Biological Control of the Soybean Aphid, Aphis Glycines Matsumura

a 3.

am...

.1.

3.x?

.. ,

1.4.“:

r,» I This is to certify that the

thesis entitled

BIOLOGICAL CONTROL OF THE SOYBEAN APHID, APHIS GLYCINES MATSUMURA

(HOMOPTERA: APHIDIDAE)

presented by

TYLER BRYCE FOX

has been accepted towards fulﬁllment of the requirements for.

M. S. degree in Entomology

Major professor

Date 13 December, 2002

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution UBRARY Michigan State University

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

DATE DUE DATE DUE DATE DUE

JQMQZW

[JCT @4207

0928 08.

6/01 cyclRC/DatoDuo.p65

Tyler Bryce Fox

A THESIS

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

MASTER OF SCIENCE

Department of Entomology

2002 ABSTRACT

BIOLOGICAL CONTROL OF THE SOYBEAN APHID, APHIS GLYCINES MATSUMURA (HOMOPTERA: APHIDIDAE)

Tyler Bryce Fox

Aphis glycines Matsumura is an invasive pest of soybean [Glycine max (L.) Merrill] ﬁrst

discovered in the United States in 2000. The impact of predation on A. glycines

establishment and population increase was studied in Ingham County, Michigan during

2001-2002. Direct observation and pitfall traps were utilized to determine the species

composition and abundance of potential predators in soybean ﬁelds. Laboratory feeding

assays showed that most of the common foliar-foraging and ground-dwelling predators

present in soybean ﬁelds were capable of consuming A. glycines. The impact of

predation on A. glycines establishment in the ﬁeld was examined using clip cages.

Predation signiﬁcantly reduced A. glycines 24 h survival in four out of six trials over both

years. In a ﬁeld study during July-August in 2001- 2002, l m2 predator exclusion or

sham cages were used to assess predation impacts on established A. glycines populations.

Cages were initially infested with A. glycines adults and sampled for approximately four

weeks. The 2001 trial was inconclusive, however, in 2002 foliar-foraging predators prevented population growth in sham cages and dramatically reduced A. glycines populations where exclusion cages were opened after two weeks. These studies demonstrate that predator communities can signiﬁcantly impact A. glycines establishment and population density in soybean and should be further investigated as a potential means of managing A. glycines populations in United States soybean production systems. Dedicated with love to my wife Mari Igarashi Fox, for always loving, encouraging and supporting me while I was pursuing this degree, and to Naotsugu Igarashi for becoming a

second father and a best friend.

iii ACKNOWLEDGEMENTS

I must ﬁrst thank Doug Landis, for showing me that science is a challenge worth

investigating. I greatly appreciate his dedication to investigate unique interactions within

the insect world, and his willingness to share his wealth knowledge with others. I thank

Chris DiFonzo for always assisting me with my numerous aphid-related questions. I also

thank Suzanne Thiem and Kurt Thelen for their guidance during every stage of my

experiments over the past two years. I also thank Fabian Menalled for his guidance with

another project. Fernando Cardoso of the Statistical Consulting Center also deserves and enormous amount of thanks for his unquestionable devotion to helping me with statistical questions throughout these studies.

I am indebted to Chris Sebolt for his ﬁeld assistance and friendship. I also thank my

fellow lab mates Matt O’Neal, Tammy Wilkinson, and Alejandro Costamagna for their friendship, understanding, and encouragement at all times. I also thank Sandra Clay,

Alison Gould, Andrea McMillian, Andre Ball, Meghan Burns, Michelle Smith, Christy

Hemming, Kathy McCamant, Allison Lewinski, Ben Nessia, and Kevin Newhouse for their help with ﬁeldwork. Collectively, these folks probably counted more aphids than there are stars in the sky.

iv I also thank the Niles APHIS lab for supplying live aphids for my studies. Early season research would have been nearly impossible without the use of their colony. This work funded by a Cooperative Agreement with USDA APHIS PPQ.

I am greatly indebted to my family for their overwhelming support and encouragement. I am very fortunate to have two loving sets of parents in both the United States (Dan

“Fodge” and Cheryl Fox) and Japan (Naotsugu and Kiyo Igarashi) who showed their love and devotion at all times. TABLE OF CONTENTS

LIST OF TABLES viii

LIST OF FIGURES xii

KEY TO SYMBOLS AND ABBREVIATIONS xvi

CHAPTER 1 Introduction to soybean production, Aphis glycines Matsumura biology and the importance of biological control

List of References 14

CHAPTER 2 The impact of indigenous predators on preventing establishment of the soybean aphid, Aphis glycines Matsumura

Introduction 20

Materials and Methods 23

Results 28

Discussion 34

List of References 64

CHAPTER 3 Evaluating the impact of indigenous and exotic predators of the soybean aphid, Aphis glycines Matsumura, in laboratory no-choice feeding assays

Introduction 67

Materials and Methods 69

Results 71

Discussion 73

List of References 8O

CHAPTER 4 Predator effects on soybean aphid, Aphis glycines Matsumura, population growth

Introduction 82

Materials and Methods 85

vi Results 89

Discussion 97

List of References 145

APPENDIX A Effects of soybean varieties on Aphis glycines Matsumura 148 survival and reproduction

APPENDIX B Record of deposition of voucher specimens 154

vii LIST OF TABLES

Table 1.1. Soybean management strategies in four counties of Michigan. Rotation percentages show the percentage of farmers that use each rotation within the three counties. Tillage percentages show the percentage of farmers within the three counties that use particular tillage practices. Row spacing percentages show the percent of farmers that use speciﬁc row spacing within each county. All surveys were conducted between April 23 and March 13, 2001 13

Table 2.1. Species composition, total number and percent of total potential A. glycines predators in ﬁve-minute observations in l m x 30 cm areas of soybean during trial one (7-8 June 2001), trial two (14-15 June 2001), and trial three (24- 25 June 2001), East Lansing, Michigan 39

Table 2.2. Species composition, total number, and percent of total potential A. glycines predators in three-minute observations in l m x 30 cm areas of soybean during trial one (18-19 June 2002), trial two (24-25 June 2002), and trial three (2-3 June 2002), East Lansing, Michigan 40

Table 2.3. Species composition, total number, and percent of total of predators collected in pitfall traps in refuge and control plots during trial one (7-8 June 2001), trial two (14-15 June 2001), and trial three (24-25 June 2001), East Lansing, Michigan 41

Table 2.4. Species composition, total number, and percent of predators collected in pitfall traps during trial one (18-19 June 2002), trial two (24-25 June 2002), and trial three (2-3 July 2002), East Lansing, Michigan 42

Table 2.5. Probability of a greater F -statistic based on the proportion of live adult A. glycines remaining in refuge and control plots under different cage treatments during 2001, East Lansing, Michigan 43

Table 2.6. Probability of a greater F -statistic based on the number of A. glycines nymphs produced per cage in refuge and control plots under different cage treatments during 2001, East Lansing, Michigan

Table 2.7. Probability of a greater F -statistic based on the total number (adults + nymphs) of live A. glycines in refuge and control plots under different cage treatments during 2001, East Lansing, Michigan 45

Table 2.8. Probability of a greater F -statistic based on the proportion of live adult A. glycines remaining under different cage treatments during 2002, East Lansing, Michigan 46

viii Table 2.9. Probability of a greater F —statistic based on the number of A. glycines nymphs produced per cage in soybean plots under different cage treatments during 2002, East Lansing, Michigan 47

Table 2.10. Probability of a greater F -statistic based on the total number (adults + nymphs) of live A. glycines under different cage treatments during 2002, East Lansing, Michigan 48

Table 3.1. Number of A. glycines surviving and percent mortality of A. glycines in 24 h no-choice feeding trials consumed by foliar-foraging predators from 8 June to 10 August 2001, East Lansing, Michigan 76

Table 3.2. Number of A. glycines surviving and percent mortality of A. glycines in 24 h no-choice feeding trials consumed by ground-dwelling predators from 8 June to 10 August 2001, East Lansing, Michigan 77

Table 3.3. Estimated seasonal occurrence of foliar-foraging predators during 2001, East Lansing, Michigan 78

Table 3.4. Estimated seasonal occurrence of ground-dwelling predators during 2001, East Lansing, Michigan 79

Table 4.1. Species composition, total, and percent predators observed during 3- minute non-intrusive and intrusive observations in exclusion and open cages during 2001. Exclusion cages were sampled once a week from 26 June to 31 July and open cages were sampled from 26 June to 4 September, East Lansing, Michigan 102

Table 4.2. Species composition, total, and percent ground dwelling predators captured per seven days in exclusion cages versus those in the surrounding ﬁeld during 2001. Pitfalls were sampled once a week from 26 June to 31 July, 2001, East Lansing, Michigan 103

Table 4.3. Probability of a greater F -statistic based on results of temperature difference (°C interior minus oC exterior) of open, exclusion, and frame cages during trial one, 2002 104

Table 4.4. Probability of a greater F -statistic based on results of temperature difference (°C interior minus °C exterior) of open, exclusion, and frame cages during trial two, 2002 105

Table 4.5. Probability of a greater F -statistic based on results of relative humidity difference (interior minus exterior) of open, exclusion, and frame cages during trial one, 2002 106

ix Table 4.6. Probability of a greater F -statistic based on results of relative humidity difference (interior minus exterior) of open, exclusion, and frame cages during trial two, 2002 107

Table 4.7. Probability of a greater F -statistic based on results of plant height (based on measurement of ﬁve random plants) in exclusion, open, and frame cages during trial one, 2002 108

Table 4.8. Probability of a greater F -statistic based on plant height (based on measurement of ﬁve random plants) in exclusion, open, and frame cages during trial two, 2002 109

Table 4.9. Species composition, total, and percent of ground dwelling predators captured per seven days in exclusion cages (2 pitfall traps/cage) versus those placed randomly throughout the study area in trial one (28 June to 29 July) and trial two (12 July to 12 August), 2002, East Lansing, Michigan 110

Table 4.10. Probabilities of a greater F -statistic based on results of total pitfall captures of ground-dwelling predators in exclusion cages versus those in the study area (plot) during trial one, 2002 111

Table 4.11. Probability of a greater F -statistic based on results of total pitfall captures of ground-dwelling predators in exclusion cages versus those in the study area (plot) during trial two, 2002 112

Table 4.12. Probability of a greater F -statistic based the number of adult A. glycines in 1 1112 open, exclusion and frame cages during trial one, 2002 113

Table 4.13. Probability of a greater F -statistic based on the number nymph A. glycines in l m2 open, exclusion and frame cages during trial one, 2002 114

Table 4.14. Species composition, total, and percent predators observed during 3-minute non-intrusive and intrusive observations in exclusion, open, and frame cages prior to cage switch during trial one, (28 June to 12 July), 2002, East Lansing, Michigan 115

Table 4.15. Species composition, total, and percent of predators observed during 3-minute non-intrusive and intrusive observations in exclusionl, open, and frame cages after cage switch (15 July to 29 July) during trial one, 2002 116

Table 4.16. Probability of a greater F -statistic based on predators observed during 3-minute non-intrusive and intrusive observations in exclusion, open, and frame cages during trial one, 2002 117 Table 4.17. Probability of a greater F -statistic based on number adult A. glycines in l m2 open, exclusion and frame cages during trial two, 2002 118

Table 4.18. Probability of a greater F -statistic based on results of nymph A. glycines produced in 1 m2 open, exclusion and frame cages during experiment two, 2002 119

Table 4.19. Species composition, total, and percent of total predators observed during 3 minute non-intrusive and intrusive foliar observations in exclusion, open, and frame cages prior to cage switch during (12 July to 26 July) tn'al two, 2002, East Lansing, Michigan 120

Table 4.20. Species composition, total, and percent of total predators observed during 3-minute non-intrusive and intrusive foliar observations in exclusionl, open, and frame cages after cage switch (29 July to 12 August) during trial two, 2002. This part of the experiment was conducted from during 2002, after open and exclusion cages were switched, East Lansing, Michigan 121

Table 4.21. Probability of a greater F -statistic based on the number of predators observed during 3-minute non-intrusive and intrusive foliar observations in exclusion, open, and frame cages during trial two, 2002 122

Table 1. Mean number (1'. SEM) of adult A. glycines remaining and offspring produced on six different varieties of soybean at 24 h (day 1), 48 h (day 2), and 72 h (day 3) during trial one 151

Table 2. Mean number (_-I_- SEM) of adult A. glycines remaining and offspring produced on six different varieties of soybean at 24 h (day 1), 48 h (day 2), and 72 h (day 3) during trial two 152

Table 3. Mean number (i SEM) of adult A. glycines remaining and offspring produced on six different varieties of soybean at 24 h (day 1), 48 h (day 2), and 72 h (day 3) during trial three 153

xi LIST OF FIGURES

Figure 1.1. Distribution of A. glycines in Michigan. Dotted counties indicate the presence of A. glycines during sampling. Dark colored counties indicate infestation levels of 100 or more aphids per leaf during sampling. Dr. Chris DiFonzo, Michigan State University, Personal Communication. 12

Figure 2.1. Layout of the 2001 ﬁeld location on the Michigan State University Entomology Farm, East Lansing, Michigan. The ﬁelds were separated into four blocks, each containing a refuge and control plot. Enlarged plots illustrate where pitfall traps and foliar sampling areas were located, and where clip cages were placed. 49

Figure 2.2. Mean number (i SEM) of predators observed in l m x 30 cm areas during; A. 5- minutes in trial one (7-8 June 2001), trial two (14-15 June 2001), and trial three (24-25 June 2001) and; B. 3-minutes during trial one (18-19 June 2002), trial two (24-25 June 2002), and trial three (2-3 July 2002), East Lansing, Michigan. 50

Figure 2.3. Mean number (i SEM) of predators per trap per day collected in 32 pitfall traps; A. located in refuge and control plot during trial one (7-8 June 2001), trial two (14-15 June 2001), and trial three (24-25 June 2001); B. in 30 pitfall traps in ﬁve blocks during trial one (18-19 June 2002), trial two (24-25 June 2002), and trial three (2-3 July 2002), East Lansing, Michigan. 51

Figure 2.4. Percentage of live adult A. glycines remaining at 15 and 24 h for the open, leaky and exclosure treatments in refuge and control plots during; A. trial one (7-8 June 2001); B. trial two (14-15 June 2001); and C. trial three (24-25 June 2001), East Lansing, Michigan. Different letters above bars within an hour denotes signiﬁcant (P g 0.1) differences among open, leaky, and exclosure treatments (Poisson regression test). 53

Figure 2.5. Number of live A. glycines nymphs at 15 and 24 h in three cage treatments during; A. trial one (7-8 June 2001); B. trial two (14-15 June 2001); and C. trial three (24-25 June 2001), East Lansing, Michigan. Different letters above bars within an hour denote signiﬁcant (P 5 0.1) differences in the number of nymphs present in open, leaky, and exclosure treatments (Poisson regression (log-linear model) test). 55

Figure 2.6. Total number of live adult and nymph A. glycines at 15 and 24 h in h in three cage treatments during 1) trial one (7-8 June 2001), 2) trial two (14-15 June 2001) and 3) trial three (24-25 June 2001), East Lansing, Michigan. Different letters above bars within an hour denote signiﬁcant (P 5 0.1) differences among open, leaky, and exclosure treatments (Poisson regression (log-linear model) test). 57

xii Figure 2.7. Percentage of live adult A. glycines remaining at 15 and 24 h for the open, leaky and exclosure treatments in during; A. trial one (18-19 June 2002); B. trial two (24-25 June 2002); and C. trial three (2-3 July 2002), East Lansing, Michigan. Different letters above bars within an hour denote signiﬁcant differences among open, leaky, and exclosure treatments (Poisson regression test). 59

Figure 2.8. Number of live nymph A. glycines at 15 and 24 h in three cage treatments during; A. trial one (18-19 June 2002); B. trial two (24-25 June 2002); and C. trial three (2-3 July 2002), East Lansing, Michigan. Different letters above bars within a given hour denote signiﬁcant (P 5 0.1) differences in the number of nymphs present in open, leaky, and exclosure treatments (Poisson regression (log- linear model) test). 61

Figure 2.9. Total number of live adult and nymph A. glycines at 15 and 24 h in three cage treatments during; A. trial one (18-19 June 2002); B. trial two (24-25 June 2002) and C. trial three (2-3 July 2002), East Lansing, Michigan. Different letters above bars within a given hour denote signiﬁcant (P 5 0.1) differences among open, leaky, and exclosure treatments (Poisson regression (log-linear model) test). 63

Figure 4.1. Design of 1 m2 cages used during 2001-2002. Both cage types contain the same cage materials to account for potential cage effects. A. open and B. exclusion. 123

Figure 4.2. Mean (i SEM) temperature difference (°C interior minus the oC exterior) of open and exclusion cages during 2001, East Lansing, Michigan. 124

Figure 4.3. Mean (i SEM) relative humidity difference (interior minus the exterior) of open and exclusion cages during 2001, East Lansing, Michigan. 125

Figure 4.4. Mean height (t SEM) in cm of ﬁve randomly selected soybean plants in the open and exclusion treatments during 2001, East Lansing, Michigan. 126

Figure 4.5. Mean (i SEM) abundance of foliar-foraging predators based on a combination of 3-minute non-intrusive visual examination followed by hand examinations of foliage in the Open and exclusion treatments during 2001, East Lansing, Michigan. 127

Figure 4.6. Mean (1- SEM) abundance of Harmonia axyridis adults and larvae based on a combination of 3-minute non-intrusive visual examination followed by hand examinations of foliage in the open and exclusion treatments during 2001, East Lansing, Michigan. 128

xiii Figure 4.7. Mean (i SEM) abundance of Orius insidiosus adults and nymphs based on a combination of 3-minute non-intrusive visual examination followed by hand examinations of foliage in the open and exclusion treatments during 2001, East Lansing, Michigan. 129

Figure 4.8. Mean (i SEM) adult A. glycines population per 5 plant tips during 2001 in open and exclusion cages during 2001, East Lansing, Michigan. 130

Figure 4.9. Mean (i SEM) nymph A. glycines population during 2001 in Open and exclusion treatments during 2001, East Lansing, Michigan. 131

Figure 4.10. Mean (1- SEM) abundance Of Harmonia axyridis (Pallas) found on sticky traps in the exclusion cages versus those found in the entire plot during 2001, East Lansing, Michigan. 132

Figure 4.11. Mean (i SEM) abundance of Orius insidiosus (Say) found on sticky traps in the exclusion cages versus those found in the entire plot during 2001, East Lansing, Michigan. 133

Figure 4.12. Mean (i SEM) predator/pitfall trap/day of ground beetles in pitfall traps in exclusion cages versus those in the study area during 2001, East Lansing, Michigan. 134

Figure 4.13. Mean (i SEM) temperature difference (°C interior minus the oC exterior temperature) in open, exclusion, and frame cages during 2002. The arrow indicates when open and exclusion cages over 1 m2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 135

Figure 4.14. Mean (i SEM) relative humidity difference (interior minus the exterior) in open, exclusion, and frame cages during 2002. The arrow indicates when open and exclusion cages over 1 in2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 136

Figure 4.15. Mean (i SEM) plant height per ﬁve plants in open, exclusion, and frame cages during 2002. The arrow indicates when open and exclusion cages over 1 m plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 137

xiv Figure 4.16. Mean (1 SEM) abundance of carabid beetles collected per seven days in 10 total pitfall traps from exclusion cages (2 traps/ cage) versus 10 total pitfall traps placed randomly in the study area during 2002: A. Trial one; and B. Trial two. The arrow indicates when open and exclusion cages over 1 m2 plots were switched, at which time eight pitfall traps were in exclusion cages and eight within the plot. 138

Figure 4.17. Mean (i SEM) adult (A.) and nymph (B.) A. glycines population (log scale) per ten whole plants in Open, exclusion, and frame cages in trial one during 2002. The arrow indicates when open and exclusion cages over 1m2 plots were switched. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 139

Figure 4.18. Mean (i SEM) adult (A.) and nymph (B.) A. glycines population (log scale) per ten whole plants in open, exclusion, and frame cages in trial one during 2002. An arrow indicated when Open and exclusion cages over 1m2 plots were switched. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 140

Figure 4.19. Mean (i SEM) abundance of foliar-foraging predators based on a combination of 3-minute non-intrusive visual examination followed by hand examinations of foliage in the open, exclusion, and frame treatments during 2002. The arrow indicates when open and exclusion cages over 1 in2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 141

Figure 4.20. Mean (1- SEM) abundance of Harmonia axyridis (Pallas) adults based on a combination of 3-minute non-intrusive visual examination followed by hand examination of foliage in the open, exclusion, and frame treatments during 2002. The arrow indicates when open and exclusion cages over 1 m2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 142

Figure 4.21. Mean (i SEM) abundance of Leucopis midge larvae in based on a combination of 3-minute non-intrusive visual examination followed by hand examinations of foliage in the open, exclusion, and frame treatments during 2002. The arrow indicates when open and exclusion cages over 1 m2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 143

Figure 4.22. Mean (i SEM) abundance of Orius insidiosus (Say) based on a combination Of 3-minute non-intrusive visual examination followed by a hand examinations of foliage in the open, exclusion, and frame treatments during 2002. The arrow indicates when open and exclusion cages over 1 m2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 144

XV KEY TO SYMBOLS AND ABBREVIATIONS

AN OVA analysis of variance

0C degrees Celsius cm centimeters

days df degrees of freedom

F -statistic

hours ha hectare

liter

L:D light: dark hours

meters

number of observations ns not signiﬁcant

probability under the assumption that the null hypothesis is true

RH relative humidity

SEM standard error of the mean spp. species

t-statistic

xvi CHAPTER 1

Introduction to soybean production, Aphis glycines Matsumura biology and the

importance of biological control

Introduction

Prior to the summer of 2000, no aphid species was known to regularly colonize

soybean, Glycine max (L.), in the United States (DiFonzo 2000). However, in 2000

Aphis glycines Matsumura, an invasive aphid species from Asia, was discovered infesting

soybean plants in the United States. First observed in Wisconsin in July 2000, the aphid was initially thought to be the melon aphid, Aphis gossypii (Glover). The positive identiﬁcation of A. glycines prompted surveys in surrounding states in the summer of

2000 and 2001. As of 2001, the current distribution of A. glycines included MN, IL, IA,

IN, MI, ND, KY, OH, WI, and NY (University of Minnesota Web Site, North Dakota

State University). In 2002, it had spread to NE, KS, DE, MN, GA, and MS (Ragsdale

Pers. Comm.). In Michigan A. glycines is distributed primarily in the central and southern portions of the state, with the most severe infestations occurring in the southwestern counties during 2000 and the southeastern counties during 2001 (DiFonzo

2001, Figure 1.1). Aphis glycines populations in Michigan during 2002 presumably occurred in most soybean growing regions of the states; however, populations remained low.

Aphis glycines is a serious pest of soybean in Asia and has the potential to negatively impact soybean production in Michigan and the entire United States. Aphis glycines feeding can cause up to a 20.2 cm reduction in growth and a 27.8 percent reduction in seed yield (Wang et a1. 1996). In 2000, up to a 13 percent yield reduction occurred in replicated ﬁeld plots in Wisconsin during the year 2000 (University of

Minnesota Extension Service Web Site). In 2001, A. glycines caused up to 40 percent yield loss in Michigan (DiFonzo, Pers. Comm). This pest can indirectly harm soybean plants by vectoring persistent viruses, such as soybean dwarf virus, millet red leaf and non-persistent viruses such as soybean stunt virus, soybean mosaic virus, bean yellow mosaic virus, beet mosaic, and mung bean mosaic (Halbert et al. 1986, Van den Berg et al. 1997, Fletcher and Desborough 2000). Epidemics of soybean mosaic potyvirus in summer-sown soybean ﬁelds in Jiangsu, China are closely related to the timing of A. glycines immigration (Li and Pu 1991). Virus transmission by aphids has not yet been observed in United States (DiFonzo Pers. Comm.). In addition, aphids also cause indirect damage by excreting honeydew onto foliage which promotes the growth of sooty molds that have the potential to reduce the photosynthetic capacity of the darkened leaves

(Lenné and Trutmann 1994, Hirano et a1. 1996, DiFonzo 2000, Ostlie and Hutchinson

2000).

Due to its recent arrival in the United States, the ideal way to manage A. glycines in North America is not known. Insecticides have been used as a short-term solution on other aphid species, but studies show that insecticide treatments often fail to signiﬁcantly reduce aphid populations and can disrupt control by predatory insects (Asin and Pons

1999, Wilson et a1. 1999, Jansen 2000). The potential for predators to control this pest in the United States is also not well understood, though they have shown to be important in

Asia (Van den Berg et a1. 1997, Chang et a1. 1994). Many sources have illustrated that aphidophagous predators can reduce aphid populations (Grasswitz and Burts 1995, Star)?

1995, Van den Berg et a1. 1997, Obrycki and Kring 1998). However, generalist natural enemies can also play crucial roles in biological control (Hance 1987; Landis and Van der Werf 1997; Chang and Kareiva 1999, Symondson et a1. 2002). Moreover, conserving predatory insects through habitat management may enhance both specialist and generalist predator activity (Tonhasca 1993, Menalled et al. in review, Carmona and Landis 1999,

Landis et a1. 2000, Lee and Landis 2002, Lee et a1. 2001). Therefore, the potential for indigenous and exotic predators to aid in the control of this pest with and without conservation enhancement were examined.

Introduction to Soybean Value and Ecology

Glycine max (L.) is a very important crop in Michigan and other parts of the

United States. As a source of protein, soybean is often more inexpensive on a cost-per- kilogram basis than animal protein (Hymowitz and Newell 1981). Moreover, soybean contains 20 to 23 percent oil, a component commonly used in margarine, shortening, mayonnaise, and dressing. Soybean meal is commonly used as a major source of protein in animal feeds (Hymowitz and Newell 1981). The production of soybean in Michigan totaled 74.9 million bushels during the year 2000 harvest and the average yield was 36 bushels per acre (Michigan Agricultural Statistics 2000-2001). According to Michigan

Agricultural Statistics (2000-2001), the price of soybean per bushel received by farmers in 2000 was $4.75 in Michigan. While this is lower than historical prices, it is still an attractive alternative to corn ($1.90/bu) and wheat ($2.10/bu). Soybean Management

Soybean production methods vary from region to region within Michigan.

Typical soybean production practices for several of the aphid infested soybean-producing regions of the state are presented in Table 1.1. Overall soybean ﬁts into several different rotations, frequently following corn. Most Michigan producers seem to be following reduced tillage with a large proportion of the acreage being produced in drilled, narrow rows (< 76 cm). These production practices have the potential to inﬂuence management

Of the A. glycines.

Biology and Life Cycle of Aphis glycines Matsumura

A major challenge in this research is that while extensive literature on A. glycines exists, a majority of it is in Japanese, Korean, and Chinese language journals, and only a limited number of manuscripts have been translated to English at this time (Kansas State

University Web Site 2001). The following contains the majority of the literature currently available in English. Unfortunately, some of the literature is only available through secondary sources and translated abstracts, putting limitations on knowledge of previous research.

Like all insects, A. glycines development is affected by temperature. Laboratory research showed the intrinsic rate of increase of A. glycines was higher at 27° C than at

22° C (Hirano et al. 1996). Hirano et a1. (1996) also found that mean daily fecundity rose at a more rapid rate, peaked earlier, and declined faster at 27° C than at 22° C. It also has been stated that the optimum temperatures for A. glycines development in late June to early July are between 22 and 25° C, with a relative humidity less than 78 percent in

China (Wang et a1. 1962).

In Asia, A. glycines requires soybean and Rhamnus davurica (buckthom) to complete its life cycle (Blackman and Eastop 2000). Four Rhamnus spp. are currently distributed in Michigan, R. alnifolia ‘Her’ (Alderleaf buckthom), R. cathartica L.

(Common buckthom), R. utilis Dcne (Chinese buckthom), and R. frangula L. (Glossy buckthom) (Voss 1998). Another source suggests a ﬁfth species, R. davurica Pallas, is also present in the state (USDA Natural Resources Conservation Services Plants Data

Base).

The life cycle of A. glycines is complex. In late fall, female oviparae lay eggs on

Rhamnus species, after mating with males. Nymphs hatch in spring and the species undergoes two generations of Wingless females, followed by a winged generation of females. These winged females leave Rhamnus in search of soybean, their summer host.

Once on soybean, there are generations of Wingless and winged females, the latter leaving their natal site in search of new soybean plants when crowding occurs. In fall, males develop on soybean plants, fly to Rhamnus, and mate with oviparae on Rhamnus.

These mated females then lay eggs on the inner margins of buds on Rhamnus twigs

(Blackman and Eastop 2000).

Guang-xue and Tie-sen (1982) observed hybridization between Aphis glycines and Aphis gossypii, the melon aphid, under artiﬁcial conditions. The authors found that the hybrid offspring reproduced both parthenogenetic and sexual forms to complete their life cycle. They hypothesized that this potential for hybridization may result in aphid lines more resistant to certain chemical insecticides, but this has not been tested. Aphidophagous Predators

Aphidophagous predators can greatly contribute to the reduction of a wide variety of aphid species (Grasswitz and Burts 1995, Star)? 1995, Van den Berg et a1. 1997,

Obrycki and Kring 1998). For example, it was found that a complex of natural enemies, including Chrysopa nigricomis Burmeister, Orius spp., Coccinella transversoguttata

Faldermann, Hippodamia convergens Guerin, and various species of Syrphidae and

Chamaemyiidae all contributed to the reduction of Aphis pomi De Geer in a ﬁeld study using apple trees (Grasswitz and Burts 1995). Similarly, Star)? (1995) found that

Coccinella septempunctata L., Adalia bipunctata (L.), and Episyrphus balteatus (De

Geer) were common predators that consumed, and thus aided in control, of the Spiraea aphid, Aphis spiraephaga Muller in Czechoslovakia.

Furthermore, it is known that in Asia, coccinellid (Harmonia spp.) predators play an important role in suppressing A. glycines populations in soybean ﬁelds in tropical areas of Southeast Asia (Van den Berg et al. 1997). Other aphidophagous predator species, such as Nabis spp., Harmonia axyridis (Pallas), Coccinella septempunctata L., and Chrysopa spp. aided in its control during the mid to late season in China (Han 1997).

Schneider (1971) reported the predaceous syrphid larva Ischiodon escutellaris (F.) in the

Philippines was another aphidophagous predator that fed on a wide variety of aphids, including A. glycines. However, low syrphid abundance in comparison to the number of aphids has not resulted in suppression of aphid populations by syrphids, though they do contribute aphid reduction (Alﬁler and Calilung 1978, Van den Berg et a1. 1997). In Asia, coccinellid predators (Harmonia spp.) and to a much lesser extent syrphid predators, depend on aphids as a food source and reduce A. glycines populations, but their effects are not as signiﬁcant in temperate versus tropical climates (Van den Berg

1997). At lower temperatures it was found that the reproductive rate of A. glycines exceeded the predation rate of the coccinellid predators (Van den Berg et a1. 1997). This allowed A. glycines populations to rapidly develop in temperate regions, particularly during the early portion of the season (Van den Berg et al. 1997). A similar result has been documented for Coccinnella spp. attacking pea aphids, Acyrthosiphon pisum

(Harris), British Columbia (Frazer and Gilbert 1976)

Parasitoids of A. glycines have also been identiﬁed in Taejon, Korea and reports suggest that Ephredrus plagiator (Hymenoptera: Braconidae) might be a potential biological control agent because they are less frequently attacked by hyperparasitoids

(Chang et a1. 1994). Foraging predators searching for aphids accidentally consume some larvae of parasitoids (Taylor et a1. 1998). Taylor et a1. (1998) found that aphid parasitoids respond to the presence of predators by reducing residence time and oviposition in patches containing aphid predators. This interaction is not completely understood, but it is thought that the parasitoids perceive chemical cues produced by either the predators or aphids. Even when the introduction of speciﬁc parasitoids resulted in spectacular success of biological control of the pest, the importance of other predators should not be ignored (Takagi 1999). Generalist Predators

The role of aphidophagous predators in control of A. glycines in the United States is not currently known. One of the main beneﬁts of generalist predators is that they may help to reduce pest populations when specialized predators are not available (Symondson et al. 2002). One example is that of the Colorado potato beetle, Leptinotarsa decemlineata (Say), which has few specialized natural enemies. Hilbeck et a1. (1997) found that generalist predators have the ability to signiﬁcantly reduce the abundance of

Colorado potato beetle eggs during their three-year ﬁeld study. Similarly, it was found that immigration of generalist lycosid spiders and carabids into a garden area exerted sufﬁcient biological control of the striped cucumber beetle, Acalymma vittata F., the spotted cucumber beetle, Diabrotica undecimpunctata howardi Barber, and the squash bug, Anasa tristis De Geer, to increase the fruit production of squash (Snyder and Wise

1999). Furthermore, three carabid species, Asaphidion ﬂavipes (L.), Agonum dorsale

(Pont.) and Pterostichus melanarius (111.) all reduced populations of Aphisfabae Scopoli, a pest of sugar beet, in caged experiments. Further, it has been shown that a complex of ground-dwelling predators, such as carabids, staphylinids, and spider, all contributed to the reduction of the bird cherry-oat aphid, Rhopalosiphum padi (L.), populations (Ostman et a1. 2001).

While generalist predators may not be as effective as specialists, they can make up for this by being present earlier in the season (Chang and Kareiva 1999). Early in the growing season, pest density is low but populations build up and may reach outbreak status late in the season (Takagi 1999). Thus, the early season might be an important time for generalists to reduce pest density before they have an opportunity to initiate large colonies. For example, Landis and Van der Werf (1997) found many species of carabids in sugar beet ﬁelds in the Netherlands, along with spiders and a dominant generalist predator, Cantharis lateralis (Coleoptera: Cantharidae) early in the season. They found that early season predation by generalists played a key role in suppressing the spread of aphid-vectored viruses and allowed for season-long reduction of aphid populations.

Early season predation by generalist predators may also be important because these predators are capable of locating alternate food sources when the pest population is low

(Chang and Kareiva 1999, Symondson et a1. 2002).

Effects of Pesticides on Aphid Predators

The negative effects of insecticides on beneﬁcial insects are well documented

(Wilkinson et al. 1979, Bellows and Morse 1993, Epstein et al. 2000). In Belgium, use of insecticides using ﬂuvalinate and esfenvalerate to control cereal aphids in wheat was detrimental to coccinellids and pirimicarb was found to be harmful to syrphid larvae

(Jansen 2000). Jansen (2000) also reported that cyﬂuthrin, deltamethrin and phosalone reduced catches of both syrphid and coccinellids in wheat ﬁelds. Similarly, higher numbers of melon aphids, A. gossypii, occurred in cotton treated with thiodicarb than in unsprayed cotton due to its negative effect on aphidophagous predators (Wilson et al.

1999). Similarly, it was found that at planting treatments of carbofuran negatively affected populations of carabid beetles and that maize aphid, Rhopalosiphum padi (L.), populations became higher in carbofuran treated plots compared to control plots. In conventional and no-tillage situations, chloropyrifos, DPX-43898, fonofos, and terbufos, all granular insecticides, negatively affected carabid beetles (Reed et a1. 1992). Certain herbicides may also negatively affect predators. Brust (1990) found that glyphosate and paraquat acted as repellents to certain species of carabid beetles.

Importance of Conservation Biological Control

The aim of conservation biological control is to enhance the overall survival impact of natural enemies (Debach and Rosen 1991). This can be accomplished by manipulating the environment to reduce harmful factors and to increase favorable conditions (Landis et a1. 2000). Reducing direct mortality, providing supplementary resources, controlling secondary enemies, and manipulating host plant attributes are the main approaches to natural enemy conservation (Debach and Rosen 1991).

Importance of Habitat Management

Habitat management is considered a subset of conservation biological control

(Pickett and Bugg 1998, Landis et a1. 2000). A number of studies show the signiﬁcance of habitat management in increasing predator abundance. For example, in the United

Kindom, species richness of generalist predators (spiders, carabids, and heteropterans) was greater in non-cropped headlands on the outer six meters of arable fields (Hassall et al. 1992). Canola flowers in an adjacent crop area to wheat crops were an important food source for adult syrphid flies, which flew to the wheat crop to oviposit (Bowie et a1.

1999). Furthermore, carabid beetles in upland rice fields were significantly more abundant in plots that contain piles or strips of weeds and staphylinid beetles were more abundant in mulched plots (Afun et a1. 1999). Undisturbed semi-natural habitats and extensively managed field margins were found to be important overwintering sites for

10 predatory arthropods, including staphylinids, carabids, and spiders (Pﬁffner and Luka

2000).

Management practices of the crops themselves can also have an inﬂuence on predators. In Michigan, researchers have been working to increase generalist predators in

Michigan corn, soybean and wheat rotations (Carmona and Landis 1999; Landis et al.

2000; Lee and Landis 2002; Lee et al. 2001). They have found that grassy refuge strips with cover crops and reduced tillage can increase the number of species and the overall abundance of generalist predators in crop ﬁelds adjacent to the refuge habitats, by providing overwintering sites, alternate food sources and cover from crop management practices.

Objectives

The purpose Of these studies was to evaluate the impacts of indigenous and exotic predators on A. glycines with and without habitat management. The speciﬁc objectives were to:

1. To determine the importance of early-season predation on A. glycines by exotic and indigenous predators.

2. To determine which exotic and indigenous foliar and ground-dwelling predators are capable of consuming A. glycines.

3. To determine the ability of exotic and indigenous predators to reduce establishing aphid colonies during mid to late season.

III?

Distribution as of Distribution as of October 2000 October 2001

DiFonzo, Michigan State University, Personal Communication. Table 1.1. Soybean management strategies in four counties of Michigan. Rotation percentages show the percentage of farmers that use each rotation within the three counties. Tillage percentages show the percentage of farmers within the three counties that use particular tillage practices. Row spacing percentages show the percent of farmers that use speciﬁc row spacing within each county. All surveys were conducted between April 23 and March 13, 2001

County: Saginaw (M. Gratiot (D. Rossman Monroe (N. Berrien (M. Seamon Pers. Pers. Comm.) Birkey Pers. Staton Pers. Comm.) Comm.) Comm.)

Rotations C-S 30 % C-S 40 % C-S 35 % C-S 60% C-S-W 10% C-S-W 20 % C-S-W 50 % C-S-W 6-10% S-SB-S-C 30% 8-8 < 10 % 8-8 40% S-S 34-30% S-W-SB-C 15% SB-C-DB-W-S 20 % C-S-Vegetable S-W-S-W 10 % 15 % Tillage No till 50% No till 35 % No till 50 % No till 40% (Preceding Fall chisel plow 40 % Fall chisel plow including Minimum tillage Fall chisel soybeans) Moldboard plow 10% ﬁeld cultivation in spring 50% plow 60% 60 % Moldboard plow and all Others 5% Row 7 1/2” 50% (70% 7 V2” 60 % (use no till) 7 V2” 50 % (No 7 V2” 70% (No spacing conventional tillage, 15" 5 % till) till) 30” 30% 30 % no till) 30” 35 % (drilled) 15”, 20” and 30” (chisel/disc 20 or 22” 10% (90% equals the combo, though conventional, 10% no remainder a few also till) incorporate no 28 or 30" 40% (40% tillage) conventional, 60% no till)

Key: S=soybean W=wheat DB: dry beans C: corn SB: sugar beets

13 Literature Cited

Afun, J. V. K., D. E. Johnson, and A. Russell-Smith. 1999. The effects of weed residue management on pests, pest damage, predators and crop yield in upland rice in COte d’lvoire. Biological Agriculture and Horticulture. 17:47-58.

Alﬁler, A. R. R., and V. J. Calilung. 1978. The life history and voracity of the syrphid predator, Ischodiodon escutellaris (F.) (Diptera: Syrphidae). Philippine Entomologist. 4(1-2)105-117.

Asin, L., and X. Pons. 1999. Effects of soil insecticide treatments on maize aphids and aphid predators in Catalonia. Crop Protection. 18: 389-395.

Bellows, T. S., and J. G. Morse. 1993. Toxicity of pesticides used in citrus to Aphytis melinus Deback (Hymenoptera: Aphelinidae) and Rhizobius lophanthae Blaisd (Coleoptera: Coccinellidae). Canadian Entomologist. 125: 987-994.

Blackman, R. L., and V. F. Eastop. 2000. Aphids on the world’s crops: an identiﬁcation and information guide. Second Edition, Wiley, Chichester.

Bowie, H. M., G. M. Gurr, Z. Hossain, L. R. Baggen, and C. M. Frampton. 1999. Effects of distance from ﬁeld edge on aphidophagous insects in a wheat crop and observations on trap design and placement. International Journal of Pest Management. 45: 69-73.

Brust, G. E. 1990. Direct and indirect effects of four herbicides on the activity of carabid beetles (Coleoptera: Carabidae). Pesticide Science. 30: 309-320.

Carmona, C. M., and D. A. Landis. 1999. Inﬂuence of refuge habitats and cover crops on seasonal activity-density of ground beetles (Coleoptera: Carabidae) in ﬁeld crops. Environmental Entomology. 28(6): 1145-1153.

Chang, G. C., and P. Kareiva. 1999. The case for indigenous generalists in biological control, pp. 103-115. In B. A. Hawkins and H. V. Cornell, ed. Theoretical approaches to biological control. Cambridge University Press, UK.

14 Chang, Y. D., J. Y. Lee, and Y. N. Youn. 1994. Primary parasitoids and hyperparasoids of the soybean aphid, Aphis glycines (Homoptera: Aphididae). Korean Journal of Applied Entomology. 33(2): 51-55.

Debach, P., and D. Rosen. 2001. Biological control by natural enemies. Cambridge University Press, Cambridge.

DiFonzo, C. 2000. Aphids amuck: soybean aphids discovered in US. Field Crop Advisory Team Alert. 15(15): http://www.msue.msu.edu/ipm/CATOO ﬁeld/FC08-l7- 00.htm

DiFonzo, C. 2001. Soybean aphid in Michigan: Update from 2001 Season. Michigan State University. Extension Bulletin E-2748

Epstein, D. L., R. S. Zack, J. F. Brunner, L. Gut, and J. J. Brown. 2000. Effects of broad-spectrum insecticides on epigeal arthropod biodiversity in Paciﬁc Northwest apple orchards. Environmental Entomology. 29: 340-348.

Fletcher, M. J ., and P. Desborough. 2000. The soybean aphid, Aphis glycines, present in Australia: http://www.agric.nsw.govau/Hort/ascu/insects/aglycin.htm

Frazer, B. D., and N. Gilbert. 1976. Coccinellids and aphids: a quantitative study of the impact of adult ladybirds (Coleoptera: Coccinellidae) preying on ﬁeld populations of pea aphids (Homoptera: Aphididae). Journal of the Entomological Society of British Columbia. 73: 33-56.

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

Guang-xue, Z., and Z. Tie-sen. 1982. Experimental studies on some aphid life-cycle patterns. Sinozoologica. 2:7-17.

Halbert, S. E., G-X. Zhang, and Z-Q. Pu. 1986. Comparison of sampling methods for alate aphids and observations on epidemiology of soybean mosaic virus in Nanjing, China. 109: 473-483.

15 Han, X. 1997. Population dynamics of soybean aphid Aphis glycines and its natural enemies in ﬁelds. Hubei Agricultural Sciences. 2: 22-24.

Hance, T. 1987. Predation impact of carabids at different population densities on Aphis fabae development in sugar beet. Pedobiologia. 30: 251-262.

Hassall, M. A. Hawthorne, M. Maudsley, P. White, and C. Cardwell. 1992. Effects of headland management on invertebrate communities in cereal ﬁelds. Agriculture, Ecosystems and Environment. 40: 155-178.

Hilbeck, A., C. Eckel, and G. Kennedy. 1997. Predation on Colorado potato beetle eggs by generalist predators in research and commercial potato plantings. Biological Control. 8: 191-196.

Hirano, K., K. Honda, and S. Miyai. 1996. Effects of temperature on development, longevity and reproduction of the soybean aphid, Aphis glycines (Homoptera: Aphididae). Applied Entomology and Zoology. 31(1): 178-180.

Hymowitz, T., and CA. Newell. 1981. Taxonomy of the genus Glycine, Domestication and uses of Soybeans. Economic Botany. 35 (3): 272-288.

Jansen, J. P. 2000. A three-year ﬁeld study on the short-term effects of insecticides used to control cereal aphids on plant-dwelling aphid predators in winter wheat. Pest Management Science. 56: 533-539.

Kansas State University Web Site. 2001. Soybean aphid literature translation project. http://www.lib.ksu.edu/depts/issa/aphids/reporthtml/aphidcitations.html

Landis, D. A., and W. Van der Werf. 1997. Early-season aphid predation impacts establishment and spread of sugar beet yellows virus in the Netherlands. Entomophaga. 42: 499-516.

Landis, D. A., S. D. Wratten, and G. Gurr. 2000. Habitat manipulation to conserve natural enemies of arthropod pests in agriculture. Annual review Of entomology. 45: 173-199.

16 Lee, J. C., and D. A. Landis. 2002. Noncrop habitat management for carabid beetles. Pp: 279-303. In J. Holland, ed. Carabid beetles and agriculture, intercept, Adover. .

Lee, J. C., F. D. Menalled, and D. A. Landis. 2001. Refuge habitats modify insecticide disturbance on carabid beetle communities. Journal of Applied Ecology. 38: 472-483.

Lenné, J. M., and P. Trutmann. 1994. Diseases of tropical pasture plants. CAB International, Wallingford, UK. P. 69.

Li, W. M., and Z. O. Pu. 1991. Population dynamics of aphids and epidemics of soybean mosaic virus in summer sown soybean ﬁelds. Acta Phytophylactica Sinica. 18(3): 123-126.

Menalled, F. D., J. C. Lee, and D. A. Landis. 2000. Herbaceous ﬁlter strips in agroecosystems: implications for carabid beetle conservation and invertebrate weed seed predation. Journal of Economic Entomology. in review.

Michigan Agricultural Statistics. 2000-2001. Michigan Department of Agriculture. USDA.

North Dakota State University Web Site: http://www.ag.ndsu.nodak.edu/aginfo/entomology/entupdates/Soybean_Aphid/sybn_aph d.htm.

Obrycki, J. J., and T. J. Kring. 1998. Predaceous coccinellidae in biological control. Annual Review of Entomolgy. 43: 295-321.

Ostlie, K., and B. Hutchinson. 2000. New soybean pest appears in the Midwest: is it in Minnesota? (2) 15.

Ostman, O., E. Ekbom, and J. Bengtsson. 2001. Landscape heterogeneity and farmingpractice inﬂuence biological control. Basic and Applied Ecology. 2: 365-371.

17 Pickett, C. H., and R. L. Bugg. 1998. Enhancing biological control: habitat management to promote natural enemies of agricultural pests. Berkeley: University of California Press. 422 pp.

Pﬁffner, L., and H. Luka. 2000. Overwintering of arthropods in soils of arable ﬁelds and adjacent semi-natural habitats. Agriculture, Ecosystems and Environment. 78: 215-222.

Reed, J. P., F. R. Hall, and H. R. Krueger. 1992. Contact and volatile toxicity of insecticides to black cutworm larvae (Lepidoptera: Noctuidae) and carabid beetles (Coleoptera: Carabidae) in soil. Journal of Economic Entomology. 85(1): 256-261

Schneider, F. 1971. Bionomics and physiology of aphidophagous syrphidae. Annual Review of Entomology. 12: 103.

Snyder, W. E., and D. H. Wise. 1999. Predation interference and the establishment of generalist predator populations for biocontrol. Biological Control. 15: 283-292.

Stary, P. 1995. Natural enemy spectrum of Aphis spiraephaga (Hom: Aphididae), an exotic immigrant aphid in central Europe. Entomophaga. 40(1): 29-34.

Syndmondson, W.O.C., K.D. Sunderland, and M.H. Greenstone. 2002. Can generalist predators be effective biocontrol agents? Annual Review of Entomology. 47: 561-594.

Takagi, M. 1999. Perspective of practical biological control and population theories. Researches on population ecology. 41: 121-126.

Taylor, A. J ., C. B. Muller, and C. J. Godfray. 1998. Effect of aphid predators on oviposition behavior of aphid parasitoids. Journal of Insect Behavior. 11(2): 297-302.

Tonhasca, A. 1993. Carabid beetle assemblage under diversiﬁed agricultural systems. E Entomologia Experimentalis et Applicata. 68: 279-285.

University of Illinois Web Site: http://www.inhs.uiuc.edu/cbd/aphid/

18 University of Minnesota Extension Service Web Site: http://wwwsoybeansumn.edu/crop;prod/insects/aphid/aphid.htm

USDA Natural Resources Conservation Services Plants Data Base: http://plants.usda.gov/plants/index.html

Van den Berg, H., D. Ankasah, A. Muhammad, R. Rusli, H. A. Widayanto, H. B. Wirasto, and I. Yully. 1997. Evaluating the role of predation in population ﬂuctuations of the soybean aphid Aphis glycines in farmers’ ﬁelds in Indonesia. Journal of Applied Ecology. 34, 971-984.

Voss, E. G. 1998. Michigan ﬂora part II: dicots: a guide to the identiﬁcation and occurrence of the native and naturalized seed-plants of the state. University of Ann Arbor, Michigan

Wang, C. L., N. I. Siang, G. S. Chang, and H. F. Chu. 1962. Studies on the soybean aphid, Aphis glycines Matsumura. Acta Entomologica Sinica. 11: 31-44.

Wang, S. Y., X. Z. Boa, Y. J. Sun, R. L. Chen, and B. P. Zhai. 1996. Study on the effects of the population dynamics of soybean aphid (Aphis glycines) on both growth and yield of soyabean. Soybean Science. 15(3): 243-247.

Wilkinson, J. D., K. D. Biever, and C. M. Ignoffo. 1979. Synthetic pyrethroid and organophosphate insecticides against parasitoid Apanteles marginiventris and the predators Geocoris punctipes, Hippodamia convergens, and Podisus maculiventris. Journal of Ecomomic Entomology. 72: 473-475.

Wilson, L. J ., L. R. Bauer, and D. A. Lally. 1999. Insecticide-induced increases in aphid abundance in cotton. Australian Journal of Entomology. 38: 242-243.

19 CHAPTER 2

The impact of indigenous predators on preventing establishment of the soybean

aphid, Aphis glycines Matsumura

Introduction

The soybean aphid, Aphis glycines Matsumura, is an exotic pest from Asia that was discovered in the United States in 2000. In Asia, A. glycines feeding can cause up to a 20 cm reduction in growth and a 27.8 percent reduction in seed yields (Wang et al.

1996). In 2000, A. glycines caused up to a 13 percent yield reduction in replicated ﬁeld plots in Wisconsin (University of Minnesota Extension Service Web Site). In 2001, it was found that A. glycines caused up to a 40 percent sprayed versus unsprayed yield loss in Michigan (Difonzo, Pers. Comm). This pest can also indirectly harm soybeans by vectoring persistent viruses, such as soybean dwarf virus and non-persistent viruses, such as soybean stunt virus, soybean mosaic virus, and bean yellow mosaic virus (Van den

Berg et al. 1997). Epidemics of soybean mosaic potyvirus in summer-sown soybean

ﬁelds in Jiangsu, China were found to be closely associated to the timing of A. glycines immigration (Li and Pu 1991). In addition, aphids also cause indirect damage by excreting honeydew onto foliage, which promotes the growth of sooty molds that reduce the photosynthetic capacity of the leaves (Lenné and Trutmann 1994, Hirano et al. 1996).

Aphidophagous predators can greatly contribute to the reduction of a wide variety of aphid species (Grasswitz and Burts 1995, Stary 1995, Van den Berg et al. 1997,

Obrycki and Kring 1998). For example, a complex of natural enemies, including

20 Chrysopa nigricornis Burmeister, Orius spp., Coccinella transversoguttata Faldermann,

Hippodamia convergens Guerin, and several species of Syrphidae and Chamaemyiidae all contributed to reductions in Aphis pomi De Geer in a ﬁeld study on apple (Grasswitz and Burts 1995). Similarly, Stary (1995) found that Coccinella septempunctata L.,

Adalia bipunctata (L.), and Episyrphus balteatus (De Geer) were common predators that all consumed and thus aided in control of the spiraea aphid, Aphis spiraephaga Muller in

Czechoslovakia.

Furthermore, it is known that in Asia that coccinellid (Harmonia spp.) predators play an important role in suppressing A. glycines populations in soybeans (Van den Berg et al. 1997). Other predator species, such as Nabis spp., Harmonia axyridis (Pallas),

Coccinella septempunctata L., and Chrysopa spp. have been shown to aid in A. glycines control during the mid to late season in China (Han 1997).

Schneider (1971) reported the predaceous syrphid larva Ischiodon escutellaris (F .) in the Philippines was another aphidophagous predator that fed on a wide variety of aphids, including A. glycines. However, low syrphid abundance in comparison to the number of aphids has not caused suppression of aphid populations by syrphids alone, although they do contribute aphid reduction (Alﬁler and Calilung 1978, Van den Berg et al. 1997).

In Asia, coccinellid predators (Harmonia spp.) and to a much lesser extent syrphid predators are associated with A. glycines populations (Van den Berg et al. 1997).

Aphidophagous predators clearly depended on aphids as a food source and reduced A. glycines populations in tropical areas of Southeast Asia, but the effects were not as pronounced in temperate climates. At lower temperatures it was found that the

21 reproductive rate of A. glycines exceeded the predation rate of the coccinellid predators

(Van den Berg et al. 1997). This allowed A. glycines populations to develop rapidly in temperate regions, particularly during the early portion of the season (Van den Berg et al.

1997). A similar result has been documented for Coccinnella spp. attacking pea aphids,

Acyrthosiphon pisum (Harris) in British Columbia (Frazer and Gilbert 1976).

Much consideration has been given to the importance that more generalist predators may play in biological control (Hance 1987, Hilbeck et al. 1997, Landis and

Van der Werf 1997, Snyder and Wise 1999, Chang and Kareiva 1999, Symondson et al.

2002). For example, it has been shown that ground beetles (Carabidae) species are capable of reducing populations of Aphis fabae Scopoli, a pest of sugar beet (Hance

1987). It has also been demonstrated that many species of carabid beetles in sugar beet

ﬁelds in the Netherlands, along with spiders and a dominant cantharid generalist predator,

Cantharis lateralis (Coleoptera: Cantharidae) aided in A. fabae aphid control early in the season (Landis and Van der Werf 1997). Further, it has been shown that a complex of ground-dwelling predators, such as carabids, staphylinids, and spider, all contributed to the reduction of the bird cherry-oat aphid, Rhopalosiphum padi (L.), populations (Ostman et al. 2001). While generalist predators are not as effective per capita as specialized predators, they can often compensate by being present earlier in the season (Chang and

Kareiva 1999), when pest densities are low and specialist predators scarce (Takagi 1999).

It is now known that A. glycines establishment in Michigan occurs in rrrid-June, prior to the arrival of specialized natural enemies. Thus, it is the more generalist predators that will likely make an impact when A. glycines establishment occurs. However, the role of predation to control A. glycines in the United States is not currently known.

22 The positive aspects of conservation biological control have been well documented in the literature for both specialist and generalist predators (Bowie et al.

1999, Carmona and Landis 1999, Landis et al. 2000, Debach and Rosen 2001, Lee and

Landis 2002; Lee et al. 2001). Manipulating the environment through habitat management to favor natural enemies can reduce direct mortality, provide supplementary resources, and help to control secondary enemies by manipulating host plant attributes

(Debach and Rosen 2001). One potentially important role of habitat management is to increase the abundance of generalist predators early in the season.

The objectives of these studies during 2001 was to: 1) determine the abundance and species composition of potential A. glycines predators in plots bordering natural enemy refuge habitats versus control plots; 2) to study the impacts of predators on A. glycines establishment in these soybean plots. The objective in 2002 was to: 3) examine the impact predation on A. glycines establishment in soybean ﬁelds without refuge habitats.

Materials and Methods

Studies during 2001 were conducted at the Michigan State University

Entomology Farm, Ingham County, MI. The site consisted of 30 x 30 m plots arranged in a randomized complete block design with four replications containing “refuge” and control plots established in 1994 (Figure 2.1). Refuge plots contained a central 3.2 m refuge strip consisting of a combination of orchard grass (Dactylis glomerata L.), white clover (Trifolium repens L.), and several ﬂowering perennial plants (Carmona and Landis

1999). The control plots had no refuge and were planted to soybean. One half of each

23 plot was randomly selected and used for either predation studies or predator sampling during June of 2001. Predation of A. glycines was studied in the two rows approximately

8 m from the refuge or control strip (Figure 2.1). Sampling for predators occurred at 4 and 8 m from the refuge, and consisted of pitfall traps in rows and a predator observation area at 8 m from the refuge or control strip (Figure 2.1). The outer 6 m of the ﬁeld was not used to minimize edge effects. The crop area was managed using reduced primary tillage (chisel plow, disc) followed by secondary tillage (ﬁeld cultivation). The herbicide metolachlor (Dual 11) was applied at a rate of 2 l/ha. Potash was applied at a rate of 168 kg/ha to meet soil test requirements. Soybeans (Mycogen 5251RR) were planted on 5

May 2001 in 38 cm rows at a rate of 70,822 seeds/ha.

In 2002, studies were conducted in June at Michigan State University Crop and

Soil Sciences Research Farm, Ingham County, Michigan. The study field measured approximately 36 m x 117 m and was split into five 12 m x 24 m blocks. The outer 6 m of the field was not used to minimize edge effect. Each block was divided in half, with each 12 m x 12 m area randomly designated for either predator sampling or predation studies. Pitfall traps were placed in two rows 4 m apart, with a distance of 3 m between pitfall traps. A l m x 30 cm predator observation area was randomly placed within each

12 m x 12 m sampling area throughout the experiment. The crop area was managed using reduced tillage (fall chisel plow, secondary tillage with cultivator). Planting took place on 5 May, 2002 in 38 cm rows. The cultivar used was Mycogen 5251RR at a population density of 68,825 seeds/ha. Alpine liquid fertilizer (6-24-6 Fortiﬁed) was applied on 5 May at a rate of 38 l/ha.

Insects.

24 In both 2001 and 2002, infested soybean plants containing Aphis glycines

Matsumura were obtained from a laboratory colony maintained by the USDA-APHIS-

PPQ facility located in Niles, Michigan, where they were reared on Asgrow #3303 cultivar soybeans at 26 °C and 60 percent RH with a photoperiod of 16:8 (L:D). Once at

Michigan State university, they were placed in grow chambers at 25°C, 70 % RH, and

16:8 (L:D) photoperiod until used.

Predator Abundance and Sampling.

In both 2001 and 2002, ground dwelling predator abundance and composition was assessed using six 8.5 cm wide by 13 cm deep pitfall traps placed in two rows in each half of the plots with rims at the substrate level. The six pitfall traps were placed at a distance of 4 m and 8 m from the refuge or control strip, and at a distance of 3.7 m away from the previous trap. Pitfall samples were collected each 48 h during trials. In addition, there was one 1 m x 30 cm area in each sampling plot where ground and foliar predator density and species composition were recorded during 5 min. visual observations. Foliar predator observations were recorded when aphids were sampled at

15 and 24 h. The methodology was similar during 2002, except pitfall traps were placed in two rows in each block half and the l m x 30 cm predator observation area was randomly placed over a pitfall trap in each block.

Aphid Survival and Reproduction.

Early season A. glycines survival and reproduction was studied by conﬁning A. glycines adults on soybeans in clip cages allowing variable access to predators. Larger A. glycines with a visible cauda were considered adults. Three treatments (Open, exclosure, and leaky) of 1 cm (interior) circumference clip cages constructed of 1.8 cm (outside)

25 diameter Cresline® PVC pipe were used in three trials during 2001; 7-8 June (plants in

V2), 14-15 June (plants in V3), and 24-25 June (plants inV5), and three trials during

2002; 18-19 June (plants in V3), 24-25 June (plants in V5), and 2-3 July (plants in V6).

The aim was to do this study before, during, and immediately after natural A. glycines infestation. Fine-mesh brass screen (96 threads per in.) on cages allowed air exchange but kept aphids enclosed. Natural A. glycines were ﬁrst observed on 12 June 2001 at an initial density of 7.8, and on 14 June 2002 at a lower initial density of 1.2 aphids per m2.

In 2001, three adult female A. glycines were placed in each clip cage for trial one and then four per cage during the second and third trials. For all treatments, aphids were initially transferred into clip cages using a ﬁne camel hair brush and then clipped on the underside surface of soybean leaves. Four replicates of each cage treatment were placed on soybean leaves in each refuge or control plot that were each replicated in four blocks.

Cages were placed on the uppermost trifoliate of selected plants at 8:00 AM. In 2002,

ﬁve adult A. glycines were placed in six of each cage treatment replicated in ﬁve blocks.

A. glycines were given a 6 h acclimation period to settle before differential cage treatments were applied. At this point, the cages were completely removed from open treatments and the leaky cages had a 3 mm corks removed from the side. This allowed A. glycines to exit but prevented entry by predators larger than 3 mm. During 2001, cages were placed 4 m away from control or refuge strips and were placed on the uppermost leaves of plants large enough to support clip cages. In 2002, cages were randomly placed on plants with leaves large enough to support the clip cages. In both years, at 8:00 AM

(15 h) and 4:00 PM (24 h) the following day, cages were momentarily opened and data

26 on the proportion of live adult A. glycines remaining and the number of nymphs produced in each cage were recorded.

In 2001, an additional test was conducted as checks on cage performance. The

ﬁrst consisted of leaky and exclosure cages clipped onto wooden pot stakes in the ﬁeld.

Comparison of A. glycines numbers in these treatments allowed us to evaluate the propensity for A. glycines to escape or emigrate when conﬁned on a non-food source.

These internal controls were used as a check on the methods and were not statistically analyzed.

In general A. glycines confined to soybean leaves in clip cages in the field settled and began to feed within several hours. Aphis glycines in exclosure treatments were totally confined and could not leave the cages, while those in leaky cages could leave if they desired. Differences in numbers of A. glycines remaining in exclosure versus leaky cages thus represents those aphids that emigrated from the cage. Aphis glycines in open cages may be removed by predators or emigration, so a comparison of A. glycines remaining in open versus leaky cages reveals the impact of predation.

Data Analysis.

Data from 2001 from predator observation areas were analyzed as the mean number of predators per 5 min. per 1 m x 30 cm in refuge versus control plots. Data from

2001 pitfall traps were analyzed as the mean number of predators/trap/day in refuge versus control plots. A type III F -test for overall treatment effect determined statistical signiﬁcance of treatment effect of pitfall and direct predator observation data. Data from

2002 predator observations were summarized as the mean number of predators per 3 min. per 1 m x 30 cm in the ﬁve blocks. Data from 2002 pitfalls were summarized as the

27 mean number of predators/trap/day. Neither foliar or pitfall data was statistically analyzed during 2002 because no treatment was applied. During 2001 and 2002, data on the proportion of live adult aphids and the number of nymphs produced were analyzed by logistic and Poisson regression, respectively, using the GLIMMIX Macro link of SAS statistical program (SAS Institute 2000). When a signiﬁcant treatment*hour interaction existed, data was further sliced to reveal individual signiﬁcant comparisons within the LS

MEANS. When analyzing nymph data, a signiﬁcant effect of the number of A. glycines at the beginning of trial was sometimes encountered. This indicates that the number of adults entering the trial at the beginning differed between treatments. This can be explained by the fact that not all Of the adult A. glycines settled on plants when cages were removed. The number of adult A. glycines at the beginning of the trial was therefore applied as a covariate to equalize treatments. A P value of 0.1 was accepted as signiﬁcant in this study for several reasons. First, trials were designed to be conducted in

24 h to minimize potential loss of data due to rainfall, or temperature extremes.

However, this leaves little time for predation to occur. Thus, the risk of rejecting the Ho of predation effect is already quite substantial. Second, as the ﬁrst test of its kind on A. glycines in North America, there was no preliminary data with which to predict the power of the tests used. Thus, less restrictive criteria were considered appropriate. Graphs in this study were based on actual means and standard errors and not the estimated means and conﬁdence intervals from the Poisson Regression.

Results

Refuge Habitat Effects 2001.

28 Proximity to refuge habitats did not signiﬁcantly (F = 2.00, df = l, 18, P: 0.20; F =

0.10, df= 1, 18, P: 0.90; F = 0.34, df= l, 18, P: 0.60) affect the number of predators

Observed in the 1 m x 30 cm plots during any of the three trials. Pitfall trap catches were signiﬁcantly (F = 2.90, df=: 1, 18, P: 0.04) higher in control versus refuge plots during trial one; however, this effect did not persist into trials two or three (F = 0.08, df= l, 18,

P: 0.93; F: 0.40, df= 1, 18, P: 0.54). Similarly, there was no signiﬁcant refuge effect on predation of A. glycines adults or nymphs during any trial during 2001 (Table 2.5,

Table 2.6). Though there was a signiﬁcant interaction between refuge*treatment*hour during trial three for adults, slicing of the analysis showed that this interaction was between an individual cage at 15 and 24 h and was not associated with refuge. Therefore, refuge and control plot data was pooled for subsequent analysis.

Predator Density 2001-2002.

In 2001, predator density in predators Observations (1 m x 30 cm per 5 min.) was greatest in trial one, lowest in two, and intermediate in trial three (Figure 2.2), with ground dwelling predators predominant in all but trial three (Table 2.1). In trial one, the carabid beetle, Elaphropus anceps (Le Conte) was the most abundant predator, accounting for 70.7 percent of the total number of predators observed. During trial two,

E. anceps was still the most abundant predator, accounting for 55.6 percent of predators observed, with foliar-foraging predators making their ﬁrst appearance. Coccinella septempunctata (L.) adults accounted for 27.8 percent of the predators and was the second most abundant predator. In contrast, during trial three, foliar-foraging predators were overall more abundant. Harmonia axyridis (Pallas) larvae, 39.5 percent, H. axyridis adults, 18.6 percent, and Orius insidiosus (Say), 14 percent, made up a large portion of

29 the observed predators, but the carabid beetle E. anceps, 16.3 percent, were still abundant

Predator density in 2002 was much lower than in 2001, peaking at 0.8 predators per 1 m x 30 cm per 3 min. in trial two (Figure 2.2). During trial one, the two most abundant predators were spiders and C. septempunctata, both of which accounted for 40 percent of the predators observed during this trial. During trial two, spiders were again one of the more abundant predators, accounting for 87.5 percent of the predators observed. In trial three, though, the diversity of foliar-foraging predators increased.

Hippodamia convergens Guerin, Nabis spp., and 0. insidiosus adults all accounted for 20 percent of the predators observed, while spiders once again were the most abundant and made up 40 percent of the observed predators, though the abundance of predators during this year was quite low in all three trials (Table 2.2).

Pitfall Captures 2001-2002.

Pitfall captures in 2001 were greatest in trial one and lowest in trial three (Figure

2.3). Almost all of the ground-dwelling predators captured were carabid beetles. In trial one, E. anceps predominated in refuge and control plots and accounting for 51.1 percent of the total catches. In trial two, however, Scarites quadriceps Chd., 35.2 percent, was the most abundant. During trial three, Scarites subterraneus F., 33.3 percent, was the most abundant (Table 2.3).

In 2002, total trap captures (Figure 2.3) were very similar to 2001, and the species collected differed (Table 2.4). Notably, E. anceps was missing from 2002 collections and in all three trials. Clavina impressefrons Le Conte was the most abundant predator

30 comprising 22 percent of the catches in trial one, 29.7 percent in trial two, and 50 percent in trial three (Table 2.4).

While much remains to be learned about the consumption of aphids by ground dwelling predators, E. anceps and C. impressefrons have been shown to be capable of consuming A. glycines in laboratory feeding assays (Chapter 3). It is possible that these predators foraged on plants at night, but this was not observed because of the beetles’ nocturnal activity. Thus, the effectiveness of carabid feeding on establishing aphid colonies needs to be further investigated. However, in the case of foliar-foraging predators such as C. septempunctata, H. axyridis, H. convergens, and Nabis spp., they readily feed on A. glycines in feeding assays (Chapter 3) and were each observed to consume A. glycines in the ﬁeld (Fox Pers. Obs).

Predation on A. glycines.

Test of cage efﬁcacy:

Conﬁnement of A. glycines in leaky and exclosure clip cages on a wooden surface

(i.e. no food or water resources) showed that they readily disperse from leaky cages under unfavorable conditions. In one trial, 100 percent of the A. glycines left the leaky cage by

15 h, while in a second, 89.3 percent left by 15 h and 98.2 percent by 24 h. The survival in the exclosure cages at 24 h was very low (0-7 percent). This test indicates that A. glycines are unable to escape the exclosure cages on a ﬂat surface of wood versus a leaf surface but can readily emigrate from leaky cages.

Clip cage trials:

Figures 2.4 and 2.7 reveal the overall trends in adult A. glycines survival under the differing cage treatments. As expected, the number of A. glycines remaining in all cage

31 types was greater at 15 versus 24 h. At 15 h, the greatest percentage of A. glycines remained in the exclosure treatments with a mean of 84.7 :1; 2.3 percent across all trials

(n=6). The mean number remaining in the leaky cages was 77.8 i 4.7 percent (n=6), with 69.2 :1; 3.9 percent (n=6) remaining in the open treatment. At 24 h, the greatest percentage of A. glycines remained in the enclosure treatments with a mean of 73.5 i 2.0 percent (n=6). The number remaining in the leaky cages was intermediate at 62.6 i 3.5 percent (n=6), with only 49.5 i 5.5 percent (n=6). In both years, more aphids survived to

24 h in the ﬁrst trial with fewer remaining in the second and third trials.

2001 Adult Survival.

Analysis of predator exclosure treatments revealed a significant treatment effect in trials one and two (Figure 2.4). There was a significant treatment*hour interaction in trials one and three, due to differential numbers remaining in open versus exclusion treatments at 15 or 24 h. There was also a highly significant hour effect, with fewer adult

A. glycines remaining at 24 h than at 15 h in all trials. In trial one, signiﬁcantly fewer A. glycines remained in open versus exclosure cages and open versus leaky cages at 15 h

(Table 2.5, Figure 2.4). In trial two, significantly fewer A. glycines remained in open versus exclosure cages and leaky versus exclosure cages at 15 h. At 24 b there were significantly fewer adult A. glycines in open versus exclosure cages, and leaky versus exclosure cages in trial two. In trial three no significant effects occurred.

2001 Nymph Production.

In 2001, the greatest number of A. glycines nymphs per cage were generally produced in the exclosure treatments (Table 2.6, Figure 2.5). In trial one there was no signiﬁcant treatment effect during any trial (Table 2.6, Figure 2.5). All three trials

32 showed a significant hour effect, with more nymphs present at 24 versus 15 h. In trial two, there was a significant treatment*hour interaction, but further slicing of the data did not reveal further significance based on pair wise comparisons.

2001 Total Aphid Numbers.

A comparison of the total A. glycines number (adults + nymphs) remaining in the different treatments per cage is shown in Figure 2.6 and Table 2.7. There was a significant effect of treatment in trials one and two, and there was also a significant treatment*hour interaction in trial two. In all three trials there was a significant hour effect, with fewer adults remaining and more nymphs being produced. In trial one, fewer total A. glycines were found in open versus leaky treatments at 15 h fewer A. glycines in

Open versus both leaky and exclosure treatments at 24 h. In trial two there were signiﬁcantly fewer aphids in open versus leaky and exclosure cages at 24 h. In trial three, there was no signiﬁcant treatment effect.

2002 Adult Survival.

In 2002, there was evidence for predation on adult A. glycines (Table 2.8, Figure

2.7). There was a significant treatment effect in trials two and three. There was a highly significant hour effect in all three trials. In trial one there was no significant treatment effect (Table 8, Figure 2.7). During trial two, significantly fewer A. glycines were found in open versus exclosure cages and leaky versus exclosure cages at 15 h. At 24 h, significantly fewer A. glycines remained in open versus exclosure and open versus leaky cages at 24 h. During trial three there were significantly fewer A. glycines in open and leaky cages versus exclosure at 15 h. At 24 h, open cages contained fewer A. glycines than leaky or exclosure cages.

33 2002 Nymph Production.

In 2002 the greatest number of nymphs were generally produced in leaky or exclosure treatments, but differences were not consistent (Table 2.9, Figure 2.8). There was no significant treatment effect in trial one or three, but there was a significant treatment effect in trial two. In addition, there was a significant hour effect in trials one and two (Table 2.9). In trial two there were significantly more nymphs in exclosure versus Open or leaky cages at 15 h and 24h. There was no significant treatment effect in trial three, though there was a significant treatment*hour interaction, but this interaction did not reveal significant pair wise comparisons.

2002 Total Aphids.

In 2002, total A. glycines numbers per cage (adults + nymphs) was again higher in leaky and exclusion cages versus open cages (Table 2.10, Fig. 2.9). In trials one and two, there was a significant treatment effect. A significant treatment*hour interaction was found in trials two and three. In trial one this interaction showed a significant treatment effect at 24 h, but this was not the case in trial two, where the interaction revealed no further significance. In all trials there was a significant hour interaction. In trial one, there were significantly more A. glycines in leaky and versus exclosure treatments at 15 h, and leaky versus open treatments at 24 h. In trial two, there were significantly more A. glycines in exclosure versus open, and exclosure versus leaky cages at 24 h. In trial three, there was no significant treatment effect.

Discussion

34 Overall refuge plots did not alter the abundance or species composition of foliar- foraging or ground-dwelling predators and impact predation. One possible reason for

ﬁnding no signiﬁcant impact is that the plot size was small compared to the mobility of the dominant predators. The importance of habitat manipulation cannot be discounted, and results may have been different if the plot size was larger. This part of the study should be repeated in a larger area.

Examining adult, nymph and total A. glycines remaining in 2001 revealed several trends. First, the number remaining to 15 h was generally rather high. Although a trend of reduced numbers remaining in the more exposed treatments (leaky and open) was observed, it was seldom significant. More pronounced differences are seen at 24 h where a consistent trend for reduced numbers of aphids remaining in open treatments was observed. The strongest evidence of predation occurs when the numbers in open treatments were significantly less than exclosure treatments and when there was a significant difference between the open and leaky cages. When significantly more A. glycines remained in leaky versus open cages, there was evidence that A. glycines were also being lost to predation in addition to those that may have emigrated.

The case for predation is weakened but not excluded in cases where there was no signiﬁcant difference between open and leaky cages. During 2001, there were signiﬁcantly fewer A. glycines in open versus exclosure treatments at 15 h but not 24 h.

While trends for lower numbers of remaining nymphs did exist, there was no signiﬁcant treatment effect during any trial. Evidence based on the total number of A. glycines remaining suggests predation was likely responsible this reduction during trial one of

2001, where there were signiﬁcantly fewer A. glycines present in open versus leaky and

35 exclusion cages at 24 h. There is also evidence for predation during trial two because there were signiﬁcantly fewer aphids present in open versus leaky cages and open versus exclusion cages at 24 h.

Data from 2002 yields overall stronger evidence for predation, with signiﬁcant reductions in the number of adult A. glycines remaining in two out of three trials at 24 h.

In trial two there was evidence for predation because there were signiﬁcantly fewer A. glycines remaining in Open versus leaky cages and open versus exclusion cages at 24 h.

A similar trend existed in trial three, where significantly fewer A. glycines remained in open versus leaky and exclosure cages at 24 h. There did not seem to be significant predation on nymphs during 2002. The reason for this is not known. Perhaps due to their small size, nymphs may be more difficult for predators to locate. In the case of the total number of A. glycines, there was no statistical evidence for predation.

In summary, there was strong evidence for predation on adult A. glycines in one out of six trials at 15 h and two out of six trials at 24 h. There was no evidence for predation of nymphs in any trial. Furthermore, there was strong evidence for predation based on the total number of live A. glycines in two out of six trials at 24 h. Thus, as a result of these trials, the Ho that predators do not impact A. glycines establishment is rejected. In contrast, these studies establish that predation can be a signiﬁcant factor inﬂuencing establishment of A. glycines populations in soybean.

Our studies also shed light on which predators are responsible for reductions in A. glycines establishment. Signiﬁcant predation occurring at the 15 h mark may in part be attributed to the activity of nocturnal predators. While many carabid beetles were present in the plots and our no-choice tests show that most of the carabid species could consume

36 A. glycines (Chapter 3), results show this predation to be limited. This is in contrast to

Hance (1987), who attributed reductions in aphid populations to predatory carabids and

Ostman et al. (2001), who also found carabid beetles and other ground-dwelling predators to successfully reduce low density aphid populations. Other investigators have noted an interaction between foliar predators that disturb aphids and cause them to drop and predation by ground-dwelling predators (Losey and Denno 1998). A preliminary study indicated that A. glycines do not drop in response to disturbance. Thus, this was an important factor in our result. Since A. glycines does not readily drop from plants, ground-dwelling predators would have to climb plants to encounter most A. glycines.

However, this study is in agreement with the ﬁndings of Chang and Kareiva (1999), and

Takagi (1999), who suggest that generalist predators may be most effective by reducing initial pest populations before they reach outbreak levels. Even a low level of nocturnal feeding may thus reduce initial numbers so that diurnal predators foraging on soybean leaves would be better able to reduce populations.

During the later spring trials, more specialized (aphidophagous) predators became predominant and were responsible for aphid reduction. During the second and third trials in both years aphids were reduced the most at 24 h. During these studies predators were observed consuming A. glycines on eight separate occasions feeding on aphids in the open treatment. This included observations of predation by early instar H. axyridis larvae, and 0. insidiosus nymphs (Fox unpub. data). Considering the very low density Of

A. glycines in the ﬁeld at this time, these observations of early season predation on A. glycines should not be underestimated and are consistent with the known ability of

37 aphidophagous predators to search for and ﬁnd prey at low densities (Grasswitz and

Burts 1995, Stary 1995, Van den Berg et al. 1997, Obrycki and Kring 1998).

Overall, this research points to the importance of early season predators in reducing the establishment of A. glycines in soybean. A complex of predators including both generalist and more specialized natural enemies contributed to these effects.

Establishment success was reduced to a greater degree later in the spring. This may indicate a potentially important interaction between crop planting date, A. glycines arrival and conservation of predators in the ﬁeld. Further research should concentrate on means to improve the reliability of predators in reducing early season A. glycines survival and establishment.

38 x m l

% 2.3

4.7 4.7 in 18.6 14.0 16.3

39.5

Lansing, 100.0%

Three East

1 8 36

2 2 7 6

Trial

Total 2001), observations

% June

5.6

11.1

27.8

55.6

100.0%

Two ﬁve-minute (24-25 in

1 10

Trial

Total three trial predators

9.2

4.6 15.4

and

70.7 100.0%

One glycines

2001),

65 Trial 46

Total A. June potential

(14-15

adult

adults Say two

total

larvae

adults (L.)

of Guerin

trial

Conte)

adults

(Le

(Pallas) (Pallas)

percent (Say) 2001),

and

convergens anceps

June

spp.

septempunctata quadrimaculatum

axyridis axyridis

(7-8 insidiosus number one

total

Spiders

Coccinella

Orius

Hippodamia Harmonia

Harmonia Elaphropus

Bembidion Species Bembidion trial during composition,

soybean

Anthocoridae

Carabidae Coccinellidae Of Species areas 2.1.

Other

Heteroptera:

Coleoptera: Coleoptera:

Family TOTALS 30 Michigan Table

39 m 1

in Lansing, % 20.0 20.0 40.0

Three East Trial 1 1 1 2 5 Total

observations

2002), July % - -

- (2-3 12.5 87.5 Two 100.0%

three-nrinute three in 1 Trial [\OO -

Total

trial and predators % - 20.0

40.0 40.0

One 2002),

100.0%

glycines June Trial A.

1 N W 2 - -

Total (24-25

potential two adults adults

total trial of (L.)

Guerin 2002), adults

percent June (Say) and convergens

septempunctata (18-19

number, one spp. insidiosus

total trial Coccinella Orius Spiders Hippodamia Species

Nabis during

composition,

soybean of Anthocoridae Nabidae Coccinellidae

Species areas

2.2.

cm 30 Heteroptera: Heteroptera: Others

TOTALS

x Family Coleoptera: Michigan Table

40 plots

- -

- - - -

8.3

16.7

33.3

25.0

100.0% control

Three and

1 12

- -

- - - 3 - - - - -

Total

Trial Michigan refuge in

- - -

1.4 1.4

1.4

2.8

5.6

2.8 7.0

4.2

11.3

14.1 12.7 Lansing,

35.2

100.0% traps

Two East pitfall

10 Trial 00

25 OVF‘v—‘MNV‘I 2 71 - - -

Total in 2001), June

% collected

1.0

1.9 1.0

1.9 1.0 8.7

7.8 2.9

2.9 2.9

4.9

11.7

51.5

100.0% (24-25

One three predators

Trial

103

12 1 1 CNN—‘00 3 3

- 3

2 5 53

Total of trial

) total

Say and Of

(F.

Conte

(III.)

(DeG.)

Conte)

)

Le 2001),

(Say)

Say

(F.)

(F.

Chd. percent

(Le

(Say) June

(Schrank) and

santaecrusis

(DeG.)

melanarius

comma

anceps

spp.

quadrimaculatum (14-15

herbivigus

aﬁim's

pensylvanicus

lucublandus

chalcites

cupripenne

quadriceps subterraneus

impressefrons

bipustulata

aenea number, two total trial

Clavina

Harpalus Poecilus Poecilus Pterostichus Scarites Scarites Stenolophus

Hatpalus

Bembidion Elaphropus Harpalus

Species

Bembidion

Agonum Anisodactylus

Amara 2001), June composition,

Carabidae (7-8 Species one trial 2.3.

TOTALS

Family Coloeptera:

Table during

2002),

_ -

8.3 8.3

8.3

6.7

50.0

100.0%

June

Three

Total

Trial

(18-19

one

trial

- -

1.4

8.1

5.8

8.7 7.2

7.2

17.4

11.6

31.9

Two

100.0%

during

Trial

5 -

Total

traps

pitfall

3.5

2.5

3.7

2.5

4.9

12.3

13.6 14.8

16.0

23.5

One

100.0%

Michigan

collected

Trial

12 19

3 2

- 2

Total

Lansing,

predators

)

East

Say

(F.

Conte

(III.)

)

2002),

(Say)

percent

(F.)

(LeC.)

Chd.

(F.

(Say)

July

and

(Schrank)

santaecrusis

(2-3

melanarius

comma

rapidum

spp.

quadrimaculatum

aﬂinis

lucublandus

chalcites

number,

quadriceps

subterraneus

impressefrons

bipustulata

three

total

trial

Clavina Clavina

Stenolophus

Scarites

Scarites Pterostichus

Poecilus Poecilus

Harpalus

Bembidion Bembidion Bembidion

Species

Anisodactylus

and

2002),

composition,

June

Carabidae)

Species

(24-25

2.4.

two

TOTALS (Coloeptera:

Family

trial Table

42 plots

control

0.41 0.38 0.74 0.01* 0.83 0.05* 0.09*

and

Three F

3.12 2.47

0.96 0.30 0.05 0.00

26.51

refuge

Trial

remaining * 0.1 P

glycines

0.13

0.47 0.59 0.04* 0.34 0.44 0.001* 0.57 0.75 0.01* 0.07* 0.01 P

<0.001*

adult Indicates

live

hhhbh * Of \oxooxgxoo (43v—‘v—1 F1 v—Qv—In

0 0‘ :-

mmm—J—rmm line. Michigan

proportion

dotted *

the

0.88 0.44 0.43 0.66 0.02* 0.12

0.80 009* 0.001 0.49 0.04* 0.37 0.03*

on below < Lansing,

One

based F shown East 1.12 2.45 3.31 0.22 0.73 0.02 27.63

Trial are 2001,

-statistic

158 158

158 158 df MEANS

1,158 1,158 during

2, 2, 2, 2, LS

greater

Open Open

of from Open

Exclosure Exclosure treatments

Open

versus versus cage versus versus

versus

Probability comparisons

Exclosure Leaky

Leaky Exclosure Leaky

Leaky

2.5. different wise h 15 Pair Refuge*Treatment*Hours Hours Refuge*Hours Refuge Refuge*Treatment Treatment*Hours under Effect Treatment

Table

43 control

------

and

0.80 0.20 0.44 0.005* 0.69 0.29

0.001* 0.43

Three

refuge

1.63 1.23

8.8

0.08 0.85

0.83 0.16

32.84

Trial

7 7

7 7 7 7 7

cage 6 6 6 6 6 6 16

per 1 1 l, 1 2 2 2 2

produced

0.1

- -

- - - -

_<_ 0.001* 0.19 0.80 0.47 0.17 0.04* 0.001* 0.78

P < <

nymphs

Two

F 1.75 3.18 0.22 0.25 0.75 4.50 10.53 40.90

Indicates

Trial

glycines

A. 166 166 166 166 166 166 3

line. 1, 1 1, 1, 2 2, 2,166 2, Michigan

number

dotted Lansing,

the

- - - - - 100 0.73 0.001* 0.42 0.33 0.13 0.56 0.11

below < East

One

based

shown 2001, 1.11 2.27 2.03 0.00 0.12 0.66 0.42 24.86

Trial

are during

-statistic

F 57 57 3 157 157 157 157 157

df , ,

MEANS 1, l, 1 l 2, 2, 2, 2

greater treatments

a Open Open

from

cage

Exclosure

Exclosure Open

versus versus

versus

versus versus different

Probability

comparisons

Beginning

Exclosure Leaky-Open

Exclosure Leaky

Leaky Leaky

2.6. under

wise h 15

Pair Refuge*Hours Refuge*Treatment Refuge Hours Refuge*Treatment*Hours Number Treatment*Hours Treatment

Table plots Effect

44 control P and ------0.003* 0.10 0.25 0.75 0.20 0.73 0001* 0.65 < refuge Three F in 1.38 1.62 2.31 0.32 0.21 0.13 13.43 24.74 Trial 167 167 167 167 167 167 167 3 glycines df l, 1, 1, 1 2, 2, 2, 2, A. live

of 0.1

P g 0.005* 0.05* 004* 0.69 0.36 0.71 0.02* 0.28 0.001* 0.53 0.19

0.09*

0.76 0.001* P < < nymphs) + Two F 1.19 2.82 2.48 5.47 0.27 0.63 13.67

25.36 Indicates Trial

(adults * 166 166 166 166 166 166 166 3

(If line. 1, 1, 1, 1 2, 2, 2,

2, Michigan number

total dotted

the Lansing, P 0.67 0.07* 0.20 0.26 0.03* 0.03* 0.18 0.82 0004* 0.22 0.63 0.39 0.96 0.06*

on below

< East One based

shown 2001, 1.73 1.52 8.66 2.95 0.24 0.95

0.00

0.20 are

Trial during -statistic F 157 157 157 157 157 157 157 157

df MEANS 1, 1, 1, 1, 2, 2,

2, 2, LS greater

a treatments Open Open

from cage Exclosure Open Exclosure Open versus versus versus versus versus

versus different

Probability comparisons Beginning Exclosure Exclosure Leaky Leaky Leaky Leaky

2.7.

under wise h

15 Pair Refuge*Hours Refuge*Treatment*Hours Hours Number

Refuge*Treatment Treatment*Hours Refuge

Effect plots Treatment Table

45 cage P

0003* 002* 009*

0.50 0006* 0.31 0002* 0001*

<0001* Three different

7.46

52.88 Trial under

73.2

1,804

remaining 0.1 P

glycines _<_

0001*

0.30 0.22 0002* 0002* 0011*

0003* P

<0001*

A. <0001*

Two adult

12.49

97.40 Indicates

Trial

live * of

83.9 line.

1,974

proportion dotted

the below on

One

based shown

Trial

are Michigan

-statistic °9°l"1

F O\O\—‘ [\ONON 0

(xi—TN MEANS

Lansing, LS greater a Open

Open East

of from Exclosure Open

Exclosure Open 2002, versus versus versus versus versus versus

Probability

during comparisons Exclosure Leaky Leaky Exclosure Leaky Leaky

2.8. wise h h

15 Pair

24 Hours

Effect

Treatment*Hours

Treatment treatments Table

46 under plots

003*

010* 0.75 0.92

Three soybean F 2.42 0.08 0.10 4.82 in

Trial cage

75.6 77.0 76.7 76.2 df per

2, 2,

produced 0.1

P g

003* 0002* 010* 0007* 0.19

0003* 0.32

001*

0002* 0.56 P nymphs

Two F 1.16 8.84

6.48 6.46 Indicates

Trial

glycines * A.

86.0 82.9

81.8

of df

1,860 line.

2, 2,

number dotted the

Michigan 0.11 0001*

0.33

0.29 below

on <

One based

F shown

1.12 1.11 Lansing, 2.26

24.23

Trial

are East -statistic F 87.9 87.7 77.3

1,988

2002, MEANS 1,

2, 2, LS greater

a during Open Open

of from Open Exclosure Exclosure Open versus

versus treatments versus versus versus versus

Probability

comparisons cage beginning Leaky Exclosure Exclosure Leaky Leaky Leaky

2.9. wise h h

15 Pair 24 Hours Number Treatment*Hours

Effect Treatment Table different

47 different P

0.14 006* 0001*

0.50

under

Three

F present

8.7

2.9 2.2

0.5 Trial

76.7

75.9 glycines (If

1,778

2,773

2, 2,

live

of 0.1

0.20

0003* 001*

0005* 007* 008*

0001*

P +nymphs)

Two

F 2.78 7.13 0.77

13.31

(adults Indicates

Trial * 84.8 83.1 81.2 84.8

df line. 1, 2,

2, 2,

number

total

dotted the

P on

0.26 003*

0.25 0.21

0.84 006* 008*

009* 008* 0.22 below

One

based shown

1.52 Michigan 3.12 3.16

2.41 are

Trial

-statistic

Lansing,

87.7

87.9 79.1

(If

1,988

MEANS

2, 2,

East

greater

Open

2002,

from

Exclosure

Open

Exclosure

Open

versus

during

versus

versus versus versus

Probability

comparisons

beginning

Exclosure Leaky

Exclosure Leaky Leaky

Leaky

2.10.

treatments

wise

15h

Pair

Number

Hours

Treatment*Hours

Effect

Treatment

Table cage

48 32m

64m

Block 1 Block 11 Block 111 Block 1V

III—:1 Refuge strip 1:1 Control strip

Control or Refuge 3.2m Control or Refuge

4m 1 a o o 1

300m I'm-“w 12m * ** ** * ** 8m 0 L 0 8m * ** * 1m

Crop area 15m * = Clip cage area

<— + 0 = Pitfall traps [:1 = Visual sampling area

Figure 2.1. Layout of the 2001 ﬁeld location on the Michigan State University

Entomology Farm, East Lansing, Michigan. The ﬁelds were separated into four blocks, each containing a refuge and control plot. Enlarged plots illustrate where pitfall traps and foliar sampling areas were located, and where clip cages were placed.

49 .52 IA. .9 , I I i i i 7’ ’ i * T2001? tn 5 4 . E U a l ’ >< 3 "

8 L. T a 1 c: J. “3 i E i 0 L ., 7., LL;

. Trial One Trial Two Trial Three C ._ 5 .._ - M-,—H rev . _ , .- . .L' E .B. 2002; m l l E 4 ‘ f o O m >< 3 . I E i 1 § 2 ‘ 1 N '8 . ‘5 1 l T f s. I—I‘j 1 [—1—] Q) . . 2 0 - » . ____ I i Trial One Trial Two Trial Three

Figure 2.2. Mean number (i SEM) of predators observed in 1 m x 30 cm areas during;

A. 5- minutes in trial one (7-8 June 2001), trial two (14-15 June 2001), and trial three

(24-25 June 2001) and; B. 3-minutes during trial one (18-19 June 2002), trial two (24-25

June 2002), and trial three (2-3 July 2002), East Lansing, Michigan.

50 ; IA T200T I E Q I S ‘5 I I § ‘ T I N E 1' .1. I a Q I T S I I I ‘51O I I

5d) I 2 OI -C- CLEO

Trial One Trial Two Trial Three

: 2713. T T T' T T T T T $02 T 3;.'0 I g T I 5 l T l a“? 1 l a I '0 aE I. at l 1: a 65 0 If:

Trial One Trial Two Trial Three

Figure 2.3. Mean number (i SEM) of predators per trap per day collected in 32 pitfall traps; A. located in refuge and control plot during trial one (7-8 June 2001), trial two (14-

15 June 2001), and trial three (24-25 June 2001); B. in 30 pitfall traps in ﬁve blocks during trial one (18-19 June 2002), trial two (24-25 June 2002), and trial three (2-3 July

2002), East Lansing, Michigan.

51 Figure 2.4. Percentage of live adult A. glycines remaining at 15 and 24 h for the open, leaky and exclosure treatments in refuge and control plots during; A. trial one (7-8 June

2001); B. trial two (14-15 June 2001); and C. trial three (24-25 June 2001), East Lansing,

Michigan. Different letters above bars within. an hour denotes signiﬁcant (P _<_ 0.1) differences among open, leaky, and exclosure treatments (Poisson regression test).

52 110 100 - 3,. 90

80 - 70 '

50 O

h 0

and 0

a b

remaining

i '.

glycines

.24“-

adult

. m

Percent

a a

833838885888883888‘Nw“

15 H 15H 24 H 24 H I‘ 24 H Leaky Exclosure Open Leaky Exclosure I

Figure 2.5. Number Of live A. glycines nymphs at 15 and 24 h in three cage treatments during; A. trial one (7-8 June 2001); B. trial two (14-15 June 2001); and C. trial three

(24-25 June 2001), East Lansing, Michigan. Different letters above bars within an hour denote signiﬁcant (P g 0.1) differences in the number of nymphs present in open, leaky, and exclosure treatments (Poisson regression (log-linear model) test).

h . N

24 w

and

O‘

nymphs

glycines

Number

N O

T dm

.1. .1.

15h 15h 24h

Open Leaky Exclosure Open Exclosure

Figure 2.6. Total number of live adult and nymph A. glycines at 15 and 24 h in h in three cage treatments during 1) trial one (7-8 June 2001), 2) trial two (14—15 June 2001) and 3) trial three (24-25 June 2001), East Lansing, Michigan. Different letters above bars within an hour denote signiﬁcant (P g 0.1) differences. among open, leaky, and exclosure treatments (Poisson regression (log-linear model) test).

56 a . I Adults C1 Offspring I

N h

and

present

glycines

nymph

and

adult

number

O‘

Total

Leaky l Exclosure Open Leaky Exclosure ‘

Figure 2.7. Percentage of live adult A. glycines remaining at 15 and 24 h for the open, leaky and exclosure treatments in during; A. trial one (18-19 June 2002); B. trial two (24-

25 June 2002); and C. trial three (2-3 July 2002), East Lansing, Michigan. Different letters above bars within an hour denote signiﬁcant differences among open, leaky, and exclosure treatments (Poisson regression test).

58 1 10 00 a 90 _I. 80 0 6O 50 4o 30

I 20

24 1 10

and 100 15 90 4 b at 80 70 60 remaining 50 40 30

glycines

A. 20 1 10 adult 00

of 90 80 b

Percent

15H 15H 15H 241-1 24H 24H

Open Leaky Exclosure Open l Leaky Exclosure '

Figure 2.8. Number of live nymph A. glycines at 15 and 24 h in three cage treatments during; A. trial one (18-19 June 2002); B. trial two (24-25 June 2002); and C. trial three

(2-3 July 2002), East Lansing, Michigan. Different letters above bars within a given hour denote signiﬁcant (P g 0.1) differences in the number of nymphs present in Open, leaky, and exclosure treatments (Poisson regression (log-linear model) test).

60 A.

10 ~

8 f

a 6 .

I a a 4 l a T iii h T .I. a I

2 . .L 24 T i 'L 5? :t if. if and 0 . 7

at 12 B.

10 ~ a menu nymphs

8 I 11 the:

glycines 6" h b

;\ en 110:1 A.

T T

of 4 1 1 . Ital}.

2 . Number

0 12 - ~- C. 10 -

8 ' a a a a T T 6 T l. a

4 T l

2 . I

0 . I 15H 1511 15 H 24 H I 24H

Open Leaky Exclosure Open Leaky Exclosure

Figure 2.9. Total number Of live adult and nymph A. glycines at 15 and 24 h in three cage treatments during; A. trial one (18-19 June 2002); B. trial two (24-25 June 2002) and C. trial three (2-3 July 2002), East Lansing, Michigan. Different letters above bars within a given hour denote signiﬁcant (P g 0.1) differences amongopen, leaky, and exclosure treatments (Poisson regression (log-linear model) test).

62 .P- -—'

iAdults T I I I Cl Nymphs I

h 10

24 8

6 .g

and

4 4

1__,

at 2

I,“

I 18 i

present

t‘,

16 41......

14 f

glycines

1*“ h—D—t

A. 1

.I_¥

nymph

and

ON-P-OOOON

adult of I

a a number I a

Total I I

15 H 15H 15 H 24 H 24 H I 24 H I I Open Leaky Exclosure Open I Leaky I Exclosure

References Cited

Alﬁler, A. R. R., and V. J. Calilung. I978. The life history and voracity of the syrphid predator, Ischodiodon escutellaris (F.) (Diptera: Syrphidae). Philippine Entomologist. 4(1-2)105-117.

Bowie, H. M., G. M. Gurr, Z. Hossain, L. R. Baggen, and C. M. Frampton. 1999. Effects Of distance from ﬁeld edge on aphidophagous insects in a wheat crop and observations on trap design and placement. International Journal of Pest Management. 45: 69-73.

Chang, G.C., and P. Kareiva. 1999. The case for indigenous generalists in biological control, pp. 103-115. In B. A. Hawkins and H. V. Cornell, ed. Theoretical approaches to biological control. Cambridge University Press, UK.

Debach, P., and D. Rosen. 2001. Biological control by natural enemies. Cambridge University Press, Cambridge.

Han, X. 1997. Population dynamics of the soybean aphid Aphis glycines and its natural enemies in ﬁelds. Hubei Agricultural Sciences. 2: 22-24.

Hance, T. 1987. Predation impact of carabids at different population densities on Aphis fabae development in sugar beet. Pedobiologia. 30: 251-262. Hilbeck, A., C. Eckel, and G. Kennedy. 1997. Predation on Colorado potato beetle eggs by generalist predators in research and commercial potato plantings. Biological Control. 8: 191-196.

Landis, DA, and W. Van der Werf. 1997. Early-season aphid predation impacts establishment and spread of sugar beet yellows virus in the Netherlands. Entomophaga. 42: 499-516. Landis, D. A., S. D. Wratten, and G. Gurr. 2000. Habitat manipulation to conserve natural enemies of arthropod pests in agriculture. Annual review of entomology. 45: 173-199.

Lee, J. C., and D. A. Landis. 2002. Noncrop habitat management for carabid beetles. Pp: 279-303. In J. Holland, ed. Carabid beetles and agriculture, Intercept, Adover.

Lee, J. C., F. D. Menalled, and D. A. Landis. 2001. Refuge habitats modify insecticide disturbance on carabid beetle communities. Journal of Applied Ecology. 38: 472-483.

Lenné, J.M., and P. Trutmann. 1994. Diseases of tropical pasture plants. CAB International, Wallingford, UK. P. 69.

Li, W.M., and Z.Q. Pu. 1991. Population dynamics of aphids and epidemics of soybean mosaic virus in summer sown soybean ﬁelds. Acta Phytophylactica Sinica. 18(3): 123- 126.

Losey, J. E., and R. F. Denno. 1998. The escape response of pea aphids to foliar- foraging predators: factors affecting dropping behaviour. Ecological Entomology. 23: 53-61.

Obrycki, J. J. and T. J. Kring. 1998. Predaceous Coccinellidae in biological control. Annual review of entomology. 43: 285-321.

65 Ostman, O., E. Ekbom, and J. Bengtsson. 2001. Landscape heterogeneity and farming practice inﬂuence biological control. Basic and Applied Ecology. 2: 365-371.

Losey, J E and RF. Denno. 1998. The escape response of pea aphids to foliar-foraging predators: factors affecting dropping behaviour. 23: 53-61. Ecological Entomology.

SAS Institute. 2000. SAS/STAT user’s guide, release 8.1 led. SAS Institute, Cary, NC.

Schneider, F. 1971. Bionomics and physiology of aphidophagous syrphidae. Annual Review of Entomology. 12: 103.

Snyder, W.E., and DH. Wise. 1999. Predation interference and the establishment of generalist predator populations for biocontrol. Biological Control. 15: 283-292.

Syndmondson, W.O.C., K.D. Sunderland, and M.H. Greenstone. 2002. Can generalist predators be effective biocontrol agents? Annual Review of Entomology. 47: 561-594.

Stary, P. 1995. Natural enemy spectrum of Aphis spiraephaga (Horn: Aphididae), an exotic immigrant aphid in central Europe. Entomophaga. 40(1): 29-34.

Takagi, M. 1999. Perspective of practical biological controll and population theories. Researches on population ecology. 41: 121-126.

University of Minnesota Extension Service Web Site: http://www.soybeans.umn.edu/crop prod/insects/aphid/aphid.htm

Van Den Berg, H., D. Ankasah, A. Muhammad, R. Rusli, H.A. Widayanto, H.B. Wirasto, and I. Yully. 1997. Evaluating the role of predation in population ﬂuctuations of the soybean aphid, Aphis glycines in farmers’ ﬁelds in Indonesia. Journal of Applied Ecology. 34, 971

66 CHAPTER 3

Evaluating the impact of indigenous and exotic predators of the soybean aphid, A.

glycines, in laboratory no-choice feeding assays

Introduction

The soybean aphid, Aphis glycines Matsumura, is an exotic pest from Asia that was discovered in the United States in 2000. In Asia A. glycines feeding can cause up to a 20 cm reduction in growth and a 27.8 percent reduction in seed yields (Wang et al.

1996). In 2000, it was found that up to a 13 percent yield reduction occurred in replicated ﬁeld plots in Wisconsin (University of Minnesota Extension Service Web

Site). In 2001, this aphid caused a 40 percent yield loss in (Difonzo, Pers. Comm). This pest may also indirectly harm soybeans by vectoring persistent viruses, such as soybean dwarf virus and non-persistent viruses, such as soybean stunt virus, soybean mosaic virus, bean yellow mosaic virus (Van den Berg et al. 1997). Epidemics of soybean mosaic potyvirus in summer-sown soybean ﬁelds in Jiangsu, China were found to be closely related to the timing of A. glycines immigration (Li and Pu 1991). In addition, aphids also cause indirect damage by excreting honeydew onto foliage, which promotes the growth of sooty molds that reduce the photosynthetic capacity of the leaves (Lenné and Trutmann 1994, Hirano et al. 1996).

Foliar-foraging predators can contribute to the control of a wide variety of aphid species (Grasswitz and Burts 1995, Stary 1995, Obrycki and Kring 1998). For example, it was found that a complex of natural enemies, including Chrysopa nigricornis

67 Burmeister, Orius spp., Coccinella transversoguttata Faldermann, Hippodamia convergens Guerin, and various species of Syrphidae and Chamaemyiidae all contributed to the reduction of Aphis pomi De Geer in a ﬁeld study using apple trees (Grasswitz and

Burts 1995). Similarly, Stary (1995) found that Coccinella septempunctata L., Adalia bipunctata (L.), and Episyrphus balteatus (De Geer) were common predators that aided in control of the spiraea aphid, Aphis spiraephaga Muller in Czechoslovakia.

There is currently no published data on the ability of folia-foraging predators to consume A. glycines in United States. In Asia it is known that coccinellid (Harmonia spp.) predators play an important role in suppressing A. glycines populations in soybean

ﬁelds in the tropical Southeast at higher temperatures (Van den Berg et a1. 1997).

Another study suggested that Nabis spp., Harmonia axyridis (Pallas), Coccinella septempunctata L., and Chrysopa spp. aid in A. glycines control during the mid to late growing season in China (Han 1997). At this time it is not known if these species will consume A. glycines in the Michigan, or which other common predators in soybean ﬁelds will aid in its control.

Ground-dwelling predators may also play an important role in aphid biological control (Hance 1987, Landis and Van der Werf 1997, Snyder and Wise 1999, Chang and

Kareiva 1999, Symondson et al. 2002). For example, it has been shown that ground beetles (Carabidae) species can of reduce populations of Aphis fabae Scopoli, a pest of sugar beet (Hance 1987). It was also demonstrated that many species of carabid beetles in sugar beet ﬁelds in the Netherlands, along with spiders and a dominant cantharid generalist predator, Cantharis lateralis aided in aphid control early in the season (Landis

and Van der Werf 1997). No data exists to assess the ability of ground-dwelling predators to consume A. glycines in North America.

Petri dish or small cage laboratory studies provided an avenue for quickly assessing predation on numerous pest species by a wide variety of potential predators

(Bilde and Toft 1997, Sebolt 2000, Schmaedick and Shelton 2000). Landis and Van der

Werf (1997) used Petri dish assays to determine potential predators of A. fabae, and their possible preference for nymph or adult aphids. Therefore, the objective of this study was to examine the capability of common predators in Michigan soybean ﬁelds to consume A. glycines adults in laboratory no-choice feeding assays using techniques similar to Landis and Van der Werf (1997). Using a no-choice test, it was determined which species can feed on A. glycines adults and nymphs, and if these predators have a preference for aduls or nymph A. glycines. Attention was focused on these predators as potential predators of

A. glycines in the ﬁeld.

Materials and Methods Insects.

A colony of A. glycines was Obtained from the USDA-APHIS-PPQ facility, Niles,

Michigan and reared on greenhouse grown soybeans (Mycogen 5251RR) in growth chambers at 25° C, 70 % RH, and 16:8 (L:D) photoperiod. Plants were changed on a weekly basis, and adults were transferred from old to new plants with a ﬁne camel hair brush.

Predator Collection.

Foliar-foraging and ground-dwelling predators used for feeding assays were collected from soybean ﬁelds located at the Michigan State University Entomology

69 Research Farm, Ingham County, Michigan. From 7 June to 9 August 2001 foliar- foraging predators were collected with sweep nets from soybean foliage. Ground- dwelling predators were collected in dry pitfall traps placed in soybean ﬁelds. Pitfall

traps were opened for 62 h each week, and checked every 24 h.

Feeding Assay.

Predators were returned to the laboratory, held at approximately 24¢ 2° C and placed into individual Petri dishes (90 mm diam. with a 4 mm screened hole in the lid for

ventilation) lined with moistened ﬁlter paper for a 24 h acclimation/starvation period. A

2 cm wet cotton wick in each dish moistened with distilled water provided moisture.

Predator availability varied between species and over time; the number tested varied from a minimum of ﬁve to a maximum of 60 individuals. All feeding assays were conducted from 8 June to 10 August 2001. After 24 h of acclimatization for predators, ten adult aphids were placed onto a soybean leaf and then transferred to each experimental dish containing a predator. Larger aphids with a visible cauda were considered adults in thsi study. During each trial, ﬁve to ten control dishes without predators were also established to evaluate survival in the absence of predation. Aphis glycines and predators were left in experimental and control dishes for 24 h after which predators were removed and the number of remaining (out of ten) A. glycines were counted. At this time nymph aphids that were produced were also counted.

Analysis.

At the end of the season, adult and nymph aphid numbers in experimental and control dishes were pooled and compared for each predator species tested (SAS Institute 2000).

Since only several of a given species were often collected on a given date, and survival of

aphids in control dishes did not differ signiﬁcantly by date, individual predator dishes and their controls were pooled throughout the season, which yielded a stronger statistical analysis. A pooled t-test with equal variance was used on the number of aphids remaining in dishes to evaluate feeding on A. glycines for all predator species.

Experimental and control dishes were not paired for analysis, since control dishes were compared to all experimental dishes. We calculated percent mortality (= predation rate) for both adults and nymphs that were produced during the trial. Nymphal mortality was deﬁned as an expected reduction in both control and experimental dishes. Expected reduction was calculated by averaging the number of nymphs produced in experimental dishes divided by the average beginning and ﬁnal adult aphid number in experimental dishes. When the same procedure was done in control dishes, nymphal mortality could be determined by dividing expected reduction in predator dishes by that in control dishes.

After this, nymphal mortality was contrasted to adult mortality, allowing the measurement of preference. A mortality ratio greater than one indicates preference for nymphs over adults.

Results

All foliar predators consumed signiﬁcant (P 3005) numbers of A. glycines adults when compared to survival in the control (Table 3.1). The mean numbers of A. glycines surviving were consistently low for almost all coccinellid species, including Coccinella septempunctata (L.), Harmonia axyridis (Pallas), and Hippodamia convergens Guerin, but C. septempunctata larvae consumed the most, with an average of 1.1 i 0.3 A. glycines surviving at the end of the 24 h period. Other foliar predators that consumed 70 percent

71 or more of the A. glycines presented, included Chrysopa spp. larvae, and Nabis spp. adults and nymphs. Several foliar-foraging predators consumed less than 50 percent of the A. glycines presented, including Coleomegilla maculata De Geer adults, which allowed the most A. glycines to survive (6.4 i 0.7) at 24 h. Chrysopa spp. adults, Orius insidiosus adults and nymphs each consumed fewer A. glycines.

Foliar—foraging predators also consumed a signiﬁcant (P 3005) number of A. glycines nymphs in all but one case when compared to survival in the control (Table 3.1).

The only exception was Orius insidiosus adults, which left 6.6 i 0.9 nymph aphids in the dish, and was not significantly different from the control. While Orius insidiosus nymphs and Chrysopa spp. adults also left over five nymph A. glycines in the dish, though their feeding was significant in both cases. All other predators left no more than three remaining A. glycines in the dish at 24 hours.

Four of the species tested had a preference for nymph over adult aphids.

Coccinella septempunctata larvae, and H. axyridis adults and nymphs had a slight preference, while C. maculata had a strong preference, for nymphs. The remaining predators had mortality ratios less than one and preferred adults over nymphs, though several species (C. septempunctata adults, H. convergens, and Nabis spp.) were borderline and therefore it can be assumed that their preference was minimal.

Most ground-dwelling predators also always significantly reduced adult aphid survival in contrast to the control (Table 3.2). Poecilus chalcites, Poecilus lucumblandus, and Pterostichus melanarius were significant predators (P g 0.05). Forficula auricularia each consumed more than 50 percent of the A. glycines presented (P g 0.05). Two

72 predators, Clavina impressefrons and Philonthus thoracicus, did not show signiﬁcant

feeding on adult A. glycines.

Ground-dwelling predator feeding on nymphal A. glycines was also signiﬁcant in

most cases. Both P. chalcites and P. lucublandus left the fewest remaining nymphs, 1.5

i 0.5 and 2.2 i 0.5, respectively. Clavina impressefrons, Harpalus herbivagus, and

Philonthus throracicus did not show signiﬁcant feeding on nymph A. glycines.

The mortality ratio for ground-dwelling predators differed from that of foliar-

foraging predators. All predators with the exception of one, F. auricularia, had a

preference for nymphs over adults. This predator had a borderline mortality ratio of 0.9.

The mortality ratio was greatest, indicating stronger preference, for smaller predators,

including Elaphropus anceps, Harpalus herbigus, and Philonthus thoracicus, despite the

fact that feeding was not signiﬁcant for the latter two species.

Discussion

While all foliar species tested showed signiﬁcant feeding on adult A. glycines,

some species (C. septempunctata, H. axyridis, H. convergens, Chrysopa spp. larvae, and

Nabis spp.) appeared to feed more readily on A. glycines than other species (Chrysopa

spp. adults, C. maculata, and 0. insidiosus) tested. In many cases C. septempunctata and

H. axyridis began immediately feeding on A. glycines and consumed most of the A. glycines in the dish within the ﬁrst few hours of the study. In this case the apparent

preference for nymphs is artifact of the equation used to calculate nymphal mortality

slightly overestimates the number of nymphs available for predation. The predators

tested have a differing seasonal occurrence, with Orius insidiosus, Harmonia axyridis,

73 and Coccinella septempunctata appearing in early June, when aphid establishment occurs

(Table 3.3, Chapter 4). Although 0. insidiosus adults and nymphs left more adult and

nymph A. glycines than other predators tested, their abundance in the ﬁeld may make up

for their small size and relatively low feeding capacity. These predators appearing

around the time of A. glycines establishment likely will help to control it before colonies

become large. Those appearing later may help to reduce the rise of A. glycines colonies

that survive initial reduction from other predators.

As for ground-dwelling predators, most of which are nocturnal, immediate

feeding was rarely observed. Most predators hid under the soybean leaf in the dish when

they were introduced. However, F. auricularia began immediately consuming A.

glycines when it was introduced into dishes. During the course of the trial most predators

consumed signiﬁcant numbers of adult and nymph A. glycines. The reason for a

preference of nymphs in most cases cannot easily be explained. One would think that

voracious predators such as larger carabid beetles would prefer larger prey. The fact that

Elaphropus anceps had a preference for nymph A. glycines is not surprising, since this

predator is approximately the size of adult A. glycines. Many of these ground-dwelling

predators are present from June to August, and may encounter A. glycines during their

nocturnal foraging (Table 3.4), so their importance as aphid predators should not be

discounted. However, their importance as A. glycines predators needs to be assessed

further. Further studies are needed to determine if ground-dwelling predators actively

climb plants to feed on aphids, or consume aphids that may fall from plants when disturbed.

74 Any successful biological control program for A. glycines in part depends on existing predators present in soybean ﬁelds to both prevent aphid establishment and reduce populations of established colonies. As such, the results of this study contribute to implementation of a successful biological control program for the soybean aphid. These studies indicate that both foliar and ground-dwelling predators readily consume both adult and nymph A. glycines in no—choice tests. The role of these predators in future predation needs to be further assessed.

75 foliar- by

ratio

1.8 1.0 1.1

1.0 1.1

0.2 0.9 0.9 0.9 0.8

0.7 0.6 Mortality consumed l9

87 88 86 27 66 29 69 70 94 77 98 Percent 0.05 Mortality g trials P at 0.9NS i feeding

SEM)

18:05* 12i03*

17:02*

87:06

51:08*

06:02*

22:05* OJiOJ*

50i08*

07:05*

03:0J* 6.6 07:04* Mean

number

remaining (+ signiﬁcant *, no-choice h 6 24

Percent Mortality in signiﬁcant; not glycines

SEM)

Mean NS, 14:02* 14i02* 13:02* 15:02* 1J:03*

number 1Ji03* 57:04* 2Ji03* 22:04* 94i01 57106* 6A:0]* 62:03* A.

(+ remaining of

Wines—Narm—

1 Michigan 17 14 17 84 37 24 50 27 24 22 5 60 46 N dishes. mortality

control Lansing, adults adults adults

and East percent adults (L.) larvae Geer Guerin

and 2001, De adults predator in larvae (Pallas) (Say)

nymphs August larvae adults surviving

maculata 10 convergens septempunctata septempunctata nymphs

adults to axyridis axyridis spp. spp. surviving # spp. spp. insidiosus insidiosus

glycines

June 8 A.

of Species Chrysopa Coccinella Coccinella Coleomegilla Chrysopa Orius Orius Control Nabis

Harmonia Harmonia Hippodamia Nabis

from Predators contrasting Number

t-test predators 3.1. abidae Unpaired N

Anthocoridae Chrysopidae Coccinellidae

Family

foraging Foliar-Foraging Table

76 ground-

Ratio

1.7 1.3 1.6 1.3 1.2 1.0

2.2 2.3 2.7

0.9

Mortality by 0.05 consumed

32 64 60

53 78 44 46

42 43

Percent

Mortality g P trials at feeding SEM)

15:05*

87:06

27:1J* 37:05* 22:05*

53:10NS 06:06* 44:17NS

43:06* 46:16NS 43:05* Mean number remaining (+ signiﬁcant *, no-choice h 24 in signiﬁcant;

19 17

33 25 24 50 90 6

Percent

Mortality not

- NS,

1.3*

1.1NS

0.1 glycines

04* 04* 0.5NS

07* 03* 08* 05*

03*

i i i -_I-_ i i

i :t i A.

SEM)

Mean of

1.0

8.3

3.8

8.2 7.4

5.2 7.6 7.7 9.4

6.7

4.9

number Michigan

(+ AdultAlecineLanhids._Nxmnh_—

dishes. remaining 1 12 8 2 5 5 5 6 6 4 4 control mortality Lansing, and East

percent

)

Say (F. 2001, and

predator

(111.) Conte)

LeC

(Grav.)

Linnaeus

(Say)

Say (Le

August (Say) surviving

10 santaecrusis

surviving to

melanarius anceps

thoracicus

quadrimaculatum

auricularia

herbivigus

lucublandus

chalcites impressefrons June glycines 8

A. Species of Predators from

contrasting

Forﬁcula

Clavina

Philonthus

Pterostichus

Control

Poecilus Poecilus

Elaphropus

Harpalus

Bembidion Anisodactylus t-test Number predators 3.2. Unpaired

Foriculidae

Staphylinidae Carabidae Table Ground-Dwelling Family dwelling

77 September Michigan Almust Lansing,

East

Julv

2001,

during June

predators

adults

larvae

(L.) Geer

foliar-foraging

Guerin

De adults

larvae

(Pallas)

(Say) nymphs

larvae

adults

maculata

convergens

septempunctata septempunctata

nymphs adults

occurrence

axyridis axyridis

spp.

spp. spp.

insidiosus insidiosus

seasonal

Species

Coccinella Coccinella Coleomegilla

Chrysopa Chrysopa

Orius Orius

Harmonia Harmonia

Hippodamia Nabis Nabis Estimated Predators

3.3.

Nabidae

Chrysopidae Anthocoridae Coccinellidae Foliar Table Family

September Michigan

August

Lansing, East July

_ 2001,

during

June predators ) Say

(F. ground-dwelling (111.) Conte) LeC

(Grav.) of (Say) Linnaeus Say (Le (Say) santaecrusis melanarius anceps thoracicus

quadrimaculatum occurrence auricularia herbivigus lucublandus chalcites impressefrons

Predators seasonal Species Forﬂcula Clavina Pterostichus Philonthus Elaphropus Poecilus Poecilus Bembidion Harpalus Anisodactylus

Foliar

Estimated 3.4. Staphylinidae Foriculidae Carabidae

Table

Family

79 Literature Cited

Bilde, T., and S. Toft. 1997. Limited predation capacity by generalist arthropod predators on the cereal aphid, Rhopalosiphum padi. Entomological Research in Organic Agriculture. 15: 143-150.

Chang, GO, and P. Kareiva. 1999. The case for indigenous generalists in biological control, pp. 103-115. In B. A. Hawkins and H. V. Cornell, ed. Theoretical approaches to biological control. Cambridge University Press, UK.

Han, X. 1997. Population dynamics of soybean aphid Aphis glycines and its natural enemies in ﬁelds. Hubei Agricultural Sciences. 2: 22-24.

Hance, T. 1987. Predation impact of carabids at different population densities on Aphis fabae development in sugar beet. Pedobiologia. 30: 251-262.

Landis, D.A., and W. Van der Werf. 1997. Early-season aphid predation impacts establishment and spread of sugar beet yellows virus in the Netherlands. Entomophaga. 42: 499-516.

Lenné, J .M., and P. Trutmann. 1994. Diseases of tropical pasture plants. CAB International, Wallingford, UK. P. 69.

Li, W.M., and Z.Q. Pu. 1991. Population dynamics of aphids and epidemics of soybean mosaic virus in summer sown soybean ﬁelds. Acta Phytophylactica Sinica. 18(3): 123- 126.

Obrycki, J.J., and T.J. Kring. 1998. Predaceous coccinellidae in biological control. Annual Review of Entomolgy. 43: 295-321.

SAS Institute. 2000. SAS/STAT user’s guide, release 8.1 led. SAS Institute, Cary, NC.

Sebolt, D. C. 2000. Predator effects on Galerucella calmariensis L. (Coleoptera: Chrysomelidae), classical biological control agent of Lythrum salicaria L. (Myrtales: Lythraceae). MS. Thesis, Michigan State University.

Snyder, W.E., and DH. Wise. 1999. Predation interference and the establishment of generalist predator populations for biocontrol. Biological Control. 15: 283-292.

Syndmondson, W.O.C., K.D. Sunderland, and M.H. Greenstone. 2002. Can generalist predators be effective biocontrol agents? Annual Review of Entomology. 47: 561-594.

Schmaedick, M. A., and A. M. Shelton. 2000. Arthropod predators in cabbage (Cruciferae) and their potential as naturally occurring biological control agents for Pieris rapae (Lepidopera: Pieridae). The Canadian Entomologist. 132: 655-675.

Stary, P. 1995. Natural enemy spectrum of Aphis spiraephaga (Horn: Aphididae), an exotic immigrant aphid in central Europe. Entomophaga. 40(1): 29-34.

University of Minnesota Extension Service Web Site: http://wwwsoybeansumn.edu/crop prod/insects/@hid/aphidhtm

Van den Berg, H., D. Ankasah, A. Muhammad, R. Rusli, H.A. Widayanto, H.B. Wirasto, and I. Yully. 1997. Evaluating the role of predation in population ﬂuctuations of the soybean aphid Aphis glycines in farmers’ ﬁelds in Indonesia. Journal of Applied Ecology. 34: 971-984.

Wang, S.Y., X.Z. Boa, Y.J. Sun, R.L. Chen, and B.P. Zhai. 1996. Study on the effects of the population dynamics of soybean aphid (Aphis glycines) on both growth and yield of soyabean. Soybean Science. 15(3): 243-247.

81 CHAPTER 4

Predator effects on soybean aphid, Aphis glycines Matsumura, population growth

Introduction

The soybean aphid, Aphis glycines Matsumura, is an exotic pest from Asia that was discovered in the United States in 2000. In Asia A. glycines feeding on soybean can cause up to a 20 cm reduction in growth and a 27.8 percent reduction in seed yield (Wang et al. 1996). In 2000, A. glycines caused up to a 13 percent yield reduction occurred in replicated ﬁeld plots in Wisconsin during the year 2000 (University of Minnesota

Extension Service Web Site). In 2001, it was found that A. glycines caused up to a 40 percent yield loss in Michigan (Difonzo, Pers. Comm). Aphis glycines pest can also indirectly harm soybeans by vectoring persistent viruses, including soybean dwarf virus and non-persistent viruses, such as soybean stunt virus, soybean mosaic virus, and bean yellow mosaic virus (Van den Berg et al. 1997). Epidemics of soybean mosaic potyvirus in summer-sown soybean ﬁelds in Jiangsu, China were found to be closely associated to the timing of A. glycines immigration (Li and Pu 1991). In addition, aphids also cause indirect damage by excreting honeydew onto foliage, which promotes the growth of sooty molds that reduce the photosynthetic capacity of the leaves (Lenné and Trutmann

1994, Hirano et al. 1996).

Foliar-foraging predators contribute to the reduction of a wide variety of aphid species (Grasswitz and Burts 1995, Star)’I 1995, Van den Berg et al. 1997, Obrycki and

Kring 1998). For example, it was found that a complex of natural enemies, including

82 Chrysopa nigricornis Burmeister, Orius spp., Coccinella transversoguttata Faldermann,

Hippodamia convergens Guerin, and several species of Syrphidae and Chamaemyiidae all contributed to the reduction of Aphis pomi De Geer in a ﬁeld study using apple trees

(Grasswitz and Burts 1995). Similarly, Stary (1995) found that Coccinella septempunctata L., Adalia bipunctata (L.), and Episyrphus balteatus (De Geer) were common predators that all consumed and thus aided in control of the spiraea aphid, Aphis spiraephaga Miiller in Czechoslovakia. Chen and Hopper (1997) found in Europe that

Russian wheat aphid, Diuaphis noxia (Mordvilko), populations were reduced at critical times by several species of coccinellid and syrphid predators.

In Asia, coccinellid predators (Harmonia spp.) play an important role in suppressing A. glycines populations in soybean ﬁelds (Van den Berg et al. 1997). Other aphidophagous predators, such as Nabis spp., Harmonia axyridis (Pallas), Coccinella septempunctata L., and Chrysopa spp. aided in its control during the mid to late season in

China (Han 1997). Schneider (1971) reported the predaceous syrphid larva Ischiodon escutellaris (F .) was another aphidophagous predator that fed on a wide variety of aphids, including A. glycines in the Philippines. However, low syrphid abundance in comparison to the number of aphids has not resulted in suppression of aphid populations by syrphids alone, although they do contribute aphid reduction (Alﬁler and Calilung 1978, Van den

Berg et al. 1997).

Folia-foraging predators clearly depend on aphids as a food source and reduce aphid populations in tropical areas, but the effects were not as pronounced in temperate climates. At lower temperatures it was found that the reproductive rate of A. glycines exceeded the predation rate of the coccinellid predators (Van den Berg et al. 1997). This

83 allowed A. glycines populations to rapidly develop in temperate regions, particularly

during the early portion of the growing season (Van den Berg et al. 1997). A similar

result has been documented for Coccinnella spp. attacking pea aphids, Acyrthosiphon pisum (Harris), in British Columbia (Frazer and Gilbert 1976).

In the United States, generalist predators may be the most common natural enemies of A. glycines. Much consideration has been given to the importance that generalist predators may play in biological control (Hance 1987, Hilbeck et al. 1997,

Landis and Van der Werf 1997, Snyder and Wise 1999, Chang and Kareiva 1999,

Symondson et al. 2002). For example, it has been shown that ground beetles (Carabidae) species are capable of reducing populations of Aphisfabae Scopoli, a pest of sugar beet

(Hance 1987). It has also been demonstrated that many species of carabid beetles in sugar beet ﬁelds in the Netherlands, along with spiders and a cantharid, Cantharis lateralis (Coleoptera: Cantharidae) aided in A. fabae aphid control early in the season

(Landis and Van der Werf 1997). Further, it has been shown that a complex of ground- dwelling predators, such as carabids, staphylinids, and spiders, all contributed to the reduction of bird cherry-oat aphid, Rhopalosiphum padi (L.), populations (Ostman et al.

2001). While generalist predators are not as effective per capita as specialized predators, they can often compensate by being present earlier in the season (Chang and Kareiva

1999), when pest densities are low and specialist predators scarce (Takagi 1999).

However, the role of predation in control of A. glycines in the United States is not currently known.

The objectives of these studies were to: 1) determine the abundance and species

COmposition of potential A. glycines predators in Michigan soybean ﬁelds following

84 aphid establishment (2001) and prior to A. glycines establishment (2002), and 2) study the impacts of predators on A. glycines with and without predator exclusion.

Materials and Methods

Field Site.

Experiments took place at the Michigan State University Entomology Farm,

Ingham County, Michigan. During 2001, a 134 m x 99 m ﬁeld was planted to soybean variety Mycogen (5251RR) in 38 cm rows. All plots were located at least 9 m from any

ﬁeld borders to minimize edge effects. The crop area was managed using reduced primary tillage (chisel plow, disc) followed by secondary tillage (ﬁeld cultivation) after metolachlor (Dual 11) was applied at a rate of 2 1/ha to control weeds. Potash was applied at a rate of 168 kg/ha to meet soil test requirements. Soybeans were planted on 5 May

2001 at a rate of 70,823 seeds/ha.

During 2002, a 99 m x 94 m field was planted to soybean Pioneer (93882) in 38 cm rows. Experiments were conducted in this larger field with no plot closer than 9 m from any edge to minimize edge effect. The field was planted on 20 May 2002 at a rate of 70,900 seeds/ ha. The crop area was again managed using reduced primary tillage

(chisel plow, disc) followed by secondary tillage (ﬁeld cultivation). A tank mix of the herbicides, of lactofen (Cobra) (1 l/ha), bentazon (Basagran) (2 l/ha), sethoxydim + dash

(Poast Plus) (2 l/ha), and crop oil concentrate (2 l/ha) was applied on 5 June 2002.

2001 Predator Exclusion Trial.

In 2001, A. glycines population growth was contrasted in treatments consisting of cages designed to exclude all predators versus Open (sham) cages that allowed predator

85 access (Figure 4.1). The experiment was laid out as a completely randomized design with ﬁve replications and one cage of each cage type placed per 9 m x 9 m area. Cage frames were constructed of 1.3 cm outside diameter Cresline PVC pipe. Frames for exclusion cages were 1 m2 (top view) with legs 0.90 m long, with 20 cm to be placed in the soil and 0.70 m to extend above the soil line (Figure 4.1). The top of the frame was covered with 1 m2 white no-see-um fine mesh netting that draped 27 cm down the side of cages. The bottom of the frame was surrounded by a thin 3 mil plastic sheet that extended 10 cm below the ground and upward 27 cm to meet the bottom of the screen.

Where they met, the screen and the plastic were joined with a 2 cm strip of Velcro that allowed cages to be opened for sampling. Frames for open cages were also 1 m2 (top view) but legs were 1 m long, with 0.10 m below the soil surface. This allowed for a 18 cm gap between the plastic and the soil line to allowed predators to enter and a 10 cm gap between the plastic and the screen for foliar-foraging predators to enter. Both exclusion and open cages contained the same amount of screen and plastic materials above the soil level to control for potential cage effects. The exclusion cage was designed to exclude all predators and was completely sealed. Two 8.5 cm by 13 cm pitfall traps with rims at the soil level were placed in exclusion cages to remove ground-dwelling predators and two standard 14 cm x 23 cm yellow sticky traps were placed in each cage to remove foliar- foraging predators. Twenty sticky traps and twenty pitfall traps (one in each treatment plot) were placed in the ﬁeld to allow comparisons to be made between predator composition and abundance in exclusion treatments versus those in the entire plot area.

Pitfall and sticky traps were changed weekly.

86 Sampling was initiated on 26, June 2001, after natural A. glycines infestation took place, and was completed on 31 July 2001. Sampling was continued in the open cage treatment until 4 September 2001. At weekly intervals, the temperature inside and outside of cages were determined using a mercury thermometer held inside cages for 1 rrrinute and then outside cages for 1 nrinute. A hygrometer was similarly used to collect relative humidity data. Foliar observations in the form of a 3-minute non-intrusive visual examination of soybeans in each cage was followed by a hand examination of foliage were conducted to determine predator abundance and species composition. After this,

ﬁve soybean tips per cage (consisting of the uppermost node) were collected to assess A. glycines populations. Tips were placed in plastic bags and returned to the laboratory for counting. Then plant height was deterrrrined by measuring from the soil to the top of the canOpy for ﬁve random plants in each cage.

2002 Predator Exclusion Trial.

In 2002, three cage treatments were used, open and exclusion, as described above and a frame alone treatment, to control for the possible effects that plastic and screen netting may have on conditions in cages. This treatment had the same dimensions as the open treatment but lacked plastic and screen netting. The plot layout was a completely randomized block design, with one cage per 6 m x 6 m area. There were five replications with three cages in each replication, for a total of 15 cages. Two trials of this experiment were conducted during 2002, and each trial was split into two parts (before and after cage switch), with five sample dates for each part. The first trial was initiated on 26 June, prior to natural A. glycines infestation in this field, when an average of 110 _+_ 10 adult A. glycines were introduced to random plants in each cage by transferring them from

87 infested soybean sprigs to plants within cages using a thin camel hair brush. Large A. glycines with a visible cauda were considered adults. The ﬁrst trial was sampled from 28

June t012 July. On 12 July, four replicates were randomly selected then open and exclusion cages were switched in these replications. The frame cages were never moved.

The trial was then conducted for another ﬁve sample dates from 15 July to 29 July. The unchanged open and exclusion cages served as a control to determine A. glycines population trajectories. These cages were sampled from 15 July to 29 July. A second experiment was initiated on 10 July, when we artiﬁcially infested plants in each cage with 131 i 11 adult A. glycines using the same methodology described above. This second trial of this experiment ran from 12 July to 26 July. At this time, open and exclusion cages were once again switched. Data was collected again from 29 July to 12

August. To accommodate increased soybean height, when cages were switched, exclusion cages were raised approximately 10 cm. This left enough plastic to provide a barrier to prevent entry by predators.

In both trials, data was collected every Monday and Friday. Temperature and relative humidity data was collected as in 2001 and three-minute visual examinations were taken using the same methodology described above. Then 10 randomly selected whole plants were counted inside each replicate in the ﬁeld to assess the adult and nymph

A. glycines population. Five plants were selected and measured in each cage and their stage recorded on each sample date. Exclusion cages each had two 8.5 cm by 13 cm pitfall traps with rims at the substrate level. These traps were placed in opposite comers of the cage to reduce populations of ground-dwelling predators. Therefore, there were ten total pitfall traps in the exclusion cages in each plot. There were also ten pitfall traps

88 placed randomly within each 6 m x 6 m plot to compare the abundance of predators in cages and in surrounding crop. Pitfall traps were ﬁlled with 50 percent ethylene glycol and samples were collected every seven days.

Analysis.

Data from 2001 was considered prelirrrinary and not statistically analyzed. In

2002 a type III F -test for overall treatment effect determined the statistical significance of treatment effects on temperature, humidity, and plant variables in open, exclusion, and frame cages by an analysis of variance (ANOVA). Adult and nymph A. glycines, predator, and pitfall trap counts were analyzed with a type III F -test for overall treatment effect with the Poisson regression for counts, using the GLIMMIX Macro link of SAS statistical program (SAS Institute 2000). From this analysis, treatment, date, and treatment*date interaction will be reported. A significant treatment effect indicates that means between treatments differed (P g 0.05), and a significant date effect indicates that means were higher or lower than other dates. A significant treatment*date interaction occurs when the effect of treatment on the date that it is being observed and vice-versa, i.e. the effect of date will be different for each treatment. When this occurred, data was sliced to reveal differences on particular dates. When treatment and date were both significant, individual pair wise comparisons from LS MEANS from SAS output were used to determine significance between individual treatments on that date. When there was no interaction, treatment was reported as an overall treatment effect, because it was then significant regardless of the date.

Results

89 2001 Predator Exclusion Trial.

Neither cage type appeared to inﬂuence temperature as the mean difference between interior and exterior temperatures varied by -2 i 07° C in exclusion cages and -

0.6 i 04° C in open cages (Figure 4.2). Relative humidity also varied little with the exception of 17 June, after a rainfall, where the humidity inside the exclusion cage was approximately nine percent higher than the outside humidity (Figure 4.3). Plant growth did not vary in any cage due to A. glycines or cages effects, as plant heights were similar in all treatments on all sample dates (Figure 4.4).

Prior to the implementation of cage treatments in 2001, A. glycines and predators were well established in plots. While removal of all predators from the exclusion cages was attempted, this proved to be impossible (Figure 4.5). Further, the exclusion cages were unsuccessful in preventing re-entry of some predator species. In particular,

Harmonia axyridis (Pallas) larvae and Orius insidiosus (Say) nymphs and adults were found in nearly equal numbers in both open and exclusion cages, and made up a large portion of the predators observed (Table 4.1, Figure 4.6, Figure 4.7). It was evident that coccinellid egg masses were missed in the visual searches as concentrations of young larvae were frequently found in subsequent searches in the exclusion treatment.

Had exclusion cages effectively reduced predator numbers, the expectation was that A. glycines numbers inside exclusion cages would rise in relation to those in open cages. In contrast, A. glycines adult and nymph populations in both open and exclusion plots fell in the ﬁrst two weeks of the experiment, and remained similar in open and exclusion treatments until mid-late July (Figure 4.8, Figure 4.9). The reduction in A. glycines adult and nymph numbers occurred at the time H. axyridis populations reached

90 their peak abundance on 3 July (Figure 4.6). In general, H. axyridis populations in both open and exclusion cages appeared to track populations of A. glycines with peaks in H. axyridis following aphid peaks by approximately one week (Figure 4.6, Figure 4.8,

Figure 4.9). From 24-31 July adult A. glycines numbers declined in the exclusion plots but rose dramatically in the open plots. A second peak in H. axyridis numbers subsequently followed this.

In contrast, populations of 0. insidiosus were low in late June and early July, peaking on July 17‘h (Figure 4.7). There were no differences in the number of 0. insidiosus inside the exclusion or open cages. Orius insidiosus was largely absent for the

ﬁrst peak in A. glycines and reached it peak densities in plots shortly after a low point in

A. glycines populations. Orius insidiosus was unable to prevent resurgence in A. glycines populations in the open plots but the higher numbers of 0. insidiosus in exclusion plots are associated with a decline in adult A. glycines numbers (Figure 4.8) but not nymphs

Figure 4.9).

The placement of sticky traps in the exclusion cages did not greatly influence predator numbers. While large numbers of H. axyridis were present in the field, few were caught on sticky traps in exclusion cages or field areas (Figure 4.10). Orius insidiosus populations also did not appear to have been reduced by the traps in cages, as numbers remained steady over the trial (Figure 4.11). The mean number of ground- dwelling predators generally decreased over time in exclusion plots, while mean number of ground predators in the entire plot increased slightly (Figure 4.12). Carabid abundance and species richness was lower in exclusion cages versus the surrounding study area, indicting that exclusion cages were effective at reducing these predators (Table 4.2).

91 2002 Predator Exclusion Trials.

Temperatures inside and outside of cage treatments differed very little throughout

both trials in 2002 (Figure 4.13). Statistically signiﬁcant effects, were few, and when

they did occur, they were caused by the exclusion cages differing from the open and

frame cages (Table 4.3, Table 4.4). This is likely a result of the fact that temperature in

exclusion cages was typically slightly higher, because the screen and plastic caused

ventilation to be minimal.

With one exception, relative humidity also did not vary greatly between

treatments in trial one, with the humidity difference typically slightly higher in exclusion

cages (Figure 4.14). Signiﬁcant differences were only found after the switch when open

and exclusion cages had higher relative humidity than frame cages (Table 4.5). In trial

two, the humidity difference appears to be greatest in exclusion cages (Figure 4.14).

Exclusion cages generally had higher relative humidity than open or frame cages during

both parts of the trial (Table 4.6).

Plant growth was not signiﬁcantly affected in either trial, and neither cage type

nor A. glycines caused plant heights to differ (Table 4.7, Table 4.8, Figure 4.15). During

trial one, plants were in V5 stage on the ﬁrst sample date (28 June), and R1 on the last

sample date before cages were switched (12 July). After cages were switched, plants

were in R1 stage on the first sample (15 July) date and R2 on the last 29 July. During trial two, plants were in R1 stage on the first sample date (15 June), and R2 on the last sample date before cages were switched (26 July). After cages were switched, plants were in R2 stage on the first sample date (29 July) and R4 on the last sample date (9

August).

92 Abundance and species composition of ground-dwelling predators was consistently low in both trials and therefore their potential A. glycines feeding was likely minimal (Table 4.9, Figure 4.16). In both trials the species richness was lower in exclusion cages than in the ﬁeld as a whole. In trial one, captures were signiﬁcantly higher on all sample dates in the crop area (Table 4.10). In trial two, captures were low

(Table 4.9, Figure 4.16) and not signiﬁcantly different in cages and the ﬁeld (Table 4.11), because abundance was similar in exclusion cages and in the crop area (Table 4.9, Figure

4.11). Even though there was not statistical signiﬁcance, the captures were extremely low in this part of the experiment (Table 4.11).

Aphid and Predator Response Trial One.

After cage establishment, adult A. glycines populations in trial one increased steadily, reaching approximately 100 aphids per ten plants (Figure 4.17). In contrast, in open and frame cages populations remained below 10 aphids per ten plants in both cages.

Similar trends occurred for nymphal A. glycines (Figure 4.17). Statistically, exclusion cages contained greater numbers of adult and nymphal A. glycines than open and frame cages on most dates prior to cages being switched (Table 4.12, Table 4.13). Prior to cages being switched, the abundance of predators was lowest in the exclusion cages, although 0. insidiosus adults and Chrysoperla spp. were found at a very low density

(Table 4.14, Figure 4.19). Orius insidiosus adults were abundant and made up 61.7 and

55.2 percent of the observed predators in open and frame cages, respectively (Table

4.14). More predators were observed in the open and frame cages, where the species composition was greater and consisted of a combination of foliar predators, including H. axyridis at a low density (Figure 4.20).

93 The un-switched replicates demonstrated what occurs with A. glycines adult and

nymph populations when cage treatments were left in place (Figure 4.17). The A.

glycines population increased dramatically in exclusion cages, peaking at 1,534 adult and

2,492 nymph A. glycines per ten plants on 26 July. When cages were switched, high

populations of A. glycines were exposed to predator colonization and the low populations

in former open treatments were protected from predation. In the newly opened cages, the

adult and nymphal A. glycines populations remained constant for approximately one

week and then decreased from 103 adult A. glycines per ten plants on 15 July, to 11 per

ten plants on 26 July, and from 159 A. glycines nymph per ten plants to 40 per ten plants

on the same dates (Figure 4.17). For both adult and nymph A. glycines, there were

statistically more A. glycines in exclusion versus open or frame treatments on most days

prior to the switch (Table 4.13, Figure 4.17). This trend was signiﬁcantly reversed by the

end of the test on 26 July.

When predators were excluded from formerly open cages, there was an almost

immediate increase in both adult and nymph A. glycines populations, indicating that

predators had been playing a role in keeping these A. glycines populations low (Figure

4.17). There were signiﬁcant differences in A. glycines adults and nymphs when comparing the now open versus exclusion treatments. As the A. glycines population decreased in the open treatments, it increased in the exclusion treatments (Table 4.12,

Table 4.13).

On 19 July, there was a similar predator community in most treatments, but an enormous increase in predator abundance in the former exclusion, now newly opened treatments, where predator abundance jumped from below 20 predators per 3 min. to over

94 60 predators per 3 min. (Figure 4.19). The most abundant predators in this part of the trial were H. axyridis adults, Leucopis spp. midge larvae, and 0. insidiosus (Say) adults, which made up 22.9, 30.6, and 14 percent of the total predator counts (Table 4.15). For

H. axyridis and Leucopis midges, there was a striking increase on 19 July, when the initial decrease of A. glycines began (Figure 4.20, Figure 4.21). The population of H. axyridis remained high as the A. glycines population decreased, but the Leucopis spp. larvae increase was ephemeral. The 0. insidiosus population also increased on this date, but appears to have also increased equally in exclusion and frame cages (Figure 4.22).

There were signiﬁcantly more predators in the now open treatment versus the other treatments on 19 July (Table 4.16).

Aphid and Predator Response Trial Two.

In trial two, results were qualitatively similar to trial one for the ﬁrst half of the trial. The statistically signiﬁcant differences were detected on many dates (Table 4.17,

Table 4.18). The main difference was that adult and nymph A. glycines populations were uniformly lower. Exclusion cages had higher adult and nymphal populations than open or frame cages on most dates. Adult A. glycines population peaked at about 30 A. glycines per ten plants, and the nymph populations peaked at about 70 A. glycines per ten plants. Predator counts appeared to be slightly higher in open cages than in the other treatments (Figure 4.19). During this part of the trial, the predator community was similar to that of trial one (Table 4.19). The most abundant predator in all cages were 0. insidiosus adults, which made up 69.9, 58.5, and 61.5 percent of predators observed in exclusion, open, and frame cages. There were few statistical differences between treatments, and the only difference occurred on 19 July (Table 4.21). While the

95 exclusion cages prevented entry by most predators, 0. insidiosus was more difﬁcult to

remove in this experiment because as the plant canopy became more dense, visibility and

efﬁciency of removing predators decreased. The number of 0. insidiosus was similar in

all cages (Table 4.19, Figure 4.22). Harmonia axyridis was the second most abundant in

open and frame cages, though its population remained low (Table 4.21, Figure 4.20)

compared to trial one (Figure 4.20).

In contrast to trial one, aphid numbers did not decrease in the now open cages in

the second half of the trial (Fig. 4.18). In addition, the un—manipulated control cage, the adult and nymph A. glycines populations never increased beyond those in the new exclusion cage. The reason was not explained by difﬁculty excluding predators, since

very few predators were encountered in this cage (Table 4.20). Adult and nymph A. glycines populations continued to increase in open, frame and exclusion cages, with the population open cages being higher until the last sample date, when open and exclusion cages became similar in all cages (Figure 4.18). In both open and exclusion cages, open cages differed from the other treatments on most dates (Table 4.17, Table 4.18). Again, predator abundance increased in open cages, but the increase was not as obvious as in experiment one (Figure 4.19). The same three species were again the most abundant in open cages (H. axyridis, Leucopis midge larvae, and 0. insidiosus), and again these predators became most abundant approximately one sample date after cages were switched (Table 4.19, Figure 4.20, Figure 4.21, Figure 4.22). This time, though,

Leucopis nridges were not prevalent, and 0. insidiosus was more abundant in open cages.

These predator differences between exclusion and the other cages were signiﬁcant on

96 most dates (Table 4.21). Even with signiﬁcantly more predators in the now open cages, adult and nymph A. glycines numbers did not decrease rapidly.

Discussion

Exclusion treatments were not effective at reducing the abundance of foliar- foraging predators in 2001. Even though any predators seen in the exclusion cages were killed or physically removed each time they were found during foliar observations this apparently had only a temporary effect on predator abundance. Implementing this study after A. glycines colonization and predator appearance made predator removal very difﬁcult. From the observations of 1st instar coccinellid larvae on plants in exclusion cages, it was clear that coccinellid egg masses were missed. In addition, 0. insidiosus nymphs were essentially impossible to remove, as they tended to colonize the folded tips of newly emerging soybean leaves. These two predators made up the majority of those found in exclusion cages. More difﬁcult to explain is the appearance of adult coccinellids and various other predators in exclusion cages. Even though every attempt was made to seal the cage after sampling, it is likely that some found crevices to enter the cages at the seams where the Velcro joined the plastic and screen parts of the cage. In spite of the failure to create treatment differences in predator numbers, this test did reveal the potential importance of H. axyridis and 0. insidiosus as predators of A. glycines. The likely cause for A. glycines reduction from 26 June to 10 July was predation, since 0. insidiosus and H. axyridis clearly responded to A. glycines populations. However, the failure of cages to create differences in predator populations limits the understanding of their role in A. glycines population dynamics. In 2001, temperature, relative humidity,

97 and plant height did not differ greatly among cage treatments, and were not likely

associated with the observed trends in A. glycines populations.

As in 2001, in 2002 temperature, relative humidity, and plant height did not differ

greatly among treatments. While average temperature differentials were slightly higher

inside versus outside of exclusion cages in contrast to open or frame treatments, it is

unlikely that this contributed to increased A. glycines populations inside the cages. The

optimum temperature for A. glycines development is reported to be between 22 and 25° C

in late June to early July (Wang et al. 1962). During both of our trials, the average

temperature inside exclusion cages was above 30° C by which exceeds the optimum

temperature for aphid development. Thus, it might be expected that A. glycines population growth may have been slightly slowed in the exclusion cages in relation to open or frame treatments.

In addition to excluding predators, exclusion cage treatments could also unnaturally conﬁne A. glycines on the plants. It is known that production of alate aphids increases under crowding and plant stress conditions (Dixon 1985). Thus, emigration may be an alternative explanation for a decrease in aphid abundance after the cages are switched. Several lines of evidence argue against this. First, alate A. glycines were rarely observed in exclusion cages during aphid counts and predator observations. No alates were observed in trial one until 10 July, at this point a total of three alates were found.

Alates only became abundant in the exclusion control on 22 July, when adult aphid populations were averaging over 1,000 per plant. In trial two there was only one observation of alates in exclusion cages on 26 July. Coupled with observations of a large

98 increase in predator abundance following cage switch emigration is unlikely to be the sole explanation for A. glycines reductions in the former exclusion treatments.

In contrast to 2001, in 2002, most predators were successfully reduced in exclusion cages and tests provided strong evidence that predators were effective at keeping A. glycines densities low. Actively foraging predators, particularly C. septempunctata, H. axyridis, and 0. insidiosus, continually removed A. glycines in open and frame treatments keeping populations low in comparison to exclusion treatments.

Having these treatments side by side eliminates the possibility that environmental conditions were the cause of limited A. glycines populations. In contrast, where predators had free access to plants they prevented an aphid outbreak while adjacent plots with predator exclusions reached thousands of aphids per ten plants. Collectively, predators showed the ability to keep initial A. glycines populations low (both trials) and quickly decrease high A. glycines populations when cages were switched (trial one only). This study also bolsters the ﬁndings of Van den Berg et al. (1997), where Harmonia spp. contributed to the reduction of A. glycines populations.

Overall, the ﬁndings also agree with Grasswitz and Burtz (1995) and Chen and

Hopper (1997), who found that aphid densities were controlled by a community of foliar- foraging predators. During these trials, H. axyridis and 0. insidiosus were consistently the most abundant predators. Leucopis midge larvae were only abundant during one sample date. It is possible that they have an ephemeral period of abundance, or that they provided a secondary food source for H. axyridis and 0. insidiosus. It is unlikely that 0. insidiosus alone can successfully suppress A. glycines populations, since pOpulations were similar in all cage treatments after cages were switched in trial one. Despite this

99 similarity in 0. insidiosus populations, there was an increase in exclusion cage aphids within one week. However, it is possible that 0. insidiosus and H. axyridis may interact in a positive way to increase A. glycines control.

The overall increase in A. glycines populations in the absence of predators and the reduction of A. glycines populations when predators were allowed access was greater in the ﬁrst trial when plants were smaller and at an earlier phenological stage. Van den

Berg et al. (1997) showed that A. glycines intrinsic rate of increase decreased as soybeans age. It may be that in the ﬁrst trial of 2002, A. glycines populations were able to respond to the lack of predators due to more suitable host plant quality. In turn, predators may have had an easier time foraging for A. glycines on these smaller plants with less foliage.

When plants are smaller and the canopy less dense, prey have fewer places to hide and predators have less foliage to search. The opposite may be true when plants mature, as in trial two, where plants were both taller and had thicker foliage. A study by Garcia and

O’Neil (2000) bolsters this argument, because they found that predation by the coccinellid C ryptomaemus montrouzieri Mulsant on the citrus mealybug, Planococcus citri Risso, decreased as Coleus plants increased in size. They suggested that plant characteristics were the most likely reason for this decrease in predator efﬁciency found in their study.

These studies indicate an important role for existing predators in A. glycines control in Michigan. With apparently strong A. glycines suppression by existing predators already in place, introduction of additional biological control agents should be done with caution. It is likely that existing predator communities would interact with introduced parasitoids (Colfer and Rosenheim 2001) and may slow or prevent their

100

establishment. In contrast, combinations of natural enemies may be found that provide more complete of more reliable aphid suppression. Consideration of these potential interactions prior to the release of additional natural enemies is recommended.

101 in were cages Observations open % 1.0 1.2 1.5 1.7 1.2 1.5 8.3 3.4 2.2 0.2 0.2 4.9 15.2 34.2 23.2 and 100.0% intrusive July O “- V) 40 and N 34 I‘O‘F‘m—‘OWO‘I—‘WDV' 62 Open 409 TotaL 31 to June % 26 -

a- -

1.3 2.0 7.6 6.0 0.3 0.7 0.3 non-intrusive 13.4 17.9 50.5 100.0% from 1 152 18 1 301 - 23 VN - 54 6 Exclusion 40 TmaL week 3-minute a adults once during adults adults adults larvae (L.) Geer Guerin sampled larvae De adults observed (Pallas) larvae were adults nymphs

(Say)

nymphs Michigan maculata maculata larvae spp. spp. convergens spp. predators cages septempunctata septempunctata adults nymphs axyridis axyridis spp. spp. spp.

insidiosus

insidiosus Lansing, percent Exclusion

and East Chrysoperla Chrysoperla Coccinella Coccinella Coleomegila Coleomegila Orius Syrphid Orius Hippodamia Harmonia Species Hemerobius Harmonia Nabis Nabis total,

2001. September,

during

4 to composition, cages

abidae June Anthocoridae N Hemerobiidae Chrysopidae

Coccinellidae 26 open Syrphidae Species Family

and from 4.]. Diptera: TOTALS Heteroptera: Neuroptera: Neuroptera: Coleoptera: Heteroptera: Order:

Table sampled exclusion

102 Table 4.2. Species composition, total, and percent ground dwelling predators captured per seven days in exclusion cages versus those in the surrounding ﬁeld during 2001.

Pitfalls were sampled once a week from 26 June to 31 July, 2001, East Lansing,

Michigan

E Order Coleoptera Exclusion Entire Plot Family Species Total % Total %

Carabidae Agonum cupripenne Say 2 6.3 2 0.3 Agonum placidum (Say) - - 136 22.4 Amara aenea (De G.) 2 6.3 35 5.8 Amara apricaria (Payk.) - - l 0.2 Amara rubrica Haldman - - 8 1.3 Anisodactylus rusticus (Say) - - l 0.2 Anisodactylus santaecrusis (F.) 9 28.1 96 15.8 Bembidion quadrimaculatum Say 1 3.1 30 4.9 Bembidion rapidum (Le C.) - - 2 0.3 Bradycellus rupestris (Say) - - 1 0.2 Chlaenius pusillus Say - - 3 0.5 Chlaenius tricolor Dej. - - 2 0.3 Clavina bipustulata (F.) l 3.1 4 0.7 Clavina impressefrons Le C. - - 15 2.5 Colliutis pensylvanicus - - l 0.2 Cyclotrachelus sodalis (Le C.) 1 3.1 11 1.8 Elaphropus anceps (Le C.) 4 12.5 63 10.4 Harpalus afﬁnis (Schrank) - - 3 0.5 Harpalus pensylvanicus (De G.) - - 5 0.8 Poecilus chalcites (Say) - - 23 3.8 Poecilus lucublandus (Say) - - 11 1.8 PIerostichus commutable (Motsch.) - - 1 0.2 Pterostichus melanarius (III.) 3 9.4 41 6.8 Scarites quadriceps Chd. 7 21.9 25 4.1 Stenolophus comma (F.) 2 6.2 81 13.2 Lampyridae Photinus spp. - - 6 1.0 TOTALS 33 100.0% 606 100.0%

103 °C

Indicates minus

which interior indicated.

July (°C 0.67 0.62 0.91 date 29 determine the to at

date difference cages July 0001* 0.09 0002* 26 < frame

excluding temperature

and of MEANS

open, results LS

July on 0.52 0.88 0.48 the 22 exclusion,

used based of we

2002 results

July P P MEANS date. one, 002* 0.22 0.35

0.35 0.29 0.99

001* 001* 002* 19 LS the signiﬁcant,

trial ANOVA

of F F 1.40 1.14 from 4.32

was

3.87

2.96 4.61

effect from during 0.44 0.006 0.04 56 56

56 if

39 39 July regardless cages 0.23 0.92 0.23 2, 2,

2, 4, treatment

15 -statistic Treatment

comparisons F 005* only frame S wise P Open Open

and greater at

When a Exclusion Open Exclusion Open

pair of versus versus the 005*. versus versus versus versus are exclusion, 5 signiﬁcant

P line at Switch Exclusion Frame open, Frame Frame Frame

Exclusion

Probability Switch were of

dotted 4.3.

Bgmm

After

Date Treatment Date

Effect Treatment*Date Treatment*Date Treatment signiﬁcance treatments Below

Table exterior)

104 * line. exterior) °C dotted the minus below interior (°C shown are difference indicated date temperature the at of cages results on frame based and P P 0.10 0.79 0.31 0.17 009* 0.06 OVA open, 2002 AN F F 1.87 1.53 1.22 3.79 2.91 0.23 two, from exclusion, trial 35.6 35.6 38.4 55 55 55 df of if 8, 8, 2, 2, 4, 4, -statistic during F MEANS cages 0.05 L8 greater g a P frame of from at and Switch Probability Switch comparisons signiﬁcance exclusion, 4.4. Before After Date Treatment Treatment*Date Effect Date Treatment*Date Treatment wise open, indicates Pair Table of

105 * which line. exterior) dotted minus determine the to July below 0.06 0008* 0.39 date (interior 12 shown are excluding difference MEANS indicated humidity LS date the the relative at of used

July

5 we cages results on frame signiﬁcant, and based P P

July 0002* 0.26

005* 0.33 0.73 004* was date.

1 open, 2002 the ANOVA

1.36

1.19

7.49

2.06

4.38 0.32 one, treatment from exclusion, trial

Eﬁ‘ect

37.5

38.9 37.5 only

60 60 60

June

41"

8, 2,

regardless

2, 4, 0.54 28 0009* 0.0005* -statistic during F When 0.05 S MEANS Treatment P cages 0.05. LS at greater g a P frame of from at and signiﬁcant

Switch

Frame-Open

Frame-Exclusion Frame-Open Frame-Exclusion Exclusion-Open Exclusion-Open

Probability Switch were comparisons signiﬁcance exclusion,

4.5. After

Date

Before Treatment

Treatment*Date Date Effect Treatment Treatment*Date wise open, treatments indicates Pair

of Table

106 * which

line. exterior)

dotted minus determine the to August 002* 0002* 0.18 below 12

date (interior shown are

excluding difference 0006* 0006* 0.98 August 9 MEANS

indicated humidity LS date the

the relative

at of used 0.39 0.76 0.60 August we 5

cages

results on frame signiﬁcant,

and

based P August 0.42

0.36

0007* 0.93 0004* 005* 0.30 0.30 0003* P

was date. 2 open,

2002

the

ANOVA

of F F 1.22

1.28

3.20

3.18 4.50 6.67 two,

treatment from

Eﬁ‘ect exclusion,

trial

36.7 36.7 38.2 56 56 56 July only

of df 003* 005* 0.97

0003* 0002*

0.84

8, 2,

4, 2, regardless 4, 4f

29 -statistic

during F

When 0.05

Treatment S

MEANS P cages 0.05.

at greater

_<_ a P

frame of from at

and signiﬁcant Switch Exclusion-Open Frame-Open Frame-Exclusion

Frame-Open

Exclusion-Open Frame-Exclusion

Probability

Switch were comparisons signiﬁcance

exclusion,

4.6. After

Before

Date

Date Treatment Treatment*Date Effect Treatment wise Treatment*Date

open, treatments Pair indicates

of Table

107 *

line. random

ﬁve dotted of

the

below measurement

shown on

are (based

indicated height

date plant

the

at of

cages results on

frame

0.33 003* 0001* 0.42 0001* 0.20 P

P < <

based and 2002 one, OVA

open, 8 5

F AN l.

1.28 2. 2.44 trial F

25.19 106.34 from

exclusion, during

8 527 40 40 48 48

2, 2, 4, 4, 4, df cages -statistic F

MEANS

frame 0.05.

LS greater

and _<_

a P of

from at open, Probability Switch

exclusion,

comparisons signiﬁcance in 4.7.

Date After Treatment Treatment Treatment*Date

Effect Date Treatment*Date

wise indicates

Pair Table plants)

_BEIQLeSwitch

108 * in line. plants) dotted the random ﬁve below of shown are measurement on indicated date (based the at height cages plant on frame P

1.0 0.08 0001* 0.43

1” 0001* 0.10 <

< based and open, F

F 1.02 ANOVA

1.75 3.75

0.01 2002 70.05

120.56 from two, exclusion,

11.9 39.2 39.2 6.76

46.2 46.2 of trial 41'

if 8,

8, 2,

2, 4,

4, -statistic F during MEANS 0.05. LS greater cages g a P of from at frame and

Switch Probability Switch open, comparisons signiﬁcance 4.8.

Date After

Treatment*Date Treatment Date

Effect Treatment Treatment*Date

Before wise Pair indicates exclusion,

Table

109 pitfall (2 12 to cages

July % - - - 1.8 1.8 1.8 1.8

5.3

7.0 5.3 7.0 3.5 7.0 7.0 14.1

36.8 12 100.0%

( ’ l

21 3 3 - 56

4 VN - vh—«v—Iv—

two Total

exclusion Two in trial days - - - -

- % -

2.8 2.8 2.8

2.8 Trial

and °

11.1 11.1 11.1 16.6-

38.9 100.0% seven July) v

14 1 1 1 36 - - - -

- VG 4 Total 29 per to June captured (28 -

- % -

1.2 1.2 1.2

3.7 3.7- 7.4 9.9 3.7 2.5 6.1 4.9

16.0 38.3 100.0% one

v-IIWU—lv—lmo 3 81 2 3 3 31 - 5

predators v—v Total trial

in One

area .O. dwelling - - -

- % - - - -

Trial as 9.5 14.3 47.6 100.0% study ground the 10 21 2 - - - - 3 - NV - -

Exclrrs' - Total of .) (F C.) percent C. throughout (Say) (DeG.) (Le ) Le and (Say) Say (F.) (LeC.) (F. Chd. Dej. (Say) (Say) Duft. total, sodalis randomly santaecrusis

(DeG.) Michigan comma ochropezus rapidum spp. tricolor herbivigus

pensylvanicus __ chalcites lucublandus placidum quadriceps impressefrons

bipustulata _ placed aenea

familiaris Lansing, composition, those yclotrachelus Clavina Clavina C Chlaenius Harpalus Poecilus Scarites Stenolophus Stenolophus Harpalus Poecilus

Bembidion Species Bembidion Anisodactylus Amara Agonum

Amara East Species

versus

2002, Coleoptera 4.9.

TOTALS

Family

Carabidae Order August), traps/cage)

Table

110

dotted

the

exclusion in below

shown predators are

indicated July

date ground-dwelling

001*

of 26 the at

Switch captures (plot) After

area

pitfall July

003* 19

study

total of the

and

2002

results

July on

cages one,

005* 12 P

0003* 0.79 0001* trial

< based Switch exclusion 5.13 0.34 F

22.89 during of

Before,

-statistic F

0.05. (plot)

66 66 66 July 5 df 0001*

1, 5 3, 3, P

MEANS

area greater

at a

of study

from

the in

signiﬁcance

those Probabilities Exclusion-Plot comparisons

indicates

4.10. versus * Date Effect

Treatment Treatment*Date 5 wise line.

Pair

Table cages

111 dotted the exclusion in below shown predators are indicated date ground-dwelling the of at (plot) captures area pitfall study total the of and 2002 results cages two, on

0.99 0.06 0.23 trial based exclusion

1.48

3.64

0.05

F during of -statistic F

0.05. (plot)

62 62

df g

P MEANS area

at greater LS a study of from the in

signiﬁcance those Probability comparisons

indicates 4.11. versus

Date

Treatment*Date Effect Treatment wise

line. Pair Table cages

112 *

line.

during

July.

cages

dotted

the

frame July July 0001* 0001* 0.22

and

below 0.39

< <

0006* 0003*

12 29

interaction

shown

exclusion

are

open,

July July 0001* 0001* 0.42

0.31

treatment*date

003* 004* 8 < <

indicated

date

the

signiﬁcant

glycines

July

0002* 0002* 0.70

was

cages

adult

there

frame

that

and July

July 002*

0009* 0.17 P

19 0004* 0.40

005* 002* 0002*

number

P <0.001*

open, 0.11 0.84 0006*

the

determined

3.46 7.88 F

2.74 31.33 0.35 4.41

based

data

exclusion,

1 1

of 10.2

41.7 41.7

44.5 47.8

df of

June

July 4r

0.1 0.98

0.1 8,

8, 2,

4, 4.44.5 001* 0.54 004*

-statistic

28 15

Slicing

MEANS

greater

0.05.

from

Probability Switch

Frame-Open

Frame-Exclusion Exclusion-Open

Frame-Open Frame-Exclusion

Exclusion-Open Switch

comparisons

signiﬁcance 2002

4.12.

After

Date

Date Treatment

Effect

Treatment*Date

Treatment*Date Treatment

wise one,

indicates

Pair

Table trial

_Befnre

113 *

line.

July.

cages

dotted

the

frame

July

and July

0.11 0003* 0003* 0.14

below

<0.001* <0.001* 29

interaction

shown

exclusion

are

open,

July July

0.16 005* 0.22 0001* 8 0001* 0.57

treatment*date

1 < <

indicated

date

the

signiﬁcant

glycines

July

5 0001* 0001* 0.95

was

cages

nymph

there

frame

that

and

number P

July 001* 0.57 0001* 0001* 0001* 0003* P

July

0.19 001*

0.12

the < <

0.12

001*

open,

determined 3.52 9.49

3.44

0.75 4.19

46.18 F

based

data

exclusion,

of July

10.5.8 10

48.5 41.2 41.6 48.3

June

of df 001* 0.52 003* df

0.1 0.07 0.73

8, 2,

2, 9, 4, 4, 15

-statistic

Slicing

MEANS

greater

0.05.

from

at 2002

Probability

Switch

Frame-Open Frame-Exclusion

Frame-Open

Exclusion-Open

Frame-Exclusion

Exclusion-Open one,

Switch

comparisons

signiﬁcance trial

4.13.

Before Date After

Treatment

Treatment*Date Effect Treatment Date

Treatment*Date

wise

Pair indicates

Table during

114

in -

- 1.5

1.5

3.0 3.0

7.5 3.0

9.0 4.5

observations

11.8

55.2 100.0%

Michigan

2 - 69 v—IWM - \ONMOOV" Frame Total intrusive and

Lansing,

- -

- 1.8

1.8

3.6 5.5

7.3

5 5.5 %

7.3 61.7

East

100.0%

- -

twat—1mm 34 m 4 Open Total 2002, non-intrusive 5° N

July), 9.

- - -

- - - - - . hi % 12 77.8 or § 3-minute

- — -

- N ox

- - Exclusion Total June during (28 one,

obsertved

adults

adults adults

trial

adults

larvae (L.)

Geer Guerin

during

De adults

predators

Stephens

(Pallas)

larvae adults

switch

nymphs (Say)

percent

maculata

spp.

spp. convergens

cage

cognatus

septempunctata septempunctata

nymphs

adults axyridis and

spp. spp.

insidiosus insidiosus total,

prior

Chrysoperla

Orius

Coleomegila Coccinella Coccinella

Nabis

Harmonia

Philonthus

Nabis Hippodamia Species cages composition, frame

and

Anthocoridae Nabidae

Species

Chrysopidae

Coccinellidae Cantharidae open, Family

4.14.

Neuroptera: Heteroptera: Heteroptera:

TOTALS

Coleoptera: Coleoptera: Order: Table exclusion,

115 in cage). observations

3.5 2.1

0.7 0.7

% 2.8

0.7 0.7 18.2

22.4 48.3 100.0% exclusion 141 v—(I—ﬂ l 1 3 intrusive 32

26 5 69

Frame

Total (now and cage 1.7 1.5 3.1 3.7 7.5

2.1 % 5.9 0.2 0.4 0.6 0.6 0.4 0.2 4.6 14.0 30.6 22.9 100.0% open 17 1 15 146 10 l 109 8 36 3 3 28 2 67 477

7 2 22

Open Total non-intrusive former 2002

M the

1.9 1.3

one, 2.6

0.6

0.6 0.6

31 58.7

3-minute 100.0%

trial

1 155 l

NN 3

48 4 during

Exclusion

Total containing during plot July) observed 29

adults

adults to physical

adults

larvae

(L.) the

Geer July predators

Guerin

adults in

larvae

nymphs

adults of (15

Stephens

(Pallas) larvae

(Say)

larvae

adults placed

(Say) (Say) switch percent

maculata

larvae

spp. spp.

convergens were

cognatus septempunctata

septempunctata

munda adults nymphs

axyridis axyridis and cage

spp.

spp. spp. traps

insidiosus

insidiosus total, after

Chrysoperla

Orius Orius

Coleomegila Coleomegila Cycloneda Unidentiﬁed

Coccinella Coccinella Syrphid

Leucopis

Harmonia Harmonia Hippodamia Nabis Nabis

Philonthus Species pitfall cages frame composition, switched, and

were

Nabidae Anthocoridae

Species

Chrysopidae

Coccinellidae Cantharidae

open,

Syrphidae Chamaemyiidae cages 4.15. When

Neuroptera:

Heteroptera:

TOTALS Diptera: Diptera:

Coleoptera:

Coleoptera: Table Order:Family exclusion’,

116 Table 4.16. Probability of a greater F -statistic based on predators observed during 3-minute non-intrusive and intrusive observations

in exclusion, open, and frame cages during trial one, 2002

Effect Before Switch 41f F P

Treatment 2, 15.9 0.93 0.41 Date 4, 50.5 6.35 0003* Treatment*Date 8, 50.7 1.37 0.23

After Switch df_ F P Treatment NVOO ,8 .73 31.52 0001* Date ,3 8 1.78 0.15 Treatment*Date ,3 8.1 2.20 0.05 *

117 15 July 19 July 22 July 26 July 29 July Exclusion-Open 0.09 < 0001* 0.13 0.65 0.53 Frame-Exclusion 0.34 0.55 0.69 0.21 0.30 Frame-Open 0.32 < 0001* 005* 0.09 0.10

Pair wise comparisons from LS MEANS of exclusion, open, and frame cages at the date indicated are shown below the dotted line. *

indicates signiﬁcance at P 5 0.05.

trial g

line.

durin es

dotted

g ca

the July August 0.35

0.22

004*

0.93

0002*

26 frame 12

below and shown

are

July

exclusion

0.25

0004*

0001* 22 0.040 0002* 0002*

August open, 9

indicated

m2 1

date in the

at July August 5 0001*

0.57

0001*

0.14

0005*

003*

19 glycines < <

cages A.

frame adult

and

July P P August 0001*

0.63

0001*

0004*

0001*

0.91

0.21 0.52 0.19

0003* 0003*

number 15 2 < <

open,

on F

4.50 4.00 0.91 4.63

10.75

35.381 based exclusion,

10.9

36.7 7.96

50.1 42

50.2 July July

of df 0.60 0.001

0.001

0.76

0.41

0.28

8, 2,

8, 4, 2,

4, 12

29 -statistic < <

4f F

MEANS

reater

g 0.05.

P of

from

at Probability Switch Frame-Exclusion Exclusion-Open

Frame-Open

Frame-Exclusion

Exclusion-Open Switch

comparisons

signiﬁcance

4.17. After

Date

Treatment Date

Treatment*Date 2002

Effect Treatment*Date Treatment

wise indicates Pair

two, Table

_EQLQLQ

118 * line. frame dotted and the July August 001* 0.55 0.06 0.64 0001* 0001* 12 26 exclusion below < < open, shown m2 are 1 in July August 001* 0.87 002* 0.44 0001* 0003* 22 9 indicated produced date the at glycines July August 0.53 0001* 0001* 0.25 0001* 0006* A. 19 5 < < cages nymph frame of and P P July 002** 0001* 0001* August 0002* 0.75 0005* 0.56 0.09 0.15 0.56 0002* 0001* results < 15 2 open, on F F 8.30 2.58 5.28 0.93 19.13 35.75 based exclusion, 11.2 8.52 50.2 50.2 42 42.1 July July of df 0.64 0002* 0006* 0.73 0.24 0.17 8, 8, 2, 2, 4, 4, 12 29 -statistic 4f F

MEANS 2002 greater 0.05. LS a

5 two, P of from

at experiment Probability Switch Frame-Open Frame-Exclusion Exclusion-Open Frame-Open Frame-Exclusion Exclusion-Open Switch comparisons signiﬁcance

4.18. during After Treatment Before Date Date Treatment*Date Effect Treatment Treatment*Date wise indicates Pair Table

cages

119 foliar Lansing, - - intrusive % - 1.1 2.3 5.7 0.6 0.6 0.6 16.7 10.9 61.5 East 100.0% and 174 107 1 10 1 19 l 29 - - 2 2002, - 4 Frame Total two, trial non-intrusive - - - - 1.3 % 0.4 5.1 0.4 0.9 0.4 32.9 58.5 100.0% July) minute l 1 12 l 137 3 - 234 77 - 2 - - Open Total 26 3 to July during ------2.0 2.0 % 0.7 25.7 69.6 (12 100.0% 103 148 1 - 3 38 3 ------Exclusion Total observed during switch adults adults adults predators larvae adults larvae cage Geer total to Guerin of De adults nymphs prior (Pallas) (Pallas) larvae adults (Say) (Say) percent cages maculata spp. spp. convergens septempunctata(L.) septempunctata and adults nymphs axyridis axyridis frame spp. spp. insidiosus insidiosus total, and Chrysoperla Chrysoperla Orius Orius Coleomegila Coccinella Coccinella Nabis Harmonia Species Harmonia Hippodamia Nabis open, composition, exclusion, Species in Nabidae Anthocoridae Chrysopidae Coccinellidae Family 4.19. Table Heteroptera: Neuroptera: Heteroptera: TOTALS Coleoptera:

Michigan. Order: observations

120

the

of foliar

cage).

part

3.3 %

0.8

4.9

0.8

34.1

38.2

This

100.0% intrusive

exclusion

and

121

Frame

Total 2002. Michigan

(now two,

cage

8.9

8.1

0.4

3.9

4.3

0.4

0.8

0.4

trial

23.6 48.8

Lansing,

100.0% non-intrusive open

East

during

126

258

Open Total

former 3-minute

the

- -

switched,

1.1

1.2

2.2

6.7

0.3

August)

55.1

100.0%

during 12

were

Exclusion Total

containing July

cages observed

plot

(29

adults adults

exclusion switch

physical

predators

adults (L.) the

and

Geer

cage total

larvae

nymphs

adults of

open

after

(Pallas)

larvae

larvae adults

placed

(Say) (Say)

after

percent

cages

maculata

spp.

spp. spp.

were

septempunctata

nymphs

adults

axyridis

axyridis and

2002,

spp.

frame

spp. spp.

traps

insidiosus

total, and

during

Hemerobius

Chrysoperla

Orius

Coleomegila

Coccinella

Nabis

Leucopis

Harmonia

Nabis Species

pitfall open,

from composition, switched,

conducted exclusion’,

were

Nabidae

Anthocoridae

Species

Hemerobiidae

Chrysopidae Coccinellidae

was

Chamaemyliidae Family

cages 4.20. When

Neuroptera:

Heteroptera:

TOTALS

Heteroptera:

Diptera:

Coleoptera:

Order:

observations

experiment Table

121 * line. intrusive and dotted the August July 002* 0.46 002* 0.33 0.37 0.07 12 26 below non-intrusive shown are 3-minute July

0.83 0.13 0.09 August during

0.068

0003* 0007*

22 9 indicated date Observed the 1 at July

0.97 0001*

0.008 0002* 002* 0.1

August 19

5 2002 cages predators

of two,

frame trial and number

July 0001* 0.13 002* 0001* 002* 0.34 0.20 0001*

0.67 0001*

0.91 0.59 P P the August < <

15 during < < open, 2 on 1.74 1.17 3.25 5.78 F

F cages 10.81 55.18 based

exclusion, frame 10.1 36.2 32.2 6.73 48.3 July

48.3 July of df 4r 0.89 0.60

0.73

0.60 0.70 0.89 8, 2, 8, 2, 4,

12 and 29 -statistic

F open, MEANS

greater 0.05. LS

5 P of

from

exclusion, in

Probability

Frame-Open

Frame-Exclusion Exclusion-Open

Frame-Exclusion

Frame-Open Exclusion-Open Switch

comparisons signiﬁcance

4.21.

observations

Treatment Date After

Date Effect Treatment*Date Treatment*Date

Treatment

wise indicates

Pair foliar Table

_BeforeSwitch

122 > 1 m A A Fine mesh netting

0.9 m i, A A A {5

Plastic barrier

0.7 m

Pitfall

Figure 4.1. Design of 1 m2 cages used during 2001—2002. Both cage types contained the

same cage materials to account for potential cage effects. A. open and B. exclusion.

123 LID-Open j

difference

p— +____ 92599409 I _4A

(°C)

exterior)

0 I I.

minus

temperature

(interior

SEM)

Mean -2 - -. L i , _ 7_ : “r—' 3 7»_ D —..""' 3 '_—.. 3 2 00 a C0 2 50 2 DD a Q- m O I‘ V "‘ 1 1 I I é. — —‘ N "‘1 f\ V '— co — N N

Figure 4.2. Mean 0: SEM) temperature difference (°C interior minus the °C exterior) of open and exclusion cages during 2001, East Lansing, Michigan.

124 I +0969 I I 2713:6999 j I

difference

exterior)

humidity

rnrnus

relative

interior

(

SEM)

(_+_

Mean I

‘

’

I_-

“

3-Jul

4-Sep

lO-Jul

l7-Jul

31-Jul

24-1111

7-Aug

14-Aug

2l-Aug 28-Aug

Figure 4.3. Mean (i SEM) relative humidity difference (interior minus the exterior) of open and exclusion cages during 2001, East Lansing, Michigan.

125

80‘

3 75 T T-U-Open TI 70 ‘ I I 5 I-O'j'Exclusion I a 65 . In 60 . 8.$— 55 I g 50 -‘ 5,1 45 :51 40 ~« "I :5) 35 E 30 .1 25 I: _‘S 20 o. I sc ,015 ‘. 2 5 l

O ,i? v 2‘

I ' ‘

" ”

6-Jul

l3-Jul

20-Jul

27-Jul 3-Aug 29-Jun 10-Aug l7-Aug

Figure 4.4. Mean height Ct SEM) in cm of ﬁve randomly selected soybean plants in the open and exclusion treatments during 2001, East Lansing, Michigan.

126 .0 <15 0 .TT LLL _ a I-D-Open I

E 20 ‘ I+Exclusion I "=3 I 5 I .E I 2 15 . m 33' I ' ' 2A 0" 10 ~ I LU - U} +1 ' .' ' T: ,I T ' I e 9 5 . -‘ . .I E :' a. I t: I 8 0 . a 1 .. 7 *7 _ r_ 2 ,2 : _.2 s;'—' 2.'—' 2.'_' '_.2 00= =00 00= =50 g o- . «a :5 I; 4 4 <.< <5 3‘ ‘F - 2?; — ~ N m I\ v —- oo ‘1' v— (\I N

Figure 4.5. Mean (i SEM) abundance of foliar-foraging predators based on a combination of 3-minute non-intrusive visual examination followed by hand examinations of foliage in the open and exclusion treatments during 2001, East Lansing,

Michigan.

127 I I I I I I I N M M ai-r I-o-ToT—T" .2 20~ T I pen. I :3 I—ﬁ_ExclusnonI Sek U) z . E 15 a “d —-I o 49 A 2 "210 - LL] 3 I W c: 23': .. o‘ 5 I. I I I: I g I .‘ l . Q) I

2 0 T T — 1T 'TT _T I I c: "‘ "‘ "" "‘ 60 DD 00 DO 00 -= :1 a 3 :4 3 a 3 2 3 < :1 < :3 < :3 < :1 a: :1 83 - N N «'4 o' (\l 4 ... _q — N M

Figure 4.6. Mean (:1; SEM) abundance of Harmonia axyridis adults and larvae based on a combination of 3-minute non-intrusive visual examination followed by hand examinations of foliage in the open and exclusion treatments during 2001, East Lansing,

Michigan.

128

combination Figure Michigan. examinations U) “a 2 m E O % m a .l‘ a c: 8 c: o- g ”I g =9 .2 2 'c '9 S a ‘= :3 o m 10

4.7. — O 5 U1 ,b (<3. _=, b I l I s: . Mean 43‘ of of foliage «a -— 2. - 3 (i R SEM) minute in / .5: -- —. :5 I the abundance open non-intrusive I -- i; a .— I and of exclusion é —- N a : l‘ l Orius 129 I visual "f m a —- :— insidiosus I I treatments «7 examination l‘ ‘F I on = adults during v -' ‘F . I on = I I and — followed 2001, l-O-Exclusion -D-Open nymphs N ‘F — '— on = “_h East based N 00

I-D-Open tn . . . E m - " ‘-0- __ Exclusnon . E Q I- 3.» 10 co <2 2 :1 '0 J. «3m MD: 0'5 2:,Au I m0.LU" . H v $— .8 ' _v E I I t:3 I I I I ___ g ' - é) o- +— -— - —— - . r . II _=. c: a:—- -:7: 5.:- --—.=.’ -47‘ 2 on 2 on :5 on g on 6 m C l\ V '-' n u n I

N ._ —— N m h E (T. :3 4-Sep

Figure 4.8. Mean (i SEM) adult A. glycines population per 5 plant tips during 2001 in open and exclusion cages during 2001, East Lansing, Michigan.

130 40 ‘ ..

LT.- Exclusion ‘

glycines/

nymph

tips of

plant

SEM)

(i l

l ‘i

I I

‘

number on on a. g '5 3 '5 '3 3 :9 = :s o v-; '7‘ '7 '7 '7‘ '7 < < <2 ”.1 b M O t‘ V "‘ | N v—q —— N m [\ .1. 05 v

l4—Aug N N

Mean

Figure 4.9. Mean (i SEM) nymph A. glycines population during 2001 in open and exclusion treatments during 2001, East Lansing, Michigan.

131 two

per S-Exclusion cagei fx— Entire plot I \l .O u:

found

days

seven

axyridis 0.5 -

per

H. T

traps .I. T 1-

SEM) 0.25 I1 1 I

sticky

(-_+-_ I T., J I .L

no. I |

0 IE a +—- a .0 v

Mean 3 E E E .25 m :5 Sf. x' «7"»

Figure 4.10. Mean (i SEM) abundance of Harmonia axyridis (Pallas) found on sticky traps in the exclusion cages versus those found in the entire plot during 2001, East

Lansing, Michigan.

132 per I;Exclusion cage I l I I I I-5- Entire plot

found b.)

days

seven

insidiosus

per

traps

SEM)

sticky

(i l T 1':

two

no. L A '1- l I

Mean l I I

0 k1 A1 ' if —* I 7 h Tqi _ ___—‘i” 7 T“ I '5 '5' '3 E '5 '7 '7’ '7‘ '7’ '7‘ m o t\ v ~— -— —u N m

Figure 4.11. Mean (i SEM) abundance of Orius insidiosus (Say) found on sticky traps in the exclusion cages versus those found in the entire plot during 2001, East Lansing,

Michigan.

133 'o 28 ———*Ae — . ::: -'X- Exterior pitfalls o0 +Exclu510n . cage pitfalls . I i3 ‘9. ‘ V ' * a g ..o a) r: “O °>’ T T 2g 333 1 1I :1" °- 1 ; o g 11: "‘ b u T 2 a Li 3‘, o. 3 J TT 1 ‘ _¥3 ._,, #3 7 A 7 air A _ __ 2 0 7‘ V .' A ":5 ":3 '5 ":3 '7 7 7 '7

Figure 4.12. Mean (i SEM) predator/pitfall trap/day of ground beetles in pitfall traps in exclusion cages versus those in the study area during 2001, East Lansing, Michigan.

134 difference

(°

exterior) -4 r. , W, , -._,_-, ,. +Exclusion then Open

rmnus -6 —'D—Open then Exclusion temperature ——A—— Frame -8 . . — -O- -Exclusron control

(interior SEM) _10 I — -E.I- -Open control (i -12 . ”--,iiﬁip “a. Mean —. g —.~'5' ’1''5 7‘3' '5'7 7'5 7‘2' '3'"r 50 N \o o v 00 m \o N I—t —- I-! N N

difference

(°C)

exterior)

minus

temperature

(interior

SEM)

Mean

I 7 f l6-Jul lZ-Jul

l-Aug

20-Jul 28-Jul 24-Jul

Figure 4.13. Mean (i SEM) temperature difference (°C interior minus the °C exterior) in open, exclusion, and frame cages during 2002. The arrow indicates when open and exclusion cages over 1 m2 plots were switched during: A. Trial one; and B.

Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched.

135 30-A.* L 7 7 ¢ f A a.

25 I 'r I ~4Excltfsibnmen Open I I

difference per 20 I—D—Open then Exclusion I i —-A-——Frame 15 1 - -O - Exclusion control

humidity exterior) 10 - :- Open control

minus 5 I

relative

0 ‘

SEM)

(interior

(i -5 I

db V -10 = * ﬂ . Mean g 3 3 ‘5 ‘5 3 '5 3 0;,-—. N'7’ o'7 o'7 '7‘v co"I" m'7' <5'3 N u— v—I —‘ N N | | J b) 0 on N U!

difference

per N O 4 _ U!

humidity

exterior)

ES l

minus M

relative O

SEM)

(interior r I M (i

Mean 3 '5 '7 7 N \0 ~ _

l-Aug

ZO-Jul 24-Jul

28-Jul

S-Aug

9-Aug

One randomly selected replicate had cages that were not switched to show the trend if cages were not switched.

136

' A. 110 ' ' ” 3 —o—E.xclosure then Open II 100 ' —c)—Open then Exclusion

plants %

5 1 ' -‘ '— Franc I 80 — + - Exclusion control

per — -CI- — Open control 5’ m.

height

50-

plant 40 :I

Mean 30‘ zoI . ~ ~ I ~9 5 E E E E E E E. _a l a I o l i o 33 N ‘0 2 3 E 8 :9.

120 - ~~ —— 3110 : 5100‘ cu. .n h 90 . a 80 ,4

.20E 70 . g 60 ‘5' 50- "5.. 40. 5 E 30* 20 ~ ,, —~ ~ , f — — ~- —— 2.'5 3;'5 a'5 3'5 :2“:5 5.3 on 5.5 °° 3.5 °° .— --I N N N —- V3 0‘

Figure 4.15. Mean (i SEM) plant height per ﬁve plants in open, exclusion, and frame cages during 2002. The arrow indicates when open and exclusion cages over 1 m2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched.

137 l A. 7 if. _, ,i 7" 7,—7 ﬁe. .

*Extegorpitfalls I

pitfall ‘-0—Exclusion cage pitfalls

day

predator/ trap/

SEM)

Mean 0 * iﬁ 7—7——— ~ . - '57 '5*9 '37 '5“r 3'7 '37 V} 0‘ C" l\ —i m "" —‘ N N

pitfall

day predator/

trap/

SEM)

Mean

l I 0 hj—i‘A—+~W~7~——-~——~ T :3 :3 3 D

O\ M (\ '-' I I v-I N N M V 00

Figure 4.16. The mean (i SEM) abundance of carabid beetles collected per seven days in 10 total pitfall traps from exclusion cages (2 traps/ cage) versus 10 total pitfall traps placed randomly in the study area during 2002: A. Trial one; and B. Trial two.

The arrow indicates when open and exclusion cages over 1 m2 plots were switched, at which time eight pitfall traps were in exclusion cages and eight within the plot.

138 10000 ._A. 77,, 7* v i , J 7, i 7 i ' i‘ 7 A. ‘ﬁ:‘— Exclusion'thenOOpen 1000 } —O—Open then Exclusion : z ’

adult ——A—- Frame . I

of ‘I — 9- Exclusion control / X

plants 100 E,- '0‘.’ 'Qgsniontrol , I 10 SEM) (i elvcmes/ number

Mean 0.1 r — . ,, -. a _- - _ ., 2-Jul 6-Jul l4-Jul l8-Jul lO-Jul

22-Jul 26-Jul

28-Jun

10000 . w 77 .55

A. ‘ 1000 -- , I

nymph

plants §

SEM)

(i O

elvcmes/

number

Mean '

* 0.1 -

2-Jul

6-Jul

18-Jul lO-Jul

l4-Jul

22-Jul 26-Jul 28-Jun

Figure 4.17. Mean (i SEM) adult (A.) and nymph (B.) A. glycines population (log scale) per ten whole plants in open, exclusion, and frame cages in trial one during

2002. The arrow indicates when open and exclusion cages over 1 m2 plots were switched. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched.

139 10000 : ._ ___,i , —— . W, - , .7, , vi ,A.

g 1000 . —O—Exclusion then Open ‘05 .2 I —D—Open then Exclusion E g I — A—- Frame m g 100 I - -O— ‘Exclosion control (2' : — 6- 0an control V 8 ’ii' i... E 10 3 a S a s: 1 + c: I a a) 2 0.1 . i - - _ — if _ ~— '5 '5 '5 '3 ‘5 g0 g0 g0 -—.a. -se -s:5 t-ia Hob <5 ‘F at w— --I N N N F— W 0‘

10000 -» —— - — -— — -— — — — —— <:' B. 4: a. E 3 >» s: 1000 I ‘7- .2 “— c: 3. .9. 2 g 100 53 3 ' +| 9. v h )3 10 ~ 3 g aE z'8 1 . g G! E 0.1 i i if T_ 'A 7 *7 __ 7’ a“ “— 7’ fl *~_ fi‘" 'T‘_‘_—‘_ —‘ Y‘ ‘_‘ " '— _'T_‘ ‘ "7" '3‘ '5 '5 '5 '5 g0 g0 g0 3'} 12 3 2'3 22 ‘F fi< ‘F — '— N N N — W 0‘

Figure 4.18. Mean (1- SEM) adult (A.) and nymph (B.) A. glycines population (log scale) per ten whole plants in open, exclusion, and frame cages in trial one during

2002. An arrow indicated when open and exclusion cages over 1m2 plots were switched. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched.

140

120 .7 7 it —7 7—7 — 7 — ﬂ ’ .5 A.

:9: _ i ., “o 100 7 +Exclusron then Open g —D—Open then Exclusion ‘ . 'E I +Frame l 2 80 I _ ‘ m . ' ° 0 ' - Exclusron control g 7:7‘Efg9pen control ma 3 o 6° ' m \tl 40

$— .9 a: "o 20 « 8 a. c :3 0 a) I~ T ﬂ 2 3 3 3 '5 3 3 '5 '5 -i '7‘ 7 '7 '7' .7 _I, .7 N .., ..q —- N N

S 120 ~ - 7 7w f *A' _q. k" ’** “I '—= B. e y I‘ g 100 « I .s’ I 2 80 m I g I A m 60 I 2 '8 i 5; I

,_ 9 I ('3 I “o i Q) I L: t:==~ a ... E 3 "s 3 '5 ” ' —.. -.~ -.~ -.» '5’-.~ g0 :3? 2? In 0\ M 1‘ "‘ it it it — --' N N M V w N

The arrow indicates when open and exclusion cages over 1 m2 plots were switched during: A. Trial one; and B. Trial two. One randomly selected replicate had cages that were not switched to show the trend if cages were not switched. 141 examination on not A. Figure arrow

a? 2 t {3’ —t 'o .53 3.5 53]

o~ ”a? 2

+I g:

('3

S <6 = g a

>2 73 E f; t' 3'" {J 2

33 c

3 g E k 9: as Trial a

2 c... '0 '5 .... g switched

o. 0 £—

‘Y c a v;

m ‘o— 2 "g g ,8

8. 3— c: "‘- 0 o combination

. indicates

25 30

10 15

20 25

15 10

4.20. O s- OJ 0

5 one; C

I: . ‘

I 1 ~ '

. I

. - IA. :

2 V3

N 0;, 2.

I I ; of to and

—D—Open —0—Exclusion

--EI--Open --+--Exclusion +Frame Mean foliage show when

—* B. of

m _ a: (i Trial

_ 3-minute

a 0‘ the open

. SEM) in | trend the

control then two. and open,

c —~=.’ __ abundance

control

then

Exclusion non-intrusive

V7 if

:=. m exclusion One

. cages

Open exclusion, randomly |

o :74 _

._ were

i of

. 142 cages

2 l‘

. Harmonia visual not and selected over

a: __

—- v switched. I frame examination

2‘

. 1 axyridis m2 replicate treatments

00 a: __

if c— plots I

< (Pallas) 2?

I followed were had

N m __ 5.: during cages switched adults

I that 2002. hand

N o -=-.= __ based during: were The

- I

~ . I 1 I ; I

I i ; I I . "o 30 . ..,, 7 , 7 , v E i g IA.

E 70 I -—O—Exclusion then Open I T m 60 -. —Cl—Open then Exclusion 5.. I 8‘ +Frame g 50 " - - -O - - Exclusion control E2 :3 40 i , - 0:996" “"1901 10' :3 g ::: 30 - E «2 -—-.2 20 ' gQ 10 - 3 .

'3 O .. I § 5 '3 '3 3 '5' 3' s '3 E ‘U 0,3H “‘'7 ‘°'7' 2'7‘ '72 1°.'7‘ 5:}'7 3H

'0 80 - — ifi i g 5 B. -— ::' 70 . E m ,_ 60 . a I E - ' m 3 4o ._ dj 0 mg ._.5. 30 _I I a 8 I 5. 20 ~ .52 . Q. . . 8 10 j- a: I E 3 0 IH—W—l—m—G' f"; T? L? 37: :3 3 3’ 3° tn 0\ m 1‘ ~ . . 2 —. —- N N m v 00

lZ-Aug

per 20 I r:.—Exclusion then O_pen I 55 I —D—Open then Exclusion I I I SEM) :15 i +Frame :

(1; 2:: I . ' '0 --Exclusion control I ,' g I ‘49u II II -__:='_'_- 'Ommrol__ I .' I a s '. r: 10 ~ . cc: .5: insidiosus 2

Mean --. g 7'5' '57 'ﬁ'3 7'5 '5'-.~ ".~'3 '7'3 0;, m o :5 v 00 N o

25---. ,h, , , , - I

per

obs

SEM)

foliar

and

insidiosus

Mm 0.

I Mean l-Jul l9-Jul lS-Jul 27-Jul 3 23-Jul 8-Aug 4-Aug 12-Aug

Figure 4.22. Mean (i SEM) abundance of Orius insidiosus (Say) based on a combination of 3-minute non-intrusive visual examination followed by a hand examinations of foliage in the open, exclusion, and frame treatments during 2002.

144 Literature Cited

Alﬁler, A. R. R., and V. J. Calilung. 1978. The life history and voracity of the syrphid predator, Ischodiodon escutellaris (F.) (Diptera: Syrphidae). Philippine Entomologist. 4(1-2) 105-117.

Chen, K., and KR. Hopper. 1997. Diuraphis noxia (Homoptera: Aphididae) population dynamics and impact of natural enemies in the Montpellier region or southern France. Environmental Entomology. 26: 865: 875.

Colfer, R.G, and J .A. Rosenheim. 2001. Predation on immature parasitoids and its impact on aphid suppression. Oecologia. 126: 292-304.

Dixon, A. F. G. 1985. Aphid ecology. Blackie and Son. New York.

Garcia, J .F ., and RJ. O’Neil. 2000. Effect of Coleus size and variegation on attack rates, searching strategy, and selected life history characteristics of Cryptolaemus montrouzieri (Coleoptera: Coccinellidae). Biological Control. 18: 225-234.

Han, X. 1997. Population dynamics of soybean aphid Aphis glycines and its natural enemies in ﬁelds. Hubei Agricultural Sciences. 2:22-24.

Hance, T. 1987. Predation impact of carabids at different population densities on Aphisfabae development in sugar beet. Pedobiologia. 30: 251-262.

145 Hilbeck, A., C. Eckel, and G. Kennedy. 1997. Predation on Colorado potato beetle eggs by generalist predators in research and commercial potato plantings. Biological Control. 8: 191-196.

Landis, D.A., and W. Van der Werf. 1997. Early-season aphid predation impacts establishment and spread of sugar beet yellows virus in the Netherlands. Entomophaga. 42: 499-516.

Lenné, J .M., and P. Trutmann. 1994. Diseases of tropical pasture plants. CAB International, Wallingford, UK. P. 69.

Li, W.M., and Z.Q. Pu. 1991. Population dynamics of aphids and epidemics of soybean mosaic virus in summer sown soybean ﬁelds. Acta Phytophylactica Sinica. 18(3): 123-126.

Obrycki, J .J ., and T.J. Kring. 1998. Predaceous coccinellidae in biological control. Annual Review of Entomolgy. 43: 295-321.

Ostman, O., E. Ekbom, and J. Bengtsson. 2001. Landscape heterogeneity and farming practice inﬂuence biological control. Basic and Applied Ecology. 2: 365- 371 .

SAS Institute. 2000. SAS/STAT user’s guide, release 8.1 led. SAS Institute, Cary, NC.

Snyder, W.E., and DH. Wise. 1999. Predation interference and the establishment of generalist predator populations for biocontrol. Biological Control. 15: 283-292.

Syndmondson, W.O.C., K.D. Sunderland, and M.H. Greenstone. 2002. Can generalist predators be effective biocontrol agents? Annual Review of Entomology. 47 : 561-594.

Schneider, F. 1971. Bionomics and physiology of aphidophagous syrphidae. Annual Review of Entomology. 12: 103.

146 Stary, P. 1995. Natural enemy spectrum of Aphis spiraephaga (Horn: Aphididae), an exotic immigrant aphid in central Europe. Entomophaga. 40(1): 29-34.

Takagi, M. 1999. Perspective of practical biological control and population theories. Researches on population ecology. 41: 121-126.

University of Minnesota Extension Service Web Site: http://www.soFLbeansumn.edu/crop_prod/insects/aphid/aphid.htm

Van den Berg, H., D. Ankasah, A. Muhammad, R. Rusli, H.A. Widayanto, H.B. Wirasto, and I. Yully. 1997. Evaluating the role of predation in population ﬂuctuations of the soybean aphid Aphis glycines in farmers’ ﬁelds in Indonesia. Journal of Applied Ecology. 34: 971-984.

Wang, C. L., N. I. Siang, G. S. Chang, and H. F. Chu. 1962. Studies on the soybean aphid, Aphis glycines Matsumura. Acta Entomologica Sinica. 11: 31-44.

Wang, S.Y., X.Z. Boa, Y.J. Sun, R.L. Chen, and BF. Zhai. 1996. Study on the effects of the population dynamics of soybean aphid (Aphis glycines) on both growth and yield of soyabean. Soybean Science. 15(3): 243-247.

147 APPENDIX A

Effects of soybean varieties on A. glycines survival and reproduction

This test was conducted to determine how adult A. glycines survival and reproduction on the soybean varieties Mycogen (5152RR) and Pioneer (93382) that were used in our 2001 and 2002 ﬁeld and laboratory studies, compared to other common soybean varieties used in Michigan.

Materials and Methods

Insects.

During late June 2002, soybean aphids, Aphis glycines Matsumura, were collected from soybean ﬁelds in East Lansing, Michigan. Aphids were transferred to greenhouse grown Mycogen (5152) soybean plants and reared in a growth chamber held at approximately 25 0C and 70 percent relative humidity and a photoperiod of

16:8 (L:D).

Plant Preference.

Six varieties of commonly grown soybean in Michigan were used for this study: Mycogen (5152RR), Pioneer (93B82), Michigan Grown (Jock), Asgrow

(AG2602), DSR 232 (Nematode Resistant Variety), and Garst (D261RR). Plants were grown from seed in the greenhouse in 10 cm2 diameter pots. When plants reached the V3 stage, they were transfer to the lab for testing. Apterous adult A. glycines were conﬁned on plants in 1 cm circumference clip cages constructed of 1.8 cm diameter Cresline® PVC pipe and their survival and reproduction measured over

3 days.

148 Five adult A. glycines were placed on each plant, two in one cage, and three in

another, attached to the underside of the middle (V2) trifoliate soybean leaf. Six

replicates (plants) of each variety were used, and a set of three replications of each

variety were randomly selected and placed in a one of two growth chambers held at

approximately 25 i l 0C degrees Celsius and 80 percent relative humidity. Three complete trials of this experiment were conducted.

Aphids were conﬁned to these cages for 72 h (three days), and at 24 h (1 day),

48 h (2 days), and 72 h (3 days), data was collected on the total number of adults remaining and the offspring produced on each variety to detennine if there was a difference in adult survivorship or offspring production between varieties.

Analysis.

Adult survival and nymph production over the course of three days was determined by a Chi-square test at P g 0.05, using the GENMOD procedure of SAS

(SAS Institute 2001).

Results

There was no significant effect of soybean variety on adult A. glycines survival in any trial (Table l, 2, 3). Adult numbers generally decreased slightly during the three days of the trial due to either natural mortality, accidental crushing of aphids when replacing clip cages, or a combination of both. There was a significant effect of day (P: 0.02) only in trial three, where aphid survival decreased significantly from day one to three (Table 3).

In contrast, soybean variety signiﬁcantly affected nymph production and survival. There was a signiﬁcant variety effect in all three trials (P < 0.002; P < 0.001;

P < 0.001), as well as a signiﬁcant block effect (P < 0.002; P < 0.004) in trials two

149 and three. There was also a signiﬁcant (P < 0.001; P < 0.001; P < 0.001) day effect in

all three trials, as nymph numbers generally increased over time.

Looking across the trials several trends emerge. Asgrow (AG2602) had the

highest ﬁnal nymph production at three days in two of the three trials while DSR 232

had the lowest nymph production in two of the three trials. Mycogen (5152RR) and

Pioneer (93B82) were never signiﬁcantly different from each other in nymph

production or survival. They were also never different from the variety with the

highest nymph production/survival in any trial. These results indicate that these two

varieties are suitable host for A. glycines nymph production/ survival.

Discussion

These tests showed that A. glycines adults are capable of surviving on a variety of soybean cultivars used in Michigan. We found no evidence that adult survival differs among any of the six varieties tested. We did detect differences in nymph production/survival among the six cultivars. The cultivars selected for field trials did not differ from the best variety in their suitability for nymphs; therefore, it is highly unlikely that the field trial results are biased due to poor adult survival or low offspring production/survival. This experiment shows that in relation to A. glycines adult survival and reproduction, the varieties selected were broadly representative of what Michigan farmers may experience. It was also observed that the nematode resistant variety DSR 232 may be less suitable for A. glycines and should be explored further in lab and field trials to determine its utility in an A. glycines management program.

150 l), _<_

main (day

h 24 Below 3.1a 3.9ab 3.43 2.9b 2.3ab

2.7ab at square i : i i i i 3 line. (Chi 30.0 26.8 22.2 23.2 27.5 27.5

Day soybean

above of different

3.0ab 3.0b

2.2ab 3.4ab 2.3b

2.4a 0001*

i -_+-_

i -_+-_ i

0002* 0.80

numbers

0.80

varieties P:

16.2

18.8 18.2

17.5

22.7 20.7

P: P:

Day

of P:

right 6.13,

signiﬁcantly

284.59,

24.59,

0.06,

Offspring different the

not )6:

78: )8: )6:

1.6b 1.2bc 1.0abc

1.7ac six

2.2abc

0.8ac to

10, 1,

are

5, 2, *

:1; i -_t-_ i

i on a (s)

8.0

7.0

9.8

Day df= df= df= df=

12.0

11.3

10.2 by

letter produced 0.6 0.2 0.3 0.3 0.3 0.3 same indicated 3 i -_I-_ i i i i is the 3.4 4.2 Day

4.2 4.0 4.5 4.0 offspring

with and Signiﬁcance means 0.3 0.2 0.4 0.2 0.3 0.3 2 1.0 :L' i i :1: i i 0.67

0.82 0.97 remaining 4.5 0.05. P: Day 4.5 4.2 4.5 4.2 4.5 where P: P: P: g P 0.19, at 0.88, 0.80,

0.05, glycines 76: MEANS, test

76: 78: )6:

0.2

0.4

0.2 0.3 0.3

0.2 1 1, 10, 2,

it; i

i i

one

adult

Day

4.7

4.7 4.2 4.7 4.3 4.5 df= df= df= df=

the

of trial chi-squared from

SEM) during

Nymphs (day

resistant)

comparisons

number and

(Jock)

determined 72

wise

mean and

was Adults

Grown

(5152RR)

(nematode

pair 2),

(AG2602) The

(93882)

(D261RR) are One

232 l.

(day h

Pioneer Mycogen DSR Asgrow Garst

Variety*Day Variety Michigan Day

Block Variety Trial

0.05). Signiﬁcance

effects

Table 48

151 1), 5

main (day

h 24 Below 1.7ab 3.9a 3.6ab 3.3a 3.0b

3.73 at square i i i _+_; i i 3 line. (Chi 13.5 17.2 21.7 20.8 21.7 22.8

Day soybean

above of

different

1.8a

1.03

0001*

3.1a

2.2ab

2.5a

2.0b

0001*

11;

i i _-l;

i :1;

numbers

0.90

varieties 0002*

7.0

10.2

12.5

10.5

12.7 13.8

Day P: right

signiﬁcantly different the

not

78:4.92,

Offspring six

x2=281.26, 38:49.84, 38:924.

1.2a

1.4a 1.2a 1.1b

2.33 0.43 are

10,

2, 5, on

i i i

i i a (s)

2.7

6.8

5.0 5.8

7.0 7.4

Day

df=

df= df= df=

letter produced same indicated 0.4 0.4 0.2 0.3 0.4 0.4 3 is i :t _+_- :1; i i

the offspring 3.5 3.5 4.4 4.0 4.8 Day 4.0

with and Signiﬁcance means 0.3 0.3 0.3 0.4 0.2 0.4

2 1.0 i

i _-1_-_ i i i

0.76

0.72

0.66 remaining 0.05.

where P: 4.1 4.1 4.2 Day 4.8 4.0 4.0

_<_

P: P: P: P

0.65,

Adults

2.62,

0.67,

0.19, glycines test

MEANS,

76:

)6: 78: 0.3 0.2 0.4 0.3 0.0 0.3 1

10,

5,12: 2, i i _+_ i i i

two adult the 5.0

Day

df= df=

4.3 4.2 4.5 4.3 4.2

df= df= of trial chi-squared from

a SEM) during

by (i

3) Nymphs (day

resistant) comparisons

number and

(Jock) determined 72

wise mean and

was Adults pair

Grown

(5152RR) 2),

(nematode The

(AG2602)

(93B82)

are 2.

(D261RR) Two

232 (day h

Signiﬁcance Pioneer Asgrow 0.05).

DSR Variety*Day

Variety Michigan Mycogen effects

Garst

Day

Variety Block

Trial Table 48

152 l),

main (day

(P h 24

Below

2.8ac 3.4b

2.9ab

5.5ac 2.1ac

4.9ac at

square

:t 3; i

i i i

line.

(Chi

17.3

Day

23.2 34.8 27.0

23.0 25.8 soybean

above of

different

1.7b

2.3ac 2.3a 3.7ac 2.23c

4.0ac 0001* 0001*

i i :1;

3; :1; i

2 < 0.24

numbers 0004* < varieties P: P:

12.3

21.0 29.7 21.8 20.8 25.3

Day P: P:

right

signiﬁcantly 12.76, 12.68, different 85.36, 270.34,

the Offspring

not 12: six

to 76: 76: 12:

are

1.1b 1.1b

2.1b 3.1b 2.1a

2.7ab 1, 10, on 2, 5,

i 1; i i

;l-_ i

(8)

by df= df= df= df=

8.0

6.5

9.2

9.7

14.7

11.8

Day

letter produced

same

indicated

0.3 0.6 0.3 0.5 0.2 0.6

i i i i i

the offspring

3.5 2.8 3.7 3.8 3.5

Day

4.0

with and

Signiﬁcance

means

0.2 0.3 0.0 0.3 0.3 0.3

i i i

1; i 1.0 0.36 0.90 002* remaining

0.05.

where

5.0

Day

4.7 4.2 4.5 4.5 4.5

_<_ P: P: P: P:

P 1.0,

at 1.63, 8.19, 0.85, glycines Adults

test MEANS, )6: A. )8: )6: 76:

0.0

0.2 0.2 0.0 0.0 0.2

LS 1, 10, 5. 2,

i i i i

i -_1-_

three adult

the

5.0

5.0 5.0 df= df= df= df=

4.7 4.7

Day 4.7 of

trial

chi-squared

from

a SEM)

during

by (i

3) Nymphs

(day

resistant) and

comparisons number

(Jock)

determined

wise mean

and

was Adults

pair

Grown

(5152RR)

2),

(nematode The

(AG2602)

(93B82)

are 3.

(D261RR) Three

232

(day

Signiﬁcance

Mycogen 0.05).

Garst DSR Pioneer effects

Variety*Day Variety

Asgrow Michigan Block Variety Day Trial Table

153 Appendix 1

Record of Deposition of Voucher Specimens*

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

Voucher No.: 2002-11

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

BIOLOGICAL CONTROL OF THE SOYBEAN APHID, APHIS GLYCINES MATSUMURA (HOMOPTERA: APHIDIDAE)

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

Entomology Museum, Michigan State University (MSU)

Other Museums:

Investigator’s Name(s) (typed) Tyler 8. Fox

Date 12/9/02

*Reterence: 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.

154 Appendix 1.1

Voucher Specimen Data

Page 1 of 4 Pages

Museum where

deposited MSU

Other

of: Adults 6‘

v—iv—dv—i—ItMNNNv-(mv—Iv—INN Adults S?

for

Number Pupae

Nymphs University

Larvae specimens

Eggs State 1

listed

2001

2002

2001 2001

2001

Michi

2002

above

2002 2002

2002

2001

2002-1

2001

the

July

No.

Aug.

July

Aug.

July June

June July June

July

June

May

19 11

11 11

collected

Received

deposit Voucher

MSU

specimens

Farm,

Co.

for

" "

deposited

Fox

data

and

Lansing

Ingham

used Ml Entomology East

Label T.

G.)

C.)

(typed) 12/5/2002

Guerin

(De.

(111.)

Conte

(Le necessary)

(Say)

Stephens

Le if (Say)

(Say)

Say

Dej.

G.)

(Duft.) Name(s)

sodalis taxon

harisii sheets

(De

melanarus

convergens Fox

cognatus

munda

rugulosus

other B.

compar

tricolor

pensylvanicus

pusillus

lucublandus

placidum or

aenea

familiaris

additional

Date

Investigators Tyler

(Use

Poecilus Pterostichus

Philonthus Podabrus Chaenius Cycloneda Cyclotrachelus

Chaenius

Hippodamia

Harpalus Harpalus Species

Anisodactylus

Amara Amara Agonum

155

Number of: where deposnted

Adults Other Museum Adults Nymphs Pupae Larvae Eggs Label data for specimens collected or SpeCIes or other taxon used and deposited 6‘ 9

Anisodactylus santaecrusis (F.) MI Ingham Co. Mt. Hope, MSU 11 June 2001 3 MSU East Lansing T. B. Fox Bembidion quadrimaculatum Say 31 July 2001 NN'ﬁWVMV‘lmcr-‘N—‘OOOO Voucher Bembidion rapidum (Le C.) 16 Aug. 2002 Page Bradycellus repestris (Say) 11 June 2001 spp. 17 July 2002 Appendix

Chrysoperla 2 156

Clavina bipustalata (F.) 11 June 2002 Specimen Clavina impressefrons Le C. 11 June 2002 of (L.) 19 June 2002 Coccinella septempunctata 4 Elaphropus anceps (Le Conte) 11 June 2002 1.1 17 July 2002 Pages Harmonia axyridis (Pallas) Data Harpalus aﬂinis (Shrank) 22 June 2002 Hippodamia convergens Guerin 19 July 2002 Hippodamia parenthesis (Say) 11 June 2002 Nabis spp. 17 July 2002 Orius insidiosus Say 17 July 2002

(Use additional sheets if necessary) Investigators Name(s) (typed) Voucher No. 2002-11 Tyler B. Fox Received the above listed specimens tor deposit in the Michigan State University

Entomology Museum.

1 2/5/2002 Date Curator Date Appendix 1.1

Voucher Specimen Data

Page 3 of 4 Pages

Museum where

deposited MSU

Other

of: Adults 6‘

VF‘WMV} 3

Adults 9 4 4

for Number Pupae

Nymphs University NM? NWNF’

Larvae specimens

Date

Eggs State 1

listed

Michigan

above

2002 2002-1

2002 2002

2002

Museum.

2002 2002 2001

2002 2001 2002

2002 2002

2002

the

No.

June

July July

July

July July July

July July June

June

Aug.

July

18 19

11 11

collected

20 30

Entomology Received

Curator

deposit Voucher

specimens

Co.

MSU

for

" " "

deposited

Fox

data

and

Lansing

Hope,

Ingham

East

Mt.

MI Label used T.

12/5/2002

(L.)

Geer

(typed)

necessary)

(F.) Chd.

(Say)

(Pallas)

(Say)

Name(s)

Matsumura

taxon

sheets

maculata

Fox

comma spp.

septempunctata

axyridis spp.

chalcites

other

spp.

auricularia

spp.

subterraneus quadriceps

insidiosus

glycines additional

Date

Investigators Tyler

Hemerobius Coleomegilla

Orius

(Use

Spiders

Coccinella

Chrysopa

Harmonia Forﬁcula

Leucopis

Syrphid Stenolophus Scarites Scarites

Poeciulus

Species Aphis

157 Appendix 1.1

Voucher Specimen Data

Page 4 of 4 Pages

Museum where

deposited MSU

Other

of: Adults 6‘

Adults 9 for

Number Pupae

Nymphs University

Larvae specimens

Date Eggs State

listed

2002

Michigan

above

2002-11

Museum.

the

June

No.

June

collected

Entomology

Received Curator

deposit Voucher

MSU

specimens

Farm,

Co.

MSU

for

deposited

Fox

data

and

Lansing

Hope,

Ingham

East

MI East

Entomology Mt.

T. used Label MI

(typed)

2/5/2002

(Say)

necessary)

Name(s)

taxon

sheets

permundus

Fox

other

spp.

additional

Date

Investigators Tyler

(Use

Syrphid Pterostichus Species

158