F = 0.11; df = 1,27; p = 0.74) (Figure 5). However, the standard errors,
especially in the ‘no lacewing’ treatment, were relatively high. Similarly, the difference
between egg density in the closed cages with or without a lacewing was not signiﬁcant (F
= 0.35; df = 1,28; p = 0.56). The variability in the closed cages, as indicated by the
standard errors, was extremely high.

Objective 3: Field test of C. ruﬁlabris. Pre-treatrnent aphid density in Ingham
County was relatively high (mean of 35 i 2.4 aphids sampled per tree) compared to
Grand Traverse County (Figure 6). In Ingham County, there was no signiﬁcant
difference in aphid density between the two treatment groups on 10 May (t = -1.35, df =
38, p = 0.19) (Figure 6). A week after treatment, on 21 May, no signiﬁcant difference
was apparent between the two treatment groups (t = 0.16, df = 38, p = 0.87). On 4 June,
however, aphid numbers on the lacewing-release trees were 36% lower than aphid

numbers on the control trees (t = -2.45, df = 38, p = 0.02). In Ingham County, M.

65

abietinus egg density was also signiﬁcantly lower on the trees treated with C. ruﬁlabris
(Kruskal-Wallis H = 38.9, df = 1; p < 0.0001) (Figure 7).

In Antrim County on 14 May 2001, aphid density was also relatively high (32 :1:
4.5 aphids per tree). No signiﬁcant difference in aphid density between treatment groups
was observed on 14 May (four days before the ﬁrst lacewing release) or on 23 May, ﬁve
days after the ﬁrst lacewing release (Figure 6). On 6 June, however, aphid numbers on
the control trees were 66% higher than either lacewing treatment (F = 12.05; df = 2,37; p
<0.0001). This relationship remained consistent on 14 June (F = 9.44; df = 2,37; p =
0.0005). On both 6 June and 14 June, no difference in aphid sample numbers was
observed between the trees that received one or two applications of lacewing larvae
(Figure 6). The density of M. abietinus eggs was signiﬁcantly lower on the trees treated
with C. ruﬁlabris (F = 21.9; (if = 2,160; p < 0.0001) than on untreated trees, although
there was no signiﬁcant difference in egg density between the trees treated with one or
two applications of C. ruﬁlabris (t = -0.39, df = 82, p = 0.70) (Figure 7).

In Grand Traverse County on 14 May 2001, average aphid density was lower and
relatively more variable than in the other two counties (mean of 15 d: 4.1 aphids per tree).
although differences in aphid density among treatment groups were not signiﬁcant
(Figure 6). Differences among treatments were not observed until 14 June, two weeks
after the second lacewing release. On 14 June, aphid numbers on the control trees were
signiﬁcantly lower than either lacewing treatment (F = 4.02; df = 2,37; p = 0.03), but the
difference was small, approximately 7 aphids per tree. Egg density was not signiﬁcantly

different among the three treatment groups (F = 1.83; df = 2,74; p = 0.17).

66

Discussion

Objective 1. Natural enemy complex. Mindarus abietinus is associated with a
diverse complex of natural enemies in Michigan Christmas tree ﬁelds. Our observations
in the three Christmas tree ﬁelds conﬁrm other reports in the literature of known
predators of M. abietinus (Berthiaume et al. 2000, Kleintjes 1997b, Nettleton and Hain
1982, Ehler and Kinsey 1985). However, while several predators attack M. abietinus,
they do not always keep the aphid below damaging levels (Rather and Mills 1989). One
reason for this may be the early start of the aphid life cycle; the stern mothers
(fundatrices) are present at a time when few predators are active. When fundatrices begin
to reproduce, the rate of increase is so high (average fecundity 22-41 offspring (Varty
1968)) that predators may not be able to reduce the population before damage occurs.

Another reason for the lack of early season natural enemy activity may be the
ability of M. abietinus to survive cold temperatures and late frosts in early spring.
Mindarus abietinus exhibits a remarkable ability to survive in adverse conditions; for
example, M. abietinus likely evolved in the boreal forests, and the brief life cycle of most
Mindarus probably reﬂects an adaptation to short subarctic summers (Hille Ris Lambers
1966). In Michigan, the growing season is relatively longer than in the far north, and the
complex of natural enemies that prey on M. abietinus in Michigan tend to have longer life
cycles and generation times. Also, most are generalists not tightly tied in to the M.
abietinus cycle. One exception may be the coccinellid Anatis mali, an indigenous

coccinellid that is associated with coniferous forests and tends to become active relatively

67

early in the spring (Berthiaume 1998). Another may be the syrphid Allograpta obliqua,
which we observed ovipositing directly in active colonies of M. abietinus.

The brief life cycle of Mindarus presents a challenge for conservation biological
control strategies. With an aphid that completes the cycle from egg hatching to
oviposition only once per year in a relatively brief time span, predators may not have the
opportunity to mount a density-dependent response to a rapid increase in prey density.
However, we did observe that the average number of aphid predators tended to increase
as the aphid population grew, appearing too late to reduce damage but potentially
affecting the aphid population for the next year. When we monitored uncaged branch
tips in 2001, the density of aphid eggs on these branches was signiﬁcantly lower (by
63%) than the egg density in the cages where natural enemies were excluded.

The apparent synchrony between aphid numbers and syrphid larvae on infested
branch tips presents interesting possibilities. Adult syrphid ﬂies use aphid colonies as
cues for oviposition (Sadeghi and Gilbert 2000), and we observed this occurring on
several occasions in the ﬁeld. Syrphid ﬂies may have a signiﬁcant effect on aphid
populations—larvae of Syrphus ribesii can eat up to 52 pea aphids per day (Chambers et
al. 1983). Possibly, the impact of syrphid larvae has been underestimated, as they are not
readily visible when feeding on aphids within curled needles, and tend to be nocturnal.
Other studies on aphid predators have also recognized the potential to underestimate the

impact of syrphid larvae preying on aphids (Chambers et al. 1983).

68

Objective 2. Effectiveness of C. ruﬁlabris larvae in the lab and field cages.

Our results indicated that under some conditions, C. ruﬁlabris is capable of
reducing M. abietinus population levels. In the laboratory, C. ruﬁlabris proved a
voracious predator of M. abietinus. This agrees with Ehler and Kinsey’s ﬁndings (1985).

Results in the ﬁeld cages were varied. Chrysoperla ruﬁlabris larvae did
signiﬁcantly reduce egg density in the moderately infested cages in 2000, but not
signiﬁcantly in the heavily infested cages, or in any cage in 2001. However, the mean
egg density in the cages with no lacewing was often highly variable, which may have
prevented us from seeing signiﬁcant differences.

Although ﬁeld cages have the beneﬁt of conﬁning a known number of predators
to a given space, they also have disadvantages. In 2000, the heavily infested cages were
literally sticky with honeydew and crowded with aphids; such extreme densities were
probably much higher than would normally be observed. This may have prevented us
from seeing an effect ﬁom the lacewing larvae in the cage. The very high density of
aphids may have affected the reproductive capacity of the sexuparae (Chambers et al.
1983), which would also have affected the results. In 2001, we caged the branch tips
early in the season, but were unable to add the lacewings until later, so the aphid density
inside the cages grew to unnaturally high levels. We might have avoided this situation if
we had added the lacewings earlier.

The uncaged, tagged branches in 2001 had signiﬁcantly lower (by 33%) egg
densities than caged branches with a lacewing. Of course, uncaged branches were
exposed to natural enemies in the ﬁeld, and the alatae produced on the uncaged branches

had the opportunity to disperse while those on the caged branches did not. Nevertheless,

69

the mean egg density in cages with one lacewing was consistently lower than in cages
with no lacewing, if not statistically signiﬁcant. Further work is needed to describe the
relationship of M. abietinus egg density to aphid populations the next year.

Objective 3. Open field release of C. ruﬁlabris. In our open ﬁeld releases, C.
ruﬁlabris effectively reduced the M. abietinus population in two out of three ﬁelds. The
two ﬁelds with signiﬁcant effects (Ingham and Antrim Counties) had relatively high
populations of M. abietinus, indicating perhaps that the observable effect of C. ruﬁlabris
may be limited to higher densities. However, the effect may also have been apparent
because variability among aphid densities in Ingham and Antrim Counties was lower than
in the Grand Traverse County ﬁeld. In addition, the small lacewing larvae may not have
been able to search and disperse efﬁciently enough to have an effect on the relatively
sparse population of M. abietinus in Grand Traverse County. .

Another possible reason for the lack of signiﬁcant effects in the Grand Traverse
County ﬁeld is the timing of our release, when aphids were already well into the second
generation. We would have liked to release the lacewing larvae earlier, to test our
hypothesis that early season release would be effective. However, practical
considerations prevented this. Further research is needed to determine if C. ruﬁlabris
larvae can effectively reduce the aphid population if applied early in the year when
ﬁmdatrices are present and fewer natural enemies are active. In the early spring in
Christmas tree plantations, M. abietinus may be the only prey readily available to C.
rufilabris larvae. Mindarus abietinus has a relatively short life cycle (about eight weeks
ﬁ‘om egg hatch to oviposition) making a relatively short-lived augmentative control

method more feasible for M. abietinus than for an aphid that has several generations.

70

The lack of signiﬁcant differences between one or two lacewing applications may
have been due to several factors. The second release of lacewings took place a week
later, possibly too late to cause a signiﬁcant reduction in the aphid population. More
likely, Chrysoperla larvae are well known to be cannibalistic, and the large number of
larvae probably resulted in intraspeciﬁc predation. The higher competition pressure may
also have resulted in lowered survival of the lacewings in the second release. On the
other hand, this indicates that there may be a ‘threshold’ beyond which additional
lacewing larvae will not increase the effectiveness. Further studies may be warranted to
determine where this threshold occurs.

Practical applications. Although many studies have documented the potential of
chrysopids to consume insect pests in the laboratory, the release of chrySOpid larvae at
commercially feasible rates has not been tested extensively (Tauber et al. 2000). Despite
its potential for effective control, we recognize that augmentative releases of C. ruﬁlabris
larvae may not yet be economical for large plantations. However, they may be useful for
smaller farms or even ornamental trees with patchy infestations of M. abietinus.
Possibly, the emerging interest in organically grown Christmas trees may eventually
provide a lucrative market for Christmas trees grown without pesticides.

The bottles of 1,000 C. ruﬁlabris larvae that we used retail for an average of US
$30, including shipping. This is relatively expensive, since we used one bottle to treat 20
infested trees, and Christmas trees average 1200 trees per acre. However, C. rufilabris
can be purchased as eggs (a card of 1,000 eggs sells for approximately $5), and wholesale
prices on bulk orders would likely be lower. We did not test the effectiveness of using

eggs of C. ruﬁlabris for control in the ﬁeld, partly because other workers have reported

71

less success with eggs than with larvae (Daane et a1. 1996, Daane and Yokota 1997,
Ridgway and Murphy 1984), and eggs are vulnerable to predation by ants (Dreistadt et al.
1986). However, the lower cost of using eggs may warrant further investigation.

Mechanical methods such as backpack Sprayers and tractor-pulled Sprayers can be
used for the release of C. ruﬁlabris (Gardner and Giles 1996, 1997). Methods for the
release of C. ruﬁlabris larvae mixed with sawdust have been developed for use in cotton
(Ridgeway et al. 1977, Kinzer 1976). Currently, a wide variety of release methods are
being tested (Tauber et al. 2000), some of which may be potential candidates for use in
Christmas trees. However, the effectiveness of mass-released lacewings can vary
signiﬁcantly with the release rate, release method, and even intraguild predation in the
ﬁeld (Tauber et al. 2000). Further reﬁnement of the release methods and ﬁeld tests are
needed to determine an appropriate and cost-effective method of using lacewings in
Christmas tree ﬁelds.

An interesting possibility with the augmentative use of Chrysoperla includes the
judicious use of chemical insecticides, since many Chrysoperla species have shown
resistance to common insecticides (Tauber et al. 2000, Rumpf et al. 1997, Ridgway and
Jones 1968). Conservation of existing natural enemies can also be achieved by reducing
the number of insecticide sprays. Many natural enemies such as adult syrphid ﬂies
consume ﬂower nectar, and providing ﬂowering plants nearby may help to encourage
populations of natural enemies. An integrated management program combining
conservation of existing natural enemies, effective timing of insecticide applications, and
occasional use of C. ruﬁlabris larvae can work to provide effective and environmentally

sound control of M. abietinus in ﬁr Christmas trees.

72

References Cited

Batra, S.W.T. 1993. Opportunistic bumblebees congregate to feed at rare, distant alpine
honeydew bonanzas. J. Kansas Entomol. Soc. 66(1): 125-127.

Berthiaume, R. 1998. Les ennemis naturels du puceron des pousses du sapin, Mindarus
abietinus Koch (Homoptera: Aphididae), avec une emphase particuliere sur les
coccinelles Anatis mali Say et Harmonia axyridis Pallas. Mémoire de maitrise,
Université Laval, Quebec, Canada.

Berthiaume, R., Ch. Hébert and C. Cloutier. 2000. Predation on Mindarus abietinus
infesting balsam ﬁr grown as Christmas trees: the impact of coccinellid larval predation
with emphasis on Anatis mali. BioControl 45: 425-438.

Berthiaume, R., Ch. Hébert and C. Cloutier. 2001. Podabrus rugosulus (Coleoptera:
Cantharidae), an opportunist predator of Mindarus abietinus (Hemiptera [sic]: Aphididae)
in Christmas tree plantations. Can. Ent. 133: 151-154.

Bowie, M.B., G.M. Gurr, Z. Hossain, L.R. Baggen, and CM Frampton. 1999.
Effects of distance from ﬁeld edge on aphidophagous insects in a wheat crop and
observations on trap design and placement. Int. J. Pest Mgrnt. 45(1): 69-73.

Bradbury, R.L., and E.A. Osgood. 1986. Chemical control of balsam twig aphid,
Mindarus abietinus Koch (Homoptera: Aphididae). Maine Agric. Exp. Stn. Tech. Bull.
124.

Brooks, S. J. 1994. A taxonomic review of the common green lacewing genus
Chrysoperla (Neuroptera: Chrysopidae). Bull. Br. Nat. Hist. (Ent.). 63(2): 137-210.

Brooks, SJ. and RC. Bernard. 1990. The green lacewings of the world: a generic
review (Neuroptera: Chrysopidae). Bull. Br. Mus. Nat. Hist. (Ent.) 59(2): 117-286.

Carter, CL, and J.F.A. Nichols. 1985. Some resistance features of trees that inﬂuence
the establishment and development of aphid colonies. Zeitschrift fuer Angewandte
Entomologie 99: 64-67.

Chambers, R.J., K.D. Sunderland, I.J. Wyatt and G.P. Vickerman. 1983. The
effects of predator exclusion and caging on cereal aphids in winter wheat. J. Applied
Ecology 20: 209-224.

Daane, KM. and G.Y. Yokota. 1997. Release strategies affect survival and
distribution of green lacewings (N europtera: Chrysopidae) in augmentation programs.
Environ. Entomol. 26(2): 455-464.

Daane, K.M., G.Y. Yokota, Y. Zheng, and KS. Hagen. 1996. Inundative release of
common green lacewings (Neuroptera: Chrysopidae) to suppress Erythroneura variabilis

73

and E. elegantula (Homoptera: Cicadellidae) in vineyards. Environ. Entomol. 25(5):
1224-1234.

DiFonzo, C.D. and D.G. McCullough. 1998. Potential impacts of the Food Quality
Protection Act on the Michigan Christmas tree industry. Michigan Christmas Tree
Joumal 45(1): 26-29.

Dreistadt, S.H., K.S. Hagen and D.L. Dahlsten. 1986. Predation by Iridomyrmex
humilis (Hymenoptera: Fonnicidae) on eggs of Chrysoperla cornea (Neuroptera:
Chrysopidae) released for inundative control ofIllinoia liriodendri (Homoptera:
Aphididae) infesting Liriodendron tulipifera. Entomophaga 31(4): 397-400.

«AL-9N5"
1 .

Ehler, LE. and M.G. Kinsey. 1995. Ecology and Management of Mindarus kinseyi
Voegtlin (Aphidoidea: Mindaridae) on White-Fir Seedlings at a California Forest
Nursery. Hilgardia 62:1.

Fondren, K. and D.G. McCullough. 2001. “GREEEN Research Update—Pine Needle
Scale”. Michigan Christmas Tree Journal 48(1):26-29.

M.- ”{1811 m r I up“.

Gardner, J. and K. Giles. 1996. Handling and environmental effects on viability of
mechanically dispensed green lacewing eggs. Biological Control 7: 245-250.

Gardner, J. and K. Giles. 1997. Mechanical distribution of Chrysoperla ruﬁlabris and
T richogramma pretiosum: survival and uniformity of discharge after spray dispersal in an
aqueous suspension. Biological Control 8: 138-142.

Gordon, RD. 1985. The Coccinellidae (Coleoptera) of America north of Mexico. J.
New York Ent. Soc. 93(1): 1-912.

Grasswitz, T.R., and EC. Burts. 1995. Effect of native natural enemies and
augmentative releases of Chrysoperla ruﬁlabris Burmeister and Aphidoletes aphidimyza
(Rondani) on the population dynamics of the green apple aphid, Aphis pomi De Geer.
Int]. J. Pest Management, 41(3): 176-183.

Hille Ris Lambers, D. 1966. Polymorphism in Aphididae. Ann. Rev. Entomol. 11:47-78.

Kinzer, R. E. 1976. Development of techniques for using Chrysopa cornea Stephens to
control Heliothis spp. in cotton. Ph.D. dissertation, Texas A&M University, College
Station, Texas.

Kleintjes, P.K. 1997a. Effects of integrating cultural tactics into the management of the
balsam twig aphid Mindarus abietinus Koch (Aphididae: Homoptera) in balsam ﬁr
Christmas tree plantations. pp. 112-121 in J.C. Gregoire, A.M. Liebhold, F.M. Stephen,
K.R. Day, and SM. Salom, eds. Proceedings: Integrating cultural tactics into the
management of bark beetle and reforestation pests. USDA Forest Service General
Technical Report NE-236.

74

Kleintjes, P.K. 1997b. Midseason insecticide treatment of balsam twig aphids
(Homoptera: Aphididae) and their aphidophagous predators in a Wisconsin Christmas
tree plantation. Environ. Entomol. 26: 1393-1397.

Kleintjes, P.K., E.E. Lemoine, J.Schroeder, and M.J. Solensky. 1999. Comparison of
methods for monitoring Mindarus abietinus (Homoptera: Aphididae) and their potential
damage in Christmas tree plantations. J. Econ. Ent. 92(3): 638-643.

Klimaszewski, J. and D.K.McE. Kevan. 1985. The brown lacewing ﬂies of Canada
and Alaska (N europtera: Hemerobiidae). Part I. The Genus Hemerobius Linnaeus:
systematics, bionomics and distribution. McGill University, MacDonald College, Lyman
Entomological Museum and Research Laboratory, Memoire No. 15:iv + 1-119.

Mattson, W.J., R.A. Haack, R.K. Lawrence and D.A. Herms. 1989. Do balsam twig
aphids (Homoptera: Aphididae) lower tree susceptibility to spruce budworm? Can.
Entomol. 121: 93-103.

McCullough, D.G. 1999. Biological control in Christmas tree plantations. In: D.G.
McCullough, S.A. Katovich, D.L. Mahr, D.D. Neumann, C.S. Sadof, and M.J. Raupp
[eds.], Biological control of insect pests in forested ecosystems: a manual for foresters,
Christmas tree growers and landscapers. Michigan State University Extension Bulletin
E-2679, pp. 75-91.

McCullough, D.G. and K. Fondren. 1998. F QPA and insecticide use in Michigan
Christmas tree ﬁelds: preliminary data from 1998 survey. Michigan Christmas Tree
Journal 45(3): 22-29.

McGugan, B.M. and BC. Coppel. 1962. Biological control of forest insects, 1910-
1958. In: A Review of the Bioloigical Control Attempts against Insects and Weeds in
Canada. Tech. Commun. Commonw. Inst. Biol. Contr., Trinidad 2:35-216.

Nettleton, W.A., and RP. Hain. 1982. The life history, foliage damage and control of
the balsam twig aphid, Mindarus abietinus Koch (Homoptera: Aphididae) in Fraser ﬁr
Christmas tree plantations of western North Carolina. Canadian Entomologist 114: 155-
162.

Neuenschwander, P. 1984. Sampling procedures for chrysopid populations. pp. 205-
212. In Canard, M., Y. Séméria, and TR. New [eds.], Biology of Chrysopidae. Dr W.
Junk Publishers, The Hague, Netherlands.

New, T.R. 1975. The biology of Chrysopidae and Hemerobiidae (N europtera), with

reference to their usage as biocontrol agents: a review. Trans. R. Ent. Soc. Lond. 127(2):
1 15-140.

75

Nordlund, D.A. and R.K. Morrison. 1990. Handling time, prey preference, and
functional response for Chrysoperla ruﬁlabris in the laboratory. Entomol. Exp. Appl. 57:
237-242.

Oswald, JD. 1993. Revision and cladistic analysis of the world genera of the family
Hemerobiidae(1nsecta: Neuroptera). J. New York Entomol. Soc. 101(2): 143-299

Raupp, M.J., C.S. Koehler and J .A. Davidson. 1992. Advances in implementing
integrated pest management for woody landscape plants. Ann. Rev. Entomol. 37:561-
585.

Rﬁther, M. and NJ. Mills. 1989. Possibilities for the biological control of the Christmas
tree pests, the balsam gall midge, Paradiplosis tumifex Gagne (Diptera: Cecidomyiidae)
and the balsam twig aphid, Mindarus abietinus Koch (Homoptera: Mindaridae), using
exotic enemies from Europe. Biocontrol News and Information, Vol. 10 no. 2.

Ricci, C. 1986. Beneﬁcial Coccinellidae caught in yellow traps in some Italian regions.
pp. 441-447 In I. Hodek (ed.), Ecology of Aphidophaga. Dr W. Junk Publishers,
Dordrecht, The Netherlands.

Ridgway, R.L. and S.L. Jones. 1968. F ield-cage releases of Chrysopa carnea for
suppression of populations of the bollworm and the tobacco budworm on cotton. J. Econ.
Entomol. 61(4): 892-898.

Ridgway, R.L., and W.L. Murphy. 1984. Biological control in the ﬁeld. pp. 220-228
In Canard, M., Semeria, Y., New. T.R. [eds.], Biology of Chrysopidae. Dr W Junk
Publishers, The Hague, The Netherlands.

Ridgeway, R.L., E.G. King and J.L. Carrillo. 1977. Augmentation of natural enemies
for control of plant pests in the western hemishpere. Chapter 13. Pp. 379-416. In:
Ridgeway, R.L. and SB. Vinson [eds.], Biological Control by augmentation of natural
enemies: Insect and mite control with parasites and predators. Plenum Press, New York.

Rumpf, S., C. Frampton, and B. Chapman. 1997. Acute toxicity of insecticides to
Micromus tasmaniae (N europtera: Hemerobiidae) and Chrysoperla cornea (N europtera:
Chrysopidae): LCso and LC90 estimates for various test durations. J. Econ. Entomol.
90(6): 1493-1499.

Sadeghi, H., and F. Gilbert. 2000. Oviposition preferences of aphidophagous
hoverﬂies. Ecological Entomol. 25: 91-100.

SAS Institute. 1999. SAS/STAT user’s manual, version 8. Cary, NC.

Saunders, J.L. 1969. Occurrence and control of the balsam twig aphid on Abies grandis
and A. concolor. J. Econ. Entomol. 62: 1106-1109.

76

Sokal, RR. and F.J. Rohlf. 1995. Biometry: The principles and practice of statistics in
biological research. New York, Freeman. 3rd ed.

Stelzl, M., and D. Devetak. 1999. Neuroptera in agricultural ecosystems. Agriculture,
Ecosystems and Environment 74: 305-321.

Tauber, M.J., C.A. Tauber, K.M. Daane, and KS. Hagen. 2000. Commercialization
of Predators: Recent lessons from green lacewings (Neuroptera: Chrysopidae:
Chrysoperla). Am. Entomol. 46(1): 26-38.

Throne, A.L. 1971. The Neuroptera-Suborder Planipennia of Wisconsin. Part II—
Hemerobiidae, Polystoechotidae and Sigyridae. Mich. Ent. 4: 79-87.

Varty, I.W. 1966. The seasonal history and population trends of the balsam twig aphid,
Mindarus abietinus Koch, in New Brunswick. Internal report M-12, Forest Research
Laboratory, Canada Department of Forestry, Fredericton, NB.

Varty, I.W. 1968. The seasonal history and population trends of the balsam twig aphid,
Mindarus abietinus Koch, in New Brunswick: polymorphism, rates of development and
seasonal distribution of populations. Internal report M-24, Forest Research Laboratory,

Canada Department of Forestry, Fredericton, NB.

Vockeroth, J.R. 1969. A revision of the genera of the Syrphini (Diptera: Syrphidae).
Mem. Ent. Soc. Canada 62: 1-176.

Welch, B. 1938. The signiﬁcance of the difference between two means when the
population variances are unequal. Biometrika 29: 350-362.

77

£2 :2 2:00 8 0000000

 

 

 

05: mN 05: M: 00:05
0025008022

0:2. on 0:3. On 0:554

>02 om >02 2.. 0000:0332

0:3. _ 0:3. _ 0000602
0:3 mm 0:3. _ 00:00.; 0:000 002003.20
$53025 :32
>02 cm >02 om SEES NE» 0.20.3000ERM.
0:3. _ >02 2N 00:02:02 .ﬁanwﬁ NE“. 0.000000%
0:3 _ 0:3. _ 0000602 Q0? 033.03
00: 8 02 cm 8205 2000 3.00% 002800
00039.3 .0201—
0::_. Om 0:2. on SE: 0005308 00.050on
0:3. mm 0:3. mm 03080—32 0&0 0.2006002?
>02 3 >02 2 :anwﬂ Nae. 0.000023%
830.06 .33..
>02 on >02 on 08000.2. 0:000 E0032 0R0~E0NN 0002000100 0:05.809
0:3. 2 0:3. 2 0000; 0:000 A>000Uv 002000000 0000000202
E? NN Ea< NN SEE. €85.00 00...: assess:
>02 cm >02 on 080005. 0:000 A>0mv 00030.53 0.000?
>02 om >02 2 0000; 0:000 A>0mv .0000 0.000?

:38

>02 cm =25 R :00st 00000:. 0:000 0000000000.: 0000006203000 0-05000W

20: N E... E 80:02
. 600050—32 .5935 .852 .0000; 0:000 Am0=0ac 0.528 0.30052
02:050000 :55.

as Ea

0800:00 0009 530000. 20E 0000.:—

 

0::_. On 3 Em< 2 80¢ £03800 0000 :0 0:0: >9 @3— :2 20m 05 :_ 0080:00 000.0030 .2 .20 0080000: 30:08; A 030,—.

78

 

VOOCNVMWNNOF‘ﬁ

VB \0
mmo‘m

m
o—
N—

o

w
2

oo—soo—‘ooooooommoooooo
oomﬂvm—noooommxoomomoomm

O~hm~E§BOO~BM~NWF~~O~N

vaxoomSEvv-zrvooooxoxohoovvv

.32 0.33. mm
:23. 3-2
=.a< 03
2:: 3-:
as: 8-:
2:: E-m

2:: 9.02 8
.02 00-:
002 2-2
002 2-0

02 209... mm
:23. m2:
:2? 03.

33 3.25: «N
2:: $-23 a
0:3 0->02 mm
02 00-:
32 t-»

.02 SE? a
E?- 3.00
E? 2.2

05.5 83.2.

.00 0230...”. 0:020

>.::0U E02»:—

 

m2=§800 ”208020 352080: ”2&5

0:0: 2

00:0Q

E03

 

doom 30¢ £000 3 300 >203 05 :0 0:00.“ 0:000:05 =33 .n 030%

79

 

 

w o o v v as: 8-:

v _ o v v as: 3-:

N o o _ v as: :6

ﬂ v ﬂ m w 83 932 3

ﬂ o o o v 32 3-:

_ o m o v :2 2-2

V o m w v :2 OZ beau 83.2
£3588 muaomEu £3825: €293 8.3 u 3:5 22m

 

6.2—3 N 03:8

80

E? N

 

 

 

 

 

 

 

N
E? a _ Sacaam
0:3. _N N
E? N _ 8.5.2
0:3. m _
:2 N _
$2 2 m
32 N m 8.5%; @520
$2 2 _
ba: N _
E? N N 8235 53 $23 s§§£
0:3 a _
32 N _ Osman—«M
0:3. c _
:2 ON _
N82 2 _ 85.2 Axmmv 3255 unmzﬁuab
.93 «N ﬂ SEES ﬁcowono 83322: $235830
in? a ~ 8.6%; €80 Ham—=2 uxﬁﬁmm SEQEFN £35809
>22 3 ~ ooumEaEM
:2 N N
E? mN N SEE
05;. 3 ~
>22 cN _ oﬂoZﬁH 6:80 =2,on .azcanéﬁt Suaawoﬁwanah £35,980
:2 2 _
.32 N _ 8.63:. EEO “magnum £23596: £35980
0:3. m N 8.52
0:3. m _
N32 3 _ oﬁogﬁ. EEO Axmmv m3: Macaw
So Hum 95 Ban 3.00:8 Non—5.2 >830 860%

 

.ooON .22» :28 E mash 305 no v5.0a 32:05800 mo 86on .n 035—.

81

 

 

 

 

 

32 N _ SE?
32 2 N
:22 N N 00.620; 2:20 30985 80.3. 32:00.33
:93 mN _ 0E0>8H @505 30009 8:28.33: 3:223:33
32 0N H
:2 N _
E? S N 8.52.
ENE mN : 00003; 96va
ENE aN : Emnwﬁ 90$ 0.20520qu SEQEQNRNE
05; a :
:02 a: : CONNQEEM 00:0: .333 DESEQI
So :00 nab 0:09 380:8 “0:852 3580 028%

 

3:8 N 29¢

82

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Grand Traverse County -----aphids
1800 1 31 158-, — + - Syrphid larvae
1500 -- 1534 -~ 30 a,
(U
£1200 q~ * 25 E
c «1 20
8 900 ~- E
o' ‘” ‘5 E
Z 600 1* m
w 10 -
O
300 -_ ~_ 5 z
0 0
27-Apr 06-May 12-May 18-May 26-May 01-Jun 08-Jun 15—Jun
Ingham County —-—aphids
2500 1 :3 — + - Syrphid larvae
/ \
2000 —— 1630 I \ -0 15 g
0)
E1500 ~- 5
3 —~ 10 .2
o- 1000 ~~ '8
Z s.
500 ~ "’ 5 "3
111 g
0 - 0
22-Apr 02-May 10-May 16-May 23-May 29-May 07-Jun 14-Jun
Antrim Coun .._+__ '
300 -- W 15 3pm? 1— 16
242 12 / A - + - Syrphid larvae
-- 12 m
(U
E
__ 3 <_0
.9
it
\ 23 "F 4 5‘
0 I — -’ '
0 J 1 1 1 1 1 v 1 2 1 1 0 2
27-Apr 06-May 12-May 19-May 26-May 02-Jun 08-Jun

 

 

Figure 1. Mindarus abietinus and syrphid larvae counts in clipped branch
tip samples, 1999.

83

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    
   

 

 

 

 

 

 

 

 

 

Ingham County —O—Adult coccinellids
20 ‘ - - I - - Larva coccinellids
,'.“ +Lacewing larvae
'8 15 ‘ ——x—syrphid larvae
Z .
a)
3 10 —
O
o'
2
17-May 24-May 30-May 07-Jun 14-Jun 21-Jun
Grand Traverse County —O—Adult coccinellids
140
I. -~l--Larval
120 ~ ' ‘. coccinellids
—A—- Lacewrng larvae
100 - ' .
§ ‘-
o 80 - .\
B
o 60 .
o“
2 40 l
20 -
O __ ‘ ~. ,
18-May 26-May 01-Jun 08-Jun 15-Jun 22-Jun
Antrim County —O—Adult coccinellids
12 - - - I - - Coccinellid larvae
q —A— lacewing larvae
11,10
CD —-)(-—syrphid larvae
Z 8 —
a)
m
.o
o
o’
2

 

19-May 26-May 02-Jun O8-Jun

 

15-Jun 22-Jun

 

 

Figure 2. Numbers of coccinellids, lacewings and syrphids
found in beat and visual examinations of sample trees, 1999.
In Antrim Co. in 1999 no survey was done on 2 June
because of rain. On 15 June a survey was done but no

predators were observed.

84

1 "“733

“-1 .‘lﬂﬂnl ‘8". ' _

 

 

 

 

 

 

 

 

 

 

  
 
 
 

 

 

 

30 _ —l—Lacewing larvae
25 " -t—Coccinellid larvae
20 .
15 ~ . .
Aphid hatch begins
10 . ,0 ~o \
,Q \0- ‘
5 . . / \\ / '
X """" x- ' - T A
0* 4‘— 1~~‘ r I x‘.r-- ..I
26- 06- 13- 22- 27- 03- 11- 17- 24— 01- 08- 16-
Mar Apr Apr Apr Apr May May May May Jun Jun Jun
20 ﬂ - 0- Syrphid larvae Grand Traverse County
—l—— Lacewing larvae
+ Coccinellid larvae
15 i ---x - - Hemerobiid adults
10 — . .
Aphid hatch begins
5 _
0 a . I j r ' '1' 1

 

 

 

28-Mar 04-Apr18-Apr 25-Apr 02- 10- 19- 25- 05-Jun14-Jun
May May May May

 

 

 

 

16 _ - O- Syrphid larvae Antrim County
+ Lacewing larvae

+ Coccinellid larvae
- - ->< - - Adult hemerobiids

  
  

121

 

 

 

8 — Aphid hatch begins

 

 

28- 04- 18- 25- 02- 10- 19- 26- 06- 13- 20-
Mar Apr Apr Apr May May May May Jun Jun Jun

 

Figure 3. Predators found in beat samples, 2000. Data are presented
as the number of each taxon collected per sample date.

85

 

 

 

 

.h

O
01

‘A)

'u
l

 

O
-‘ 01 N
1 I l I

.0 . .0 .
0949800)?

01
l

 

 

eggs/cm new foliage

 

 

 

 

O

 

 

1.4~ B)

 

 

 

 

 

0 1 5

No. lacewing larvae added to cage

Figure 4. Number of M. abietinus eggs per cm of current-year shoot
growth (mean +/- 1 SE) on branch tips from ﬁeld cages in 2000. A)
Heavily infested branches (n = 26). B) Moderately infested branches
(n = 23).

86

120

 

 

 

 

A 'Closed' ca as

8100 - ) g
g
«9 801
3
C 601
8
g 40 -
(I)
O)
8’ 20 ~

0 _

100

 

B) 'Open' cages

 

     

 

 

eggs/cm2 new foliage
N 4s o: oo
o o o o o
l 1_ 1 l I

No lacewing lacewing Not caged

Figure 5. Mean (+/- SE) number of aphid eggs/cm2 of current-year
foliage in 2001 ﬁeld cages. Letters above bars indicate signiﬁcant
differences. A) 'closed' cages were kept closed before lacewings
were added; B) 'open' cages had one end left open until lacewings
were added.

87

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ingham County —0— Lacewings
45 ~ l - No lacewings
a
8 30 _
g a
m
2 15 -
0
10'May 21-May 04-Jun
Antrim County +1 lacewing release
300 +2 lacewing releases
' - i - -no lacewings
a

1. ,L
:9 200 — .' . . . . .
E .' g a
m b
29 100 - b

b

O

 

 

 

 

14-May 23-May 06-Jun 14-Jun

 

 

Grand Traverse County _ o— .1 lacewing release

-—I—— 2 lacewing releases
- - t - - no lacewings

60

 

 

 

 

No. aphids
A
O

N
o
1

 

 

 

 

14-May 23-May 06-Jun 14-Jun

 

 

 

Figure 6. Mean (+/— SE) aphid counts before and after the
lacewing applications in the open ﬁeld releases in 2001. Data
within a column marked with different letters are signiﬁcantly
different (p < 0.05).

88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A) Ingham County lno lacewing

8: application
g 0'16 a Ellacewing
9 application

3 0.12 _

a)

C

E 0.08 1
E

8 0.04 -

(D
2 0.00 -

g B) Antrim County 21 No lacewings
'55“ 1.5 a IOne release
; 1.2 1 ElTwo releases
,af,

2 0.9 J 52:." ‘, b

E. 0.6 - :5 b

m in:

CS 0.0 .1”. *3"

2

C) Grand Traverse County El No lacewings

E, 0.10 IOne release
6;! 0.08 g DTwo releases
2 0.06 — a a

E 0.04 - a i

o- 0.00 ' 1

z

 

 

 

Figure 7. Mean (+/- SE) density of M. abietinus eggs on new foliage
in open releases of C. ruﬂlabris larvae in 2001.

89

Chapter 3

Phenology and natural enemies of pine needle scale (Chionaspis heterophyllae

(Fitch)) (Homoptera: Diaspididae) in Christmas tree ﬁelds

Abstract. Our objectives in this study were to: 1) determine the phenology of the
second generation of pine needle scale in calendar days and degree-day accumulations; 2)
characterize the natural enemy complex acting on the 2"d generation of pine needle scale
in Christmas tree ﬁelds in Michigan; and 3) determine the rates of predation and
parasitism on pine needle scale infestations on commercial Scotch pine (Pinus sylvestris
L.) Christmas tree ﬁelds. We monitored scale populations in three counties in Lower

Michigan for three years. The development of the second generation of pine needle scale
was more consistently associated with cumulative degree-days base 10°C (50°F) (DD50)

than calendar date. Egg hatching extended for a period of approximately three weeks.

The second instar (the hyaline stage) was the predominant stage present at roughly 1500
DD50. The natural enemy complex acting on pine needle scale was similar in all ﬁelds

and years. The coccinellids Chilocorus stigma (Say) and Microweisia misella (LeConte)
were most ﬁ'equently found in all ﬁelds. An endoparasitic wasp, Encarsia bella Gahan
(Hymenoptera: Aphelinidae), and a hyperparasitoid, Marietta mexicana Howard
(Hymenoptera: Aphelinidae), were also present. Predaceous mites were observed but not
monitored closely. The percentage of predation was high in unsprayed ﬁelds in 1999,
averaging 70% predation and leaving few adult females to overwinter. Parasitism rates

varied among ﬁelds and in different months.

90

Introduction

The pine needle scale (Chionaspis pimfoliae (Fitch)) is native to North America
and has a relatively wide host range, including 28 species of conifers in the genera Pinus,
Picea, T suga, Abies, Cedrus, T axus, and T oreya (Zahradm'k 1990, Luck and Dahlsten
1974). A nearly identical insect, C. heterophyllae (Cooley) (‘pine scale’) shares at least
14 host plant species, primarily Pinus species (Shour 1986). Chionaspis heterophyllae is
native to the eastern United States, including Michigan (Shour and Schuder 1987). The
two species are nearly morphologically identical, and have extremely similar life
histories; several papers that have been published using the species name C. pimfoliae
were later determined to actually be C. heterophyllae (i.e. Nielsen and Johnson 1972 and
1973, Walstad et al. 1973; per Burden and Hart 1989). Our study populations were
originally designated C. pinifoliae, but have since been identiﬁed as C. heterophyllae
(courtesy of Douglass Miller, Systematic Entomology Laboratory, Agricultural Research
Service, USDA). We will refer to our study organism as ‘pine needle scale’, to be
consistent with the literature and in light of the extreme similarity of the two insects.

In managed settings such as nurseries, tree plantations, and omamentals, pine
needle scale can become a major problem (Tooker and Hanks 2000, Johnson and Lyon
1988, Sheffer and Williams 1987). Pine needle scale can be found in the forest, but it
rarely if ever becomes a noticeable pest there (Burden and Hart 1993, Cooper and
Cranshaw 1999, Ruggles 1931). It is one of the most common insect pests of Scotch pine

(Pinus sylvestris L.) grown for Christmas trees in Michigan (McCullough and Fondren

91

 

1998). In a 1998 survey of Michigan Christmas tree growers, the number of acres treated
and number of sprays for this insect alone exceeded those for any other insect pest in
Scotch pine Christmas trees (McCullough and Fondren 1998). On Christmas trees, light
to moderate populations of pine needle scale can reduce the value of trees due to the
white scale armor that remains on the needles even after the insects die. High
populations can cause needles or even branches to die (Kosztarab 1990, Walstad et a1.
1973, Cumming 1953).

Life cycle. The number of generations of pine needle scale can vary with
geographic location and other factors such as temperature, host plant, density, and local
climactic conditions (Shour 1986, Nielsen and Johnson 1973, Luck and Dahlsten 1974).
Univoltine populations are reported from relatively cold climates such as upstate New
York and Saskatchewan (Cumming 1953, Gambrell 1938). Bivoltine populations are
most common throughout most of the United States including Lower Michigan (Shour
1986). Most commonly, the scale insects overwinter as eggs that hatch in early spring,
around the time that lilacs bloom in May (Mussey and Potter 1997, Herms 1990). The
pinkish newly hatched crawlers move about for a few days before “settling” or inserting
their stylets into a needle on the host plant. After the crawlers settle, they undergo a ﬁrst
. molt aﬁer approximately ten days. This stadium, the second instar, is also called the
‘hyaline’ (transparent) stage. After about ﬁve to seven days, the female second instar
nymphs begin to secrete a thin transparent covering at their posterior end while the males
begin to secrete a white waxy covering (Cumming 1953, Nielsen and Johnson 1973).
Like most diaspidids, the females undergo a total of three instars while the males have

ﬁve instars, including a pupal stage (Beardsley and Gonzalez 1975). The males reach

92

adulthood at about the same time as females reach the third instar, and mating occurs
shortly after the males emerge. Following mating, the females usually begin to secrete
the familiar white waxy covering. In Indiana, both C. pinifoliae and C. heterophyllae
have been occasionally observed to have a partial third generation (Shour 1986). This
may occur in southern Michigan in especially warm years, but was not observed during
our study or in previous studies in Michigan (Eliason and McCullough 1997).

The crawler stage of pine needle scale has often been recommended as the ideal
stage to target with insecticides (Beard and McLeod 1992, Burden and Hart 1989,
McCullough et al. 1998, Gambrell 1938). This strategy is suited to the ﬁrst generation of
pine needle scale, which tend to hatch within a few days’ time. However, when control is
needed on the second generation, targeting the population when most are in the hyaline
stage will be most effective, as by then most of the eggs would have hatched from
underneath the protective covering (Nielsen and Johnson 1972, Martel 1972)

We focused our efforts on monitoring the phenology of the second generation
because it poses the most challenges for pest control. Phenology of the second
generation of pine needle scale is more difﬁcult to predict and monitor than the ﬁrst
generation. Second generation crawlers begin to hatch in midsummer, usually in July in
Lower Michigan, and continue to hatch for several weeks. After the insects secrete their
protective waxy covering, control with insecticides is less effective (McCullough et al.
1998, Beard and McLeod 1992, Martel 1972).

Natural enemies. Many natural enemies have been reported to prey on C.
pinifoliae and C. heterophyllae. Most coccinellids associated with pine needle scale and

other armored scales (Diaspididae) are specialists on this family (DeBoo and Weidhass

93

1976, Drea and Gordon 1990). The coccinellid Chilocorus stigma (Say) (=C.
bivulnerus), the twice-stabbed ladybeetle, was noted in the original description of C.
pinifoliae (Fitch 1856, from Shour 1986). Several other workers have also reported this
coccinellid to be an important or common predator on both C. pinifoliae and C.
heterophyllae (Cumming 1953, Shour 1986, Nielsen 1970, DeBoo and Weidhaas 1976).
Most known Chilocorus species are specialists on diaspidid scales (Greathead and Pope
1977, Drea and Gordon 1990). Other coccinellids found in association with C. pinifoliae
include: Coccidophilus (as Microweisia or Cryptoweisia) marginata (LeConte),
Microweisia (as Cryptoweisia) atronitens (Casey), and Microweisia misella (LeConte)
(Nielsen 1970, Luck and Dahlsten 1974, Gorham 1921).

The parasitoid complex associated with pine needle scale appears to vary in
different geographic regions (Martel and Sharma 1975, Burden and Hart 1990, Cooper
and Cranshaw 1999). Ten species of primary parasitoids and two species of
hyperparasitoids are reported to attack either C. pinifoliae or C. heterophyllae in North
America, and are usually found at densities of about 15-30% of the host population
(Burden and Hart 1993). Most of the parasitoids found to attack pine needle scale have
multiple hosts (Martel and Sharma 1975, Burden and Hart 1993).

One predatory mite, Hemisarcoptes maIus (Shimer) (Astigmata:
Hemisarcoptidae), was reported to feed on C. heterophyllae in Indiana (Nielsen 1970,
from Shour 1986). Gerson et al. (1990) note that H. malus is known to be phoretic on C.
stigma.

Control methods. Currently, pine needle scale is usually controlled in Christmas

tree plantations with broad-spectrum insecticides (McCullough and F ondren 1998,

94

Eliason and McCullough 1997, Beard and McLeod 1992). Many of these products are
undergoing intensive review under the federal Food Quality Protection Act of 1996, and
may eventually lose registration in minor-use crops such as Christmas trees (DiFonzo and
McCullough 1998). Developing other options for control of pine needle scale will give
growers an alternative to conventional insecticides, encourage conservation of natural
enemies, and reduce worker exposure to pesticides.

Christmas tree plantations are a relatively stable ecosystem compared with annual
crops (McCullough 1999) which may increase the potential for using biologically —based
control methods (Raupp et al. 1992). Plantations can provide overwintering sites for
insect natural enemies such as coccinellids, which usually overwinter as adults in leaf
litter. The structural heterogeneity of tree plantations encourages a diversity of insect
species, providing refuge for predators and alternate prey sources if necessary.

Our objectives in this study were to: 1) determine the phenology of the second
generation of pine needle scale in terms of calendar days and degree-day accumulations;
2) characterize the natural enemy complex acting on the 2nd generation of pine needle
scale in Christmas tree ﬁelds in Michigan; and 3) determine the rates of predation and

parasitism on pine needle scale infestations in commercial Christmas tree ﬁelds.

Materials and Methods

Field sites. F our ﬁeld sites were used throughout the course of this study from

1999 to 2001. All trees were Scotch pine (Pinus sylvestris L.) planted in a standard 1.3 x

95

1.3-m (6 by 6 ﬁ) spacing and sheared annually. Three sites were on commercial
Christmas tree plantations: Montcalm County in 1999 (43°14’N, 85°03’W), Van Buren
County in all three years (42°22’N, 85°52’W), and another ﬁeld in Montcalm County in
2001 (43°21 ’N, 85°14’W). The fourth site was on MSU’s Tree Research Center in
Ingham County, monitored only in 1999 and 2000 (42°40’N, 84°27’W).

Age and variety of trees varied among the ﬁelds (Table 1). All trees selected for
phenology and natural enemy sampling were selected in cooperation with the growers.
Trees were selected nonrandomly based on the presence of a noticeable scale insect
infestation. In 1999, no insecticides were applied in any ﬁeld during our study. In 2000
and 2001, some trees in the ﬁelds were used for a related study on insecticides, but we
monitored scale phenology only on unsprayed trees.

Objective 1. Phenology of the second generation. Sampling began before
initial hatch of second generation scale insects and continued weekly until second
generation scales began ovipositing in late summer. For the ﬁrst few weeks, at least 25
adult female scales (when possible) from each ﬁeld were examined to estimate the
progression of crawler hatch and number of eggs present. As the crawlers hatch, they
leave white chorions behind under the female scale armor. These chorions, however, are
very ﬂimsy and could not usually be counted accurately enough to determine an exact
percentage of egg hatch, so the mean number of eggs was used as an indicator of the
progression of egg hatch.

Once eggs began to hatch, samples of scale insects were taken in each ﬁeld by
removing approximately 20 current-year and previous year fascicles (2 needles per

fascicle) from infested areas in each cardinal direction on each tree. Samples were placed

96

individually in coin envelopes, kept in coolers for transport to the laboratory and held at
4°C to limit insect development. On each sample date, ﬁve to ten needles per tree were
examined under a microscope to determine the density (scales/cm of needle), sex, life
stage and cause of mortality of each scale insect present. In 2000 and 2001, three to ﬁve
trees were sampled each week (rather than 6 to 10) and 3-5 needles per tree (rather than
10) were examined under a microscope. At least 100 scales were examined from each
sample on each sample date, when possible, to monitor the phenology of the population.
Objective 2. Natural enemies. Visual counts of coccinellid adults and larvae on

each sample tree (Table 1) were conducted weekly in 1999. Number and stage (adult, I

 

larvae, or pupae) of each insect was recorded. In 2000 and 2001, we monitored sample
trees at biweekly intervals to record predator activity, scale development, mortality, and
to collect scale predators. Natural enemy activity was recorded on each ﬁeld visit and
while processing samples in the lab. Although the precise impact of each natural enemy
could not be determined, we recorded the presence and type of natural enemies collected
from each sample. Representative specimens of each taxon were identiﬁed using the
keys in Gordon (1985). Voucher specimens were deposited in the A.J. Cook Arthropod
Research Collection at Michigan State University, voucher number 2002-01.

To look for overwintering parasitoids, ten randomly selected branch tips were
clipped from infested trees in the Van Buren County ﬁeld on 11 March 2000. Fifty
randomly selected needles with a total of 641 scales were examined under a microscope
to search for overwintering parasitoids. When a parasitized female scale was found, it
was isolated in a microfuge tube and the parasitoid was allowed to emerge. In the

summer of 2000, we also reared out parasitoids found while inspecting our weekly needle

97

samples. Parasitoids were allowed to air dry and were mounted on slides. Parasitoids
were identiﬁed to genus by Dr. John Luhman, Minnesota Department of Agriculture, St.
Paul, MN, and to species by Dr. Michael Gates, Systematic Entomology Laboratory,
Agriculture Research Service, US. Department of Agriculture.

Objective 3. Rates of predation and parasitism. Predation or parasitism on
each scale insect was recorded when needle samples were examined weekly. Predation
and parasitism rates were recorded as the percentage of the total number of scales
examined. On most sample dates, and on every sample date in 1999, we noted the instar
of each predated scale body when possible, to obtain an estimate of the rate of predation
or parasitism on different instars. Obvious signs of predation on 3rd instar and adult
scales included jagged holes or tears in the scale armor. Stage-speciﬁc predation on the
earlier instars was difﬁcult to assess accurately after the scale body had dessicated.
Because our sampling was destructive, the rates of predation recorded each week
represent different sets of needles.

Rates of parasitism were determined by ﬂipping over the white scale armor of
adult females and looking for scale mummies. Scales parasitized by endoparasitoids
appeared rigid and orange-brown in color, and ﬁlled the whole armor (Nielsen and
Johnson 1973). Dead female scales were typically dark and shriveled up in the anterior
end of the armor. Only the adult females appeared to be attacked by parasitoids. Rates
of parasitism could only be determined for Van Buren County in 2000.

Statistical Analyses. Data for this chapter are presented as either tables or
ﬁgures that list percentages of scale insects in various life stages, numbers of scale eggs,

percentages of predated scales, and numbers of natural enemy species collected.

98

Results

Objective 1. Phenology of the second generation. Second generation eggs
hatched over a period of approximately three weeks, usually beginning the second or
third week of July and continuing until early August. As eggs began to hatch, the
average number of eggs we found beneath female armor declined, giving us an indicator

of the progression of hatch (Table 2). In 1999, eggs hatched ﬁrst in Van Buren County,
the southernmost county, beginning on or just prior to 5 July (1228 degree-days (DD50)).

By 13 July in Van Buren County, the number of eggs per female had declined by 59%
(Table 2). On that date, both ﬁrst and second instar scale nymphs were present in the

ﬁeld (Table 3). In Montcalm County in 1999, eggs hatched between 7 July and 12 July
(1227-1315 DDSO), as shown by the appearance of ﬁrst instar scale nymphs on the foliage
(Table 4). In 2000, egg hatch was ﬁrst observed in Van Buren County on 12 July (1280
DD50), and in Ingham County, the beginning of egg hatch took place between 11 July and
24 July (1232-1436 DDso). Egg hatch in Montcalm County was not observed in 2000,

but in 2001, we did monitor egg hatch in Montcalm County in conjunction with another

study (see Chapter 4). The second-generation egg hatch in Montcalm County began
between 12 July and 27 July (1124-1465 DD50), but since we could not sample between
those dates we were unable to determine the duration of egg hatch more precisely. In

Van Buren County on 19 July 2001 (1337 DD50), crawlers had already begun to hatch.

99

Based on these data, the initiation of second-generation egg hatch consistently occurred
at approximately 1230-1300 DD50, and generally progressed from Van Buren County

(the southernmost county) northward to Montcalm and Ingham Counties.

Development of the scale population. Our ﬁrst observations of crawlers on the
foliage were the most reliable evidence that egg hatch had started, and throughout the
sampling period we determined the length of egg hatch in part based on observations of
live crawlers. On 5 July 1999 in Van Buren County, we observed crawlers (ﬁrst instars)

in the ﬁeld (Table 2), although we did not take rigorous counts of the number of scales
per needle. By 13 July (1392 DD50), the second-generation scale p0pulation consisted of

31% ﬁrst instar and 69% second instar (hyaline stage) nymphs (Table 3). This indicated
that egg hatch had been continuing for at least 12 days, but our observations of adult

females with eggs showed that on average 11 eggs still remained beneath the armor, so
egg hatch was likely not yet complete (Table 2). On 19 July (1509 DD50), 76% of the

second generation scales had reached the second instar, while 13% had molted to the
third instar and 11% were ﬁrst instars (Table 3). Roughly six eggs remained per female
(Table 2), but they were probably not viable.

In 1999 in Montcalm County, 76% of the second generation had reached the
second instar on 22 July (1520 DD50), or three days later than in Van Buren County.
Abundance of crawlers decreased as the population molted to the hyaline stage, and by

30 July (1695 DD50), roughly 38% of the population had reached the third instar.

In 2000 and 2001, we did not monitor the development of the scale population as
intensively, and instead focused on the duration of egg hatch and beginning of the second

instar (hyaline stage) (Table 5). In Van Buren County in 2000, the crawlers began to

100

hatch on 12 July, and by 20 July, approximately 25% were second instars. In Ingham

County in 2000, we found crawlers just beginning to emerge on 11 July (1232 DD50), and
by our next sample date on 24 July (1436 DD50), 23% were crawlers and 42% were

already second instars (Table 2). On 31 July (1566 DD50) in Ingham County, 59% of the
second generation was in the hyaline stage. In 2001 in Van Buren County, we found
66% of the population had reached the second instar on 25 July (1588 DDSO).

Approximately 20% of the population had molted to at least the third instar at that time
(Table 5), although some eggs still remained beneath the armor. On 2 August, 49% of
the population was in the second instar (hyaline stage) while the rest were adults. Scales
overwintered as eggs until the following spring. In early 2000, a random sample of 641
scales taken from the Van Buren County ﬁeld on 11 March showed less than 1% of the
females overwintered as mated adults.

Objective 2. Natural enemies. In 1999, two species of coccinellids were
abundant throughout the season in both ﬁelds (Figure 2). The twice-stabbed lady beetle
(Chilocorus stigma (Say)) and a small black coccinellid (Microweisia misella LeConte)
were the two most common insect predators observed on infested needles. In Van Buren
County, C. stigma appeared to complete at least two generations. The number of adults
observed declined from 27 July to 11 August, when pupae were ﬁrst observed. The
abundance of larval C. stigma peaked on 3 August. We saw an increase in the number of
C. stigma adults in late August as the second generation matured. Adult M. misella were
observed throughout the sampling period, appearing to peak on 3 August (Figure 2).

The parasitic wasps we collected were identiﬁed as Encarsia bella Gahan

(Hymenoptera: Aphelinidae) and Marietta mexicana Howard (Hymenoptera:

101

Aphelinidae). Rates of parasitism observed in Van Buren County in 2000 ﬂuctuated
throughout the season (Table 6).

Objective 3. Rates of predation and parasitism. In the 1999 ﬁeld studies on
unsprayed trees, rates of predation reached better than 70% in each ﬁeld (Tables 7 and 8).
In Van Buren County, the predation rate averaged roughly 70% from 3 August to 27
August. Predation rates in the Van Buren County ﬁeld jumped ﬁom 23% to 73%
between 27 July and 3 August. This corresponded with a signiﬁcant increase in the
numbers of C. stigma and M. misella larvae and M. misella adults observed in the ﬁeld
(Figure 2).

In both ﬁelds, the scale covers of the adult males appeared nearly 100% predated
by the end of the season. This is not a result of the males emerging because they back out
from underneath their scale armor when mature and do not tear the armor (Cumming
1953). Their cover is also thinner than females, which may provide easier access for
predators, especially while the scales are in the pupal stage.

By deﬁnition, the number of adult female scales remaining at the end of the
season indicates the potential for increase of the population the following year, since they
are sessile and the eggs remain under the scale armor throughout the winter. In Van
Buren County in 1999, 51% of the adult females had already been killed by predators on
20 September (Table 7). On 9 September in Montcalm County, roughly 17% of 30
remaining females had been predated (Table 8). On 23 September in Montcalm County,

only ﬁve adult females could be found in a sample of 278 scales (Table 8).

102

 

Discussion

The extended hatching period of the second generation of pine needle scale is
interesting in light of its contrast to the ﬁrst generation. Although we did not intensively
sample the ﬁrst generation hatch, other workers have found the duration of the ﬁrst
generation hatch to be shorter than that of the second generation (i.e. Beard and McLeod
1992, Nielsen and Johnson 1973). An extended hatching period allows some eggs to
remain relatively protected beneath the female’s armor while other hatched crawlers
emerge to seek places to settle on the new needles. The second generation of pine needle
scale generally moves out to the new current-year needles to settle. The extended hatch
period may be an adaptive strategy to allow the maximum number of scale crawlers to
reach new growth; as the new needles extend, more crawlers emerge to take advantage of
the resource. Since Scotch pine typically carries only two year’s growth of needles, the
scale insects on the newest growth have the best chance of remaining on the tree through
the winter.

In terms of the survival of the population, the second generation is especially
important since they must successfully settle in a place where they will overwinter.
Protection from natural enemies may not be the reason for an extended hatching period,
since scales are vulnerable to natural enemies such as Chilocorus stigma in nearly every
life stage (eggs, crawlers, adults). In case of adverse weather conditions, which can have
a devastating effect on the new crawlers (Nielsen and Johnson 1973), spreading out the
hatching period in late summer may avoid strong summer storms or other acute

disturbances. Another possible advantage of spreading out the hatch period may be

103

increasing the genetic variability of the population. Males typically mature and mate at
about the same time as the females reach the third instar (Brown 1959, Nielsen and
Johnson 1973). With different segments of the population maturing at different rates,
mixing of the gene pool may be encouraged.

Overall, however, we did ﬁnd consistent associations between degree-days and
major events in the life cycle (at least :t 100 degree-days). Degree—days have been shown
to be a good predictor for the phenology of armored scales in general (Beardsley and
Gonzalez 1975, McClure 1990), although individuals may be affected by microclimatic
factors (Burden and Hart 1989).

In general, we found that natural enemies, especially coccinellids and, to a lesser
extent, endoparasitoids, could be found in unsprayed ﬁelds throughout the development
of the second generation. The ubiquitous presence of Chilocorus stigma and Microweisia
misella indicate a high potential for mortality caused by natural enemies in unsprayed
Michigan plantations. These coccinellids were found in every ﬁeld and year, but were
especially prevalent in 1999, when our ﬁeld plots were left unsprayed by chemical
insecticides. Chilocorus stigma seems well adapted to the pine needle scale life cycle,
with at least two generations per year coinciding with the pine needle scale hatching
periods. Larval M. misella are very small and were difﬁcult to observe, although a
pattern of increasing larval numbers around 3 August was observed in 1999 in Van Buren
County, and 6 August 1999 in Montcalm County. In contrast to many coccinellids that
overwinter as adults, including C. stigma, larvae of coccinellids in the tribe Microweisini
are known to overwinter beneath the scale armor of C. pinifoliae (Eliason and

McCullough 1997). These coccinellids may be amenable to conservation strategies. If

104

C. stigma and M. misella are observed in a Christmas tree plantation, avoiding the use of
a broad-spectrum insecticide is likely to improve control of the pine needle scale.

The rates of predation and parasitism observed in unsprayed ﬁelds in 1999 were
high, effectively leaving a relatively small nrunber of live female scales available to
produce a new generation in the next season. Adult male scales consistently displayed a
very high rate of predation, which may be due to their relatively thin armor. However, in
most armored scales, one male may inseminate several females (Beardsley and Gonzalez
1975), so a high predation rate on males may not signiﬁcantly affect the population.
Although we did not make a direct comparison, predation rates in ﬁelds sprayed for
control of pine needle scale would probably have been lower, given other factors equal.
Predation data in 2000 and 2001 were confounded by a simultaneous insecticide trial
occurring in the same ﬁelds; the trial was unlikely to affect the phenology of the scale but
it may have affected the natural enemy population.

Armored scale insects are often affected by a complex of natural enemies that
becomes ineffective if insecticides are repeatedly applied (Luck and Dahlsten 197 5,
Ripper 1956). The resurgence of phytophagous insects such as armored scales following
applications of broad-spectrum insecticides has been frequently documented (Roberts et
a1 1973, Luck and Dahlsten 1975, Ripper 1956, McClure 1977, Sheffer and Williams
1987). Reducing the rate of disturbance in a tree plantation and increasing the diversity
of the natural enemies present will discourage widespread outbreaks.

Implications for management strategies. Documenting the relationship between

second generation phenology and degree-days is an important step toward managing the

pine needle scale population in Christmas tree plantations. Approximately 1500 DD50,

105

when hatch is complete and the majority of the population is still in the second instar,
would be an appropriate time for applying controls such as an insecticide or horticultural
oil (Nielsen and Johnson 1972, Martel 1972). This information may help to show
growers that delaying a spray application until most eggs have hatched may be more
efﬁcient and effective than spraying as soon as crawlers are visible.

Our results will help managers to develop control strategies for pine
needle scale that will not exacerbate the problem. An integrated approach to
management that relies on scouting appropriately and preventing unnecessary insecticide
applications is likely to be an effective long-term strategy. Further research on the
natural enemies of this pest is needed to elucidate the impact of each species on the pine
needle scale population and the potential for using natural enemies to prevent scale

outbreaks.

106

‘33: s ofﬂi“:ﬁg 5.41.19“
0:

WW’Tp—ti'i

References Cited

Beard, K.K. and D.M. McLeod. 1992. Pine needle scale control on Michigan Christmas
trees. Down to Earth 47(1):16-19.

Beardsley, Jr., J.W. and RH. Gonzalez. 1975. The biology and ecology of armored
scales. Ann. Rev. Entomol. 20: 47-73.

Brown, CE. 1959. Reproduction of the pine needle scale, Phenacaspis pinifoliae
(Fitch), (Homoptera: Diaspididae). Can. Entomol. 91(9): 529-535.

Burden, DJ. and ER. Hart. 1989. Degree-day model for egg eclosion of the pine
needle scale (Hemiptera: Diaspididae). Environ. Entomol. 18(2): 223-227.

Burden, DJ. and E. R. Hart. 1990. Parasitoids of Chionaspis pinifoliae (Homoptera:
Diaspididae) in Iowa. Great Lakes Entomologist 23 (2): 93-97.

Burden, DJ. and E.R. Hart. 1993. Parasitoids associated with Chionaspis pinifoliae
and Chionaspis heterophyllae (Homoptera: Diaspididae) in North America. J. Kansas
Ent. Soc. 66(4):383-391.

Cooper, D.D. and W.S. Cranshaw. 1999. The natural enemy complex associated with
the pine needle scale, Chionaspis pinifoliae (Fitch) (Homoptera; Diaspididae), in North
Central Colorado. J. Kansas Ent. Soc. 72(1):131-133.

Cumming, M.E.P. 1953. Notes on the life history and seasonal development of the pine
needle scale, Phenacaspis pinifoliae (Fitch). Can. Entomol. 89(9): 347-352

DeBoo, RF. and J .A. Weidhaas, Jr. 1976. Studies on the predation of pine needle
scale, Phenacaspis pinifoliae Fitch, by the coccinellid, Chilocorus stigma Say. Plantation
research: XIV. Report CC-X-l 19. Chemical Control Research Institute, Canadian
Forestry Service, Department of the Environment, Ottawa, Ontario.

DiFonzo, C.D. and D.G. McCullough. 1998. Potential impacts of the Food Quality
Protection Act on the Michigan Christmas tree industry. Michigan Christmas Tree
Journal 45( 1): 26-29.

Drea, JJ. and RD. Gordon. 1990. Coccinellidae. pp 19-40 In: D. Rosen [ed.],
Armored scale insects: their biology, natural enemies and control, Vol. B. Elsevier
Science Publishers, Amsterdam.

Eliason, E.A. and D.G. McCullough. 1997. Survival and fecundity of three insects
reared on four varieties of Scotch pine Christmas trees. J. Econ. Entomol. 90(6): 1598-
1608.

107

Fatzinger, C.W., H.O Yates III, and LR. Barber. 1992. Evaluation of aerial
applications of acephate and other insecticides for control of cone and seed insects in
southern pine seed orchards. J. Entomol. Sci, 27(2): 172-184.

Gambrell, EL. 1938. Experiments for control of the pine needle scale, Chionaspis
pinifoliae (Fitch). J. Econ. Entomol. 31(2): 183-186.

Gerson, U. and R. Schneider. 1981. Laboratory and ﬁeld studies on the mite
Hemisarcoptes coccophagus Meyer (Astigmata: Hemisarcoptidae), a natural enemy of
armored scale insects. Acaralogia, 22(2): 199-208.

Gerson, U., B.M. O’Connor and M.A. Houck. 1990. Acari. pp. 77-97 In: D. Rosen
[ed.], Armored scale insects: their biology, natural enemies and control, Vol. B. Elsevier
Science Publishers, Amsterdam.

Gordon, RD. 1985. The Coccinellidae (Coleoptera) of America north of Mexico. J.
New York Ent. Soc. 93(1): 1-912.

Gorham, RP. 1921. Notes on the pine needle scale and one of its enemies. Proc.
Acadian Entomol. Soc. (1921) 7:37-41.

Greathead, DJ. and RD. Pope. 1977. Studies on the biology and taxonomy of some
Chilocorus spp. (Coleoptera: Coccinellidae) preying on Aulacaspis spp. (Hemiptera:
Diaspididae) in East Africa, with the description of a new species. Bull. Ent. Res. 67:
259-270.

Herms, D.A. 1990. Biological clocks. Am. Nurseryman 172(8):56-63.

Johnson, W.T., and H.H. Lyon. 1988. Insects that feed on trees and shrubs, 21nd ed.
Comstock University Press, Ithaca, NY.

Kosztarab, M. 1990. Economic Importance. pp. 307-310. In D. Rosen [ed.], Armored
Scale Insects: Their Biology, Natural Enemies, and Control, Vol. B. Elsevier Science
Publishers, Amsterdam, The Netherlands.

Luck, RF. and D. L. Dahlsten. 1974. Bionomics of the pine needle scale, Chionaspis
pinifoliae, and its natural enemies at South Lake Tahoe, Calif. Ann. Ent. Soc. Am. 67(3):
309-3 16.

Luck, RF. and D.L. Dahlsten. 1975. Natural decline of a pine needle scale (Chionaspis
pinifoliae (F itch)) outbreak at South Lake Tahoe, California following cessation of adult
mosquito control with malathion. Ecology 56: 893-904.

Martel, P. 1972. Le malathion et le diméthoate dans la lutte contre la cochenille du pin,

Phenacaspis pinifoliae (Fitch) (Homoptera: Diaspididae). Ann. Soc. Ent. Qué. 17: 20-23.

108

Wmm‘J-Etm“ “.1 mum-51?“

Martel, P. and M.L. Sharma. 1975. Parasites de la cochenille du pin, Phenacaspis
pinifoliae (Fitch), dans la région de Sherbrooke, Quebec. Ann. Soc. Ent. Qué. 20: 11-14.

McClure, M.S. 1977. Resurgence of the scale, F iorinia externa (Homoptera:
Diaspididae), on hemlock following insecticide application. Env. Entomol. 6(3): 480-
484.

McClure, M.S. 1990. Seasonal history. pp. 315-318 In D. Rosen [ed.], Armored Scale
Insects: Their Biology, Natural Enemies, and Control, Vol. A. Elsevier Science
Publishers, Amsterdam, The Netherlands.

McCullough, D.G. 1999. Biological control in Christmas tree plantations. In: D.G.
McCullough, S.A. Katovich, D.L. Mahr, D.D. Neumann, C.S. Sadof, and M.J. Raupp
[eds.], Biological Control of insect pests in forested ecosystems: a manual for foresters,
Christmas tree growers and landscapers. Michigan State University Extension Bulletin
E-2679. East Lansing, MI. pp. 75-91.

McCullough, D.G. and Fondren, K. 1998. What’s Bugging You—FQPA and
Insecticide Use in Michigan Christmas Tree Fields: Preliminary Results from 1998
Survey. Michigan Christmas Tree Journal 45(3):22-29.

McCullough, D.G., S.A. Katovich, M.E. Ostry, and J. Cummings-Carlson [eds.].
1998. Christmas tree pest manual, 2"d ed. Michigan State University Extension Bulletin
E-2676. East Lansing, MI. 143 pp.

Mussey, GJ. and D.A. Potter. 1997. Phenological correlations between ﬂowering
plants and activity of urban landscape pests in Kentucky. J. Econ. Entomol. 90(6):1615-
1627.

Nielsen, D.G. 1970. Host impact, population dynamics, and chemical control of the pine
needle scale, Phenacaspis pinifoliae (Fitch), in central New York. Ph.D. thesis,
Department of Entomology and Limnology, Cornell University, Ithaca, New York.

Nielsen, D.G. and NE. Johnson. 1972. Control of the pine needle scale in central New
York. J. Econ. Entomol. 65(4): 1161-1164.

Nielsen, D.G. and NE. Johnson. 1973. Contribution to the life history and dynamics of
the pine needle scale, Phenacaspis pinifoliae, in central New York. Ann. Ent. Soc. Am.
66: 34-43.

Raupp, M.J., C.S. Koehler and J.A. Davidson. 1992. Advances in implementing

integrated pest management for woody landscape plants. Ann. Rev. Entomol. 37:561-
585.

109

‘15“?!

 

Ripper, W.E. 1956. Effect of pesticides on the balance of arthropod populations. Ann.
Rev. Entomol. 1: 403-438.

Roberts, F.C., R.F. Luck, and D.L. Dahlsten. 1973. Natural decline of a pine needle
scale population at South Lake Tahoe. California Agriculture October 1973: 10-12.

Ruggles, A.G. 1931. Preliminary notes on the biology and control of the pine leaf scale,
Chionaspis pinifoliae Fitch. J. Econ. Entomol. 24:1 15-1 19.

Sheffer, BJ. and M. L. Williams. 1987. Factors inﬂuencing scale insect populations in
southern pine monocultures. Fla. Entomol. 70(1): 65-70.

Shour, M.H. 1986. Life history studies of the pine scale, Chionaspis heterophyllae
Cooley, and the pine needle scale, Chionaspis pinifoliae (Fitch). PhD. dissertation,
Department of Entomology, Purdue University, West Lafayette, IN.

Shour, M.H. and D.L. Schuder. 1987. Host range and geographic distribution of
Chionaspis heterophyllae Cooley and C. pinifoliae (Fitch) (Homoptera: Diaspididae).
Indiana Academy of Science 96: 297-304.

Tooker, J.F. and L.M. Hanks. 2000. Inﬂuence of Plant Community Structure on
Natural Enemies of Pine Needle Scale (Homoptera: Diaspididae) in Urban Landscapes.
Environ. Entomol. 29(6): 1305- l 31 1.

Walstad, J.D. , D.G. Nielsen and NE. Johnson. 1973. Effect of the pine needle scale
on photosynthesis of Scots pine. Forest Science 19(2): 109-1 1 1.

Zahradnik, J. 1990. Conifers, pp. 633-644. In D. Rosen [ed.], Armored Scale Insects:
Their Biology, Natural Enemies, and Control, Vol. B. Elsevier Science Publishers,
Amsterdam, The Netherlands.

110

' 1ZW.‘......' ’Imlv.mwa._ge.nm

So» £5 5 wow: 8: 803 83 82:1

 

 

 

2523: 568
.28 a as» 2
“U 98 m 3:80
580
298: 8 cocci .Eom E ban: 8 38» m
3 380—622 88% c ”< @380 on 880on .33 8mm a a a Sam
menu» 88.»
n 6:5 N" 6:5
298: 3 Suﬁsm 859m
a a a s. 63682 83 2 am: 83 SON
was?»
5505?: S .025
but? ”Bo 88% m hotoasm
e 56m we“; a. .653. S 6,6: .83 BE N a»: 83 82
wow: _o>o_ wow: _o>o_ 5093 wow: :32 bus?
305 bots, use ow< . . . .
.oZ :ousmomﬁ 80b 02 5.588?“ 98 ow< 805 oz aocﬁmom£ was ow<
5550 53282 >6on :oSm §> 5880 Emnwﬁ

 

use... 288 86 5 SN was SON .82 6 83 was Sam .2 285

111

.8ng

 

Em an “88 £82 088 zum H 3.0 ow oowﬁ “mam—.2 m
waESm: =5 wwwo 088 . .

cogs BE? mo 5.320301 cw o H _ 3 vw emf ﬁsw=< ﬂ
omsm 2%: .33 6.0 a 52 K a: 32 E
Maegan“. ”mam 25%: S ._ a N2 3 mm: 33 om
maﬁa £256 86 a «.2 2: 82 33 t
:22 .255 Mo 8:33 a; a «.8 mm 82 .3: S
a: a 9mm 2: we: 33 o

w; a m3 2: m3 as: 8 88 .8 55m §>
:5 =5 ”.2380 6.5%: 3? NE a E S 82 33 Nu
2%? :8 use $2380 3: a 0.: a $2 33 2

E. 8:22 B; 252 MS a 02 on RS 33 n 82 .8 83282
232 8.2: x: 23 a «m t 82 same... m
:2; bag 8:2: was“? :2 go a an E 82 .23 2
own; 8:3 .xao 3 a 2: om 82 33 2

22.. 5 £82.. 38% e235 0. 3 3: mm 3.: b3 m 82 .8 55m S~>

mac—automac 2.25.. 605888 8A:— oaan 23> 22h

. .3.— mmwo 6 Z 3.55.— 62 «>52:an .

 

v3.8.— uE-Ea: was 358%.»: «£31 38.3 938 0:25.. .82.: .58.:— muuo no non—~52 .N «EFF

112

 

3:2 33:3 .82 £.~ a S a B: 3&2 2
3% 2333. 38.3 m; w 5: S cm: 3&3. m
:5 £258 32 a 0.: mm 8: 33 R
$31.2 2 3: 33 S
gaugmzmm 336.8%
BEE $2320 3233 “WW—”MM H no.3 m3 oov .32 NN
Via 32. £3 62%... mmﬁmw a 8 2 own 3.2 S 88 .8 838252
39:96an “2.3K
mod u." m; E 2: 3%? w
8:5 223 Mo 3233830 i woo A. 8.2 a. 82 33 F.
own; 23333 38.3 mi H B: 3 32 33 R
via 83% Ea MMHWMM :3 a 9.2 3 cm: 33 3
mom 3 EN 2 RS 33 :
3% a mi om E: 33 m
o W? 2.3 R 88 358 .5233
macaaiomac 2.3.3 62:88:. SA:— 8:: 23> 29E
. hon away 6 Z 835$ 62 252.550

 

.5253 N 233.

113

 

 

:23 -
8 SEE Em 8380.3 .838 3:33 ammo 2 n m Gem “ma—mam 2
39a 22: . . 3%:
98 8380.3 332: Em .85?wa 3 o a w w mm Swﬁ H < o
3% 25%; 38.. who a 3:3 3 mm: 33 mm
. . 3 3.58
So E2380 mo o a a S we RS :3 2 Sow :oSm §>
2580.. 35888 3::
3:3an.50 .3.— muwo 6 Z 8:25.. 62 952.550 8.5 30> 23m

 

.5253 N 033.

114

h
1.
.E (45. Irrig- f sﬂ‘...

..... .... Aalr‘iﬂels. 1!

O—NG a

an: a

828m @8305 02:03 E8 2 :33 mo 88qu 2a 835 0.3 .«o mowﬂaoouom

 

 

#03 83. 3.3 ah.w_ 24 $3 -\+ aé 83 $2 .Eom om
mmd 3.3 3.2 mm._m Sam nmnd .\+ on” mi ovom .Eom n
vawm 36 ac. 2 26m Sum Sac 4+ 006 mm: :3 #392 mm
~08 26 3.2 3.2 $6 md -\+ :3 38 comm ~m=w=< E
3.? mod 8:2 Gem cﬁm $3 -\+ Sum .83 88 Hmzms< Z
Sém o 3.3 oo.m~ cod md -\+ $6 3% 82 Hm=w=< m
mmdm o 8.2 3.3” 8.x £3 -\+ 36 omom v3; 33. um
$4 0 co.- o9? m6— ."ch -\+ mm.m won 32 b3. 9
o o Ed mode Shem $6 -\+ $6 :2 Na? 33. 2
meow 3.3
$38 :3an meadow “33% 53. no\ 25939 who—3809 SEHE 0:50: Ame—RE Z 996 oohwov 8.5
o a . 2 a . o o Ba Serving m>< dean?

 

633:9? 9.2.? mm m 4+ £325 33 .bESU uohsm =a> .3332; 2:8 0:80: 2:.— .m «Bah

115

.5.

338 32305 3.32: 28 35888 Benz: :32 05 .«o :6 08 mowsm 0.5 «o mowsseom

 

 

 

2 Zn N2 Sm ed 8T5 who 4+ 3 ME <42 .23 a
3 2+ 2; 3m 3 3 4+ ed 88 £3. .23 a
a? $3 8.: 2.: 2.: m3 4+ 3 98 3mm .25 m
a; 8.2 3: ~62 8.: 3 4+ 3 82 a: 3593 mm
o a? 2: 2 m on w.o4+ 3 SS 28 an»? 2
ad can 92 n: 0.2 one 4+ 3 82 ct: 53:2 0
o ma 2m 0: 3; 3o 4+ 3+ 82 32 23 2
go o 3 w? 92 81586 4+ 3 EN 82 23 S.
o m; to an Ni smug 3.0 4+ 3 o: 32 b3 2
o o we a: 3m 2 4+ 2 3% 22 33 S
o o o c o ‘ c o RS 4:3 h
me «So no a £8qu

 

333%.? 22.3 Hm ﬂ 4+ 2525 93— .43580 E13252 Sue—26.:— o_a3. 2.52. 2:.— .v 035.

116

. t p . v O ‘
F .54 I... n.i£a;5egt... J.,“ 4 4‘23. ..‘1. ,LI. .

N 033+ 08 ”mo—mom cowauocow “am box: 608 0.8 085a

£83853 3 3:2 83% 09:2: 93 :38 .«o 38qu 0.8 momma 0.5 we mowScoeom

 

ad». 92 3: S: 2 3+ 8: 88 3&3 a 55m 5>
3:. Z we a? o a: E: 88 2&3 o 83m 5>
3». 2 mm 3:. o :5 8: 88 as»? N 55m 5>
3 S: 3 as v.2 as $2 88. b3 2 :25 89
as 2:. M: 1% 5 am 82 88 b3 2 82?.
3 3m 2 a? SN 3 £3 88 33 E 823:
0.2 n: no 3m 3 5 22 88 b3 : 8935
«SN 3.8 .2 ma SN 8v Na: 88 b3 8 55m 5>
.98 .2: .3 L: S cm: 82 88 b3 2 58m 5>
Goo:
838 :35 ..\o mug—«Sum :33 o\.. :35 En Ax. 3:93 ..\o "co—380 o\o Ame—38$ Z moor. 83 98c Ban— .8332 23m

nonwow .xoaq<

SON Ea 88 328% 28m 288 2E .m as;

117

.1 - :3 . ﬂ’u k» .u-brn

1:-

 

3988 8: mm? Boom": «mo: Mo .8385: on?

 

 

3% N ﬂ _ acacia: “5653: oo-w=<- S

N N _ mN Nam 8-3-3
Emacs 8oz 9- mNm 8-3-8

3% v NN 8 8 8-3-:

N. a g : _N 3N 8-3-2
2: o Nm 8 8 8-3-8
3%on 28 Z n N. oo-c:—.-o~

m? N on NB 2N Ave-$2-:
818m cousin-8a 880:8

EESNNM :8ch 83.233 .02 3.282% .02 3&5an .oZ ESSA 8mm

 

doom E 88288 03% Bow @2382 $6383 .e «Zak.

118

 

32. .8 BEBE 3.83% 85 3:0 C32: 9883 owﬁm 25m»: 05 mo owﬁcoouomn
BEBE 3.80%? 35 53230; 38 05 .«o owﬁaoobmn

 

 

GNTE N? Ens Sm §Nu5 QNN 8:1; 2: :No 8: is 62 E80.%m oN
ASHE 25 8W5 08 EYE MEN ANNHE 3% 3.8 NNN ovoN speeaom N
GNNTE 0.; Gus o ANVNna 2m aﬁua SN 92 NEN RNN Ems-N NN
EEHE v. a 2N5 mN GNNHE So 26.1.5 oNv 3N5 WEN SNN Ems-<- M:
Awning N3 ANHE o ANNE; v.8 sans EN 8.: momN SON mam? :
GNNTE ZN S2 5W5 3o ANwmna Ev 8.2 2% 82 “2&3- N
:2 T5 SN <2 Awe-us NN 83 W5 NeN WNNN $8 3: bi NN
5H5 ta $2 EVE 3N ANNE; 2m $3 a: 82 23 2
<2 ans o $2 23132: S2 S2 82 33 2
33on 333a BEBE $5333 «8885 Ame—33 axemmw 0:5
8:2: 2:? .x. muBEo-H :23 .x. .52: Em ..\o 2:33 .x. Eco-Em Z 08on

 

.83 295% some so owsm um: :08 E mm? 35 coax—anon 2: no owﬁcoﬂom
05 N8 m 2an com .33 bSSU 55m §> .owﬁm um: :98 E 038 0:50: 2:9 Mo 53305 Evens .h 03.;

119

30w 8 00:30:. 003250 09: Dee Gwﬁm 051025.505 98000 0E .«o 0wﬂs080mp

0030000 00.80%? 8:“ 832389 38 05 mo 0mﬁc080ma

 

 

 

Gua odm ASHE 0.? £100.? 32.1.5 30 0.8 ms .25 0oz 383$ mm

amna 5.2 $310 35 sins 3m Gonna 3m 8 $2 VSN Esoaom a

a: o 8:10 25 EYE Gm sale new 58 m? an” 382% m

mans ZN somua ms? 5 W5 0: Emua :0 an 82 as 030$ 2

é o gang 3w @2qu N3 Gels 2a In SS 28 7503 2

ans 9o axis 3% Ezua Q? Emma a? 0.? $2 on: 5:03 0

o 82% a8 $2.15 3m Senna 5.3 an 82 82 33 om

Qua o <2 8T5 iv @8ch n? 02 SW 82 :3 mm

:us o Qua 30 Gina W8 SN 0: E2 33 2

Anna 02 amucv 09 M2: SN 22 33 S

o o 82 bi n

0380.“ 032 @2305 983005 «BEBE @2003 $80.53 025
@2305 :53 .x. 0805 Emﬁ. 05:93 °\° E083 Z 980 00%0Q

 

.008 2983 £000 no 0930 :80 E E 005 552393 05 no
0330800 05 8m v 0309 00m .33 .3550 8:02:02 dwﬂm 0E :80 as 2000 0:000: 0:150 20:30.5 E080m .w 03:.

120

IFﬁE .. F Sch... LE -. Ii. .

.98 29:8
:08 :o 3.9.. 3.500 89.9.. new :93 cm> m5 5 ooom 5 SEE
28m m_mE£ 05 £8ch Emmma mama .6 $92.: ma 4.... cams. .F 9:9“.

 

 

u3< .2. .2.. .2. .2. .2. .2. .2. .2. .2. .2. :2.
.mo .5 SN -VN -VN -ON -5 -N_. -2. .00 -mo -oN

_ p h >

— p — — ~ _ . — p o

l-l-i
'0'
I0
c;
N
9669 0N

0 0 ,3

.0m

 

bcaoo Emcmc. III

 

3:300 :95 cm> o - mm

 

 

 

 

 

121

 

Van Buren County

 

 

 

 

 

 

 

 

 

lSlAdult C. stigma
ElAdult M. misella

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

160
F" El Larva C. stigma
.0 ; ILarva M. misella
"£120 T # "—‘§* 2; _, A '2 Elpupa C. stigma
g I
.. 5
3 80 «~— *-——4'/~* A 2e _. 2* - —#4
g 5
§ ’ '
z 40 ‘ ”Er —# — 7
O " I 1 T ‘ I 4—1 E I a
27-Jul O3-Aug 11-Aug 18-Aug 27-Aug O7-Sep 20-Sep
Montcalm County ElAdult C. stigma
EIAdult M. misella
100 ElLan/a C. stigma ”‘
80 .- ILarva M. misella
g 1 2 i 777 7 _ Elpupa C. stigma
$60M 42- u— v
.D
2 l.
g 40 a — 2
E
3
Z 20 T24 — _. —
0 , '~ . D ‘ [m

 

 

 

 

 

 

 

 

 

22-Jul 30-Jul 06-Aug 16-Aug 03-Sep OQ-Sep 23-Sep

 

 

Figure 2. Number and stage of Chilocorus stigma and Microweisia
misella observed in our 1999 ﬁeld study. The total number of each

observed on ten trees in Van Buren County and six trees in Montcalm

County is recorded.

122

911

 

Chapter 4

Potential efficacy of horticultural oil for control of pine needle scale (Chionaspis

heterophyllae (Fitch)) (Homoptera: Diaspididae) in Christmas tree fields

Abstract. Pine needle scale ((Chionaspis pinifoliae (Fitch) and C. heterophyllae
(Cooley)) is an important pest of Scotch pine (Pinus sylvestris L.) grown for Christmas
trees in Michigan. The most common method of control, application of broad-spectrum
insecticides, is most effective before the scale secretes its protective covering. Timing of
spray applications is critical especially on the second generation because of its
asynchronous hatching period. We also investigated the potential for using horticultural
oil as an alternative control method for pine needle scale in Christmas tree ﬁelds. Our
objectives were to 1) monitor the phenology of second generation pine needle scale to
identify optimal timing for insecticide applications in Christmas tree ﬁelds; 2) determine
the mortality rate of pine needle scale when horticultural oil is applied with a backpack
mist blower; and 3) assess the effectiveness of applying horticultural oil with a tractor-
mounted airblast sprayer on a large commercial Christmas tree plantation. We monitored
the phenology of the second generation of pine needle scale weekly on a minimum of six

trees on each of three plantations in Lower Michigan. We found that egg hatch began at
approximately 1230 degree-days base 10°C (50°F) (DDso) and the peak of the second

instar coincided with 1500-1600 DD50. In 2000 and 2001, we compared the efﬁcacy of a

highly reﬁned horticultural spray oil and conventional insecticides applied with a

123

ﬁ-

 

backpack mist blower at 1500-1600 DD50. In both years, the spray oil did as well or

better than the chemical insecticide, and scale mortality on trees treated with oil ranged
from 66% to 80%. In 2001, horticultural oil was applied with a tractor-mounted airblast
sprayer to a portion of a ﬁeld. Scale mortality ranged from. 36% to 56% and was not

signiﬁcantly different from conventional insecticides.

Introduction

Pine needle scale (Chionaspis pim'foliae (Fitch) and pine scale (C. heterophyllae
Cooley) are important insect pests of Scotch pine (Pinus sylvestris L.) and several other
conifer species grown in nurseries, landscapes, and tree plantations (Beard and McLeod
1992, Burden and Hart 1990, Johnson and Lyon 1988). Chionaspis pinifoliae has a wide
geographic distribution, including most of the United States, Canada, Mexico, the
Caribbean, and England (Nakahara 1982), while C. heterophyllae is native to the eastern
and midwestern United States (Shour and Schuder 1987). The two species are nearly
morphologically identical, and have extremely similar life histories; several papers that
have been published using the species name C. pinifoliae were later determined to
actually be C. heterophyllae (i.e. Nielsen and Johnson 1972 and 1973, Walstad et a1.
1973). The literature using the species name pinifoliae is more extensive than that of
heterophyllae, but because the ecology of the two species is so similar, we will refer to

the species collectively as pine needle scale.

124

 

While pine needle scale can be found in forested areas, it rarely if ever becomes a
noticeable pest there (Cooper and Cranshaw 1999, Burden and Hart 1993, Luck and
Dahlsten 1975, Ruggles 1931). In contrast, in highly managed settings such as Christmas
tree plantations or landscapes, populations can reach high densities (Tooker and Hanks
2000, Johnson and Lyon 1988, Sheffer and Williams 1987). Even moderate populations
of pine needle scale can cause Christmas trees to be unsaleable and high populations can
cause needles or branches to die or reduce grth and tree vigor (Nielsen and Johnson
1973, Dahlsten et a1. 1969, Walstad et al. 1973, Cumming 1953).

Life cycle. The number of generations of pine needle scale can vary with
geographic location and other factors such as temperature, host plant, density, and local
climactic conditions (Shour 1986, Nielsen and Johnson 1973, Luck and Dahlsten 1974).
In Michigan, C. pinifoliae has two generations per year (Eliason and McCullough 1997).
In Indiana, both C. pinifoliae and C. heterophyllae have been occasionally observed to
have a partial third generation (Shour 1986), but this was not observed during our study
or in past studies in Michigan (Eliason 1996).

Pine needle scale overwinters as eggs that hatch in early spring, typically just after
lilac (Syringa vulgaris L.) is in full bloom in May (Mussey and Potter 1997, Herms
1990). The ﬁrst generation hatching period extends for about seven days (Beard and
McLeod 1992, Nielsen and Johnson 1973). Newly hatched crawlers move about for a
few days before settling and inserting their stylets into the host plant. About a week to
two weeks after settling, the crawlers molt, pushing the exuviae up on top of their bodies
(Nielsen and Johnson 1973, Cumming 1953). This stage (the second instar) is termed the

hyaline (transparent) stage.

125

 

After about ﬁve to seven days, the female second instar nymphs begin to secrete a
thin transparent covering at their posterior end while the males begin to secrete a white
waxy covering (Cumming 1953, Nielsen and Johnson 1973). Females undergo three
instars in total and remain immobile once they begin feeding. Males complete three
instars, then pupate and emerge as winged adults (Cumming 1953, Shour 1986). Mating
begins ahnost immediately after the males emerge, usually just after the females reach the
third instar (Brown 1959). Aﬁer mating, the females secrete the familiar white waxy
covering and produce eggs that remain under the white scale armor until hatching occurs.
The adult females die after oviposition is complete (Brown 1959). The white wax
covering can protect the insects from many insecticides (Nielsen and Johnson 1972,
Martel 1972), but not always (Beard and McLeod 1992).

The second generation of pine needle scale usually begins to hatch in early July in
Michigan. The hatching period extends for up to three weeks, making it difﬁcult to time
insecticide applications or other control measures (Eliason and McCullough 1997). By
the time the last eggs hatch, the earliest hatching nymphs may have already molted to the
third instar (Beard and McLeod 1992, Nielsen and Johnson 1972, 1973).

Effective control of pine needle scale requires application of insecticide when the
insect is most susceptible, and for the ﬁrst generation, this usually means targeting the
crawler stage (Beard and McLeod 1992, Burden and Hart 1989, Nielsen and Johnson
1972, McCullough et al. 1998, Gambrell 1938). However, when controlling the second
generation, the second instar (hyaline stage) is a good target because it is also susceptible
to insecticide sprays, and its peak occurrence on the needles usually indicates that egg

hatch is complete (Nielsen and Johnson 1973, Shour 1986). Insecticide sprays applied

126

 

after the scale insect produces its protective covering are usually ineffective (Martel
1972, Nielsen and Johnson 1972).

Control methods. Pine needle scale is commonly controlled in large Christmas
tree plantations with applications of broad-spectrum insecticides (Beard and McLeod
1992, McCullough and Fondren 1998). Results of a survey of Michigan Christmas tree
growers showed that the nmnber of acres treated and the number of insecticide sprays for
pine needle scale alone exceed those for any other insect pest of Scotch pine
(McCullough and F ondren 1998). Scotch pine may be sprayed as many as four times per
year in Michigan plantations for control of pine needle scale (Beard and McLeod 1992,
McCullough and Fondren 1998). Broad-spectrum insecticides such as chlorpyrifos and
malathion have been recommended for use against pine needle scale for some time
(Beard and McLeod 1992, Martel 1972), and are among the most common insecticides
used on pine needle scale in Michigan (McCullough and F ondren 1998). Applications of
broad-spectrum insecticides may decimate natural enemy populations and often result in
outbreaks of armored scales (Roberts et a1. 1973, Luck and Dahlsten 1975, Ripper 1956,
McClure 1977, Sheffer and Williams 1987). Often, this situation leads to further
applications of insecticides, e.g. the ‘pesticide treadmill’ (Rose 1990).

Horticultural spray oils are often used to control insect pests in landscape or
ornamental situations (Raupp et al. 1992, Nielsen 1990), but are rarely used on Christmas
tree plantations (McCullough and F ondren 1998, Nielsen 1990). Horticultural spray oils
have been recommended for use against armored scale insects since the early 1920’s
(Riehl 1990, Pearce et al. 1941, Gambrell and Hartzell 1939, Doane 1926). Modern

spray oils are not phytotoxic under most conditions (Riehl 1990, Nielsen 1990), and are

127

 

relatively harmless to humans and natural enemies such as coccinellids (Smith and
Krischik 2000). After more than ﬁfty years of use, there have been no known cases of
insecticide resistance to horticultural oils among armored scales (Riehl 1990). The mode
of action of spray oils is simple—the oil clogs the spiracles of the insect, effectively
suffocating it. Because of this mode of action, adequate coverage is essential. Some of
the concerns raised by growers have included the efﬁcacy of oils and the difﬁculty of
obtaining adequate coverage efﬁciently in a large-scale setting.

Objectives. Our objectives in this study were to 1) monitor the phenology of
second generation pine needle scale to identify optimal timing for insecticide applications
in Christmas tree ﬁelds; 2) determine the mortality rate of pine needle scale when
horticultural oil is applied with a backpack mist blower; and 3) assess the effectiveness of
applying horticultural oil with a tractor-mounted airblast sprayer on a large commercial

Christmas tree plantation.

Materials and Methods

Study sites. This research was conducted in 2000 and 2001 in three Scotch pine
ﬁelds on commercial Christmas tree plantations in Lower Michigan (Table 1). Two
study sites were located on a farm in Van Buren County (42°22’N, 85°52’W). Another
site, used in 2000 only, was on MSU’s Tree Research Center in Ingham County
(42°40’N, 84°27’W). A fourth site was located in Montcalm County (43°21’N, 85 °27’

W) and was used only in 2001.

128

 

All trees used in the study were selected nonrandomly on the basis of a visible
pine needle scale infestation on at least part of the tree. The number of trees used each
year ranged from 13 to 50 trees per ﬁeld, depending on pine needle scale abundance and
the number of trees that could be provided by growers (Table 1).

The Montcalm County ﬁeld was used only for a test with a grower’s commercial

airblast sprayer in 2001. In May 2001, the grower notiﬁed us that he intended to use

rmn

horticultural spray oil on part of his ﬁeld. He allowed us to sample the trees before and
after the spray application to assess its effectiveness when applied with an airblast

sprayer. Therefore, the ﬁrst generation of pine needle scale was monitored in this ﬁeld,

 

rather than the second generation as in the other ﬁeld studies. In May 2001, before ,_
sprays were applied, 15 trees with obvious pine needle scale infestation were selected

with the grower’s help. At least four trees were located in each of three sections. The

three sections were named Groups A, B, and C (Table 1).

Objective 1. Phenology of second generation scales. In 2000 and 2001, we
monitored the duration of second generation egg hatch and crawler emergence in Van
Buren County. Ingham County was monitored in 2000 only, because the scale infestation
was too low in 2001. In each ﬁeld, up to 100 adult female scales were collected weekly
beginning in late June or early July, before any crawlers had hatched. Samples were
taken by removing approximately 20 infested needles from each sample tree on each
date. Needles were taken from the midcrown level in all four cardinal directions, when
the infestation level permitted. Uninfested needles were not chosen because our
objective was to monitor phenology of the scale insect rather than density of the

population on the tree. We examined the ﬁrst 100 female scales from each collection

129

under a microscope for evidence of oviposition and egg hatch. We determined the
duration of egg hatch by counting the number of live eggs per female on each sample
date. Each week, the number of scale insects on an additional sample of three to ﬁve
needles per tree was also examined under the microscope. The stadium and apparent
cause of mortality, if applicable, was recorded for each scale insect (stadia were
identiﬁed after Cumming 1953 and Nielsen and Johnson 1973). We determined when the
development of hyaline stage nymphs peaked by monitoring the proportion of hyaline
nymphs in the population each week until all had molted to the third instar.

Cumulative degree-days were obtained weekly from the Michigan State
University Agricultural Extension website (http://www.agweather.geo.msu.edu/) for the
closest weather station to each study site. Degree-day data from Lansing were used for
the Ingham County ﬁeld, Paw Paw for the Van Buren County ﬁelds and Grand Rapids
for the Montcalm County ﬁeld. Cumulative degree-days expressed as base 50°F (10°C)
were used because this is the best threshold for development of pine needle scale
(Mussey and Potter 1997, Burden and Hart 1989), and because degrees Fahrenheit are
usually more accessible than degrees Celsius to growers and extension personnel in the
United States (Mussey and Potter 1997, Herms 1990, Pruess 1983). Cumulative degree-
days are indicated by the symbol DDSO.

Objective 2. Efﬁcacy of horticultural oil. In 2000, the 40 selected trees in the
Van Buren County ﬁeld were randomly assigned to one of four spray treatments: 3 2%
solution of highly reﬁned horticultural oil (SunSpray 6E, Sunoco, Philadelphia, PA),
chlorpyrifos (a common organophosphate)(Lorsban 4E, Dow AgroSciences LLC,

Indianapolis, IN), water, and no spray (control). In Ingham County, we assigned ten trees

130

 

each to chlorpyrifos, oil, or water application. Because only 30 trees with adequate scale
populations were available in this ﬁeld, we used water-sprayed trees as a control to
ensure that we could differentiate effects of the oil and insecticide from any physical
effects of the spray itself. Treatments were applied in both ﬁelds with a backpack mist
blower (Solo model 423, Sindelﬁngen, Germany) at recommended ﬁeld rates (7.57 liters

oil/378.5 liters water applied at 935.4 liters/ha; 0.946 liters chlorpyrifos/757.1 liters water

 

a?
at 467.7 liters/ha) (English units: 2 US. gal/100 gal water at 100 gal/acre; 1 pt 1'
chlorpyrifos/ 100 gal water at 50 gal/acre). We attempted to simulate the use of an i
airblast sprayer as closely as possible by using a gasoline powered mist blower (spray ;
output of 24.5 ml/s) rather than a hand pump sprayer. To calibrate the mist blower, we a

determined the number of ml of solution that should be applied to each tree based on the
typical density of 2,963 trees per ha (1,200 trees per acre). We calculated the amount of
time needed per tree based on the spray output per 5. This was 6.4 3 per tree for
chlorpyrifos and 12.9 5 per tree for oil. We applied the spray to all sides while circling
the tree until time ran out and the foliage was wet, but not dripping heavily.

In 2001, we used a similar spray protocol in Van Buren County, although a
different brand of highly reﬁned horticultural oil was used (Superior Miscible Spray Oil,
Universal Cooperatives, Inc., Minneapolis, MN). The entire ﬁeld was sprayed aerially
with chlorpyrifos two days before treatments were applied. Forty of the trees we had
selected were enclosed and protected with white plastic bags during the aerial spray (per
Beard and McLeod 1992), and the remaining ten trees were left exposed.

Pre—treatment scale mortality. Approximately 20 infested needle fascicles were

pulled from each sample tree immediately before sprays were applied. Needles were

131

removed from all sides of the tree when infestation levels permitted. Samples were
placed in coin envelopes and held in the laboratory at 4°C until they were examined under
a microscope the following day. We examined enough needles to count 25 to 50 scale
nymphs per tree. We determined if each nymph was live or dead by poking it with a
minuten pin to see if liquid oozed out. Dead scale insects typically appeared to be dry,

ﬂattened, or otherwise desiccated. Samples were collected weekly for three weeks in the

an- m I
:4

same manner and the mean percentage mortality of scale nymphs was determined for
each tree. r

Objective 3. Effectiveness of commercial application of horticultural oil. The

 

grower applied different spray treatments in each of three sections, to test the difference
between using horticultural oil or a conventional insecticide. In this trial, a true control
(unsprayed or water-sprayed) could not be established because of the grower’s immediate
need to achieve control of the population. Treatments were assigned to each section
based on the logistics of using the airblast sprayer and other considerations, including
expected harvest date, and designated Groups A, B, and C (Table 1). Trees in Group A
received a 4% horticultural oil treatment (Damoil, Drexel Chemical Co., Memphis, TN)
(7 .57 liters oil/ 189.3 liters water at the rate of 467 liters/ha ) (2 gal oil/50 gal water at 50
gal per acre). Trees in Group B received 2% oil and trees in Group C were sprayed with
Metasystox R (S-[2-(Ethylsulﬁnyl)O,-O-dimethyl phosphorothioate)(Gowan Co., Yuma,
AZ), along with the fungicide Bravo (Zeneca Ag Products, Wilmington, DE), at label
rates (Metasystox R: 2.34 liters/ha at the rate of 467 liters/ha; Bravo at 3.36 g/ha in 467
liters water/ha) (1 qt/acre in 50 gal H20 and 3 lbs/acre in 50 gal H20). Treatments were

applied on 17 May 2001 with an AgTec airblast sprayer (model 2004). On 29 May 2001,

132

the grower reapplied all sprays. Samples were taken on 17 May before treatment, 22
May, 1 June, 4 June, and 9 June.

Statistical Analyses. Data were tested for normality using the Shapiro-Wilk test.
Data were transformed when necessary to meet the assumptions of normality and
homogeneity of variance. A one-way analysis of variance (AN OVA) was used to test for
differences in scale mortality in the oil and insecticide trials. When the ANOVA was
signiﬁcant, treatment means were separated using Fisher’s Least Signiﬁcant Difference.
For the Montcalm County 2001 trial, transformed data did not meet assumptions of

normality, so data were analyzed using the nonparametric Kruskal-Wallis test. For these

W‘ﬂa‘lgau‘ﬁf : mean “on Hilﬁgq

data, exact p-values were calculated using the Monte Carlo estimation. All tests were
conducted at a signiﬁcance level of a = 0.05. Data were analyzed using SAS v. 8 (SAS

Institute 1999).

Results

Objective 1. Phenology of second generation scales. Phenology of the second
generation egg hatch in 2000 and 2001 was generally similar among locations (Table 2).

In 2000 we sampled at the start of egg hatch in Van Buren County on 12 July (1280

DD50). In other cases, we did not sample at the very beginning of egg hatch, but crawlers
were already present by roughly 1230 to 1300 DD50 in all ﬁelds. Once egg hatch began,

eggs continued to hatch for approximately 2-3 weeks. By roughly 1800 DD50 in each

ﬁeld and year, the last crawlers had emerged, and egg hatch appeared to be complete

133

except for an average of ﬁve to seven eggs that were probably nonviable. While egg
hatch was occurring, the newly settled scale nymphs matured, often reaching the third

instar before hatch was completed.

The second instar, or hyaline stage, began to appear at approximately 1400 DD50,

and peaked between 1500-1600 DD50 in each ﬁeld (Figure 1). The calendar date of the

peak hyaline stage varied slightly, but occurred in the last week of July in Ingham and
Van Buren Counties, and the ﬁrst week of August in the more northern Montcalm
County. Our estimation of the peak hyaline stage depended on our sample date, which
also varied from year to year depending on the logistics of sampling several ﬁelds
simultaneously.

Objective 2. Efﬁcacy of horticultural oil. Horticultural oil applied with a
backpack mist blower performed at least as well as the conventional broad-spectrum
insecticide chlorpyri fos. In 2000 in Van Buren County, the pre-treatrnent mortality rate
averaged about 40% and there were no signiﬁcant differences among trees in the four
treatment groups (F = 0.31; df = 3, 36; P = 0.81) (Figure 2). After treatments were
applied on 25 July, the mortality rates on the unsprayed and water spray control trees did
not increase, but mortality in the oil and chlorpyri fos treatment groups was signiﬁcantly
higher than the controls (F = 8.58; df = 3,35; p = 0.0002) (Figure 2). Mortality of scales
on trees treated with oil and chlorpyrifos did not signiﬁcantly differ.

In 2000 in Ingham County, pretreatment mortality averaged roughly 26% for all
trees with no signiﬁcant differences among treatment groups (F = 0.07; df = 2,27; p =
0.93) (Figure 3). One week after treatments were applied, the average mortality

increased to 53.6% (Figure 3). Initially, oil appeared to perform slightly better than

134

WWW—.54» ,m .m—‘n 1
, .

chlorpyrifos, but differences among treatment groups were not signiﬁcant (F = 1.57; df =
2,27; p = 0.23). Four weeks after treatments were applied, the mortality rate averaged
29% for all trees, and did not differ signiﬁcantly among the treatment groups (F = 1.22;
df= 2, 27; p = 0.31).

In 2001 in Van Buren County, pretreatment mortality on 25 July averaged 25%,
with no signiﬁcant differences among treatment groups (F = 0.21; df = 4,45; p = 0.93)
(Figure 4). On 13 August, one week after treatments were applied, mortality rates in the
unsprayed, water-sprayed, and aerial ly sprayed treatment groups increased to an average
of 52.7%, but scale mortality on these three groups of trees did not differ signiﬁcantly
(Fisher’s LSD). Mortality rates on the trees treated with oil and chlorpyrifos increased to
77.1% on average by 19 August. On 19 August, two weeks after treatments were
applied, relative differences in scale mortality among treatments were consistent:
mortality rates in the oil and chlorpyrifos treatment groups were signiﬁcantly greater than
the control and aerially-sprayed groups, and mortality on the trees treated with oil and
chlorpyrifos were not signiﬁcantly different ﬁom each other (Figure 4). Notably,
mortality of scale insects on trees treated aerially with chlorpyrifos did not differ
signiﬁcantly from the two control groups.

Objective 3. Effectiveness of commercial application of horticultural oil. In
May 2001, we monitored the ﬁrst generation hatch in Montcalm County to help the
grower time spray applications for this generation. We found that by 22 May, most of the
eggs had hatched, and crawlers were beginning to turn yellow, an event which
immediately precedes the molt to the hyaline stage (per Nielsen and Johnson 1973)

(Table 2).

135

Pretreatment mortality of ﬁrst generation scale insects averaged 6.4% before the
ﬁrst spray application on 17 May, and did not differ signiﬁcantly among the three
treatment groups (Kruskal-Wallis H = 0.30; df = 2; p = 0.88) (Figure 5). After treatment,
on 22 May, the mean mortality rate for the 2% oil treatment was 55.8%, compared with
mean mortality rates of 47.4% and 35.9% for the Metasystox R and 4% oil treatments,
respectively. Differences in mean mortality rates among the three treatment groups were

not signiﬁcant (Kruskal-Wallis H = 1.28; df = 2; p = 0.58). The repeat spray applications

1“." ".1.de r‘.

on 29 May, when most scale insects were in the second instar, did not substantially

increase scale mortality. Mean mortality rates on 4 June were 40% for 2% oil, 36.5% for

:73! ‘77. WNW} A 1

 

Metasystox R, and 23.5% for 4% oil, and differed signiﬁcantly among treatments

(Kruskal-Wallis H = 6.04; df = 2; p = 0.0378).

Discussion

The phenology of the second generation of pine needle scale was complicated by
the extended hatching period. Nielsen and Johnson (1973) also found an extended
second generation hatching period in C. heterophyllae in New York. All populations
studied were biparental (Stimmann (1969) had observed a uniparental population).
Although several factors can inﬂuence the phenology of the scale, our study showed that
it was more predictable by use of degree-day accumulation than calendar date. This

correlation between degree-days and peak crawler hatch in the second generation is an

136

important tool for determining the best time to spray and to avoid costly and ineffective
sprays. The phenology of the second generation was closely matched to degree-day
accumulation in both years. This close relationship is not unusual among diaspidid scales
(Beardsley and Gonzalez 1975).

Preserving natural enemies, if they are present in a ﬁeld, can be an important
addition to a program using oils as an alternative to broad-spectrum chemical
insecticides. Horticultural oils have been shown to be harmless to coccinellids, an
important predator of armored scales (Smith and Krischik 2000). A reduction in the use

of broad-spectrum insecticides may increase the potential for control by natural enemies

 

of the scale (i.e. Luck and Dahlsten 1975). Mortality caused by the oil spray combined
with conservation of beneﬁcial predators and parasitoids may provide adequate control in
most situations.

Our results indicated that using horticultural oil for control of pine needle scale on
Scotch pine Christmas trees was at least as effective as using broad-spectrum chemical
insecticides. This conﬁrms the observations of other workers (Nielsen 1990, Neilsen
1970, Gambrell 193 8) who tested the efﬁcacy of horticultural oil to control armored
scales. In each of our trials, we carefully timed our spray applications to coincide with
the completion of second-generation egg hatch and highest proportion of hyaline stage
nymphs. This would be an ideal time because the hyaline stage is still susceptible to
insecticides, and the eggs are no longer protected under the white waxy scale armor
(Nielsen and Johnson 1972). In our study, the application of horticultural oil increased

pine needle scale mortality by roughly 40%, regardless of the initial level of mortality.

137

In the ﬁeld, the timing and manner of spray oil application are critical for success;
because of its mode of action, adequate coverage must be achieved for the oil to be
effective. Our results ﬁom Montcalm County showed that a common tractor-mounted
airblast sprayer was capable of achieving adequate spray coverage to cause signiﬁcant
mortality on trees with moderate pine needle scale infestation. However, no signiﬁcant
differences were observed among treatment groups either before or after treatment,
illustrating that the oil treatments were as effective as the chemical insecticide. The small
sample size is an artifact of using an active commercial operation for the test rather than
an experimental plot, but the use of a working farm demonstrates that horticultural oil can
be effective if applied with an airblast sprayer. At least, horticultural oil is no worse than
chemical insecticides—~neither product achieved much more than 55% mortality. The
difﬁculty of achieving adequate coverage regardless of the product used may be one
reason for this low level of mortality.

Concerns of Christmas tree growers about phytotoxicity of horticultural oils may
stem largely from anecdotal reports or past experience. Early horticultural spray oils did
result in phytotoxicity, depending on the properties of the oil and the tree species used
(Riehl 1990). The unsulfonated residue (UR) content of spray oils is associated with
damage to foliage on citrus trees (Riehl 1990). Currently, there is no evidence of injury
to plant foliage from oils with a 92% or higher UR (Riehl 1990). The UR content of
spray oils used today is mandated at a minimum of 92% (Riehl 1990). The highly reﬁned
oils we used in our study had a UR of 98% or better, reducing the risk of phytotoxicity.

We did not observe phytotoxic effects in our ﬁelds, nor did any grower bring this to our

138

attention. However, several varieties of Scotch pine tend to have yellowish foliage
naturally (Eliason 1996) which may have masked any phytotoxic effects.

The cost of using horticultural oil has also been perceived as an obstacle to
increasing its use as an insecticide in Christmas tree plantations. In Montcalm County,
the second oil spray by the grower on 29 May did not signiﬁcantly increase scale
mortality, indicating that a single spray application would have been equally effective
and less expensive. As an example, the grower reported that Damoil cost US $5.19 per
US. gallon (gal), chlorpyrifos was US $44.50/gal, and MSR was US $78.00/gal. This

corresponds to $10.38 to $15.57 per acre for Damoil, $11.12 for chlorpyrifos and $19.50

 

If." u'l: I

for MSR, given the standard delivery rates used in Christmas tree production (per the
product labels). On a per-acre basis, the cost of using horticultural oil was similar to the
costs of the broad-spectrum insecticide products for this grower.

Control recommendations. Several factors can make pine needle scale difﬁcult
to control in commercial Christmas tree ﬁelds. Its capacity to rapidly increase its
population size, its protective covering, and small inconspicuous size all contribute to the
difﬁculty of detecting and controlling pine needle scale before it becomes a notable
problem (Eliason and McCullough 1997, Nielsen and Johnson 1972).

Careful timing of insecticide application is important to adequately control pine
needle scale, regardless of the insecticidal product used (Nielson 1970, Martel 1972).
Our data showed that an ideal window for spray application occurs between 1500-1600
degree-days, coinciding with the maximum number of second instar or hyaline stage

nymphs (Nielsen and Johnson 1972, Martel 1972). Using the published degree-day

139

accumulations available in newsletters or on the web would be helpful to growers
planning to control pine needle scale.

The mortality rates following treatment applications at our study sites did not
result in complete scale mortality in any ﬁeld. The perception among growers that total
mortality is necessary to achieve adequate control has often led to overuse of pesticides,
not only in Christmas trees but in many agricultural crops (Rose 1990). The
establishment of an economic threshold level for pine needle scale would be an important
addition to an integrated control program. This would demonstrate that 100% mortality
is not necessary to achieve an acceptable level of control (i.e. Sadof et al. 1987). The use
of horticultural oils has beneﬁts that render it a viable option for control of pine needle

scale, especially when populations are at low to moderate densities.

140

References Cited

Beard, K.K. and B.M. McLeod. 1992. Pine needle scale control on Michigan Christmas
trees. Down to Earth 47(1):16-19.

Beardsley, J r., J.W. and RH. Gonzalez. 1975. The biology and ecology of annored
scales. Ann. Rev. Entomol. 20: 47-73.

Brown, CE. 1959. Reproduction of the pine needle scale, Phenacaspis pinifoliae
(Fitch), (Homoptera: Diaspididae). Can. Entomol. 91(9): 529-535.

Burden, DJ. and ER. Hart. 1989. Degree-day model for egg eclosion of the pine
needle scale (Hemiptera: Diaspididae). Environ. Entomol. 18(2): 223-227.

Burden, DJ. and E. R. Hart. 1990. Parasitoids of Chionaspis pinifoliae (Homoptera:
Diaspididae) in Iowa. Great Lakes Entomologist 23 (2): 93-97.

Burden, DJ. and ER. Hart. 1993. Parasitoids associated with Chionaspis pinifoliae
and Chionaspis heterophyllae (Homoptera: Diaspididae) in North America. J. Kansas
Ent. Soc. 66(4):383-391.

Cooper, D.D. and W.S. Crenshaw. 1999. The natural enemy complex associated with
the pine needle scale, Chionaspis pinifoliae (Fitch) (Homoptera; Diaspididae), in North
Central Colorado. J. Kansas Ent. Soc. 72(1): 13 1-133.

Cumming, M.E.P. 1953. Notes on the life history and seasonal development of the pine
needle scale, Phenacaspis pinifoliae (Fitch). Can. Entomol. 89(9): 347-352

Dahlsten, D.L., R. Garcia, J.E. Prine, and R. Hunt. 1969. Insect Problems in Forest
Recreation Areas: Pine Needle Scale. . .Mosquitoes. California Agriculture (July 1969):
4-6.

Doane, RW. 1926. Controlling mealybugs on ornamental plants. Pan-Paciﬁc Entomol.
2(4):213-214.

Eliason, E.A. 1996. Evaluation of the susceptibility of four Scotch pine Christmas tree
varieties to insect pests. Master’s thesis, Department of Entomology, Michigan State
University, East Lansing, Michigan.

Eliason, E.A. and D.G. McCullough. 1997. Survival and fecundity of three insects
reared on four varieties of Scotch pine Christmas trees. J. Econ. Entomol. 90(6): 1598-
1608.

141

 

Gambrell, EL. 1938. Experiments for control of the pine needle scale, Chionaspis
pinifoliae (Fitch). J. Econ. Entomol. 31(2): 183-186.

Gambrell, EL. and F1. Hartzell. 1939. Dormant spray mixtures on conifers. J. Econ.
Entomol. 32(2): 206-209.

Herms, D.A.1990. Biological clocks. Am. Nurseryman 172(8):56-63.

Johnson, W.T., and H.H. Lyon. 1988. Insects that feed on trees and shrubs, 2“d ed.
Comstock University Press, Ithaca, NY.

Luck, RF. and D. L. Dahlsten. 1974. Bionomics of the pine needle scale, Chionaspis
pinifoliae, and its natural enemies at South Lake Tahoe, Calif. Ann. Ent. Soc. Am. 67(3):
309-316.

Luck, RF. and D.L. Dahlsten. 1975. Natural decline of a pine needle scale (Chionaspis
pinifoliae(Fitch)) outbreak at South Lake Tahoe, California following cessation of adult
mosquito control with malathion. Ecology 56: 893-904.

Martel, P. 1972. Le malathion et le diméthoate dans la lutte contre la cochenille du pin,
Phenacaspis pinifoliae (Fitch) (Homoptera: Diaspididae). Ann. Soc. ent. Québec 17: 20-
23.

McCullough, D.G. and Fondren, K. 1998. What’s Bugging You—FQPA and
Insecticide Use in Michigan Christmas Tree Fields: Preliminary Results from 1998
Survey. Michigan Christmas Tree Journal 45(3):22-29.

McCullough, D.G., S.A. Katovich, M.E. Ostry, and J. Cummings-Carlson [eds.].
1998. Christmas tree pest manual, 2nd ed. Michigan State University Extension Bulletin
E-2676. East Lansing, MI. 143 pp.

McClure, M.S. 1977. Resurgence of the scale, Fiorinia externa (Homoptera:
Diaspididae), on hemlock following insecticide application. Env. Entomol. 6(3): 480-
484.

Mussey, G.J. and D.A. Potter. 1997. Phenological correlations between ﬂowering
plants and activity of urban landscape pests in Kentucky. J. Econ. Entomol. 90(6) 1615-
1627.

Nakahara, S. 1982. Checklist of the armored scales (Homoptera: Diaspididae) of the
conterminous United States. US. Dept. Agric. Animal Plant Health Insp. Serv.-Plant
Protec. Quar. Publ. No. 1089.

Nielsen, D.G. 1970. Host impact, population dynamics, and chemical control of the pine
needle scale, Phenacaspis pinifoliae (Fitch), in central New York. Ph.D. thesis,
Department of Entomology and Limnology, Cornell University, Ithaca, New York.

142

Nielsen, D.G. 1990. Evaluation of biorational pesticides for use in arboriculture. J.
Arboric. l6(4):82-88.

Nielsen, D.G. and NE. Johnson. 1972. Control of the pine needle scale in central New
York. J. Econ. Ent. 65(4): 1161-1164.

Nielsen, D.G. and N .E. Johnson. 1973. Contribution to the life history and dynamics of
the pine needle scale, Phenacaspis pinifoliae, in central New York. Ann. Ent. Soc. Am.
66: 34-43.

Pearce, C.W., A.W. Avens and P.J. Chapman. 1941. The use of petroleum oils as
insecticides. J. Econ. Entomol. 34(2):202-212.

Pruess, K.P. 1983. Day-Degree methods for pest management. Environ. Entomol. 12:
613—619.

Raupp, M.J., C.S. Koehler, and J .A. Davidson. 1992. Advances in implementing
integrated pest management for woody landscape plants. Ann. Rev. Entomol. 37: 561-
585.

Rich], LA. 1990. Control Chemicals. pp 365-392. In D. Rosen, [ed.], Armored Scale
Insects: Their Biology, Natural Enemies and Control, vol. B. Elsevier Science Publishers,
Amsterdam, The Netherlands, 1990.

Ripper, W.E. 1956. Effect of pesticides on the balance of arthropod populations. Ann.
Rev. Entomol. 1: 403-438.

Roberts, F.C., R.F. Luck, and D.L. Dahlsten. 1973. Natural decline of a pine needle
scale population at South Lake Tahoe. California Agriculture October 1973: 10—12.

Rose, M. 1990. Periodic colonization of natural enemies, pp. 433-440. In D. Rosen,
[ed.], Armored Scale Insects: Their Biology, Natural Enemies and Control, vol. B.
Elsevier Science Publishers, Amsterdam, The Netherlands, 1990.

Ruggles, A.G. 1931. Preliminary notes on the biology and control of the pine leaf scale,
Chionaspis pinifoliae Fitch. J. Econ. Entomol. 24:115-119.

Sadof, C.S., M.J. Raupp and J.A. Davidson. 1987. Survey ﬁnds defoliated plants
won’t sell. Am. Nurseryman, August, 37-39.

SAS Institute. 1999. SAS/STAT user’s manual, version 8.1. Cary, North Carolina.

Sheffer, B.J. and M. L. Williams. 1987. Factors inﬂuencing scale insect populations in
southern pine monocultures. Fla. Entomol. 70(1): 65-70.

143

Shour, M.H. 1986. Life history studies of the pine scale, Chionaspis heterophyllae
Cooley, and the pine needle scale, Chionaspis pinifoliae (Fitch). PhD. dissertation,
Department of Entomology, Purdue University, West Lafayette, IN.

Shour, M.H. and D.L. Schuder. 1987. Host range and geographic distribution of
Chionaspis heterophyllae Cooley and C. pinifoliae (Fitch) (Homoptera: Diaspididae).
Indiana Academy of Science 96: 297-304.

Smith, S.F. and V.A. Krischik. 2000. Effects of biorational pesticides on four
coccinellid species (Coleoptera: Coccinellidae) having potential as biological control
agents in interiorscapes. J. Econ. Entomol. 93(3): 732-736.

Stimmann, M.W. 1969. Seasonal history of a unisexual population of the pine needle
scale, Phenacaspis pinifoliae. Ann. Entomol. Soc. Am. 62(4): 930-931.

Q (T. ‘ ﬂ‘ﬂq

Tooker, J.F. and L.M. Hanks. 2000. Inﬂuence of plant community structure on natural
enemies of pine needle scale (Homoptera: Diaspididae) in urban landscapes. Environ.
Entomol. 29(6): 1305-131 1.

 

Walstad, J.D., D.G. Nielsen and N .E. Johnson. 1973. Effect of the pine needle scale on '7
photosynthesis of Scots pine. Forest Science 19(2): 109-111.

144

ooh 85 .8 82m 885 8 N 8 8:88: .89 8.5.8 o. A u .983 ”ooh 05 .8 82m 8.5 .85 an 8
8:82. .89 838 07m M 888.82 ”nob 85 .8 02m 28 .38 no 8 6:80: 8a .02QO m 52... 83$ u .82on .3221:

 

 

a 850on 8 Em: 589 m. 9833: noaoom

v .98.. 8 888.82 58.» o .820 :2.on

on .98.. 9 888.82 589 m 8?... 8.5

ow .98: 9 88.8.82 58» 5 82m 888m 815

on am: as... : 85 8.55m 8.3

com: womb 82 28.8. 853805 owe cob. .385; BC.

0

as m 888.38

< 580.5%
.3580 5.8822
Sow.
coon
.3580 5.5m 5~>

ooom
.3580 55%.:

 

RE :28 5 com: moot .8 58.55: 85. moms £8.85; .— «Emu.

145

. . . . . J . . . . 1
F 3.911... .0503... hump—H.H.. 7k:- dd‘. I 33:? .L

 

9% a 3. cm 8: be. m
.3580
o 28 as. R 88 5.83
855 Rm 8
.88 £22 2:8 :3. a 3.8 9. 8.: 583. w
8225.
:5 wwwo 858
”85.8
323.. 85830 .2 £3 a I: a. 98. same... _
ems... 8:8. sea Gd n. Z: we a: .2.: am
25552. 6
owe... 8:5: o: a NE 3 mm: 33 om M
maze... 222.8 8.0 a «.2 2: 82 a: 2
:22
.2388 8:55 .2 a «.2 mm 85 .23 S
3.. a 9mm 8. we: 33 e
.3580
.l 3o a $6 2: m8 8.: em 88 55m 8.>
682588 ROG
mﬂcﬂatomA—O o—NEQH .31 wwwo .OZ maid-“07.“ viva—55:0 gun hao> 63m.”—

 

wotom 8582 .85. 82858 maize .858 28... 8358.. 8.855 8808 wwwo .8 8852 .N 03:.

.3580

 

888:. m8.850 wad a as. we Rm. .33. a. Sam :83. 5.>
oweﬁ . . 8»:
:8... 5 8.88.. .83. o. N a n a m mmw. . < o.
owe. 28.3 8.3.. w; a n. : 2. 98. 58:... m
8888 58.380 .N. a 88. mm no... .33 R
a... 3.2 2 8: be. 2
828.88% 28.8%
32.8.. .853 .
$232... 8.8... .0 XS n. B _ m. 8.. a: 8
58.268 8.5m Bo: .3580
3mm 82:3 80.380 a 0.. m . omm 92>. 2 88 5.8.8.).
228.88% Raﬁ .4/4
S... a we... v. 35. same... m
85.5. . . .A 5
8.2.5.8 8.3.3.838 3.. mm o a em 0. we con. . .. .m
ems. 2.53 n.8,. we. .1. 2: E on: 33 R
58.388 8.5.. . . 9 a
.858 .85. 8533. .m . H o a. 9. on... . .. em
5 . . .A 2. .3580
mm m H h mm mm mmm. . .. 52w...
35588 San—
8833830 8.55.. 8.. ﬁne .8 7. 85......” 3.8.5530 8.5 89> 20....

 

8.88 N wish.

cunt 25 ﬂo ~50 0.5.5 "mu—Now ﬂomwmhoﬁow UGOOOm Oﬂm

 

 

8.35. -
Ea 8.2: :32 wwwo m. n m :8 mama... m.
ommsq
one as. 8.85. So a w... 2 5: saw? o
.86... Em $05.9...
55
0mg 2:2. .8. who a 5.: 3 £2 3:. mm 5:”. 5%
35883 a
3:53.330 2:89.. .5.— muuo 67. 8.35.. gun—sﬁmw 8.5 30> 29....
67. .

 

6.2.8 N 2%...

148

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00 O Montcalm 1999
' I
A) ' I O I
a, 701 I : : IVan Buren1999
8 . ,
0) ' I
g 50 . : I
E : :0 <>
.5 4O ‘ I : I 0
§’ <> i : <>
2:3 20‘ : : I I
d) I
0- 10 . O : :
I
0 L0 . . . 1 . : . r . . T . .
1227 1315 1354 1392 1509 1520 1695 1744 1876 1990 2025 2066 2197 2200
Degree days base 100
O Ingham 2000
7° . A . IVan Buren 2000
I I
8: 60 4 B) : O : AVan Buren 2001
g ' :
9 5°“ : . A
'-E- O I O I
g 40 4 I I
.C I I
.E O I I A
a: 30 1 I I
6 . : :
c 20 — I
g : . A
a 10— l l
I I
o . 1 . l l

 

 

 

 

1232 1280 1422 1436 1500 1566 1588 1700 1817 2100
Degree days base 100

 

Figure 1. Percentage of scale population in the 2nd instar, or hyaline
stage, in each ﬁeld. A) Montcalm and Van Buren Counties, 1999. B)
Ingham and Van Buren Counties, 2000 and 2001. Dotted lines indicate
approximately 1500-1600 degree days. Data were not available for
every sample date.

149

 

 

 

+Oi|

 

 

 

 

  

 

 

 

 

 

 

 

 

- - 0 - - chlorpyrifos
100 _2_ 5 DA i ’r __i T ”V .____.___~_.-_ Water
—E}— No spray

80 — —* 2 - -_
.é‘ .......... a
E
E 60 ~ ‘3 -
E 40 _ __a, . ' ‘* 4 b h
g a T b V ‘ 1f
0) \-. b
o. a

20 - — 2 ———— —— .L

0 T I 1
25-Jul-00 01-Aug-00 08-Aug-00
Date

 

 

 

Figure 2. Mean (+/- SE) pine needle scale mortality in the Van Buren
County ﬁeld in 2000. Treatments were applied on 25 July after
pretreatment samples were taken. Signiﬁcant differences among
treatment groups on each date are indicated by different letters.

N = 10 trees per treatment.

150

 

 

 

 

 

 

 

 

 

 

 

100 - -A- bit
- - o - - chlorpyrifos

80 -~ — — — ———# — —Cl—Water
g
g 60 i-
o
E
E
g 40 -2
(D
o.

20 - --

0 I T
31-Jul-00 08-Aug-00 01-Sep-00
Sample date

 

 

 

Figure 3. Mean (+/- SE) pine needle scale mortality in the Ingham
County ﬁeld in 2000. Treatments were applied on 31 July 2000 after
pretreatment samples were taken. There were no signiﬁcant differences
among treatment groups on any date. N = 10 trees per treatment.

151

 

 

 

l oil

Elchlorpyrifos (mist blower)
100 [ll chlorpyrifos (aerial)
Ewater

El control

 

 

 

 

 

  

A
O
1

Percent scale mortality

20-

 

 

 

 

 

 

 

 

25—Jul-01 13-Aug-01 19-Aug-01

 

Figure 4. Pine needle scale mortality in Van Buren County, 2001.
Treatments were applied on 6 August 2001. Means were separated
with Fisher's LSD where the global ANOVA was signiﬁcant (p <
0.05). Means with different letters are signiﬁcantly different. Data
are presented as mean +/- 1 SE.

152

 

 

 

IMSR

 

 

 

 

 

 

 

 

   

1 7-May 22-May 01-Jun O4-Jun OQ-Jun

 

  

 

100 Eloil2%
Eloil4%
80~ ----- — + - —
g
g 50 1V ‘ —-—-—
E
‘* 22
‘L ’/
20 ~—— ¥ /4
/
M 72' ‘1' 5/

 

Figure 5. Percentage of mortality (mean +/- 1 SE) in the Montcalm
County ﬁeld in 2001. Treatments were applied on 17 May and repeated
on 29 May 2001. There were no signiﬁcant differences among
treatment groups (Kruskal-Wallis; n=4) except on 4 June. Signiﬁcant
differences on 4 June are marked with different letters.

153

 

APPENDIX

154

Appendix 1
Record of Deposition of Voucher Specimens*

The specimens listed on the following sheet(s) have been deposited in the named
museum 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.: 2002-01
Title of thesis:

Alternative Control Methods for Two Important Insect Pests of Christmas Trees
Museum where deposited and abbreviations for table on following sheets:

Entomology Museum, Michigan State University (MSU)

Investigator’s Name:
Kirsten M. Fondren
17 1 ﬂ

W
Date 7[ [OZ 2&2 _.

*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.

155

Appendix 1.1

Voucher Specimen Data

 

Page I of 8 Pages

i|\

"MinaN 33K. W\

37.5.55 88m 55:22 05 E “Renew

 

INOO

Sen

 

 

.5.“ 2555on 58m: 96% 05 cozooom

 

Bacon .2 5389

 

 

 

 

 

 

 

 

 

 

5-88 .52 5:953,
5 5 _ 38 as s 88 as: E 5550 855. ﬂow 5.5535 5255...:
82 52 2 .3550 8.55. :82 355.56 5:555:
$2 5.2 S 5550 SEE :8: 5.5.535 5:555:
_ 32 .32 cm .3550 0535.5 250 zoom 55835 5.523553.
~ 35— .QE my .3550 0526; 5:50 55M 358.55 3535:
_ 33 .32 M: .5550 0526C. 5:50 55M 5.35835 3555:
_ 32 .32 c 5550 3.5.5; 5550 zoom 9.52535 .5535:
_ 32 .52 on .5550 8555 :82 5.5.5.55 525552
_ _ 82 52 am 5550 5555 zoom 5.5.555 5:555:
_ 82 >52 mm 5550 8555 zoom 5.5.635 5:555:
_ 9.52 .52 S .5550 8553 535 555535 5:555:
5 82 52 N . .5550 8555 535 5.5.035 5253...:
1s;
d
u
b
m R Mb 6 Q0
m w. (w\ .m s 3:5 5 28 5me 5:5 5 85on
w u e e e 58: .5 350:8 5055on 5.“ 93 583 .
d 10 0+ x a .g .C .
m m m m m m
.s. .w a m m... m .00. .m .m
uwhamaamm
M d m A A A S F F

 

 

in; 15 6:52

 

 

156

Voucher Specimen Data

 

Page 2 of 8 Pages

Ban—

500.50

 

.58255 38m 55:22 0.0 5 05000
5.0 5855on 03m: 05% 20 0030032

 

08D

 

5:032 awe—0590m—

 

 

 

_o-Noom .02 5:25>

:805m .2 0259

 

ﬂ'v—INNv—Iv—Iw-INv—Iv—IN

88 53 R .5550 5505

88 5:. m .5550 5555

88 05; R .5550 5555

88 553.. 2 .5550 5.50 55>
88 53 cm .5550 550 5>
552 5555.5m 2 .5550 550 5.>
aaa— HDQEBQDW h .bgonv Guam SS!
88 55.2 S .5550 58552
88 be. S .5550 58:52

58 52 S .5550 555.52

552 5552.5m 5 .5550 55562

320000 0020320000 0.0582030
320000 0303500000 5.0932030
Axe—000v 0500300000: 5.3000030
Que—000v 0303550000: 0.05.0530
920000 0300300000 0.033530
905000 030030.000: 05502030
320000 002.306.0000 0.058530
920000 00-§Q9~0~0- 5.0562030
Axe—000v 002.030.0000 0.05.0530
303000 0509390000 03.30030
920000 0330300000: 5.5502030

 

 

 

 

 

 

deposited
Other
Adults 6‘
Adults SB
Pupae
Nymphs

Museum

where

 

Larvae
Eggs

 

020.5000 05.. 003
5 088:8 505505 5.0 800 5.05

 

553 0005 .5 85on

 

 

55> 5 555.2

 

 

157

Appendix 1.1

Voucher Specimen Data

 

Page 3 at 8 Jages

859

55550

 

 

23550355 355 gmﬂoaz 05 E 05500

85m

 

550532 3058335

 

 

5.3 55085055 085: 0555 05 003083

 

_o-Noom .oZ 5:0:0>

5050.553 .2 585.53

 

v—tv—tq—ly—tv—tu—‘v—qp—‘ﬁ

:55 53 55 .5550 5.55 55>
am? 5558an cm 5550 :05m 05>
552 5553. M: .5550 855 55>
82 53 2 .5550 5.55 55>
:55 53 R .5550 552.52

:55 53 R .5550 55552

mamm Hun—EUHQDW a .5“:on 833.52
552 53 S .5550 55252
552 53 5 .5550 5525.2

35$ 05355 5550330
35$ 05355 53.80330
359 35355 5550330
359 0550.35 55.50330
35$ 05mg. 55.60330
35$ 0:585 55.80330
A253 0:585 5550330
955 55%.:5 53.50330
95$ QEME. 55.80330

 

 

 

 

 

 

 

 

 

S 552 83 5 .5550 5255 2550 555.5. 5:85.50
2 32 83 5 .5550 5.555 9550 5:03 5:85.50
2 552 552 on .5550 5555 25550 555.5,... 5585.50
2 552 .52 55 .5550 5255 9550 555.5. 5:85.50
m .w 10 0+ .m
u .n m m e 00050500 0:5 005:
m m m. m M M m. m m M.» 5 088:8 50085055 5.“ 5550 5554 5555 5:5 5 505.0%
M w d 0 A A n"... N L E

 

 

'—

“.5 05:52

 

 

158

3mm H9830

 

.8532 $038025

 

8mm

 

3633‘s 88m gwﬁoaz 05 E 389%

 

8m 2.258% wow: 95% 05 Biooom
8-88 .02 8558/

 

GOHUGOR .2 GOwng

 

 

Appendix 1.1
Page 4 of J Pages

Voucher Specimen Data

 

Sea §§< o .3960 85m §>
a¢m~ hDQEOqum h .3550 85m Gw>
32 33 Q 5950 89m 5>
82 sneoaom m 5550 ﬁasco:
8% saw? 2 5550 38302
82 33 S 5580 833.52

32 be: 2 5550 83382

3% b3 0 5550 829:

32 bi m 5550 seems

$2 2:: a 5550 52%:

8:80qu SENS 338383:
AcacoUod SEQ? uEmEEEE
Ava—50016 3:935 338333:
3:809: 3335 3&93232
Aoaaooouc £3.35 338333:
AoEOUoAV 323.2: 633333:
EEOUQHV unmet: 3.48323:
328005 33.35 3.48323:
3:80qu 53.0.2: 339323:
325096 3:32: US$323:

 

 

 

 

Museum
deposited

where
Other

vozmomov 98 c8:
8 @8828 mauﬁmoomm Bu 3% $93

 

 

 

 

Nymphs
Larvae
Eggs

 

:83 .850 3 86on

 

a...
O

 

a
O

 

Z Pupae

 

159

Appendix 1.1
Voucher Specimen Data

080 8850

 

 

080
.8332 $208859

3882.5 88m 5&9va 06 9 083%
.80 mac—98% 93m: 0258 05 9988M =298m .2 988.5—
_o-NooN .oZ Hcacao>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

~ 88 :8... mm 5550 9.85.9 880 9880 833% 3.8.3.3:
~ 33 05;. 2 .5580 080280. 98.5 30800 8933:: 35333:
_ 32 :5... R 55.8 8.5.2 90800 8.8838 essaa§
~ SON >82 3 5880 88.85. 98.5 98985 39.3 35333:
W . 88 E2 2 55.8 8.25 85.0 9856 5...... 3.8.5.32
n.” m OOON .82 N 5550 0&9»th 985 98985 88.5 3.22339
8 m 33:5 833 Bob 3.80% Bio—Esm 83$sz etmmembﬁu
f
M m SON 28:. o .3550 388C. 985 mam 3338 98.5.0
e w oooN 0:3. m .3550 0.0.885. 98.5 98m 8338 8&8580
W... N oooN >32 0N .3580 08285. 98.5 8% 833.8 3.93930
9 oooN 05:. 0 $9500 8885. 98.5 maoanoam 3&9. 33985mm
N GEN 82 0N 5:300 09085. 98.5 9.23on 35%...” 8.398280%
ﬂ oOON >82 2 5560 8.88; 9:25 98.3on Sauna... wascgmﬁem
_ oocN :5.“ N .3950 0828; 98.5 2893on 32mm»... nameoxmﬁmm
8 mm.
m e .m r w e .m. x 9.383% 98 com: :85 .850 he mug—comm
m m m. e M m. m. m % 8 988:8 9.298% 8.0 8% 984 .
M w d m A m N L E
mo .828: Z

 

 

 

160

Appendix 1.1

Voucher Specimen Data

8.5 .5850

 

 

8.5

 

5.85.2 85558.5

 

38.8.55 85m 5.58:2 on. 5 88.8..

 

.8 9858on 8.8— 8.58 85 .8883.
S-NooN 52 8.8.5.»

 

8.95.... .2 8.89

 

32 82 on 5.58 082.; .85
so. 8.2 E 5.58 ass.

88 8.8 8 5.58 8.5:...

SON .82 E .5550 888.... 985
oooN .82 o. .3550 888.8 98.5
9.5. 8...... w .5550 88.8.... 98.5
82 2.... 8 5550 082.; .88
a... 8.2 8 5550 082.8 880
93. .82 w. .b550 888.9 98.5
33 .82 N. .8550 388.... 98.5
888. 8: o... 5.58 888.3.
a... .83. w. 5550 88m 8.>
88 :5... 8 5.38 3.3.... 2.80
88 8.8 mm 55.8 8...»...

88 5.2 cm 5.58 8.59..

88 82 N 55.8 8.55.

88 8.8 m. 5.58 88.8

Page 6 of 8 Pages

 

55 338...... 8......

3.9 .8... 83....

A89 .88... 88:9

.88 .3... 8......

3.8 .3... 8......

2.8 .8... n.s....

3.8 .2.... 8......

3.8 8.... 8......

9.8 .3... n.s....

3.8 .2.... 8......

$80 .2.... 8......

38$ 885.88% 828.5%%..%%
989 885.88% 8588.35
989 888.88% 853.583..
989 885.88% 855.5%5t
9..@ 883.88% a...§~5%%..m%
A89 888.88% u...§..5%%..%%

 

 

 

 

 

 

 

 

Nc—tq—uv—tv—nv—Iv—av—I—nv—Av—nc—av—tv—ov—tv—t—d
Adultsgd‘

 

 

 

 

 

d
.m m 8.8285 98 9.5
m m m a m P w s 5 9.88:8 58.58% .5 88 o 84 .58. 8:5 .5 88on
m .pn. w. n. m. m g . m 3 n
M w d % m... N L .m.
”.5 895.2

161

.8350

 

bﬁgaa 39m 3222 ”a 5 E23
8m 2886on 3%: 025m 2.: 3380M

Snows—2 hwoﬁaoucm

 

8mm

 

 

 

Sumoom .oZ H“52:5

585m .2 seam—UH

 

v—iv-iv—dv—lI—i

—-4

Appendix 1.]

Page_]_of__8_Pages

Voucher Specimen Data
v-¢ '— N m

v—lv—I

83. as: a 5550 gamma
88 ~$2 mm 5550 oo§§§
88 as: e 5550 BE?
88 52 em 5550 8E5
88 Q: t 5550 5.55

88 $2 2 5550 85%.? 9:20
ocom >22 N .3550 3.52»? 9:20

33 32 cm .5550 080228 28.5
amﬁ E2 2 558 03025 266

33 >32 0 .3550 86>“; 93.5
82 E? R 5550 3525 case

33 32 cm .3950 our—38H 956
88 E? m 55.8 Banana

3m@ 5855 SummoNuAD
Axmmv anti: 332393
3me 3335 Semzﬁuab
Qmmv £235 33:33.9
Ahmmv 5335 362833

“madam SSSESS: SBSBQD
“3.63 889:5962 £33980

:3on haehﬁggut
SSSMQE???» £33860

Bum—=2 SSNRSQ. 83935;“ 335800
H5332 883:3 3393?» 332.880

 

 

deposited
Other
Adults 6‘
Adults £2
Pupae
Nymphs
Larvae

Museum

where

 

 

 

 

 

 

 

 

358% can wow:
8 330:8 £586on Ho.“ 83 Ban..—

 

coxﬁ .850 8 momooqm

 

 

 

CH
O

 

162

DHMQ HON—MSU

 

 

8am

 

.8332 awe—080:5

 

bangs 88m gwﬁomz 2: E :momoc

 

.8.“ £586on 36: 26% 05 Blooom
8-88 62 8:025

 

868m .2 =25

 

 

Appendix 1.1
Page 8 of 8 Pages

Voucher Specimen Data

v—lv—IN

88 a: 2 .358 8§§§
58. a: E 5580 unease 880
88 be: a 558 82?:

Axmmv 3:339 SEEMS;
3me Savage SQEMSNV
Emmy as???» GEEMSNV

 

 

 

 

 

 

 

 

 

 

 

 

 

d
10 0+
m e .m s s e .m. e 333% 23 wow: :93 .850 8 88on
e r s a h k a a s no 360:8 8088on 8% ﬁat 693 .
m m m. m a a m. m m a .
M w x o A A P N L E
mo 829.2

163

   

    
 

MICH

IIIII'“II

SSSSSSSSSSSSSSSSS

IIII3III III III III II

 

III