:4 ..
ass-nun q;

"|'
“I

'Iltﬁlm

.3? _,

NIL

wk

‘ Like»

. .

9

h
m

if

. J’s}

wﬁdwﬁﬁ’m

. E

 

 

mm. A
'3: '-,
”4;"...
#33::

’al‘

; SW39!
. Cﬁ‘rffﬁs“
ﬁg] V V '\

7:-

:3“
~€

 

‘
w.

 

3:05 LIBRARY
Mich .. State
University

 

 

 

This is to certify that the
dissertation entitled

POTENTIAL FOR BEHAVIORAL AND CULTURAL MANIPULATION
OF JAPANESE BEETLE (POPILLIA JAPONICA NEWMAN,
COLEOPTERA: SCARABAEIDAE) POPULATIONS

presented by
Zsofia Szendrei

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Entomology

ajor ProTe_ssor’s Signature

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

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

 

APR 16 2007

 

0403 (l?

 

""060 3 08‘

 

MgY7qj% 2ng

 

 

 

 

 

 

 

 

 

 

 

2/05 c:/CTRCIDateDm.lndd—p.15

POTENTIAL FOR BEHAVIORAL AND CULTURAL MANIPULATION OF
JAPANESE BEETLE (POPILLIA JAPONICA NEWMAN, COLEOPTERA:
SCARABAEIDAE) POPULATIONS
By

Zsofia Szendrei

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Entomology

2005

ABSTRACT
POTENTIAL FOR BEHAVIORAL AND CULTURAL MANIPULATION OF
JAPANESE BEETLE (POPILLIA JAPONICA NEWMAN, COLEOPTERA:
SCARABAEIDAE) POPULATIONS
By
Zsoﬁa Szendrei

The Japanese beetle, Papillia japom'ca Newman (Coleoptera: Scarabaeidae), is an
invasive insect with potential for range expansion into many of the major agricultural
production regiOns of the world. Popillia japom'ca is currently the most important insect
pest for producers of hi ghbush blueberry, Vaccinium corymbosum L., in the Midwestern
United States. At fifteen commercial hi ghbush blueberry farms in Michigan, ﬁelds with
tilled row-middles had 72% fewer P. japonica larvae than ﬁelds with grassy row-
middles. Papillia japonica larval populations were similar in the perimeters of tilled and
untilled ﬁelds. Adult beetles on bushes were significantly less abundant in tilled ﬁelds
compared to ﬁelds with grassy row-middles. The abundance of larvae inside fields during
the spring was significantly correlated with early, but not late summer abundance of adult
beetles on bushes. Grassy plots between rows of blueberry bushes tilled in spring and
kept bare all year had fewer P. japonica larvae than plots retaining perennial ryegrass.
Tillage in the spring and in the autumn caused 51% reduction in 2002 and 69% in 2003.
Over three years, signiﬁcantly fewer adult Japanese beetles were found on clover,
ryegrass or bare ground than on buckwheat, whereas larval density from natural
infestation in these plots in the fall was signiﬁcantly lower under buckwheat and bare

ground than under ryegrass and clover. Significantly fewer eggs that were artificially

placed into the soil in these plots developed into larvae under clover and bare ground,
than under ryegrass and buckwheat. Female P. japonica observed in the field spend more
time digging on ryegrass than buckwheat or clover. These ﬁeld results indicate
differential acceptance of females when choosing an oviposition site. In laboratory
behavioral bioassays, females laid more eggs in soil with artiﬁcial grass stems compared
to bare ground, and preferred for higher stem density and diameter, but did not respond to
different stem colors. These ﬁndings suggest the possibility for reducing P. japonica
populations by changing the vegetation characteristics of blueberry ﬁelds. In a survey,
Michigan blueberry growers were asked to rate the extent togvhich P. japonica is a pest
within this crop, and to reveal their responses to this pest management challenge. This
insect was of concern to the majority of growers that returned the survey (84%), causing
additional costs per acre of $72. Increased insecticide use was the major cause of the .
economic loss. Insecticides and clean cultivation were the main method for controlling
Japanese beetle. Japanese beetle has driven changes in row-middle management,
indicated by growers who have switched to clean cultivation recently. Many growers
were willing to try new cover crops, if they are shown to be effective against Japanese

beetle.

ACKNOWLEDGEMENTS

I would like to thank the members of my Ph.D. committee: Rufus Isaacs, Eric
Hanson, James Miller, and David Smitley for their guidance and relentless attention
during my studies and research at MSU. I would like to thank all the people who helped
me in the ﬁeld experiments: Tracy Anderson, Carolyn Klunzinger, Zoltan Horvath, Kelly
Bahns, Keith Mason, Natalia Botero-Garcés, and Rodrigo Mercader for assistance in
collecting and processing soil samples. Anna Wright, Lydia Langdon, Rozlyn Fulgoni
and Henry Milliner III are thanked for help in lab experiments. I also thanks Dave Trinka
(MBG Marketing) for help with locating field sites, and to TNRC farm management for
plot maintenance. This work would not have been possible without the blueberry growers
who allowed me to use their farms. Thanks to Dave Trinka and Gretchen Sonntag from
MBG Marketing, and Bob Tritten from MSU Genesee County Extension for mailing the
surveys to their growers. Thanks to all the Michigan blueberry growers who responded to
the survey. Funding for this research was made available in part by Project GREEEN,
North Central Region USDA-SARE, MBG Marketing, Michigan Agricultural
Experiment Station, and the USDA Crops at Risk program. Funding was also provided
by the Hutson Endowment (Department of Entomology, MSU). I would like to thank

MSU for providing several travel grants for international travel.

iv

TABLE OF CONTENTS

LIST OF TABLES .............................................................................. ix
LIST OF FIGURES .............................................................................. x
CHAPTER 1: Biology, ecology and management of the Japanese beetle ................ 1
Introduction ............................................................................................ 1
History and distribution of the Japanese beetle .......................................... 1
Biology of the Japanese beetle ............................................................ 3
Integrated pest management of the Japanese beetle ..................................... 7
Objectives ................................................................................... 16

CHAPTER 2: Effect of tillage on abundance of Japanese beetle larvae and adults in

highbush blueberry ﬁelds ........................................................................ 18
Introduction .......................................................................................... 1 8
Materials and methods ............................................................................. 20
Larval density in commercial fields ...................................................... 20
Soil analysis ................................................................................. 22
Relationship between larval and adult populations .................................... 22
Effect of tillage on larval establishment ................................................. 23
Effect of tillage timing in infested plots ................................................. 24
Results ................................................................................................. 26
Larval density in commercial ﬁelds ...................................................... 26

Soil analysis ................................................................................. 27

Relationship between larval and adult populations .................................... 29
Effect of tillage on larval establishment ................................................. 32
Effect of tillage timing in infested plots ................................................. 32
Discussion ............................................................................................ 35

CHAPTER 3: Cover crops in highbush blueberry inﬂuence Japanese beetle density

and oviposition behavior ......................................................................... 39
Introduction .......................................................................................... 39
Materials and methods .............................................................................. 42
Effect of cover crops on adult P. japonica ............................................. 43
Larval density under different cover crops ............................................. 43
Survival of P. japonica larvae under cover crops ...................................... 44
Effect of cover crOps on P. japonica behavior ......................................... 45
Data analysis ................................................................................ 47
Results ................................................................................................ 47
Effect of cover crops on adult P. japonica ............................................. 47
Larval density under different cover crops ............................................. 49
Survival of P. japonica larvae under cover crops ...................................... 49
Effect of cover crops on P. japonica behavior ......................................... 50
Discussion ............................................................................................ 54

vi

CHAPTER 4: Do plant cues inﬂuence the oviposition behavior of Japanese beetles?

......................................................................................................... 58
Introduction .......................................................................................... 58
Materials and methods .............................................................................. 60
Oviposition arena ........................................................................... 6O
Insects ....................................................................................... 62
Soil and plant material ..................................................................... 63
Bioassays .................................................................................... 63
Assessment of oviposition ................................................................ 66
Statistical analysis of bioassay data ...................................................... 66
Behavioral observation .................................................................... 67
Results ................................................................................................ 68
Bioassays .................................................................................... 68
Behavioral observation .................................................................... 75
Discussion ............................................................................................ 77
CHAPTER 5: Survey of Japanese beetle management practices in Michigan
blueberry ............................................................................................. 82
Introduction .......................................................................................... 82
Materials and methods .............................................................................. 83
Results and discussion .............................................................................. 84
Conclusion ........................................................................................... 99

vii

CHAPTER 6: Conclusions and future directions ......................................... 101

REFERENCES CITED ......................................................................... 106
APPENDICES .................................................................................... 1 15
Appendix 1: Survey of Japanese beetle management methods ........................ 116
Appendix 2: UCI-IRIS approval of the survey ............................................ 122
Appendix 3: Record of deposition of voucher specimens .............................. 123
Appendix 3.1: Voucher specimen data .................................................... 124

viii

LIST OF TABLES

Table 2.1. Effect of tillage timing on P. japonica larval density in plots within a
blueberry planting. Comparisons of larval densities before and after tillage are
represented by x2 values (NS = not signiﬁcant). Percent control values show average
reduction in larval density by treatment. Values in the last column are not significantly
different from each other within years (a=0.05) ................................................ 34

Table 3.1. Description of behavioral elements of P. japonica observed on four row-
middle treatments in highbush blueberry ......................................................... 46

Table 3.2. Comparison of P. japonica adult density in four different row-middle
treatments in hi ghbush blueberry. Samples were taken in the summer of 2002, 2003, and
2004. For each year, averages within a column followed by the same letter are not
significantly different (a = 0.05) .................................................................. 48

Table 3.3. Comparison of P. japonica larval density in four different row-middle
treatments in highbush blueberry. Samples were taken in the fall of 2002, 2003, and
2004. For each year, averages within a column followed by the same letter are not
signiﬁcantly different (a = 0.05) .................................................................. 49

Table 3.4. Popillia japonica larval survival in four row-middle treatments in hi ghbush
blueberry. Numbers are averages of larvae survived out of ﬁve ﬁrst instars, in the
summer of 2002, 2003, and 2004. For each year, averages within a column followed by
the same letter are not signiﬁcantly different (a = 0.05) ....................................... 50

Table 5.1. Advantages and disadvantages of cover crops and clean cultivation for
integration into blueberry production, and the percentage of survey respondents that

consider these important (in order of importance). Eighty-nine responses were given.
......................................................................................................... 97

ix

LIST OF FIGURES

FIGURES 1.1. AND 1.2 IN THIS THESIS ARE PRESENTED IN COLOR.

Figure 1.1. Distribution of the Japanese beetle in the US and Puerto Rico in 2004. (Anon.
2005) ................................................................................................... 2

Figure 1.2. Lifecycle of the Japanese beetle in Michigan ....................................... 4

Figure 2.1. Density of P. japonica larvae (average + SE) in samples taken from (a.) row-
middles and (b.) perimeters of commercial hi ghbush blueberry ﬁelds with row—middles
that were either tilled or grassy. An asterisk above the bars indicates a signiﬁcant
difference between tillage treatments (a=0.05). Numbers on the lower bars show the
number of farms in each treatment during each date for (a.) and (b.) ........................ 28

Figure 2.2. Average number (+ SE) of adult P. japonica per blueberry bush in July and
August in ﬁelds with grass (N=3) or tillage in the row-middles (N=3). Values are
presented for samples taken from (a) the field interior and (b) the field perimeter.
Adjacent bars with an asterisk are signiﬁcantly different (a=0.05) ........................... 30

Figure 2.3. Relationships between P. japonica larval density measured in the spring and
abundance of adults on blueberry bushes measured in July (a.) and August (b.) in the
interior and perimeter of ﬁelds. White data points are from tilled ﬁelds and black data
points are from fields with grassy row-middles. Circles indicate average values from
interior and squares from perimeter samples ..................................................... 31

Figure 2.4. Density of P. japonica larvae (average + SE) in tilled (bare ground) or
ryegrass row-middles. An asterisk above the bars demotes a significant difference
between treatments (a=0.05) ...................................................................... 33

Figure 3.1. Popillia japonica behavior on different cover crops (n: number of
observations within a treatment) in a blueberry ﬁeld in 2002 and 2003. Within each
behavioral element, bars with the same letter are not signiﬁcantly different (a = 0.05).

......................................................................................................... 53
Figure 4.1. Bioassay arena for measurement of oviposition decisions by P. japonica.
......................................................................................................... 61
Figure 4.2. Average number (+ SEM) of eggs per soil block testing the effect of grass
cues on ovipositional behavior (Experiments 1-3). Drawings below graphs indicate
respective treatments. Bars within a graph with the same letter above them are not
signiﬁcantly different (a= 0.05). ‘N’ represents the number of replicates in each
experiment where one beetle was used per replicate ............................................ 69

Figure 4.3. Relationships between the proportion of soil blocks receiving eggs and
average number of eggs per soil block, for different soil block treatments in Experiments
1-3 ..................................................................................................... 70

Figure 4.4. Average number (+ SEM) of eggs per soil block testing artificial grass cues
on ovipositional behavior (Experiments 4-6). Drawings below graphs indicate respective
treatments. Bars within a graph with the same letter above them are not significantly
different (a= 0.05). ‘N’ represents number of replicates in the experiment ................. 73

Figure 4.5. Relationship between the proportion of soil blocks receiving eggs and
average number of eggs per soil block, for different soil block treatments in Experiments
4-6 ..................................................................................................... 74

Figure 4.6. Frequency and duration of digging behavior by P. japonica in bioassay
arenas, as a function of treatment. Average number of digging events per day (a) and
average duration of digging bouts (b) on the four treatments. Bars in a graph with the
same letter are not signiﬁcantly different (a= 0.05) ............................................ 76

Figure 5.1. Map of Michigan’s lower peninsula with the number of responses (a.) and
median farm size (acres) (b.) per county, in 16 counties in the southern tier of Michigan.
Names of Michigan counties in the main blueberry growing region are indicated on the
left of the map (a.) ................................................................................... 85

Figure 5.2. Percent of respondents that ranked Japanese beetle as the most important pest
in blueberry production, by county ................................................................ 88

Figure 5.3. Estimated economic loss per acre from Japanese beetle at blueberry farms in
SW Michigan counties .............................................................................. 88

Figure 5.4. Causes of the economic impact due to Japanese beetle .......................... 90

Figure 5.5. Growers' responses to the question "What is the most important Japanese
beetle control method?” ............................................................................ 90

Figure 5.6. Number of farms with different types of row-middle management (top). Total
acreage of farms with different row-middle management (bottom) .......................... 92

Figure 5.7. Responses to the question "Would you be interested in using new ground
covers if they were effective against JB?“ ....................................................... 94

Figure 5.8. Relationship between farms size and the importance of Japanese beetles as
pest in blueberry production. Importance of the pest was on a scale from 1 (not
important) to 5 (extremely important); results shown in the graph are ranks 4 and 5 only.
......................................................................................................... 95

xi

CHAPTER 1:

BIOLOGY, ECOLOGY AND MANAGEMENT OF THE JAPANESE BEETLE

INTRODUCTION
History and distribution of the Japanese beetle
The Japanese beetle (Popillia japonica Newman, Coleoptera: Scarabaeidae) is endemic
to the main islands of the Japanese Archipelago. Today, this insect is established in
eastern North America and on Terceira Island (Azores, Portugal), where it escaped from
a United States air base in the early 19705 (Lacey et a1. 1994). The Japanese beetle is
currently established in all states East of the Mississippi River, except for Florida, and
parts of Wisconsin, Minnesota, Iowa, and Nebraska (Fig. 1.1). The beetle is capable of
dispersing up to 12 km/year, and has the potential to spread further into southern Canada,
and some of the western parts of the United States (Allsopp 1996). To prevent the further
spread of Japanese beetle, the United States Department of Agriculture (USDA) has
established the Japanese beetle Harmonization Plan that restricts interstate shipment of
nursery stock with soil and movement of airplanes from regulated airports. The western
states protected by the Harmonization Plan are: Arizona, California, Idaho, Nevada,
Oregon, Utah, and Washington, with Montana in the process of being added to that list
(Anon. 2003).

This exotic pest was brought from Japan into southern New Jersey in 1916 with a
shipment of nursery materials (Fleming 1972). The beetle entered the United States

without its natural enemies, and started spreading west as it found favorable climate,

large areas of turf for the development of grubs and almost 300 species of plants for the
adults to feed on (Fleming 1972).

Soon after it was identiﬁed in the US, its potential threat to agriculture was
recognized and the USDA Japanese beetle Laboratory was established in 1917 to learn
more about the biology of this insect. As soon as the ﬁrst Japanese beetle appeared in
Michigan in 1932 in Detroit, the USDA started eradication and control programs. Local
eradication efforts brought success initially (Anon. 1959), but the lack of quarantine and

the presence of the beetle in surrounding states made it clear that permanent eradication

    

Infestation status
I Established
- Not established
- Being eradicated
E] Eradicated

D Not found

Figure 1.1. Distribution of the Japanese beetle in the US in 2004. (Anon. 2005)

was impossible. Beside the beetle’s natural spread westwards, it was also accidentally
transported by car, airplane and other means of human transport to new areas in the state.
In the 19608, area-wide eradication programs were organized to eliminate infestations,
and treatments included soil insecticide against the larvae and aerial applications of
insecticide against the adults (Anon. 1962, Anon. 1963). However, these efforts have
failed to eliminate this pest. Recently, the Japanese beetle has become the most important
production problem for numerous plant commodities in Michigan, including hi ghbush
blueberry.

Michigan leads the nation in the production of highbush blueberry (Vaccinium
corymbosum L.); it produces over 40% of the US crop. However, the ﬁnancial security of
the Michigan blueberry industry has been threatened by the recent invasion of Japanese
beetle into it’s major production centers. The zero threshold for beetles (zero insects in
fruit sampled by the buyer) by processors or the fresh market demands superlative pest
control by the grower, and has led to increased insecticide applications close to harvest. It
is clear from the recent contamination problems experienced by some blueberry growers
that chemical control of adults is not wholly effective. This one-dimensional approach is
also environmentally damaging, due to the toxicity of many broad-spectrum insecticides

to non-target organisms and their persistence.

Biology of the Japanese beetle
Lifecycle and reproduction. The Japanese beetle is univoltine throughout most of its
range in the United States, but in parts of its northern range it can take two years to

develop (Vittum 1986). Adult emergence begins in early July in Michigan. Some

researchers consider the Japanese beetle protandrous (Fleming 1972, Régnier et al.
1981a, Vittum 1986), but in a recent study female emergence was skewed signiﬁcantly
earlier than that of males (Van Timmeren et al. 2000).

Emerging females carry an average of 20 eggs and begin to oviposit approximately
two weeks aﬁer emerging and mating (Fig. 1.2). Females alternate between periods of
feeding and egglaying; and they typically enter the soil 10 or more times and depositing
40-60 eggs over a 4-6 week adult life span (Fleming 1972). Eggs hatch in about 10-14
days, and development of the ﬁrst and second instars requires about 2-3 weeks and 3-4

weeks respectively (Fleming 1972).

 

 

Figure 1.2. Lifecycle of the Japanese beetle in Michigan.

Most grubs are third instars by late September in Michigan. Young larvae are more
sensitive, particularly to desiccation, whereas older larvae are less susceptible to
environmental stress, partly because of their ability to move downward in the soil

(Régniere et al. 1981b, Allsopp et al. 1992). The larvae overwinter below the frost line

and start moving upwards in the spring once soil temperatures rise above 10°C (Fleming
1972). The larvae feed for another 4-8 weeks, and then move down in the soil to pupate.
The prepupal stage lasts approximately ten days and the pupal stage lasts 7-17 days
(Fleming 1972). Newly emerged adults remain in their earthen cells for 2-14 days before
emerging (Fleming 1972).

Virgin females emit a sex pheromone ((Z)-5-(1-decenyl)dihydro-2(3H)-furanone),
which is highly attractive to males (Tumlinson et al. 1977). The sex pheromone is
produced in female-speciﬁc epithelial cells that line the inner surface of anal plates and
two apical stemites. These cells are connected through many pores to the cuticule surface
(Tada and Leal 1997). The sex pheromone has an identical chemical formula to the sex
pheromone of the Osaka beetle (Anomala osaka). These two species live in the same
geographic area in Japan, but they use opposite enantiomeric forms of the chiral
pheromone (Leal 1998). Male antennae of the Japanese beetle possess olfactory receptor
neurons within the same sensilla placodea, one tuned to the detection of the sex
pheromone (R-japonilure), and the other to the behavioral antagonist, (S-japonilure)
(Wojtasek et al. 1998, Nikonov and Leal 2002).

Virgin females are typically inseminated as they emerge from the soil (Ladd 1970).
Japanese beetles are both polygynous and polyandrynous (Barrows and Gordh 1978), and
sperm from the last mating is used to fertilize the eggs (Ladd 1966). Males guard females
and experiments have conﬁrmed sperm displacement (Ladd 1966), which is likely the
driving cause of guarding behavior. Repeated mating is unnecessary for the sustained

production of fertile eggs, therefore polyandry is likely to have evolved to increase fitness

(Ladd 1966, Ladd 1987). The food type and the age of females affect both their fecundity

and their fertility (Goonewardene 1969, Ladd 1970).

Feeding habits. The adult Japanese beetle is highly polyphagous; it feeds on the leaves,
flowers and fruits of over 300 plant species of wild and cultivated plants in 79 families
(Fleming 1972). Host-ﬁnding is said to be mainly mediated by olfaction in adult P.
japonica (Ahmad 1982), and both sexes are strongly attracted to blends of plant volatiles
released from beetle-damaged leaves (Loughrin et al. 1995, Loughrin et al. 1996,
Loughrin et al. 1997).

The stimulatory and inhibitory effects that plant chemicals, either primary or
secondary, exert on the host-plant selection behavior of herbivorous insects counteract
each other and their balance determines the possible outcome of the decision-making
process: rejection or variable degrees of acceptance (Miller and Strickler 1984). In
polyphagous species, several ubiquitous primary plant metabolites are sufﬁcient to
stimulate feeding on many plant species and rejection occurs only in those plants that
produce deterrents of such quality or in such quantity that feeding stimulation is inhibited
(Schoonhoven et al. 1998). Japanese beetles accept a range of host plants regardless of
their suitability, thus it is likely that their decisions to accept a food source are based on
non-speciﬁc plant cues like the ones that occur at the leaf surface, i.e. ubiquitous plant
sugars (sucrose, maltose, fructose and glucose), that are phagostimulants for this species
(Ladd 1986). Cucurbitacins (T allamy et al. 1997), neriifolin (Reed et a1. 1982), and

phenolic glycosides (Orians et al. 1997) are examples of secondary metabolites produced

by plants that deter feeding by Japanese beetle. The lactiferous secretion of milkweed has
also been found to be a feeding deterrent for this species (Doussard and Eisner 1987).
Japanese beetle larvae are polyphagous, but because of their limited mobility they
tend to be facultative monophages (Potter and Held 2002). The larvae feed on roots of
grasses, weeds, garden and nursery crops (Fleming 1972, Potter et al. 1992, Crutchfield

and Potter 1995, Smitley 1996).

Integrated pest management of the Japanese beetle

Integrated pest management programs emphasize ecosystem-based strategies that provide
economical long-term solutions to pest problems (Flint and Gouveia 2001). Insecticide
treatments alone may not always be economically optimal or ecologically acceptable.
Integrated pest management practices may include managing water and fertilizer,
adjusting cultivation techniques, the use of mating disruption, or disruption of the pest
habitat (Flint and Gouveia 2001). Strategies that encourage natural enemies can also help
limit pest numbers. With growing concern for the environment, methods of pest control
have become more reﬁned and require a precise understanding of the target insect’s
biology, physiology, behavior, distribution, etc. The shift in management practices from
relying solely on insecticides towards integrated pest control necessitates the combination
of different methods that complement each other in minimizing harmful effects on the

environment and economic loss to growers (Matthews 1984).

Sampling populations. As an insect threatens a crop valued by humans, it becomes a

target for control. Efﬁcient control measures require an understanding of the presence

and spatio-temporal distribution of the targeted pest population. Detection of the presence
of Japanese beetle became a primary focus in early eradication programs (Fleming 1976),
and the investigation for attractants began as early as 1919 (Fleming 1972). The female
pheromone was identified in 1977 (Tumlinson et al. 1977), and effective attractants were
developed in 1980 (Ladd and McGovern 1980). Male and female beetles can now be
monitored by deploying traps baited with ﬂoral odor lures containing phenylethyl
propionate: eugenol: geraniol (3:7:3) and sex pheromone lures ((R,Z)-5-(Dec-l-
enyl)dihydro-2(3H)furanone) (T urnlinson et al. 1977).

Assessment of Japanese beetle grub populations requires soil sampling with a spade
or a golfcup cutter. Larval populations are typically aggregated, so sequential sampling

plans have been proposed in turf systems (Dalthorp et al. 1999).

Chemical control. Management of Japanese beetle in agricultural ﬁelds is complex
because the adults are mobile and, in many cases, targeting one life stage may not
provide sufﬁcient control. There is a lack of information on the relationship between
larval and adult abundance in a given area, making it hard to estimate the connections
between controlling grubs and reduction in the number of subsequent adult beetles.
Insecticides are widely used in turf and home lawns against Japanese beetle larvae, where
toxicity of the residues in food products is not an issue, and where larval control is the
primary concern. The soil-applied chloronicotinyl, imidacloprid, and the thianicotinyl,
thiarnethoxam, are effective insecticides against white grubs when applied to the soil as
preventative treatments during or immediately after egg laying (Grewal et al. 2001). They

have also been effective in preventing the downward movement of the overwintering

grubs at the onset of winter, thus affecting their survival (Grewal et al. 2001). The recent
registration of the soil-applied formulation of imidacloprid to control P. japonica larvae
in blueberry provides a new alternative for growers to reduce pest pressure in and around
their ﬁelds, especially in highly infested areas.

Insecticide applications against adults continue to be the foundation of management
strategies for Japanese beetle in blueberries (Z. Szendrei, unpublished; Chapter 5) and
many other fruit crops, as growers strive to meet the market’s demand for contamination-
free fruit. Broad-spectrum insecticides are effective against beetles that are treated
directly (Isaacs et al. 2004), but many of the effective insecticides have long pre-harvest
intervals, creating a potential for immigrating beetles to re-infest ﬁelds as residue activity
declines. Use of a short-lived insecticide with rapid ‘knock-down’ activity, such as a
pyrethrum (Casida and Quistad 1995), may be an effective approach for removing live
beetles from bushes immediately before harvest. Extracted from Chrysanthemum ﬂowers,
these botanical insecticides immediately paralyze treated insects and cause beetles to
drop to the ground. However, the expense, environmental impact, and food safety
concerns related to insecticide treated fruit mandate the search for non-chemical

approaches.

Biological control. When the Japanese beetle was introduced into the United States,
biological control agents of this species were limited to generalist predators such as birds,
toads, mammals, and insects (Fleming 1968). These failed to reduce populations or limit
the spread of the beetle. To enhance population control by natural enemies, predator and

parasite species were imported from Japan starting as early as the 19203, and reared for

releasing into areas with known Japanese beetle infestation (Fleming 1968). Out of the 49
imported species, only a few became established. The most widely distributed were:
Tiphia vernalis, a wasp that parasitizes overwintering grubs; Tiphia popilliavora, which
attacks young grubs in late summer; and Istocheta aldrichi, a tachinid fly that parasitizes
newly emerged adults (Fleming 1968). None of these species has been found in Michigan
(Cappaert and Smitley 2002).

Japanese beetles with entomophagous nematodes were first found in New Jersey in
1929 (Fleming 1968). Some like Steinemema glaseri and Heterorhabditis bacteriophora
can be effective biocontrol agents of Japanese beetle in turf and nursery stock (Potter and
Held 2002). New species of nematodes, such as Steinemema kushidai (Koppenhdfer et
al. 2000a) and Steinemema scarabaei (Koppenhdfer and Fuzy 2003), which are relatively
speciﬁc to scarab larvae, may prove to be more effective. There is a synergistic
interaction between certain species of nematodes (Steinemema glaseri and
Heterorhabditis bacteriophora) and imidacloprid, where the action of the insecticide
decreases the activity of the grubs, thus exposing them to increased nematode attack
(Koppenhbfer et al. 2000b). Despite their potential, entomopathogenic nematodes were
found to be uncommon in recent surveys of P. japonica larvae in Michigan (Cappaert and
Smitley 2002).

Milky disease, a bacterial pathogen, was ﬁrst discovered in central New Jersey in
1933, when a few abnormally white grubs were found in the ﬁeld (Fleming 1968).
Programs to disseminate Paenibacillus spores were conducted in the 1940s and although
this effort suppressed Japanese beetle populations in treated areas, milky disease

remained a factor that contributed to population suppression but was not by itself an

10

effective microbial insecticide (Fleming 1968). Bacterial pathogens infected less than 1%
of the Japanese beetle larvae in Michigan (Cappaert and Smitley 2002).

Pathogenicity tests with different species of fungi found attacking white grubs have
been conducted since 1921 (Fleming 1968). Metarhizium anisopliae prevalence in natural
Japanese beetle populations is low, and in Michigan and Connecticut this pathogen was
detected in a very small percentage (<0.1%) of the samples (Cappaert and Smitley 2002).

Ovavesicula popilliae, a microsporidian, was shown not to affect larval weight or
development of Japanese beetle in the laboratory. However, under ﬁeld conditions, adult
emergence was delayed and female fecundity was reduced at high infection rates (Hanula
1990). This pathogen was present in P. japonica grubs collected from a few locations in
Michigan (Cappaert and Smitley 2002). Generally, gregarines have been considered weak
pathogens having little affect on their hosts (in Cappaert and Smitley 2002). However,
few gregarines have been studied carefully enough to determine the effect of infection on

their hosts.

Behavioral and cultural pest control. Insect populations may be reduced by
manipulation of the target insect’s behavior. Stimuli that either inhibit or change the
pestilential behavior are used, and for the protection of a crop it is necessary to
understand the behavior (or aspect of behavior), the method by which it can be
successfully manipulated, and the implementation of this method (Foster and Harris
1997). The basis by which the behavior for manipulation is chosen is usually driven by

the available methods, and is often not linked to the type of damage caused.

11

Mechanism of cultural pest control. An important aspect of behavioral manipulation of
plant pests is the process of resource acquisition by the insect in the context of the insect-
plant interaction (Opp and Prokopy in Miller and Miller 1986). In the case of the
Japanese beetle, plants provide essential resources for feeding, mating, egg-laying and
refugia. As the insect through its environment, it meets a multitude of stimuli, mainly
visual and chemical plant cues (Schoonhoven et al. 1998). Olfaction is a key factor in
host ﬁnding by Japanese beetle (Ahmad 1982, Loughrin et al. 1995, Loughrin et al.
1997). Insects receive information from the environment through their peripheral sensory
system, which is then processed to shape a response. Thus, there are two major phases
that can be the target of behavioral manipulation: changing the inputs from the
environment, or, altering the processing of these inputs (Foster and Harris 1997). Insect
behavior is usually manipulated by changing the normal environmental stimuli leading to
pest interaction with a crop. The stimulus chosen for behavioral manipulation has to be
accessible to the insect, well deﬁned for reproducibility, easily controlled, and practical
to apply (Foster and Harris 1997). The more speciﬁc a stimulus is to a behavior, the more
precisely the particular behavior can be controlled, and although precise deﬁnition of a
stimulus allows for better control of a behavior, it is not essential.

Natural objects such as plants are frequently used to manipulate behavior to reduce
pest pressure (Prokopy 1994, Bugg and Waddington 1994, Hartwi g and Ammon 2002),
and in this approach, the stimulus is generally not deﬁned precisely. Behavioral
manipulation of an insect population within an agricultural field can be achieved through
increasing the species diversity and changing the composition of the plant community

within the field by for example planting cover crops. Vegetation designs and other habitat

12

manipulation practices such as cover crop management, crop row spacing, or trap-
cropping have been extensively studied in annual crop systems (in Aluja et al. 1997), and
less well studied in perennial crops, because in these systems it is more difficult to
experimentally test the impact of orchard design (Bugg and Waddington 1994). Risch et
a1. (1983) reviewed 150 studies of insect populations in intercropped systems and found
that of 198 insect species reported in these studies, 53% had lower populations in the
intercrops, 18% had higher populations, 9% showed no signiﬁcant difference, and 20%
had a variable response. A meta-analysis by Tonhasca and Byme (1994) found lower
herbivore densities in polycultures in 60-70% of the studies. Andow (1991) indicated that
insect pest densities were lower in polycultres in 52% of the examined cases when
compared to monocultures.

The two most common mechanisms proposed to explain lower densities of insect
populations in intercropped systems are the “Natural enemies” and the “Resource
concentration hypothesis” (Root 1973). The “Natural enemy hypothesis” stipulates that,
due to increased diversity of the ecosystem, there is enhanced natural enemy abundance
and heightened rates of predation and parasitism, leading to pest suppression. Beneficial
insects often require pollen and nectar as adult food sources, which are easier to ﬁnd in a
varied plant composition. According to the “Resource concentration hypothesis” it will
become difficult for an insects to ﬁnd host plants in a diverse system and therefore they
will be more likely to leave. This is particularly important for controlling a specialist
herbivore, where the cues from the plants inﬂuence colonization rate and behavior.
Vanderrneer (1989) has proposed the “Trap crop hypothesis” in addition to the “Resource

concentration hypothesis” to explain how vegetation diversity can affect herbivore

13

abundance. This supposes that pests will be attracted by certain plants, and thus will be

more attracted to trap crops and will be less likely to move into the crop to be protected.

Approaches to cultural control. Cultural controls are the modification of normal crop or
landscape management practices to decrease pest establishment, reproduction, dispersal,
and survival (Flint and Gouveia 2001). Cultural practices include some of the oldest pest
management tactics used, often exploiting weak links in the pest’s life cycle or behavior.
Habitat modiﬁcation intentionally changes the managed ecosystem to limit availability of
one or more of the requirements for survival and reproduction, making the environment
less suitable for pest populations and improving survival and effectiveness of the pest’s
natural enemies (Flint and Gouveia 2001). Using cover crops can have long term
economic benefits, by increasing soil fertility, suppressing weeds, preventing soil erosion,

conserving soil moisture, and reducing the need for herbicides (Flint and Gouveia 2001).

Cultural controls and Japanese beetle. In the case of Japanese beetle the mechanisms of
ovipositional decision-making are not clear. Japanese beetle adults are extremely
polyphagous, and host selection may be influenced by relatively nonspeciﬁc stimuli
common to many different hosts (Foster and Howard 1998). Compared to food-searching
behavior, fewer critical cues may be involved in choosing an ideal egglaying site, since
the larvae are facultative monophages. Because of limited larval mobility, the mother’s
decision is crucial for hatching and survival of the offspring. In the literature, oviposition
preference and speciﬁcity is most often studied on species that lay eggs on above-ground

plant parts (Wiklund 1981, Foster and Howard 1998, Bossart and Scriber 1999, Showler

14

2001). Since Japanese beetle lays eggs into the soil, qualities of the soil and the plants
that grow in the soil may be correlated with cues that drive female ovipositional choice.
In laboratory assays, females responded differentially to soil qualities such as texture,
moisture, and organic matter before ovipositing (Allsopp et al. 1992). The incorporation
of Metarhizium anisopliae (a fungal pathogen for the biological control of scarab grubs)
into the soil increased oviposition (Villani et al. 1994), suggesting that female beetles
could detect the pathogen. Data from these laboratory experiments reveal an important
aspect of Japanese beetle searching behavior: namely that the female makes post-
alighting decisions on where to lay eggs based on soil-bom cues. Currently, no
information is available on how the evaluation of the environment takes place by the
insect and to what extent plant cues play a role in female ovipositional choice.
Circumstantial evidence for female choice is available via the distribution of grubs. On
this basis females prefer wet, sandy soils and low—mown grass, in sunny locations
(Fleming 1972, Potter et al. 1996). They do not lay eggs into bare soil under natural
conditions (Fleming 1972), but will readily do so in a conﬁned arena (Allsopp et al. 1992,
Van Timmeren et al. 2000). Based on these findings, it could be possible to use ground
covers as natural barriers to avert ovipositing females from an area that provides ideal
soil conditions for oviposition and larval development.

A suite of approaches is required by growers that can be integrated into their farming
operations for reducing the likelihood of beetle infestation, fruit contamination, as well as
to create a suppressive environment for Japanese beetle. Grubs potentially provide a
better target for control than adults, because they stay in the soil for eight months. During

most of this time growers are not actively involved in fruit production, thus grub control

15

would be time saving, because there would be reduced need for Japanese beetle adult
control during the harvest season. Grub control can also be preventive, because the
beetles are eliminated as grubs, so they never emerge to cause further damage. In
highbush blueberry production, on the other hand, the adult stage is responsible for the
economic loss, since the grubs feed only on the roots of grasses and weeds, but not of
blueberries. Thus the relationship between the density and location of grubs and adults
needs to be established before effective grub control measures can be recommended or
implemented.

Because of the zero-threshold for beetle contaminants in harvested blueberries, the
goal for the growers should be to reduce the presence of both adults and larvae of this
insect in their ﬁelds. The main reasons for Michigan blueberry growers’ interest in
planting cover crops in their ﬁelds are weed suppression, reduction of soil erosion, and to
permit driving machinery through the field during wet conditions. Since Japanese beetle
is currently the most important insect pest in blueberries in Michigan (Z. Szendrei,
unpublished, Chapter 5), there is a need for testing cover crop species that satisfy
grower’s horticultural needs and do not attract Japanese beetle to their ﬁelds. In practice,
the naturally occurring weed population is kept under control by mowing or different
species of grasses are seeded in the row-middles. Blueberries are grown in low-lying
areas with high soil moisture and ﬁelds are irrigated in the summer. The wet soil
conditions and the vegetation in and around the blueberry fields create an ideal condition
for Japanese beetle oviposition. The blueberry industry provides a model system for
understanding how cultural and behavioral manipulation of an insect can be achieved for

control in an integrated management program.

16

Objectives

The overall goal of this research was to evaluate behavioral control methods for reducing
Japanese beetle adult and larval populations within blueberry ﬁelds. Ideally, these
methods should be economically viable, practical, and environmentally harmless. In
order to investigate this broad question, I undertook a set of investigations into the
ecology and behavior of the Japanese beetle. The first objective was to determine the
distribution of larvae within and around commercial blueberry ﬁelds, and the effect of
soil tillage on Japanese beetle larval density. Since adult Japanese beetles are responsible
for the economic loss in highbush blueberry production, the effectiveness of larval
control for the protection of blueberries is dependent on whether there is a relationship
between the adult and larval life stages within a ﬁeld. If larval control translates into
reduction of adult abundance within the ﬁeld, then other options for larval control, such
as rotovation and planting cover crops in row-middles that are not ideal for larval
development and survival, could be considered for reducing adult Japanese beetle
abundance. The second objective was to test if cover crops have the potential to reduce
oviposition and larval and adult abundance within a blueberry ﬁeld. If this approach were
successful in reducing oviposition within ﬁelds, this would decrease the subsequent larval
abundance, and the number of adults that emerge from the ﬁeld. The third objective
aimed to understand the underlying behavioral mechanisms that lead to the selection of
grassy areas. This knowledge would be indispensable for developing a pest-control
strategy that growers could use reliably and which would be based on this behavioral

feature of the insect.

17

My fourth objective was to conduct a survey to quantify management tactics that growers
have adopted to control Japanese beetle and to determine willingness of growers to
incorporate alternative management strategies, such as the use of cover crOps, in their

future management programs if they are proved effective against Japanese beetle.

18

CHAPTER 2:
EFFECT OF TILLAGE ON ABUNDANCE OF JAPANESE BEETLE LARVAE

AND ADULTS IN HIGHBUSH BLUEBERRY FIELDS

INTRODUCTION

The Japanese beetle, Popillia japonica Newman (Coleoptera: Scarabaeidae), is an
invasive insect with potential for range expansion into many of the major agricultural
production regions of the world (Allsopp 1996). Popillia japonica is currently the most
important pest for producers of hi ghbush blueberry, Vaccinium corymbosum L., in the
Midwestern United States (Anon. 2001, Isaacs et al. 2004). It is a univoltine insect; adults
are most abundant from mid-July to August (Fleming 1972), when beetles feed on leaf
tissue and ripe blueberry fruits. If beetles are not controlled, fruit can be contaminated
when harvesting machines knock beetles off the bushes. Over 70% of Michigan’s 7,300
ha of blueberry production is harvested mechanically (Kleweno and Matthews 2002), and
because the market demands fruit completely free of insect contamination, strategies are
needed to minimize the risk of adult beetles being present during harvest. Foliar
insecticide applications continue to be the foundation of P. japonica management in
blueberries and many other fruit crops, as growers strive to meet exacting quality
standards. Additional strategies targeting larvae, which develop in the soil, may help
growers reduce populations of P. japonica within infested blueberry production regions
and minimize the risk of beetles spreading into uninfested areas. Such strategies could
also have the long-term beneﬁt of reducing the number of foliar applications of

insecticides.

l9

Larval insect populations are inﬂuenced by initial adult distribution and density,
oviposition preference and host plant acceptability, egg survival, and host plant suitability
(Singer 1986, Renwick 1989). Conditions in and around crop ﬁelds can favor or inhibit
development of P. japonica depending on whether requirements for population
development are met (Vittum et al. 1999). Within many Michigan blueberry ﬁelds and
along the perimeter of these ﬁelds, ground covers of seeded grass or the naturally-
invading mix of grass and broadleaved weeds are commonly used to maintain soil
structure, provide conditions where agricultural machinery can be driven during wet
conditions, reduce soil erosion, and prevent pesticide and fertilizer runoff (Pritts and
Hancock 1992). Fields are often irrigated to minimize crop stress and to maintain fruit
quality and yield during periods of insufﬁcient rainfall. Together, these factors provide
ideal habitat for oviposition, egg hatch, growth, and survival of P. japonica larvae
(Régniére et al. 1981, Allsopp et al. 1992, Potter et al. 1996).

Cultural practices such as tillage can have strong negative effects on arthropod
pests, by modifying the soil habitat where many insects reside during at least a part of
their life cycle (Funderburk et al. 1990). Immature stages of P. japonica feed on the roots
of grasses, so mechanical manipulation of their habitat may alter their survival and
development (Fleming 1976). In 1955 Cory and Langford (in Fleming 1976) reported
that rototilling destroyed up to 90% of the P. japonica grubs. Fleming (1976 and
references therein) reported that reduction of larval population in fallow and sodlands
was greatest when the soil was rototilled in early autumn and late spring, when grubs are
near the soil surface. Previous studies have also investigated the potential for control of

white grubs through some cultural practices (mowing height, irrigation, application of

20

organic fertilizer, aeriﬁcation, heavy rolling) that are appropriate in turf systems (Potter et
al. 1996). Tonhasca and Stinner (1991) explored the effect of tillage on adult P. japonica
abundance in soybean fields, but there is a lack of information on cultural manipulation
of P. japonica in perennial fruit crops and woody omarnentals.

To determine the response of P. japonica to tillage in the row-middles of
hi ghbush blueberry ﬁelds, we compared the relative abundance of larvae at ﬁfteen
commercial blueberry ﬁelds in row-middles that were either tilled or covered with
mowed permanent sod comprised of seeded grass and local weed populations. The
relationship between larval density and some soil parameters was examined to determine
whether variation in grub abundance is explained by soil characteristics. Selected
parameters describe conditions for nutrient uptake in the soil for the plants that serve as
food sources for the grubs. Thus, these factors could have an indirect effect on
herbivorous insect abundance. At six commercial blueberry ﬁelds (three tilled and three
grassy), larval and adult densities were compared at the perimeter and interior of the
ﬁelds. Spring larval populations were compared to subsequent summer adult populations,
to determine whether there is a relationship between the two life stages within blueberry
ﬁelds. In addition to these population studies, the effects of (1) the presence or absence of
a grass ground cover and (2) different tillage timings on larval abundance of P. japonica
were investigated experimentally over two growing seasons in a research planting of

highbush blueberry.

21

MATERIALS AND METHODS

Larval density in commercial ﬁelds

The density of larval P. japonica populations was determined across the primary
blueberry production region of Southwest Michigan (Ottawa, Allegan, Van Buren, and
Berrien counties) in 2001 and 2002. Fifteen ﬁelds in 13 commercial blueberry farms with
a history of P. japonica infestation were sampled. The sizes of the ﬁelds ranged from 1 to
5 ha.

Larval density was determined in April 2001, September 2001, May 2002, and
September 2002 by taking 15 cm deep soil cores from the perimeter of each ﬁeld and
from the area between the rows of blueberry bushes (row-middles) of each ﬁeld. This
was done using a cylindrical golf cup cutter (area = 95 cm2) (Parmenter & Andre Inc.,
Grand Rapids, MI). In each ﬁeld, 80 samples were taken approximately 1 m from the
edge of the blueberry ﬁeld (in the drive lane around the ﬁeld), by sampling in 20
positions spread out along each of the four sides of the ﬁeld. Ten row-middles were
selected that were spaced evenly across the ﬁelds, but more than 5 m away from the ﬁeld
border. In each selected row, six samples were taken equidistant between rows of bushes
along the length of the field without sampling within 5 m from the row ends (60 interior
samples / ﬁeld). Soil cores were examined in the ﬁeld, and all beetle larvae were placed
in plastic bags with a small amount of soil. The bags were labelled with date, location,
and number of larvae and transported back to the laboratory in a cooler containing an
icepack. Larvae were identiﬁed to species using the diagnostic rastral patterns and other

morphological features (Vittum et al. 1999).

22

For each ﬁeld and sampling date, whether the ﬁelds were tilled or had grassy row-
middles was recorded. Because row-middles were tilled based on growers’ management
decisions, the number of ﬁelds that had tilled or grassy row-middles varied by sampling
date. Field perimeters were not tilled and had a grass-weed mix cover throughout the
study, regardless of the type of row-middle management within the ﬁeld. Foliar
insecticides were applied to the bushes according to standard practices (Wise et al. 2003)
at all of the farms.

The average numbers of P. japonica larvae in samples taken in the row-middles
and in the perimeter of the ﬁfteen ﬁelds were compared between tilled and non-tilled
ﬁelds. Data were transformed to \/(y+0.5), before analysis to meet normality
assumptions, and were analysed with a two-way ANOVA (tillage and date as factors)
(SAS Institute 2001). T-tests were used at each sample date to compare samples from
tilled and untilled fields. In cases where the population variance of the two compared
groups were different, the Satterthwaite’s approximation was used for calculating degrees

of freedom for the t-tests (SAS Institute 2001).

Soil analysis

Soil cores (area 95 cm2, 15 cm deep) were taken for soil quality analysis from all 15
ﬁelds in autumn 2002. One sample was taken from each of the four sides of each ﬁeld,
and four samples were taken from row-middles, distributed across the ﬁeld. Each sample,
was analysed for: pH, organic matter (%), and the concentration of P, K, Mg, and Ca (all
in ppm). Samples were analyzed by A&L Great Lakes Laboratories, Inc. (Fort Wayne,

IN). Multiple regression was used to determine the relationship between the six measured

23

soil parameters and average larval densities at the 15 ﬁelds in autumn 2002 (SAS
Institute, 1999). Samples taken from the interiors and perimeters of the ﬁelds were

analysed separately.

Relationship between larval and adult populations

In 2003, ﬁelds from the study described above that had consistent row-middle
management history during 2001-2002 were selected. This group was comprised of three
tilled ﬁelds and three with grass in their row-middles. Larval populations of P. japonica
were sampled at all six ﬁelds in May using the methods described above. Eighty samples
were taken from the interior and 80 from the perimeter of ﬁelds. Perimeter samples were
taken approximately 1 m away from the bushes. Twenty samples were taken evenly on
each of the four sides of the ﬁeld. From the row-middles of ﬁelds, 8 samples were taken

from 10 row-middles in a similar manner as described in the previous larval study.

The number of adult P. japonica per bush was sampled once in July, and once in
August at each ﬁeld by cOunting the number of adult beetles on 80 bushes on the
perimeter (20 on each side of the ﬁeld) and on 80 inside (10 bushes from 8 rows). All
inside samples were taken at least 5 m from the edges of the ﬁeld and the 10 sampling
points were evenly spread out through the rows. Adult beetles were counted on sunny
days with low wind velocity, and observers moved carefully through the ﬁeld so as not to

disturb beetles.

The abundance of adult P. japonica in July and in August was compared between

grassy and tilled fields with the Mann-Whitney U test (SAS Institute 2001). Multiple

24

regression analysis was used to determine relationships between larval and adult
populations of P. japonica within fields (SAS Institute 2001). The regression model
consisted of larval and adult numbers as variables and position as an indicator variable
(perimeter = 0, interior = 1). Statistical analyses were performed on average numbers of
adults per bush and average numbers of larvae per sample and were considered

signiﬁcant at a=0.05.

Effect of tillage on larval establishment

This experiment was conducted in a four-year-old 0.4 ha field of V. corymbosum cv.
Bluecrop, at the Trevor Nichols Research Complex (TNRC) in Fennville, MI. The
planting was established two years prior to the start of this experiment, on a 3.6 x 1.2 m
spacing with 12 bushes in each row. The row-middles were kept clean of plants. Popillia
japonica larvae were not found in soil samples taken before the experiment in April,
2002. Ten row-middles (14 x 2.5 m area between two rows of bushes) were tilled in May
10, 2002 and assigned to be grass or bare ground treatments. Treatments were assigned
within the planting in a randomised complete block design with ﬁve replicates. Grass
plots received perennial ryegrass seed at 27 kg/ha (Lolium perenne L., Michigan State
Seed Solutions, Grand Ledge, M1) on May 23, 2002. Plots were rolled after planting to
improve germination. This grass species was chosen because of it tolerates soil acidity.
Bare-ground plots received further treatments of herbicide (Roundup Ultra 4WSL at
31.3ml/1) spot-applied with a 3.8 l hand-held sprayer when needed to keep the soil free of
vegetation during the growing seasons. Fertilizers and pesticides were not used in the

plots during this experiment. Samples were taken in September 2002, as well as May and

25

October 2003 to assess larval density of P. japonica. Five soil cores were taken with a
golf cup cutter (area = 95 cm2) from each plot, and collected larvae were identified to
species as described above.

Larval densities in the two treatments were compared for each sample date with a
Mann-Whitney U test, since normality assumptions were not met even after

transformation (SAS Institute 2001; Ott and Longnecker 2001).

Effect of tillage timing in infested plots

This study was conducted in two 0.07 ha sections of a 1 ha field of V. corymbosum cv.
Rubel, at TNRC. The blueberry plants were 10 years old on a 3.6 x 1.2 m plant spacing,
with 12 bushes in each row. The 14 x 2.5 m row-middles between the rows of blueberry
were comprised of naturally occurring weeds dominated by grass species, and were kept
mowed. The experiment was conducted in 21 adjacent row-middles in 2002 and was
repeated in 2003 in 21 adjacent row-middles located in a different section of the same
blueberry ﬁeld. These two sections of the ﬁeld were selected based on prior sampling to
verify the presence of P. japonica larvae across the ﬁeld sections. The following three
treatments were applied during both years to the different sections: tillage once in the
spring, tillage once in the autumn, or tillage once in the spring and once in the autumn.
Treatments were assigned to the row-middles in a completely randomised design with
seven replicates. To apply tillage treatments, 3 tractor-powered BushHog rotovator
(Model H72, Allied Products Corporation, Selma, Alabama) was driven between the bush
rows twice, at 15 cm depth. Plots tilled in the spring were kept weed-free in the growing

season by applying herbicide (as described above). Tillage treatments were applied on

26

May 10 and September 9 in 2002 and on May 2 and Oct 10 in 2003. All plots were
sampled with a golf cup cutter as described above in the spring and the autumn, before
and after tillage treatments. Popillia japonica larvae were sampled 2-4 (1 before and 10-
14 d after tillage. Three samples were taken per row-middle in 2002 and six in 2003, and
collected larvae were identified to species.

Larval count data were Poisson distributed, so pre- and post-tillage densities were
compared using a log-linear model (PROC GENMOD, SAS Institute 2001). Comparisons
of the percent reduction in P. japonica density between treatments were made after
arcsine transformation with a one-way ANOVA procedure followed by Tukey’s means

separation (SAS Institute 2001).

RESULTS

Larval density in commercial ﬁelds

In spring 2001, the average density of P. japonica larvae across all commercial blueberry
farms sampled was 0.28 i 0.01 larvae/sample. This increased slightly in autumn 2001
and then decreased at each sampling time to 0.14 i 0.01 larvae/sample by autumn 2002.
Larval density was greater on the perimeter than in the row-middles of ﬁelds throughout
this study, regardless of whether the row-middles were tilled. During the two sampling
seasons in 2001, larval densities at the perimeter were approximately 4-fold greater than
those in the interior, and by spring 2002 there was a 3-fold difference between these
positions which decreased to 0.4-fold in autumn 2002. The highest average larval density
in the perimeter of ﬁelds occurred in autumn 2001 (0.48 :f: 0.02 larva/sample) and the

lowest in autumn 2002 (0.17 i 0.01 larva/sample). Eighty-seven percent of all the larvae

27

were found in ﬁelds with grassy row-middles compared to those that were cultivated; the
largest difference between the two treatments occurred in spring 2001 (0.14
grubs/sample) and the smallest difference occurred in autumn 2001 (0.09 grubs/sample)
(Fig. 2.1). Larval density was signiﬁcantly lower in tilled fields than in those with grassy
row-middles in spring 2001 (t = -3.41, P < 0.01) and autumn 2001 (t = -2.94, P < 0.01)
(Fig. 2.1a). At the other two sampling times, the differences between the two types of
row-middle management were not statistically signiﬁcant, even though on average there
were twice as many larvae in grassy row-middles compared to tilled ones in autumn 2001
and 13 times more in spring 2002. Some growers changed row-middle management
practices during the study, so between 6 and 13 of the 15 ﬁelds sampled had a grass
ground cover, depending on the date of the sample (Fig. 2.1).

On average, larval density was 12 times greater in the perimeter than in the row-
middles of tilled fields, whereas density was only 2.4 times greater at the perimeter of
grassy ﬁelds (Fig. 2.1 a and b). Across the whole study, 78.5% of the larvae detected in
soil samples were found in the perimeter of the ﬁelds; the highest value was 85% in
autumn 2001, and the lowest value was 65% in autumn 2002. Popillia japonica larval
density at ﬁeld perimeters was not signiﬁcantly different between ﬁelds with grassy or
tilled interiors (F 13: 3.44, NS); but, in all of the samples, a greater density of larvae was
found in the perimeter of grassy ﬁelds (Fig. 2.1 b).

In addition to P. japonica, larvae of European chafer (Rhizotrogus majalis
Razoumowsky, Coleoptera: Scarabaeidae) and Asiatic garden beetle (Maldera castanea
Arrow, Coleoptera: Scarabaeidae) were also found in the samples, but in very low

numbers.

28

Soil analysis

Average organic matter content was similar in samples taken from the row-
middles and from ﬁeld perimeters, at 3.3 :l: 0.1% and 3.5 i 0.2%, respectively. The
largest difference between positions was found for calcium, which was 668 ppm in the
perimeter on average, compared to 558 ppm in the row-middles, but this difference was
not statistically signiﬁcant. The average concentrations of potassium and magnesium ions
were not signiﬁcantly different between the perimeter and the row-nriddles. There was no
signiﬁcant relationship between the measured soil parameters and larval densities, either
for the soil samples taken in the row-middles (F63: 0.57, R2 = 0.3, NS) or in the

perimeters (F63: 2.85, R2 = 0.68, NS).

29

0.3-

[:| tilled (a)
grassy
*

0.2r

    
 

Average larvae per sample (+ SE)

 

ilil

'spring 2002
autumn 2001 autumn 2002

Figure 2.1. Density of P. japonica larvae (average + SE) in samples
taken from (a) row-middles and (b) perimeters of commercial hi ghbush
blueberry ﬁelds with row-middles that were either tilled or grassy. An
asterisk above the bars indicates a signiﬁcant difference between tillage
treatments (a=0.05). Numbers on the lower bars show the number of
farms in each treatment during each date for (a) and (b).

30

Relationship between larval and adult populations

Adult beetles were less abundant in July (0.88 i 0.41 beetles/bush) than in August (1.68
i 0.59 beetles/bush). At both sampling times, greater numbers of beetles were detected
on bushes in blueberry fields with permanent sod than in tilled fields (Fig. 2.2). This
pattern was consistent across the positions of the sampled ﬁelds; but statistically
signiﬁcant differences in average adult beetle numbers were only detected in the
perimeters (Z = -1.96, n = 3, P = 0.049 for both sample times). Fields that had greater
larval abundance in the spring generally had more adult beetles on bushes in July,
regardless of the position in the ﬁeld (Fig. 2.3). Regression analyses showed that larval
populations measured in the spring predicted 84% of the variability in July adult
populations (F29: 25.19, P < 0.01). In contrast to the July adult samples, beetle

abundance in August was not correlated with earlier larval abundance (F29 = 0.60, NS)

(Fig. 2.3).

31

5 I |:| filled (a)
- grassy

A

 

Average adults per bush (+ SE)

 

 

July August

Figure 2.2. Average number (+ SE) of adult P. japonica per
blueberry bush in July and August in ﬁelds with grass (N=3) or
tillage in the row-middles (N=3). Values are presented for
samples taken from (a) the ﬁeld interior and (b) the ﬁeld
perimeter. Adjacent bars with an asterisk are signiﬁcantly
different (a=0.05).

32

Average adults per blueberry bush

y = 3.85x - 0.01 (a)
R2: 0.84

 

 

 

 

 

(b)

5‘ . I y=12.63x+1.30
4- R=oo5
3 q I
2 ‘//

_ El
‘cpo 1:1 '
o O . . . .

0 0.2 0.4 0.6 0.8 1

Average larvae per sample

Figure 2.3. Relationships between P. japonica larval density
measured in the spring and abundance of adults on blueberry
bushes measured in July (a.) and August (b.) in the interior
and perimeter of fields. White data points are from tilled
fields and black data points are from ﬁelds with grassy row-
middles. Circles indicate average values from interior and
squares from perimeter samples.

33

Effect of tillage on larval establishment

Densities of P. japonica larvae increased during this experiment; the greatest increase
occurred in plots that were planted with ryegrass (Fig. 2.4). Larvae in ryegrass plots were
three times more abundant in autumn 2003 than in the previous autumn (Z = -3.6, n = 10,
P < 0.01). During each of the three sampling dates, significantly fewer larvae were found
in plots kept free of plants compared to those with ryegrass. The largest difference

between the two treatments in the average number of larvae was in autumn 2003 (Z = -

3.8, n =10, P < 0.01).

Effect of tillage timing in infested plots

The change in P. japonica larval abundance due to tillage in the spring was not consistent
over the two years; this treatment was associated with a signiﬁcant decline in larval
density in 2003, but not in 2002 (Table 2.1). Tillage in the autumn caused a decrease in
P. japonica during both years when comparing pre- and post-tillage larval densities, but
this trend was significant only in 2003 (Table 2.1). Tillage of the same plot twice, in the
spring and in the autumn, significantly reduced larval density in both years, when
comparing overall reduction from spring to autumn; by 51 :0.- 12 % in 2002 and 69 i 15 %
in 2003. The effect of tilling plots once was not consistent across the two years: although
most of the treatments caused a similar reduction in P. japonica larval density in the row-
middles, in each year one of the treatments did not provide any control. The percent
control values (Table 2.1) were not significantly different in either year (F330 = 0.65, P =

N3 in 2002; F320 = 0.13, P = NS in 2003).

34

Average larvae per sample (+SE)

0.7—

0.5—

0.3—

0.1—

 

 

 

 

tilled

ryegrass

 

 

0.0

autumn 2002

 

 

 

 

 

 

 

l

 

spring 2003

35

E—muih,

autumn 2003

Figure 2.4. Density of P. japonica larvae (average + SE) in tilled
(bare ground) or ryegrass row-middles. An asterisk above the bars
demotes a signiﬁcant difference between treatments (a=0.05).

 

 

 

one mz so Renee Sumo cease .
Suﬁ adv 5 Eng no“: were were
Same adv 2: 23¢ no.4: 5:33 5:23
EH8 adv a; Shoo 33.. men... macaw

88
anew mz em :3 .o Round saga -
Suﬁ m2 3 dong 386 were macs
once“ - - oduod Ndﬁmd 5:28 5558

one m2 8 mouse 3.4.8 mace... were

«8....

.md H .838 Q» an ax ewe—=32:— owﬂabum wag: Eamon.—
mm H 2953 8a 23.3— owauo>< «38am owe—Fm.

 

.Amoduev B8» 55:3 850 88
Bob Engage 32onch Ho: one 5:28 “2.: 2: E 82m> 82585 ,3 58:8 REE E 8888 ems—oi» 255
82? 35:8 “cookomdcmocEwE Ho: u mZV 82? Nx 3 83889: 8w owe—E 8% 28 883 823% RE—

Lo mnemngoO .wEEmE b82255 a £53, 829 5 58:8 Eta— eomzomex .& :o wEE: owe—E mo Hootm A.” 033—.

36

DISCUSSION

This study demonstrates that tillage can reduce larval P. japonica abundance in
highbush blueberry fields. At commercial farms, tilled row-middles had lower larval and
adult density of P. japonica than row-middles with grass cover. Smitley (1996) found
lower densities of P. japonica larvae and adults in nurseries with clean ﬁelds compared to
weedy ones, and this pattern is expected to be a general response of P. japonica to the
lack of suitable hosts, leading to lower rates of oviposition and/or larval survival.

Tillage of the row-middles did not signiﬁcantly reduce P. japonica larval
populations in the uncultivated perimeters. Therefore additional control strategies, such
as targeting early instars with soil-applied insecticides (Mannion et al. 2001) could be
used to control P. japonica in regions of the farm that are not tilled. As the results
depicted in Figure 2.4 suggest, the removal of grassy groundcover should reduce the
number of beetles emerging in the ﬁeld, providing fewer gravid females to move to the
perimeter for oviposition. The higher larval density of P. japonica in the perimeters of
ﬁelds, independent of the ground cover management regime, could be because gravid
females arriving from outside the ﬁeld are arrested at the ﬁeld perimeters, and/or because
females may leave insecticide-treated areas inside the field to lay eggs in the untreated

soil around the ﬁelds.

The signiﬁcant relationship between larval populations in the soil during the
spring and adult populations on bushes in July (Fig. 2.3) suggests that spring larval
sampling can be used to estimate local adult populations. This may allow growers to
predict the risk of P. japonica adult presence across their farm and provide information to

focus management inputs to ﬁelds with the greatest spring larval populations. Smitley

37

(1996) found a positive correlation between the number of P. japonica adults trapped and
the number of larvae in nursery fields. Choo et al. (2002) found a similar positive
relationship between adult and larval populations of oriental beetle (Exomala orientalis
Waterhouse) in golf course fairways in Korea; adults were more abundant in fairways
that had high larval densities. In their study of P. japonica spatial dispersion in golf
courses, Dalthorp et al. (2000) suggested that grub dispersion patterns are dependent on
location of adult feeding aggregations. Popillia japonica emerge from the soil in early
July in Michigan, and the newly emerged adults may initially feed and aggregate close to
their larval habitat if there is suitable food nearby. This may explain the greater
predictability of adults in July compared to August. After 4-6 weeks of adult emergence
and movement between fields, spring larval density no longer predicted the number of
adult beetles on bushes.

The reduction of adult P. japonica densities by tillage may result fewer larvae
emerging from these ﬁelds, but it could also be the case that fewer adult beetles may be
attracted to blueberry ﬁelds with bare soil in the row-middles. In the small plot trials,
tillage in row-middles was effective in preventing the establishment of P. japonica larval
populations (Fig. 2.4). Tilling row-middles both in the spring and autumn reduced larval
populations by 59%. The mechanism of larval population reduction through rotovation
needs to be determined in the future.

The lack of relationship between soil parameters and larval density indicates that
these factors do not affect the suitability of the environment in a blueberry ﬁeld for P.
japonica. Previous ﬁndings by Dalthorp et al. (2000) in turf found no relationship

between mean P. japonica larval density and soil organic matter, but perennially low

38

grub densities were associated with high organic matter content. Potter et al. (1996) have
shown that the response of larval populations to spatial gradients in soil moisture may
override the effects caused by soil pH, making soil micronutrient parameters of low value

for predicting risk from P. japonica.

Targeting management tactics to the parts of the field or farm with the highest P.
japonica larval population may be an effective and efﬁcient way to help manage the adult
pest populations. There are several advantages to controlling larvae in addition to adults.
First, larval P. japonica are present for a longer period than adults, so there may be more
opportunities for control. Second, larvae are not capable of moving large distances, so it
would be easier to target controls against them, compared to the mobile adults. A recently
available option for controlling P. japonica larvae in blueberry ﬁelds is the use of
imidacloprid (Bayer CropSciences, Kansas City, MO) applied to the soil during the egg
laying period (Mannion et al. 2001). Biological control methods are also being
investigated, but parasites and pathogens of P. japonica are currently at low levels and
cause only minor mortality in Michigan populations (Cappaert & Smitley 2002).

The presence of a groundcover, and the type of plant, are important in inﬂuencing
larval P. japonica populations in turf (Potter et al. 1992, Crutchﬁeld & Potter 1995), as
well as in nursery (Smitley 1996) and Christmas tree plantations (Kard & Hain 1988). In
sites where tillage is not appropriate, cover crops may be an alternative for management
of P. japonica. Cover crops that are tolerant of the acidic soil conditions of blueberry
ﬁelds are also being evaluated currently for their potential in reducing P. japonica

populations.

39

CHAPTER 3:
COVER CROPS IN HIGHBUSH BLUEBERRY INFLUENCE JAPANESE

BEETLE DENSITY AND OVIPOSITION BEHAVIOR

INTRODUCTION

Cultural control of insect populations in agricultural fields can be achieved by planting
cover crops that affect pest dispersal, colonization, or reproduction (Bugg and
Waddington 1994). A cover crop is any living groundcover that is planted in the main
crop (Hartwi g and Ammon 2002) for its positive horticultural effects, such as weed
control, reduction of soil erosion, improvement of the biological, physical and chemical
properties of the soil, or pest control. Increasing the diversity of plants in agricultural
production can reduce pest abundance (Root 1973), but the mechanisms are dependent on
the biology and behavior of the insect species. In many different agricultural systems,
insect behavior is inﬂuenced by cover crops planted with a main crop, such that the pest
is repelled or reduced in abundance (Risch 1981, Tonhasca and Stinner 1991, Bugg 1992,
Bugg and Waddington 1994, Prokopy 1994, Hartwig and Ammon 2002). Despite its
polyphagous feeding behavior as an adult, the more selective oviposition behavior of
gravid females (Szendrei and Isaacs, in press, Chapter 4) may provide an opportunity for
behavioral manipulation of the Japanese beetle, Popillia japonica Newman (Coleoptera:
Scarabaeidae). Following the steps provided by Foster and Harris (1997), development of
such an approach will require an understanding of the behavior for manipulation, the
method by which it can be successfully manipulated, and the implementation of this

method for the protection of a commercial crop.

40

Selection of host plants for feeding by adult P. japonica may be inﬂuenced by
relatively nonspeciﬁc stimuli common to many different hosts (Foster and Howard 1998).
In contrast, the critical cues involved in choosing an ideal oviposition site are expected to
be more speciﬁc since the female beetle is presumingly searching for specific habitat
requirements. Insect pests are confronted with an array of potential hosts that vary in their
suitability for larval development, and so most insects show some form of egg-layin g
preference (J aenike 1978). The larvae of P. japonica are facultative monophages because
of their limited mobility, thus the mother’s decision is crucial in the hatching and survival
of the offspring. Female preference for ovipositional sites, and larval performance on
hosts chosen during egg laying by the adult are important in determining subsequent
population densities in a field. Egg laying preference might be exploited for behavioral

manipulation of P. japonica.

Popillia japonica is most likely to lay eggs into soil under monocotyledonous plants
(Fleming 1972). Soil parameters such as moisture, texture, and organic matter content
(Régnier et al. 1981, Allsopp et al. 1992), short plant cover (Hawley 1944 in Fleming
1972) and light intensity (Dalthorp et al. 1999) can inﬂuence the number of eggs laid by
P. japonica. In a comparison of larval abundance in nursery ﬁelds, P. japonica were 10-
fold more abundant in weedy fields compared to clean ﬁelds (Smitley 1996), and clean
cultivation of row-middles of highbush blueberry ﬁelds reduces the number of P.
japonica larvae in the soil by 72% compared to ﬁeld with permanent sod (Szendrei et al.,
in press, Chapter 2). These ﬁndings suggest that the presence of plants in the larval
habitat is essential for development of P. japonica. The adult females of this species are

also capable of evaluating plant cues on the soil surface to adjust oviposition investment

41

(Szendrei and Isaacs, in press, Chapter 4). Although clean cultivation between the rows
of blueberry plants can reduce P. japonica larval abundance, the negative horticultural
aspects of this management method such as wet ﬁelds and the potential for pesticide
runoff and erosion require development of alternative strategies to reduce egg laying
and/or larval survival by P. japonica. One such alternative is to plant ground covers in

blueberry row-middles that reduce the egg-laying and/or larval development of this pest.

Foliar insecticide applications continue to be the foundation of Japanese beetle
management in bluebenies and many other fruit crops, as growers strive to meet the
exacting quality standards and are looking for alternative solutions to reduce P. japonica
pressure in their ﬁelds. Because of the zero-threshold for beetle contaminants in
harvested blueberries, the growers’ goal is to reduce the presence of both adults and
larvae of this insect in their ﬁelds. The main reasons for Michigan blueberry growers’
interest in planting cover crops in their ﬁelds are: weed suppression, reduction of soil
erosion, and to drive machinery through the ﬁeld during wet conditions. Since P.
japonica is currently the most important pest in blueberries in Michigan, there is a need
for testing cover crop species that satisfy grower’s horticultural needs and do not attract
P. japonica to their ﬁelds. Ideally, cover crops could be used as a natural barrier to avert
ovipositing females from an area that otherwise provides ideal soil conditions for

oviposition and larval development.

In this study, two acid-tolerant ground covers were evaluated for their potential to
reduce the suitability of blueberry fields for P. japonica. Alsike clover (Trifolium
hybridum L.) and buckwheat (Fagopyrum esculentum Moench) were chosen because

their morphological and biological characteristics are different from the preferred

42

perennial ryegrass (Lolium perenne L.), and these plants are recommended in other
perennial fruit systems (Anon. 1992). These were compared to perennial ryegrass and
bare ground within a blueberry planting to determine their ability to manipulate: l)
attraction to adult female beetles, 2) female beetle behavior after landing on the ground,
3) abundance of P. japonica larvae in the soil, and 4) survival of P. japonica from eggs to

mature larvae.

MATERIALS AND METHODS

Experiments were conducted in a replicated ﬁeld experiment at the Trevor Nichols
Research Complex, Fennville, MI. The 4050 m2 highbush blueberry planting (Vaccinium
corymbosum var. Bluecrop) was established using two year old bushes in 2001, on a 3.7
m x 1.4 m plant spacing with 12 bushes in each row. Four row-middle treatments were
established with either perennial ryegrass, alsike clover, buckwheat, or bare ground
between the rows of bushes. Cover crops were seeded in the row-middles (53 m2 / row-
middle) in May and June 2002 in a completely randomized design. Buckwheat was
reseeded in the third week of June in 2003 and 2004. One plot consisted of a set of two
row—middles with the same treatment on either side of a row of blueberry bushes, and
treatments were replicated five times. The seeding rates were 27 kg/ha for perennial
ryegrass, 22 kg/ha for clover, and 67 kg/ha for buckwheat (Michigan State Seed
Solutions, Grand Ledge, MI). The bare ground plots received herbicide (Roundup Ultra
4WSL at 31.3ml/l), spot-applied with a 3.8 l hand-held sprayer when needed, to keep the
soil free of vegetation during the growing seasons. Overhead irrigation of all the row-

middles was started from the time of seeding, and continued weekly or as needed

43

throughout the growing season in each of the years of this study to create conditions
typical of a commercial blueberry ﬁeld. Ryegrass and clover plots were mowed
throughout the growing season as needed. Measurements of larval density, adult

abundance and individual adult beetle behavior were conducted in this site.

Effect of cover crops on adult P. japonica

One of the row-middles from adjacent pairs of the same cover treatment was selected at
random, and the numbers of P. japonica adults that naturally occurred in these plots were
counted weekly in July and August in 2002, 2003, and 2004. Observations were made
from 12 - 2 PM on sunny days with low winds. Each visual survey of a row-middle lasted
two rrrinutes, and was measured with a timer. The observer walked down one side of the
row-middle and back on the other side, focusing on the area in the row-middle and
counting the number of beetles. The observer walked on the periphery of the area covered
by the crops, wore neutrally colored clothes, and moved evenly and slowly so as to
minimize disturbance of the beetles. At the end of the two minutes the total number of

beetles per row-middle was recorded on a data sheet.

Larval density under different cover crops

The naturally-established population of P. japonica was surveyed by taking soil samples
in May 2002, before the cover crop seeds were sown in the plots. Subsequently, larval
density was sampled in October 2002, September 2003, and October 2004 in the same
plots. Larvae were sampled by taking five 15 cm-deep soil cores from each row-middle

using a cylindrical golf cup cutter (area = 95 cm2) (Parmenter & Andre Inc., Grand

Rapids, MI). Soil cores were examined in the ﬁeld, and all beetle larvae were placed in
plastic bags with a small amount of soil. The bags were labeled with date, treatment, and
number of larvae. Bags were transported to the laboratory in a cooler containing an
icepack. Larvae were identiﬁed to species using the diagnostic rastral patterns and other

morphological features of the labrum (Vittum et al. 1999).

Survival of P. japonica larvae under cover crops

Plots for this experiment were established in a different part of the same blueberry field
as above, to create row-middles with bare ground, clover, buckwheat, or ryegrass in a
completely randomized design with ﬁve replicates (same seeding rates as above). To
obtain ﬁrst instar P. japonica, adults were collected from a grassy ﬁeld (East Lansing,
MI) with Japanese beetle Expando Traps baited with ﬂoral and sex lures (Great Lakes
IPM Inc., Vestaburg, MI). Traps were deployed on July 1, 2002, 2003, and 2004. Adults
were transported to the laboratory daily and approximately 500 (~50:50 sex ratio) were
put into each of ten 36 x 24 x 12 cm plastic containers. The lid of each container had a 10
x 15 cm hole for ventilation, which was covered with nylon mesh to stop beetles from
escaping. A 5 cm deep layer of moist sand (Dixie Cut Stone & Marble Inc., Bridgeport,
MI) was provided as an ovipositional medium in each container. Slices of apple were
placed on top of the sand as ad libitum food and to provide moisture. Containers were
kept at constant environmental conditions (16L:8D, 30% RH, 24-26 °C) and fresh slices
of apples were added as needed. The sand was sifted for ﬁrst instars by placing a handful
of sand in a sieve (#18 mesh size) and washing the sand with running tap water. Larvae

were transferred using a paintbrush (Loew-Comell®, #12) to moist ﬁlter paper (90 mm

45

diameter, Whatman International Ltd., Maidstone, UK) in Petri dishes (Becton Dickinson
and Co., Franklin Lakes, NJ) (10 x 1.5 cm) kept at room temperature in the laboratory
until further use in the ﬁeld, but not longer than 12 h.

First instars were transported from the laboratory to the field plots in a cooler.
Five PVC pipes (15 x 8 cm) were pounded into the soil in each row-middle, for a total of
25 pipes per treatment. A 10 cm-deep hole was made in the soil in the middle of the PVC
pipe with the end of the paintbrush; ﬁve larvae were carefully placed in the hole with the
brush. Then the hole was covered with soil and irrigated with 500 ml water to wash soil
particles close to the larvae and to provide moisture. After placement of larvae in the soil,
the PVC pipes were covered with mesh to prevent oviposition by naturally occurring P.
japonica. First instars were placed in the ﬁeld in the last week of July and PVC pipes
were removed in the ﬁrst week of September in 2002, 2003, and 2004. Soil was shaken
from the pipes onto a tray and examined for P. japonica larvae. The number of P.

japonica larvae per PVC pipe was recorded.

Effect of cover crops on P. japonica behavior

The behavior of individual adult female P. japonica was observed in row-middles planted
with clover, buckwheat, or ryegrass. Beetles were very rarely observed on bare ground
plots. Gender was distinguished by observing the shape of the front tibia (Fleming 1972)
without disrupting the insect’s natural behavior. Data were collected in July and August
2002 and 2003 weekly from 10 am — 4 pm on sunny days with low winds. A total of 35
observations were made in buckwheat and ryegrass and 39 in clover during the two years

of this study, and observations were recorded by the same observer throughout. The order

46

.-

of observations on the three cover crop treatments was randomized. An observation
began as soon as a beetle was located by the observer on one of the cover crops, and
lasted for a maximum of one hour or until the beetle left the plot. The observer wore
neutrally colored clothes, and moved evenly and slowly so as to minimize the disturbance
of the beetles. Continuous focal sampling, in which all of the behaviors of the subject are
recorded, was conducted using The Observer 3.0 (2002) and 4.0 (2003) (Noldus
Information Technologies, The Netherlands) loaded onto a laptop computer that was
carried by the observer. Behaviors that were observed and recorded are defined in Table

3.1.

Table 3.1. Description of behavioral elements of P. japonica observed on four row-
middle treatments in hi ghbush blueberry.

 

Behavior

 

Stationary Standing motionless

Walk Moving on a surface by means of leg motions

Groom Using legs or mouthparts to clean abdomen or antennae

Feed Moving mouthparts for the purpose of ingesting food

Dig Moving ﬁrst pair of legs to loosen the soil surface, associated with

oviposition behavior

Position
Crop At least two thirds of insect’s body is in contact with plant
Soil At least two thirds of insect’s body is in contact with soil surface

 

47

Data analysis

The natural larval and adult densities of P. japonica, and the larval survival values were
transformed to \ly+0.l to meet the assumptions of normality and homogenous variance.
Sub-samples from each row-middle taken for the natural larval infestation experiment
and larval survival experiment were averaged. Data for these three experiments were
analyzed separately for each year with a one-way AN OVA using PROC GLM (SAS
Institute 2001). Adult density data were analyzed with a two-way ANOVA (time and
treatment) using PROC GLM (SAS Institute 2001). Tukey’s test was used to determine
signiﬁcant differences between means at the 5% probability level. Average rate
(frequency of behavior/minute), average duration (min) of each behavior, and duration as
a percent of the total observed time (% observation) of behavioral elements on the three
cover crops were compared with a multivariate analysis of variance (MANOVA, SAS
Institute 2001) and contrasts of mean values were made to compare behavior of different
cover crops. The proportion of time spent on plant or soil surfaces was compared with a
two-way ANOVA (PROC GLM, SAS Institute 2001), with type of surface (plant vs. soil)

and cover crop treatments as factors.

RESULTS

Effect of cover crops on adult P. japonica

The number of adult P. japonica observed was signiﬁcantly different among the row-
middle treatments throughout this study (2002: F11“ = 11.29; 2003: F1153 = 47.09; 2004:
F11” = 6.95; P < 0.01 for all dates). The general trend in P. japonica density on the four

treatments in this study was, in decreasing order: buckwheat, clover, ryegrass, bare

48

ground. On average there were 7 times more beetles on buckwheat than on the other
planted row—middle treatments (Table 3.2). Bare ground had consistently fewer beetles
(0.06 beetles per sample) than the other row-middles with plants (2.24 beetles per
sample) over the three years (Table 3.2). In 2002 and 2003 there were significantly fewer
beetles on clover and ryegrass than on buckwheat (Table 3.2). In 2004, beetle abundance
on clover was not signiﬁcantly different from that on buckwheat, but ryegrass still had
signiﬁcantly fewer beetles than buckwheat. In 2002 and 2004, no adult beetles were
observed on bare ground, but during the higher overall population levels of 2003, a few
beetles were observed on bare ground at a density signiﬁcantly lower than the number
seen on all the other treatments (Table 3.2). There were signiﬁcantly more beetles on

clover than ryegrass in 2003 and 2004.

Larval density under different cover crops

The larval density of P. japonica varied signiﬁcantly among the four row-middle
treatments throughout the three years of this study (2002: F339 = 5.33; 2003: F339 =
45.52; 2004: F339 = 6.32; P < 0.01 for all dates). Larval density was highest on ryegrass
and clover in the three years, and the average larval densities were not statistically
different in any of the years between these two treatments (Table 3.3). Across the three
years, larvae were 3 times more abundant on average in ryegrass and clover than in
buckwheat. In 2002 and 2003 bare ground had signiﬁcantly fewer larvae than the other
treatments, while in 2004 the larval density in bare ground and buckwheat was

significantly lower than in the two other row-middle treatments.

49

Survival of P. japonica larvae under cover crops

Popillia japonica larval survival was signiﬁcantly different among the four row-middle
treatments in 2002 (F119 = 3.76, P = 0.03), but not in the other two years (Table 3.4).
There was high variability in larval survival within treatments, but survival was
numerically the highest in plots with ryegrass throughout this experiment. In 2002, a
similar number of larvae survived in clover and ryegrass, and buckwheat and bare ground
had signiﬁcantly fewer larvae per PVC pipe than ryegrass. This trend was not consistent
throughout the study; in 2003 there was no signiﬁcant response to row-middle treatments,
but the lowest larval survival was in clover and the highest in ryegrass and bare ground.
In 2004 ryegrass still had the highest numbers of surviving larvae, but in this year the
lowest values were recorded in buckwheat. Variation between years was the lowest when

larvae were placed in plots of ryegrass (Table 3.4).

50

Table 3.2. Comparison of P. japonica adult density in four different row—middle
treatments in highbush blueberry. Samples were taken in the summer of 2002, 2003, and
2004. For each year, averages within a column followed by the same letter are not
signiﬁcantly different (a = 0.05).

 

Avery-Le (1 SE) P. japonica adults/row-middle/2 min

 

 

treatment 2002 2003 2004 grand mean
Bare ground 0.00 a: 0.00 b 0.20 1 0.12 c 0.00 4.- 0.00 b 0.07 a 0.12
Buckwheat 2.07 i 0.39 a 11.45 i 1.78 a 0.67 i 0.27 a 4.70 i 2.44
Clover 0.87 a 0.47 b 2.50 a 0.53 b 0.37 a 0.18 b 1.30 a 1.20
Ryegrass 0.60 4.- 0.34 b 1.55 a 0.38 b 0.03 a 0.03 b 0.70 a 0.80

 

Table 3.3. Comparison of P. japonica larval density in four different row-middle
treatments in hi ghbush blueberry. Samples were taken in the fall of 2002, 2003, and
2004. For each year, averages within a column followed by the same letter are not
signiﬁcantly different (a = 0.05).

 

Average (:1: SE) P. japonica larvae/sample

 

 

treatment 2002 2003 2004 grand mean
Bare ground 0.00 i 0.00 c 0.10 :t 0.04 c 0.06 t 0.03 b 0.05 1: 0.05
Buckwheat 0.30 i 0.08 ab 0.64 i 0.14 b 0.00 t 0.00 b 0.31 :l: 0.22
Clover 0.88 t 0.31 ab 2.06 :t 0.19 a 0.22 i 0.07 a 1.05 i 0.57
Ryegrass 0.88 :1: 0.19 a 2.58 i 0.24 a 0.30 i 0.08 a 1.25 i 0.51

 

Table 3.4. Popillia japonica larval survival in four row-middle treatments in highbush
blueberry. Numbers are averages of larvae survived out of ﬁve ﬁrst instars, in the
summer of 2002, 2003, and 2004. For each year, averages within a column followed by
the same letter are not signiﬁcantly different (a = 0.05).

 

 

 

treatment Average (:1: SE) P. japonica larvae/PVC pipe
2002 2003 2004 grand mean
Bare ground 0.41 i 0.24 c 1.21 i 0.27 a 0.40 :t 0.17 a 0.67 :t 0.68
Buckwheat 0.40 i 0.09 bc 1.11 i 0.19 a 0.13 :1: 0.06 a 0.55 i 0.34
Clover 1.04 :l: 0.19 ab 0.46 1 0.13 a 0.20 i 0.09 a 0.56 i 0.40
Ryegrass 1.33 i 0.36 a 1.21 i 0.59 a 0.57 z 0.15 a 1.04 -_1- 1.10

 

51

Effect of cover crops on P. japonica behavior
Beetle behavior was affected by the different cover crops tested in blueberry. Beetles
were stationary at least twice as frequently on buckwheat than on the other two
treatments (F1110 = 10.47, P < 0.01) (Fig. 3.13). Beetles walked almost 50% less
frequently on clover than on the ryegrass or buckwheat (132,110 = 3.89, P = 0.02) (Fig.
3.1a), and fed signiﬁcantly more frequently on buckwheat and clover than on ryegrass
(F2,110 = 12.49, P < 0.01). The rate of digging was affected by the ground cover (F1110 =
3.95, P = 0.02), and was six times higher in ryegrass than in buckwheat. Ryegrass and
clover treatments were not signiﬁcantly different (F 1,1 10 = 2.56, P > 0.05) (Fig. 3.1a).
The average durations of each bout of stationary, walking and grooming
behaviors were not significantly different among the ground covers (F1110 = 0.07; F1110 =
0.71, F1110 = 0.01 respectively, P > 0.05) (Fig. 3.1b). In contrast, the durations of feeding
and digging behaviors were highly affected by ground covers, in opposite ways. Feeding
bouts lasted 1.65 i 0.73 and 1.87 i 0.45 minutes on buckwheat and clover respectively,
but P. japonica spent signiﬁcantly less time feeding on ryegrass (0.006 :t 0.006 min)
(F1110 = 11.60, P < 0.01). The total number of feeding behaviors throughout all the
observations was only 12 in the case of ryegrass, 63 for buckwheat and 112 for clover.
Digging bouts lasted an average of 1.32 i 0.44 minutes in ryegrass, signiﬁcantly longer
than in clover (0.52 i 0.24 min) or in buckwheat (0.05 i 0.05) (F1110 = 6.18, P < 0.01).
The proportion of time allocated by P. japonica females to particular behaviors
was also affected by ground cover (Fm 10 = 4.11, P = 0.02) (Fig. 3.1c). A signiﬁcantly
higher prOportion of their time was spent stationary on buckwheat (34.2 t 6 %) than on

clover (21.0 t 7 %), but this proportion was not significantly different between ryegrass

52

(25.6 t 5 %) and the other two treatments (Fig. 3.1c). A significantly greater proportion
of time was spent walking on ryegrass (36.3 :1: 6 %) than on either buckwheat or clover
(F1110 = 3.97, P = 0.02), but grooming was not affected by cover crop. The proportion of
time spent feeding was significantly different among the three cover crops (F2110 = 20.19,
P < 0.01). The greatest proportion of time spent feeding occurred on clover (47.8 i 9 %);
this was 2.5 times higher than the proportion of time spent feeding on buckwheat, and 15
times higher compared to ryegrass. The proportion of time spent digging was
signiﬁcantly different among treatments (Fm 10 = 4.26, P = 0.02); beetles spent most of
their time digging in ryegrass (16.4 i 7 %), and approximately 8% was spent digging in
clover, but this was not signiﬁcantly different from the other two treatments. Only one
beetle was observed digging (< 1% of time spent) in buckwheat.

On average, P. japonica spent signiﬁcantly more time on the plants than on the
soil on all of the treatments (F5220 = 51.58, P < 0.01). Beetles spent 99.4 1 0.5 % of their
time on buckwheat plants, 82.6 i 7 % on clover plants and 73.9 i 8 % on ryegrass. The
amount of time spent on clover and buckwheat was not statistically different (P = 0.21),

but less time was spent on ryegrass than on the other two treatments (P < 0.05).

53

Average rate + SE
(number/ min)

Average duration + SE (min)

Average duration + SE

1.6- a (A)
1.4‘

  
 
 

lilillil buckwheat, n=35

  

1.2 ,

1_ I ryegrass, n=35
0.8 E] clover, n=39
0.6 r
0.4 ,

0.2 ‘ 555

2.5 -

0.5 -

 

 

(% observation)
to 00 J:- UI 0) O
‘L 9 C? 3 9 .

_I.
.°

 

 

O

stationary walk groom feed dig

Figure 3.1. Popillia japonica behavior on different cover crops
(n: number of observations within a treatment) in a blueberry ﬁeld
in 2002 and 2003. Within each behavioral element, bars with the
same letter are not signiﬁcantly different (a = 0.05).

54

DISCUSSION

Cover crops can reduce pest damage in agricultural ﬁelds, by serving as a sink for pests,
or confusing pests visually or olfactorally and thus reducing colonization, or changing
microclimate within the field, thereby reducing pest success (Bugg 1992). In this study,
the adult and larval density and behavior of P. japonica were inﬂuenced by different
cover crops in row-middles of highbush blueberry ﬁelds. While P. japonica feeds on
many hosts, its choice for an oviposition site is more restricted; the results presented here
indicate that ryegrass is a preferred host for ovisposition and larval development, while
the other tested plants are less suitable or are not utilized for oviposition. This knowledge
can provide information when making management decisions about when and how to use
cover crops to reduce pest populations.

Larval P. japonica density was highest in ryegrass and clover, whereas adult
population density was lowest on ryegrass and highest on buckwheat throughout three
years. In a choice test conducted on potted plants in the greenhouse, P. japonica larvae
preferred to feed on roots of perennial ryegrass compared to five other turfgrass species
(Crutchﬁeld and Potter 1994). Previous studies have also found that survival varies with
plant species: third instar P. japonica survived better in a greenhouse study on large
crabgrass than on dandelion, white clover and Kentucky bluegrass (Crutchﬁeld and Potter
1995). Thus, selection of cover crop species can inﬂuence pest abundance.

Bare ground row-middles harbored the fewest larvae, indicating that P. japonica
density is increased by the presence of plants in the row-middles. This result is similar to
that of Crutchﬁeld and Potter (1995) who found in a greenhouse study that survival of

third instar P. japonica larvae was higher in the presence of plants than in no-plant

55

treatments. The inconsistency between years in P. japonica larval survival on the
different treatments could be due to abiotic factors, such as temperature differences
between the years. The results of this experiment are not consistent with the outcome of
the larval density study, in that naturally occurring larvae were not found in bare ground
plots, but ﬁrst instars survived and developed in PVC pipes in bare ground treatments.
This means that there is enough organic matter naturally in the soil of the experimental
ﬁeld to provide developing larvae with food. Inconsistencies between the results of the
larval density study and the larval survival experiment could also be because the PVC
pipes altered the microclimate, or that it prevented natural enemies from killing the
larvae.

Fewer adult beetles were found in bare ground plots than in plots with cover
crops, so it is likely that adult beetles avoid landing on brown (non-green) surfaces
(Kostal and Finch 1994); alternatively, tenure time on bare ground could be brief. This
would help to explain the lower abundance of larvae in bare ground plots. Among the
three cover crops, the highest numbers of adult beetles were found on buckwheat
throughout this study, indicating that this plant attracts beetles to land on them, making it
unsuitable for reducing adult populations in blueberry fields. Observations of female
beetle behavior under ﬁeld conditions provide insight into the mechanisms driving the
differences in beetle abundance and larval density in the three tested ground covers.
These are likely to be due to a combination of plant suitability for feeding and
oviposition. The observations revealed that although beetles spent more time and more
frequently fed on buckwheat and clover than on ryegrass, they dug more frequently and

for longer durations on ryegrass than on the other two treatments. Since P. japonica

56

adults fed only minimally on ryegrass, but laid eggs in this treatment, it is clear that host
plant evaluation for oviposition involves a different mechanism from food plant
evaluation. Duration and frequency data revealed the dominant behaviors on the different
plants: stationary on buckwheat, feeding on clover, walking and digging on ryegrass.

Popillia japonica utilize plants in different ways, and this ultimately translates
into variation in larval density under the different ground covers. The high larval density
on ryegrass is associated with a high proportion of digging behavior on this plant, and the
low larval density in buckwheat is due to the lack of digging behavior. Interestingly, the
rate and duration of digging behavior on clover was not statistically different from either
of the other two treatments (Fig. 3.1), but larval densities were similar to those in
ryegrass (Table 3.3). Clover is different from the other two plants in that it is a suitable
host for feeding, oviposition, and for larval survival and these qualities together will
likely make this plant unsuitable as a cover crop in blueberry fields. Buckwheat is not a
host for oviposition and is not ideal for larval survival, but adults used it for basking and
sometimes fed on the ﬂowers (Z. Szendrei, personal observations). Ryegrass is not
consumed by the adults, but is ideal for oviposition, and larval survival. These qualities
should be considered when selecting plants that have a potential to supress P. japonica
populations within agricultural ﬁelds.

The morphological differences among the tested plants may explain the behavior
patterns observed. Buckwheat is 50-60 cm tall during the beetle ﬂight period, whereas the
other two plants are 10—30 cm tall; so the taller plants could serve as a barrier separating
the beetles from the soil surface. Perennial ryegrass is a monocotyledonous plant with a

thin leaf blade, while the other two plants have broad leaves; this feature could be

57

signiﬁcant in inﬂuencing P. japonica behavior if the leaf shape is used as an orientation
or landing cue. Some of these morphological features could divert beetles from landing
on the cover crops or after landing, could avert them from approaching the soil surface
for digging. In a study with artiﬁcial grass blades, gravid females responded differently to
stem density and diameter, suggesting that P. japonica uses visual cues to evaluate
potential oviposition hosts (Szendrei and Isaacs, in press, Chapter 4). Visual cues have
been shown to be important factors in host orientation, location, and choice for many
species of Diptera, Lepidoptera, Hemiptera and Coleoptera (Prokopy and Owens 1983).
In P. japonica, vision has already been proven important during the development of an
effective monitoring trap (Fleming 1969). Olfactory and/or tactile cues may also be
involved in oviposition, but the chemical basis of oviposition decisions needs further
investigation.

Our small-plot trials were effective for measuring behavioral responses of P.
japonica to candidate cover crops, but larval and adult density studies should be repeated
at larger, commercial farm scales. Future studies should focus on adult density on
blueberry bushes and P. japonica damage to berries in farms that have different row-
middle treatments, to get a direct estimate of how cover crops affect P. japonica control
and of the economic value of having a certain species of cover crop. These ﬁndings could
be useful in other agricultural systems where cover crops are used for the control of an ,
insect pest, because they provide valuable insight into the principles of insect-plant
interactions that is critical in the selection of the appropriate plant species. For example,
depending on the type of main crop and on growers’ pest control goals, certain cover crop

species are more suited to control adults while others reduce larval abundance.

58

CHAPTER 4:
DO PLANT CUES INFLUENCE THE OVIPOSITION BEHAVIOR OF

JAPANESE BEETLES?

INTRODUCTION

Vegetation characteristics of agricultural landscapes can be manipulated to control insect
pest populations (Tonhasca and Byme 1994, Gurr et al. 2004), but this requires a
thorough understanding of the mechanisms underlying the insect-plant interaction
targeted for management. Plants are frequently used to manipulate the behavior of
herbivores to reduce pest pressure and this is often achieved by changing the plant
composition (Prokopy 1994, Bugg and Waddington 1994, Hartwi g and Ammon 2002).
This type of habitat modiﬁcation changes the managed ecosystem to limit availability of
one or more of the resources needed for insect development and reproduction, making the
environment less suitable for pest populations (Flint and Gouveia 2001).

As an adult, the Japanese beetle (Popillia japonica Newman, Coleoptera:
Scarabaeidae) has an extremely wide diet breadth; it feeds on the leaves, ﬂowers and
fruits of more than 300 plant species (Fleming 1972, Potter and Held 2002). This has
made this species an important pest of horticultural crops in the eastern United States.
Host selection by generalist species is likely to be inﬂuenced by relatively nonspeciﬁc
stimuli common to many different hosts (Thorsteinson 1960, Renwick 1983, Renwick
1989, Foster and Howard 1998), which for P. japonica are likely to be green leaf odors
(Ahmad 1982, Loughrin et al. 1995) and color (Held and Potter 2004). Compared to

food-searching behaviors, fewer cues may be involved in choosing an ovipositional site

59

in this species, since the larvae are facultative monophages (Potter and Held 2002).
Because of the limited mobility of larvae, the mother’s ovipositional site selection
decisions are crucial for hatching and survival of the offspring. However, despite the
economic importance of P. japonica to agricultural systems and its potential for invasion
into uninfested regions, biotic cues that are involved in the selection of ovipositional sites
have not been investigated.
Ovipositional preference and specificity are most often studied in species that lay

eggs on the above-ground sections of plants (W iklund 1981, Foster and Howard 1998,
Bossart and Scriber 1999, Showler 2001), in part because of the challenge of measuring
ovipositional preference of insects that lay eggs in the soil. Circumstantial evidence for
ovipositional site selection can be gathered from the distribution of larvae in the soil; P.
japonica larval density is greatest in moist, sandy soils with grass cover in sunny
locations (Fleming 1972, Potter et al. 1996). Evidence for P. japonica ovipositional
response to specific abiotic cues is also available from laboratory experiments; females
were found to be selective for soil qualities such as texture, moisture and organic matter
(Régniere et al. 1981b, Allsopp et al. 1992). Laboratory bioassays have demonstrated that
incorporation of Metarhizium anisopliae (a fungal pathogen for the biological control of
scarab grubs) into soil inﬂuences oviposition by P. japonica (Villani et al. 1994). Taken
together, these studies suggest that female P. japonica use soil-borne cues in post-
alighting ovipositional decisions, but the role of speciﬁc plant cues in ovipositional
decision-making remains unclear.

Female P. japonica have a strong preference to oviposit in grass, so cues from

vegetation are expected to play a key role in female choice. We developed a bioassay

60

method to allow quantification of P. japonica oviposition and to determine the inﬂuence
of visual, tactile, and olfactory plant characteristics on ovipositional investment
decisions. Oviposition was compared between grass and clover, which are both
commonly used cover crops, to determine beetle response to plants. These plants were
chosen because P. japonica larvae are more abundant in ryegrass compared to clover
(Fleming 1972), suggesting female preference for grass. However, I wanted to determine
whether preference is caused by differential suitability for oviposition. To investigate
mechanisms underlying P. japonica oviposition, observations of digging behavior were

made in ovipositional choice bioassays.

MATERIALS AND METHODS

Oviposition arena

Bioassays were conducted in a four-choice oviposition arena made of a Plexiglas box
(internal dimensions 15 x 18 x 18 cm) (Fig. 4.1). The arena had a removable Plexiglas
top, while the other sides were permanently glued together. To supply beetles with an
inert ﬂoor for walking and provide access to the top of the soil blocks, a rectangle of
Styrofoam (15 x 18 x 2.5 cm) was inserted into each arena. The Styrofoam square had
four 4 x 4 cm openings cut into it, with two along each of the longer sides, so adjacent
openings were 5 cm apart. The openings were cut with a Styrofoam-cutter (Flora Craft
Co., Ludington, MI, USA) using a cardboard template to guide the cutter. A 2 cm
diameter hole was cut in the middle of the Styrofoam insert to facilitate handling and
once in position, a 7 cm diameter piece of plastic sheet was placed over this hole with a

fresh piece of apple on top for the beetles to feed on. To assemble the bioassay chamber,

61

-r-.

the Styrofoam insert was wedged 8 cm above the bottom of the box to supply beetles
with an inert surface for walking and provide access to the top of the soil blocks, and the
potential ovipositional resources described below were inserted into each of the four
openings. All arenas were washed between experiments with a weak detergent solution

and thoroughly rinsed with water.

Insects

To extend the availability of adult P. japonica, beetles were collected once in June 2004
(before P. japonica emergence in Michigan), from Auburn University, Piedmont
Substation, Camp Hill, AL, USA, and this group of beetles was used in Experiment 1.
For Experiments 2-6 and the behavioral observations, adult P. japonica were collected
from Michigan State University campus Entomology Research Farm in East Lansing, MI,
USA from J ulyl through September 2004. Japanese beetle Expando Traps were baited
with a ﬂoral lure (Great Lakes IPM Inc., Vestaburg, MI, USA) that attracts both sexes.
Traps were placed out into open, grassy fields and were emptied after one day of
deployment. Beetles were transported to the laboratory as needed for the experiments. To
ensure that all females were mated, males and females were placed together in a 1 m3
screen cage in the laboratory at room temperature, without an oviposition medium and
were provided with cut apple slices ad libitum for 2-3 days prior to use in the
experiments. Females for the bioassays were taken out of the screen cage just before use
in the experiments. Beetles were sexed based on morphological differences on the foreleg

tibia (Fleming 1972), and a single female beetle was placed on the slice of apple in the

62

center of each oviposition arena at the start of each assay. Beetles were used only once in

the bioassays.

 

63

 

:“zl .
A’ Plexrglas box

 

ryegrass
\

 

apple it" ~‘ 9

artificial ‘
grass \

 

K Styrofoam floor

 

 

 

 

 

Figure 4.1. Bioassay arena for measurement of oviposition decisions by P. japonica.

Soil and plant material
The soil used in all experiments and for all of the treatments was a 1:1 (vol/vol) mixture
of organic potting soil (Scotts garden soil, The Scotts Company, Marysville, OH, USA)
and sand (Dixie Cut Stone & Marble Inc., Bridgeport, MI, USA). After thorough mixing,
the soil was placed into 4 x 4 x 8 cm light material paper bands (Monarch Manufacturing
Inc., Salina, CO, USA), and seeds were sown on top as needed for the experiments.
Perennial ryegrass, (Lolium perenne L.) (The Scotts Company) and alsike clover,
(Trifoilum hybridum L.) (Michigan State Seed Solution, Grand Ledge, MI, USA) were
grown in a greenhouse at 22-30 °C and ambient light. Enough water was added daily to
the soil blocks from the top to moisten the column completely. New plants were seeded
each week and ca. 3 week-old plants were used for the experiments; new plants were
used for each experiment. Soil blocks used for the experiments had uniformly healthy,
vigorous, and densely growing plants. The bare ground treatment consisted of soil alone
with an even, ﬂat surface, in the paper sleeves.

Artiﬁcial grass stems were arranged in the soil after the blocks had been placed
into position in the oviposition arenas. Green ﬂoral wire stems (Panacea, Columbus, OH,
USA) were washed with 70% ethyl alcohol and air-dried prior to use in the experiments.

Thereafter all artiﬁcial stems were handled using gloves.

Bioassays
To start each assay, the soil blocks were installed into a cage in randomized positions,
and a beetle was placed in the center of the arena on a slice of apple. Once all arenas had

a beetle, they were closed with the lid and transferred to an environmental chamber at 25

65

°C, 60% r.h., and L16: D8 for 7 days. The orientation of each arena was randomized in
the environmental chamber to compensate for any directional bias. Oviposition arenas in

an experiment were set up at the same time.

Experiment 1. To determine the relative ovipositional investment in artificial resources in
the presence of a natural resource, beetles were assayed with bare ground, artiﬁcial grass,
grass roots, and perennial ryegrass trimmed to 5 cm. For the artiﬁcial grass, 0.7 mm
diameter green ﬂoral stern wire was cut into 5.5 cm long pieces with a wire cutter. Ten
wire pieces were then inserted vertically ca. 0.5 cm into the soil, in a random
arrangement. Grass roots were obtained by cutting perennial ryegrass at soil level with

scissors. A total of 48 arenas was established for this experiment.

Experiment 2. Arenas were set up with bare ground, artificial grass, bare ground with
holes and bare ground with holes and grass puree to determine whether beetles respond to
plant cues from the soil surface. The artiﬁcial grass treatment was prepared as described
above. To imitate grass roots loosening the soil, ten holes were punctured 0.5 cm deep
into the bare ground with a ﬂoral stem wire. Holes were randomly distributed on the
surface of the soil blocks. For the treatment with holes and grass purée, both the above .
and below ground plant parts of 40 g fresh perennial ryegrass were homogenized with
200 ml deionized water using a kitchen blender, and 5 ml of this was poured on the
surface of soil blocks. Then ten holes were punctured in a similar manner as described

above. Thirty replicates were set up for this experiment.

66

Experiment 3. In this experiment, the acceptance by P. japonica females of clover or
grass as ovipositional resources was compared. A soil block containing clover was placed
diagonally across from another soil block containing ryegrass in each of the arenas and
the two other positions were ﬁlled with soil blocks containing bare ground. Thirty—one

arenas were set up for this experiment.

Experiment 4. The aim of this experiment was to determine whether oviposition by

female P. japonica is affected by artiﬁcial grass stem diameter. Ten wires made of either
0.4, 0.9, or 1.5 mm diameter green ﬂoral stem wire (Panacea, Columbus, OH, USA) were
positioned vertically and were randomly distributed across the soil blocks. A bare ground

soil block served as the control. Twenty-nine arenas were established for this experiment.

Experiment 5. To determine ovipositional response by P. japonica to the density of
artiﬁcial grass stems, the number of wire stems per soil block was varied. Either zero,
ﬁve, 10 or 15 artiﬁcial stems made from 5 .5 cm lengths of 0.7 mm diameter green wire
were randomly spread out on the surface of the soil blocks. Thirty oviposition arenas

were set up for this experiment.

Experiment 6. The response by P. japonica to the color of artificial plants at an
oviposition resource was assayed by providing beetles with a choice between 0.9 mm
diameter steel wire (Anchor Wire, Goodlettsville, TN, USA) stems painted with either
white, green, blue or yellow spray paints (Aco Hardware, Farrnington Hills, MI, USA).

Steel wire was used instead of the green ﬂoral stem wire to provide an inert surface for

67

 

the paint. After painting, wires were left to dry in a fume hood for 24 h. The wire was
then cut into 5.5 cm long pieces and 10 wire pieces of the same color were vertically
inserted approx. 0.5 cm deep into a soil block, arranged randomly on the surface. Thirty
arenas were set up for this experiment, with the four color treatments arranged randomly

in the four positions.

Assessment of oviposition

After one week in the environmental chambers, soil blocks were removed individually
from the arenas and visually inspected for eggs by gently breaking apart the soil. Eggs
were picked from the soil with a moist paintbrush to verify their identity. The number of

eggs found in each soil block was recorded.

Statistical analysis of bioassay data

The probability that the mean number of eggs laid in a treatment was different from that
expected by chance (25% in each soil block) was evaluated using a logistic model (PROC
GENMOD, SAS Institute 2001), with cage as the repeated subject. The mean number of
eggs per soil block was compared among treatments using a 12 test. Differences among
treatments in the proportion of females choosing a resource, indicated by the presence of
eggs in a soil block, were analyzed using a 12 test (Ott and Longnecker 2001).
Ovipositional preference was determined by correlation analysis between the proportion
of soil blocks with eggs and the average number of eggs per soil block (Spearman’s
correlation, PROC CORR, SAS Institute 2001). The critical value for signiﬁcance for all

these tests was determined at a = 0.05 level.

68

Behavioral observation

Detailed observation of oviposition behavior by P. japonica was performed using
simultaneous video recordings of individual females placed in each of four bioassay
arenas that each contained soil blocks with perennial grass, artificial grass, grass roots
and bare ground. Arenas were set up as described in Experiment 1. Experiments were
conducted between July 20 and September 2, 2004, producing a total of 9 recordings
from four arenas. Each recording lasted 24 days based on the condition of the insect,
plant and food material. On each day of the experiment, the video recording began at
10.00 hr and was terminated at 19.00 hr. Four closed circuit security cameras (Shebar,
Burton, MI, USA) simultaneously ﬁlmed the bioassay chambers from above and were
connected to a time-lapse video recorder (Shebar), which recorded for 24 hr on an 8 hr
videotape. Video tape recordings were then converted into digital format using Broadway
Pro Capture Software (Data Translation, Inc., MA, USA) for quantiﬁcation of beetle
behavior.

Beetle behavior was scored using The Observer 5.0 (N oldus Information
Technologies, Wageningen, The Netherlands). The number of digging events per day and
the proportion of time spent digging (digging on the surface plus time spent under the soil
surface) were recorded. The total number of digging events was adjusted for the length of
each observation by dividing the total number of events by the number of days of the
respective observation. Seventy-eight percent of observations lasted for more than 2 days.
The effect of the four treatments on digging behavior was analyzed with a general linear
mixed model (GLIMMIX macro, SAS Institute 2001) suitable for analyzing non-

normally distributed data, with camera as a random factor and observational date as the

69

block in the model. Treatment means were compared using the Tukey-Kramer adjusted
LSMEAN S t — test (SAS Institute 2001). The critical value for significance for these tests

was determined at 01 = 0.05 level.

RESULTS

Bioassays

Experiment 1. The average number of eggs per soil block was significantly different
among all of the treatments (Fig. 4.2a); 60% of the eggs found in the arenas were under
ryegrass. Bare ground had the fewest eggs on average; it received only 2.5% of the eggs
laid. The maximum number of eggs in a soil block showed a similar trend; a maximum of
25 eggs were found in ryegrass soil blocks and only seven in the bare ground treatment.
Grass roots received signiﬁcantly more eggs on average than artiﬁcial grass (x2 = 6.72, P
< 0.01), and the maximum number of eggs per soil block was also higher for grass roots
(20) than for artiﬁcial grass (15). The proportion of soil blocks with eggs was
signiﬁcantly different from a random distribution in bare ground (10%), and in perennial
ryegrass (71%) (x2 =10.78 for both tests, P < 0.05). Ovipositional preference was not
signiﬁcantly correlated between average numbers of eggs and proportion of soil blocks

with eggs (r2 = 0.8, P = 0.2) (Fig. 4.3a).

70

average number of eggs / soil block (+ SEM)

 

0|

&

(J

_.

 

 

a.
Experiment 1
N = 48

d
l—hl
bare ground artifice ii?

Experiment 2 ab
N= 30

grass grass roots ryegrass

9)
CT

:l—e

' bare ground holes purée&holes alttflclal grass

 

C.
Experiment 3
N = 31

 

bare ground clover ryegrass

lab”

Figure 4.2. Average number (+ SE) of eggs per soil block testing the effect of grass cues
on ovipositional behavior (Experiments 1-3). Drawings below graphs indicate respective
treatments. Bars within a graph with the same letter above them are not signiﬁcantly
different (01: 0.05). ‘N’ represents the number of replicates in each experiment where one
beetle was used per replicate.

71

a

‘ Experiment 1 .

ryegrass

   
  
 

y = 10.28x
8:076

O
. grass roots
bare ground

artificial grass

 

 

  
  

 

 

    
 

 

 

‘é , 2 . . .
B 0 0.2 0.4 0.6 0.8
9 3 r
\ b- . holes artificial grass
8: 2.5 - Experlment 2
O) O
(D 2 4 .
“5
h 1'5 l = 5.9x - 0.35
8 purée & holes :12 _ 0 7 4
E 1 ‘ . " '
:
g 0.5 « bare ground
or
9 0 I V Y I l
g 0 0.1 0.2 0.3 0.4 0.5
<0 4 —
C. .
3-5 ‘ Experiment 3 ryegrass
3 ..
25 , y = 7.64x + 0.4
2, 3:072
15 - bare ground clover
O
1 O
0.5
0 a a . .
0 0.1 0.2 0.3 0.4

proportion of soil blocks with eggs

Figure 4.3. Relationships between the proportion of soil blocks receiving eggs and
average number of eggs per soil block, for different soil block treatments in Experiments

1-3.

72

Experiment 2. Thirty-five percent of the eggs in this experiment were found in artiﬁcial
grass, signiﬁcantly greater than the investment expected by chance (x2: 14.72, P < 0.01).
Artiﬁcial grass received signiﬁcantly more eggs on average than bare ground ()8 = 4.47,
P = 0.03) (Fig. 4.2b). The addition of grass purée to soil with holes caused no signiﬁcant
change in the average number of eggs per soil block compared to soil with holes ()8 =
3.0, P = 0.08). Artiﬁcial grass was also not signiﬁcantly different from soil with holes (x2
= 3.32, P = 0.97). Artiﬁcial grass and soil with holes received a maximum of 13 eggs,
puree with holes in the soil received nine, and bare ground received eight. Ovipositional

preference was not statistically signiﬁcant (r2 = 0.74, P = 0.26) (Fig. 4.3b).

Experiment 3. Although 3.6 times as many eggs were laid in ryegrass than in bare
ground, and 3.0 times as many in ryegrass than in clover, there was no signiﬁcant
difference among the treatments (Fig. 4.2c). Maximum number of eggs per soil block was
highest in ryegrass (25), and was lowest in bare ground (11). More than twice as many
soil blocks with ryegrass had eggs in them than bare ground ones, but there was no
signiﬁcant difference between the proportions of soil blocks with eggs. Ovipositional

preference was not statistically signiﬁcant (r2 = 0.76, P = 0.051) (Fig. 4.3c).

Experiment 4. The number of eggs laid per soil block increased with increasing stem
diameter (Fig. 4.4a), but there was no significant difference among the stem diameters.
Medium diameter artificial stems received signiﬁcantly more eggs on average than bare
ground (x2 = 4.43, P = 0.03) (Fig. 4.4a). Maximum egg number of eggs per soil block did

not differ markedly among treatments, 16 being the highest in medium stem diameter,

73

and 10 the lowest in bare ground. Fifty two percent of beetles laid eggs in soil with the
largest stem diameter. Ovipositional preference was not signiﬁcantly correlated between

the two variables (r2 = 0.71, P = 0.2) (Fig. 4.5a).

Experiment 5. Average number of eggs per soil block increased with increasing stem
density, although there was no statistical difference between the stem-containing
treatments (Fig. 4.4b). Treatments with 10 and 15 stems had signiﬁcantly more eggs on
average than bare ground (x2 > 4.62, P < 0.03). Soil blocks with ﬁve stems had 6.5 times
as many eggs on average compared to bare ground. Ovipositional preference increased

nearly linearly with increasing stem density (r2 = 0.98, P < 0.01) (Fig. 4.5b).

Experiment 6. Stem color had no effect on the average numbers of eggs per soil block
(Fig. 4.4c) and the proportion of soil blocks with eggs was also similar across the
different treatments. Twenty-seven percent of the blocks with blue and green stems
received eggs, whereas 30% of the yellow and 33% of the white had eggs in them (Fig.
4.40). The correlation between proportion of soil blocks with eggs and average number of

eggs per soil block was the weakest in this experiment (r'2 = 0.27, P = 0.26) (Fig. 4.50).

74

4 Experiment4
3.5 N :29

3 ab

2.5 FL
2

1.5 b
1 .

0.5

0 bare ground small large

I U
b

3 Experiment 5 a

N = 30 a

2.5

2 ab

1 r
0.5 b

0 I l , , .
bare ground 5 1‘? TI)

”HI '1' y]: I‘l”

2 c

Experiment 6
N = 30

l—a

 

 

 

 

 

 

 

 

 

 

average number of eggs / soil block (+ SEM)

1.5

0.5

   
 

 

   

white yellow blue green

I ‘ W W

Figure 4.4. Average number (+ SE) of eggs per soil block testing artiﬁcial grass cues on
oviposition behavior (Experiments 4-6). Drawings below graphs indicate respective
treatments. Bars within a graph with the same letter above them are not signiﬁcantly
different (cl= 0.05). ‘N’ represents number of replicates in the experiment.

75

3.5 -

   
 

 

 

a. medium. large
3 1 Experiment 4
2.5 .
y : 8.78x - 1.36
2 d r2 = 0.71 small
1.5 -
1 a
bare ground
0.5 -
0 l I r r 1 ﬂ
0 0.1 0.2 0.3 0.4 0.5 0.6
3 "b

2.5 ‘ Experiment 5

    
   

15 stems
2‘ y=4.27x-O.13
,5, 3:098

1 5 stems

 

 

average number of eggs / soil block

 

 

0.5 .
0 bare ground
0 0.1 0.2 0.3 0.4 0.5 0.6
1.6 ~ 6.
1-4 ‘ Experiment 6 O
1.2 i green
1~ y=6.18x-0.94
0.8 ~ r2 = 0.27 brie
0.6 « yel ow
, O
0-4 white
0.2 ~
0 I T I 7
0 0.1 0.2 0.3 0.4

proportion of soil blocks with eggs

Figure 4.5. Relationship between the proportion of soil blocks receiving eggs and
average number of eggs per soil block, for different soil block treatments in Experiments
4-6.

76

Behavioral observation

Digging behavior of P. japonica was significantly different among the four treatments,
for both the number of digging bouts (F33 = 4.17, P < 0.01) and the duration of the
digging bouts (F33 = 7.86, P < 0.01). Beetles dig signiﬁcantly more frequently on
ryegrass than on bare ground (t = -3.01, d.f. = 16, P = 0.02) (Fig. 4.6a). Although no
statistically significant differences were documented in any of the other mean
comparisons, the number of digging events on grass roots was almost three times higher
than on bare ground. Beetles started digging twice as many times on ryegrass (0.77 / day)
than on artiﬁcial grass (0.38 / day).

Female P. japonica spent signiﬁcantly more time digging and under the soil
surface in ryegrass than in artificial grass (t = 3.94, d.f. = 16, P < 0.01) or grass roots (t =
3.16, d.f. = 16, P = 0.01) (Fig. 4.6b). Time spent digging in bare ground (36%) was
similar to that in grass roots (33%) and artiﬁcial grass (34%), whereas beetles spent 92%
of their time digging on or under the soil surface of ryegrass. Despite these numerical

differences, the proportions were not signiﬁcantly different.

77

0.8 ~ ab

0.6 . ab

0.4 ~

average number of digging
events / day

0.2 -

 

0 . , . . .
bare ground artificial grass grassroots ryegrass

 

12
b. a
1
9E
52
5g o.a<
8’8
--.0
go
2.9
C
8g 0.4-
92
%e
0.2-
o

 

 

 

 

bare ground artificial grass grassroots ryegrass

Figure 4.6. Frequency and duration of digging behavior by P. japonica in bioassay
arenas, as a function of treatment. Average number (+SE) of digging events per day (a)
and average duration of digging bouts (b) on the four treatments. Bars in a graph with the
same letter are not signiﬁcantly different (a= 0.05).

78

DISCUSSION

Bioassays conducted in the oviposition arena developed for this study revealed that
female P. japonica are selective in their oviposition investment decisions. Natural grass
elicited the strongest response, and isolated structural and chemical cues also stimulated
females to lay eggs in the soil. Plant-related ovipositional preference has been examined
previously in ﬁeld studies with insects that lay eggs in the soil, including scarabaeid
(King et al. 1981 and references therein) and chrysomelid beetles (Boetel et al. 1992,
Rondon and Gray 2004), although these studies did not investigate the mechanisms
involved in ovipositional site selection. The bioassay developed in this research could be
useful for studying ovipositional behavior of generalist or specialist insect species that lay
their eggs in the soil, as well as in determining the factors involved in oviposition
decision-makin g.

The number of eggs per soil block increased as an increasing number of plant
cues was present on the soil surface. This trend in P. japonica oviposition has been
inferred from grub abundance in soil samples in different agricultural systems (Fleming
1972, Smitley 1996, Szendrei, unpublished data), but, to date, it has not been measured
directly. When female P. japonica were offered a choice between natural and artiﬁcial
ovipositional cues at ovipositional sites, beetles preferred ryegrass more than artiﬁcial
stimuli (Experiment 1), demonstrating either that the artiﬁcial grass was lacking some
visual, tactile, or olfactory quality that enhances oviposition. Manipulation of the soil
surface by punching holes (Experiment 2) increased the average number of eggs laid
compared to that in bare ground, suggesting that female P. japonica use soil texture in the

evaluation process of the oviposition medium. The presence of grass puree on soil blocks

79

with holes did not increase egg deposition. This may have been because P. japonica do
not use volatile or tactile chemical cues from grass to select an ovipositional site, or
because of the higher moisture content of these soil blocks with the addition of the grass
puree. Alternatively, the odors released after our grass-extraction procedure are different
from those of real grass. Future investigations with different grass parts and plant
extraction methods are needed to examine the role of volatile and non-volatile chemical
cues in oviposition decisions by P. japonica.

In Experiments 1, 2, 4, 5, and 6, soil blocks with artificial grass received
consistently more eggs than bare ground, revealing that vertical objects on the soil
surface serve as post-alighting pro-ovipositional cues. This kind of response to vertical
plant cues on the soil surface was also observed to be one of the important cues
stimulating onion ﬂy oviposition behavior (Harris and Miller 1984). In the presence of
artiﬁcial stems (Experiments 4-6), P. japonica were able to discriminate between the
quality and quantity of the stimuli on the soil surface; more eggs were laid in soil blocks
that had large stem densities or a higher number of stems. It is possible that females
initially oriented towards a resource because there were more or larger diameter stems in
the soil, which was perceived as a larger soil surface area covered by potential resources
for the soil-dwelling offspring. After encountering the resource, the positive response in
ovipositional investment to stem diameter (Experiment 4) and density (Experiment 5)
could be a result of the greater probability of encounter with the cues that are
proportional to the diameter or density of the stems (Visser 1988). It could also be that
elevated number of eggs in artificial grass treatments compared to bare ground occurred

because the wire stems impaired the beetles’ movement, causing them to stay below

80

ground longer, although this is unlikely, since the stems reached 0.5 cm into the soil and
the sandy soil allowed the stems to fall if they were pushed by the beetles. In Experiment
3, more eggs were found in ryegrass than in clover, suggesting perception of visual,
tactile, or olfactory differences between the two plants. However, in the context of our
present ﬁndings, morphological (i.e. leaf shape) differences between the two plants may
play a signiﬁcant role in female ovipositional choices.

Popillia japonica are not only capable of evaluating the plant characteristics on
the soil surface, but our results also suggest that they adjust their investment decisions
according to the strength of the resource cues, i.e. more eggs were laid in resources that
were accepted more frequently. Correlations between the proportion of soil blocks with
eggs and the average number of eggs per soil block were not statistically signiﬁcant in
most cases, but in Experiments 1-5, the regression coefﬁcients were all positive and
above 0.7. This ﬁnding is further supported by the maximum egg numbers, which were
typically the highest in those resources that were more often chosen by the females
throughout all the bioassays. Additional investigation is needed to verify that choice of an
ovipositional resource translates into increased larval survival and development, as has
been found in other herbivorous insects (J aenike 1978, Craig et al. 1989, Bemays and
Chapman 1994, Showler 2001). Popillia japonica larval survival and growth are highest
on perennial ryegrass (Crutchfield and Potter 1995) compared to other cool-season
turfgrass species, but the degree of correlation between P. japonica female preference
and larval performance is not known. The use of sub-optimal resources as ovipositional
sites suggests that spreading offspring survival risks (Root and Kareiva 1984, Seger and

Brockman 1987, Freese and Zwolfer 1996, Roitberg et al. 1999) could be an important

81

strategy where larvae have little mobility, as in the case of P. japonica. Although this
might be an artifact of the amount of time the beetles were enclosed in the arenas, in 39%
of all the bioassay arenas across the six bioassay experiments, females laid eggs in more
than one treatment. Interestingly, the highest propensity to lay eggs in more than one
resource was found in Experiments 1 and 4, where the correlations between the
proportion of soil block with eggs and the numbers of eggs per soil block was the highest.
The variation in ovipositional activity between experiments could be because females of
varying ages were used in the different experiments, although Van Timmeren et al.
(2001) found that in laboratory bioassays P. japonica ovipositional rate was not affected
by the age of females.

On a highly stimulating resource, P. japonica may accumulate sufﬁcient stimulus
for oviposition in less time than on a lower quality resource, therefore enabling quick
decisions to lay eggs (Finch and Collier 2000). This may allow additional time for other
behaviors, such as feeding and mating. Some insects can adjust their oviposition behavior
based on resource quality, for example by reducing the number of eggs laid on a poor
quality resource (Awmack and Leather 2002), and based on the present results (Fig. 4.3,
4.5) this might be the case for P. japonica. An implication of these ﬁndings is that most
grass species used in lawns may be equally suitable for oviposition by P. japonica. Future
investigations into potential ground covers for control of P. japonica through reducing
oviposition should focus on plants that have broad leaves or other morphological or
chemical features that markedly distinguishes them from turfgrass species.

While the bioassays revealed cues that stimulate P. japonica females to lay eggs

in soil, the observations showed how these surface cues might play a role in inﬂuencing

82

digging behavior and the time spent under the surface. Beetles that engaged in digging
either disappeared under the surface in a short time or ceased digging, so the longer
digging bouts correspond to longer time spent in the soil. The number of digging events
was not signiﬁcantly different among the three treatments with surface cues (ryegrass,
grassroots and artificial grass) (Fig. 4.6a), but beetles spent signiﬁcantly more time in the
soil under ryegrass than under any of the other treatments (Fi g. 4.6b). Popillia japonica
may perceive contact cues for oviposition from ryegrass leaf blades on the surface that
not only determine whether they oviposit in a resource, but also increase time spent in the
soil, and the number of eggs laid. Alternatively, there may be ovipositional cues detected
after entering the soil under ryegrass that provide further indication of resource
suitability, modulating the number of eggs laid. These cues may be physical or chemical,
and may involve interactions between soil particles and grass roots.

Further investigation of P. japonica oviposition behavior is expected to provide
insight into potential avenues for disruption of this insect that is polyphagous as an adult,
but more speciﬁc in its oviposition behavior. In annual and perennial agricultural
systems, ground covers may be successful as natural barriers to avert P. japonica from
ovipositing in areas that would otherwise provide suitable conditions for larval

development.

83

CHAPTER 5:
SURVEY OF JAPANESE BEETLE MANAGENIENT PRACTICES IN

MICHIGAN BLUEBERRY

INTRODUCTION

Michigan leads the nation in production of hi ghbush blueberry (Vaccinium corymbosum
L.), with 33% of the US crop produced in this state. In Michigan in 2003, there were
15,900 bearing acres of blueberries producing an average of 3,900 pounds of berries per
acre (Kleweno and Matthews 2004). Of this, 24 million pounds were for the fresh market,
receiving $1.30 per pound, and 38 million pounds were for the processed food market,
receiving $0.84 per pound. In the past ﬁve years in Michigan, farm gate value of the
whole industry has ranged from $49 (2001) to $63 (2003) million (Kleweno and
Matthews 2004).

This industry has been challenged by the recent invasion of the Japanese beetle
(Popillia japonica Newman, Coleoptera: Scarabaeidae) into it’s major production
centers. Although this pest was ﬁrst detected in Michigan in 1932 (Anon. 1962),
populations of Japanese beetle in west Michigan have increased in the past ten years, and
some growers are applying increasing amounts of insecticides to maintain beetle-free
fruit. Japanese beetle adults start emerging in early July and feed and mate on the bushes
until mid—September (Fig. 1.2), which coincides with the period of blueberry harvest. The
primary economic effect comes not from feeding damage, but from the risk of
contamination of harvested fruit. About 75% of Michigan's blueberries are sold for

processing, and for this purpose berries are harvested by machines that shake the fruits

84

off the bushes. If beetles are present on the fruit or foliage as the harvester passes over the
bushes, there is a risk of beetles being collected along with the berries. Buyers require the
harvested fruit to be completely free of insect contamination, and although post-harvest
techniques are available for removal of beetles and other items from the fruit, this
situation demands superlative pest control by the growers facing this pest in their ﬁelds.
In 2001, Japanese beetle was identiﬁed as the most important production problem for
blueberry by commercial growers in Michigan (Anon. 2001), and has become a major
focus for research and extension programs.

This survey was developed in 2003 to collect information from commercial
blueberry growers in Michigan about the extent and magnitude of the problem Japanese
beetles cause in blueberry production. The survey was also conducted to quantify
management tactics that growers have adopted to control Japanese beetle and to
determine willingness of growers to incorporate alternative management strategies, such
as the use of cover crops, in their future management programs if they are proved

effective against Japanese beetle.

MATERIALS AND METHODS

The survey (see Appendix 1) was developed to cover both sides of a letter sheet of paper
and to be quick to complete (ca. 10-15 min). It was approved by the University
Committee on Research Involving Human Subjects (UCRIHS) at Michigan State
University (see Appendix 2) prior to being mailed to commercial hi ghbush blueberry
growers in South Michigan in 16 counties across the main blueberry production regions

of the state. Surveys were mailed to 215 growers in the winter of 2003 by the Michigan

85

Blueberry Growers’ Association, Grand Junction, MI and by Genesee County Extension,
Flint, MI who used their address lists to mail the surveys to growers. Reply envelopes
with postage were enclosed with the surveys, and reminders to return surveys were made
at MSU Extension meetings in December 2003. Each survey that was returned was given
a unique identification number used for the data organization and analysis. The name and
address ﬁelds were optional to ﬁll out, thus the answers could remain anonymous, based
on the decision of the respondent. Data were entered into a Microsoft Excel spreadsheet,
with the answers for each question entered on a separate spreadsheet. Some growers had
farms in multiple counties, and these were treated as separate entries for the analysis of
the ﬁrst three questions. This was the case for 7 farms, and for these the total acreage was
equally divided among the counties the farm was located in. Median and maximum
values were calculated for values such as farm size and economic losses. Economic loss
per acre was calculated by dividing the estimated loss per year by the total acreage from
each farm. Percentage of respondents answering questions in a certain way were
calculated based on the number of respondents for each question. Answers were analyzed

with a x2 test, to determine signiﬁcant departures from that expected by chance.

RESULTS AND DISCUSSION
Surveys were received from commercial blueberry growers in 16 counties of the southern
tier of Michigan (Fig. 5.1a). A total of 99 surveys were received, and these were used in

the analyses. The median farm size throughout this study was 21 acres (range: 1-375

acres) (Fig. 5.1b).

86

Figure 5.1. Map of Michigan’s lower peninsula with the number of responses (a.) and
median farm size (acres) (b.) per county, in 16 counties in the southern tier of Michigan.
Names of Michigan counties in the main blueberry growing region are indicated on the
left of the map (a.).

87

 

 

 

 

 

 

 

 

 

 

 

E

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2
9 1 2 ,
Muskegon 1
34 ‘ 4 L1 ‘
Ottawa " ‘
Allegan 13 I

 

 

 

 

 

 

 

 

 

 

 

 

Van Buren 29 1
Berrien 5 1
\b (b.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3“, 11.2 » 3°

35 l 2.5
60 l 3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

88

Eighty-four percent of respondents considered Japanese beetle a pest on their
farms. On a 1-5 scale, (l-not important, 5-extremely important), 31% found Japanese
beetle an extremely important production problem, and this value is signiﬁcantly higher
than expected (x2=11.52, P < 0.05). Only 6% of the respondents thought it was not an
important insect pest. Responses ranged between growers considering this pest to be the
most important production problem on their farm, to not having seen a Japanese beetle.

Among the major blueberry producing counties (in southwest Michigan), 53% of
respondents considered Japanese beetle as an extremely important production problem in
Allegan county, whereas only 22% of respondents in Muskegon county (Fig. 5.2). The
reason for fewer growers considering Japanese beetle an extremely important pest in
Muskegon county could be because of its geographic location: it is in the northern part of
the main blueberry growing region, and Japanese beetles are not as abundant in the
northern counties in Michigan as in the south part of the state. This could be because this
beetle was ﬁrst introduced into the state near Detroit, so it initially established in the
southeast and/or because the climatic conditions in the northwestern part of the state are

less suitable for sustaining Japanese beetle populations.

89

Percent respondents

A 0) oo 3
o o o 0

Median economic loss ($) / acre
N
O

50

4O

30

20

10

 

o . . .
Muskegon Van Buren Allegan
Ottawa Berrien

 

Figure 5.2. Percent of respondents that ranked Japanese
beetle as the most important pest in blueberry production, by
county.

 

_ $96 $99
$44 $44
$33 I I
0 J . . .

 

Muskegon Ottawa Berrien Allegan Van Buren

Figure 5.3. Estimated economic loss per acre from Japanese
beetle at blueberry farms in SW Michigan counties.

90

The median economic loss per acre due to Japanese beetle as estimated by
respondents was $72 per year, and this value varied greatly among counties (overall
range: $10-1000) (Fig. 5.3). In order to understand the causes of the ratings respondents
gave to the magnitude of importance of Japanese beetle as a pest, growers were asked
about some of the production factors that might be responsible for this loss. More
respondents were likely to consider Japanese beetle a pest on their farm if they had
overhead irrigation, than if they didn’t: 64% of growers who ranked Japanese beetle as an
extremely or very important production problem have overhead irrigation (x2 =124.19, P
< 0.05). More people rated Japanese beetle a very or extremely important production
problem who didn’t cultivate the soil around their ﬁelds (76%, x2 = 5.38, P < 0.05), than
if they cultivated the soil around their fields (23%).

Seventy-nine responses were received to the question investigating the reason for
the economic loss due to Japanese beetle. Signiﬁcantly more than expected (63%, x2 =
67.12, P < 0.05) of the respondents indicated increased insecticide use as the major
reason (Fig. 5.4). According to others, the damaged fruit reduced product value, damaged
bushes yield less than undamaged bushes, there is an increased need for cultivation,
harvest time is prolonged, or more labor and time are needed to remove beetles from
harvested fruit. The risk of load rejection was considered to be a reason for increased cost
by only 11% of growers, perhaps because if this occurs, it would be after the fruit have

left the farm.

91

 

 

 

120 *
- other
increased
§ - insecticide use
C H . .
(a, 80 lower frurt price
CD
.: El load rejection
0
E
E 40 r
3
Z
0
Figure 5.4. Causes of the economic impact
due to Japanese beetle.
0
87 4 A) El foliar insecticides

ﬂ soil insecticides

cover crop

E clean cultivation

   

63% I herbicide killed sod

 

Figure 5.5. Growers’ responses to the question "What is the most important
Japanese beetle control method?”

92

The two most widely used methods of Japanese beetle control were foliar
insecticides (63%) and clean cultivation (16%) (Fi g. 5.5). The other methods listed were
used by fewer than 21% of respondents. These include herbicide-killed sod, cover crops,
and the use of soil insecticides. Four growers used traps to reduce the number of beetles
(although this strategy could potentially attract more beetles into the crop), while others
(6 responses) removed non-crop plants from the field that attract Japanese beetle. These
results show that growers rely heavily on foliar insecticides for their Japanese beetle
control, but that they are adopting additional approaches to help reduce the incidence of
this pest.

Eighty—five of the respondents answered question 6 about the type of ground cover
used at their farm. Thirty-one out of the 85 respondents (37%) had permanent sod
throughout the row-middles across their farms. The second most commonly used row-
middle treatment was clean cultivation, but only 24% of the growers who answered the
question had this type of row-middle on their whole farm, suggesting that different ﬁelds
have different row-middle management strategies. Almost twice as many acres were
clean cultivated (2,067 acres) as were kept with permanent sod (1,254 acres) (Fig. 5.6).
Only 15% of the 85 answers indicated the use of multiple ground covers at an individual

farm.

93

Number of farms

Acres

35-

 

    

30 i
25 ~
20 -
15 -
10 ~
5 _
100% 100% clean 100% 100% other
permanent herbicide
sod killed sod
2500
2000 ““jllllllllll
[inlllll‘i‘i
will“
1 500 llm'lll‘l“
rmIWll twill
l'lmlllllll‘ lllll‘llll“
WW“ mu
100° mutillllllll .‘l‘lkitll
Willi-ill l‘il‘ll
500 tiﬁﬂlllllﬂ
11111111111 1‘. it

 

 

 

 

l .
clean permanent herbicide other
sod killed sod

Figure 5.6. Number of farms with different types of row-middle
management (top). Total acreage of farms with different row-
middle management (bottom).

94

Of the growers that replied to the survey, there were approximately 544 acres of farms
with herbicide-killed sod, and 207 acres of other types of ground covers. These included
rye and herbicide-killed rye.

Conditions in and around crop ﬁelds can favor or inhibit development of Japanese
beetle depending on whether fundamental requirements for the pest population
development are met (Vittum et al. 1999). Within many Michigan blueberry ﬁelds and
along the perimeter of these ﬁelds, ground covers of seeded grass or the naturally-
invading mix of grass and broadleaved weeds are commonly used to maintain soil
structure, provide conditions where agricultural machinery can be driven during wet
conditions, reduce soil erosion, and prevent pesticide and fertilizer runoff.

These areas provide ideal conditions for Japanese beetle where both the egg-laying and
larval developmental requirements are met. Changing these parameters, as some growers
already are doing, may reduce pest pressure over the long-term by affecting the insect’s
biology.

I asked whether growers had changed row-middle management practices to
control Japanese beetle. Ninety-two responses were received to the question and 42% of
these growers changed row-middle management practices, with their changes starting on
average in 2000. Out of 38 growers who changed management methods, 27 switched to
clean cultivation and eight changed to herbicide-killed sod. In one case the grower started
narrowing the sod strip in row-middles, and in another the grower switched to mowing
more often. This clearly shows a tendency to move away from grassy, living row-middle
ground covers. Fifty percent of growers who had made a change said that the changes

reduced Japanese beetle pressure, and 45% were not sure. When asked whether they

95

would be willing to incorporate new cover crops into their management strategies to
control Japanese beetle if they were effective, 59% of the 90 respondents answered “yes”

(Fig. 5.7), indicating a potential for integrating alternative ground covers into Michigan

blueberry fields.

I yes

n no
[:1 don’t know

59%

 

Figure 5.7. Responses to the question "Would you
be interested in using new ground covers if they
were effective against Japanese beetle?"

96

Eighty-nine responses were received to questions 9 and 10 about the perceived
advantages and disadvantages of cover crops and clean cultivation in blueberry ﬁelds
(Table 5.1). The two most important advantages of cover crops according to growers
were prevention of soil erosion (65%, 12: 34.5, P < 0.05) and preservation of soil
moisture (42%, x2: 16.0, P < 0.05). The most important disadvantage of having cover
crops was that they compete with the blueberry bushes (44%, NS) and they might serve
as an alternative host for pests (40%, NS). Seventy-six percent (68 answers out of 89) of
the growers answered that having clean cultivation provides control of pests (x2: 31.1, P
< 0.05), and half as many answers (34) indicated that clean cultivation is advantageous
because it decreases frost damage in the ﬁelds. The main disadvantage of clean
cultivation according to the survey was that it creates muddy and dusty conditions (x2:
16.1, P < 0.05), and the second disadvantage of clean cultivation was that it facilitates
soil erosion. Interestingly increased cost, such as the expense of cover crop seeds or
cultivation (as opposed to maintaining row-middles with cover crop) was never
considered of high importance. Growers’ answers indicated that the issues they consider
when deciding what soil management method to apply in their ﬁelds are mostly
horticultural, based on the effect on bushes (competition of cover crops for soil moisture)
and insect pests (cover crops could provide alternate hosts for pest and clean cultivation
is a method of pest control). Complaints about mud, dust and erosion in blueberry ﬁelds
are common among growers who have clean cultivation (Z. Szendrei, personal
communication with growers), and most of the blueberry ﬁelds in west Michigan have

light, sandy soils that are prone to erosion and create dust when dry.

97

Table 5.1. Advantages and disadvantages of cover crops and clean cultivation
for integration into blueberry production, and the percentage of survey
respondents that consider these important (in order of importance). Eighty-
nine responses were given.

 

 

ADVANTAGES ”:33?“ DISADVANTAGES “ﬁg?“
Cover cr0p
l. Prevents soil erosion 65 1. Competition with bushes 44
2. Conserves soil moisture 42 2. Alternative host for pests 41
3. Controls pests 36 3. Too expensive 30
4. Cuts fertilizer costs 12 4. Difﬁcult to seed 21
Clean cultivation
1. Controls pests 76 1. Mud and dust in field 76
2. Decreases frost damage 38 2. Soil erosion 54
3. Easy to manage 30 3. Soil moisture loss 34
4. Low cost of maintenance 12 4. Need for herbicides 25

 

Seventy two percent of respondents did not cultivate the soil around their fields.
Growers maintain this area to have permanent sod cover in the driveways and around the
ﬁelds for the maneuverability of tractors and harvesters. These areas therefore cannot be
switched to clean cultivation and there is a great need to develop effective approaches to
control Japanese beetle grubs that are developing here in this area of the blueberry farms.

I asked growers if there were any changes to their current Japanese beetle
management programs that they were planning for the future. Twenty-two out of 73
answers (30%) were “yes”. The types of changes growers were planning included:
switching to clean cultivation, planting cover crops, applying herbicide to row-middles,
use more foliar insecticides, removing weeds, spraying the ﬁeld perimeters, or using soil
insecticides. One grower was planning on planting grapes around the ﬁeld as a trap crop

and spraying them with insecticide (although this may not be beneficial, because of the

98

potential for attracting beetles to the area of the ﬁeld). Thirteen out of the 30 respondents
who said that Japanese beetle was an extremely important pest on their farm would not
change their current management program. Twelve said they would change their
management program, and the most common form of change was to spray more foliar
insecticide.

Seventy-two answers were received to the question about whether there are other
equally or more important pest management problems besides Japanese beetle. Blueberry
maggot (Rhagoletis mendax Curran) was the most common other pest problem for
growers (23%), but other common answers were cranberry fruitworm (Acrobasis vaccinii
Riley) (17%), diseases (9%), weeds (7%), birds (6%), mammals (groundhog and deer,
6%). There were a few other sporadic problems with other insect pests mentioned, these
included rose chafer (Macrodactylus subspinosus (Fabricius)), tussock moths (Orgyia
leucostigma (J .E. Smith», blueberry gall wasp (Hemadas nubilipennis), blueberry aphids
(Illinoia pepperi (McGillivary)), and oblique banded leafrollers (Choristoneura
rosaceana (Hanis)). One grower said that government regulations were an equally or
more important problem than Japanese beetle.

Among the growers who listed Japanese beetle as an extremely important pest on
their farm, cranberry fruitworm and blueberry maggot were the most common other
production problems, while weeds and cherry fruitworm (Grapholita packardi) were the
second most common responses.

When comparing farm size and the importance of Japanese beetle in production
we found that 67% of the owners of large farms (60—375 acres) said that Japanese beetle

is an extremely important pest on their farm, 43% of the 13-20 acre and 53% of the less

99

than 13-acre farm owners said the same (Fig. 5.8). Twice as many large farm owners (60-
375 acres) gave rank 5 (extremely important pest) than rank 4 (very important pest),
whereas this difference was only 1.3 -fold in the case of small farm owners (<13 acres).
This could mean that growers who could potentially suffer a large economic loss due to
Japanese beetle are more aware of the importance of this pest. It may also be that more
Japanese beetles are attracted to a large patch of blueberries than a small one, or that the
pest management program is easier to manage on a smaller farm than in a large one,

particularly around harvest time.

‘3 Japanese beetle is an extremely important pest (5)

- Japanese beetle is a very important pest (4)

 

375-60

 

 

 

UT
‘9
N
_.L

 

 

0 5 10 15 20 25

Number of responses

Farm size (acres)

 

 

 

c1

Figure 5.8. Relationship between farms size and the importance of Japanese
beetles as a pest in blueberry production. Importance of the pest was on a
scale from 1 (not important) to 5 (extremely important); results shown in the
graph are ranks 4 and 5 only.

100

CONCLUSION

This grower survey shows that Japanese beetle is an important pest management
challenge for blueberry producers in southern Michigan, with signiﬁcant economic
impact. Increased insecticide use was the major cause of the economic loss due to this
pest for the producers. For the blueberry industry, the major causes of the economic loss
included lost customers and sales due to grade reduction of berries that were chewed by
beetles and secondarily infected with fungus (Dave Trinka, personal communication). It
is estimated that at least half of the fruit loss is caused by this factor, which translates into
an economic loss of $7-10 million per year (Dave Trinka, personal communication).
Considering the $63 million dollar value of the industry in 2003, this translates into a
~l4% value loss for growers.

The most widely used methods of controlling Japanese beetle in Michigan
highbush blueberry fields are insecticides and clean cultivation. Szendrei et al. (in press)
showed that clean cultivated ﬁelds had fewer Japanese beetle larvae than fields with
permanent sod. In Michigan blueberry ﬁelds, more farms have permanent sod coverage
than clean cultivation, but the acreage of clean cultivated farms is higher. This means that
larger farms tend to be clean cultivated, and this could be correlated with the risk factor
and size of economic loss a grower could suffer if a Japanese beetle was detected in the
harvested fruit. It is likely that this trend is because growers who have larger farms and
rely on blueberry farming for income are taking more precautions to minimize the chance
of contamination of fruit with Japanese beetle. Currently, Japanese beetle is driving
changes in row-middle management, as indicated by growers recently switching to clean

cultivation. The changes implemented to control Japanese beetle are considered to be

101

effective by growers responding to this survey, and most growers are not planning on
making more changes to their pest management programs. Many growers would be
willing to try new cover crops if they are shown to be effective against this pest.

This type of survey is important because the results point out a number of further
opportunities for research and extension toward the adoption of new strategies to control
Japanese beetle in hi ghbush blueberry. The information gathered here could be used to
develop control strategies that rely on realistic management needs and existing methods.
Usefulness of grower surveys in developing future research directions has also been
found in the case of strawberry growers in Ohio (Scheerens and Brenneman 1990) and
New York (Stivers-Young and Tucker 1999). In the latter survey, the authors identiﬁed
the types of cover crops that were currently being planted and the advantages and
disadvantages that growers perceived when using them. In the future, participatory
research that actively includes blueberry growers may speed up appropriate adoption of

new cover crops or other strategies to help control Japanese beetle and other pests.

102

CHAPTER 6:

CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS

The research presented in this thesis has, in the broad sense, provided new
information in the study of polyphagous insects and their interaction with plants, with
relevance for pest management. Currently, there is an increased need for alternative
methods of pest control to reduce the use of pesticides in agricultural production and this
demands research into approaches that provide environmentally sustainable solutions to
reduce insect pest damage. Behavioral manipulation of polyphagous pests, such as the
Japanese beetle, through changing the resource composition within an agricultural ﬁeld is
challenging because of the lack of speciﬁcity in the insect’s feeding behavior, and their
ability to utilize multiple resources. The behavior chosen for manipulation is however not
necessarily the one causing injury to the plant, but rather a behavior that can be targeted
by management methods and is effective in reducing the subsequent population levels.
Results of this thesis indicate that, in the case of the Japanese beetle, a possible target for
behavioral manipulation is their oviposition, and its control thorough cultural practices.

Ultimately what matters for the success of pest control is the reduction of damage
to the crop and economic loss related to this damage. Although I have been able to prove
that spring larval populations are related to subsequent early-summer adult populations
(Chapter 2) within a blueberry ﬁeld, future research needs to investigate further the
relationship of the two life stages on different spatio-temporal scales. This information is
essential to understand the success of controlling adults through reducing the number of

larvae within an agricultural field. If this is effectively done, then the benefits of

103

disrupting oviposition could be two-fold: reducing the number of females that arrive to
the ﬁeld to oviposit and reduction of the following larval populations. This would also
mean that soil applied insecticides targeting the larvae could be used to reduce adult
density in the following summer. The larval density data in Chapter 2 provide useful
information about where to target soil applied insecticides in the ﬁeld. This information
could be made more explicit if there was additional data on the spatial distribution of
larvae and the change in distribution patterns over time. Currently we have no
information about the year-to-year variation in the location of larvae within a ﬁeld. Much
of the success of pest management on a ﬁeld scale is dependent upon the migration rates
of the targeted insect from the surrounding habitats. In the future, detailed study of the
immigration and emigration of Japanese beetle in and out of different types of habitats
will aid growers’ decisions on the use of different control strategies. This kind of
information is also useful in predicting infestation rates depending on the type of habitat
around the agricultural area.

In Chapter 3, I have investigated the interaction of cover crops and Japanese
beetle. The economic damage of this pest in blueberry is due to the presence of adults on
the blueberry bushes, so the next important step is to investigate how this interplay
between cover crops and beetles will affect the number of beetles on the harvested crop.
In this case the use of cover crops for reduction of crop damage is the goal, and this
cannot be achieved reliably without understanding the particular aspects of insect plant
interactions. Results in Chapter 3 give an indication of how some cover crop species
might affect overall Japanese beetle abundance within a ﬁeld, but the implications of

these ﬁndings could be controversial. First, if a cover crop is more attractive for the pest

104

than the main crop, such as was found for buckwheat, it could mean that the insects will
be attracted out of the valued crop, thus reducing damage. On the other hand, it is
possible that the higher overall abundance of the pest within the ﬁeld will result in
increased pest numbers on the bushes, and a higher level of damage and greater risk of
contamination. In the future, more species of cover crops should be tested for their
effectiveness to reduce Japanese beetle damage to blueberries. They also should be tested
at larger scales. When selecting a cover crop for pest management, the horticultural
aspects of adding a new plant into the agricultural ﬁeld must be given high priority, thus
the effect of the cover crops on the main crop has to be investigated. These effects can
sometimes be deleterious, such as drawing water and nutrients away form the main crop,
thus reducing yield, or it could change the microclimate within the ﬁeld so that it creates
ideal conditions for the development of diseases. In practice, weed control is one of the
main reasons for using cover crops, and so it would also be important to measure the
competitiveness of the chosen plant to weed species that are typically occurring in the
main crop.

Field observations of Japanese beetle behavior in Chapter 3 have provided
valuable information on how different plants change the behavior of Japanese beetle, or
alternatively, how Japanese beetle utilize the different plant species. Further
investigations are needed to examine what components of the plants act as cues for
initiating particular behaviors that are characteristic on a plant species. Behavioral
observations in the ﬁeld provide valuable information about the behavior of the insect in
its natural context, but to determine the effect of proximate factors inﬂuencing behavior,

controlled studies are needed in a seminatural or laboratory setting. These types of

105

experiments would allow the examination of the effect of the various components of the
environment, and also allow the use of a rigorous experimental design.

For a successful application of behavioral pest management in this system, it is
necessary to understand the mechanisms involved in ovipositional decision-making by
Japanese beetle, which lives as larva in the soil and above ground as an adult. The larvae
feed on plant material in the soil, whereas the adults feed on the above ground plant parts.
Using bioassay arenas, I was able to prove that Japanese beetle females show selectivity
in their oviposition decisions. I was interested to ﬁnd out if this decision is made based
on olfactory, tactile, or visual cues, or a combination of any of these (Chapter 4), because
this information is essential for successful behavioral manipulation. In Chapter 4, I show
the significance of visual and possibly tactile cues. The role of chemical cues from the
soil and the plants is yet to be determined, but this could be done through similar
bioassay experiments. The method developed in Chapter 4 for testing ovipositional
choice has been found appropriate to answer some fundamental questions about Japanese
beetle host selection, but other approaches to examine related research questions should
be used in the future, such as no-choice assays, or olfactometers. It would also be of
interest to investigate whether female preference is based on larval performance as has
been found in the case of other herbivores (J aenike 1978, Craig et al. 1989, Bemays and
Chapman 1994, Showler 2001). This would be useful when evaluating characters of an
agricultural field for predicting future infestation rates.

In my investigations described in this thesis, I kept in the forefront the practicality
of my ﬁndings for growers. I often presented my results at growers’ meetings and asked

for feedback from them about the nature and magnitude of the problem Japanese beetles

106

cause in highbush blueberry production and the control measures they use and would be
willing to use in their pest management regimes (Chapter 5). My goal was to interact with
growers to understand the needs and potential avenues in Japanese beetle control. I
therefore had to combine horticultural aspects of blueberry production with our basic
knowledge of Japanese beetle biology to make advancement in pest control through
behavioral pest management in this system. Cover crops seemed like an ideal ﬁt for this
complex problem, since some growers have already applied them for alternative reasons.
It is challenging to achieve a high level of control of Japanese beetle with non-chemical
methods, like behavioral manipulation, because the rates of immigration and pest
population levels are high in this species, but the incorporation of this new method into
current management practices could contribute to achieving lower crop damage levels.
Some of the current findings could be used to build a theoretical model using the
empirical data generated in this research. The integration of theoretical and empirical
approaches offers the possibility to unravel the processes underlying habitat management
strategies and generate hypotheses that can be tested in empirical studies. For example,
by combining the information in Chapter 3 with the information from the literature on
Japanese beetle biology, predictions can be made about how changing the plant
composition will alter the density of the different lifestages of Japanese beetle in an

agricultural field.

107

REFERENCES CITED

Ahmad S. 1982. Host location by the Japanese beetle: evidence for a key role for
olfaction in a highly polyphagous insect. Journal of Experimental Zoology 220: 1 17—20.

Allsopp, P.G., M.G. Klein, and EL. McCoy. 1992. Effect of soil moisture and texture
on oviposition by the Japanese beetle and rose chafer (Coleoptera: Sacarabaeidae).
Journal of Economic Entomology 85: 2194-2200.

Allsopp, RC. 1996. Japanese beetle, Popillia japonica Newman (Coleoptera:
Scarabaeidae): rate of movement and potential distribution of an immigrant pest species.
Coleopterist Bulletin 50:81-95.

Aluja, M., A. Jimenez, M. Camino, J. Pinero, L. Aldana, V. Castrejon, and ME.
Valdes. 1997. Habitat manipulation to reduce papaya fruit ﬂy (Diptera: Tephritidae)
damage: orchard design, use of trap crops and border trapping. Journal of Economic
Entomology 90: 1567—1576.

Andow, D.A. 1991. Yield loss to arthropods in vegetationally diverse agroecosystems.
Environmental Entomology 20: 1228-1235.

Anonymous. 1959. Japanese beetle control program. Annual report 1959. Michigan
Department of Agriculture.

Anonymous. 1962. Japanese beetle control program. Annual report 1962. Michigan
Department of Agriculture.

Anonymous. 1963. Japanese beetle control program. Annual report 1963. Michigan
Department of Agriculture.

Anonymous. 1992. Managing cover crops proﬁtably. Produced by Rodale Institute,
Emmaus, PA; Sustainable Agriculture Research and Education Program of CSRS, US.
Dept. of Agriculture, Washington, DC.

Anonymous. 2001. Pest Management Strategic Plan for the Michigan Blueberry
Industry. http://pestdata.ncsu.edu/pmsplpdf/MIblueberrypdf

Anonymous 2003. The National Agricultural Pest Information System Database.
http://www.ceris.purdue.edu/n_a_pis/pests/jb/itfacts.

Anonymous. 2005. National Agricultural Pest Information System
(http://www.ceris.purdue.edu/napis/pests/jb/imap/jb2004.html)

Awmack, C.S., and SR. Leather. 2002. Host plant quality and fecundity in herbivorous
insects. Annual Review of Entomology 47: 817-844.

108

Barrows, E.M., and G. Gordh. 1978. Sexual behavior in the Japanese beetle, Popillia
japonica, and comparative notes on sexual behavior of other scarabs (Coleoptera:
Scarabaeidae). Behavioral Biology 23:341-54.

Bernays, E.A., and R.F. Chapman. 1994. Host-plant selection by phytophagous insects.
Chapman and Hall, New York, NY.

Boetel, M.A., D.D. Walgenbach, G.L. Hein, B.W. Fuller, and M.E. Gray. 1992.
Oviposition site selection of the northern corn rootworm (Coleoptera: Chrysomelidae).
Journal of Economic Entomology 85: 246-249.

Bossart, J .L., and J .M. Scriber. 1999. Preference variation in the tiger swallowtail
butterﬂy (Lepidoptera: Papillionidae). Environmental Entomology 28: 628-637.

Bugg, R.L. 1992. Using cover crops to manage arthropods on truck farms. HortScience
27: 741-745.

Bugg, R.L., and C. Waddington. 1994. Using cover crOps to manage arthropod pests of
orchards: A review. Agriculture, Ecosystems and Environment 50:11-28.

Casida, J .E., and G.B.Quistad. 1995. Pyrethrum ﬂowers: production, chemistry,
toxicology, and uses. Oxford University Press, New York.

Cappaert, D.L., and D.R. Smitley. 2002. Parasitoids and pathogens of Japanese beetle

(Coleoptera:Scarabaeidae) in southern Michigan. Environmental Entomology 31: 773-
580.

Choo, H.Y., D.W. Lee, J .W. Park, H.K. Kaya, D.R. Smitley, S.M. Lee, and Y.M.
Choo. 2002. Life history and spatial distribution of the oriental beetle (Coleoptera:
Scarabaeidae) in golf courses in Korea. Journal of Economic Entomology 95: 72-80.

Craig, T.P., J .K. Itami, and P.W. Price. 1989. A strong relationship between
oviposition preference and larval performance in a shoot-galling sawﬂy. Ecology 70:
1691-1699.

Crutchfield, B.A., and D.A. Potter. 1994. Preference of Japanese beetle and southern
masked chafer (Coleoptera: Scarabaeidae) grubs among cool-season turfgrasses. Journal
of Entomological Science 29:398-406.

Crutchfield, B.A., and D.A. Potter. 1995. Feeding by Japanese beetle and southern
masked chafer grubs on lawn weeds. Crop Science 35: 1681-1684.

Dalthorp, D., J. Nyrop, and M.G. Villani. 1999. Estimation of local mean population

densities of Japanese beetle grubs (Scarabaeidae: Coleoptera). Environmental
Entomology 28:255-65.

109

Dalthorp, D., J. Nyrop, and M.G. Villani. 2000. Spatial ecology of the Japanese beetle,
Popillia japonica. Entomolgia Experimentalis et Applicata 96: 129-139.

Doussard, R.E., and T. Eisner. 1987. Vein cutting behavior: insect counterploy to the
latex defense of plants. Science 237: 898-901.

Finch, S., and R.H. Collier. 2000. Host-plant selection by insects — a theory based on
‘appropriate/inappropriate landings’ by pest insects of cruciferous plants. Entomologia
Experimentalis et Applicata 96: 91-102.

Fleming, W.E. 1969. Attractants for the Japanese beetle. USDA Technical Bulletin N o.
1399. US. Government Printing Ofﬁce, Washington, DC.

Fleming, W.E. 1968. Biological control of the Japanese beetle. USDA Technical
Bulletin No. 1383. US. Government Printing Office, Washington, DC.

Fleming, W.E. 1972. Biology of the Japanese beetle. USDA Technical Bulletin No.
1449. US. Government Printing Ofﬁce, Washington DC.

Fleming, W.E. 1976. Integrating control of the Japanese beetle — a historical review.
USDA Technical Bulletin No. 1545. US. Government Printing Ofﬁce, Washington DC.

Flint, M.L., and P. Gouveia. 2001. IPM in practice. Principles and methods of
integrated pest management. University of California. Statewide integrated pest
management project. Agriculture and Natural Resources, Publication 3418.

Foster, S.P., and MD. Harris. 1997. Behavioral manipulation methods for insect pest-
management. Annual Review of Entomology. 42: 123-146.

Foster, S.P., and AJ. Howard. 1998. Inﬂuence of stimuli from Camellia japonica on
ovipositional behavior of generalist herbivore Epiphyas postvittana. Journal of Chemical
Ecology 24: 1251-1275.

Freese, G., and H. Zwolfer. 1996. The problem of optimal clutch size in a tritrophic
system: The ovipositional strategy of the thistle gallﬂy Urophora cardui (Diptera:
Tephritidae). Oecologia 108: 293-302.

Funderburk, J .E., D.L. Wright, and LB. Teare. 1990. Preplant tillage effects on
population dynamics of soybean insect pests. Crop Science 30: 686-690.

Goonewardene, H.F., and J .E. McKay. 1969. An artificial diet for the adult Japanese
beetle. Journal of Economic Entomology 62: 964.

Grewal, P.S., K.T. Power, and D.,]. Shetlar. 2001. Neonecotinoid insecticides alter
diapause behavior and survival of overwintering white grubs (Coleoptera: Scarabaeidae).
Pest Management Science 57: 852-857.

110

Gurr, G.M., S.D. Wratten, and M. Altieri. 2004. Ecological Engineering for Pest
Management: Advances in Habitat Manipulation for Pest Management. CSIRO
Publishing, Melbourne, Australia.

Hanula, J .L. 1990. Epizootiological investigations of the microsporidian Ovavesicula
popilliae and bacterium Bacillus popilliae in ﬁeld populations of the Japanese beetle
(Coleoptera: Scarabaeidae). Environmental Entomology 19: 1552-57.

Harris, M.O., and J.R. Miller. 1984. Foliar form inﬂuences ovipositional behaviour of
the onion ﬂy. Physiological Entomology 9: 145-155.

Hartwig, N.L., and H.U. Amman. 2002. Cover crops in living mulches. 50th
Anniversary- Invited Article. Weed Science 50: 688-699.

Held, D.W. and, D.A. Potter. 2004. Floral characteristics affect susceptibility of hybrid
tea roses, Rosa x hybrida, to Japanese beetle (Coleoptera: Scarabaeidae). Journal of
Economic Entomology 97: 353-360.

Isaacs, R., Z. Szendrei, and J. Wise. 2004. Evaluation of new approaches for
management of Japanese beetle in hi ghbush blueberries. Small Fruit Review 3: 349-360.

Jaenike, J. 1978. On optimal oviposition behavior in phytophagous insects. Theoretical
Population Biology 14: 350-356.

Kard, B.M.R., and F.P. Hain. 1988. Inﬂuence of ground covers on white grub
(Coleoptera: Scarabaeidae) populations and their feeding damage to roots of Fraser ﬁr
Christmas trees in the Southern Appalachians. Environmenatal Entomology 17, 63-66.

King, P.D., C.F. Mercer, and J .S. Meekings. 1981. Ecology of black beetle,
Heteronychus arator (Coleoptera: Scarabaeidae) — inﬂuence of pasture species on
oviposition site preference. New Zealand Journal of Zoology 8: 119-122.

Kleweno, D.D., and V. Matthews. 2002. Michigan Agricultural Statistics 2001-2002.
Michigan Agricultural Statistics Service, Lansing, Michigan.

Kleweno, DD. and V. Matthews. 2004. Michigan Agricultural Statistics 2003-2004.
Michigan Department of Agriculture 2003 Annual Report. Michigan Department of
Agriculture and USDA, National Agricultural Statistics Service. Michigan Agricultural
Statistics, Lansing, MI.

Koppenhiiffer, A.M., M. Wilson, I. Brown, H.K. Kaya, and R. Gaugler. 2000a.
Biological control agents for white grubs (Coleoptera:Scarabaeidae) in anticipation of the
establishment of the Japanese beetle in California. Journal of Economic Entomology 93:
7 1-80.

111

Koppenhiiffer, A.M., P.S. Grewal, and H.K. Kaya. 2000b. Synergism of imidacloprid

and entomopathogenic nematodes against white grubs: the mechanism. Entomologia
Experimentalis et Applicata 94: 283-293.

Koppenhiiffer, A.M. and EM. Fuzy. 2003. Ecological characterization of Steinemema
scarabaei, a scarab-adapted entomophatogenic nematode from New Jersey. Journal of
Invertebrate Pathology 83: 139-148.

Kostal, V. and S. Finch. 1994. Inﬂuence of background on host plant selection and
subsequent oviposition by the cabbage root ﬂy (Delia radicum). Entomologia
Experimentalis et Applicata 70: 153-163.

Lacey, L.A., J .J . Amara], J. Coupland, and M.G. Klein. 1994. The inﬂuence of
climatic factors on the ﬂight activity of the Japanese beetle (Coleoptera: Scarabaeidae):
Implications for use of a microbial agent. Biological Control 4: 298-303.

Ladd, T.L. Jr. 1966. Egg viability and longevity of Japanese beetles treated with tepa,
apholate and metepa. Journal of Economic Entomology 59:422-425.

Ladd, T.L. Jr. 1970. Sex attraction in the Japanese beetle. Journal of Economic
Entomology 63:905-908.

Ladd, T.L. Jr. 1986. Inﬂuence of sugars on the feeding response of Japanese beetles
(Coleoptera: Scarabaeidae). Journal of Economic Entomology 79: 668-71.

Ladd, T.L. Jr. 1987. Japanese beetle (Coleoptera: Scarabaeidae): inﬂuence of favored
food plants on feeding response. Journal of Economic Entomology 80: 1014-1017.

Leal, W.S. 1998. Chemical ecology of phytophagous scarab beetles. Annual Review of
Entomology 43:39-61.

Ladd, T.L. J r., and T.P. McGovern. 1980. Japanese beetle: a superior attractant,
phenethyl propionate + eugenol + geraniol 3:7:3. Journal of Economic Entomology
73:689-691.

Loughrin, J .H., D.A. Potter, and T.R. Hamilton-Kemp. 1995. Volatile compounds
induced by herbivory act as aggregation kairomones for the Japanese beetle (Popillia
japonica Newman). Journal of Chemical Ecology 21:1457—67.

Loughrin, J .H., D.A. Potter, T.R. Hamilton-Kemp, and M.E. Byers. 1996. Volatile
compounds from crabapple cultivars (Malus spp.) differing in susceptibility to the
Japanese beetle (Popillia japonica Newman). Journal of Chemical Ecology 22:1295—305.

Loughrin, J .H., D.A. Potter, T.R. Hamilton-Kemp, and M.E. Byers. 1997. Response
of Japanese beetles (Coleoptera: Scarabaeidae) to leaf volatiles of susceptible and
resistant maple species. Environmental Entomology 26:334—42.

112

Mannion, C.M., W. McLane, M.G. Klein, J. Moyseenko, J .3. Oliver, and D. Cowan.
2001. Management of early-instar Japanese beetle (Coleoptera: Scarabaeidae) in ﬁeld-
grown nursery crops. Journal of Economic Entomology 94: 1151-1161.

Matthews, G.A. 1984. Pest management. Lon gman, New York. p. 50.

Miller, J .R., and KL. Strickler. 1984. Finding and accepting host plants, pp.127-157.
In W.J. Bell, and R.J. Carde’ (eds), Chemical ecology of insects. Chapman and Hall Ltd.,
New York, NY.

Nikonov, A.A., and W.S. Leal. 2002. Peripheral coding of sex pheromone and a
behavioral antagonist in the Japanese beetle, Popillia japonica. Journal of Chemical
Ecology 28:1075-1089.

Orians, C.M., C.H. Huang, A. Wild, K.A. Dorfman, and P. Zee. 1997. Willow
hybridization differentially affects preference and performance of herbivorous beetles.
Entomologia Experimentalis et Applicata 83: 285-94.

Opp, 8.8., and R.J. Prokopy. 1986. Approaches and methods for direct behavioral
observation and analysis of plant-insect interactions, pp. 1-22. In J .R. Miller and TA.
Miller (eds), Insect-Plant Interactions. Springer-Verlag New York Inc., New York, NY.

Ott, R.L., and M. Longnecker. 2001. Statistical methods and data analysis. Fifth
edition. Thomson Learning Ltd. USA.

Potter, D.A., C.G. Patterson, and CT. Redmond. 1992. Inﬂuence of turfgrass species
and tall fescue endophyte on feeding ecology of Japanese beetle and southern masked
chafer grubs (Coleoptera: Scarabaeidae). Journal of Economic Entomology 85: 900-909.

Potter, D.A., A.J. Powell, P.G. Spicer, and D.W. Williams. 1996. Cultural practices
affecting root-feeding white grubs (Coleoptera Scarabaeidae) in turfgrass. Journal of
Economic Entomology 89:156-164.

Potter, D.A., and D.W. Held. 2002. Biology and management of the Japanese beetle.
Annual Review of Entomology 47:175-205.

Pritts, M.P., and J .F. Hancock. 1992. Hi ghbush Blueberry Production Guide. NRAES-
55, Northeast Regional Agricultural Engineering Service, Cooperative Extension, Ithaca,
New York.

Prokopy, R.J., and ED. Owens. 1983. Visual detection of plants by herbivorous
insects. Annual Review of Entomology. 28: 337-364.

Prokopy, R.J. 1994. Integration in orchard pest and habitat management: A review.
Agriculture, Ecosystems and Environment 50:1-10.

113

Reed, D.K., B. Freedman, and T.L. Ladd, Jr. 1982. Insecticidal and antifeedant
activity of nertifolin against codling moth, striped cucumber beetle, and Japanese beetle.
Journal of Economic Entomology 75: 1093-96.

Régniére, J ., R.L. Rabb, and RE. Stinner. 1979. Popillia japonica (Coleoptera:
Scarabaeidae): a mathematical model of oviposition in heterogeneous agroecosystems.
Canadian Entomologist 111: 1271-1280.

Régniére J., R.L. Rabb, and RE. Stinner. 1981a. Popillia japonica: simulation of
temperature-dependent development of the immatures, and prediction of adult
emergence. Environmental Entomology 10: 290-296.

Régniére, J ., R.L. Rabb, and RE. Stinner. 1981b. Popillia japonica: Effect of soil
moisture and texture on survival and development of eggs and ﬁrst instar grubs.
Environmental Entomology 10: 654-660.

Renwick, J.A.A. 1983. Nonpreference mechanisms: plant characteristics inﬂuencing
insect behavior. In: P.A. Hedin (ed.), Plant resistance to Insects. American Chemical
Society, Washington DC, pp. 199-213.

Renwick, J.A.A. 1989. Chemical ecology of oviposition in phytophagous insects.
Experientia 45: 223-228.

Risch, SJ. 1981. Insect herbivore abundance in tropical monocultures and polycultures:
an experimental test of two hypotheses. Ecology 62: 1325-1340.

Risch, S.J., D Andow, and M.A. Altieri. 1983. Agroecosystem diversity and pest
control: data, tentative conclusions, and new research directions. Environmental
Entomology 12: 625-629.

Roitberg, B.D., LA. Robertson, and J .G.A. Tyerman. 1999. Vive la variance: a
functional oviposition theory for insect herbivores. Entomologia Experimentalis et
Applicata 91: 187-194.

Rondon, S. I., and M.E. Gray. 2004. Ovarian development and ovipositional preference
of the western corn rootworm (Coleoptera: Chrysomelidae) variant in east central Illinois.
Journal of Economic Entomology 97: 390-396.

Root, R.B. 1973. Organization of a plant-arthropod association in simple and diverse
habitats: the fauna of collards (Brassica oleracea). Ecological Monographs 43: 95-124.

Root, R.B., and RM. Kareiva. 1984. The search for resources by cabbage butterﬂies

(Pieris rapae): ecological consequences and adaptive signiﬁcance of Markovian
movements in patchy environments. Ecology 65: 147-165.

114

SAS Institute. 2001. Version 8. SAS Institute Inc., Cary, NC.

Scheerens, J .C., and G.L. Brenneman. 1990. Survey of Ohio strawberry growers:
present practice and future directions. Research circular - Ohio Agricultural Research and
Development Center 297: 39-49.

Schoonhoven, L.M., T. Jermy, and J .J.A. van Loon. 1998. Insect-Plant biology: From
physiology to evolution. Chapman & Hall Ltd., London, pp. 146-148.

Seger, J., and HJ. Brockman. 1987. What is bet hedging? Oxford Surveys in
Evolutionary Biology 4: 182-211.

Showler, A.T. 2001. Spodoptera exigua oviposition and larval feeding preferences for
pi gweed, Amaranthus hybridus, over squaring cotton, Gossypium hirsutum, and a
comparison of free amino acids in each host plant. Journal of Chemical Ecology 27:
2013-2028.

Singer, M.C. 1986. The deﬁnition and measurement of oviposition preference in plant
feeding insects, pp. 65-94. In J .R. Miller and TA. Miller (eds), Insect-plant interactions.
Springer-Verlag, New York, NY.

Smitley, D.R. 1996. Incidence of Popillia japonica (Coleoptera: Scarabaeidae) and other
scarab larvae in nursery fields. Journal of Economic Entomology 89: 1262-66.

Stivers-Young, L.J., and F .A. Tucker. 1999. Cover-cropping practices of vegetable
producers in western New York. HortTechnology 9: 459-465.

Szendrei, Z., N. Mallampalli, and R. Isaacs, (in press). Effect of cultivation on
abundance of Japanese beetle, Popillia japonica Newman (Coleoptera: Scarabaeidae), in
highbush blueberry ﬁelds. Journal of Applied Entomology.

Szendrei, Z., and R. Isaacs, (in press). Do plant cues inﬂuence the oviposition behavior
of Japanese beetle? Entomologia Expreimentalis et Applicata.

Tada, S., and W.S. Leal. 1997. Localization and morphology of sex pheromone glands
in scarab beetles. Journal of Chemical Ecology 23:903-15.

Tallamy, D.W., J. Stull, N.P. Ehresman, P.M. Gorski, C.E. Mason. 1997.
Cucurbitacins as feeding and oviposition deterrents to insects. Environmental

Entomology 26: 678-83.

Thorsteinson, AJ. 1960. Host selection in phytophagous insects. Annual Review of
Entomology 5: 193-218.

115

Tonhasca, A., Jr., and B.R. Stinner. 1991. Effects of strip intercropping and no-tillage
on some pests and beneficial invertebrates of corn in Ohio. Environ. Entomol. 20: 1251-
1258.

Tonhasca A. J r., and D.N. Byrne. 1994. The effects of crop diversification on
herbivorous insects: a meta-analysis approach. Ecological Entomology 19:239-244.

Tumlinson, J .H., M.G. Klein, R.E. Doolittle, T.L. Lass, and A.T. Proveaux. 1977.
Identiﬁcation of the female Japanese beetle sex pheromone: inhibition of male response
by an enantiomer. Science 197: 789-792.

Vanderrneer, J. 1989. The ecology of intercropping. Cambridge University Press,
Cambridge, 237 pp.

Van Timmeren, S.J, P.V. Switzer, and KC. Kruse. 2000. Emergence and reproductive
patterns in the Japanese beetle, Popillia japonica (Coleoptera: Scarabaeidae). Journal of
the Kansas Entomological Society 74: 17-24.

Wiklund, C. 1981. Generalist vs. specialist oviposition behaviour in Popillia machaon
(Lepidoptera) and functional aspects on the hierarchy of oviposition preferences. Oikos
36: 163-170.

Villani, M.G., S.R. Krueger, P.C. Schroeder, F. Consolie, N.H. Consolie, L.M.
Preston-Wilsey, and D.W. Roberts, 1994. Soil application effects of Metarhizium
anisopliae on Japanese beetle (Coleoptera: Scarabaeidae) behavior and survival in
turfgrass microcosms. Environmental Entomology 23: 502-513.

Vittum, PJ. 1986. Biology of the Japanese beetle (Coleoptera: Scarabaeidae) in Eastern
Massachusetts. Journal of Economic Entomology 79: 387-391.

Vittum, PJ., M.G. Villani, and H. Tashiro. 1999. Turfgrass insects of the United States
and Canada 2“(1 edition. Cornell University Press, Ithaca, NY.

Wiklund, C. 1981. Generalist vs. specialist oviposition behaviour in Papillio machaon
(Lepidoptera) and functional aspects on the hierarchy of oviposition preferences. Oikos
36: 163-170.

Wise, J .C.; L.J. Gut, R. Isaacs, A.M.C. Schilder, B. Zandstra, E. Hanson, and B.
Shane. 2003. 2004 Michigan Fruit Management Guide. Bulletin E-154, Michigan State
University Extension, East Lansing, Michigan.

Visser, J .H. 1988. Host-plant ﬁnding by insects: orientation, sensory input and search
patterns. Journal of Insect Physiology 34: 259-268.

116

Wojtasek, H., B.S. Hansson, and W.S. Leal. 1998. Attracted or repelled? A matter of
two neurons, one pheromone binding protein, and a chiral center. Biochemical and
Biophysical Research Communications 250:217—22.

117

APPENDICES

118

APPENDIX 1

Survey of Japanese Beetle Management Methods

 

 

This survey has been prepared to gather information about Japanese beetle and
strategies for its control in blueberries. Your answers can be anonymous, but the survey
results will be sent to respondents who provided their name and address. All information
that is provided is treated as confidential. No information from individual replies will ever
be used in a public way, and results will be shared with the industry at future MSU
extension meetings.

Thanks for you time,

 

 

Dr. Rufus Isaacs and Zsofia Szendrei, MSU Small Fruit Entomology

You indicate your voluntary agreement to participate by completing and returning this
questionnaire.

Completed surveys should be mailed to: Small Fruit Entomology, 202 CIPS,
Michigan State University, East Lansing, MI 48824, using the enclosed envelope.

Name*: 'Name and address are

 

optional

Address*:

 

Counties where your blueberry farms are located:

 

Total area of commercial blueberry production: acres

 

 

1. Is Japanese beetle a pest on your farm? Yes No

 

 

 

 

 

 

2. If Japanese beetle is a pest, how important is this insect in your blueberry
production?

Oextremely Qvery Osomewhat Oslightly Onot

3. What do you estimate the economic cost of Japanese beetle to be per year on
your farm?

 

$ per year
4. Is the economic impact of Japanese beetle due to:
Elload rejections I:llower fruit price 1:! increased insecticide use
Other

 

119

5. What method(s) do you use to control Japanese beetle?
Check all circles that apply. Rank them in order of importance: 1: most important

Insecticide, foliar application

 

Insecticide, soil application

 

Cover crops

 

Clean cultivation

 

Herbicide killed sod in row-middles
Other:

 

OOOO

 

 

6. What row-middle ground cover do you use at your farm?
Check all circles that apply. °/o Of acres

Permanent sod

 

Clean cultivation
Herbicide killed sod
Other:

 

 

 

 

7. Have you changed row-middle management practices to help control
Japanese beetll:

 

 

 

 

 

Yes No

If YES, How many years ago did you change?

 

What changes did you make?

 

 

 

 

 

 

 

 

1- av the changes reduced Japanese beetle pressure? Yes No
Don’t know

 

 

 

 

 

 

8. Would you be interested in using new ground covers if they were effective
against Japanese beet Yes No Don’t
know

 

 

 

 

 

 

 

 

 

 

 

 

9. What do you think are the possible advantages or disadvantages of cover

crops?
Advantages:
:1 Controls pests Disadvantages: :1 Too expensive
[:1 Prevents soil erosion l:l Difficult to seed
|:l Conserves soil moisture CI] Competition with bushes
1:] Cuts fertilizer costs [:1 Alternative host for pests

120

10. What do you think are the possible advantages or disadvantages of clean
cultivation?

 

Advantages: Disadvantaqes:
:lControls pests Cl Soil erosion
1:] Low cost of maintenance l:l Soil moisture loss
l:l Easy to manage :1 Need for herbicides
EjDecreases frost damage [:1 Mud and dust in field
11. What kind of irrigation system do you have?
Overhead
Check all circles that apply. 0 Drip

 

 

 

 

 

 

 

12. Do you cultivate the soil around your fields? Yes No

 

 

13. Are there any changes to your Japanese beetle management strategy that
you are planning for the future?

 

 

14. What other pest management problems do you have that are equally or more
important than the Japanese beetle?

 

If you have any questions regarding this survey itself, please contact Rufus Isaacs at
(517) 355-6619.

If you have any questions regarding your participation as a subject in this research, please contact:

Ashir Kumar, UCHRIS, 202 Olds Hall, Michigan State University, East Lansing, MI 48824-1046

121

APPENDIX 2

September 24, 2003

TO: Rufus ISAACS
202 CIPS

RE: mart 03-706 CATEGORY: EXEMPT 1-2

APPROVAL DATE: September 22, 2003
EXPIRATION DATEzAugust 22, 2004

TITLE: SURVEY OF JAPANESE BEETLE MANAGEMENT METHODS

The University Committee on Research Involving Human Subjects' (UCRIHS) review of this
project ls complete and I am pleased to advise that the rights and welfare of the human
subjects appear to be adequately protected and methods to obtain informed consent are
appropriate. Therefore, the UCRIHS approved this project.

RENEWALS:‘ UCRIHS approval is valid until the expiration date . listed above. Projects
continuing beyond this date must be renewed With the renewal form. A maximum of four such
expedited renewals are possible. Investigators wishing to continue a project beyond that time
need to submit a 5-year application for a complete review. '

REVISIONS: UCRIHS must review any changes In procedures involving human subjects, prior
to Initiation of the change. If this is done at the'time of renewal, please include a revision form
with the renewal. To revise an approved protocol at any other time during the year, send your
written request "with an attached revision cover sheet to the UCRIHS Chair, requesting revised
approval and referencing the project's IRB# and title. Include in your request a description of
the change and any revised instruments, consent forms or advertisements that are applicable.
PROBLEMSICHANGES: Should either of the following arise during the course of the work.
notify UCRIHS promptly: 1) problems (unexpected side effects, complaints, etc.) involving
human subjects or 2) changes In the research environment or new Information indicating
greater risk to the human subjects than existed when the protocol was previously reviewed and
approved.

If we can be of further assistance, please contact us at (517) 355-2180 or via email:

UCRIHS@msu. edu. Please note that all UCRIHS forms are located on the web:
http://www. humanresearch. msu. edu

Sincerely,
roan/«Xi

Peter Vasilenko Ill, Ph.D.
UCRIHS Chair

PV: jm

cc: Zsotia Szendrei
202 CIPS

APPENDIX 3

Record of Deposition of Voucher Specimens*

The specimens listed on the following sheet(s) have been deposited in the named
museum(s) as samples of those species or other taxa, which were used in this research.
Voucher recognition labels bearing the Voucher No. have been attached or included in
ﬂuid-preserved specimens.
Voucher No.: 2005-02
Title of thesis or dissertation (or other research projects):
POTENTIAL FOR BEHAVIORAL AND CULTURAL MANIPULATION OF
JAPANESE BEETLE (POPILLIA JAPONICA NEWMAN, COLEOPTERA:
SCARABAEIDAE) POPULATIONS
Museum(s) where deposited and abbreviations for table on following sheets:

Entomology Museum, Michigan State University (MSU)

Other Museums:

Investigator’s Name(s) (typed)
Zsoﬁa Szendrei

 

 

Date April 19. 2005

*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.

123

 

Appendix 3.1

Voucher Specimen Data

Pages

1

of

Page 1

\n

 

€24.25 kw

  

cm 22.2 as 5 88%
22 mcoEBoaw new: m>onm 9.: 8289”.

 

mo-._m<-m F San.

 

 

 

No-moow .oz Lmgozo>

86:on mcomN
63.8 @oEmz 38339:.
35889. = 93% 35:63 33

 

668% .N 58 22.
mag ocoESoca E
OEZF 6:356”.

 

 

 

 

 

 

 

 

 

 

 

 

 

3w: 3 m: .00 cmao=< .5 52.262 8.220qu £5..qu
m m 10 ,9 m e oaoao cm om:
w m m a m .lst m o. a s u a. on U :oxmtocaoco 86on
w e o. h u u D m w 9 no 88260 mcoEanwcoaﬁmgonS
h e t d d u V. a 9
M w d O A A P N L E
coconEnz

124

 

1111111111111